Flac3d 600 Theory

FLAC3D 6.
0
Theory and Background
(an excerpt from FLAC3D Help)
© 2017 Itasca Consulting Group, Inc.

break
First Edition (FLAC3D Version 2.1) April 2002
Second Edition (FLAC3D Version 3.0) September 2005
Third Edition (FLAC3D Version 3.1) December 2006
Fourth Edition (FLAC3D Version 4.0) December 2009
Fifth Edition (FLAC3D Version 5.0) October 2012
Sixth Edition (FLAC3D Version 6.0) April 2017
break
FLAC3D 6.0
Precis
This document is a reproduction of the section “FLAC3D Theory and
Background” from the FLAC3D Help. It is provided as a convenience to FLAC3D
users who need access to the documentation in easily printable form.
There is no substantial difference between the content presented here and that
which appears in the FLAC3D Help. In cases of variance, the difference is either
a change made to accommodate format in going from a screen-based media
to one intended for paper, or, since the Help is updated concurrent to code
revisions, that the Help file content is more up-to-date than this document.
In any case of variance between the two, precedence should always be given to
the Help.
Printing Tip
This document uses a facing pages layout. Users who wish to print the
document are advised that printing double-sided, if possible, will produce the
best result. A double-sided copy can be bound or inserted into a ring-binder.
break
FLAC3D 6.0
This Document and Help Hyperlinks
The FLAC3D Help, from which this document is excerpted and produced, has
many cross-reference hyperlinks. These are not reproduced here. Elements in
the body-text of the Help that are hyperlinks are colored dark green in this
document. This convention has been adopted to allow for link identification
within the text of this PDF. To see the matter referenced, however, readers
will need to access the equivalent link in the Help file.
FLAC3D 6.0
FLAC3D Theory and Background: Table of Contents
FLAC3D Theory and Background: Table of

Contents
FLAC3D Theory and Background ……………………………………………………………………… 1
THEORETICAL BACKGROUND ………………………………………………………………………… 3

Formulation Of A 3d Explicit Finite Difference Model ………………………………… 5
Mathematical Model Description …………………………………………………………… 5
Conventions ……………………………………………………………………………………… 5
Stress ……………………………………………………………………………………………… 6
Rate of Strain and Rate of Rotation ………………………………………………… 6
Equations of Motion and Equilibrium ……………………………………………… 6
Boundary and Initial Conditions ……………………………………………………… 7
Constitutive Equations ……………………………………………………………………… 7
Numerical Formulation ………………………………………………………………………… 8
Finite Difference Approximation to Space Derivatives ……………………… 9
Nodal Formulation of the Equations of Motion ………………………………… 10
Explicit Finite Difference Approximation to Time Derivatives ………… 14
Constitutive Equations in Incremental Form …………………………………… 15
Large- and Small-Strain Modes ……………………………………………………… 16
Mechanical Timestep Determination for Numerical Stability …………… 16
Mechanical Damping ……………………………………………………………………… 20
Grid Discretization …………………………………………………………………………………… 27
Mixed Discretization for a Hexahedral Grid ………………………………………… 28
Nodal Mixed Discretization for a Tetrahedral Grid ………………………………… 31
Nodal mixed discretization on strain ……………………………………………… 31
Nodal mixed discretization on stress ……………………………………………… 32
Numerical Implementation ……………………………………………………………………… 35
Body Discretization ……………………………………………………………………………… 35
FLAC3D 6.0
Hexahedral Meshing ……………………………………………………………………… 35
Tetrahedral Meshing ……………………………………………………………………… 35
Initial and Boundary Conditions …………………………………………………………… 36
Main Calculation Steps ………………………………………………………………………… 36
Strain-Rate Calculation ………………………………………………………………………… 37
Stress Calculation ………………………………………………………………………………… 37
Nodal Mass Calculation ………………………………………………………………………… 38
Out-of-Balance Force and Ratio Calculation ………………………………………… 39
Maximum Out-of-Balance Force ……………………………………………………… 39
Local Maximum Force Ratio …………………………………………………………… 40
Average Force Ratio ………………………………………………………………………… 41
Maximum Force Ratio ……………………………………………………………………… 41
Velocity and Displacement Calculations ……………………………………………… 41
Geometry Update Calculation ……………………………………………………………… 42
Energy Calculation in FLAC3D ……………………………………………………………… 42
Energy Dissipation in Zones through Plastic Work …………………………… 42
References ……………………………………………………………………………………………… 45
INTERFACES ………………………………………………………………………………………………… 47
General Comments …………………………………………………………………………………… 49
Formulation ……………………………………………………………………………………………… 51
Creation of Interface Geometry ………………………………………………………………… 59
Choice of Material Properties …………………………………………………………………… 63
Interface Used to Join Two Sub-grids …………………………………………………… 63
Real Interface — Slip and Separation Only …………………………………………… 67
All Properties Have Physical Significance ……………………………………………… 69
Modeling Guidelines ………………………………………………………………………………… 79
Troubleshooting ………………………………………………………………………………… 79
Initial Stresses …………………………………………………………………………………… 79
Interface Corners ………………………………………………………………………………… 79
Overlapping Interfaces ………………………………………………………………………… 81
Interfaces and Fluid Flow …………………………………………………………………… 82
FLAC3D 6.0
Interfaces and Changing Interacting Objects in Small Strain ………………… 83

References ……………………………………………………………………………………………… 85
FACTOR OF SAFETY ……………………………………………………………………………………… 87

Introduction …………………………………………………………………………………………… 89
Factor of Safety ………………………………………………………………………………………… 91
Computational Methods for Factor of Safety Calculation of Slopes …………… 93
Strength Reduction Technique ……………………………………………………………… 93
Limit Analysis ……………………………………………………………………………………… 94
Limit Equilibrium ………………………………………………………………………………… 95
Relation of Strength Reduction Method to Limit Equilibrium
and Limit Analysis …………………………………………………………………………… 96
Strength Reduction Procedure in FLAC3D ………………………………………………… 99
Strength Reduction Properties …………………………………………………………… 102
Mohr-Coulomb Material ………………………………………………………………… 102
Ubiquitous-Joint Material ……………………………………………………………… 102
Hoek-Brown Material …………………………………………………………………… 103
Interfaces ……………………………………………………………………………………… 104
Example Factor of Safety Calculations using
the Strength Reduction Method ……………………………………………………………… 107
Failure of a Slope with a Complex Surface Profile in a
Mohr-Coulomb Material …………………………………………………………………… 109
Influence of Slope Curvature on Stability …………………………………………… 115
Simple Slope in Hoek-Brown Material ………………………………………………… 127
Automatic Calculation of a Stable Pit Slope Angle ………………………………… 133
Factor of Safety Contours …………………………………………………………………… 143
References ……………………………………………………………………………………………… 151
CONSTITUTIVE MODELS ……………………………………………………………………………… 155

Constitutive Models in FLAC3D ……………………………………………………………… 159
Incremental Formulation ………………………………………………………………………… 165
Null Model Group …………………………………………………………………………………… 169
FLAC3D 6.0
Null Model ………………………………………………………………………………………… 171
Elastic Model Group ………………………………………………………………………………… 173
Elastic (Isotropic) Model ……………………………………………………………………… 175
Anisotropic (Transversely-Elastic) Model …………………………………………… 177
Orthotropic Elastic Model …………………………………………………………………… 181
Plastic Model Group ………………………………………………………………………………… 185
Drucker-Prager Model ………………………………………………………………………… 189
Mohr-Coulomb Model ………………………………………………………………………… 199
Ubiquitous-Joint Model ……………………………………………………………………… 211
Anisotropic-Elasticity Ubiquitous-Joint Model …………………………………… 223
Strain-Softening/Hardening Mohr-Coulomb Model …………………………… 229
Bilinear Strain-Softening/Hardening Ubiquitous-Joint Model ……………… 235
Double-Yield Model …………………………………………………………………………… 251
Modified Cam-Clay Model ………………………………………………………………… 267
Hoek-Brown Model …………………………………………………………………………… 285
Hoek-Brown-PAC Model …………………………………………………………………… 293
Cap-Yield (CYSoil) Model ………………………………………………………………… 303
Simplified Cap-Yield (CHSoil) Model …………………………………………………… 327
Plastic-Hardening Model …………………………………………………………………… 341
Swell Model ……………………………………………………………………………………… 367
Mohr-Coulomb Tension Crack (MohrT) Model …………………………………… 373
Model Tests and Examples ……………………………………………………………………… 377
Oedometer Test with Mohr-Coulomb Model ……………………………………… 379
Uniaxial Compressive Strength of a Jointed Material Sample ……………… 387
Isotropic Consolidation Test with Double-Yield Model ……………………… 393
Isotropic Consolidation Test with Modified Cam-Clay Model ……………… 397
Triaxial Compression Test with Hoek-Brown Model …………………………… 401
Triaxial Compression Test with Hoek-Brown-PAC Model …………………… 407
Isotropic Compression Test with CYSoil Model …………………………………… 415
Oedometer Test with CYSoil Model ……………………………………………………… 419
Drained Triaxial Test with CYSoil Model — Constant Dilation …………… 423
Drained Triaxial Test with CYSoil Model — Dilation Hardening ………… 429
FLAC3D 6.0
Undrained Triaxial Test with CYSoil Model ………………………………………… 433

Drained Triaxial Compression Test with
Simplified Cap-Yield (CHSoil) Model ……………………………………………… 437
Comparison between Mohr-Coulomb Model and
Plastic-Hardening model ………………………………………………………………… 441
Isotropic Compression Test with Plastic-Hardening Model ………………… 445
Drained Triaxial Compression Test with Plastic-Hardening Model ……… 449
Undrained Triaxial Compression Test with Plastic-Hardening Model … 455
Oedometer Test with Plastic-Hardening Model ………………………………… 459
Single Zone Swell Test ……………………………………………………………………… 465
Single Zone Loading-Unloading Test with MohrT Model …………………… 469
References ……………………………………………………………………………………………… 479
FLUID-MECHANCIAL INTERACTION …………………………………………………………… 481

FLAC3D Fluid-Thermal-Mechanical-Formulation —
Mathematical Description …………………………………………………………………… 487
Conventions and Definitions ……………………………………………………………… 488
Governing Differential Equations ……………………………………………………… 489
Fluid Flow Boundary and Initial Conditions in FLAC3D ……………………… 492
Numerical Formulation ………………………………………………………………………… 493
Saturated Fluid Flow ………………………………………………………………………… 494
Finite-Difference Approximation to Space Derivatives ………………… 494
Nodal Formulation of the Mass Balance Equation ………………………… 495
Explicit Finite-Difference Formulation ………………………………………… 498
Stability Criterion ………………………………………………………………………… 498
Implicit Finite-Difference Formulation ………………………………………… 500
Convergence Criterion …………………………………………………………………… 502
Saturated/Unsaturated Flow ……………………………………………………………… 503
Mechanical Timestep and Numerical Stability …………………………………… 504
Total Stress Correction ……………………………………………………………………… 505
Fully Saturated Fast Flow …………………………………………………………………… 506
An Alternative Fast-Flow Algorithm ……………………………………………… 507
FLAC3D 6.0
Calculation Modes for Fluid-Mechanical Interaction ………………………………… 511
Grid Not Configured for Fluid Flow …………………………………………………… 511
Grid Configured for Fluid Flow …………………………………………………………… 512
Properties and Units for Fluid-Flow Analysis ………………………………………… 519
Permeability Coefficient ……………………………………………………………………… 519
Mass Density ……………………………………………………………………………………… 520
Fluid Moduli ……………………………………………………………………………………… 521
Biot Coefficient and Biot Modulus ………………………………………………… 521
Fluid Bulk Modulus ……………………………………………………………………… 522
Fluid Moduli and Convergence ……………………………………………………… 522
Fluid Moduli for Drained and Undrained Analyses ………………………… 523
Porosity ……………………………………………………………………………………………… 524
Saturation ………………………………………………………………………………………… 525
Undrained Thermal Coefficient …………………………………………………………… 525
Fluid Tension Limit …………………………………………………………………………… 526
Fluid-Flow Boundary Conditions, Initial Conditions, Sources and Sinks … 527
Solving Flow-Only and Coupled-Flow Problems ……………………………………… 531
Time Scales ………………………………………………………………………………………… 531
Selection of a Modeling Approach for a Fully Coupled Analysis …………… 535
Time Scale …………………………………………………………………………………… 536
Nature of Imposed Perturbation to the Coupled Process ………………… 537
Stiffness Ratio ……………………………………………………………………………… 538
Recommended Procedure to Select a Modeling Approach ……………… 539
Fixed Pore Pressure (Used in Effective Stress Calculation) ………………… 544
Flow-Only Calculation to Establish a Pore-Pressure Distribution ………… 544
No Flow — Mechanical Generation of Pore Pressure ………………………… 546
Coupled Flow and Mechanical Calculations ………………………………………… 550
Verification Examples ……………………………………………………………………………… 561
Unsteady Groundwater Flow in a Confined Layer ……………………………… 563
One-Dimensional Filling of a Porous Region ……………………………………… 569
Steady-State Fluid Flow with a Free Surface ……………………………………… 575
Spreading of a Groundwater Mound …………………………………………………… 587
FLAC3D 6.0
One-Dimensional Consolidation ………………………………………………………… 593

Consolidation Settlement at the Center of a Strip Load ………………………… 611
Transient Fluid Flow to a Well in a Shallow Confined Aquifer ……………… 621
Pressuremeter Test …………………………………………………………………………… 631
Semi-confined Aquifer ……………………………………………………………………… 645
Verification of Concepts, and Modeling Techniques
for Specific Applications ………………………………………………………………………… 649
Solid Weight, Buoyancy and Seepage Forces ……………………………………… 649
A Simple Example Illustrating Solid Weight, Buoyancy and
Seepage Forces ……………………………………………………………………………… 650
Pore Pressure Initialization and Deformation …………………………………… 661
Heave of a Soil Layer …………………………………………………………………… 662
Effect of the Biot Coefficient ……………………………………………………………… 667
Undrained Oedometer Test …………………………………………………………… 668
Pore Pressure Generation in a Confined Sample …………………………… 669
Pore Pressure Generation in an Infinite Layer ………………………………… 671
Input Instructions for Fluid-Flow Analysis ……………………………………………… 673
Fluid Commands ……………………………………………………………………………… 673
Fluid FISH Variables ………………………………………………………………………… 674
Zone-Based Pore Pressure ………………………………………………………………… 675
References ……………………………………………………………………………………………… 679
FLAC3D 6.0
FLAC3D 6.0
FLAC3D Theory and Background 1
FLAC3D Theory and Background

This section contains background information and descriptions of the theoretical
basis for the basic components of FLAC3D. It is composed of five major sections,
described as follows.
The theoretical formulation for FLAC3D is described in the first section,

Theoretical Background. This includes an overview of the explicit finite
difference method and the numerical implementation in FLAC3D.
The next section, Interfaces, describes the FLAC3D interface logic. It includes
several example applications. Interfaces represent planes on which sliding and
separation can occur.
A Factor of Safety calculation, based upon the “strength reduction method,” is

described in the third section. Several example applications of this calculation for
stability analyses are also presented in this section.
The fourth section provides a description of Constitutive Models in FLAC3D. The

basic models in the program range from linearly elastic models to highly
nonlinear plastic models. All are introduced here.
The last section describes the implementation of Fluid-Mechanical Interaction in

FLAC3D models, with illustrative examples.
FLAC3D 6.0
2 FLAC3D Theory and Background • Theoretical Background
FLAC3D 6.0
Theoretical Background 3
Theoretical Background
The theoretical background for FLAC3D is presented in this section. Much of
FLAC3D is a direct extension from the two-dimensional version of FLAC. Refer to
the FLAC User’s Manual for further background on the explicit finite difference
method, which is the basis for both FLAC and FLAC3D.
There are, however, distinct differences between the two- and three-
dimensional formulations, specifically in the 2D-to-3D extension of the
mathematical model that describes the mechanics of the system. This section
describes the three-dimensional formulation and its implementation in FLAC3D.
Section Outline
• Formulation of a 3D Explicit Finite Difference Model
◦ Mathematical Model Description
▪ Conventions
▪ Stress
▪ Rate of Strain and Rate of Rotation
▪ Equations of Motion and Equilibrium
▪ Boundary and Initial Conditions
▪ Constitutive Equations
◦ Numerical Formulation
▪ Finite Difference Approximation to Space Derivatives
▪ Nodal Formulation of the Equations of Motion
▪ Explicit Finite Difference Approximation to Time
Derivatives
▪ Constitutive Equations in Incremental Form
▪ Large- and Small-Strain Modes
▪ Mechanical Timestep Determination for Numerical Stability
▪ Mechanical Damping
▪ Local Nonviscous Damping
▪ Combined Damping
• Grid Discretization
◦ Mixed Discretization for a Hexahedral Grid
◦ Nodal Mixed Discretization for a Tetrahedral Grid
FLAC3D 6.0
▪ Nodal mixed discretization on strain

▪ Nodal mixed discretization on stress
• Numerical Implementation
◦ Body Discretization
▪ Hexahedral Meshing
▪ Tetrahedral Meshing
◦ Initial and Boundary Conditions
◦ Main Calculation Steps
◦ Strain-Rate Calculation
◦ Stress Calculation
◦ Nodal Mass Calculation
◦ Out-of-Balance Force and Ratio Calculation
▪ Maximum Out-of-Balance Force
▪ Local Maximum Force Ratio
▪ Average Force Ratio
▪ Maximum Force Ratio
◦ Velocity and Displacement Calculations
◦ Geometry Update Calculation
◦ Energy Calculation in FLAC3D
▪ Energy Dissipation in Zones through Plastic Work
• References
FLAC3D 6.0
Formulation of a 3D Explicit Finite Difference Model 5
Formulation of a 3D Explicit Finite Difference

Model
FLAC3D is an explicit finite difference program to study, numerically, the
mechanical behavior of a continuous three-dimensional medium as it reaches
equilibrium or steady plastic flow. The response observed derives from a
particular mathematical model on one hand, and from a specific numerical
implementation on the other. These two topics are addressed below.
Mathematical Model Description

The mechanics of the medium are derived from general principles (definition of
strain, laws of motion), and the use of constitutive equations defining the
idealized material. The resulting mathematical expression is a set of partial
differential equations, relating mechanical (stress) and kinematic (strain rate,
velocity) variables, which are to be solved for particular geometries and
properties, given specific boundary and initial conditions.
An important aspect of the model is the inclusion of the equations of motion,

although FLAC3D is primarily concerned with the state of stress and deformation
of the medium near the state of equilibrium. The numerical implementation
section shows that the inertial terms are used as a means by which to reach the
equilibrium state in a numerically stable manner.
Conventions
In the Lagrangian formulation adopted in FLAC3D, a point in the medium is
characterized by the vector components xi, ui, and dvi /dt, i = 1,3 of position,
displacement, velocity and acceleration, respectively.
As a notation convention, a bold letter designates a vector or tensor, depending

on the context. The symbol ai denotes component i of the vector [a] in a
Cartesian system of reference axes; Aij is component (i,j ) of tensor [A]. Also, αi is
the partial derivative of α with respect to xi. (α can be a scalar variable, a vector
or tensor component.)
By definition, tension and extension are positive.
FLAC3D 6.0
The Einstein summation convention applies, but only on indices i, j and k, which
take the values 1, 2, 3.
Stress
The state of stress at a given point of the medium is characterized by the
symmetric stress tensor σij. The traction vector [t] on a face with unit normal [n]
is given by Cauchy’s formulae (tension positive):
(1)
Rate of Strain and Rate of Rotation

Let the particles of the medium move with velocity [ v]. In an infinitesimal time,
dt, the medium experiences an infinitesimal strain determined by the
translations vidt, and the corresponding components of the strain-rate tensor
may be written as
(2)
where partial derivatives are taken with respect to components of the current
position vector [x].
For later reference, the first invariant of the strain-rate tensor gives a measure
of the rate of dilation of an elementary volume. Aside from the rate of
deformation characterized by the tensor ξij, a volume element experiences an
instantaneous rigid-body displacement, determined by the translation velocity
[v], and a rotation with angular velocity.
(3)
where eijk is the permutation symbol, and [ω] is the rate of rotation tensor whose
components are defined as
(4)
Equations of Motion and Equilibrium

Application of the continuum form of the momentum principle yields Cauchy’s
equations of motion:
FLAC3D 6.0
(5)
where ρ is the mass-per-unit volume of the medium, [b] is the body force per
unit mass, and d[v]/dt is the material derivative of the velocity. These laws
govern, in the mathematical model, the motion of an elementary volume of the
medium from the forces applied to it. Note that, in the case of static equilibrium
of the medium, the acceleration d[v]/dt is zero, and Equation (5) reduces to the
partial differential equations of equilibrium:
(6)
Boundary and Initial Conditions

The boundary conditions consist of imposed boundary tractions (see Equation
(1)) and/or velocities (to induce given displacements). In addition, body forces
may be present. Also, the initial stress state of the body needs to be specified.
Constitutive Equations
The equations of motion Equation (5), together with the definitions (Equation
(2)) of the rates of strain, constitute nine equations for fifteen unknowns (the 6
+ 6 components of the stress- and strain-rate tensors and the three components
of the velocity vector). Six additional relations are provided by the constitutive
equations that define the nature of the particular material under consideration.
They are usually given in the form
(7)
in which is the corotational stress-rate tensor, [H] is a given function, and κ

is a parameter that takes into account the history of loading. The corotational
stress rate is equal to the material derivative of the stress as it would appear
to an observer in a frame of reference attached to the material point and rotating
with it at an angular velocity equal to the instantaneous value of the angular
velocity [Ω] of the material. Its components are defined as
(8)
in which dσij/dt is the material time derivative of [σ], and [ω] is the rate of
rotation tensor.
FLAC3D 6.0
Numerical Formulation
The method of solution in FLAC3D is characterized by the following three
approaches:
1. finite difference approach (First-order space and time derivatives

of a variable are approximated by finite differences, assuming
linear variations of the variable over finite space and time
intervals, respectively.);
2. discrete-model approach (The continuous medium is replaced by

a discrete equivalent — one in which all forces involved (applied
and interactive) are concentrated at the nodes of a three-
dimensional mesh used in the medium representation.); and
3. dynamic-solution approach (The inertial terms in the equations

of motion are used as numerical means to reach the equilibrium
state of the system under consideration.)
The laws of motion for the continuum are, by means of these approaches,
transformed into discrete forms of Newton’s law at the nodes. The resulting
system of ordinary differential equations is then solved numerically using an
explicit finite difference approach in time.
The spatial derivatives involved in the derivation of the equivalent medium are
those appearing in the definition of strain rates in term of velocities. For the
purpose of defining velocity variations and corresponding space intervals, the
medium is discretized into constant strain-rate elements of tetrahedral shape
whose vertices are the nodes of the mesh mentioned above. A tetrahedron is
represented in Figure 1, as an illustration:
FLAC3D 6.0
Figure 1: Tetrahedron.
Finite Difference Approximation to Space Derivatives

The finite difference formulation of the strain-rate tensor components for the
tetrahedron are derived below as a preliminary to the nodal formulation of the
equations of motion. The tetrahedron nodes are referred to locally by a number
from 1 to 4 and, by convention, face n is opposite node n. (See Figure 1.)
By application of the Gauss divergence theorem to the tetrahedron, we can write
(9)
where the integrals are taken over the volume and the surface of the tetrahedron,
respectively, and [n] is the exterior unit vector normal to the surface.
For a constant strain-rate tetrahedron, the velocity field is linear, and [ n] is

constant over the surface of each face. Hence, after integration, Equation (9)
yields
(10)
where the superscript ( f ) relates to the value of the associated variable on face f,
and is the average value of velocity component i. For a linear velocity variation,
we have
(11)
FLAC3D 6.0
where the superscript l relates to the value at node l.
Substitution of Equation (11) in (10) yields, reorganizing terms by node

contribution,
(12)
If we replace vi with 1 in Equation (9) we obtain, by application of the divergence

theorem,
(13)
Using this last relation, and dividing Equation (12) by V, we get
(14)
and the components of the strain-rate tensor may be expressed as
(15)
Nodal Formulation of the Equations of Motion

The nodal formulation of the equations of motion will be derived below by
application of the theorem of virtual work, at any instant in time, to an
equivalent static problem. Approximations on the form of the nodal inertial
terms will be made by using those terms as means to reach the solution
corresponding to the equilibrium equations, Equation (6).
Fixing time, t, we consider an equivalent static problem governed at any instant

in time by the equilibrium equations
(16)
with body forces defined as (Equation (5))
(17)
FLAC3D 6.0
In the framework of the finite difference approximation adopted here, the

medium is represented by a continuous assembly of constant-strain tetrahedra
subjected to body forces [B]. The nodal forces [f] n , with n = (1,4), acting on a
single tetrahedron in “static” equilibrium with the tetrahedron stresses and
equivalent body forces, are derived by application of the theorem of virtual work.
After application of a virtual nodal velocity δ[v] n (it will generate a linear velocity
field δ[v] and a constant strain-rate δ[ξ] inside the tetrahedron), we equate the
external rate of work done by the nodal forces [f] n and body forces [B] with the
internal work rate done by the stresses σij under that velocity.
Following the notation conventions (a superscript refers to the nodal value of a

variable) and Einstein summation convention on indices i and j, the external
work rate may be expressed as
(18)
while the internal work rate is given by
(19)
Using Equation (15), we can write, for a constant strain-rate tetrahedron,
(20)
The stress tensor is symmetric and, defining a vector T l with components
(21)
we obtain
(22)
After substitution of Equation (17) in (18), the external work rate may be
expressed as
(23)
FLAC3D 6.0
where E b and E I are the external work-rate contributions of the body forces ρbi
and inertial forces, respectively. For a constant-body force ρbi inside the
tetrahedron, we can write
(24)
while E I may be expressed as
(25)
According to the finite difference approximation done earlier, the velocity field
varies linearly inside the tetrahedron. To describe it, we adopt a local system of
reference axes xʹ1, xʹ2, xʹ3, with origin at the tetrahedron centroid, and write
(26)
where N n (with n = 1,4) are linear functions of the form
(27)
and c0n , c1n , c2n , c3n (with n = 1,4) are constants determined by solving the
systems of equations
(28)
where δnj is the Kronecker delta. By definition of the centroid, all integrals of the
form vanish, and substitution of Equation (26) for δvi in Equation (24)
yields, using Equation (27),
(29)
Using Cramer’s rule to solve Equation (28) for c0n , we obtain, taking advantage
of the properties of the centroid,
(30)
From Equation (29) and (30), we may write
FLAC3D 6.0
(31)
Also, substitution of Equation (26) for δvi in Equation (25) gives
(32)
Finally, with expressions Equation (31) for E b and Equation (32) for E I, Equation
(23) becomes
(33)
For static equilibrium of the tetrahedron in the framework of the equivalent

problem, the internal work rate (see Equation (22)) equals the external work rate
expressed by Equation (33) for any virtual velocity. Hence, we must have,
regrouping terms,
(34)
For small spatial variations of the acceleration field around an average value
inside the tetrahedron, the last term in Equation (34) may be expressed as
(35)
Also, for constant values of ρ inside the tetrahedron, and using the properties of
the centroid mentioned above (see Equation (27) and (30)) we may write
(36)
In the context of this analysis, the mass ρV/4 involved in the inertial term above
is replaced by a fictitious nodal mass m n, whose value will be determined below,
in order to ensure numerical stability of the system on its route to equilibrium.
Accordingly, Equation (36) becomes
(37)
and Equation (34) may be written as
FLAC3D 6.0
(38)
The equilibrium conditions for the equivalent system may now be established by
requiring that, at each node, the sum of the statically equivalent forces, -[ f], of
all contributing tetrahedra and nodal contributions [P] of applied loads and
concentrated forces be zero. To express those conditions, we adopt a notation
where a variable with superscript <l> refers to the value of the variable at a node
with label l in the global node numbering. The symbol [[.]] <l> is used to represent
the sum of the contributions at global node l of all tetrahedra meeting at that
node. With those conventions, we may write the following expressions of
Newton’s law at the nodes:
(39)
where nn is the total number of nodes involved in the medium representation, the
nodal mass M <l> is defined as
(40)
and the out-of-balance force [F] <l> is given by
(41)
This force is equal to zero when the medium has reached equilibrium.
Explicit Finite Difference Approximation to Time Derivatives

Taking into consideration the constitutive equations, Equation (7), and the
relation Equation (15) between deformation rate and nodal velocities, Equation
(39) may be expressed formally as a system of ordinary differential equations of
the form
(42)
where the notation {} <l> refers to the subset of nodal velocity values involved in
the calculation at global node (see Equation (39)) In FLAC3D, this system is
solved numerically using an explicit finite difference formulation in time. In this
approach, the velocity of a material node is assumed to vary linearly over a time
interval Δt, and the derivative on the left-hand side of Equation (42) is evaluated
FLAC3D 6.0
using central finite differences, whereby velocities are stored for times that are
displaced by half-timesteps with respect to displacements and forces. Nodal
velocities are computed using the recurrence relation
(43)
In turn, the node location is similarly updated using the central finite difference
approximation,
(44)
First-order error terms vanish when the finite difference scheme embodied in
Equation (43) and (44) is used (i.e., the scheme is second-order accurate).
Nodal displacements are calculated in the code from the relation
(45)
with .
Constitutive Equations in Incremental Form

In FLAC3D, it is assumed that velocities remain constant in the time interval, Δt.
An incremental expression of the constitutive Equation (7) with the following
form is used:
(46)
where is the corotational stress increment, and Hij* is a given function.
For small displacements and displacement gradients during Δt, we may write
(47)
where Δij is the change of strain related to the configuration at time t.
The stress increment Δσij is computed from , using
(48)
where Δσ C
ij is a stress correction defined as (see Equation (8))
(49)
FLAC3D 6.0
The components of the rate-of-rotation tensor are calculated using Equation (4)
and the finite difference formulation Equation (14):
(50)
Specific forms of the constitutive function H * are described in Constitutive

Models, where the subject of their numerical implementation in FLAC3D is also
discussed.
Large- and Small-Strain Modes

The numerical formulation described above provides a description of large-strain
deformation, involving large displacements, displacement gradients and
rotations. This is termed the large-strain mode in FLAC3D.
In applications in which the rotations are sufficiently small that the components
ωij - δij are small compared to unity, [ω] may be replaced by [I], and the stress
correction in Equation (48) may be omitted. Also, for small displacements and
displacement gradients, the spatial derivatives involved in the definition
Equation (2) of the strain-rate tensor may be evaluated with respect to the initial
configuration, and the node coordinates need not be updated (see Equation (44)).
In FLAC3D, the small-strain mode assumes small displacements, displacement

gradients and rotations. In that mode, node coordinates are not updated, and
stress rotation corrections are not taken into consideration.
Mechanical Timestep Determination for Numerical Stability

The difference equations (Equation (43)) will not provide valid answers unless
the numerical scheme is stable. Some physical insight on this topic may be
gained by viewing the idealized medium as an assembly of point masses (located
at the nodes) connected by linear springs, a conceptualization which may be
made on the following grounds. The equations of motion for a mass-spring
system may be expressed, in matrix notation, as
(51)
FLAC3D 6.0
where the braces denote vectors of nodal values, {P*} are exterior forces, [K] is
the matrix of rigidity of the springs, and [M ] is a diagonal nodal mass matrix.
The analogy with the idealized medium is immediate if we interpret the out-of-
balance forces in Equation (39) as the applied and spring reaction forces in
Equation (51).
In studying the oscillating mass-spring system with a finite difference scheme, a

timestep that does not exceed a critical timestep related to the minimum
eigenperiod of the total system must be used. Hence, the stability criterion for
the numerical scheme must provide an upper bound for the values of the
timesteps used in the finite difference scheme.
Derivation of a relation providing a measure of critical timesteps for the system

requires knowledge of the eigenperiods of the system. However, in practice,
global eigenvalue analyses are impractical and not commonly used for this
purpose (see Press et al. 1986). In FLAC3D, as will be shown below, a local
variation of this stability analysis is performed. An important aspect of the
numerical method is that a uniform unit timestep Δt is adopted for the whole
system:
(52)
And the nodal masses on the right-hand side of Equation (39) are taken as
variables, and adjusted to fulfill the local stability conditions.
Figure 2: Mass-spring system 1.
First, consider the one-dimensional mass-spring system represented in Figure 2.

The motion of the point mass, with a given initial displacement, is governed by
the differential equation
(53)
FLAC3D 6.0
where k is the stiffness of the spring, and m is the point mass. The critical
timestep corresponding to a second-order finite difference scheme for this
equation is given by (Bathe and Wilson 1976):
(54)
where T is the period of the system — i.e.,
(55)
Figure 3: Mass-spring system 2.
Now consider the infinite series of masses and springs in Figure 2-a. By
symmetry, the behavior of this assembly may be analyzed by studying the
system sketched in Figure 2-b which, in turn, is equivalent to a single mass-
spring system with stiffness 4k (Figure 2-c) The limit-stability criterion, derived
from Equation (54) and eqtheory-springt, has the form
(56)
By selecting Δt = 1, the system will be stable if the magnitude of the point mass is
greater than or equal to the spring stiffness. In the local analysis, the validity of
(56) is extended to one tetrahedron by interpreting m as the nodal mass
contribution, m l, at local node, l, and k as the corresponding nodal stiffness
FLAC3D 6.0
contribution, k l. The nodal mass contribution as derived from the infinite series
criterion provides an upper-bound value for the system under consideration. The
nodal stiffness contribution is derived from a simple diagonalization technique of
the local stiffness matrix, as follows.
The tetrahedron internal-force contribution at local node l is equal to Ti l /3 (see

Equation (41)) This force is interpreted as a spring reaction force of the form
-kijl uj l (see Equation (51)) Taking variations over a time interval, dt, we write
(57)
Using (21), this relation becomes
(58)
With a unit-velocity component in direction q at node l, and all other nodal

velocity components set to zero, we obtain from Equation (58) the diagonal
terms of the local stiffness matrix
(59)
where, by convention, no summation is implied on repeated index q, which runs

from 1 to 3.
Adopting the small-strain incremental form of Hooke’s law to describe the

stress-strain constitutive relation in the small time interval, we write
(60)
in which α1 = K + 4/3G, K is the bulk modulus and G is the shear modulus.
Using the selected nodal-velocity values in the finite difference formulation

Equation (15) for ξ, we have
(61)
Substitution of this expression in Equation (60) gives
(62)
We now define an upper-bound value for the nodal stiffness contribution as
FLAC3D 6.0
(63)
which yields, from Equation (56) with Equation (52), the expression for the
tetrahedron mass contribution at node l:
(64)
to provide a numerically stable solution.
Mechanical Damping
The equations of motion must be damped to provide static or quasi-static (non-
inertial) solutions. Local nonviscous damping is the default damping algorithm for
static analysis in FLAC3D, and is described in this section. Local damping can be
controlled with the zone mechanical damping local <f> command, where f is the
local damping (default is 0.8). Local damping can also be implemented in a
dynamic analysis (see the section Dynamic Analysis).
An alternative damping algorithm is provided in FLAC3D for situations in which

the steady-state solution includes a significant uniform motion. This can occur,
for example, in a creep simulation, or in the calculation of the ultimate capacity
of an axially loaded pile. This damping is called combined damping, and is also
described in this section. Combined damping is more efficient than local
damping at removing kinetic energy for this special case. Combined damping is
invoked with the zone mechanical damping combined <f> command.
For dynamic analyses, Rayleigh damping and artificial viscosity damping are also
available to control damping. These algorithms are described in the section on
Mechanical Damping in the Theory and Background section.
Local Nonviscous Damping
The local nonviscous damping used in FLAC3D is similar to that described in

Cundall (1987). A damping-force term is added to the equations of motion, given
by Equation (39), such that the damped equations of motion can be written as
(65)
is the damping force and is given by
FLAC3D 6.0
(66)
where
(67)
expressed in terms of the generalized out-of-balance force, F <l>i

, and the
<l> . The damping force is controlled by the damping
generalized velocity, v (i)
constant, α, whose default value is 0.8.
This form of damping has three advantages:
1. Only accelerating motion is damped; therefore, no erroneous

damping forces arise from steady-state motion.
2. The damping constant, α, is nondimensional.
3. Since damping is frequency-independent, regions of the

assembly with different natural periods are damped equally,
using the same damping constant.
The local nonviscous damping formulation in FLAC3D is similar to hysteretic

damping, in which the energy loss per cycle is independent of the rate at which
the cycle is executed. This similarity is demonstrated in the following analysis.
Upon examination of Equation (66), it is evident that the damping force always
opposes motion, but it is scaled to the resultant generalized force, unlike viscous
damping, which is scaled to the magnitude of velocity. Two forms of Equation
(66) may be written, depending on the relative signs of and . When they
are of the same sign, the equation of motion is
(68)
and when they are of opposite signs, the equation of motion is
(69)
These equations may be written in terms of apparent mass as
FLAC3D 6.0
(70)
where and .
Figure 4: Motion of single mass-spring system.
The motion of the single degree-of-freedom mass-spring system from Figure 2,

when given an initial displacement such that x = a at t = 0, is shown in Figure 4.
For such motion, the apparent mass is increased at the two times in each
cycle when the velocity is zero, and reduced at the two times when the
velocity is maximum. The damping operates by removing kinetic energy
twice per cycle: at the two points of peak velocity, part of the moving
mass is removed and discarded. Note that there is no discontinuity in
acceleration, since the acceleration is zero at the instant when mass is
removed.
FLAC3D 6.0
The amount of energy removed per cycle is given by twice the drop in kinetic
energy at point B, or
(71)
and the mean kinetic energy at the instant of removal is
(72)
where is the generalized velocity given by Equation (66).
The ratio of energy lost per cycle to maximum energy stored is called the
“specific loss” (Kolsky 1963). The specific loss can be written in terms of the
damping constant in FLAC3D by noting that, for a single degree-of-freedom
system, or for oscillation in a single mode, the peak kinetic energy is the same as
the peak stored energy. Thus, the specific loss can be written as
(73)
The fraction of critical damping, D, can be written in terms of the specific loss
for small values of damping as
(74)
For the default value of α = 0.8, the approximate fraction of critical damping is
therefore 0.25.
Equation (73) can be verified by loading one zone and a column of three zones,
respectively, with a suddenly applied gravity load and allowing each system to
oscillate to equilibrium. Both tests give approximately the same value for Δ W/W,
confirming that the damping is independent of frequency. However, for a system
where multiple modes are active simultaneously, the response computed by
FLAC3D may not show the same damping for each mode.
Combined Damping
The damping formulation described by Equation (66) is only activated when the
velocity component changes sign. In situations where there is significant
uniform motion (in comparison to the magnitude of oscillations that are to be
damped), there may be no “zero crossings,” and hence no energy dissipation. An
FLAC3D 6.0
alternative (but less efficient) formulation is derived by noting that, for a single
degree-of-freedom system executing harmonic motion, the derivative of the
unbalanced force is proportional to minus the velocity. If
(75)
then
(76)
Hence is proportional to . However, the mean value of is zero even when

the mean value of is nonzero, for an oscillating system. An alternative form of
damping called “combined damping,” is provided in FLAC3D. The scheme
consists of a damping force composed of equal contributions from expressions
based on velocity and :
(77)
This form of damping should be used if there is significant rigid-body motion of

a system in addition to oscillatory motion to be dissipated. However, combined
damping is found to dissipate energy at a slower rate than local damping based
on velocity. Therefore, local damping is preferred in most cases.
The effect of rigid-body motion is demonstrated in ElasticBlock.f3dat, which

models a block under gravity. The corresponding project file is
“ElasticBlock.f3prj,” located in “datafiles\theory\elastic-block.” However, the
lower boundary is constrained to move upwards at constant velocity, and all
gridpoints are fixed in the horizontal direction. The final “equilibrium” state
should be one in which gravity-induced stresses act in the body but, in addition,
all gridpoints should move upwards at the same velocity. As seen from the
velocity history of one top-surface gridpoint (shown in Figure 5), the body
continues to oscillate indefinitely and does not reach the predicted steady state.
ElasticBlock.f3dat
model new
fish automatic-create off
zone create brick size 5 5 5
zone cmodel elastic
zone property shear 1 bulk 2
zone initialize density 1
model gravity 0 0 -10
FLAC3D 6.0
zone gridpoint fix velocity-x

zone gridpoint fix velocity-y
zone gridpoint fix velocity range position-z 0
zone gridpoint initialize velocity-z 1.0
history interval 10
zone history velocity-z position 3 3 5
model save 'ini'
;
model step 1000
model save 'undamped'
;
model restore 'ini'
zone mechanical damping combined
model step 1000
model save 'damped'
;
return
Figure 5: z-velocity history at top of block — local damping.
FLAC3D 6.0
When the zone mechanical damping combined command is specified, the system
then converges to the steady state in which all velocities are equal to the imposed
velocity (see Figure 6), and the internal stresses exhibit the same gravitational
gradient as the static case.
Figure 6: z-velocity history at top of block — combined damping.
FLAC3D 6.0
Grid Discretization 27
Grid Discretization
Among three-dimensional constant strain-rate elements, tetrahedra have the
advantage of not generating hourglass deformations (i.e., deformation patterns
created by combinations of nodal velocities producing no strain rate, and thus no
nodal force increments). However, when used in the framework of plasticity,
these elements do not provide for enough modes of deformation (see Nagtegaal
et al. 1974). In particular situations, for example, they cannot deform
individually without change of volume as required by certain important
constitutive laws. In those cases, the elements are known to exhibit an overly
stiff response compared to what is expected from theory. To overcome this
problem, a process of mixed discretization is applied in FLAC3D.
The principle of the mixed discretization technique is to give the element more
volumetric flexibility by proper adjustment of the first invariant of the tetrahedra
strain-rate tensor. [*] Application of the technique differs depending on whether
primary discretization of the body is done into hexahedral zones or into
tetrahedra. In the approach pioneered by Marti and Cundall (1982), a coarser
discretization in zones is superposed to the tetrahedral discretization, and the
first strain-rate invariant of a particular tetrahedron in a zone is evaluated as the
volumetric-average value over all tetrahedra in the zone. The method is
illustrated in Figure 1. In the particular mode of deformation sketched there,
individual constant strain-rate elements will experience a volume change
incompatible with a theory of incompressible plastic flow. In this example,
however, the volume of the assembly of tetrahedra (i.e., the zone) remains
constant, and application of the mixed discretization process allows each
individual tetrahedron to reflect this property of the zone, hence reconciling its
behavior with that predicted by the theory.
FLAC3D 6.0
Mixed Discretization for a Hexahedral Grid
Figure 1: Mode of deformation for which mixed discretization would be most

efficient.
In FLAC3D, a zone corresponds to an assembly of nt tetrahedra, as illustrated in

Figure 2 for the case nt = 5. Consider a particular zone where the strain-rate
tensor of a tetrahedron locally labeled l in that zone is first estimated, and then
decomposed into deviatoric and volumetric parts:
(1)
FLAC3D 6.0
where η [l] is the deviatoric strain-rate tensor, and ξ [l] is the strain-rate first
invariant,
(2)
Figure 2: An 8-node zone with 2 overlays of 5 tetrahedra in each overlay.
The first invariant for the zone is then calculated as the volumetric average value
of the first invariant over all tetrahedra in the zone:
(3)
FLAC3D 6.0
where V [k] is the volume of tetrahedron k. F inally, the tetrahedron strain-rate

tensor components are calculated from
(4)
Dilatant constitutive laws will produce changes in mean normal stress when
yielding occurs. For a consistent technique, the first invariant of the stress
tensor, derived after application of the strain-rate increment, must also be
evaluated as a volumetric average for the zone. In this process, the stress tensor
of a particular tetrahedron l in a zone is first estimated and decomposed into
deviatoric and volumetric parts:
(5)
where [s] [l] is the deviatoric strain-rate tensor, and σ [l] is the mean normal
stress.
(6)
The first invariant for the zone is calculated as the volumetric average value over
all tetrahedra in the zone:
(7)
Finally, the tetrahedron stress-rate tensor components are calculated using
(8)
In FLAC3D, the discretization process starts with the coarser grid: zones are
defined and then discretized (internally) into tetrahedra. An eight-noded zone,
for instance, can be discretized into two (and only two) different configurations
of five tetrahedra (corresponding to overlay 1 and 2 in Figure 2. The calculation
of nodal forces (based on evaluation of strain rates and stresses) can be carried
out using one overlay or a combination of two overlays. The advantage of the
two-overlay approach is to ensure symmetric zone response for symmetric
loading. In this case, mixed discretization is carried out over the combination of
two overlays, and nodal forces computations are evaluated by averaging over the
two overlays.
FLAC3D 6.0
Nodal Mixed Discretization for a Tetrahedral Grid

Nodal mixed discretization (NMD) is a variation on the mixed discretization
scheme, in which averaging of the volumetric behavior is carried out on a node
basis instead of a zone basis. The procedure is applied on the triangle or
tetrahedral-based mesh; it does not require the assembly of elements into zones.
Also, the constitutive model is called on an element basis, as usual (no call is
made on a node basis for the volumetric behavior, as in the average nodal
pressure (ANP) formulation described by Bonet and Burton 1998). The procedure
involves a nodal mixed discretization on strain, and one on stress. Both steps are
considered below.
First we recall the general calculation sequence embodied in FLAC and FLAC3D:
1. Nodal forces are calculated from stresses, applied loads and body
forces (velocity and displacement vary linearly; stress and strain
are constant within an element).
2. The equations of motion are invoked to derive new nodal

velocities and displacements.
3. Element strain rates are derived from nodal velocities.
4. New stresses are derived from strain rates, using the material
constitutive law.
In the NMD technique, the calculation sequence is respected. However, an

averaging procedure is carried out on strain rates (end of step 3) and on stress
increments (end of step 4), as described below.
Nodal mixed discretization on strain

The strain rate is derived from nodal velocities, as usual. The strain rate is
then partitioned into deviatoric, , and volumetric, , components:
(9)
where is the Kroenecker delta.
A nodal volumetric strain rate (defined as the weighted average of the

surrounding element values) is calculated using the formula
FLAC3D 6.0
(10)
where mn are the elements surrounding node n, and Ve is the volume of element
e.
After nodal volumetric strain rate values are obtained, a mean value for the
element is calculated by taking the average of nodal values:
(11)
where d = 3 for a triangle and 4 for a tetrahedral.
Finally, the element strain rate is redefined by superposition of the deviatoric

part and volumetric average:
(12)
The constitutive model is called to derive new stresses (from strain rates) and
previous stresses.
Nodal mixed discretization on stress

Consider an incremental volumetric constitutive stress-strain law which, for
small increments, can be linearized in the form
(13)
where stands for plastic volumetric strain increment, and the value is nonzero
for dilatant/hfilbreak contractant material. The associated nodal forces must be
consistent with the assumptions made to define the element kinematics. To
enforce this, a nodal mixed discretization procedure is applied on the term ,
as described below. For convenience, we refer to the term as ; With this
convention, Equation (13) may be expressed as
(14)
The value is a standard quantity evaluated in the constitutive model

procedure.
The technique for nodal mixed discretization on stress is similar to the one
applied for strain. First, nodal values for are calculated as the weighted average
of the surrounding element values:
FLAC3D 6.0
(15)
After the nodal values are obtained, a mean value for the element is
calculated by taking the average of nodal values:
(16)
where, again, d = 3 for a triangle and 4 for a tetrahedra.
Finally, the stresses calculated by the constitutive model are corrected by

substituting for in the returned value:
(17)
Clearly, the nodal mixed discretization on stress will only be relevant for
dilatant/contractant materials.
It is important to note that the averaging manipulations involved in the NMD

technique are done independent of the constitutive model formulation, which
remains unaffected. (The plastic volumetric “stress correction” calculated by
the constitutive model is simply passed back from the constitutive model as one
of the state variables.)
Also, because of the way the averaging procedure is carried out, the resulting
tetrahedral formulation gives the correct behavior in patch test situations where
is uniform, and thus equal, for all elements.
The nodal averaging procedure involved in the NMD technique removes the
excessively constrained kinematics of the linear velocity element.
Endnote
[*] This invariant gives a measure of the rate of dilation of the constant strain-
rate tetrahedron.
FLAC3D 6.0
FLAC3D 6.0
Numerical Implementation 35
Numerical Implementation
Body Discretization
By default, discretization of the volume under study is done into hexahedral
zones. The user can also import a tetrahedral mesh (generated by an external
mesher) into FLAC3D. The two types of body discretization are addressed:
Hexahedral Meshing
In FLAC3D, the general discretization of the body into hexahedral zones is
performed by the user. Several available procedures to help in this task are
described in the section Grid Generation. Each zone is discretized automatically
by the code into sets of tetrahedra. The user can decide to carry out the
calculation using one overlay or a combination of two overlays. The use of two
overlays is recommended in regions of the medium where high gradients of
stresses and deformations are expected. By default, two overlays are prescribed
for all zones in a model. Also, the technique of mixed discretization is applied to
overcome the overly stiff behavior of uniform strain tetrahedra during plastic
flow.
Tetrahedral Meshing
The NMD calculation carried out on a tetrahedral grid imported into FLAC3D uses
only one overlay. Implementation of the nodal mixed discretization technique on
strain is done using three calculation loops, performed just before calling the
constitutive model:
1. One loop on zones to distribute volumetric strain rate (as

calculated from gridpoint velocities) and area/volume to
gridpoints.
2. A loop on gridpoints to evaluate the volume-weighted average of

volumetric strain rate at the gridpoints.
FLAC3D 6.0
3. A loop on zones to calculate an average of volumetric strain rate

using the gridpoint values, and assign the value to the zone.
Note that advantage can be taken of existing calculation loops to perform the
task described above.
In case constitutive models which rely on an elastoplastic volumetric behavior

are used, similar loops (based on stress) are used, after calling the constitutive
model, to implement the nodal mixed discretization on stress.
Initial and Boundary Conditions

The boundary conditions of the problem consist of surface tractions,
concentrated loads and displacements. In addition, body forces may be given and
initial stress conditions imposed. For implementation in the code, all stresses
and nodal velocities are initially set to zero; then, initial stresses are specified.
Concentrated loads are specified at given surface nodes, and imposed boundary
displacements are prescribed in terms of nodal velocities. Body forces and
surface tractions are transformed internally into a set of statically equivalent
nodal forces. This constitutes the initial state of the numerical scheme.
Main Calculation Steps

FLAC3D uses an explicit “time-marching” finite difference solution scheme; for
every timestep, the calculation sequence can be summarized:
1. New strain rates are derived from nodal velocities.
2. Constitutive equations are used to calculate new stresses from the

strain rates and stresses at the previous time.
3. The equations of motion are invoked to derive new nodal

velocities and displacements from stresses and forces.
The sequence is repeated at every timestep, and the maximum out-of-balance

force in the model is monitored. This force will either approach zero, indicating
that the system is reaching an equilibrium state, or it will approach a constant,
nonzero value, indicating that a portion (or all) of the system is at steady-state
(plastic) flow of material. The calculation may be interrupted at any point, in
order to analyze the solution.
FLAC3D 6.0
A summary of equations (derived in Numerical Formulation) that are used in the

calculation sequence is presented below, for clarity. In the notation, superscripts
are used to particularize a node or tetrahedron only when needed in the local
context. Also, as before, the notation [[f]] <l> represents the sum of contributions
of type f at global node l of all tetrahedra having that node in common.
Strain-Rate Calculation
Starting from a known velocity field, the components of the strain-rate tensor
are computed, for each tetrahedron in a zone, using the finite difference
formulation (15):
(1)
For the default hexahedral grid, a procedure of mixed discretization is then

applied, and new diagonal strain-rate tensor components are calculated from
Equations (3) and (4). For a particular tetrahedron, now locally labeled l in the
zone, these equations yield
(2)
where ξ z, the average value for the zone of the first strain-rate invariant, is
calculated from
(3)
where nt is the total number of tetrahedra involved in the zone computation, and
V [k] is the volume of a tetrahedron labeled k.
For a tetrahedral grid, a technique of nodal mixed discretization on strain is
applied.
Stress Calculation
The constitutive equations are used in their incremental form Hij* to calculate
stress increments for each tetrahedron in a zone (see Equations (46), (47), and
(48)):
(4)
FLAC3D 6.0
where
(5)
and is computed (using Equation (15)) from
(6)
The stress correction Δσ C

ij
is not taken into consideration in small-strain mode.
In large-strain mode, its value is given by Equations (49) and (50):
(7)
where
(8)
New stress values are then derived by addition of the stress increments.
For the default hexahedral grid, a technique of mixed discretization is applied,

and the diagonal components of the stress-tensor components are adjusted
using Equations (7) and (8). For a particular tetrahedron, locally labeled l in the
zone, these equations are
(9)
and
(10)
For a tetrahedral grid, a procedure of nodal mixed discretization on stress is

used.
Nodal Mass Calculation

The tetrahedron contribution m l to the nodal mass at local node l is evaluated
using Equation (64):
(11)
noting that the magnitude of the timestep is set equal to unity.
FLAC3D 6.0
The nodal mass M <l> at global node l is computed as the sum of contributions
Equation (11) at the node of all tetrahedra having that node in common:
(12)
In the small-strain mode, the nodal mass calculation is only carried out once
before cycling. In large-strain mode, values are updated every 10 steps.
Out-of-Balance Force and Ratio Calculation

The tetrahedron contribution [p] l to the out-of-balance force at local node l is
computed from the stresses and body forces using
(13)
(see Equations (21) and (41)).
The sum of contributions Equation (13) at global node l of all tetrahedra having
that node in common is evaluated. Averaging over two overlays is carried out
when appropriate, and the contributions [P] <l> of applied loads and concentrated
forces are added, thus providing the expression for the out-of-balance force
[F] <l> at global node l:
(14)
The out-of-balance force is monitored to detect whether the system has reached
a state of equilibrium or steady flow.
Any applied damping terms will be added to the out-of-balance force, e.g., for
local damping, the term evaluated from Equation (66) as
(15)
is added to the out-of-balance force.
Maximum Out-of-Balance Force

The maximum out-of-balance force is the maximum absolute value of any one
component of the remaining force of any global node l after all forces are
summed together:
FLAC3D 6.0
(16)
where . This is the force that drives the gridpoint in the

direction of equilibrium. If the force is zero, the gridpoint can be thought of as
being converged, or at equilibrium. As a measure of overall convergence, the
major drawback of maximum out-of-balance force is that it is not a unitless
measure. The appropriate target value is dependent on the model size, zone size,
material properties, and stress magnitude.
Local Maximum Force Ratio

The local force ratio is a unitless measure of local convergence at a gridpoint. The
idea is simple, divide the out-of-balance component remaining after summing
of forces by a measure of the total forces being applied to the gridpoint. For
performance purposes, we use the “Manhattan norm” or “Taxicab norm” of a
vector, defined as , or the sum of the absolute values of the
components of the vector, as the measure. The local force ratio can then be
defined as:
(17)
where fi are all the force vectors being added to a gridpoint and , and
the summation loops over all zones with the common gridpoint.
The local maximum forced ratio is the maximum local ratio of any gridpoint in
the model.
(18)
This is a very conservative measure of the convergence of a model. However it

has a weakness when there is less than two forces being applied to a gridpoint,
or when all forces being applied to a gridpoint act in the same direction. In this
circumstance, the only “equilibrium” value possible is for both numerator and
denominator to converge towards zero. However the ratio between them will
tend towards 1.
FLAC3D 6.0
Average Force Ratio

The average force ratio is defined as the sum of all out-of-balance force
components at every gridpoint divided by the sum of all total forces applied at a
gridpoint. Or:
(19)
This measure serves as a reliable measure of the overall convergence of the

system. For models that are relatively uniform it performs very well. However,
for very non-uniform models the effects of localized convergence problems can
be lost in the average, making the user unaware that an area that might be of
most interest still needs computation. This can occur either because of
disparaties in zone size, or because of large differences in stiffness. For large
complex models, this measure should be checked carefully before being trusted.
See the discussion following.
Maximum Force Ratio

The maximum force ratio is defined as the maximum out-of-balance force
divided by the average total force acting on all gridpoints:
(20)
The average total force is defined as:
(21)
where N is the total number of gridpoints in the model.
Velocity and Displacement Calculations

Equation (43) is invoked, taking the damping into account, to derive new nodal
velocities from known out-of-balance forces:
(22)
FLAC3D 6.0
In turn, nodal displacements are evaluated using Equation (45):
(23)
Geometry Update Calculation

In large-strain mode, node locations are updated using Equation (44), and the
grid then deforms accordingly. For global node l, we have
(24)
In small-strain mode, the geometry adjustment is neglected.
Energy Calculation in FLAC3D

Elastic strain energy and dissipated plastic energy can be tracked for zones
containing a mechanical model. FLAC3D uses an incremental solution procedure:
the equations of motion (at the gridpoints) and the stress-strain calculations (at
the zones) are solved every timestep. In the stress-strain equations, the
incremental change in energy components is determined and accumulated as the
system attempts to reach equilibrium.
Energy Dissipation in Zones through Plastic Work

Several plasticity models are available in FLAC3D which can describe the
deformability of the zones. Energy is dissipated through plastic work as the
zones undergo irreversible deformation. The strain in any zone can be divided
into elastic and plastic parts.
The elastic strain energy (decomposed into deviatoric and volumetric

components) is given by
(25)
where V is sub-tet volume, σijd is deviatoric stress and is mean stress.
The change in elastic strain energy before and after the stress-strain equations
are solved in a constitutive model is given by
(26)
FLAC3D 6.0
Where We is the current elastic strain energy and Wʹe is the previous elastic strain
energy. Equation (26) can be decomposed into shear and volumetric components:
ΔWes and ΔWev.
The total shear energy change can be found from the total deviatoric strain and
deviatoric stress:
(27)
The total volumetric change can be found from the total mean strain and mean
stress:
(28)
The total shear plastic work dissipated during a timestep is the difference
between the total shear energy and the elastic shear energy change at any
timestep:
(29)
Similarly, the total volumetric plastic work dissipated during a timestep is
(30)
Each zone contains overlays of constant stress, strain tetrahedra. A volumetric

average of the above energy components is calculated for the zone for an average
energy dissipation increment for that zone. Each zone has four additional
variables to accumulate dissipated plastic energy: Wpv, Wps, Wev, Wes.
Energy calculations in FLAC3D can be accessed through the zone mechanical

energy active on command and through zone-based FISH functions
zone.work.plastic.vol, zone.work.plastic.shear, zone.work.plastic.total,
zone.work.elastic.vol, zone.work.elastic.shear, and zone.work.elastic.total.
FLAC3D 6.0
FLAC3D 6.0
References 45
References
Bathe, K. J., and E. L. Wilson. Numerical Methods in Finite Element Analysis.
Englewood Cliffs: Prentice-Hall Inc. (1976).
Bonet, J., and A. J. Burton. “A Simple Averaged Nodal Pressure Tetrahedral

Element for Nearly Incompressible Dynamic Explicit Applications,” Commun.
Numer. Meth. Engng., 14, 437-449 (1998).
Cundall, P. A. “Distinct Element Models of Rock and Soil Structure,” in Analytical

and Computational Methods in Engineering Rock Mechanics, Ch. 4, pp. 129-163. E. T.
Brown, ed. London: Allen & Unwin. (1987).
Kolsky, H. Stress Waves in Solids. New York: Dover Publications (1963).
Marti, J., and P. Cundall. “Mixed Discretization Procedure for Accurate Modelling
of Plastic Collapse,” Int. J. Num. & Analy. Methods in Geomech., 6, 129-139 (1982).
Nagtegaal, J. C., D. M. Parks and J. R. Rice. “On Numerically Accurate Finite

Element Solutions in the Fully Plastic Range,” Comp. Meth. Appl. Mech. & Eng., 4,
153-177 (1974).
Press, W. H., et al. Numerical Recipes: The Art of Scientific Computing. Cambridge:
Cambridge University Press (1986).
FLAC3D 6.0
FLAC3D 6.0
Interfaces 47
Interfaces
The theoretical background for FLAC3D interfaces is presented in this section.
Section Outline
• General Comments
• Formulation
• Creation of Interface Geometry
• Choice of Material Properties
◦ Interface Used to Join Two Sub-grids
◦ Real Interface — Slip and Separation Only
◦ All Properties Have Physical Significance
• Modeling Guidelines
◦ Troubleshooting
◦ Initial Stresses
◦ Interface Corners
◦ Overlapping Interfaces
◦ Interfaces and Fluid Flow
◦ Interfaces and Changing Interacting Objects in Small Strain
• References
FLAC3D 6.0
48 FLAC3D Theory and Background • Interfaces
FLAC3D 6.0
General Comments 49
General Comments
There are several instances in geomechanics in which it is desirable to represent
planes on which sliding or separation can occur. For example:
1. joint, fault, or bedding planes in a geologic medium;
2. an interface between a foundation and the soil;
3. a contact plane between a bin or chute and the material that it

contains;
4. a contact between two colliding objects; and
5. a planar “barrier” in space, which represents a fixed, non-

deformable boundary at an arbitrary position and orientation.
FLAC3D provides interfaces that are characterized by Coulomb sliding and/or

tensile and shear bonding. Interfaces have the properties of friction, cohesion,
dilation, normal and shear stiffnesses, and tensile and shear bond strength.
Although there is no restriction on the number of interfaces or the complexity of
their intersections, it is generally not reasonable to model more than a few
simple interfaces with FLAC3D because it is awkward to specify complicated
interface geometry. The program 3DEC (Itasca 2007) is specifically designed to
model many interacting bodies in three dimensions; it should be used instead of
FLAC3D for the more complicated interface problems.
Interfaces may also be used to join regions that have different zone sizes. In
general, the zone attach command should be used to join grids together.
However, in some circumstances, it may be more convenient to use an interface
for this purpose. In this case, the interface is prevented from sliding or opening
because it does not correspond to any physical entity.
FLAC3D 6.0
FLAC3D 6.0
Formulation 51
Formulation
FLAC3D represents interfaces as collections of triangular elements (interface
elements), each of which is defined by three nodes (interface nodes). Interface
elements can be created at any location in space. Generally, interface elements
are attached to a zone surface face; two triangular interface elements are defined
for every quadrilateral zone face. Interface nodes are then created automatically
at every interface element vertex. When another grid surface comes into contact
with an interface element, the contact is detected at the interface node and is
characterized by normal and shear stiffnesses, and sliding properties.
Each interface element distributes its area to its nodes in a weighted fashion.
Each interface node has an associated representative area. The entire interface is
thus divided into active interface nodes representing the total area of the
interface. Figure 1 illustrates the relation between interface elements and
interface nodes, and the representative area associated with an individual node.
FLAC3D 6.0
Figure 1: Distribution of representative areas to interface nodes.
It is important to note that interfaces are one-sided in FLAC3D. (This differs from
the formulation of two-sided interfaces in two-dimensional FLAC (Itasca 2011).)
It may be helpful to think of FLAC3D interfaces as “shrink-wrap” that is
stretched over the desired surface, causing the surface to become sensitive to
interpenetration with any other face with which it may come into contact.
The fundamental contact relation is defined between the interface node and a
zone surface face, also known as the target face. The normal direction of the
interface force is determined by the orientation of the target face.
During each timestep, the absolute normal penetration and the relative shear
velocity are calculated for each interface node and its contacting target face. Both
of these values are then used by the interface constitutive model to calculate a
normal force and a shear-force vector. The constitutive model is defined by a
linear Coulomb shear-strength criterion that limits the shear force acting at an
interface node, normal and shear stiffnesses, tensile and shear bond strengths,
and a dilation angle that causes an increase in effective normal force on the
FLAC3D 6.0
Formulation 53
target face after the shear-strength limit is reached. By default, pore pressure is
used in the interface effective stress calculation. This option can be activated/
deactivated using the command zone interface effective command by setting
effective = on/off. Figure 2 illustrates the components of the constitutive model
acting at interface node (P):
Figure 2: Components of the bonded interface constitutive model.
The normal and shear forces that describe the elastic interface response are
determined at calculation time using the relations
(1)
(2)
where: is the normal force at time [force];
FLAC3D 6.0
is the shear force vector at time [force];
is the absolute normal penetration of the interface node into

the target face [displacement];
is the incremental relative shear displacement vector

[displacement];
is the additional normal stress added due to interface stress

initialization [force/displacement];
is the normal stiffness [stress/displacement];
is the shear stiffness [stress/displacement];
is the additional shear stress vector due to interface stress

initialization; and
is the representative area associated with the interface node

[length2].
The inelastic interface logic works in the following way:
1. Bonded interface — The interface remains elastic if stresses

remain below the bond strengths; there is a shear bond strength,
as well as a tensile bond strength. The normal bond strength is
set using the tension interface property keyword. The command
zone interface node property shear-bond-ratio = sbr sets the
shear bond strength to sbr times the normal bond strength. The
default value of property shear-bond-ratio (if not given) is 100.0.
The bond breaks if either the shear stress exceeds the shear
strength, or the tensile effective normal stress exceeds the
normal strength. Note that giving property shear-bond-ratio
alone does not cause a bond to be established—the tensile bond
strength must also be set.
FLAC3D 6.0
Formulation 55
2. Slip while bonded — An intact bond, by default, prevents all yield

behavior (slip and separation). There is an optional property
switch (bonded-slip) that causes only separation to be prevented
if the bond is intact (but allows shear yield, under the control of
the friction and cohesion parameters, using as the normal
force). The command to allow/disallow slip for a bonded interface
segment is zone interface node and by setting bonded-slip on or
off.
The default state of bonded-slip (if not given) is off.
3. Coulomb sliding — A bond is either intact or broken. If it is

broken, then the behavior of the interface segment is determined
by the friction and cohesion (and of course the stiffnesses). This
is the default behavior, if bond strengths are not set (zero). A
broken bond segment cannot take effective tension (which may
occur under compressive normal force, if the pore pressure is
greater). The shear force is zero (for a nonbonded segment) if the
effective normal force is tensile or zero.
The Coulomb shear-strength criterion limits the shear force by the relation
(3)
where:
is the cohesion [stress] along the interface;
is the friction angle [degrees] of the interface surface; and
is pore pressure (interpolated from the target face) provided the

keyword effective = off has not been issued for the interface.
If the criterion is satisfied (i.e., if ), then sliding is assumed to occur,

and , with the direction of shear force preserved.
During sliding, shear displacement may cause an increase in the effective normal
stress on the joint, according to the relation
(4)
where:
FLAC3D 6.0
is the dilation angle [degrees] of the interface surface; and
is the magnitude of shear force before the preceding correction is

made.
On printout (see the zone interface node list command ) the value of tension
denotes whether a bond is intact or broken (or not set) — nonzero or zero,
respectively.
The normal and shear forces calculated at the interface nodes are distributed in
equal and opposite directions to both the target face and the face to which the
interface node is connected (the host face). Weighting functions are used to
distribute the forces to the gridpoints on each face. The interface stiffnesses are
added to the accumulated stiffnesses at gridpoints on both sides of the interface
in order to maintain numerical stability.
Interface contacts are detected only at interface nodes, and contact forces are
transferred only at interface nodes. The stress state associated with a node is
assumed to be uniformly distributed over the entire representative area of the
node. Interface properties are associated with each node; properties may vary
from node to node.
By default, the effect of pore pressure is included in the interface calculation by

using effective stress as the basis for the slip condition. (The interface pore
pressure is interpolated from the target face.) This applies in model configure
fluid mode, or if pore pressures are assigned with the zone water or zone
gridpoint initialize pore-pressure command without specifying model
configure fluid. The user can switch options for interface s by using the zone
interface effective command and by setting effective on or off. By default in
the FLAC3D logic, fluid flow (saturated or unsaturated) is carried across an
interface, provided the interface keyword maximum-edge is not used for that
particular interface. The permeable interface option can be deactivated/
reactivated for interface s by using the zone interface permeability command
and by setting effective on or off. Note that if the keyword maximum-edge is used
after the zone interface element command, and permeability is on for a
particular interface, a warning is issued to inform the user that this interface will
be considered as impermeable to fluid flow. (Note that for fluid flow calculation
only, a mechanical model must be present. Also, the model cycle 0 command
with model mechanical active on should be used to initialize the weighting factors
FLAC3D 6.0
Formulation 57
used to transfer fluid flow information across the interface.) No pressure drop
normal to the joint and no influence of normal displacement on pore pressure is
calculated.
Also, flow of fluid along the interface is not modeled.
FLAC3D 6.0
FLAC3D 6.0
Creation of Interface Geometry 59
Creation of Interface Geometry

Interfaces are created with the zone interface create command. For cases in
which an interface between two separate grids in the model is required, the zone
interface create by-face command should be used to attach an interface to one
of the grid surfaces. This command generates interface elements for interface s
along all surface zone faces with a center point that fall within a specified range.
Any surfaces on which an interface is to be created must be generated initially; it
must be possible to specify an existing surface in order to create the interface
elements. A gap must be specified between two adjacent surfaces, unless the zone
interface create by-face command and the separate keyword are given. In this
case, the separate sub-grids may have surface gridpoints at the same location in
space.
By default, two interface elements are created for each zone face. The number of
interface elements can be increased by using the zone interface s element
maximum-edge v command. [1] This causes all interface elements with edge lengths
larger than v to subdivide into smaller elements until their lengths are smaller
than v. This command can be used to increase the resolution and decrease
arching of forces in portions of a model that have large contrasts in zone size
across an interface.
Several rules should be followed when using interface elements in FLAC3D:
1. If a smaller surface area contacts a larger surface area (e.g., a

small block resting on a large block), the interface should be
attached to the smaller region.
2. If there is a difference in zone density between two adjacent

grids, the interface should be attached to the grid with the greater
zone density (i.e., the greater number of zones within the same
area).
3. The size of interface elements should always be equal to or

smaller than the target faces with which they will come into
contact. If this is not the case, the interface elements should be
subdivided into smaller elements.
FLAC3D 6.0
4. Interface elements should be limited to grid surfaces that will

actually come into contact with another grid.
A simple example illustrating the procedure for interface creation is provided in

“DippingJoint.f3dat”. The corresponding project file, “DippingJoint.f3prj”, is
located in the folder “datafiles\Interface\DippingJoint.” The example is a block
specimen containing a single joint dipping at an angle of 45°.
DippingJoint.f3dat
model new
model title "Model with a dipping joint"
zone create brick size 3 3 3 ...
point 0 (0,0,0) point 1 (3,0,0) point 2 (0,3,0) ...
point 3 (0,0,1.5) point 4 (3,3,0) point 5 (0,3,1.5) ...
point 6 (3,0,4.5) point 7 (3,3,4.5) ...
group 'Base'
point 0 (0,0,1.5) point 1 (3,0,4.5) point 2 (0,3,1.5) ...
point 3 (0,0,6) point 4 (3,3,4.5) point 5 (0,3,6) ...
point 6 (3,0,6) point 7 (3,3,6) ...
group 'Top'
model save 'int0'
; Create interface elements on the top surface of the base
zone interface 'joint' create by-face separate range group 'Top' group 'Base'
model save 'int1'
return
Figure 1 shows the grid before the interface is created. Two sub-grid groups are
defined: a Base grid and a Top grid. Figure 2 shows the model with the interface
elements attached to the Base grid.
A uniaxial compression test with this model is described later in All Properties
Have Physical Significance.
FLAC3D 6.0
Creation of Interface Geometry 61
Figure 1: Initial geometry before creation of the interface.
Figure 2: Final geometry.
FLAC3D 6.0
[1] Note that if model configure fluid is invoked, and permeability is on for a
particular interface, specifying maximum-edge for that interface will
automatically make it impermeable. Do not specify maximum-edge if flow
across the interface is desired.
FLAC3D 6.0
Choice of Material Properties 63
Choice of Material Properties

Assignment of material properties (particularly stiffnesses) to an interface
depends on the way the interface is used. Three possibilities are common. The
interface may be:
1. an artificial device to connect two sub-grids together;
2. a real interface that is stiff compared to the surrounding material,

but which can slip and perhaps open in response to the
anticipated loading. (This case also encompasses the situation in
which stiffnesses are unknown or unimportant, but where slip
and/or separation will occur, e.g., flow of frictional material in a
bin.); or
3. a real interface that is soft enough to influence the behavior of

the system (e.g., a joint with soft clay filling or a dike containing
heavily fractured material).
These cases are examined in detail.
Interface Used to Join Two Sub-grids

If possible, sub-grids should be joined with the zone attach command. It is more
computationally efficient to use the zone attach command than the zone
interface create command to join sub-grids. See Connecting Adjoining Primitive
Shapes for a description of, and restrictions on, the zone attach command.
Under some circumstances, it may be necessary to use an interface to join two

sub-grids. This type of interface is assigned high strength properties with the
zone interface node property command, thus preventing any slip or separation.
(This is the equivalent of a “glued” interface in FLAC.) Shear and normal
stiffnesses must also be provided; values of friction and cohesion are not needed.
It is tempting (particularly for people familiar with finite element methods) to
give a very high value for these stiffnesses to prevent movement on the interface.
However, FLAC3D does “mass scaling” (see Mechanical Timestep Determination
for Numerical Stability) based on stiffnesses, and the response (and solution
FLAC3D 6.0
convergence) will be very slow if very high stiffnesses are specified. It is

recommended that the lowest stiffness consistent with small interface
deformation be used. A good rule-of-thumb is that and be set to ten times
the equivalent stiffness of the stiffest neighboring zone. The apparent stiffness
(expressed in stress-per-distance units) of a zone in the normal direction is
(1)
where:
& are the bulk and shear moduli, respectively; and
is the smallest width of an adjoining zone in the normal direction

(see Figure 1).
The [ ] notation indicates that the maximum value over all zones adjacent to
the interface is to be used (e.g., there may be several materials adjoining the
interface).
Figure 1: Zone dimension used in stiffness calculation.
FLAC3D 6.0
To illustrate the approach, consider Figure 2, in which two sub-grids of unequal

zoning are joined by the commands in “Subgrids.f3dat”, and are loaded by a
pressure on the left-hand part of the upper surface. The corresponding project
file, “Subgrids.f3prj”, is located in the folder “datafiles\Interface\Subgrids”.
Subgrids.f3dat
model new
model title "Joining two sub-grids"
zone create brick size 4 4 4 point 0 (0,0,0) point 1 (4,0,0) ...
point 2 (0,4,0) point 3 (0,0,2)
zone create brick size 8 8 4 point 0 (0,0,3) point 1 (4,0,3) ...
point 2 (0,4,3) point 3 (0,0,5)
zone interface 'joint' create by-face range position-z 3
zone interface 'joint' node property stiffness-normal 300e9 ...
stiffness-shear 300e9 ...
tension 1e10 ...
shear-bond-ratio=1
zone gridpoint initialize position-z -1.0 add range position-z 3 5
zone cmodel assign elastic
zone property bulk 8e9 shear 5e9
zone gridpoint fix velocity-z range position-z 0
zone gridpoint fix velocity-x range union position-x 0 position-x 4
zone gridpoint fix velocity-y range union position-y 0 position-y 4
zone face apply stress-zz -1e6 ...
range position-z 4 position-x 0,2 position-y 0,2
model solve
model save 'join.f3sav'
return
The value of is 15 GPa, and the minimum zone size adjacent to the
interface is 0.5 m. Hence, we choose both shear stiffness and normal stiffness to
be 150 × 109/0.5 (i.e., Pa/m). The resulting contours of
-displacement are shown in Figure 3.
Compare this result to that for a single grid shown in this figure. This plot is at
the same scale and contour intervals as Figure 3. The two plots are almost
identical, which indicates that the interface does not affect the behavior to any
great extent.
The prescription given in Equation (1) is reasonable if the materials on the two
sides of the interface are similar and variations of stiffness occur only in the
lateral directions. However, if the material on one side of the interface is much
stiffer than that on the other, then Equation (1) should be applied to the softer
FLAC3D 6.0
side. In this case, the deformability of the whole system is dominated by the soft
side; making the interface stiffness ten times the soft-side stiffness will ensure
that the interface has minimal influence on system compliance.
Figure 2: Two unequal sub-grids joined by an interface.
FLAC3D 6.0
Figure 3: Vertical displacement contours—two joined grids.
Real Interface — Slip and Separation Only

In this case, we simply need to provide a means for one sub-grid to slide and/or
open relative to another sub-grid. The friction (and perhaps cohesion, dilation,
and tensile strength) is important, but the elastic stiffness is not. The approach
in Interface Used to Join Two Sub-grids is used here to determine and .
However, the other material properties are given real values (for advice see All
Properties Have Physical Significance).
As an example, we can allow slip in a bin-flow problem, as shown in Figure 4,

corresponding to the data file in “BinFlowSlip.f3dat”. The corresponding project
file, “BinFlowSlip.f3prj,” is located in folder “datafiles\Interface\BinFlowSlip.”
The bond strengths are not set (i.e., they default to zero); the interface
stiffnesses are set to approximately ten times the equivalent stiffness of the
neighboring zones.
FLAC3D 6.0
Figure 4: Flow of frictional material in a “bin”.
BinFlowSlip.f3dat
model new
model title "Real interface - slip and separation only"
; Create Material Zones
point 0 (0,0,0) point 1 (3,0,0) point 2 (0,3,0) point 3 (0,0,5) ...
point 4 (3,3,0) point 5 (0,5,5) point 6 (5,0,5) point 7 (5,5,5)
zone create brick size 5 5 5 point 0 (0,0,5) edge 5.0
zone group 'Material'
; Create Bin Zones
point 0 (3,0,0) point 1 add (3,0,0) point 2 add (0,3,0) ...
point 3 add (2,0,5) point 4 add (3,6,0) point 5 add (2,5,5) ...
point 6 add (3,0,5) point 7 add (3,6,5) group 'Bin1'
FLAC3D 6.0

; Assign models to groups
zone cmodel assign mohr-coulomb range group 'Material'
zone cmodel assign elastic range group 'Bin1' or 'Bin2' or 'Bin3' or 'Bin4'
; Separate zones, and create two interfaces
zone separate by-face new-side group 'Interface' ...
range group 'Material' group 'Bin1' or 'Bin2'
zone interface 1 create by-face range group 'Interface' group 'Bin1'
zone interface 2 create by-face range group 'Interface' group 'Bin2'
; Subdivide interface elements
zone interface 1 element maximum-edge 0.75
zone interface 2 element maximum-edge 0.75
; Assign properties
zone property shear 1e8 bulk 2e8 fric 30 range group 'Material'
zone property shear 1e8 bulk 2e8 range group 'Bin1' or 'Bin2'
zone property density 2000
zone interface 1 node property stiffness-shear 2e9 ...
stiffness-normal 2e9 friction 15
stiffness-normal 2e9 friction 15
; Assign Boundary Conditions
zone gridpoint fix velocity-z range position-z 0 group 'Bin1' or 'Bin2'
; Monitor histories
zone history displacement-z position (6,6,10)
; Settings
model largestrain on
model gravity 10
; Cycling
model step 3000
model save 'bin'
All Properties Have Physical Significance

In this case, properties should be derived from tests on real joints [1] (suitably
scaled to account for size effect), or from published data on materials similar to
the material being modeled. However, the comments in Interface Used to Join
Two Sub-grids also apply here with respect to the maximum stiffnesses that are
reasonable to use. If the physical normal and shear stiffnesses are less than ten
times the equivalent stiffness of adjacent zones, then there is no problem in
using physical values. If the ratio is much more than ten, the solution time will
be significantly longer than for the case in which the ratio is limited to ten,
FLAC3D 6.0
without much change in the behavior of the system. To improve solution

efficiency, serious consideration should be given to reducing supplied values of
normal and shear stiffnesses.
There may also be problems with interpenetration if the normal stiffness, , is

very low. A rough estimate of the joint normal displacement that would result
from the application of typical stresses in the system ( ) should be made.
This displacement should be small compared to a typical zone size. If it is greater
than, for example, 10% of an adjacent zone size, then there is either an error in
one of the numbers, or the stiffness should be increased if calculations are to be
done in large-strain mode.
Joint properties are conventionally derived from laboratory testing (e.g., triaxial
and direct shear tests). These tests can supply physical properties for joint
friction angle, cohesion, dilation angle, and tensile strength, as well as joint
normal and shear stiffnesses. The joint cohesion and friction angle correspond to
the parameters in the Coulomb strength criterion, [2] described in Formulation.
Values for normal and shear stiffnesses for rock joints typically can range from
roughly 10 to 100 MPa/m for joints with soft clay in-filling, to over 100 GPa/m
for tight joints in granite and basalt. Published data on stiffness properties for
rock joints are limited; summaries of data can be found in Kulhawy (1975), Rosso
(1976), and Bandis et al. (1983).
Approximate stiffness values can be back-calculated from information on the

deformability and joint structure in the jointed rock mass and the deformability
of the intact rock. If the jointed rock mass is assumed to have the same
deformational response as an equivalent elastic continuum, then relations can be
derived between jointed rock properties and equivalent continuum properties.
For uniaxial loading of rock containing a single set of uniformly spaced joints
oriented normal to the direction of loading, the following relation applies:
(2)
or
(3)
where:
= rock mass Young’s modulus;
FLAC3D 6.0
= intact rock Young’s modulus;
= joint normal stiffness; and
= joint spacing.
A similar expression can be derived for joint shear stiffness:
(4)
where:
= rock mass shear modulus;
= intact rock shear modulus; and
= joint shear stiffness.
The equivalent continuum assumption, when extended to three orthogonal joint

sets, produces the following relations:
(5)
(6)
Several expressions have been derived for two- and three-dimensional

characterizations and multiple joint sets. References for these derivations can be
found in Singh (1973), Gerrard (1982a and 1982b), and Fossum (1985).
Published strength properties are more readily available for joints than for
stiffness properties. Summaries can be found, for example, in Jaeger and Cook
(1979), Kulhawy (1975), and Barton (1976). Friction angles can vary from less
than 10° for smooth joints in weak rock, such as tuff, to over 50° for rough joints
in hard rock, such as granite. Joint cohesion can range from zero to values
approaching the compressive strength of the surrounding rock.
It is important to recognize that joint properties measured in the laboratory

typically are not representative of those for real joints in the field. Scale
dependence of joint properties is a major question in rock mechanics. Often, the
only way to guide the choice of appropriate parameters is by comparison to
similar joint properties derived from field tests. However, field test observations
are extremely limited. Some results are reported by Kulhawy (1975).
FLAC3D 6.0
The following example illustrates an application of the interface logic to simulate

the physical response of a rock joint subjected to normal and shear loading. The
model represents a direct shear test, which consists of a single horizontal joint
that is first subjected to a normal confining stress, and then to a unidirectional
shear displacement. Figure Figure #fig-directsheartest-model shows the model.
Figure 5: Direct shear test model.
First, a normal stress of 10 MPa that is representative of the confining stress

acting on the joint is applied. A horizontal velocity is then applied to the top sub-
grid to produce a shear displacement along the interface. For demonstration
purposes, we only apply a small shear displacement of less than 2 mm to this
model.
The average normal and shear stresses, and normal and shear displacements
along the joint, are measured with a FISH function. With this information, we can
determine the shear strength and dilation that are produced. The data file for
this test is shown in “DirectShearTest.f3dat”. The corresponding project file,
“DirectShearTest.f3prj”, is located in the folder “datafiles\Interface\
DirectShearTest.”
FLAC3D 6.0
DirectShearTest.f3dat
model new
model title "Direct shear test"
zone create brick size 12 1 10 point 0 4 0 5 point 1 16 0 5 ...
point 2 4 1 5 point 3 4 0 10 group 'top'
point 3 0 0 5 group 'bot' merge off
zone interface 1 create by-face range group 'bot' position-z 5.0
zone interface 1 node property stiffness-shear 4e4 stiffness-normal 4e4 ...
friction 30 dilation 6 ;tension 1e10 bslip=on
zone face apply stress-normal -10 range position-z 10
model step 0
model solve
model save 'dsta.f3sav'
;
zone gridpoint fix velocity-x 5e-7 range group 'top'
call 'fishFunctions'
@ini_jdisp
history interval 10
fish history @sstav
fish history @nstav
fish history @sjdisp
fish history @njdisp
zone gridpoint initialize displacement (0,0,0)
model step 2500
model save 'dst.f3sav'
return
The average shear stress versus shear displacement along the joint is plotted in
Figure 6, and the average normal displacement versus shear displacement is
plotted in Figure 7. These plots indicate that joint slip occurs for the prescribed
properties and conditions. The loading slope in Figure 6 is initially linear, and
then becomes nonlinear as interface nodes begin to fail until a peak shear
strength of approximately 5.8 MPa is reached. As indicated in Figure 7, the joint
begins to dilate when the interface nodes begin to fail in shear.
FLAC3D 6.0
Figure 6: Average shear stress versus shear displacement.
Figure 7: Average normal displacement versus shear displacement.
FLAC3D 6.0
We can illustrate the influence of pore pressure on slip failure in a joint, by

conducting a uniaxial compression test on the FLAC3D model of a specimen with
a single joint that we created in “DippingJoint.f3dat”. The properties of the joint
specimen are
normal stiffness ( ) 1 GPa/m

shear stiffness ( ) 1 GPa/m
cohesion ( ) 1 kPa
friction angle ( ) 30°
dilation angle ( ) 0°
tensile strength ( ) 10 GPa
The properties of the intact material are
Young’s modulus ( ) 170 MPa

Poisson’s ratio ( ) 0.22
cohesion ( ) 2 kPa
tensile strength ( ) 2.4 kPa
Slip will occur in a triaxial test, provided 0 (Jaeger and Cook

1979), where is the angle between the vertical direction and the joint plane.
Assuming failure occurs due to slip along the joint, the peak stress, , for the
uniaxial compression test is related to the strength properties of the joint by the
following expression (see Jaeger and Cook 1979):
(7)
where is the constant pore pressure applied to the specimen (i.e., an undrained
compression test, but with zero fluid modulus); = 45° in this case. We perform
two simulations to evaluate the influence of the pore pressure, , on the peak
stress . For the given material properties, and with = 0, = -4.7 kPa. If we
specify a pore pressure of 1.0 kPa, then = -2.0 kPa.
The data file to run a uniaxial test on the jointed specimen is contained in
“UniaxialCompTest.f3dat”. The project file, “UniaxialCompTest.f3prj”, is located
in the folder “datafiles\Interface\UniaxialCompTest.f3dat”
FLAC3D 6.0
Two FISH functions (stress and strain) are used to calculate the stress-strain
response during the test. The results are compared to the values calculated from
Equation (7), which are stored as the FISH parameter analytical. Note that the
tensile strength of the joint is set to a high value, but the property switch bonded-
slip is set to on to allow slip along the joint.
UniaxialCompTest.f3dat
model new
model title "Influence of pore pressure on slip failure in a joint"
; Create Base
point 0 (0,0,0) point 1 (3,0,0) point 2 (0,3,0) ...
point 3 (0,0,1.5) point 4 (3,3,0) point 5 (0,3,1.5) ...
point 6 (3,0,4.5) point 7 (3,3,4.5) group 'Base'
; Create Top - 1 unit high for initial spacing
point 0 (0,0,1.5) point 1 (3,0,4.5) point 2 (0,3,1.5) point 3 (0,0,6) ...
point 4 (3,3,4.5) point 5 (0,3,6) point 6 (3,0,6) point 7 (3,3,6) group 'Top'
; Create interface elements on the top surface of the base
zone interface 1 create by-face separate range group 'Base' group 'Top'
; Assign Material model, continuum properties, and interface properties
zone cmodel assign mohr-coulomb
zone property bulk 1e8 shear 7e7 friction 40 cohesion 2e3 tension 2400
zone interface 1 node property stiffness-normal 1e9 stiffness-shear 1e9 ...
friction 30 cohesion 1e3 tension 1e10 ...
bonded-slip on
; Fix upper and lower boundaries, and assign velocity
zone gridpoint fix velocity-z -1e-8 range position-z 6
zone gridpoint fix velocity-z 1e-8 range position-z 0
[global zpnt1 = gp.near(0,0,0)]
[global zpnt2 = gp.near(0,0,6)]
[global analytical = 4.732e3]
;
model history mechanical unbalanced-maximum
fish history @stress
fish history @analytical
fish history @strain
model save 'state1'
;
model cycle 12000
model save 'slip_p0'
The stress-strain results for zero pore pressure are plotted in Figure 8, and for
= 1.0 kPa in Figure 9. The figures also show that the peak stresses calculated by
FLAC3D compare well with the peak stresses calculated from Equation (7). The
FLAC3D 6.0
FLAC3D results can be further improved if a servo-control is used with the

applied velocity to restrict the maximum unbalanced force as the joint begins to
slip.
Figure 8: Stress-strain response of a jointed specimen, p = 0.0.
FLAC3D 6.0
Figure 9: Stress-strain response of a jointed specimen, p = 1.0 kPa.
Endnotes
[1] “Joint” is used here as a generic term.
[2] The Coulomb yield surface provides a reasonable approximation for joint
strength for most engineering calculations. More complex joint models are
available (for example, effects of continuous yielding and displacement
weakening). For analysis with other joint models, the user is referred to
UDEC (Itasca 2011).
FLAC3D 6.0
Modeling Guidelines 79
Modeling Guidelines
Troubleshooting
The geometry of the model should be checked after the grid has been generated
and the interface elements added, but before the grid is modified to bring the
faces into contact. The “zone” plot item should be combined with the “interface”
plot item, given with a different color, to show that interface elements have been
placed on the right surfaces.
After the two sides of the interface have been brought into contact, execute a
model step 0 command to calculate contact locations. The zone interface i list
information command should be given to list the target faces for all nodes in
contact on interface i, along with the weighting factors for distribution of forces
to the target face. During the calculation process, the “interface” plot item
colored by normal stress or shear stress can be used to monitor the development
of shear and normal stresses along the interface. Note that interface property
values will default to zero if not given. For example, zero normal stresses may
indicate that the normal stiffness was not specified.
Initial Stresses
Interface stresses will not be initialized automatically when stresses are
initialized in the grid (with the initialize-stresses keyword). The zone
interface node initialize-stresses commands can be used to initialize the
interface normal stress and shear stress vector components. Alternatively, the
model containing interfaces can be stepped to an initial equilibrium state;
displacements may be set to zero after the initial state is reached.
Interface Corners
Pay special attention to interfaces that will result in sharp corners (see Figure 1).
The interface node at the corner has a representative area that is twice the value
that is actually in contact. If only the bottom surface is to come into contact (as
FLAC3D 6.0
indicated in the figure), then only that surface should have interface elements
applied. If both surfaces will come into contact, two separate interfaces should be
specified.
Figure 1: Effective area at an interface corner.
There is a separate problem for interior corners (see Figure 2). In this case, the
interface elements could be placed either on side A (facing into the corner) or
side B (facing away from the corner). If the elements are placed on side A, then
the corner node will move into an area that is not “behind” the faces on either
side, and a contact will not be detected. Instead, the interface elements should be
placed on side B. Additionally, two different interfaces should be created for the
two edges of the corner, as described in Figure 1.
FLAC3D 6.0
Figure 2: Interface at an interior corner.
Overlapping Interfaces
Interface nodes can only register contact with one face at a time. This can cause
problems in the areas of multiple intersecting interfaces (see Figure 3). Interface
node A on the bottom surface will register contact with either target face b or
face c. The order in which the interfaces are specified, or the effect of rounding
off, determines this. This means that either gridpoint B or C will not register an
appropriate contact and resisting force, causing arching at the interface
connection. One solution is to declare two interfaces facing each other, each with
half the stiffness and strength of the whole; another is to increase the number of
interface nodes in that area. Either of these solutions will reduce the problem,
but may not eliminate it.
FLAC3D 6.0
Figure 3: Intersection of overlapping interfaces.
Interfaces and Fluid Flow

The complete interaction of a fluid with an interface is not modeled. Fluid may
flow across an interface in the normal direction without resistance, provided that
the two surfaces are in contact; flow within the interface (in the parallel
direction) is not modeled. The zone interface i permeability = on/off command
can be used to turn this option on or off in interface i (the default is on). If model
configure fluid is invoked, changing the interface element size by specifying the
maximum-edge keyword with the zone interface element command will
automatically make the interface impermeable, even if the keyword permeability
= on has been issued. Do not specify maximum-edge if flow across the interface is
desired.
The fluid pressure at interface nodes may be used to determine the effective
stress for the purpose of computing slip or tensile failure conditions, but the
fluid does not exert any mechanical force on the sides of the interface. The zone
interface i effective = on/off command may be used to work in effective or total
stress, respectively, for interface i (the default is on). There is no coupling
between joint movement and volume changes in the fluid.
FLAC3D 6.0
Interfaces and Changing Interacting Objects in Small Strain

Interfaces operating in small-strain mode derive their forces from a comparison
of “virtual positions” of the two interacting faces, where a virtual position of a
point is the original coordinate of the point plus the accumulated displacement
(displacement-small) to date. If one contacting object is removed (e.g., backfill
within a tunnel), and another substituted (e.g., a liner is installed) using the
same interface, then large initial stresses may appear because the two sides of
the interface appear to interpenetrate, because the virtual position of the old
object has changed with respect to the specified position of the new object.
The stored displacements associated with a particular interface n may be

multiplied by a factor v, using the keyword displacement-small combined with
the zone interface node command. Normally, this factor will be zero, thus
restoring the virtual positions of the faces making up the interface “host” to
their original locations. In this case, a new interacting object will not cause initial
stresses to be generated when its coordinates match the original coordinates of
the interface host faces. However, it is possible to give v a nonzero value, in order
to simulate an initial lack-of-fit between the two contacting objects.
FLAC3D 6.0
FLAC3D 6.0
References 85
References
Bandis, S. C., A. C. Lumsden and N. R. Barton. “Fundamentals of Rock Joint
Deformation,” Int. J. Rock Mech. Min. Sci. & Geomech. Abstr., 20(6), 249-268 (1983).
Barton, N. “The Shear Strength of Rock and Rock Joints,” Int. J. Rock Mech. Min.
Sci. & Geotech. Abstr., 13, 255-279 (1976).
Fossum, A. F. “Technical Note: Effective Elastic Properties for a Randomly

Jointed Rock Mass,” Int. J. Rock Mech. Min. Sci. & Geomech. Abstr., 22(6), 467-470
(1985).
Gerrard, C. M. “Elastic Models of Rock Masses Having One, Two and Three Sets
of Joints,” Int. J. Rock Mech. Min. Sci. & Geomech. Abstr., 19, 15-23 (1982a).
Gerrard, C. M. “Equivalent Elastic Moduli of a Rock Mass Consisting of

Orthorhombic Layers,” Int. J. Rock Mech. Min. Sci. & Geomech. Abstr., 19, 9-14
(1982b).
Itasca Consulting Group, Inc. FLAC (Fast Lagrangian Analysis of Continua), Version
7.0. Minneapolis: Itasca (2011).
Itasca Consulting Group, Inc. 3DEC (Three-Dimensional Distinct Element Code),

Version 4.1. Minneapolis: Itasca (2007).
Itasca Consulting Group, Inc. UDEC (Universal Distinct Element Code), Version 5.0.
Minneapolis: Itasca (2011).
Jaeger, J. C., and N. G. W. Cook. Fundamentals of Rock Mechanics, Third Ed.

London: Chapman and Hall (1979).
Kulhawy, F. H. “Stress Deformation Properties of Rock and Rock

Discontinuities,” Eng. Geol., 9, 327-350 (1975).
Rosso, R. S. “A Comparison of Joint Stiffness Measurements in Direct Shear,

Triaxial Compression, and In Situ,” Int. J. Rock Mech. Min. Sci. & Geomech. Abstr., 13,
167-172 (1976).
Singh, B. “Continuum Characterization of Jointed Rock Masses: Part I — The

Constitutive Equations,” Int. J. Rock Mech. Min. Sci. & Geomech. Abstr., 10, 311-335
(1973).
FLAC3D 6.0
FLAC3D 6.0
Factor of Safety 87
Factor of Safety
The theoretical background for FLAC3D factor of safety calculation is presented in
this section.
Section Outline
• Introduction
• Factor of Safety
• Computational Methods for Factor of Safety Calculation of Slopes
◦ Strength Reduction Technique
◦ Limit Analysis
◦ Limit Equilibrium
◦ Relation of Strength Reduction Method to Limit Equilibrium and
Limit Analysis
• Strength Reduction Procedure in FLAC3D
◦ Strength Reduction Properties
▪ Mohr-Coulomb Material
▪ Ubiquitous-Joint Material
▪ Hoek-Brown Material
▪ Interfaces
• Example Factor of Safety Calculations using the Strength Reduction
Method
◦ Failure of a Slope with a Complex Surface Profile in a Mohr-
Coulomb Material
◦ Influence of Slope Curvature on Stability
◦ Simple Slope in Hoek-Brown Material
◦ Automatic Calculation of a Stable Pit Slope Angle
◦ Factor of Safety Contours
• References
FLAC3D 6.0
88 FLAC3D Theory and Background • Factor of Safety
FLAC3D 6.0
Introduction 89
Introduction
A factor-of-safety calculation can be performed for stability analyses in FLAC3D.
This calculation is based upon the strength reduction method to determine a
factor of safety. The strength reduction method is an increasingly popular
numerical method to evaluate factor of safety in geomechanics (e.g., see Dawson
and Roth 1999, and Griffiths and Lane 1999). Although the method has been used
extensively in the context of Mohr-Coulomb material, there are a few references
available in the literature that extend the approach to nonlinear failure criteria in
general (i.e., Dawson et al. 2000, Shukha and Baker 2003, Hammah et al. 2005,
and Fu and Liao 2009). An overview of factor of safety and the strength reduction
technique is given in the section of Factor of Safety.
The strength reduction method can be applied to calculate the safety factor for a
variety of different above-ground and underground structures (e.g., slopes,
retaining walls, tunnels, etc.). In this section, the focus is on the factor of safety
of slopes because this is the most common practical application of the method. In
the section of computational methods for Factor of Safety, the strength reduction
method is described and compared to other computational methods (i.e., limit
analysis and the limit equilibrium technique) commonly used to determine a
safety factor for slopes.
The strength reduction procedure can be conducted manually in FLAC3D by

reducing selected strength properties until failure occurs. The method can also be
performed automatically by issuing the model factor-of-safety command in
FLAC3D. From FLAC3D Version 5.0, model factor-of-safety can be applied to
strength properties for the Mohr-Coulomb material model ( zone cmodel assign
mohr-coulomb), the ubiquitous-joint model (zone cmodel assign ubiquitous-
joint), and the Hoek-Brown model (zone cmodel assign hoek-brown). It also can
be applied to strength properties for interfaces. The properties affected by model
factor-of-safety are described Strength Reduction Properties.
Example factor of safety calculations based upon the strength reduction

technique are also provided in this section. These examples include comparisons
to solutions from limit analysis and the limit equilibrium technique. The
examples are described and data files are listed in Factor of Safety Calculation
Examples.
FLAC3D 6.0
FLAC3D 6.0
Factor of Safety 91
Factor of Safety
A “factor of safety” index can be defined for any relevant problem parameter by
taking the ratio of the calculated parameter value under given conditions to the
critical value of the parameter at which the onset of an unacceptable outcome
manifests itself. A relevant problem parameter could be a dimensionless group
that governs the problem at hand (e.g., a stability number). Examples of
(dimensional) parameters for slope stability include slope height, water level,
applied load, and strength property.
Unacceptable outcome relates to “safety” (and is usually taken as shear failure),

but other possibilities, such as displacement above a given threshold,
convergence beyond an acceptable level (such as in a tunnel excavation),
toppling failure, slope raveling (cyclic freezing/thawing, weathering), etc. can
also be considered.
By convention, a factor of safety index larger than one indicates stable

conditions. Thus, factor of safety index is taken as the actual over the critical
parameter value if the parameter value above critical is acceptable (e.g., material
cohesion), and as the inverse of this ratio otherwise (e.g., slope height). Note
that, with the exception of simple cases, the calculated factor of safety index will
not in general be linearly related to the selected problem parameter for which it
is defined. Also, different measures will give different values of factor of safety
for the same problem. Factor of safety index is most valuable when used on a
comparative basis, in analyses using the same index definition (e.g., use of the
index may produce the following statement: this slope with wider benches has a
higher index than that with higher benches).
The effort involved in computing the factor of safety index (once the definition is
established) consists of identifying actual as well as critical parameter values. In
the most general case, the actual parameter value is evaluated by direct
resolution of field and constitutive equations governing the problem, and this
often is done using a numerical method. On the other hand, an inverse boundary
value problem needs to be solved to estimate the critical value of the parameter.
In principle, this can be achieved using a trial-and-error technique, whereby
numerical simulations are performed for a range of parameter values until the
critical value is found. We refer to this general approach as the “parameter
FLAC3D 6.0
reduction technique.” Any appropriate geo-mechanical software (e.g., finite

difference, finite element, and distinct element method) can be used to perform
this task for problems involving various levels of complexity (e.g., geometry,
material constitutive law, discrete fracture network, slope reinforcement,
support systems, mechanical structures, etc.).
FLAC3D 6.0
Computational Methods for Factor of Safety Calculation of Slopes 93
Computational Methods for Factor of Safety

Calculation of Slopes
Three different computational methods are commonly employed in numerical
analyses programs to calculate a factor of safety for slopes: strength reduction
method, limit analysis (upper- and lower-bound solutions), and limit
equilibrium method (upper-bound solution). The strength reduction method is
used in FLAC3D, and can be executed automatically via the model factor-of-
safety command. This implementation is described below, followed by
Numerical Limit Analysis and Limit Equilibrium Analysis.
Strength Reduction Technique

The “strength reduction technique” is typically applied in factor of safety
calculations by progressively reducing the shear strength of the material to bring
the slope to a state of limiting equilibrium. The method is commonly applied
with the Mohr-Coulomb failure criterion (e.g., see applications by Zienkiewicz et
al. 1975, Naylor 1982, Donald and Giam 1988, Matsui and San 1992, Ugai 1989,
and Ugai and Leshchinsky 1995). In this case, the safety factor is defined
according to the equations
(1)
(2)
A series of simulations are made using trial values of the factor to reduce
the cohesion, , and friction angle, , until slope failure occurs. (Note that if the
slope is initially unstable, and will be increased until the limiting condition is
found.) One technique to find the strength values that correspond to the onset of
failure is to monotonically reduce (or increase) the strengths in small increments
until a failure state is found. Alternatively, in FLAC3D, a bracketing approach
similar to that proposed by Dawson, Roth and Drescher (1999) is used when the
model factor-of-safety command is executed. With this technique, stable and
FLAC3D 6.0
unstable bracketing states are found first, and then the bracket between the
stable and unstable solution is progressively reduced until the difference between
stable and unstable solutions falls below a specified tolerance.
The strength reduction method implemented in FLAC3D will always produce a

valid solution: in the case of an unstable physical system, FLAC3D simply shows
continuing motion in the model. An iteration solution, which is often used in the
finite element method, is not used here. The FLAC3D solution is a dynamic, time-
marching simulation in which continuing motion is as valid as equilibrium.
Neither is there iteration in the use of elastic-plastic constitutive laws: the stress
tensor is placed exactly on the yield surface (satisfying equations, such as the
flow rule and elastic/plastic strain decomposition) if plastic yield is detected. The
stress state in FLAC3D at a safety factor = 1 is the actual stress state that
corresponds to the yielding mechanism, not an arbitrary pre-yield stress state or
an elastic stress state.
The detection of the boundary between physical stability and instability is based
on an objective criterion in FLAC3D that decides whether the system is in
equilibrium or a state of continuing motion. Finer incremental changes that may
affect the solution in an iterative solution scheme are not needed in a time-
marching scheme and do not affect the solution. In order to determine the
boundary between physical stability and instability, a set of completely separate
runs is made with different strength-reduction factors. Each run is then checked
to determine whether equilibrium or continuing plastic flow is reached. The
point of failure can be found to any required accuracy (typically 1%) by
successive bracketing of the strength-reduction factors. This process should not
be confused with taking finer solution steps; the solution scheme is identical for
each run of the set (whether it results in equilibrium or continuing motion).
Limit Analysis
Limit analysis relies on the construction of solutions which obey upper- and
lower-bounds theorems developed in the theory of plasticity. These theorems
(presented in most textbooks on plasticity) provide rigorous limits on the
collapse conditions of a system consisting of a perfectly plastic material obeying
normality (associated flow rule). Of particular interest is the lower-bound
theorem, which states (Davis and Selvadurai 2002) that “Collapse will not occur
if any state of stress can be found that satisfies the equations of equilibrium and
the traction boundary conditions and is everywhere ‘below yield’.”
FLAC3D 6.0
In this theorem, the words “equations of equilibrium” pertain to local

equilibrium. Any stress field that satisfies the criteria of the lower-bound
theorem is referred to as a statically admissible stress field. Also, in a factor-of-
safety calculation, a statically admissible stress field provides a lower-bound
(conservative) estimate for the FOS.
It is also useful to recall the upper-bound theorem, which states that (Davis and
Selvadurai 2002) “Collapse must occur if, for any compatible plastic deformation,
the rate of working of the external forces on the body equals or exceeds the rate
of internal energy dissipation.”
In this statement, “compatible plastic deformation” means any deformation that

satisfies all displacement boundary conditions and is possible kinematically
according to the associated flow rule, which governs admissible dilation. Any
deformation field that satisfies the criteria of the upper-bound theorem is
referred to as kinematically admissible deformation.
Stability charts for homogeneous simple slopes (in “cohesive” material) are still
used in practice as a first estimate of slope safety. Typically, values in the chart
obtained using limit analysis (upper- and lower-bound solutions) are presented
in the form of stability numbers (see, e.g., Taylor 1937, Dawson et al. 2000,
Michalowski 2002, and Li et al. 2008). These numbers are dimensionless
quantities that relate slope height, material unit weight, and the material
strength property of cohesion for a Mohr-Coulomb material, or intact
unconfined compressive strength for a Hoek-Brown material. Stability numbers
have been associated with nontraditional FOS measures (e.g., for Mohr-Coulomb
(Michalowski 2002), and for Hoek-Brown (Li et al. 2008)).
Limit Equilibrium
Limit equilibrium (LE) methods are approximate methods that assume the
existence of a slip surface of various simple shapes: plane, circular, or logspiral.
The methods are based on the additional assumption that the soil or rock mass
can be divided into slices. The problem is reduced to one of finding the most
critical position for the slip surface of the chosen shape. Various methods exist,
including Fellenius’ (1936), Bishop’s (1955), Lowe and Karafiath’s (1960),
Janbu’s (1968), Morgenstern and Price’s (1965), and Spencer’s (1967). One of the
main differences between methods concerns assumptions made about side force
directions between slices, with potential implications for equilibrium. A
comparative description summary of methods with assumptions and limitations
may be found in TRB Special Report (1996) and Abramson et al. (2002).
FLAC3D 6.0
Note that none of the equations of solid mechanics is explicitly satisfied inside or
outside of the failure surface (assumed slip surface). Also, according to Chen
(2007):
Although the limit equilibrium technique utilizes the basic philosophy of the
upper-bound rules of limit analysis, that is, a failure surface is assumed and a
least answer is sought, it does not meet the precise requirements of the upper-
bound rules so that it is not an upper bound. The method basically gives no
consideration to soil kinematics, and equilibrium conditions are satisfied only in
a limited sense. It is clear then that a solution obtained using the limit
equilibrium method is not necessarily an upper or a lower bound.
Relation of Strength Reduction Method to Limit Equilibrium

and Limit Analysis
As mentioned in Limit Equilibrium Analysis, a limit equilibrium (LE) solution is
never a lower bound for the load because, although global equilibrium is satisfied
by the LE solution, local equilibrium is not guaranteed (none of the LE solutions
are statically admissible).
Also, a strong statement made in the literature (e.g., Davis and Selvadurai 2002)
is that the results from LE will always be the same as those from the upper-
bound theorem for any translational collapse mechanism (meaning system of
rigid soil blocks separated by thin shear surfaces). Thus, there are cases for
which an LE solution gives an upper bound for the load (Drescher and Detournay
1993).
One may ask then why an LE solution “works” since not only is it not guaranteed
to provide a lower bound for the FOS, but in some cases it is even proven to give
an upper bound for the FOS. An answer, provided by Wa-I-Fah Chen in his book
Limit Analysis and Soil Plasticity, rests on the observation that most FOS analyses
are concerned with slopes, and apparently, for most slopes, the LE solution
provides an FOS value which is close to the exact solution.
On the other hand, consider the last stable state calculated by FLAC3D (the last
lower bracket, which is typically 0.005 less than the final FOS) for an associated
problem. FLAC3D will provide an approximate exact solution to the problem at
that state, in the sense that local equilibrium may not be satisfied everywhere at
the boundary between zones, but if the zone size is reduced to zero, local
equilibrium will be satisfied to the limit. In particular, the limit stress field
satisfies the lower-bound theorem. Also, the deformation field at the “failure
FLAC3D 6.0
state” calculated by FLAC3D (the last upper bracket) is a kinematically admissible

deformation (it fulfills all the criteria of the upper-bound theorem). Thus, one
may say that if the calculated FOS tends to a limit as the grid size is reduced, this
limit may be considered to be very close to (within 0.005) the exact FOS for the
problem.
In summary then, in most cases FLAC3D (on a fine grid) and an LE solution will
give factors of safety that are very similar. In some cases, FLAC3D will give a
safety factor on a fine grid that is lower than that provided by a limit equilibrium
(LE) solution. This implies that the LE solution provides an upper bound for the
FOS. In other cases, FLAC3D will give a safety factor on a fine grid that is higher
than that provided by a limit equilibrium (LE) solution. This does not mean that
FLAC3D is nonconservative, but instead that we have encountered a case where
the LE solution cannot be relied upon (since it can never correspond to a lower
bound for the load).
Note that the limit-analysis bound theorems apply to an associated flow rule (see
Davis and Selvadurai 2002). This rule may not be very realistic in some cases, as
it provides far too much dilation. However, nonassociated flow rules do not
guarantee unique solutions. Without this assurance, a collapse load is no longer
unique. Apparently, the only useful result that can be obtained is that a
nonassociative material can be no stronger than an associative one. This follows
from the observation that, at collapse, the actual stress field in a nonassociative
soil is statically admissible. Therefore, by the lower-bound theorem, the collapse
load for a nonassociative material cannot exceed that for the corresponding
material with the associated flow rule.
FLAC3D 6.0
FLAC3D 6.0
Strength Reduction Procedure in FLAC3D 99
Strength Reduction Procedure in FLAC3D

The strength reduction method can be applied to essentially any material failure
model to evaluate a factor of safety based upon the reduction of a specified
strength property or property group. The method has been used extensively in
the context of Mohr-Coulomb material and, principally, the simultaneous
reduction of cohesion and frictional strength. From FLAC3D Version 5.0, in
addition to Mohr-Coulomb strength properties (assigned with zone cmodel
assign mohr-coulomb), the method is also applied automatically to ubiquitous-
joint strength properties (assigned with zone cmodel assign ubiquitous-joint),
and to Hoek-Brown strength properties (assigned with zone cmodel assign hoek-
brown) when the model factor-of-safety command is given.
The strength reduction method can also be applied when model factor-of-safety
is executed for interface strength properties, friction and cohesion (assigned via
zone interface node property).
The procedure for implementing the strength reduction technique in FLAC3D via
the model factor-of-safety command is as follows.
First, the code finds a “characteristic response time,” which is a representative

number of steps (denoted by ) that characterizes the response time of the
system. is found by setting the material strength (for Mohr-Coulomb
material, the cohesion and tensile strength) to a large value, making a large
change to the internal stresses (by default, a perturbation factor of 2 is applied to
the stress state), and finding how many steps are necessary for the system to
return to equilibrium.
A maximum limit of 50,000 is set for by default. If the model does not reach
equilibrium within 50,000 steps, the run will stop and the factor-of-safety
solution cannot be completed. If this happens, the user should review the
parameters selected for the model. For example, if the user has selected
structural support with a high value for Young’s modulus, this may affect the
solution convergence time.
FLAC3D 6.0
It is also possible to set the value for manually by using the characteristic-
steps keyword to specify a value for . Alternatively, the initial perturbation to
the internal stresses can be changed by specifying a different perturbation factor
using the perturbation keyword. Note that these manual controls should be used
with caution.
After is determined, for a given strength reduction factor, , steps are

executed. If the unbalanced force ratio [1] is less than 10-5 after steps, then
the system is in equilibrium. If the unbalanced force ratio is greater than 10 -5,
then another steps are executed, exiting the loop if the force ratio is less than
-5
10 . The mean value of force ratio, averaged over the current span of steps, is
compared with the mean force ratio over the previous steps. If the difference
is less than 10%, the system is deemed to be in nonequilibrium, and the loop is
exited with the new nonequilibrium, . If the above-mentioned difference is
greater than 10%, blocks of steps are continued until: (1) the difference is less
than 10%; or (2) 6 such blocks have been executed; or (3) the force ratio is less
than 10-5. The justification for case (1) is that the mean force ratio is converging
to a steady value that is greater than that corresponding to equilibrium; the
system must therefore be in continuous motion.
The following information is displayed during the solution process:
1. the operation currently being performed;
2. the number of calculation steps completed to determine a given

value of ;
3. twice the characteristic response time, (note that the change

in unbalanced force ratio is first checked at two times );
4. unbalanced force ratio;
5. the number of completed solution stages (i.e., tests for

equilibrium or nonequilibrium); and
6. current bracketing values of .
The factor of safety solution stops when the difference between the upper and
lower bracket values becomes smaller than 0.005 times the mean value. (This
resolution limit can be changed with resolution, an optional keyword to the
model factor-of-safety command.)
FLAC3D 6.0
The bracketing solution approach invoked with the model factor-of-safety

command may perform a large number of (stable and unstable) solutions before
determining a factor of safety. If an approximate range for the factor is known,
then the number of solutions (and total solution time) can be reduced by
specifying the starting bracket values. This can be done with bracket v1 v2, an
optional phrase to the model factor-of-safety command. If the calculated factor
falls outside the specified brackets, a warning message will be issued. It is also
possible to test whether a specified factor is above or below the actual factor, by
setting bracket v1 v2.
The following conditions should be noted when using model factor-of-safety.
1. The model state must be saved before a model factor-of-safety

calculation is performed.
2. The initial stress state can either be at a zero stress state or at

stress equilibrium for the model factor-of-safety calculation. If
the model is at a zero stress state, only gravity loading is applied
to determine .
3. The factor of safety calculation is performed in small-strain

calculation mode when model factor-of-safety is issued.
4. If a factor of safety calculation is performed for a coupled fluid

flow-mechanical model (with model configure fluid specified),
the fluid flow calculation will be turned off and fluid bulk
modulus will be set to zero when model factor-of-safety is
issued.
5. The factor of safety calculation assumes nonassociated plastic

flow with model factor-of-safety. The keyword associated can be
added for an associated plastic flow calculation.
Three files are saved during the model factor-of-safety calculation: the initial-
condition state, the latest stable factor of safety calculation, and the latest
unstable factor of safety calculation. By default, the initial state is saved to a file
named “FOS-Init.f3sav”, the stable state is saved to a file named “FOS-
Stable.f3sav”, and the unstable state to a file named “FOS-Unstable.f3sav”. Each
factor of safety calculation stage starts from the “FOS-Init.f3sav” state.
FLAC3D 6.0
A different file name can be specified for each of these three files (see the model
factor-of-safety filename command).
Strength Reduction Properties

The strength properties that can be reduced when using model factor-of-safety
are described in the following sections.
Mohr-Coulomb Material
If the Mohr-Coulomb model failure criterion is prescribed, cohesion, , and
friction angle, , are selected by default to be included in the safety factor
calculation when executing model factor-of-safety. The reduction equations for
these properties are
(1)
(2)
with the strength reduction procedure. These strengths can optionally be

excluded from the model factor-of-safety calculation with the keyword phrase
exclude cohesion or exclude friction.
Tensile strength, , can also be included with the optional phrase include
tension. The trial properties for tensile strength are calculated in a manner
similar to that used for material friction and cohesion. The reduction equation
for the tensile strength is
(3)
Ubiquitous-Joint Material
If the ubiquitous-joint model is used, strength values for the intact material,
and , and strength values for the ubiquitous joints, and , are included by
default in the model factor-of-safety calculation. Tensile strengths, and ,
can also be selected for reduction by adding the keyword phrases include tension
and include joint-tension, respectively. The reduction equations for the intact
material are the same as Equations (1) through (3), and for the ubiquitous joints
are
FLAC3D 6.0
(4)
(5)
(6)
Ubiquitous-joint cohesion and friction can be excluded from the safety factor
calculation with include joint-cohesion and exclude joint-friction,
respectively.
Hoek-Brown Material
The Hoek-Brown model supports factor of safety calculations with model factor-
of-safety. Strength reduction is performed with respect to shear strength flag-
fos = 0).
Note that, although the softening/hardening capabilities of the Hoek-Brown

model can be activated before the factor of safety calculation is performed, they
should be disabled (by removing the table property assignment) during the
strength reduction procedure because the value of the evolution parameter is
then ill-defined.
Factor of Safety with respect to Shear Strength,
The Hoek-Brown criterion can be approximated locally by a Mohr-Coulomb

criterion:
(7)
where apparent cohesion and friction are given in terms of the local value of
by
(8)
(9)
where (for compressive stresses positive) if :
(10)
FLAC3D 6.0
(11)
and, if :
(12)
(13)
A pragmatic approach to evaluate a factor of safety for slopes (based on the

strength reduction technique) is used, whereby local cohesion, , and friction
coefficient, , are divided by a factor until active slope failure is detected. The
factor directly applies to the maximum allowable value of shear stress (see
Equation (7)). The reduction factor at the verge of slope collapse is defined as the
FOS based on the proposed (local strength reduction) technique.
Although, in theory, it is possible to find a best fit to match the reduced envelope
with a Hoek-Brown type equation (see, e.g., Hammah et al. 2005), this step is
not required with this particular model implementation because the logic relies
on the direct use of envelope tangent (there is no need to define a curve and then
the tangent when the tangent is available in the first place—see above). Also, the
proposed local strength reduction technique provides a means by which to
quantify the shear stress allowance to collapse, as one would expect. In this case,
the reduction factor does not apply directly on model parameters (there is no
absolute reason why it should).
Factor of Safety with Respect to Unconfined Compression Strength,
The Hoek-Brown model supports an alternative factor of safety calculation with

model factor-of-safety; strength reduction is performed with respect to
unconfined compression strength by setting flag-fos = 1. The intact unconfined
compressive strength is reduced by a reduction factor until active failure is
detected to allow comparison with stability charts for simple slopes obtained by
Li et al. (2008), using limit analysis.
Interfaces
Interface strengths can be included in the safety factor calculation by adding
include interface to model factor-of-safety. For the interface strength values
and , the equations are
(14)
FLAC3D 6.0
(15)
[1] The unbalance force is the net force acting on a FLAC3D gridpoint. The ratio of
this force to the mean absolute value of force exerted by each surrounding
zone is the unbalanced force ratio. The limiting value for the unbalanced force
ratio can be changed with the optional keyword ratio to the model factor-
of-safety command.
FLAC3D 6.0
FLAC3D 6.0
Example Factor of Safety Calculations using the Strength Reduction Method 107
Example Factor of Safety Calculations using the

Strength Reduction Method
This section includes several examples that validate and demonstrate the
application of the strength reduction method in FLAC3D.
List of Examples
• Failure of a Slope with a Complex Surface Profile in a Mohr-Coulomb
Material
• Influence of Slope Curvature on Stability
• Simple Slope in Hoek-Brown Material
• Automatic Calculation of a Stable Pit Slope Angle
• Factor of Safety Contours
FLAC3D 6.0
FLAC3D 6.0
Failure of a Slope with a Complex Surface Profile in a Mohr-Coulomb Material 109
Failure of a Slope with a Complex Surface Profile

in a Mohr-Coulomb Material
Problem Statement
This example illustrates the ability of FLAC3D to identify the critical failure state
of a slope with a complex surface profile (i.e., a slope profile defined by more
than one inclination). The slope in this exercise is composed of two inclinations:
a 45° inclination in the lower part and a 26.7° inclination in the upper part. The
slope geometry is shown in Figure 1. The problem conditions are taken from
Cheng et al. (2007). Cheng et al. determine five local minima failure surfaces for
this slope using the Morgenstern-Price limit equilibrium method, as shown in
Figure 2. They identify two possible critical failure surfaces from a series of limit
equilibrium simulations: one at a factor of 1.383 and one at 1.3848.
Figure 1: Slope with complex surface profile.
FLAC3D 6.0
Figure 2: Local minima failure surfaces and factors of safety calculated by limit
equilibrium method (from Cheng et al. 2007).
The material properties of the slope material in this problem include the
following:
Elastic modulus, E = 14.0 MPa

Poisson’s ratio, ν = 0.3
Unit weight, γ = 20.0 Kn/m3
Cohesion, c = 10 kPa
Friction angle, ϕ = 30°
Dilation angle, ψ = 0.0
Tensile strength, σt = c /tan ϕ
FLAC3D Model
The FLAC3D model is constructed based upon the problem dimensions given by
Cheng et al. (2007). The slope geometry is oriented in the - plane and is one
zone thick in the -direction. The model geometry is shown in Figure 3.
Simulations were made with progressively finer zoning in order to determine the
mesh size that provides the most accurate solution. The calculated factor of
safety decreases as the mesh size is decreased, leveling off at a mesh created for
144 zones in the horizontal direction.
FLAC3D 6.0
Gravity is applied, and the model is solved to an equilibrium state to establish the
initial stresses in the model. Note that some isolated zones have reached a shear
failure state at this stage, as indicated in the plasticity state plot shown in Figure
4. The slope is stable.
Figure 3: Slope model geometry.
FLAC3D 6.0
Figure 4: Plasticity state in slope at initial force equilibrium.
The factor of safety calculation time can be reduced by selecting lower-bound

and upper-bound bracketing values that are close to the actual safety factor.
Based upon the range of factors shown in Figure 2, the lower-bound bracketing
factor is set to 1.30 and the upper-bound to 1.45. The bracketing factors are set
with bracket, an optional keyword to the model factor-of-safety command.
The calculated factor of safety is 1.38 and the failure surface develops as defined
by a shear-strain contour plot shown in Figure 5. This surface closely matches
the local minima surface shown for the FOS = 1.3848 failure surface in Figure 2.
The failure state calculated by FLAC3D is the global minimum stability state.
FLAC3D 6.0
Figure 5: Factor of safety results.
Reference
Cheng, Y.M., T. Lansivaara and W.B. Wei “Two-dimensional slope stability
analysis by limit equilibrium and strength reduction methods,” Computers and
Geotechnics, 34, 137-150, 2007.
Data File
ComplexSlope.f3dat
;-------------------------------------------------------------
; complex surface profile in Mohr-Coulomb material
;-------------------------------------------------------------
model new
model title 'Slope with complex surface profile'
; create model
zone create brick point 0 0,0,0 point 1 10,0,0 point 2 0,0.25,0 ...
point 3 0,0,10 size 40,1,40
point 3 10,0,10 size 144,1,40
FLAC3D 6.0
point 3 16,0,16 point 4 46,0.25,10 ...

point 5 16,0.25,16 point 6 46,0,16 ...
point 7 46,0.25,16 size 144,1,24
point 3 26,0,21 point 4 46,0.25,16 ...
point 5 26,0.25,21 point 6 46,0,21 ...
point 7 46,0.25,21 size 144,1,20
zone face skin ; Label model boundaries
; initialize gravity
model gravity 10
; assign Mohr Coulomb model and properties
zone property density 2000.0 young 1.4e7 poisson 0.3 ...
fric 30 cohesion 10000.0 ten 1e10
; boundary conditions
zone face apply velocity-normal 0 range group 'East' or 'West1'
zone face apply velocity-normal 0 range group 'North' or 'South'
zone face apply velocity (0,0,0) range group 'Bottom'
model save 'initial'
; Solve to initial equilibrium
model solve elastic ratio-local 5e-3
model save 'slope-1'
model factor-of-safety bracket 1.30 1.45 ratio-local 5e-3 filename 'FOS-slope'
FLAC3D 6.0
Influence of Slope Curvature on Stability 115
Influence of Slope Curvature on Stability
Problem Statement
Actual slopes are not infinitely long and straight: usually, they are curved in both
plan and elevation. The effect of slope curvature can really only be analyzed with
a three-dimensional model.
Hoek and Bray (1981) observed that the lateral restraint provided by material on
either side of a potential slope failure will increase as the slope becomes more
concave. They recommend that when the radius of curvature of the slope is less
than the height of the slope, the allowed slope angle can be 10° steeper than the
angle suggested by conventional two-dimensional stability analyses. Further, for
radii of curvature greater than twice the slope height, the maximum slope angle
given by a two-dimensional analysis should be used.
The model shown in Figure 1 represents a quarter-section of an open pit. The

height of the slope is 25 m, and the slope angle is 2 vertical to 1 horizontal
(approximately 63°). It is expected that plane-strain conditions will prevail along
the plane y = -30 (see Figure 2), while axisymmetric conditions will be
predominant at the plane x = 0 (see Figure 3).
The free water surface imposed in this problem is shown in Figure 4. This
surface intersects the top of the model 50 m behind the toe of the slope, and
there is seepage on the bottom half of the slope face. This water table, under
steady-state conditions, will lead to the pore-pressure distribution shown in
Figure 5.
FLAC3D 6.0
Figure 1: “Bathtub” model to evaluate slope curvature.
Figure 2: Vertical plane through model at y = -30.
FLAC3D 6.0
Figure 3: Vertical plane through model at x = 0.
Figure 4: Free water surface.
FLAC3D 6.0
Figure 5: Pore-pressure contours.
The strength parameters chosen for this model are selected for comparison of
FLAC3D results to circular failure charts published by Hoek and Bray (1981).
Figure 6 shows which chart to use as a function of the groundwater flow
conditions. In our case, the chart used is number 4. For example, if we assume a
friction angle of 45° (tanϕ = 1) and a factor of safety F = 1, then we can draw a
horizontal line in chart number 4 (see Figure 7) until we intersect the slope angle
of 63°. If we draw a vertical line, we obtain a value of 0.06 for c/γHF. For a
specific weight, γ, of 25,000 N/m3 and a height, H, of 25 m, we obtain a cohesion
of 37.5 kPa.
For our analysis, we select a cohesion value of 100 kPa in order to start with a
stable slope. The value for c/γHtanϕ is then 0.16 and, using Figure 7, the value for
c/γHF is 0.1 and F is 1.61.
FLAC3D 6.0
Figure 6: Chart number as a function of groundwater flow conditions (adapted

from Hoek and Bray 1981).
FLAC3D 6.0
Figure 7: Circular failure chart number 4 (Hoek and Bray 1981).
Modeling Procedure
The FLAC3D model shown in Figure 1 was created interactively using the Building
Blocks pane. The State Record pane was used to convert the result to a data file
called “geometry.f3dat”. The contents of the Building Blocks pane can be be
viewed after restoring any of the resulting data files.
The free water surface is imported from a DXF file. In this case, the DXF file is
created via FISH in the file “create-water.f3dat”, but it could just as easily have
come from any source.
The model is assigned a Mohr-Coulomb material model and several properties:
bulk modulus 200 MPa

shear modulus 100 MPa
FLAC3D 6.0
friction angle 45°

cohesion 100 kPa
tension limit 100 kPa
The mass density of the dry material is 2500 kg/m3, the mass density of the
saturated material is 2600 kg/m3, and gravity is specified at 10 m/sec2 acting in
the negative z-direction. The geometry-count range element is used to assign the
saturated density to zones below the water table surface. Roller boundaries are
placed along the sides of the model, and the bottom of the model is pinned.
The factor of safety is calculated by the strength reduction method using the
model factor-of-safety command. The convergence criteria is changed to a
local-ratio of 1e-3, as this is a stricter condition than the default.
A value of 1.70 is calculated for F. This is slighly higher than the factor of safety
produced by the circular failure chart, which suggests that there is a slight effect
of slope curvature on the stability. The resulting failure surface is depicted by the
displacement contour plot shown in Figure 8; the plot is made after restoring the
file “slopefos-Unstable.f3sav”. This plot shows that a “scoop-shaped” failure
surface develops along the long side of the bathtub, but the slope is stable at the
end.
FLAC3D 6.0
Figure 8: Displacement contours in the FLAC3D model at the failure state.
This problem was also run with the two-dimensional program FLAC in both
plane-strain mode (see data file “SL-PS.DAT”) and axisymmetry mode (see data
file “SL-AXI.DAT”). The model geometry was created to match that in the
vertical section through the FLAC3D model (see Figure 9).
The calculation for factor of safety in the plane-strain model matches that from
the circular failure chart, F = 1.61. The displacement contour and vector plot at
failure shows a similar failure surface to that from FLAC3D. Compare Figure 8 to
Figure 11, which plots displacement contours and vectors on a vertical plane
through the FLAC3D model at y = -30.
The factor of safety calculation for the axisymmetric model produces a value for
F = 2.35. This further indicates that the greater curvature produces a more stable
slope.
FLAC3D 6.0
Figure 9: The FLAC model grid.
Figure 10: Displacement contours and vectors for plane-strain FLAC model.
FLAC3D 6.0
Figure 11: Displacement contours and vectors for FLAC3D model along a vertical
plane at y = -30.
Reference
Hoek, E., and J. W. Bray. Rock Slope Engineering, 3rd Ed. London: The Institute of
Mining and Metallurgy (1981).
Data File
SlopeCurvature.f3dat
;-------------------------------------------------------------
; influence of slope curvature on stability of an open pit
;-------------------------------------------------------------
model new
model title 'Influence of slope curvature on stability of an open pit'
; Model created using Building-Blocks, data file exported from State Record
call 'geometry' suppress
zone generate from-building-blocks set 'CurvedSlope'
FLAC3D 6.0
; initialize gravity
model gravity 10
; Import water table from DXF file.

geometry import 'water.dxf'
zone water density 1000
zone water set 'water'
; assign Mohr Coulomb model and properties

zone property bulk 2e8 shear 1e8 friction 45 cohesion 1e5 tension 1e5
; boundary conditions
zone gridpoint fix velocity-y range union position-y -40 position-y 80
zone gridpoint fix velocity range position-z -15
; Initialize unsaturated density

; Initialize saturated density below water table
zone initialize density 2600 ...
range geometry-space 'water' count 1 direction (0,0,1)
; initialize stresses
zone initialize-stresses total
; histories
history interval 10
model history mechanical ratio-local
zone history displacement-x position (24.5, -40,25)
zone history displacement-z position (24.5, -40,25)
zone history displacement-x position (24.5, 0,25)
zone history displacement-y position (24.5, 0,25)
zone history displacement-z position (24.5, 0,25)
zone history displacement-y position ( 0,24.5,25)
zone history displacement-z position ( 0,24.5,25)
; calculate fos
model factor-of-safety ratio-local 1e-3
return
FLAC3D 6.0
FLAC3D 6.0
Simple Slope in Hoek-Brown Material 127
Simple Slope in Hoek-Brown Material
Problem Statement
Verification exercises are performed to validate the factor of safety calculation
using Hoek-Brown material in FLAC3D. The exercises test the strength reduction
calculation based upon shear strength, .
The factor of safety with respect to Hoek-Brown shear strength is calculated for
a simple slope geometry and compared to results based upon other methods to
calculate a safety factor for Hoek-Brown material (generalized Hoek-Brown,
equivalent Mohr-Coulomb, and Bishop and Spencer limit equilibrium methods)
reported by Hammah et al. (2005). The rock slope for this comparison calculation
has an inclination of 45° and a height of 10 m. The rock is represented as a Hoek-
Brown material with the following properties:
= 5000 MPa
= 0.3
= 2500 kg/m3
= 0.067
= 0.000025
= 0.619
= 30 MPa
The FLAC3D model mesh used for this test is shown in Figure 1. By default, when
model factor-of-safety is executed for a FLAC3D model with zone cmodel assign
hoek-brown, the factor of safety calculation is performed for Hoek-Brown
material with respect to shear strength. The calculated factor of safety for this
test is 1.15. The failure surface is shown by the shear strain contour plot in Figure
2. The result compares well with the results reported by Hammah et al. (2005).
Table 2 summarizes the safety factors reported for this test.
FLAC3D 6.0
Figure 1: Slope model mesh.
Figure 2: Factor of safety and failure surface calculated for simple slope in
Hoek-Brown material.
FLAC3D 6.0
Table 2: Factor of Safety Results for Hoek-Brown Slope
Method Factor of Safety

FLAC3D using Hoek-Brown 1.15
Generalized Hoek-Brown strength reduction* 1.15

Equivalent Mohr-Coulomb strength reduction* 1.15
Bishop’s simplified limit equilibrium* 1.153
Spencer’s limit equilibrium* 1.152
* from Hammah et al. (2005)
Hammah et al. (2005) also report the results for the case in which a horizontal
layer of Mohr-Coulomb material is located at the toe of the slope. The layer is 1
m thick and has zero cohesion and 25° friction. The slope with the Mohr-
Coulomb layer is shown in Figure 3.
When model factor-of-safety is issued, the strength reduction method is

performed concurrently for Hoek-Brown material, as described in Hoek-Brown
Material, and for Mohr-Coulomb material, as described in Mohr-Coulomb
Material. The factor of safety calculated for this model is 0.97. The results are
shown in Figure 4.
Table 3 compares the FLAC3D result with results from other methods reported by
Hammah et al. (2005).
FLAC3D 6.0
Figure 3: Simple slope in Hoek-Brown material with a Mohr-Coulomb layer.
Figure 4: Factor of safety and failure surface calculated for simple slope in
Hoek-Brown material with Mohr-Coulomb layer.
FLAC3D 6.0
Table 3: Hoek-Brown Slope with Mohr-Coulomb Layer
Method Factor of Safety

FLAC3D using Hoek-Brown and Mohr-Coulomb 0.97
Generalized Hoek-Brown strength reduction* 0.95

Bishop’s simplified limit equilibrium* 0.934
Spencer’s limit equilibrium* 0.963
* from Hammah et al. (2005)
Reference
Hammah, R. E., et al. “The shear strength reduction method for the generalized
Hoek-Brown criterion,” ARMA/USRMS 05-810, 2005.
Data Files
HoekBrownSlope.f3dat
;************************************************************************
; FoS wrt Shear Strength for HB Material
;*************************************************************************
model new
; --- geometry ---
zone create brick point 0 ( 0,0,0) point 1 (40,0, 0) ...
point 2 ( 0,1, 0) point 3 ( 0,0, 8) size 80 1 16
zone create brick point 0 (15,0,8) point 1 (40,0, 8) ...
point 2 (15,1, 8) point 3 (25,0,18) ...
point 4 (40,1,8) point 5 (25,1,18) ...
point 6 (40,0,18) point 7 (40,1,18) size 50 1 20
; --- Assign model and properties
zone cmodel assign hoek-brown
zone property density 2.5e-3 young 5000 poisson 0.3 tension 1e10
zone property constant-mb 0.067 constant-s 2.5e-5 ...
constant-a 0.619 constant-sci 30
zone property flag-evolution 1 flag-fos 0
; --- boundary conditions ---
zone face apply velocity-normal 0 range group 'Bottom'
; --- settings ---
model gravity 10
FLAC3D 6.0
; --- solution ---

model factor-of-safety bracket 1.1 1.2 ratio-local 1e-4 filename 'HBSlope'
HoekBrownSlope2.f3dat
;************************************************************************
; FoS wrt Shear Strength for HB Material
; with a Mohr-Coulomb Layer
;*************************************************************************
model new
; --- geometry ---
zone create brick point 0 ( 0,0,0) point 1 (40,0, 0) ...
point 2 ( 0,1, 0) point 3 ( 0,0, 8) size 80 1 16
zone create brick point 0 (15,0,8) point 1 (40,0, 8) ...
point 2 (15,1, 8) point 3 (25,0,18) ...
point 4 (40,1,8) point 5 (25,1,18) ...
point 6 (40,0,18) point 7 (40,1,18) size 50 1 20
zone group 'layer' range position-z 8 9
; --- Assign model and properties
zone property density 2.5e-3 young 5000 poisson 0.3 tension 1e10
zone property constant-mb 0.067 constant-s 2.5e-5 ...
constant-a 0.619 constant-sci 30
zone property flag-evolution 1 flag-fos 0
; --- mohr-coulomb layer ---
zone cmodel assign mohr-coulomb range group 'layer'
zone property young 5000. poisson 0.3 range group 'layer'
zone property cohesion 0 friction 25 tension 0 range group 'layer'
; --- settings ---
model gravity 10
model save 'initial2'
; --- solution ---
model factor-of-safety bracket 0.9 1.0 ratio-local 1e-4 filename 'HBSlope2'
FLAC3D 6.0
Automatic Calculation of a Stable Pit Slope Angle 133
Automatic Calculation of a Stable Pit Slope Angle
Problem Statement
A pit slope is excavated in strong rock that contains an 8 m thick vertical band of
weak rock. The slope is composed of two 20 m high benches with a 10 m wide
berm. Both benches are initially inclined at 60° to the horizontal. Figure 1 shows
the initial slope configuration and the location of the weak rock within the strong
rock.
The strengths of the strong rock and weak rock are prescribed by a Mohr-
Coulomb material model. The strong rock has a cohesion of 15 kPa and a friction
angle of 40°, and the weak rock has a cohesion of 7 kPa and friction angle of
120°. Both rock types are assigned a zero tensile strength, a mass density of 2000
kg/m3, a bulk modulus of 10 GPa, and shear modulus of 3 GPa.
Figure 1: Benched slope in rock with a weak vertical band.
FLAC3D 6.0
A factor of safety calculation performed for these slope conditions indicates that
the slope is unstable. A factor of 0.81 is calculated, and the failure (as shown in
Figure 2) is primarily through the weak rock in the upper bench.
Figure 2: Factor of safety and failure surface (shown by shear strain contours)
for slope with benches inclined at 60°.
The purpose of this exercise is to determine the slope angle for the upper and
lower benches that produces a stable slope condition with a factor of safety of 1.2.
This analysis is performed automatically by adjusting the slope angle in
increments until a factor of 1.2 is reached. The model is created using the
Extrusion pane in FLAC3D, and a FISH function is used to adjust the slope angle
automatically in order to calculate a safety factor for each slope angle. A series of
calculations is performed until the factor reaches 1.2.
Figure 3 shows the slope boundary that is sketched first in the construction view
of the Extrusion pane. The model width at the base is 100 m, and the model height
is 60 m. Additional edges are added to the sketch to divide the slope shape into
quadrilateral blocks in order to create a structured grid for this model. See Figure
4.
FLAC3D 6.0
The zoning is defined in the construction view by specifying a model extent of 30

zones in the Autozone dialog. The blocks in the construction view are all given the
group name “strongrock”.
Selected gridpoints are also assigned group names. These gridpoints can then be
moved to correspond to the change in slope angle that is performed in the FISH
function. The gridpoint at the front of the berm above the lower bench is
assigned the group name “pt1”. The gridpoint at the back of the berm and the
two gridpoints below this gridpoint are assigned the group name “pt2”. (This
assignment is shown in Figure 5.) The gridpoint at the top of the upper bench is
assigned group name “pt3”.
Figure 3: Benched slope boundary sketch in Extrusion pane.
FLAC3D 6.0
Figure 4: Additional edges to divide the slope shape into blocks.
Figure 5: Assignment of group name “pt2” to selected gridpoints in the

construction view.
FLAC3D 6.0
The model depth is prescribed in the extrusion view to be 40 m. The extent of the
8 m wide weak rock is also set, and the zones within this region are given the
group name “weakrock”. See Figure 6.
Figure 6: Assignment of group name “weakrock” to zones in the extrusion view.
The extruded model is shown in Figure 7. The mesh in this model is relatively
coarse for this illustrative example. The zoning can easily be changed by
returning to the Extrusion pane. The input record for this extrusion model is
saved from the State Record pane in FLAC3D. The record is named
“benchextude.f3dat”.
FLAC3D 6.0
Figure 7: FLAC3D model of a benched slope in rock with a weak vertical band.
A FISH function named benchAdjust in file “benchAdjust.f3dat” is used to move

the gridpoints with group names “pt1”, “pt2”, and “pt3” as the slope angle of
the benches is reduced from 60°. At each angle reduction, the extrude points in
the model are scanned and the points belonging to those groups have their
x-positions adjusted.
The factor of safety is calculated for each change in slope angle. A new grid is
created each time the angle is reduced (using the zone generate from-extruder
command), material properties and boundary conditions are assigned, and a
factor of safety calculation is performed. The commands to create the model for
each angle are contained in “benchFOS.f3dat”.
A factor of safety calculation cannot be initiated from inside a FISH command/

end_command statement, because the command restores an initial model state
and that state includes FISH. To work around this restriction, FISH is used to
create a data file that calculates the factor of safety for each slope angle
separately and stores the result in a table.
FLAC3D 6.0
A total of 11 factor of safety calculations are performed as the slope angle is

reduced from 60° to 35°. A table plot of factor of safety values versus slope angle
can be displayed during the calculations to monitor the effect of slope angle on
safety factor. A plot of the table at the completion of the calculations is shown in
Figure 8. From the table, the critical slope (the slope when the factor of safety is
1.0) is approximately 43.5°.
Figure 8: Factor of safety versus slope angle.
The factor of safety is 1.23 at the final inclination of 35°. The failure surface is
also confined to the upper bench for this slope geometry, as shown in Figure 9.
The table plot in Figure 8 indicates that a factor of 1.2 is reached at a slope angle
of 36°.
The FISH function in “benchAdjust.f3dat” could easily be extended to allow for

the adjustment of the upper- and lower-bench slope angles independently and
the adjustment of the bench width. Additional runs can then be made to evaluate
the effect of variation in these variables on safety factor. (Note that the model
boundary is not changed in this illustrative exercise, which limits the range of
reduction of the slope angle.)
FLAC3D 6.0
Figure 9: Factor of safety and failure surface for slope with benches inclined at
35°.
Data Files
enchExtrusion.f3dat
model new
; Created in 2D extruder and exported from state pane
call 'benchAdjust' suppress ; FISH function to modify slope
; This FISH function creates a data file to calculate the FOS
; for a range of angles
fish define createDataFile(startang,endang,increment)
local text = array.create(6)
file.open('benchrun.f3dat',1,1) ; Open text file for writing
loop local angle (startang,endang,increment) ; Global so you can see where
; you are in the symbol set
text(1) = '; Angle '+string(angle)
text(2) = '@benchAdjust('+string(angle)+')'
text(3) = 'call \'benchFOS\' suppress'
text(4) = 'model factor-of-safety ratio-local 1e-3 ... '
text(5) = 'filename \'BenchSlope'+string(angle)+'\' ; '+string(angle)
text(6) = 'table \'fos\' insert ('+string(angle)+',[global.fos])'
file.write(text,6)
endloop
FLAC3D 6.0
file.close();
array.delete(text)
end
@createDataFile(35,60,2.5)
call 'benchrun'
table 'fos' export 'fos-angle' truncate
system 'del benchrun.f3dat'
benchAdjust.f3dat
; translate gridpoints as a function of slope angle change

fish define benchAdjust(theta)
; Calculate new point position
local xa = 10 + 20 / math.tan(theta*math.degrad)
local xb = xa + 10
local xc = xb + 20 / math.tan(theta*math.degrad)
; Update x positions of points marked pt1, pt2, and pt3
local set = extrude.set.find('bench')
loop foreach local pnt extrude.point.list(set)
local v = extrude.point.pos(pnt)
if extrude.point.isgroup(pnt,'pt1') then
extrude.point.pos(pnt) = vector(xa,v->y)
endif
extrude.point.pos(pnt) = vector(xb,v->y)
endif
extrude.point.pos(pnt) = vector(xc,v->y)
endif
endloop
end
benchFOS.f3dat
;-------------------------------------------------------------
; Runs a single FOS calculation based on current extruder data
;-------------------------------------------------------------
; Clear away all old zones
zone delete
; Generate zones from extruder data
zone generate from-extruder
; Assign constitutive model and properties
zone property density 2000 bulk 1e10 shear 3e9
zone property cohesion 15000 friction 40
zone property cohesion 7000 friction 20 range group 'weakrock'
; Assign boundary conditions
zone face apply velocity-normal 0 range union position-x 0 position-x 100
zone face apply velocity-normal 0 range union position-y 0 position-y 40
zone face apply velocity (0,0,0) range position-z 0
model gravity 10
FLAC3D 6.0
FLAC3D 6.0
Factor of Safety Contours 143
Factor of Safety Contours
Problem Statement
Typically, application of the strength reduction method produces one single
factor of safety per simulation, corresponding to one global minimum stability
state. However, the ability to calculate multiple minimum states may be of
interest, for example, along a complex slope profile such as a benched cut or a
slope with a berm (e.g., see Cheng et al. 2007). A “safety map” may be
constructed through a series of analyses using the limit equilibrium method to
identify multiple possible failure surfaces for slopes of this type (Baker and
Leshchinsky 2001).
A simple procedure to determine multiple local stability states with the strength
reduction method is to exclude different regions of the slope when performing
the strength reduction calculation.
Alternatively, the explicit dynamic solution method employed in FLAC3D allows

multiple local stability surfaces to be identified in one FLAC3D simulation. When
using the model factor-of-safety command, model instability is detected by
monitoring the unbalanced force ratio throughout the model. This provides a
minimum global factor of safety for the model. In an alternative technique,
presented here, material strengths are reduced in increments by a strength
reduction factor. Unstable states for the model are identified at the global
minimum state and then beyond that state. Unstable states of the model are
identified at each stage as an assembly of gridpoints with velocities above a
specified average value. The current strength reduction factor is assigned to
unstable gridpoints for later contouring.
If the strength is reduced in small intervals, progressively more regions of

gridpoints can be identified as unstable. By monitoring the velocities, it is
possible to delineate the regions of unstable gridpoints by different strength
factors and produce a plot of factor of safety contours. This plot can be used to
locate multiple possible failure surfaces and is comparable to the safety map
developed using the limit equilibrium method.
FLAC3D 6.0
This technique is demonstrated for a slope profile consisting of two double-

inclination slopes separated by a horizontal berm. This example is taken from
Cheng et al. (2007), who produced a set of local minimum stability states for this
slope using the Morgenstern-Price limit equilibrium method. The slope
configuration and resulting local minima locations are shown in Figure 1.
FLAC3D Model
The slope geometry is drawn in the Extrusion pane, and the extruded-grid input
record is saved as data file “geometry.f3dat”. The slope is oriented in the x-z
plane and is one zone thick in the y-direction. The mesh density is specified with
the extrude set automatic-zone command.
A FLAC3D simulation is first run to determine the global minimum factor of

safety for this slope using the model factor-of-safety command. The result,
shown by the shear strain contour plot in Figure 2, is a global minimum factor of
safety of 1.32 with a multiple failure surface that corresponds to the two surfaces
with the smallest factor of safety values shown in the lower slope in Figure 1.
Figure 1: Local minima surfaces from limit equilibrium solution for slope with
beam (from Cheng et al. 2007).
FLAC3D 6.0
Figure 2: Global minimum factor of safety for slope with berm.
The safety factor from the model factor-of-safety command identifies the
starting factor to develop a set of factor of safety contours for this model. Factor
of safety contours are calculated for this slope model by using the FISH function
defined in “FOSContours.f3dat”. This function reduces the Mohr-Coulomb
strengths, cohesion, and friction incrementally. A stable or new unstable state is
determined at each strength-reduction increment, and if the state is unstable,
the portion of the model that is failing is identified by evaluating gridpoint
velocities. New gridpoints with velocities greater than a specified value are
identified as failing and are assigned the current strength-reduction increment
factor, which is stored in the gridpoint in extra variable 1. After the simulation is
complete for the selected range of strength-reduction increments, a contour plot
of gridpoint extra 1 values is produced. This is a factor of safety contour plot.
The values input as arguments for the FOSContour function are the starting value
for strength reduction, fsmin, the limiting velocity threshold for a gridpoint at
failure, vel_limit, the reduction factor increment, inc_fs, the cycle limit for each
stage, max_num_cyc, the number of stages to calculate, total_stages, and a flag
indicating whether tension should be included in the strength reduction,
inc_tension. For this example, fsmin = 1.1, inc_fs = 0.05, max_num_cyc = 30000,
FLAC3D 6.0
vel_limit = 10-4, total_stages = 10, and inc_tension = false. The maximum cycle
limit and velocity limit are problem dependent; their values may be selected after
trial runs with strength properties reduced to determine the velocity magnitude
level at which failing gridpoints are well defined.
The factor of safety contour plot produced for this example is shown in Figure 3.
The contours compare quite well with the local minima surfaces plot in Figure 1.
Note that the global minimum contour line (at a factor of 1.3) in Figure 3 closely
matches the smallest local minimum surface in Figure 1. The next contour lines,
at factors of 1.35 and 1.4 below and above the berm, also compare well with the
failure surfaces identified in Figure 3. The factor of safety contour plot also
shows a contour shape (see, for example, the 1.45 contour in Figure 3) that
curves upward beneath the berm. Note that this effect on the shape of the failure
surface is not seen with the limit equilibrium method; compare to the 1.42
surface in Figure 1.
This exercise demonstrates that the strength reduction method can be applied to
produce multiple potential failure surfaces in one simulation by monitoring
failure in terms of the development of unstable regions (defined by high
gridpoint velocities) as the strength of the material is incrementally reduced.
Figure 3: Factor of safety contours for slope with berm.
FLAC3D 6.0
References
Baker, R., and D. Leshchinsky. “Spatial distribution of safety factors,” J. Geotech.
Geoenviron. Eng., 127(2), 135-45 (2001).
Cheng, Y.M., T. Lansivaara and W.B. Wei. “Two-dimensional slope stability

Geotechnics, 34, 137-150, 2007.
Data Files
BermFOS.f3dat
model new
; Geometry created in 2D extruder, exported from state record.
extrude set automatic-zone direction construction size 160
zone generate from-extrude
; Assign constitutive model and properties
zone property bulk 6.25e6 shear 2.88462e6 density 2000
zone property cohesion 5000 friction 30 tension 1e100
zone face apply velocity-normal 0 range group 'East' or 'West'
model gravity 10
model solve elastic ratio-local 1e-4
; Solve for initial factor-of-safety
model factor-of-safety ratio-local 1e-3 filename 'Berm'
BermFOSContours.f3dat
model restore 'initial'

;
zone gridpoint initialize velocity (0,0,0)
call 'FOSContours' suppress
@FOSContours(1.1,1e-4,0.05,30000,10,[false])
model save 'FOSContours'
FOSContours.f3dat
; fsmin = Starting minimim safety factor
FLAC3D 6.0
; vel_limit = Threshold velocity below which a given gridpoint

; is considered stable
; inc_fs = Amount safety factor is increment each stage
; max_num_cyc = Number of cycles that is taken at each stage
; t_stages = Total stages taken - maximum safety factor is then
; fsmin + (total-stages-1)*in_fs
; inc_tens = If true, then tension is included in the
; general strength reduction
fish define FOSContours(fsmin,vel_limit,inc_fs,max_num_cyc,t_stages,inc_tens)
; Initialize storage in zones and gridpoints
loop foreach local gp gp.list
gp.extra(gp,1) = 0
end_loop
loop foreach local zone zone.list
zone.extra(zone,1) = zone.prop(zone,'cohesion')
zone.extra(zone,2) = zone.prop(zone,'friction')
zone.extra(zone,3) = zone.prop(zone,'tension')
endloop
; Loop through stages of target safety factor
local k = 0
local numb = 0
loop while k < t_stages
; Calculate safety factor for stage, and reduce cohesion, friction,
; and tension accordingly
local fs = fsmin + k*inc_fs
loop foreach zone zone.list
zone.prop(zone,'cohesion') = zone.extra(zone,1) / fs
zone.prop(zone,'friction') = ...
math.atan(math.tan(zone.extra(zone,2)*math.degrad)/fs)/math.degrad
if inc_tens then
zone.prop(zone,'tension') = zone.extra(zone,3)/fs
endif
endloop
;
local num_cyl_0 = global.step
command
model solve cycles @max_num_cyc ratio-local 1e-3
endcommand
loop foreach gp gp.list
if gp.extra(gp,1) = 0 then
local vel = math.mag(gp.vel(gp))
if vel > vel_limit then
gp.extra(gp,1) = fs - inc_fs
endif
endif
endloop
if (global.step-num_cyl_0) > max_num_cyc then
numb = numb + 1
endif
k = k + 1
endloop
loop foreach gp gp.list
if gp.extra(gp,1) = 0 then
FLAC3D 6.0
gp.extra(gp,1) = fs + inc_fs/2.0
endif
endloop
end
FLAC3D 6.0
FLAC3D 6.0
References 151
References
Abramson, L. W., et al. Slope Stability and Stabilization Methods, 2nd Ed. John
Wiley & Sons, Inc. New York, USA (2002).
Baker, R., and D. Leshchinsky. “Spatial distribution of safety factors,” J. Geotech.

Geoenviron. Eng., 127(2), 135-45 (2001).
Bishop, A. W. “The Use of the Slip Circle in the Stability Analysis of Earth
Slopes,” Géotechnique, 5, 7-17 (1955).
Chen, W.-F. Limit Analysis and Soil Plasticity, J. Ross Publishing (2007).
Cheng, Y.M., T. Lansivaara and W.B. Wei “Two-dimensional slope stability

Geotechnics, 34, 137-150, 2007.
Davis, R. O., and A. P. S. Selvadurai. Plasticity and Geomechanics, Cambridge

University Press (2002).
Dawson, E. M., and W. H. Roth. “Slope Stability Analysis with FLAC,” in FLAC and
Numerical Modeling in Geomechanics (Proceedings of the International FLAC
Symposium on Numerical Modeling in Geomechanics, Minneapolis, Minnesota,
September 1999), pp. 3-9. C. Detournay and R. Hart, eds. Rotterdam: A. A.
Balkema (1999).
Dawson, E. M., W. H. Roth and A. Drescher. “Slope Stability Analysis by Strength

Reduction,” Géotechnique, 49(6), 835-840 (1999).
Dawson, E., K. You and Y. Park. “Strength-Reduction Stability Analysis of Rock

Slopes Using the Hoek-Brown Failure Criterion,” in Trends in Rock Mechanics
(Proceedings of Sessions of Geo-Denver 2000, Denver, Colorado, August 2000)
Geotechnical Special Publication No. 102, pp. 65-77. J. F. Labuz, S. D. Glaser and
E. Dawson, eds. Reston, Virginia: ASCE (2000).
Donald, I. B., and S. K. Giam. “Application of the nodal displacement method to

slope stability analysis,” in Proceedings of the 5th Australia-New Zealand
Conference on Geomechanics (Sydney, Austrtalia, August 1988) pp. 456-460. Sydney:
Institution of Engineers (1988).
FLAC3D 6.0
Drescher, A., and E. Detournay. “Limit load in transitional failure mechanisms

for associative and non-associative materials,” Géotechnique, 43, 443-456 (1993).
Fellenius, W. “Calculation of the stability of earth dams,” Proceedings of the 2nd

Congress on Large Dams (Washington D. C.), Vol. 4. U. S. Government Printing
Office (1936).
Fu, W., and Y. Liao. “Non-linear shear strength reduction technique in slope
stability calculation,” Computers and Geotechnics, 37, 288-298, 2009.
Griffiths, D. V., and P. A. Lane. “Slope Stability Analysis by Finite Elements,”

Géotechnique, 49(3), 387-403 (1999).
Hammah, R. E., et al. “The shear strength reduction method for the generalized
Hoek-Brown criterion,” ARMA/USRMS 05-810, 2005.
Hoek, E., and J. Bray. Rock Slope Engineering. London: IMM (1981).
Janbu, N. “Slope stability computations,” in Soil Mech. and Found. Engrg. Rep., The
Technical University of Norway, Trondheim, Norway (1968).
Li, A. J., R. S. Merifield and A. V. Lyamin. “Stability charts for rock slopes based
on the Hoek-Brown failure criterion”, Int. J. Rock Mech. and Mining Sci., 45,
689-700 (2008).
Lowe, J., and L. Karafiath. “Stability of Earth Dams upon Drawdown,” in

Proceedings of the 1st Pan-Am. Conference on Soil Mechanics and Foundation
Engineering (Mexico City, Mexico), Vol. 2, pp. 537-552 (1960).
Matusi, T., and K. C. San. “Finite element slope stability analysis by shear
strength reduction technique,” Soils and Foundations, 32(1) 59-70 (1992).
Michalowski, R. “Stability Charts for Uniform Slopes,” J. Geotech. Geoenviron. Eng.,

128(4), 351-355 (April 2002).
Naylor, D. J. “Finite elements and slope stability,” in Numerical Methods in

Geomechanics (Proceedings of the NATO Advanced Study Institute, Lisbon, Portugal),
pp. 229-244. J. B. Martins, ed. D. Reidel Publishing Company (1982).
Shukha, R., and R. Baker. “Mesh geometry effects on slope stability calculation
by FLAC strength reduction method - linear and non-linear criteria,’’ in FLAC
and Numerical Modeling in Geomechanics - 2003 (Proceedings of the 3rd International
FLAC Symposium on Numerical Modeling in Geomechanics, Sudbury, Ontario, Canada,
October 2003), pp. 109-116. R. Brummer, et al., eds. Lisse: A. A. Balkema (2003).
FLAC3D 6.0
References 153
Spencer, E. “A method of analysis of the stability of embankments assuming

parallel interslice forces,” Géotechnique, 17(1), 11-26 (1967).
Taylor, D. W. “Stability of earth slopes,” J. Boston Soc. Civ. Eng., 24, 197-246
(1937).
Transportation Research Board. “Landslides: Investigation and Mitigation,” TRB

Special Report 247, National Academy Press, Washington D.C. (1996).
Ugai, K. “A method of calculation of total factor of safety of slopes by elasto-

plastic FEM,” Soils and Foundations, 29(2) 190-195, (in Japanese) 1989.
Ugai, K., and D. Leshchinsky. “Three-dimensional limit equilibrium and finite

element analyses: a comparison of results,” Soils and Foundations, 35(4) 1-7
(1995).
Zienkiewicz, O. C., C. Humpheson and R. W. Lewis. “Associated and non-

associated visco-plasticity and plasticity in soil mechanics,” Géotechnique, 25(4),
671-689 (1975).
FLAC3D 6.0
FLAC3D 6.0
Constitutive Models 155
Constitutive Models
Introduction
Numerical solution schemes face several difficulties when implementing
constitutive models to represent geomechanical material behavior. There are
three characteristics of geo-materials that cause particular problems.
One is physical instability. Physical instability can occur in a material if there is

the potential for softening behavior when the material fails. When physical
instability occurs, part of the material accelerates and stored energy is released
as kinetic energy. Numerical solution schemes often have difficulties at this stage
because the solution may fail to converge when a physical instability arises.
A second characteristic is the path dependency of nonlinear materials. In most

geomechanical systems, there are an infinite number of solutions that satisfy the
equilibrium, compatibility, and constitutive relations that describe the system. A
path must be specified for a “correct” solution to be found. For example, if an
excavation is made suddenly (e.g., by explosion), then the solution may be
influenced by inertial effects that introduce additional failure of the material.
This may not be seen if the excavation is made gradually. The numerical solution
scheme should be able to accommodate different loading paths in order to
properly apply the constitutive model.
A third characteristic is the nonlinearity of the stress-strain response. This

includes the nonlinear dependence of both the elastic stiffness and the strength
envelope on the confining stress. This can also include behavior after ultimate
failure that changes character according to the stress level (e.g., different post-
failure response in the tensile, unconfined, and confined regimes). The
numerical scheme needs to be able to accommodate these various forms of
nonlinearity.
The difficulties faced in numerical simulations in geomechanics—physical

instability, path dependence, and implementation of extremely nonlinear
constitutive models—can all be addressed by using the explicit, dynamic solution
scheme provided in FLAC3D. This scheme allows the numerical analysis to follow
the evolution of a geologic system in a realistic manner, without concerns about
FLAC3D 6.0
156 FLAC3D Theory and Background • Constitutive Models
numerical instability problems. In the explicit, dynamic solution scheme, the full
dynamic equations of motion are included in the formulation. By using this
approach, the numerical solution is stable even when the physical system being
modeled is unstable. With nonlinear materials, there is always the possibility of
physical instability (e.g., the sudden collapse of a slope). In real life, some of the
strain energy in the system is converted into kinetic energy, which then radiates
away from the source and dissipates. The explicit, dynamic solution approach
models this process directly, because inertial terms are included—kinetic energy
is generated and dissipated.
In contrast, schemes that do not include inertial terms must use some numerical
procedure to treat physical instabilities. Even if the procedure is successful at
preventing numerical instability, the path taken may not be a realistic one. The
numerical scheme should not be viewed as a black box that will give “the
solution.” The way the system evolves physically can affect the solution. The
explicit, dynamic solution scheme can follow the physical path. By including the
full law of motion, this scheme can evaluate the effect of the loading path on the
constitutive response.
The explicit, dynamic solution scheme also allows the implementation of

strongly nonlinear constitutive models because the general calculation sequence
allows the field quantities (forces/stresses and velocities/displacements) at each
element in the model to be physically isolated from one another during one
calculation step. The general calculation sequence of FLAC3D is described in the
section of Theoretical Background. The implementation of elastic/plastic
constitutive models within the framework of this scheme is discussed in the
Incremental Formulation section.
The mechanical constitutive models available in FLAC3D range from linearly

elastic models to highly nonlinear plastic models. The basic constitutive models
are listed below. A short discussion of the theoretical background and simple
example tests for each model follow this listing.
Section Outline
• Constitutive Models in FLAC3D
• Incremental Formulation
• Null Model Group
◦ Null Model
• Elastic Model Group
FLAC3D 6.0
Constitutive Models 157
◦ Elastic (Isotropic) Model

◦ Anisotropic (Transversely-Elastic) Model
◦ Orthotropic Elastic Model
• Plastic Model Group
◦ Drucker-Prager Model
◦ Mohr-Coulomb Model
◦ Ubiquitous-Joint Model
◦ Anisotropic-Elasticity Ubiquitous-Joint Model
◦ Strain-Softening/Hardening Mohr-Coulomb Model
◦ Bilinear Strain-Softening/Hardening Ubiquitous-Joint Model
◦ Double-Yield Model
◦ Modified Cam-Clay Model
◦ Hoek-Brown Model
◦ Hoek-Brown-PAC Model
◦ Cap-Yield (CYSoil) Model
◦ Simplified Cap-Yield (CHSoil) Model
◦ Plastic-Hardening Model
◦ Swell Model
◦ Mohr-Coulomb Tension Crack (MohrT) Model
• Model Tests and Examples
◦ Oedometer Test with Mohr-Coulomb Model
◦ Uniaxial Compressive Strength of a Jointed Material Sample
◦ Isotropic Consolidation Test with Double-Yield Model
◦ Isotropic Consolidation Test with Modified Cam-Clay Model
◦ Triaxial Compression Test with Hoek-Brown Model
◦ Triaxial Compression Test with Hoek-Brown-PAC Model
◦ Isotropic Compression Test with CYSoil Model
◦ Oedometer Test with CYSoil Model
◦ Drained Triaxial Test with CYSoil Model — Constant Dilation
◦ Drained Triaxial Test with CYSoil Model — Dilation Hardening
◦ Undrained Triaxial Test with CYSoil Model
◦ Drained Triaxial Compression Test with Simplified Cap-Yield
(CHSoil) Model
◦ Comparison between Mohr-Coulomb Model and Plastic-
Hardening model
◦ Isotropic Compression Test with Plastic-Hardening Model
◦ Drained Triaxial Compression Test with Plastic-Hardening Model
◦ Undrained Triaxial Compression Test with Plastic-Hardening
Model
FLAC3D 6.0
◦ Oedometer Test with Plastic-Hardening Model

◦ Single Zone Swell Test
◦ Single Zone Loading-Unloading Test with MohrT Model
• References
FLAC3D 6.0
Constitutive Models in FLAC3D 159
Constitutive Models in FLAC3D

The basic mechanical constitutive models provided in FLAC3D are arranged into
null, elastic, and plastic model groups:
Null model group
1. Null model
A null material model is used to represent material that is

removed or excavated.
Elastic model group
2. Elastic model
The elastic, isotropic model provides the simplest representation

of material behavior. This model is valid for homogeneous,
isotropic, continuous materials that exhibit linear stress-strain
behavior with no hysteresis on unloading.
3. Orthotropic model
The elastic, orthotropic model represents material with three

mutually perpendicular planes of elastic symmetry. For example,
this model may simulate columnar basalt loaded below its
strength limit.
4. Anisotropic model
The elastic, transversely isotropic model gives the ability to

simulate layered elastic media in which there are distinctly
different elastic moduli in directions normal and parallel to the
layers.
FLAC3D 6.0
Plastic model group
5. Drucker-Prager model
The Drucker-Prager plasticity model may be useful to model soft

clays with low friction angles. However, this model is not
generally recommended for application to geologic materials. It
is included here mainly to permit comparison with other
numerical program results.
6. Mohr-Coulomb model
The Mohr-Coulomb model is the conventional model used to

represent shear failure in soils and rocks. Vermeer and de Borst
(1984), for example, report laboratory test results for sand and
concrete that match well with the Mohr-Coulomb criterion.
7. Ubiquitous-Joint model
The ubiquitous-joint model is an anisotropic plasticity model

that includes weak planes of specific orientation embedded in a
Mohr-Coulomb solid.
7. Ubiquitous-Anisotropic
The ubiquitous-joint model is a model with a plane of weakness

and with anisotropic elasticity.
9. Strain-Softening/Hardening model
The strain-softening/hardening model allows representation of

nonlinear material softening and hardening behavior based on
prescribed variations of the Mohr-Coulomb model properties
(i.e., cohesion, friction, dilation and tensile strength) as
functions of the deviatoric plastic strain.
10. Bilinear-Softening/Hardening Ubiquitous-Joint model
The bilinear strain-softening/hardening ubiquitous-joint model

allows representation of material softening and hardening
behavior for the matrix and the weak plane based on prescribed
FLAC3D 6.0
variations of the ubiquitous-joint model properties (i.e.,

cohesion, friction, dilation, and tensile strength) as functions of
deviatoric and tensile plastic strain. The variation of material
strength properties with mean stress can also be taken into
account by using the bilinear option.
11. Double-Yield model
The double-yield model is intended to represent materials in

which there may be significant irreversible compaction in
addition to shear yielding, such as hydraulically placed backfill
or lightly cemented granular material.
12. Modified Cam-Clay model
The modified Cam-Clay model may be used to represent

materials when the influence of volume change on bulk property
and resistance to shear need to be taken into consideration, as in
the case of soft clay.
13. Hoek-Brown model
A Hoek-Brown model provides an alternative to the Hoek-

Brown-PAC model with a stress-dependent plastic flow rule,
described above. The model characterizes post-failure plastic
flow by simple flow rule choices given in terms of a user-
specified dilation angle. This model also contains a tensile
strength limit similar to that used by the Mohr-Coulomb model.
In addition, a factor-of-safety calculation based on the strength
reduction method can be run with the Hoek-Brown model.
14. Hoek-Brown-PAC model
The Hoek-Brown failure criterion characterizes the stress

conditions that lead to failure in intact rock and rock masses.
The failure surface is nonlinear and is based on the relation
between the major and minor principal stresses. The model
incorporates a plasticity flow rule that varies as a function of the
confining stress level.
FLAC3D 6.0
15. Cap-Yield (CYSoil) model
The cap-yield (CY) soil model provides a comprehensive

representation of the nonlinear behavior of soils. The model
includes frictional strain-hardening and softening shear
behavior, an elliptic volumetric cap with strain-hardening
behavior, and an elastic modulus function of plastic volumetric
strain. The model allows a more realistic representation of the
loading/unloading response of soils.
16. Simplified CYSoil (CHSoil) model
A simplified version of the CYSoil model, called the CHSoil

model, offers built-in features including a friction-hardening
law that uses hyperbolic model parameters as direct input, and a
Mohr-Coulomb failure envelope with two built-in dilation laws.
17. Plastic-Hardening model
An elasto-plasticity model with shear and volumetric hardening.
18. Swell model
A Mohr-Coulomb elasto-plasticity model considering wetting-

induced deformation.
19. Mohr-Coulomb-Tension (MohrT) model
A Mohr-Coulomb elasto-plasticity model with tensile strain

crack tracking.
Other mechanical material models are provided with the optional features in
FLAC3D. There are nine time-dependent creep constitutive models available in
the creep model option for FLAC3D. In addition, there are two pore-pressure
generation material models available in the dynamic analysis option. These
models are modified versions of the Mohr-Coulomb model that include cyclic
volume change to simulate effects of liquefaction. The models are described in
Dynamic Analysis. There is also a modified version of the Drucker-Prager model
that is provided to simulate the mechanical behavior associated with thermal
hydration, called Hydration-Ducker-Prager model.
FLAC3D 6.0
All of these models are provided as dynamic link libraries (DLLs) that are loaded
when FLAC3D is first executed, and their source code is included in the
“\pluginfiles\cmodels” sub-directory. Users can modify these models or create
their own constitutive models as DLLs by following the procedures given in
Writing New Constitutive Models.
Fluid and thermal constitutive models are also included in FLAC3D. Three fluid
constitutive models are available. These models are described in Fluid-
Mechanical Interaction. Four basic thermal constitutive models are provided.
These models are described in Thermal Analysis. A thermal hydration model is
also included.
The formulation of the basic mechanical models is addressed in general terms

first in the section of Incremental Formulation. Separate discussions of the
theoretical background and specific implementation for each model in FLAC3D
are then presented.
FLAC3D 6.0
FLAC3D 6.0
Incremental Formulation 165
Incremental Formulation
Incremental Formulation
All constitutive models in FLAC3D share the same incremental numerical
algorithm. Given the stress state at time , and the total strain increment for a
timestep, , the purpose is to determine the corresponding stress increment
and the new stress state at time . When plastic deformations are involved,
only the elastic part of the strain increment will contribute to the stress
increment. In this case, a correction must be made to the elastic stress increment
as computed from the total strain increment in order to obtain the actual stress
state for the new timestep.
Note that this incremental formulation results in effects that may be surprising
to those accustomed to more traditional implicit finite-element anlyses. First,
the displacement field is an output-only value and can be changed at any time
without affecting the stress or equilibrium state of the system. Second, changes
to material properties will also have no effect on the current stress state, they
will only change to result of future deformations that occur. This makes it easier
to model changes to a system over time.
Note also that all models in FLAC3D operate on effective stresses only; pore
pressures are used to convert total stresses to effective stresses before the
constitutive model is called. The reverse process occurs after the model
calculations are complete.
Definitions and Conventions

All stress increments in this section are corotational stress increments. For
simplicity of notation, however, the superscript will not be used. The stresses
at time are computed as “new stress values,” indicated by a superscript
in this section. However, in large-strain mode, those values must be
incremented by the stress-rotation correction (see this equation).
We define as a generalized stress vector of dimension with components ,

and as a generalized strain-increment vector with components ,
. The components of the generalized stress- and strain-increment vectors
FLAC3D 6.0
may consist of the six components of the stress- and strain-increment tensors
or other appropriately defined combinations of variables, giving a measure of
stress and strain increments in specific constitutive model contexts.
As a notation convention in this section, we use the subscript to refer to the

range of generalized components from = 1 to (e.g., is used to
represent ).
Incremental Equations of the Theory of Plastic Flow

The description of plastic flow in FLAC3D rests on several relations:
1. the failure criterion
(1)
where , the yield function, is a known function that specifies the limiting stress
combination for which plastic flow takes place. (This function is represented by a
surface in the generalized stress space, and all stress points below the surface are
characterized by elastic behavior.)
2. the relation expressing the decomposition of strain increments into the sum of
elastic and plastic parts
(2)
3. the elastic relations between elastic strain increments and stress increments
(3)
where is a linear function of the elastic strain increments,
4. the flow rule specifying the direction of the plastic-strain increment vector as
that normal to the potential surface = constant
(4)
where is a constant. (The flow rule is said to be associated if , and

nonassociated otherwise.)
5. the requirement for the new stress-vector components to satisfy the yield
function
FLAC3D 6.0
Incremental Formulation 167
(5)
This equation provides a relation for evaluation of the magnitude of the plastic-
strain increment vector.
Substitution of the expression for the elastic-strain increment derived from

Equation (2) into the elastic relation Equation (3) yields, taking into
consideration the linear property of the function ,
(6)
In further expressing the plastic strain increment by means of the flow rule,
Equation (4), this equation becomes
(7)
where use has been made of the linear property of .
In the special case where is a linear function of the components , = 1, ,

Equation (5) may be expressed as
(8)
where, as a notation convention, represents the function minus its constant

term,
(9)
For a stress point on the yield surface, = 0, and Equation (8) becomes,
after substitution of the expression Equation (7) for the stress increment, and
further using the linear property of ,
(10)
We now define new stress components and elastic guesses as
(11)
(12)
Note that the term in Equation (12) is the component of the stress
increment induced by the total-strain increment , in case no increment of
plastic deformation takes place. This justifies the name of “elastic guess” for .
FLAC3D 6.0
From the definition Equation (12), it follows, using the preceding arguments,
that
(13)
Hence, an expression for may be derived from Equation (9), Equation (10), and
Equation (13):
(14)
Using the expression of the stress increment, Equation (7), and the definition of
the elastic guess, Equation (12), the new stress may be expressed from Equation
(11) as
(15)
For clarity, recall that, in these last two expressions, is the stress
increment obtained from the incremental elastic law, where is substituted
for , .
Implementation
In FLAC3D, an elastic guess , for the stress state at time is first
evaluated by adding to the stress components at time, , increments computed
from the total-strain increment for the step, using an incremental elastic stress-
strain law (see Equation (12)). If the elastic guess violates the yield function,
Equation (15) is used to place the new stress exactly on the yield curve.
Otherwise, the elastic guess gives the new stress state at time .
If the stress point , is located above the yield surface in the generalized
stress space, the coefficient in Equation (15) is given by Equation (14), provided
the yield function is a linear function of the generalized stress vector
components. The equation Equation (15) is still valid, but is set to zero in case
, = 1, is located below the yield surface (elastic loading or unloading).
The implementation of each of the 15 basic constitutive models in FLAC3D is

described separately in the following sections. All models except the null and
elastic models potentially involve plastic deformations. Note that a wide range of
material behavior may be obtained from these models by assigning appropriate
values to the model parameters. Recommendations for applying these models are
given in Choice of Constitutive Model.
FLAC3D 6.0
Null Model Group 169
Null Model Group

• Null Model
FLAC3D 6.0
FLAC3D 6.0
Null Model 171
Null Model
A null material is used to represent material that is removed or excavated from
the model. The stresses within a null zone are automatically set to zero:
(1)
The null material may be changed to a different material model at a later stage of
the simulation. In this way, backfilling an excavation can be simulated, for
example.
FLAC3D 6.0
FLAC3D 6.0
Elastic Model Group 173
Elastic Model Group

The models in this group are characterized by reversible deformations upon
unloading; the stress-strain laws are linear and path-independent. The elastic
models include both isotropic and anisotropic elastic models. Two anisotropic
elastic models that correspond to two cases of elastic symmetry are available: the
orthotropic and transversely isotropic models. For further information on
anisotropic formulations, see Lekhnitskii (1981).
• Elastic (Isotropic) Model

• Anisotropic (Transversely-Elastic) Model
• Orthotropic Elastic Model
FLAC3D 6.0
FLAC3D 6.0
Elastic (Isotropic) Model 175
Elastic (Isotropic) Model

In this elastic, isotropic model, strain increments generate stress increments
according to the linear and reversible law of Hooke:
(1)
where the Einstein summation convention applies, is the Kroenecker delta

symbol, and is a material constant related to the bulk modulus, , and shear
modulus, , as
(2)
New stress values are then obtained from the relation
(3)
Bulk modulus, K, and shear modulus, G, are related to Young’s modulus, E, and
Poisson’s ratio, ν, by the following equations:
(4)
or
(5)
isotropic Model Properties
Use the following keywords with the zone property command to set these
properties of the elastic (isotropic) model.
bulk f
bulk modulus,
poisson f
Poisson’s ratio,
FLAC3D 6.0
shear f
shear modulus,
young f
Young’s modulus,
Notes:
Only one of the two options is required to define the elasticity: bulk
modulus and shear modulus , or, Young’s modulus and Poisson’s
ratio . When choosing the latter, Young’s modulus must be assigned
in advance of Poisson’s ratio .
FLAC3D 6.0
Anisotropic (Transversely-Elastic) Model 177
Anisotropic (Transversely-Elastic) Model

The transversely isotropic model takes a plane of isotropy into consideration. Let
the axis of rotational symmetry, normal to the plane of isotropy, correspond to
the local axis. This axis is a principal direction of elasticity. Also, any two
perpendicular directions , which are principal directions of elasticity, can be
selected in the isotropic plane. With this convention, the transversely isotropic
model may be considered as a particular case of the orthotropic model for which
(1)
(2)
(3)
(4)
If, for clarity, we write
Young’s moduli in the plane of isotropy

Young’s moduli in the direction normal to the plane of
isotropy
Poisson’s ratio characterizing lateral contraction in the plane
of isotropy when tension is applied in this plane
of isotropy when tension is applied in the direction normal to
it
shear modulus for the plane of isotropy
shear modulus for any plane normal to the plane of isotropy
The strain-stress relations in the local axes take the form
(5)
FLAC3D 6.0
The model involves the five independent elastic constants , and . The
shear modulus, , is calculated by the code from the relation (See
Equation (4)). In addition to those five properties, the user prescribes the
orientation of the isotropic plane by giving its dip and dip direction. Default
values for all properties are zero.
The cross shear modulus, , for anisotropic elasticity must be determined.

Lekhnitskii (1981) suggests the following equation, based on laboratory testing of
rock:
(6)
assuming the -plane is the plane of isotropy.
The FLAC3D implementation proceeds as described in the context of the

orthotropic model.
anisotropic Model Properties
properties of the anisotropic (transversely-elastic) model.
dip f
dip angle [degrees] of the plane of isotropy
dip-direction f
dip direction [degrees] of the plane of isotropy
normal v
normal direction of the plane of isotropy, ( )
normal-x f
x-component of unit normal to the plane of isotropy,
normal-y f
y-component of unit normal to the plane of isotropy,
normal-z f
z-component of unit normal to the plane of isotropy,
FLAC3D 6.0
Anisotropic (Transversely-Elastic) Model 179
poisson-normal f
Poisson’s ratio characterizing lateral contraction in
the plane of isotropy when tension is applied normal
to the plane, = =
poisson-plane f
Poisson’s ratio characterizing lateral contraction in
the plane of isotropy when tension is applied in the
plane, =
shear-normal f
shear modulus for any plane normal to the plane of
isotropy, = =
young-plane f
Young’s modulus in the plane of isotropy, = =
young-normal f
Young’s modulus normal to the plane of isotropy,
=
FLAC3D 6.0
FLAC3D 6.0
Orthotropic Elastic Model 181
Orthotropic Elastic Model

The orthotropic model accounts for three orthogonal planes of elastic symmetry.
Principal coordinate axes of elasticity, labeled 1’,2’,3’, are defined in the
directions normal to those planes.
The incremental strain-stress relations in the local axes have the form
(1)
The model involves nine independent elastic constants:
Young’s moduli in the directions of the local axes

shear moduli in planes parallel to the local coordinate planes
Poisson’s ratio where characterizes lateral contraction in
local direction caused by tensile stress in local direction
By virtue of the symmetry of the strain-stress matrix, we have
(2)
(3)
(4)
In addition to those nine properties, the user prescribes the orientation of the
local axes by giving the dip and dip direction of the (1’,2’) plane, and the rotation
angle between the 1’ axis and the dip-direction vector (defined in positive sense
from the dip direction vector). Default values for all properties are zero.
FLAC3D 6.0
In the FLAC3D implementation of this model, the local stiffness matrix is

found by inversion of the symmetric matrix in Equation (1). Using and
to represent the incremental stress and strain vectors present in the right and
left members of Equation (1):
(5)
In the global axes, the incremental stress-strain relations are
(6)
In FLAC3D, the global stiffness matrix is calculated by applying a

transformation of the form
(7)
where is a suitable 6 × 6 matrix involving direction cosines of local axes in

global axes (Q is derived from the relations , where is direction
cosine of local axis ).
In particular, if the local axes are obtained from the global axes by positive
rotation through an angle about the common 3 ≡ 3’ axis, we have
(8)
The matrix for rotation about the 1 ≡ 1’ or 2 ≡ 2’ axis may be obtained by cyclic
permutation of indices.
orthotropic Model Properties
properties of the orthotropic elastic model.
dip f
dip angle [degrees] of the plane defined by
axes 1’-2’
FLAC3D 6.0
Orthotropic Elastic Model 183
dip-direction f
dip direction [degrees] of the plane defined
by axes 1’-2’
normal v
normal direction of the planes of symmetry,
( )
normal-x f
x-component of unit normal to plane defined
by axes 2’-3’,
normal-y f
y-component of unit normal to plane defined
by axes 1’-3’,
normal-z f
z-component of unit normal to plane defined
by axes 1’-2’,
poisson-12 f
Poisson’s ratio characterizing lateral
contraction in direction 1’ when tension is
applied in direction 2’,
poisson-13 f
poisson-23 f
shear-12 f
shear modulus in planes parallel to axes
1’-2’,
shear-13 f
1’-3’,
FLAC3D 6.0
shear-23 f
2’-3’,
young-1 f
Young’s modulus in direction 1’,
young-2 f
young-3 f
FLAC3D 6.0
Plastic Model Group 185
Plastic Model Group

All plastic models potentially involve some degree of permanent, path-
dependent deformations (failure)—a consequence of the nonlinearity of the
stress-strain relations. The different models in FLAC3D are characterized by their
yield function, hardening/softening functions, and flow rule. The yield functions
for each model define the stress combination for which plastic flow takes place.
These functions or criteria are represented by one or more limiting surfaces in a
generalized stress space, with points below or on the surface being characterized
by an incremental elastic or plastic behavior, respectively. The plastic flow
formulation in FLAC3D rests on basic assumptions from plasticity theory that the
total strain increment may be decomposed into elastic and plastic parts, with
only the elastic part contributing to the stress increment by means of an elastic
law (see the discussion in Incremental Equations of the Theory of Plastic Flow).
In addition, both plastic and elastic strain increments are taken to be coaxial
with the current principal axes of the stresses (only valid if elastic strains are
small compared to plastic strains during plastic flow). The flow rule specifies the
direction of the plastic strain increment vector as that normal to the potential
surface; it is called associated if the potential and yield functions coincide, and
nonassociated otherwise. See Vermeer and de Borst (1984) for a more detailed
discussion on the theory of plasticity.
For the Drucker-Prager, Mohr-Coulomb, ubiquitous-joint, strain-hardening/

softening, and bilinear-hardening/softening-ubiquitous-joint models, a shear
yield function and a nonassociated shear flow rule are used. For the double-yield
and CYSoil models, shear and volumetric yield functions, nonassociated shear
flow and associated volumetric flow rules are included. The CHSoil model is a
simplified version of the CYSoil model that provides built-in friction hardening
and dilation hardening/softening laws, and does not include a volumetric cap. In
addition, the failure envelope for each of the above models is characterized by a
tensile yield function with associated flow rule.
The modified Cam-clay model formulation rests on a combined shear and

volumetric yield function and associated flow rule.
FLAC3D 6.0
The two types of Hoek-Brown model in FLAC3D provide different formulations to

represent yielding. The basic Hoek-Brown model using a nonlinear shear yield
function and a plasticity flow rule that varies as a function of the stress level is
referred to as the Hoek-Brown-PAC model. For the traditional Hoek-Brown
model, plastic flow is handled in a manner similar to that in the Mohr-Coulomb
model in which a dilation angle is specified. Also, a tensile yield function similar
to that used with the Mohr-Coulomb model is included with the Hoek-Brown
model.
Note also that all plasticity models are formulated in terms of effective stresses,
not total stresses.
The plasticity models can produce localization (i.e., the development of families
of discontinuities such as shear bands in a material that starts as a continuum).
Note that localization is grid-dependent since there is no intrinsic length scale
incorporated in the formulations. This is an important consideration when
creating a grid for a plasticity analysis, and is discussed more fully in
Localization, Physical Instability and Path-Dependence.
As discussed in Implementation, in the numerical implementation of the models,

an elastic trial (or “elastic guess”) for the stress increment is first computed from
the total strain increment using the incremental form of Hooke’s law. The
corresponding stresses are then evaluated. If they violate the yield criteria (i.e.,
the stress-point representation lies above the yield function in the generalized
stress space), plastic deformations take place. In this case, only the elastic part of
the strain increment can contribute to the stress increment; the latter is
corrected by using the plastic flow rule to ensure that the stresses lie on the
composite yield function. This section describes the yield and potential
functions, flow rules, and stress corrections for the different plasticity models in
FLAC3D.
• Drucker-Prager Model
• Mohr-Coulomb Model
• Ubiquitous-Joint Model
• Anisotropic-Elasticity Ubiquitous-Joint Model
• Strain-Softening/Hardening Mohr-Coulomb Model
• Bilinear Strain-Softening/Hardening Ubiquitous-Joint Model
• Double-Yield Model
• Modified Cam-Clay Model
• Hoek-Brown Model
• Hoek-Brown-PAC Model
FLAC3D 6.0
Plastic Model Group 187
• Cap-Yield (CYSoil) Model

• Simplified Cap-Yield (CHSoil) Model
• Plastic-Hardening Model
• Swell Model
• Mohr-Coulomb Tension Crack (MohrT) Model
FLAC3D 6.0
FLAC3D 6.0
Drucker-Prager Model 189
Drucker-Prager Model
The failure envelope for this model involves a Drucker-Prager criterion with
tension cutoff. The position of a stress point on this envelope is controlled by a
nonassociated flow rule for shear failure, and an associated rule for tension
failure.
Generalized Stress and Strain Components
The generalized stress vector involved in the definition of the Drucker-Prager

model has two components ( = 2): the tangential stress, , and mean normal
stress, , defined as
(1)
(2)
where the Einstein summation convention applies, and is the deviatoric-stress

tensor. The components of the associated generalized strain increment vector
are the shear-strain increment, , and volumetric-strain increment, ,
introduced as
(3)
(4)
where is the incremental deviatoric-strain tensor.
Incremental Elastic Law
The incremental expression of Hooke’s law, in terms of the generalized stress

and stress increments, has the form
(5)
(6)
where and are the bulk and shear modulus, respectively.
For future reference, comparing those expressions on in Incremental Formula:
FLAC3D 6.0
(7)
(8)
Composite Failure Criterion and Flow Rule
The failure criterion used for this FLAC3D model is a composite Drucker-Prager
criterion with tension cutoff as sketched in the representation of Figure 1.
The failure envelope = 0 is defined, from point to on the figure, by the
Drucker-Prager failure criterion = 0, with
(9)
and, from to , by the tension failure criterion = 0, with
(10)
where , , and are positive material constants, and is the tensile strength
for the Drucker-Prager model. Note that, for a material whose property is not
equal to zero, the maximum value of the tensile strength is given by
(11)
Figure 1: FLAC3D Drucker-Prager failure criterion.
FLAC3D 6.0
The potential function is composed of two functions, and ,

used to describe shear and tensile plastic flow, respectively. The function
corresponds in general to a nonassociated law, and has the form
(12)
where is a constant, equal to if the flow rule is associated. The function

corresponds to an associated flow rule, and is given by
(13)
Figure 2: Drucker-Prager model—domains used in the definition of the flow

rule.
The flow rule is given a unique definition by application of the following

technique. The diagonal line between the representation of = 0 and = 0 in
the plane (see Figure 2), divides the domain of the elastic guess violating
the composite yield function into two, domain 1 (shear failure) and domain 2
(tension failure), respectively (see Figure 2). If the stress point falls within
domain 1, shear failure is declared, and the stress point is placed on the curve
FLAC3D 6.0
= 0 using a flow rule derived using the potential function . If the point falls
within domain 2, tensile failure takes place, and the new stress point satisfies
= 0 using a flow rule derived using .
Plastic Corrections
First, considering shear failure, partial differentiation of Equation (12) yields
(14)
(15)
Substitution of and for and for , respectively, in Equation

(7) and Equation (8) gives
(16)
(17)
Using (see Equation (9)) then become
(18)
(19)
and
(20)
The new stress tensor components may be expressed in terms of the generalized
stresses using a rescaling technique. Formally, we may write Equation (18) as
(21)
where is a known proportionality factor ( ). From the definition

Equation (1) of in terms of the tensor components , we then have
(22)
Eliminating from the last two expressions, we obtain
FLAC3D 6.0
(23)
Finally, new stress components may be calculated from Equation (19) and
Equation (24) using the definition
(24)
We now consider tensile failure. Partial differentiation of Equation (13) gives
(25)
(26)
Using Equation (7), we obtain
(27)
(28)
With , as given by Equation (10):
(29)
(30)
and
(31)
As expected, after substitution of Equation (31) into Equation (30), we obtain
(32)
(33)
In this mode of failure, , and we may write, from the definition of

deviatoric stresses and Equation (33),
(34)
from which it follows that
(35)
FLAC3D 6.0
Apex Stress Correction
It is possible that is greater than the apex stress ( ) at an extreme

circumstance when is not zero (usually after shear plastic corrections). Once
this occurs, the stress will be forced to the apex, or and .
Implementation Procedure
In the implementation of the Drucker-Prager model in FLAC3D, an elastic guess,

, is first computed by adding to the old stress components, increments
calculated by application of Hooke’s law to the total strain increments, (see
Elastic Model). Deviatoric stresses, , are evaluated from , and the elastic
guess is derived from and using Equations (1) and (2).
If this guess violates the composite yield criterion (see Equations (9) and (10)),
then either in domain 1 or in domain 2 (see Figure 2). In the first case, shear
failure takes place, and and are evaluated from Equations (18) and (19),
using Equation (20). New deviatoric stress components are derived using
Equation (23) and, finally, new stress components are calculated from Equation
(24). In the second case, tensile failure occurs, and new stress components are
evaluated from Equation (35).
If the point is located below the composite failure envelope, no plastic

flow takes place for this step, and the new stresses are given by .
Note that the tensile strength default value is zero for a material with = 0, and
is otherwise (see Equation (11)). This last value will be substituted for if
the assigned value for the tensile strength exceeds . There is no tensile
softening in this model.
Notes on Parameters
The Drucker-Prager model in FLAC3D is described by means of four parameters:

, and . The particular case = 0, = 0 gives the von Mises criterion.
By appropriate adjustment of the parameters, the criterion may also be fitted to
approximate the geometry of the Mohr-Coulomb or Tresca criterion in the ( )-
plane. (These last criteria are addressed in the next section.)
The Drucker-Prager shear criterion = 0 is represented in the principal stress

space by a cone with axis along and apex at
, with (see Figure 3). The Mohr-Coulomb criterion,
FLAC3D 6.0
characterized by two parameters (cohesion, , and friction angle, ), is

represented there by an irregular hexagonal pyramid with the same axis and
three “outer” and three “inner” edges (see Figure 4).
The parameters and can be adjusted so that the Drucker-Prager cone will
pass through either the outer or the inner edges of the Mohr-Coulomb pyramid.
For the outer adjustment, we have
(36)
(37)
For the inner adjustment, we have
(38)
(39)
For the average adjustment (of outer and inner adjustments), we have
(40)
(41)
For a compromised adjustment so that the Drucker-Prager circle area is identical

to the area of the Mohr-Coulomb irregular hexagon in the plane, we have
(42)
(43)
FLAC3D 6.0
In the special case = 0, the Drucker-Prager criterion degenerates into the von
Mises criterion, which corresponds to a cylinder in the principal stress space. The
Tresca criterion is a special case of the Mohr-Coulomb criterion for which = 0.
It is represented in the principal stress space by a regular hexagonal prism. The
von Mises cylinder circumscribes the prism for
(44)
(45)
Figure 3: Drucker-Prager and von Mises yield surfaces in principal stress space.
FLAC3D 6.0
Figure 4: Mohr-Coulomb and Tresca yield surfaces in principal stress space.
drucker-prager Model Properties
properties of the Drucker-Prager model.
bulk f
bulk modulus,
cohesion-drucker f
material parameter,
dilation-drucker f
material parameter, . The default
is 0.0.
friction-drucker f
material parameter,
poisson f
Poisson’s ratio,
shear f
shear modulus,
FLAC3D 6.0
tension f
tension limit, . The default is 0.0.
young f
Young’s modulus,
Notes:
• Only one of the two options is required to define the
elasticity: bulk modulus and shear modulus , or
Young’s modulus and Poisson’s ratio . When choosing
the latter, Young’s modulus must be assigned in
advance of Poisson’s ratio .
• The tension cut-off is .
FLAC3D 6.0
Mohr-Coulomb Model 199
Mohr-Coulomb Model
The failure envelope for this model corresponds to a Mohr-Coulomb criterion
(shear yield function) with tension cutoff (tension yield function). The position
of a stress point on this envelope is controlled by a nonassociated flow rule for
shear failure, and an associated rule for tension failure.
Generalized Stress and Strain Components
The Mohr-Coulomb criterion in FLAC3D is expressed in terms of the principal

stresses
, and , which are the three components of the generalized stress vector for
this model ( = 3). The components of the corresponding generalized strain
vector are the principal strains , and .
The incremental expression of Hooke’s law in terms of the generalized stress and
stress increments has the form
(1)
where and are material constants defined in terms of the shear modulus, ,
and bulk modulus, , as
(2)
(3)
FLAC3D 6.0
The failure criterion used in the FLAC3D model is a composite Mohr-Coulomb

criterion with tension cutoff. In labeling the three principal stresses so that
(4)
this criterion may be represented in the plane as illustrated in Figure 1.

(Recall that compressive stresses are negative.) The failure envelope =0
is defined from point to by the Mohr-Coulomb failure criterion = 0 with
(5)
and from to by a tension failure criterion of the form = 0 with
(6)
where is the friction angle, is the cohesion, is the tensile strength, and
(7)
Note that the tensile strength of the material cannot exceed the value of
corresponding to the intersection point of the straight lines = 0 and in
the plane. This maximum value is given by
(8)
FLAC3D 6.0
Figure 1: FLAC3D Mohr-Coulomb failure criterion.
The potential function is described by means of two functions, and , used to

define shear plastic flow and tensile plastic flow, respectively. The function
corresponds to a nonassociated law and has the form
(9)
where is the dilation angle and
(10)
The function corresponds to an associated flow rule and is written
(11)
FLAC3D 6.0
The flow rule is given a unique definition by application of the following

technique. A line represented by the diagonal between the representation of =
0 and = 0 in the -plane (see Figure 2) divides the domain where an
elastic guess violating the composite yield function into two domain: domain 1
and domain 2 (see Figure 2).
If the stress point falls within domain 1, shear failure is declared and the stress
point is placed on the curve = 0 using a flow rule derived using the potential
function . If the point falls within domain 2, tensile failure takes place, and the
new stress point conforms to = 0 using a flow rule derived using .
Note that by ordering the stresses as in Equation (4), the case of a shear-shear
edge is automatically handled by a variation on this technique. The technique,
applicable for small-strain increments, is simple to implement: at each step,
only one flow rule and corresponding stress correction is involved in case of
plastic flow. In particular, when a stress point follows an edge, it receives stress
corrections alternating between two criteria. In this process, the two yield
criteria are fulfilled to an accuracy which depends on the magnitude of the strain
increment. Results obtained for the oedometric test are presented as a validation
of this approach.
FLAC3D 6.0
Figure 2: Mohr-Coulomb model—domains used in the definition of the flow

rule.
Shear Plastic Corrections
First, considering shear failure, partial differentiation of Equation (9) yields
(12)
Substitution of , , and for , , and , respectively,

in Equation (3) gives
FLAC3D 6.0
(13)
Using (see Equation (5))
(14)
and
(15)
Tension Plastic Corrections
First, we assume the case that there is only one principal stress in tension
failure. Partial differentiation of Equation (9) gives
(16)
(17)
With , as given by Equation (6)
FLAC3D 6.0
(18)
and
(19)
Substitution of Equation (15) for in Equation (18) gives
(20)
After evaluation of , , and , the stress-tensor components are evaluated

in the system of reference axes, assuming that the principal directions have not
been affected by the occurrence of a plastic correction.
Second, we assume the case that there are two principal stresses in tension
failure, or
(21)
Partial differentiation of Equation (21) gives
(22)
so
(23)
FLAC3D 6.0
From Equation (23) (and remember and after plastic

correction) we get
(24)
Substitution of Equation (24) in Equation (23) gives
(25)
Finally, for the case that all three principal stresses are in tension failure, or
(26)
similarly,
(27)
so
(28)
From Equation (28), we get
(29)
Remember , and after plastic correction, so
(30)
FLAC3D 6.0
It is possible that one or more corrected principal stresses are greater than the
apex stress ( ) at an extreme circumstance when the friction angle is not
zero (usually after shear plastic corrections). Once this occurs, all principal
stresses will be forced to the apex stress.
Implementation Procedures
In the implementation of the Mohr-Coulomb model in FLAC3D, an elastic guess (

) is first computed by adding to the stress components, increments calculated
by application of Hooke’s law to the total strain increments (see Elastic
Model). Principal stresses , , and corresponding directions are then
calculated.
If the stresses , , violate the composite yield criterion (see Equation (5)
and (6)), then it is either in domain 1 or domain 2. In the first case, shear failure
takes place, and , , and are evaluated from Equation (14), using Equation
(15). In the second case, tensile failure occurs, and new principal stress
components are evaluated from from Equation (18) through Equation (30).
If the point is located below the representation of the composite failure

envelope in the plane , no plastic flow takes place for this step, and the
new principal stresses are given by , = 1,3.
The stress tensor components in the system of reference axes are then calculated
from the principal values by assuming that the principal directions have not been
affected by the occurrence of a plastic correction.
In FLAC3D, the default value for the tensile strength, , is zero. This value is set
to (see Equation (8)) if the value assigned to the tensile strength exceeds
. By default, if tensile failure occurs in a zone, the value assigned for the
tensile strength in this zone remains constant. Alternatively, if the property
keyword flag-brittle is set true, then tensile strength is set to zero for that zone
when tensile failure occurs. This simulates instantaneous tensile softening.
FLAC3D 6.0
mohr-coulomb Model Properties
properties of the Mohr-Coulomb model.
bulk f
elastic bulk modulus,
cohesion f
cohesion,
dilation f
dilation angle, . The default is 0.0.
friction f
internal angle of friction,
poisson f
Poisson’s ratio,
shear f
elastic shear modulus,
tension f
young f
Young’s modulus,
flag-brittle b [a]
If true, the tension limit is set to 0
in the event of tensile failure. The
default is false.
FLAC3D 6.0
Notes:
Footnotes
[a] Advanced property. It should have a default value and should not be
necessary to specify for simpler applications of the model.
FLAC3D 6.0
FLAC3D 6.0
Ubiquitous-Joint Model 211
Ubiquitous-Joint Model
This model accounts for the presence of an orientation of weakness (weak plane)
in a FLAC3D Mohr-Coulomb model. The criterion for failure on the plane, whose
orientation is given, consists of a composite Mohr-Coulomb envelope with
tension cutoff. The position of a stress point on the latter envelope is controlled
again by a nonassociated flow rule for shear failure and an associated rule for
tension failure.
In this numerical model, general failure is first detected and relevant plastic
corrections are applied, as indicated in the FLAC3D Mohr-Coulomb model
description. The new stresses are then analyzed for failure on the weak plane and
updated accordingly. The FLAC3D Mohr-Coulomb model was addressed earlier;
developments related to plastic flow on the weak plane will be discussed in this
section.
Definitions
The weak-plane orientation is given by the Cartesian components of a unit

normal to the plane in the global -, -, -axes. A local system of reference axes
is defined with and in the plane and pointing in the direction of the unit
normal . (The -axis points downward in the dip direction, and the -axis is
horizontal. The local system is right-handed.)
Using a matrix notation, we define as the rotation tensor whose three

columns are the direction cosines of , , and . The stress components in the
local axes are computed using the transformation
(1)
In turn, the global stress components may be obtained from the local
components using the reverse transformation,
(2)
The magnitude of the tangential traction component on the weak plane, referred
to as , is defined as
(3)
FLAC3D 6.0
The strain variable associated with is and has the form
(4)
In what follows, we use to refer to the stress components obtained at the end
of the current timestep, after application of the plastic corrections related to
general failure only (as opposed to failure on the weak plane).
Weak-Plane Generalized Stress and Strain Components
The generalized stress vector used to describe weak-plane failure has four
components ( = 4): , , , and . The components of the corresponding
generalized strain vector are , , , and .
The incremental expression of Hooke’s law in terms of the generalized stress and
strain increments has the form (see Elastic Model)
(5)
where and are material constants defined in terms of the shear modulus, ,
and bulk modulus, , as
(6)
(7)
FLAC3D 6.0
Note that the elastic strain increments in the preceding formula may be
expressed as differences between total and plastic strain increments. Taking into
consideration that the plastic increments may contain a known contribution
associated with general failure (as opposed to weak plane failure), the derivation
leading to the expressions for the new stresses may be followed (see Incremental
Formulation), provided we interpret as the generalized stress component, ,
obtained at the end of the current timestep, after application of the plastic
corrections relating to general failure only.
The weak-plane failure criterion used in the FLAC3D model is a composite Mohr-
Coulomb criterion with tension cutoff expressed in terms of ( , ), as
illustrated in Figure 1. (Recall that compressive stresses are negative.) The failure
envelope is defined from point to by the Mohr-Coulomb failure
criterion = 0, with
(8)
and from to by a tension failure criterion of the form = 0, with
(9)
where , , and are the friction, cohesion, and tensile strength of the weak
plane, respectively. Note that for a weak plane with nonzero friction angle, the
maximum value of the tensile strength is given by
(10)
FLAC3D 6.0
Figure 1: FLAC3D weak-plane failure criterion.
The potential function is composed of two functions ( and ) used to define

shear and tensile plastic flow, respectively. The function corresponds to a
nonassociated law and has the form
(11)
where is the weak-plane dilation angle. The function corresponds to an

associated flow rule and is written as
(12)
An elastic guess violating the composite yield function is represented by a point

in the -plane located either in domain 1 or 2, partitioned by the diagonal
line between the lines defined by and (see Figure 2). If in domain 1,
shear failure is declared and the stress point is placed on the curve = 0 using a
flow rule derived using the potential function . If in domain 2, tensile failure
takes place and the stress point conforms to = 0 using a flow rule derived
using .
FLAC3D 6.0
Figure 2: Ubiquitous-joint model—domains used in the definition of the weak-

plane flow rule.
Plastic Corrections
First, considering shear failure (domain 1), partial differentiation of Equation (8)
yields
(13)
Substitution of , , , and for , , ,

and , respectively, in Equation (7), gives
FLAC3D 6.0
(14)
Using (see Equation (8)):
(15)
(16)
and
(17)
The new shear stress components on the weak plane may be derived from and
, using the relations
(18)
The stress corrections for shear failure may thus be expressed as (see Equation
(15) and Equation (18))
(19)
We now consider tensile failure (domain 2). Partial differentiation of Equation

(9) gives
FLAC3D 6.0
(20)
(21)
With as given by Equation (9):
(22)
and
(23)
Finally, the stress corrections for tensile failure may be expressed, after
substitution of Equation (23) for in Equation (22), in the form
(24)
FLAC3D 6.0
Large-Strain Update to Orientation
In large strain, the orientation of the weak plane is adjusted, per zone, to account
for rigid-body rotations and rotations due to deformations. The corrections
applied to the global components of the unit normal to the weak plane are
computed, at each step, as averages over all tetra involved in the zone.
First taking into account rotations due to deformations, the corrections in local
axes to the weak-plane normal may be expressed as
(25)
The corresponding global components are obtained using the

transformation
(26)
where is the rotation tensor introduced above.
In turn, rigid-body rotations are taken into consideration by application of the

rotation tensor to , where is the identity matrix, and is
defined in Rate of Strain and Rate of Rotation (see ). In further neglecting
second-order terms, the total corrections to be added to to obtain the new
value have the form
(27)
It is possible that is greater than the apex stress ( ) at an extreme

circumstance when is not zero (usually after shear plastic corrections). Once
this occurs, the stress will be forced to the apex, or and .
In the implementation of the ubiquitous-joint model in FLAC3D, stresses

corresponding to the elastic guess for the step are first analyzed for general
failure, and relevant plastic corrections are made, as described in the FLAC3D
Mohr-Coulomb model. The resulting stress components, labeled as , are then
examined for failure on the weak plane.
FLAC3D 6.0
Local stress components and are first evaluated using the

transformation Equation (1) and the expression Equation (3). If the stresses
and violate the composite yield criterion (see Equation (8) and Equation (9)),
then either in domain 1 (shear failure) or domain 2 (tension failure). In the first
case, shear failure takes place, and the stress corrections , , ,
, and are evaluated from Equation (14), using Equation (17). In the
second case, tensile failure occurs and the stress corrections , , and
are evaluated from Equation (24), using Equation (23). The tensor of stress
corrections is expressed in the system of reference axes using the transformation
of Equation (2) and added to to yield new stress values.
If the point is located below the representation of the composite failure

envelope in the plane , no plastic flow takes place for this step, and the
new stresses are given by .
In large-strain mode, the unit normal to the weak plane is adjusted per zone to
account for body rotations (see Equation (25) and Equation (26)).
Note that the default value for the weak-plane tensile strength is zero if = 0,
and otherwise (see Equation (10)). This last value is substituted for if the
value assigned for the weak plane tensile strength exceeds .
ubiquitous-joint Model Properties
properties of the ubiquitous-joint model.
bulk f
cohesion f
cohesion,
dilation f
dip f
dip angle [degrees] of weakness
plane
FLAC3D 6.0
dip-direction f
dip direction [degrees] of weakness
plane
friction f
poisson f
Poisson’s ratio,
shear f
tension f
young f
Young’s modulus,
joint-cohesion f
joint cohesion,
joint-dilation f
joint dilation angle, . The default is
0.0.
joint-friction f
joint friction angle,
joint-tension f
joint tension limit, . The default is
0.0.
normal v
normal direction of the weakness
plane, ( , , )
normal-x f
x-component of the normal direction
to the weakness plane,
normal-y f
y-component of the normal
direction to the weakness plane,
FLAC3D 6.0
normal-z f
z-component of the normal
flag-brittle b [a]
If true, the tension limit is set to 0 in
the event of tensile failure. The
default is false.
Notes:
elasticity: bulk modulus and shear modulus , or,
• Only one of the three options is required to define the
orientation of the weakness plane: dip and dip-direction;
a norm vector ( ); or, three norm components: ,
, and .
• The joint tension limit used in the model is the minimum
of the input and .
Footnotes
FLAC3D 6.0
FLAC3D 6.0
Anisotropic-Elasticity Ubiquitous-Joint Model 223
Anisotropic-Elasticity Ubiquitous-Joint Model

This model combines the logic of the elastic, transversely isotropic (anisotropic)
model with that of the ubiquitous-joint model (note that “weak plane” and
“joint” are used interchangeably in the text). The new model has a single
orientation of weakness, which matches the orientation of the plane of elastic
isotropy. The criterion for failure on the plane, whose orientation is given,
consists of a local Mohr-Coulomb yield criterion with tension cutoff. The model
can be used to simulate the behavior of layered material, such as clay shales, and
account for slip conditions in the direction of layering (weak plane direction).
In the FLAC3D implementation, new stresses are evaluated for each step using
the elastic-anisotropic incremental laws. The new stresses are analyzed for
yielding on the weak plane and updated accordingly using the joint flow rule. The
local shear flow rule is non-associated, and the local tension flow rule is
associated.
Consider a local system of reference axes, with and in the “weak” plane, and
the axis normal to it. This axis is a principal direction of elasticity. Also, any
two perpendicular directions and , which are principal directions of elasticity,
can be selected in the isotropic plane. The incremental elastic strain-stress
relations used in the local axes are as follows:
(1)
where
FLAC3D 6.0
(2)
and
Young’s moduli in the plane of isotropy

Young’s moduli in the direction normal to the plane of isotropy
of isotropy when tension is applied in this plane
of isotropy when tension is applied in the direction normal to it
shear modulus for the plane of isotropy
shear modulus for any plane normal to the plane of isotropy,
and .
The conversion of stress and strain tensors from local to global axes and vice-
versa is obtained by the application of the usual matrix rotation operations.
Let be defined as the magnitude of the tangential traction component on the

weak plane:
(3)
Also, is the strain variable associated with :
(4)
The local incremental stress-strain equations (Equation (1)) expressed using the
above generalized stress and strain components are:
(5)
FLAC3D 6.0
Yield and Potential Functions, Plastic Corrections, and Large-Strain Update to

Orientation
The yield and potential functions, plastic corrections, and large-strain update to
orientation are similar to those joint-related sections described in the
ubiquitous-joint model.
The coefficients of the global elasticity matrix are computed and stored in the
initialization phase. New stresses are estimated using the elasticity matrix and
the total strain increments for the step. The global stress tensor then is resolved
in the local axes of the weak plane, and the local yield conditions are tested. If
yielding is detected, a relevant local stress correction (derived using the flow
rule) is calculated as described above. The stress correction then is resolved into
global axes and added to the elastic stress estimate. Finally, in large strain,
adjustment of the weak plane orientation is performed to account for rigid body
rotations.
ubiquitous-anisotropic Model Properties
properties of the anisotropic-elasticity ubiquitous-joint model.
dip f
plane
dip-direction f
plane
joint-cohesion f
joint cohesion,
joint-dilation f
joint dilation angle, . The default is
0.0.
joint-friction f
FLAC3D 6.0
joint-tension f
0.0.
normal v
plane, ( , , )
normal-x f
normal-y f
normal-z f
poisson-normal f
contraction in the plane of isotropy
when tension is applied normal to
the plane, = =
poisson-plane f
contraction in the plane of isotropy
when tension is applied in the plane,
=
shear-normal f
shear modulus for any plane normal
to the plane of isotropy, = =
young-plane f
Young’s modulus in the plane of
isotropy, = =
young-normal f
Young’s modulus normal to the
plane of isotropy, =
FLAC3D 6.0
Notes:
a norm vector ( ); or three norm components: ,
, and .
of the input and .
FLAC3D 6.0
FLAC3D 6.0
Strain-Softening/Hardening Mohr-Coulomb Model 229
Strain-Softening/Hardening Mohr-Coulomb
Model
This model is based on the FLAC3D Mohr-Coulomb model with nonassociated
shear and associated tension flow rules, as described earlier. The difference,
however, lies in the possibility that the cohesion, friction, dilation, and tensile
strength may soften or harden after the onset of plastic yield. In the Mohr-
Coulomb model, those properties are assumed to remain constant. Here, the user
can define the cohesion, friction and dilation as piecewise-linear functions of a
hardening parameter measuring the plastic shear strain. A piecewise-linear
softening law for the tensile strength can also be prescribed in terms of another
hardening parameter measuring the plastic tensile strain. The code measures the
total plastic shear and tensile strains by incrementing the hardening parameters
at each timestep, and causes the model properties to conform to the user-defined
functions.
The yield and potential functions, plastic flow rules, and stress corrections are
identical to those of the Mohr-Coulomb model.
Hardening Parameters
The two hardening parameters for this model ( and ) are defined as the sum
of some incremental measures of plastic shear and tensile strain for the zone,
respectively. The zone-shear and tensile-hardening increments are calculated as
the volumetric average of hardening increments over all tetrahedra involved in
the zone.
The shear-hardening increment for a particular tetrahedron is a measure of the

second invariant of the plastic shear-strain increment tensor for the step, given
as
(1)
where is the volumetric plastic shear strain increment,
(2)
FLAC3D 6.0
The tetrahedron tensile-hardening increment is the magnitude of the plastic

tensile-strain increment,
(3)
Plastic-Strain Increments
The plastic-strain increments involved in the preceding formula may be derived

from the definition of the flow rule. They have the form
(4)
(5)
(6)
where and are given by the Mohr-Coulomb model.
User-Defined Functions for Cohesion, Friction, Dilation, and Tensile Strength
User-defined functions for zone yielding parameters can be determined by back-

analysis of the post-failure behavior of a material. Consider a one-dimensional
stress-strain curve, versus , which softens upon yield and attains some
residual strength (Figure 1).
FLAC3D 6.0
Figure 1: Example stress-strain curve.
The curve is linear to the point of yield; in that range, the strain is elastic only:
. After yield, the total strain is composed of elastic and plastic parts:
. In the softening/hardening model, the user defines the cohesion,
friction, dilation, and tensile strength variance as a function of the plastic
portion, , of the total strain. These functions, which could in reality be sketched
as indicated in Figure 2, are approximated in FLAC3D as sets of linear segments
(Figure 3).
FLAC3D 6.0
Figure 2: Variation of cohesion (a) and friction angle (b) with plastic strain.
Figure 3: Approximation by linear segments.
Hardening and softening behaviors for the cohesion, friction, and dilation in
terms of the shear parameter (see Equation (1)) are provided by the user in
the form of tables. Each table contains pairs of values: one for the parameter, and
one for the corresponding property value. It is assumed that the property varies
linearly between two consecutive parameter entries in the table. Softening of the
tensile strength is described in a similar manner, using the parameter (see
Equation (2)).
The implementation of the strain-softening/hardening model in FLAC3D

proceeds as indicated in the Mohr-Coulomb model description. After
determination of the new stresses for the step, the hardening parameters for the
FLAC3D 6.0
zone are updated following the procedure described above. If appropriate, these
parameters are then used to determine new values for the zone cohesion,
friction, dilation, and tensile strength.
Note that the tensile strength of the material can only decrease. Also, for
material with friction, the value cannot exceed the maximum value .
strain-softening model properties
properties of the strain-softening/hardening Mohr-Coulomb model.
bulk f
cohesion f
cohesion,
dilation f.
friction f
poisson f
Poisson’s ratio,
shear f
tension f
young f
Young’s modulus,
flag-brittle b [a]
default is false.
FLAC3D 6.0
table-cohesion i [a]
number of table relating cohesion to
plastic shear strain
table_dilation i [a]
number of table relating dilation
angle to plastic shear strain
table-friction i [a]
number of table relating friction
angle to plastic shear strain
table-tension i [a]
number of table relating tension
limit to plastic tensile strain
strain-shear-plastic f [r]
accumulated plastic shear strain
strain-tension-plastic f [r]
accumulated plastic tensile strain
Notes:
• The tension table and flag-brittle should not be active
at the same time.
Footnotes
[r] Read-only property. It should not allowed to be set by the user. Instead, it
can be listed, plotted, or accessed through FISH.
FLAC3D 6.0
Bilinear Strain-Softening/Hardening Ubiquitous-Joint Model 235
Bilinear Strain-Softening/Hardening
Ubiquitous-Joint Model
The bilinear strain-softening/hardening ubiquitous-joint model is a
generalization of the ubiquitous-joint model. In the bilinear model, the failure
envelopes for the matrix and joint are the composite of two Mohr-Coulomb
model criteria with a tension cutoff that can harden or soften according to
specified laws. A nonassociated flow rule is used for shear-plastic flow and an
associated flow rule is used for tensile-plastic flow.
The softening behaviors for the matrix and the joint are specified in tables in
terms of four independent hardening parameters (two for the matrix and two for
the joint) that measure the amount of plastic shear and tensile strain,
respectively. In this numerical model, general failure is first detected for the
step, and relevant plastic corrections are applied. The new stresses are then
analyzed for failure on the weak plane and updated accordingly. The hardening
parameters are incremented if plastic flow has taken place, and the parameters
of cohesion, friction, dilation, and tensile strength are adjusted for the matrix
and the joint using the tables.
Failure Criterion and Flow Rule for the Matrix
The criterion for failure in the matrix used in this model is sketched in the
principal stress plane in Figure 1. (Recall that compressive stresses are
negative and, by convention, .)
The failure envelope is defined by two Mohr-Coulomb failure criteria: = 0 and

= 0 for segments and , and a tension failure criterion = 0 for
segment .
The shear failure criterion has the general form = 0. The criterion is
characterized by a cohesion, , and a friction angle, , for segment , and
by a cohesion, , and a friction angle, , for segment . The tensile failure
criterion is specified by means of the tensile strength, (positive value); thus we
have
(1)
FLAC3D 6.0
(2)
where
(3)
Note that the tensile cap acts on segment of the shear envelope and, for a
material with nonzero friction angle , the maximum value of the tensile
strength is given by
(4)
Figure 1: FLAC3D bilinear matrix failure criterion.
In the model formulation, elastic guesses for the stresses are first evaluated for
the step using total strain increments. Plastic yielding is detected if the
corresponding stress point lies outside the failure surface representation
FLAC3D 6.0
in Figure 1. In this case, a stress correction must be applied to the elastic guess. It
is determined by allowing plastic flow to occur, in order to restore the condition
= 0, = 0 or = 0, depending on the position of the stress point above ,
, or . (Bisectors are used at and to delimit the domain of failure
attached to a particular segment of the yield surface.)
The usual assumption is made that total strain increments can be decomposed
into elastic and plastic parts. The flow rule for plastic yielding has the form
(5)
where = 1,3. The potential function, , for shear yielding is . This function
corresponds to the nonassociated law,
(6)
where , the dilation angle, is equal to for failure along , and along
, and
(7)
The potential function for tensile yielding is . It corresponds to the associated

flow rule,
(8)
The plastic strain increments for shear failure have the form
(9)
The stress corrections for shear failure are
(10)
where
FLAC3D 6.0
(11)
and, by definition:
(12)
In turn, the plastic strain increments for tensile failure have the form
(13)
The stress corrections for tensile failure are
(14)
where
(15)
Failure Criterion and Flow Rule for the Weak Plane
The stresses, corrected for plastic flow in the matrix, are resolved into
components parallel and perpendicular to the weak plane and tested for
ubiquitous-joint failure. The failure criterion is expressed in terms of the
magnitude of the tangential traction component, , and the
normal traction component, , on the weak plane.
The failure criterion is represented in Figure 1 and corresponds to two Mohr-

Coulomb failure criteria ( = 0 for segment , and = 0 for segment
) and a tension failure criterion ( = 0 for segment ). Each shear criterion
has the general form = 0 and is characterized by a cohesion and a friction
angle , equal to along segment , and along . The
tensile criterion is specified by means of the tensile strength, (positive value).
Thus, we have
FLAC3D 6.0
(16)
(17)
Note that for a weak plane with nonzero friction angle , the maximum value of
the tensile strength is given by
(18)
Figure 2: FLAC3D bilinear joint failure criterion.
Yield is detected and stress corrections are applied using a technique similar to
the one described in the matrix context.
Here, the flow rule for plastic yielding has the form
(19)
where is the strain variable associated with , and we have
FLAC3D 6.0
(20)
The potential function, , for shear yielding on the weak plane is . It

corresponds to the nonassociated law,
(21)
where , the dilation angle, is equal to for failure along , and along
.
The potential function for tensile yielding on the weak plane is . It

corresponds to the associated flow rule,
(22)
The local plastic strain increments for shear failure are such that
(23)
The stress corrections for shear failure are
(24)
where
(25)
and the superscript indicates values obtained just before detection of failure on
the weak plane.
The plastic corrections for the local shear stress components on the weak plane
are derived by scaling:
(26)
In turn, local plastic strain increments for tensile failure have the form
FLAC3D 6.0
(27)
The stress corrections for tensile failure are
(28)
where
(29)
Large-Strain Update of Orientation
In large strain, the orientation of the weak plane is adjusted per zone to account
for rigid-body rotations, and rotations due to deformations. The large-strain
update to orientation is identical to those described in the ubiquitous-joint
model.
Hardening Parameters
In the bilinear strain-softening/hardening ubiquitous-joint model, some or all of

the zone yielding parameters (i.e., cohesion, friction, dilation and tensile
strength) for the matrix and joint are modified automatically, after the onset of
plasticity, according to piecewise linear laws specified on input in terms of a
range of values for the hardening parameters. (See the section of user-defined
functions in the strain-softening/hardening model.) One table number must be
specified in the command of zone property for each softening parameter. (If no
table property number is specified, the parameter is taken as constant.) The
corresponding table data contain pairs of values for the parameter and the
property between which a linear variation is assumed. The last property value is
used for values of the hardening parameter beyond the last one specified in the
table.
FLAC3D 6.0
Four independent hardening parameters are used in this model:
1. measures the matrix plastic shear strain and is used to update

the matrix cohesion, friction, and dilation;
2. measures the matrix plastic volumetric tensile strain and is

used to update the matrix tensile strength;
3. estimates the joint plastic shear strain and controls the joint
cohesion, friction, and dilation update; and
4. evaluates the joint plastic volumetric tensile strain and

controls the joint tensile strength update.
The parameters are defined as the sum of incremental measures of plastic strain
for the zone. The zone-hardening increments are calculated as the volumetric
average of hardening increments over all tetrahedra involved in the zone.
The shear-hardening increment for a tetrahedron is the square root of the

second invariant of the incremental plastic shear-strain deviator tensor for the
step. For the matrix, it is given as
(30)
where is the volumetric plastic shear strain increment,
(31)
and the plastic strain increments are given by Equation (9), using Equation (11)
for .
For the joint, the formula is
(32)
where the plastic strain increments are given by Equation (23) (see Equation
(20)), using the form (25) for .
The tetrahedron tensile-hardening increment is the plastic volumetric tensile-

strain increment.
For the matrix, we have
FLAC3D 6.0
(33)
where the plastic strain increment is given by Equation (13), using Equation (15)
for .
For the joint, the expression is
(34)
where the plastic strain increment is given by Equation (27), using the Equation
(29) for .
The implementation of the bilinear model in FLAC3D proceeds as indicated above.

An elastic guess, , is first computed using stress increments for the step
evaluated by application of Hooke’s law to the total strain increments, .
Principal stresses are calculated, ordered such that , and tested for
failure in the matrix using the yield criteria defined by Equation (1) and (2). In
principle, matrix failure is declared if the representation of the stress point
falls outside the yield surface in Figure 1. In this case, stress corrections
are applied to the principal values of the elastic guess, which depend on the
position of the stress point above , or . (Bisectors are used at
and to delimit the domain of failure attached to a particular segment of the
yield surface.)
The stress corrections for shear failure in the matrix are given by Equations (10)
and (11), where the parameters of cohesion, , friction, , and dilation, , have
value for failure along , and for failure along .
The stress corrections for tensile failure in the matrix are given by by Equations
(14) and (15).
The stress tensor components in the system of reference axes, , are then
calculated from the corrected principal values by assuming that the principal
directions have not changed during plastic flow.
Local traction components on the weak plane are defined as and , with
being the normal component, and being the magnitude of the
tangential traction component. (The local axis 1สน is in the dip direction, 2สน the
strike, and 3สน the normal to the weak plane.) These stresses are resolved from
and examined for ubiquitous-joint failure using the yield criteria (Equations
(16) and (17)). In principle, ubiquitous-joint failure is declared if the
FLAC3D 6.0
representation of the stress point falls outside the yield surface in

Figure 2. In this case, local stress corrections, which depend on the position of
the stress point in the vicinity of , or , are applied. (Bisectors
are used at and to delimit the domain of failure attached to a particular
segment of the yield surface.)
The stress corrections for shear joint failure are given by Equations (24) to (26),
where the parameters of cohesion, , friction, , and dilation, , have values
for failure along , and for failure along .
The stress corrections for tensile joint failure are given by Equations (28) and
(29).
Finally, the local stress components are resolved back into global axes.
In large-strain mode, the unit normal to the weak plane is adjusted per zone to
account for body rotations.
After determination of the new stresses for the step, the hardening parameters
are incremented using Equations (30), (32), (33), and (34). These parameters are
then used to determine new values of cohesion, friction, dilation and tensile
strength for the matrix and the joint from the available input tables.
It is assumed that the tensile strength of the material can never increase. Also,
for material with friction, the value will not exceed the maximum value or
(see Equations (4) and (18)).
Note that, by default, the yield model is linear in both the matrix and the joint, in
which case only sections 1 (where = 0) and 3 (where = 0) of the yield curve
are recognized (even if properties are assigned for section 2, where = 0). To
activate the bilinear laws, the property flag-bilinear and/or the property flag-
bilinear-joint must be set to 1.
Also, if the friction angles for sections 1 and 2 become equal, the model will be
considered as linear, and section 2 will be ignored (for the matrix and/or the
joint, as appropriate). Section 2 will also be ignored if the intersection of sections
1 and 2 corresponds to a stress point that violates the tensile criterion.
FLAC3D 6.0
softening-ubiquitous model properties
properties of the bilinear strain-softening/hardening ubiquitous-joint model.
bulk f
cohesion f
matrix cohesion, =
cohesion-2 f
matrix cohesion,
dilation f
matrix dilation angle, = . The
default is 0.0.
dilation-2 f
matrix dilation angle, . The default
is 0.0.
dip f
plane
dip-direction f
plane
friction f
matrix friction angle, =
friction-2 f
matrix friction angle,
poisson f
Poisson’s ratio,
shear f
tension f
FLAC3D 6.0
young f
Young’s modulus,
joint-cohesion f
joint cohesion, =
joint-cohesion-2 f
joint cohesion,
joint-dilation f
joint dilation angle, = . The
default is 0.0.
joint-dilation-2 f
joint dilation angle, . The default
is 0.0.
joint-friction f
joint friction angle, =
joint-friction-2 f
joint-tension f.
0.0.
normal v
plane, ( , , )
normal-x f
normal-y f
normal-z f
flag-bilinear i [a]
= 0 (default) for matrix linear model
;
FLAC3D 6.0
= 1 for matrix bilinear model.
flag-bilinear-joint i [a]
= 0 (default) for joint linear model ;
= 1 for joint bilinear model ;
< 0 joint effect will be skipped, so

that the model is degenerated into a
bilinear Mohr-Coulomb model.
flag-brittle b [a]
default is false.
number of table relating matrix
cohesion = to matrix plastic
shear strain. The default is 0.
table-cohesion-2 i [a]
cohesion to matrix plastic shear
strain. The default is 0.
table-dilation i [a]
dilation angle = to matrix
plastic shear strain. The default is 0.
table-dilation-2 i [a]
dilation to matrix plastic shear
friction = angle to matrix plastic
FLAC3D 6.0
table-friction-2 i [a]
friction angle to matrix plastic
table-joint-cohesion i [a]
number of table relating joint
cohesion = to joint plastic shear
table-joint-cohesion-2 i [a]
cohesion to joint plastic shear
table-joint-dilation i [a]
dilation = to joint plastic shear
table-joint-dilation-2 i [a]
dilation to joint plastic shear
table-joint-friction i [a]
friction angle = to joint plastic
table-joint-friction-2 i [a]
friction angle to joint plastic
table-joint-tension i [a]
tension limit to joint plastic
tensile strain. The default is 0.
FLAC3D 6.0
strain-shear-plastic-joint f [r]
accumulated joint plastic shear
strain
strain-tensile-plastic f [r]
strain-tensile-plastic-joint f [r]
accumulated joint plastic tensile
strain
Notes:
a norm vector ( ); or, three norm components: ,
, and .
of the input and .
at the same time.
Footnotes
[r] Read-only property. It should not be allowed to be set by the user. Instead, it
FLAC3D 6.0
FLAC3D 6.0
Double-Yield Model 251
Double-Yield Model
Permanent volume changes caused by the application of isotropic pressure are
taken into account in this model by including, in addition to the shear and tensile
failure envelopes in the FLAC3D strain-softening/hardening model, a volumetric
yield surface (or “cap”). For simplicity, the cap surface, defined by the “cap
pressure” 0, is independent of shear stress; it consists of a vertical line on a
plot of shear stress versus mean stress. The hardening behavior of the cap
pressure is activated by volumetric plastic strain, and follows a piecewise-linear
law prescribed in a user-supplied table. The tangential bulk and shear moduli
evolve as plastic volumetric strain takes place according to a special law defined
in terms of a factor, , assumed to be constant, and defined as the ratio of elastic
bulk modulus to plastic bulk modulus.
Only two additional material parameters and a table are required, in addition to
those associated with the strain-softening model:
1. the initial value of , which corresponds to the maximum mean

pressure that the material has experienced in the past;
2. the value of , greater than unity, which controls the slope of the
stress-strain curve on volumetric unloading (the “swelling” line,
in soil mechanics terms); and
3. the table representation of the “hardening curve,” which relates

cap pressure, , to plastic volume strain, .
Hence, any laboratory-determined hardening behavior may be modeled within

the constraints imposed by a two-parameter model.
In the FLAC3D implementation of this model, principal stresses , , are

used. The principal stresses and principal directions are evaluated from the
stress tensor components, and ordered so that (recall that compressive stresses
are negative)
(1)
FLAC3D 6.0
The corresponding principal strain increments are decomposed:
(2)
where the superscripts and refer to elastic and plastic parts, respectively, and
the plastic components are nonzero only during plastic flow. (Note that
extensional strains are positive.) It is assumed that the plastic contributions of
shear, tensile, and volumetric yielding are additive, so we may write
(3)
where the superscripts , , and stand for plastic shear, plastic tensile, and
plastic volumetric strain. By convention, in this section, the symbol is used to
refer to the minus volumetric strain increment with plastic part
and elastic part . The symbol refers to minus the value of the plastic
volumetric strain .
The incremental expression of Hooke’s law in terms of principal stress and strain
has the form
(4)
where , , and and are the current tangential

bulk and shear moduli defined according to the following considerations.
Consider an isotropic compression test with increasing pressure, . As the

material becomes more compact, its plastic stiffness ( ) usually increases.
It seems reasonable that the elastic stiffness will also increase, since the grains
are being forced closer together. A simple rule is adopted in this model whereby,
under general loading conditions, the incremental elastic stiffness, , is a
constant factor, , multiplied by the current incremental plastic stiffness. The
values of bulk and shear modulus, and , supplied by the user are taken as
upper limits to and , and it is assumed that the ratio remains
constant and equal to . Using incremental notation, this law is defined by the
relations
(5)
(6)
FLAC3D 6.0
where the factor is given, and is the current slope of the table of
values.
The type of behavior exhibited by the double-yield model is illustrated in Figure

1, which shows a nonlinear volumetric loading curve with several unloading
excursions; these excursions are elastic, with slope related by to the plastic
stiffness at the point of unloading.
Figure 1: Elastic volumetric loading/unloading paths.
Yield and Potential Functions
The shear and tensile yield functions, referred to as and , have the form
(7)
FLAC3D 6.0
(8)
where
(9)
and is the friction angle, is the cohesion, and is the tensile strength.
The volumetric yield function is defined as
(10)
where is the cap pressure.
The shear potential function, , corresponds to a nonassociated flow rule; the

tensile and volumetric potential functions, and , correspond to associated
laws. They have the form
(11)
(12)
(13)
where
(14)
and is the dilation angle.
Hardening/Softening Parameters
The shear and volume yield surfaces can harden (or soften), and the tensile yield
surface can soften, according to hardening rules that are specified by user-
defined tables. Entry to the tables is by hardening parameters that record some
measure of accumulated plastic strain. In shear and tension, the hardening
parameter incremental forms are
(15)
(16)
where
FLAC3D 6.0
(17)
= 1,3, and are plastic shear and tensile strain increments in the
principal directions.
In the volumetric direction, the hardening parameter increment is
(18)
where = 1,3 are plastic volumetric strain increments in the principal

directions.
These hardening parameters are used in the tables to determine new values of
friction, cohesion, dilation, tensile strength, and cap pressure. The current bulk
and shear moduli are also calculated from the table values as per Equations (5)
and (6).
Plastic Corrections
Let the superscript be used to represent the elastic guess obtained by adding to
the old stresses , elastic increments computed using the total strain
increments. In principal axes, we then have
(19)
In the FLAC3D implementation, shear yield is detected if 0,

volumetric yield if 0, and tensile yield if 0. Corresponding
plastic corrections are evaluated using the following techniques.
We first consider the case where tensile failure is not detected for the step, but
both shear and volumetric yield conditions are exceeded. Using Equations (2) and
(3), the principal strain increments may be expressed as
(20)
The flow rules for shear and volumetric yielding are
(21)
FLAC3D 6.0
(22)
where = 1,3. Using Equations (11) to (13), these expressions become, after
differentiation:
(23)
and
(24)
Substituting in Equation (15), we obtain
(25)
With these expressions for the elastic strain increments, Hooke’s incremental
equations yield (see Equation (4))
(26)
where , = 1,3, are the initial trial stresses in Equation (19), and ,
= 1,3 are the new principal stresses for the step.
To determine the multipliers and , we require that if shear and volumetric

yielding occur, the new stresses lie on both yield surfaces and we must have
= 0 and =0. Substituting Equation (25) for = 1,3 in
Equations (7) and (10), and solving for , we obtain
(27)
FLAC3D 6.0
Hence,
(28)
In these equations, the notation stands for the function evaluated for the
initial trial stresses.
Equations (27) and (28) can now be used to evaluate the new stresses from
Equation (26); these stresses simultaneously satisfy both yield conditions and
both flow rules.
If the element is only yielding in shear, then
(29)
(30)
If the element is only yielding in volume, then
(31)
(32)
Equation (29) to (32) may be used in Equation (26), as appropriate, to compute

new stresses.
We now consider the case where tensile failure is detected by the condition
0. If volumetric failure is not detected, we use the same technique and
stress corrections as described in the Mohr-Coulomb model. If volumetric failure
is detected in addition to tensile failure, then either 0 or
0. We begin by assuming that all three yield conditions are exceeded. We assume
that the plastic contributions of shear, volumetric, and tensile yielding are
additive:
(33)
The flow rule for tensile yielding has the form
(34)
Using Equation (8), these expressions become, after partial differentiation:
FLAC3D 6.0
(35)
Using the same reasoning as above, and Equations (23) and (24) for the shear
and volumetric flow rule, we obtain
(36)
The multipliers , , and are determined by solving the system of three

equations, = 0, = 0, and = 0. This gives
(37)
Substitution of those expressions in Equation (36) yields
(38)
If only tensile and volumetric yield are detected, then = 0 in Equation (37). The
constants and are determined by requiring that conditions =
0 and = 0 both be fulfilled. After some manipulation, we obtain
(39)
Substitution of those expressions in Equation (36) gives
FLAC3D 6.0
(40)
Hardening and softening behaviors for the cohesion, friction, and dilation in
terms of the shear parameter (see Equation (15)) are provided by the user in
the form of tables. Softening of the tensile strength is described in a similar
manner using the parameter (see Equation (16)). In turn, the variation of cap
pressure is specified in a table in terms of the parameter (see Equation (18)).
Each table contains pairs of values: one for the parameter, and one for the
corresponding property value. It is assumed that the property varies linearly
between two consecutive parameter entries in the table.
In the implementation of the double-yield model in FLAC3D, new stresses for the
step are computed using the current values of the model properties. In this
process, an elastic guess,, is first computed by adding to the old stress
components, increments calculated by application of Hooke’s law to the total
strain increment for the step. Principal stresses , and corresponding
principal directions, are calculated and ordered. If these stresses violate the
composite yield criterion, corrections are applied to the elastic guess to give the
new stress state. The stress tensor components in the system of reference axes
are then calculated from the principal values by assuming that the principal
directions have not been affected by the occurrence of a plastic correction.
Plastic strain increments are evaluated from Equations (23), (24), and (35), using
relevant expressions of , and for the mode of failure taking place. Zone
hardening increments are then calculated as the surface average of values
obtained from Equations (15), (18), and (16) for all triangles involved in the zone.
The hardening parameters are updated, and new zone properties for cohesion,
friction, dilation, tensile strength, and cap pressure are evaluated by linear
interpolation in the tables. New elastic constants are derived from the cap
pressure table using Equations (36). All these properties are stored for use in the
next step. The hardening or softening lags one timestep behind the
corresponding plastic deformation. In an explicit code, this error is small because
the steps are small.
FLAC3D 6.0
For a material with friction, the maximum value of the tensile strength is
evaluated from , using the new cohesion and friction angle. This value is
retained by the code if it is smaller than the tensile strength updated from the
table.
Choice of Volumetric Properties
The “hardening curve” and ratio, , of elastic bulk modulus to plastic bulk
modulus are volumetric properties that may be derived from the results of a
triaxial test in which axial stress and confining pressure, , are kept equal. This
test (for which ) is recommended because it is best to determine the
parameters related to a particular mode of failure from a test that only involves
that failure mode.
Figure 2: Isotropic consolidation test.
FLAC3D 6.0
Consider the experimental graph of minus mean stress (pressure) versus minus
volumetric strain for an increasing stress level with a small unloading excursion,
obtained from such a test and presented in Figure 2. The volumetric strain
increment at a point of the main loading path (assuming that we are above
any initial pre-consolidation stress level) is composed of an elastic part, , and
a plastic part, . (Recall that, in this section, , , and refer to minus the
value of the volumetric strain.) The observed tangent modulus may be expressed
as
(41)
where is the plastic modulus, is the elastic modulus, and, by definition:
(42)
(43)
With the preceding notation convention, the volumetric property, , may be

defined as
(44)
and Equation (41) becomes
(45)
Expressing from this relation, we obtain
(46)
Values for and can be estimated from main loading and unloading
increments on the graph. Hence, can be calculated from Equation (44). Note
that in the context of this model, the ratio is assumed to be constant. Using
and in Equation (41), we can write, after some manipulation,
(47)
FLAC3D 6.0
From this, it follows that values of for a particular can be obtained to the
first approximation by multiplying the value on the graph corresponding to
by the ratio . For example, if = 5, then the graph curve must be
scaled by a factor of 1.2 to convert it to table values, assuming no over-
consolidation.
The double-yield model was initially developed to represent the behavior of mine
backfill material, for which pre-consolidation pressures are low. The modified
Cam-clay model is more applicable to soils such as soft clays for which pre-
consolidation pressures can have a significant effect on material behavior.
Comparison of the Double-Yield and Modified Cam-Clay Model
In the double-yield model:
• Elastic moduli remain constant during elastic loading and unloading.

• Shear and tensile failure are not coupled to plastic volumetric change, due
to volumetric yielding. The shear yield function corresponds to the Mohr-
Coulomb criterion, and the tensile yield is evaluated based on a tensile
strength.
• Material hardening or softening, upon shear or tensile failure, is defined
by tables relating friction angle and cohesion to plastic shear strain, and
tensile strength to plastic tensile strain.
• Volumetric yielding occurs; the isotropic stress at which yield occurs (the
cap pressure) is not influenced by the amount of shear or tensile plastic
deformation, but it increases (or hardens) with volumetric plastic strain.
• A tensile strength limit and tensile softening can be defined.
In the modified Cam-clay model:
• The elastic deformation is nonlinear, with the elastic moduli depending

on mean stress.
• hear failure is affected by the occurrence of plastic volumetric
deformation: the material can harden or soften depending on the degree
of pre-consolidation.
• As shear loading increases, the material evolves toward a critical state at
which unlimited shear strain occurs with no accompanying change in
specific volume or stress.
• There is no resistance to tensile mean stress.
FLAC3D 6.0
double-yield model properties
properties of the double-yield model.
bulk-maximum f
maximum elastic bulk modulus,
cohesion f
cohesion,
dilation f
friction f
angle of internal friction,
multiplier f
multiplier on current plastic cap
modulus to give elastic bulk and
shear moduli, . The default is 5.0.
pressure-cap f
current intersection of volumetric
yield surface with pressure axis,
shear-maximum f
maximum elastic shear modulus,
tension f
flag-brittle b [a]
default is false.
FLAC3D 6.0
number of table relating dilation
angle to plastic shear strain. The
default is 0.
number of table relating friction
angle to plastic shear strain. The
default is 0.
table-pressure-cap i [a]
number of table relating cap
pressure to plastic volume strain.
The default is 0.
table-tension i [a]
number of table relating tensile limit
to plastic tensile strain. The default
is 0.
bulk f [r]
current elastic bulk modulus,
poisson f [r]
current Poisson’s ratio,
shear f [r]
current elastic shear modulus,
strain-tensile-plastic f [r]
strain-volumetric-plastic f [r]
accumulated plastic volumetric
strain
young f [r]
current elastic Young’s modulus,
FLAC3D 6.0
Notes:
• If table-pressure-cap is not assigned, the model moduli
will be constant, so that = and = . In any
case, and are upper limits of moduli.
at the same time.
Footnotes
FLAC3D 6.0
FLAC3D 6.0
Modified Cam-Clay Model 267
Modified Cam-Clay Model

The modified Cam-Clay model is an incremental hardening/softening
elastoplastic model. Its features include a particular form of nonlinear elasticity
and a hardening/softening behavior governed by volumetric plastic strain
(“density” driven). The failure envelopes are self-similar in shape and
correspond to ellipsoids of rotation about the mean stress axis in the principal
stress space. The shear flow rule is associated; no resistance to tensile mean
stress is offered in this model. See Roscoe and Burland (1968) and Wood (1990)
for detailed discussions on the modified Cam-Clay model. For convenience, we
drop the qualifier “modified” in the following discussion. Also, recall that all
models are expressed in terms of effective stresses. In particular, all pressures
referred to in this section are effective pressures.
The generalized stress components involved in the model definition are the mean
effective pressure, , and deviatoric stress, , defined as
(1)
where the Einstein summation convention applies, and is the second invariant
of the effective deviatoric-stress tensor :
(2)
The incremental strain variables associated with and are the volumetric strain
increment, , and shear strain increment, , and we have
(3)
where stands for the second invariant of the incremental deviatoric-strain

tensor :
(4)
FLAC3D 6.0
In the FLAC3D plastic flow formulation, it is assumed that both plastic and elastic
principal strain-increment vectors are coaxial with the current principal stress
vector. The generalized strain increments are then decomposed into elastic and
plastic parts so that
(5)
The evolution parameter is the specific volume, , defined as
(6)
where is the volume of solid particles, assumed incompressible, contained in a

volume, , of soil. The incremental relation between volumetric strain, , and
specific volume has the form
(7)
And the new specific volume, , for the step may be calculated as
(8)
The incremental expression of Hooke’s law in terms of generalized stress and

strains is
(9)
where , and stands for the second invariant of the incremental

deviatoric-stress tensor:
(10)
In the Cam-Clay model, the tangential bulk modulus, , in the volumetric

relation (Equation (9)) is updated to reflect a nonlinear law derived
experimentally from isotropic compression tests. The results of a typical
isotropic compression test are presented in the semi-logarithmic plot of Figure 1.
FLAC3D 6.0
Figure 1: Normal consolidation line and unloading-reloading (swelling) line for

an isotropic compression test.
As the normal consolidation pressure, , increases, the specific volume, , of the

material decreases. The point representing the state of the material moves along
the normal consolidation line, defined by the equation
(11)
where is defined as the slope of the normal consolidation line (not to be

confused with the plastic volumetric multiplier, , used in the plasticity flow
rule given in the section on plastic correction), and are two material
parameters, and is a reference pressure. (Note that is the value of the
specific volume at the reference pressure.)
An unloading-reloading excursion, from point or on the figure, will move

the point along an elastic swelling line of slope , back to the normal consolidation
line where the path will resume. The equation of the swelling lines has the form
FLAC3D 6.0
(12)
where is a material constant, and the value of for a particular line depends
on the location of the point on the normal consolidation line from which
unloading was performed (i.e., for unloading from point , and for
unloading from point in Figure 1).
The recoverable change in specific volume, , may be expressed in incremental

form after differentiation of Equation (12):
(13)
After division of both members by , and using Equation (7):
(14)
In the Cam-Clay model, it is assumed that any change in mean pressure is

accompanied by elastic change in volume according to the preceding expression.
Comparison with Equation (9), therefore, suggests the following expression for
the tangent bulk modulus of the Cam-Clay material:
(15)
Under more general loading conditions, the state of a particular point in the
medium might be represented by a point, such as , located below the normal
consolidation line in the plane (see Figure 2). By virtue of the law adopted
in Equation (12), an elastic path from that point proceeds along the swelling line
through .
FLAC3D 6.0
Figure 2: Plastic volume change corresponding to an incremental consolidation

pressure change.
The specific volume and mean pressure at the intersection of the swelling line
and normal consolidation line are referred to as (normal) consolidation (specific)
volume and (normal) consolidation pressure ( and , in the case of point ).
Consider an incremental change in stress bringing the point from state to state
. At , there is a corresponding consolidation volume, , and consolidation
pressure, . The increment of plastic volume change, , is measured on the
figure by the vertical distance between swelling lines (associated with points
and ), and we can write, using incremental notation,
(16)
After division of the left and right member by , we obtain, comparing with
Equation (7),
FLAC3D 6.0
(17)
Hence, whereas elastic volume changes take place whenever the mean pressure
changes, plastic volume changes occur only when the consolidation pressure
changes.
The yield function corresponding to a particular value, , of the consolidation

pressure has the form
(18)
where is a material constant. The yield condition = 0 is represented by an

ellipse with horizontal axis and vertical axis in the -plane (see Figure
3). Note that the ellipse passes through the origin; hence, the material in this
model is not able to support an all-around tensile stress.
The failure criterion is represented in the principal stress space by an ellipsoid of

rotation about the mean stress axis (any section through the yield surface at
constant mean effective stress, , is a circle).
The potential function corresponds to an associated flow rule, and we have
(19)
FLAC3D 6.0
Figure 3: Cam-Clay failure criterion in FLAC3D.
Plastic Corrections
The flow rule used to describe plastic flow has the form
(20)
where is the plastic (volumetric) multiplier whose magnitude remains to be

defined.
Using Equation (19) for , these expressions give, after partial differentiation:
(21)
where
(22)
FLAC3D 6.0
The elastic strain increments may be expressed from Equation (3) as total minus
plastic increments. In further using Equation (21), the elastic laws in Equation
(19) become
(23)
Let the new and old stress states be referred to by the superscripts and ,
respectively. Then, by definition, for one step:
(24)
Substitution of Equation (23) gives
(25)
where the superscript is used to represent the elastic guess obtained by adding
to the old stresses; elastic increments are computed using the total strain
increments:
(26)
The parameter may now be defined by requiring that the new stress point be
located on the yield surface. Substitution of and (as given by Equation (25))
for and in = 0 gives, after some manipulation (see Equation (18)),
(27)
where
(28)
Of the two roots of this equation, the one with the smallest magnitude must be
retained.
FLAC3D 6.0
Note that at the critical point corresponding to , in Figure 3,

the normal to the yield curve = 0 is parallel to the -axis. Since the flow rule is
associated, the plastic volumetric strain-rate component vanishes there. As a
result of the hardening rule (Equation (17)), the consolidation pressure, , will
not change. The corresponding material point has reached the critical state in
which unlimited shear strains occur with no accompanying change in specific
volume or stress level.
The new stress components, , are expressed in terms of old and new
generalized stress values, using the expressions
(29)
where
(30)
and is the deviatoric stress tensor.
Hardening/Softening Rule
The size of the yield curve is dependent on the value of the consolidation
pressure, (see Equation (18)). This pressure is a function of the plastic volume
change, and varies with the specific volume as indicated in Equation (17). The
consolidation pressure is updated for the step, using the formula
(31)
where is the plastic volumetric strain increment for the step, is the current
specific volume, and and are material parameters, as given previously in
Equation (11).
Initial Stress State
The Cam-Clay model in FLAC3D is only applicable to material in which the stress
state corresponds to a compressive mean effective stress. This model is not
designed to predict the behavior of material in which this condition is not met. In
particular, the initial state of the material (just before application of the Cam-
Clay model) must be consistent with this requirement. The initial stress state
may be specified using the zone initialize command, or may be the result of a
FLAC3D 6.0
run in which another constitutive model has been used. An error message will be
issued if the Cam-Clay model is applied to a zone where the initial effective
pressure, defined as , is negative.
Over-consolidation Ratio
The over-consolidation ratio, , is defined as the ratio between the zone initial
pre-consolidation pressure (a material property) and initial pressure, :
(32)
This ratio is useful in characterizing the behavior of Cam-Clay material.
In the implementation of the Cam-Clay model in FLAC3D, an elastic guess, , is

first computed after adding to the old stress components, increments calculated
by application of Hooke’s law to the total strain increment for the step.
Elastic guesses for the mean pressure, , and deviatoric stress, , are calculated
using Equation (1). If these stresses violate the criterion for yield and 0
(see Equation (18)), plastic deformation takes place and the consolidation
pressure must be updated. In this situation, a correction must be applied to the
elastic guess to give the new stress state. New stresses and are first
evaluated from Equation (25) using the expression for corresponding to the
root of Equation (27) and (28) with smallest magnitude. Note that in this version
of the code, the expressions in Equation (22) for and are evaluated using the
elastic guess; the error associated with this technique is expected to be small,
provided the steps are small. New stress tensor components in the system of
reference axes are hence evaluated using Equation (29) and (30).
Volumetric strain increment, , and mean pressure, , for the zone are
computed as average over all involved tetrahedra (see Equation (1) and (3)). The
zone volumetric strain, , is incremented, and the zone specific volume, , is
updated, using Equation (8). In turn, the new zone consolidation pressure is
calculated from Equation (31), and the tangential bulk modulus is updated using
formula (Equation (15)).
If a nonzero value for the Poisson’s ratio property is imposed, a new shear
modulus is calculated from the expression . Otherwise,
is left unchanged as long as the condition 0.5 is satisfied; if it is not, is
assigned a value of = 0 or = 0.5, as appropriate.
FLAC3D 6.0
The new values for the consolidation pressure and shear and bulk moduli are
then stored for use in the next timestep. The material properties thus lag one
timestep behind the corresponding calculation. In an explicit code, this error is
small because the steps are small.
Determination of the Input Parameters
Frictional constant M — is the ratio of at the critical state line. Therefore,

a series of triaxial tests (drained or undrained with pore-pressure measurement)
can be used to obtain this constant. These tests should be carried out to large
strains to ensure that the final values of and are close to the critical state
line. The slope of a best-fitting line of vs will be the parameter .
is related to the effective stress friction angle, , of the Mohr-Coulomb yield

function. However, since the Cam-Clay critical state line is dependent on the
intermediate stress , while Mohr-Coulomb is not, the relation between and
will be different for different values of at yield. (This condition is similar to
the relation between Mohr-Coulomb and Drucker-Prager yield functions.) For
triaxial compression tests,
(33)
while, for triaxial extension tests,
(34)
The slopes of the normal consolidation and swelling lines ( and ) — Ideally, these
two parameters should be obtained from an isotropically loaded triaxial test ( =
0), with several unloading excursions. The slope of the normal compression line
in a versus plot will be the parameter . The slope of an unloading
excursion in the same plot will be the parameter .
These two parameters can also be derived from an oedometer test, making
certain assumptions. Let and be the vertical and horizontal stresses in an
oedometer test. In most oedometer apparatus it is not possible to measure the
horizontal stresses, , so the mean stress, , is not known.
However, experimental data show that the ratio of horizontal to vertical effective
stresses, , is constant during normal compression. Since
along the normal compression line, the slope of vs will be equal to the slope
of versus , where is the void ratio = .
FLAC3D 6.0
The compression index, , is calculated as the slope of vs . So the

parameter will be
(35)
Experimental data show that along a swelling line in an oedometer test, is not
constant, so an estimate of based on the swelling coefficient, , will only be an
approximation.
(36)
In practice, is usually chosen in the range of one-fifth to one-third of .
Location of the normal consolidation line in the versus plot — In order to

determine the location of the normal consolidation line in the versus plot, a
point , on this line must be specified. The obvious way to determine this
point is to perform an isotropic triaxial test. There is an alternative way to
determine this point based on the undrained shear strength (for details, see
Britto and Gunn 1987).
The equation of the normal consolidation line is (also, see

Equation (11)).
The specific volume intercept, , at the critical state line for , is given by
(37)
In a soil, the undrained shear strength, , is uniquely related to the specific

volume, , by the equation
(38)
Thus, the value of for a given , and therefore , can be calculated if the
undrained shear strength for a particular specific volume, , along with the
parameters , , and , is known.
Pre-consolidation pressure, — The pre-consolidation pressure determines the

initial size of the yield surface in the equation
(39)
If a sample has been submitted to an isotropic loading path, will be the

maximum past mean effective stress. If the sample has followed other non-
isotropic paths, has to be calculated from the previous maximum and ,
using Equation (39).
FLAC3D 6.0
The maximum vertical effective stress can be calculated from an oedometer test
using Casagrande’s method (for details, see Britto and Gunn 1987). Some
hypothesis has to be made about the maximum horizontal effective stress. A
common hypothesis is Jaky’s relation:
(40)
where is the coefficient of horizontal to vertical stress at rest

for normally consolidated soil. For example, if a soil with an effective friction
angle of 20° has experienced a maximum vertical effective stress, = 1 MPa.
Then, using Jaky’s relation,
and the maximum horizontal stress is
(41)
The maximum values of and are
Substituting these two values in the yield function (Equation (18)), we obtain the
pre-consolidation pressure,
Initial values for specific volume, , and bulk modulus, — Given an initial
effective pressure , the initial specific volume must be consistent with the
choice of parameters , and . The initial value is calculated to
correspond to the value of the specific volume corresponding to on the
swelling line through the point on the normal consolidation line at which .
From Figure 4, it follows that
(42)
FLAC3D 6.0
Figure 4: Determination of initial specific volume.
The initial value of the bulk modulus, , may in turn be evaluated using
Equation (15), which gives
(43)
In FLAC3D, the default values for the properties and are evaluated using
Equations (42) and (43). when the first step command is issued.
Note that the mean effective pressure, , must be initialized corresponding to the
initial stress state before stepping begins with the Cam-Clay model. This can be
done with a FISH function (e.g., in the isotropic consolidation example).
Maximum value of the elastic parameters and — In the Cam-Clay model, the
value of the bulk modulus, , changes as a function of the specific volume and
the mean stress:
(44)
FLAC3D 6.0
The input values of and are used in the mass scaling calculation
performed in FLAC3D to ensure numerical stability. This calculation is done once
every time a model step command is issued. These input values should be chosen
to give an upper bound to the sum , as evaluated by the model
between two consecutive model step commands. Values should not, however, be
set too high or the model may be slow to converge. They should be selected based
on the stress level in the problem.
or — The modified Cam-Clay model in FLAC3D allows the user to specify

either a constant shear modulus or a constant Poisson’s ratio.
If no Poisson’s ratio is specified, a constant shear modulus equal to the input

value is assumed. Then the Poisson’s ratio will vary as a function of the specific
volume and the mean stress:
(45)
If a nonzero Poisson’s ratio is specified, the shear modulus will vary at the same
rate as the bulk modulus in order to maintain a constant Poisson’s ratio:
(46)
The comparison between the double-yield model and modified Cam-clay model
are summarized here.
modified-cam-clay Model Properties
properties of the modified Cam-Clay model.
bulk-maximum f
maximum elastic bulk modulus,
kappa f
slope of elastic swelling line,
lambda f
slope of normal consolidation line,
FLAC3D 6.0
poisson f
Poisson’s ratio,
pressure-reference f
reference pressure,
pressure-effective f
mean effective stress, , required
since this is a pressure-dependent
model.
pressure-preconsolidation f
pre-consolidation pressure,
ratio-critical-state f
stress ratio at the critical state,
at the critical state
shear f
specific-volume-reference f
specific volume at reference pressure
on normal consolidation line,
bulk f [r]
specific-volume f [r]
current specific volume, ,
initialized or updated internally.
strain-volumetric-total f [r]
accumulated total volumetric strain,
stress-deviatoric f [r]
current deviatoric stress,
Notes:
• If the current bulk modulus, , is greater than , an
error message will suggest increasing .
FLAC3D 6.0
• Only one between the Poisson’s ration, , and the shear

modulus, , is required for input. If is not specified,
and a non-zero is specified, then remains constant
and will change as changes. If a non-zero is given,
then will change as changes and remains constant.
Footnotes
FLAC3D 6.0
FLAC3D 6.0
Hoek-Brown Model 285
Hoek-Brown Model
The Hoek-Brown-PAC model provides a representation for yielding that accounts
for the changing failure condition. This model works well at higher confining
stress states, but can produce excessive dilation at low confinement or under
tensile-stress conditions. The alternative version, described in this section, is
modified to include a tensile yield criterion, and also allow the user to specify a
dilation angle as an input parameter and manually control the level of dilation
that develops.
The Hoek-Brown model is derived directly from the Mohr-Coulomb model, and,
like the Mohr-Coulomb model, can be used to perform factor of safety
calculations using the model factor-of-safety command. The formulation and an
example exercise are given below.
Formulation and Implementation
The Hoek-Brown criterion is used for plastic yielding when the minor principal
stress, , is compressive. The criterion is based on a nonlinear relation between
major and minor principal stresses, and , as shown previously in the Hoek-
Brown-PAC model, and repeated here:
(1)
where is the unconfined compressive strength of the intact rock, and , ,

and are material constants that can be related to the Geological Strength Index
and rock damage (Hoek et al. 2002).
The Hoek-Brown envelope is extended for tensile by a combination of the

Mohr-Coulomb envelope (tangent to Hoek-Brown at ) and a tensile cutoff
at .
The numerical implementation of the Hoek-Brown model uses a linear

approximation (see Figure 1), whereby the nonlinear failure surface is
continuously approximated by the Mohr-Coulomb tangent at the current stress
level, . The current tangent Mohr-Coulomb criterion is
(2)
FLAC3D 6.0
where
(3)
Figure 1: Hoek-Brown failure criterion and its Mohr-Coulomb approximation.
The current (apparent) value of cohesion, , and friction, , are calculated using
(4)
FLAC3D 6.0
(5)
where
(6)
(7)
The Mohr-Coulomb envelope extension in the region of tensile is accounted

for in the logic by considering the following cap in Equations (6) and (7) (recall
that compressive stress is positive):
(8)
The tensile yield logic is the same as the one used for the strain-softening/
hardening Mohr-Coulomb model.
The plastic strain increment for shear yielding is defined using the current
Mohr-Coulomb flow rule:
(9)
where is the plastic flow increment intensity, is the plastic potential

function, and
(10)
The plastic potential function is
(11)
where is the current value of dilation and
(12)
The direction of plastic flow ( in Equation (9)) is expressed using the plastic
potential function. The plastic flow increment intensity, , is derived from the
current tangent Mohr-Coulomb yield criterion, Equation (9). There are three
choices of flow rule for the model:
FLAC3D 6.0
1. Input a dilation angle ( is a constant value) specified by

property keyword constant-dilation and flag-dilation = 0.
2. Specify associated plastic flow ( is set equal to ) by setting

flag-dilation = -1.
3. Set dilation as a fraction of the friction angle ( is set to a

constant time ); flag-dilation is a positive fraction.
Strain hardening/softening behavior can be prescribed for the Hoek-Brown

properties, , , , and . The input is via tables, in terms of an evolution
parameter. There are two choices for evolution parameter: plastic shear strain
(property keyword flag-evolution is set to 1, only acuumulated when shear
yielding occurs), or plastic strain in the direction of the least compressive
principal stress, (property keyword flag-evolution is set to 0). The evolution
parameter can be monitored via the property strain-plastic.
A simple regularization technique can be selected to address the issue of grid

dependency on softening behavior. To activate this technique, the grid zone size
used to calibrate model properties with experimental data is assigned to the
property length-calibration. This property is the calibration length. The input
softening rate is then adjusted automatically to account for a different zone size
used in the full FLAC3D model. Note that this technique is experimental and
should be used with caution.
hoek-brown Model Properties
properties of the Hoek-Brown model.
bulk f
bulk modulus,
constant-a f
Hoek-Brown parameter,
FLAC3D 6.0
constant-dilation f
dilation angle, , required only
when flag-dilation = 0. The default
is 0.0.
constant-mb f
constant-s f
constant-sci f
current-a f
current value of
current-mb f
current value of
current-s f
current value of
current-sci f
current value of
poisson f
Poisson’s ratio,
shear f
shear modulus,
tension f
current value of tensile strength, .
The default is 0.0.
young f
Young’s modulus,
flag-brittle b [a]
default is false.
FLAC3D 6.0
length-calibration f [a]
calibration length to calibrate model
properties to account for zone size,
default value is 0.0, which means
that it does not perform length
calibration to account for zone size.
flag-dilation f [a]
= 0 (default flag) to input a constant
dilation angle specified by constant-
dilation
= -1 to specify associated plastic

flow, =
= k where k is a fraction of friction

angle, , so that =
flag-evolution i [a]
= 0 (default flag) for evolution
parameter set to plastic strain in
direction of least compressive
principal stress
= 1 for evolution parameter set to

plastic shear strain, only
acuumulated when shear yielding
occurs
flag-fos i [a]
= 0 (default flag) for model factor-
of-safety solution controlled by
shear strength
= 1 for model factor-of-safety

solution controlled by unconfined
compressive strength
table-a i [a]
number of table relating to the
evolution parameter. The default is
0.
FLAC3D 6.0
table-m i [a]
0.
table-multiplier i [a]
number of table relating a multiplier
to . The default is 0.
table-s i [a]
0.
table-sci i [a]
0.
table-tension i [a]
number of table relating to plastic
tensile strain. The default is 0.
cohesion f [r]
current value of cohesion,
dilation f [r]
current value of dilation angle, , in
any cases, this value never greater
than the current friction angle.
friction f [r]
current value of friction angle,
strain-plastic f [r]
plastic strain in direction of least
compressive principal stress (if flag-
evolution = 0), or plastic shear strain
(if flag-evolution = 1)
FLAC3D 6.0
Notes:
• There are two sets of parameters for ( , , , ): one is
called constant set, which are the values assigned at the
first time and remain constant thereafter; the other is
called current set, which are automatically initialized as
the same values of the constant set, but can be updated
thereafter using tables or FISH functions if parameter
evolution/update is involved, otherwise, the current set
can be considered only for output.
at the same time.
Footnotes
can be listed, plotted or accessed through FISH.
FLAC3D 6.0
Hoek-Brown-PAC Model 293
Hoek-Brown-PAC Model
The Hoek-Brown failure criterion is an empirical relation that characterizes the
stress conditions that lead to failure in intact rock and rock masses. It has been
used very successfully in design approaches that use limit equilibrium solutions,
but there has been little direct use in numerical solution schemes. Alternatively,
equivalent friction and cohesion have been used with a Mohr-Coulomb model
that is matched to the nonlinear Hoek-Brown strength envelope at particular
stress levels Numerical solution methods require full constitutive models, which
relate stress to strain in a general way; in addition to a failure (or yield) criterion,
a “flow rule” is also necessary, in order to provide a relation between the
components of strain rate at failure. There have been several attempts to develop
a full constitutive model from the Hoek-Brown criterion: Pan and Hudson
(1988), Carter et al. (1993) and Shah (1992), for example. These formulations
assume that the flow rule has some fixed relation to the failure criterion, and
that the flow rule is isotropic, whereas the Hoek-Brown criterion is not. In the
formulation described here, there is no fixed form for the flow rule: it is assumed
to depend on the stress level, and possibly some measure of damage.
In what follows, the failure criterion is taken as a yield surface, using the
terminology of plasticity theory. Usually, a failure criterion is assumed to be a
fixed, limiting stress condition that corresponds to ultimate failure of the
material. However, numerical simulations of elastoplastic problems allow
continuing the solution after “failure” has taken place, and the failure condition
itself may change as the simulation progresses (by either hardening or
softening). In this event, it is more reasonable to speak of “yielding” than
failure. There is no implied restriction on the type of behavior that is modeled:
both ductile and brittle behavior may be represented, depending on the softening
relation used.
General Formulation
The “generalized” Hoek-Brown criterion (Hoek and Brown, 1980 and 1998),
adopting the convention of positive compressive stress, is
(1)
FLAC3D 6.0
where and are the major and minor effective principal stresses, and , ,
, and are material constants that can be related to the Geological Strength
Index ( ) and rock damage (Hoek et al. 2002). For interest, the unconfined
compressive strength is given by = , and the tensile strength by = -
/ , (see Figure 1). Note that the criterion (Equation (1)) does not depend on the
intermediate principal stress, . Thus, the failure envelope is not isotropic.
Figure 1: Hoek-Brown failure criterion.
Assume that the current principal stresses are ( ), and that initial trial
stresses ( , , ) are calculated by using incremental elasticity:
FLAC3D 6.0
(2)
where and , and ( ) is the set of

principal strain increments. If the yield criterion (Equation (1)) is violated by this
set of stresses, the strain increments (prescribed as independent inputs to the
model) are assumed to be composed of elastic and plastic parts:
(3)
Note that plastic flow does not occur in the intermediate principal stress
direction. The final stresses ( , , ) output from the model are related to the
elastic components of the strain increments. Hence:
(4)
Eliminating the current stresses, using Equations (2) and (4):
(5)
We assume the flow rule
(6)
where the factor depends on stress and is recomputed at each timestep.

Eliminating from Equation (5):
FLAC3D 6.0
(7)
At yield, Equation (1) is satisfied by the final stresses. That is,
(8)
By substituting values of and from Equation (7), Equation (8) can be solved
iteratively for , which is then substituted in Equation (7) to give the final
stresses. The method of solution is described later, but first the evaluation of is
discussed.
Flow Rules
We need to consider an appropriate flow rule, which describes the volumetric

behavior of the material during yield. In general, the flow parameter will
depend on stress, and possibly on history. It is not meaningful to speak of a
“dilation angle” for a material when its confining stress is low or tensile,
because the mode of failure is typically by axial splitting, not shearing. Although
the volumetric strain depends, in a complicated way, on stress level, we consider
certain specific cases for which behavior is well-known, and determine the
behavior for intermediate conditions by interpolation. Three cases are
considered:
Associated Flow Rule
It is known that many rocks under unconfined compression exhibit large rates of
volumetric expansion at yield, associated with axial splitting and wedging
effects. The associated flow rule provides the largest volumetric strain rate that
may be justified theoretically. This flow rule is expected to apply in the vicinity of
the uniaxial stress condition ( ). An associated flow rule is one in which the
vector of plastic strain rate is normal to the yield surface (when both are plotted
on similar axes). Thus,
(9)
FLAC3D 6.0
where the subscripts denote the components in the principal stress directions,
and is defined by Equation (8). Differentiating this expression, and using
Equation (6),
(10)
Radial Flow Rule
Under the condition of uniaxial tension, we might expect that the material would
yield in the direction of the tensile traction. If the tension is isotropically applied,
we imagine (since the test is almost impossible to perform) that the material
would deform isotropically. Both of these conditions are fulfilled by the radial
flow rule, which is assumed to apply when all principal stresses are tensile. For a
flow-rate vector to be coaxial with the principal stress vector, we obtain
(11)
Constant-Volume Flow Rule
As the confining stress is increased, a point at which the material no longer

dilates during yield is reached. A constant-volume flow rule is therefore
appropriate when the confining stress is above some user-prescribed level,
. This flow rule is given by
(12)
Composite Flow Rule
We propose to assign the flow rule (and, thus, a value for ) according to the
stress condition. In the fully tensile region, the radial flow rule ( ) will be used.
For compressive and tensile or zero , the associated flow rule ( ) is applied.
For the interval , the value of is linearly interpolated between the
associated and constant-volume limits:
(13)
Finally, when , the constant-volume value, , is used.
If is set equal to zero, then the model condition approaches a nonassociated

flow rule with a zero dilation angle. If is set to a very high value relative to ,
the model condition approaches an associated flow state.
FLAC3D 6.0
The equations presented above are implemented. One difficulty with the failure
criterion Equation (8) is that real values for do not exist if .
During an iteration process, this condition is likely to be encountered, so it is
necessary that the expression for , and its first derivatives, be continuous
everywhere in stress space. This is fulfilled by adapting the composite expression
(14)
(15)
To initialize the iteration, a starting value for is taken as the absolute

maximum of all the strain increment components. This value, denoted by , is
inserted into Equation (7), together with the value for found from the flow-rule
equations, and the resulting stress values inserted into Equations (14) and (15).
The resulting value of is denoted by . Taking the original value of as
(and the corresponding plastic strain increment of zero as ), we can estimate
a new value of the plastic strain increment, using a variant of Newton’s method:
(16)
From this, we find a new value of (call it ) and, if it is sufficiently close to

zero, the iteration stops. Otherwise, we set , , , and
and apply Equation (16) again.
Tests show that the iteration scheme converges for all stress paths tried so far,
including cases in which = 0 (material with zero unconfined compressive
strength), which led to problems in previous implementations. For high
confining stresses, the iteration converges in one step, but at low confining
stresses, up to ten steps are necessary (the limit built into the code is presently
15).
Material Softening
In the Hoek-Brown model, the material properties , , , and are assumed

to remain constant, by default. Material softening after the onset of plastic yield
can be simulated by specifying that these mechanical properties change (i.e.,
reduce the overall material strength) according to a softening parameter. The
softening parameter selected for the Hoek-Brown model is the plastic confining
FLAC3D 6.0
strain component, . The choice of is based on physical grounds. For yield

near the unconfined state, the damage in brittle rock is mainly by splitting (not
by shearing) with crack normals oriented in the direction. The parameter is
expected to correlate with the microcrack damage in the direction.
The value of is calculated by summing the strain increment values for

calculated by Equation (16). Softening behavior is provided by specifying tables
that relate each of the properties , , , and to . Each table contains pairs
of values: one for the value, and one for the corresponding property value. It is
assumed that the property varies linearly between two consecutive parameter
entries in the table.
A multiplier, (denoted as multable), can also be specified to relate the softening

behavior to the confining stress . The relation between and is also given in
the form of a table. (See Cundall et al. (2003) for an application of softening
parameters.)
hoek-brown-pac Model Properties
properties of the Hoek-Brown-PAC model.
bulk f
bulk modulus,
constant-a f
constant-mb f
constant-s f
constant-sci f
poisson f
Poisson’s ratio,
shear f
shear modulus,
FLAC3D 6.0
stress-confining-prescribed f
Hoek-Brown parameter, confining
stress at user-prescribed level,
young f
Young’s modulus,
table-a i [a]
number of table relating to . The
default is 0.
table-mb i [a]
number of table relating to to .
The default is 0.
table-multiplier i [a]
number of table relating a multiplier
to . The default is 0.
table-s i [a]
default is 0.
table-sci i [a]
default is 0.
strain-3-plastic f [r]
accumulated plastic strain in
direction of least compressive
principal stress,
number-iteration i [r]
number of iterations
FLAC3D 6.0
Notes:
Footnotes
FLAC3D 6.0
FLAC3D 6.0
Cap-Yield (CYSoil) Model 303
Cap-Yield (CYSoil) Model

The CYSoil model is a strain-hardening constitutive model characterized by a
frictional and cohesive Mohr-Coulomb shear envelope and an elliptic volumetric
cap with ratio of axes, defined by a shape parameter . The CYSoil model in
FLAC3D 6 is an updated version of the FLAC3D 5 CYSoil model with the following
built-in features:
• a cap hardening law to capture the volumetric power law behavior

observed in isotropic compaction tests;
• a friction-hardening law to reproduce the hyperbolic stress-strain law
behavior observed in drained triaxial tests; and
• a compaction/dilation law to model irrecoverable volumetric strain taking
place as a result of soil shearing.
The motivation for closing the yield surface by a cap on the mean stress axis is to
permit plastic behavior in response to an isotropic stress increase. This plasticity
effect accounts for grain crushing and rearrangement, and is particular to soils.
In the double-yield model, the cap is a plane, normal to the mean stress axis in
stress space. The impact of this particular shape on the coefficient of lateral
earth pressure, , as predicted by the model in uniaxial compression tests, has
been considered to be somewhat restrictive by some users. In that respect, the
CYSoil model is a modification of the double-yield model that addresses this
issue by accounting for a cap with an elliptic shape in the ( , ) plane. The ratio
of axes of the ellipse, , determines the value of and is a material property for
the model, which can be chosen to match a known value in uniaxial compression.
In particular, a planar volumetric cap is obtained as a special case of the CYSoil
formulation by assuming a value .
In addition, when subjected to deviatoric loading, soils usually exhibit a decrease

in stiffness, accompanied by irreversible deformation. In most cases, the plot of
deviatoric stress versus axial strain obtained in a drained triaxial test may be
approximated by a hyperbola. This feature has been used by Duncan and Chang
(1970) to formulate their well-known “hyperbolic soil” model. The hyperbolic
soil model of Duncan and Chang is a nonlinear elastic model that has been shown
to exhibit some drawbacks. These drawbacks include, for example, difficulty in
detecting and characterizing unloading/reloading and,in specific cases,
FLAC3D 6.0
producing a nonphysical bulk-modulus value that can lead to an erroneous

energy generation in the model. Because the CYSoil model is formulated in the
theory of hardening plasticity, it allows for an alternative expression of the
hyperbolic behavior (based on friction hardening), which is capable of addressing
some of these problems. A simplified version of the CYSoil model, called the
CHSoil model, also addresses the difficulties of the Duncan and Chang model and
is provided as an alternative to the Duncan and Chang model.
When tested under drained triaxial conditions, soils generally exhibit shear-
induced volume changes that are strongly dependent on soil density. Typically,
there is a tendency for the soil to contract under small shear strains, and to dilate
under larger strains, unless it is very loose (Byrne et al. 2003). In particular,
when fluid fills the pores, it is this tendency of the soil skeleton to contract and
dilate that controls its liquefaction response. Also, the shear-stress/shear-strain
response of loose soil may exhibit a softening response under undrained
conditions. It is the existence of a peak in shear strength which may lead to
instability (static liquefaction) during a monotonic load-controlled process
(Boukpeti 2001). Shear-induced volume changes can be accounted for in the
CYSoil model by means of a dilation hardening/softening law.
In this section, principal stress, , and strain, , i = 1,3, components are positive
in tension. Also, all stresses are effective by default if not explicitly stated. The
principal effective stresses are with the order , and is the most
compressive stress.
The elastic behavior is expressed using Hooke’s law. The incremental expression
of the law in terms of principal stress and strain has the form
(1)
where , , and and are current, tangent

elastic bulk and shear modulus, respectively.
In the constitutive model logic, the modulus is related to the mean effective
pressure by a power law. The user provides the value of Poisson ratio , which is
assumed to remain constant. In addition, the user can provide upper and lower
bound values for , which will then overwrite the default settings for these
FLAC3D 6.0
parameters. Note that the upper bound for is used for numerical stability: a
value close to the maximum value reached during the simulation should be
assigned to the parameter; if a much larger value is assigned, the numerical
convergence may be slow.
The bulk modulus is derived internally, using the relation:
(2)
Shear Yield Criterion and Flow Rule
Shear yielding is defined by a Mohr-Coulomb criterion (see representation in

Figure 1):
(3)
where , is the mobilized friction, a quantity

which can increase between an initial value (possibly zero), and a final value
set by the user.
Figure 1: Mohr circles at in-situ stress and at failure.
FLAC3D 6.0
Note that in the model logic, unloading is elastic, and thus cannot decrease.
The last term of Equation (3) contains mobilized cohesion, , which can
increase between an initial value and an ultimate value, , defined by the user.
The expression for is:
(4)
The shear yield envelope (see Equation (3)) can also be expressed in an
alternative form (consistent with the cap formulation) as follows:
(5)
where , is the mean effective stress, = -( + +

)/3, and is a measure of shear stress, defined as ,
where .
The potential function is nonassociated and has the form
(6)
where , ,
, and is the mobilized dilatancy angle.
Volumetric Cap Criterion and Flow Rule
Yielding on the cap is associated with the criterion:
(7)
where is a dimensionless parameter, defining the shape of the elliptical cap in

the ( , ) plane, and is the current cap pressure. Note that the ellipsoid
degenerates into a spherical cap for (default setting), and to a planar
volume cap for .
Tensile Yield Criterion and Flow Rule
The tensile yield function is the same as that used for the Mohr-Coulomb model:
(8)
where is the tensile strength. The potential function for tensile yielding is
associated.
FLAC3D 6.0
Built-in Hardening Functions
Cap Hardening
Soil stiffness usually increases in a nonlinear fashion as a function of isotropic

pressure. In the CYSoil model, soil volumetric behavior in an isotropic
compaction test is described by the following power law:
(9)
where is the volumetric strain taken positive in compression, is the

tangent elastic bulk modulus number, the product is the slope of the
laboratory curve for versus at reference effective pressure, , and is a
constant ( ).
A typical graph, representative of this law, is sketched with a small unloading

excursion in Figure 2. The equation for the curve (obtained by integration of
Equation (9) and using at is:
(10)
The usual assumption is made that the total strain increment is the sum of the
elastic and plastic contributions:
(11)
In this context, is the plastic volumetric strain contribution from cap yielding
only, not including shear yielding.
FLAC3D 6.0
Figure 2: Isotropic consolidation test: pressure versus volumetric strain.
Also, the tangent elastic modulus, , and plastic modulus, are defined as
follows:
(12)
(13)
As the material becomes more compact, its plastic stiffness usually increases; it
seems reasonable that the elastic stiffness will also increase, since the grains are
being forced closer together. The CYSoil model uses a simple rule whereby under
general loading conditions, the current elastic modulus is equal to a constant,
times the current plastic modulus:
(14)
Substitution of H from Equation (13) in Equation (14) gives:
FLAC3D 6.0
(15)
Using Equation (11) in Equation (15), we obtain
(16)
For isotropic compression, and by definition of and (see Equations

(12) and (13)), it follows from Equation (14) that
(17)
Now, substituting Equation (16) into Equation (15), one obtains
(18)
Thus, using the definition (Equation (9)) for in Equation (18), the tangent
elastic bulk modulus, is dependent on cap pressure, (or maximum effective
pressure sustained by the material in the past), according to the following power
law:
(19)
According to Equations modelcysoileqKe and (18), the tangent elastic shear

modulus obeys the following power law:
(20)
where
(21)
The input parameter , , , and must be provided by the user.
In addition, the user can provide upper and lower bound values for ( and
), which will then overwrite the default settings for these parameters. Note
that the upper bound for is used for numerical stability: a value close to the
maximum value reached during the simulation should be assigned to the
parameter. If a much larger value is assigned, the numerical convergence may be
slow.
FLAC3D 6.0
After substituting Equation (19) into Equation (14), the expression for the
hardening modulus, , becomes
(22)
Integration of Equation (22) for m < 1 by assuming = 0 at = 0 provides the

evolution law for the cap pressure, , in terms of plastic volumetric strain, :
(23)
Conversely, the plastic volumetric strain is expressed in terms of as:
(24)
Note that the code sets an upper bound of 0.99 for . However, it may be
interesting to know that in the limiting case of = 1, Equation (22) for m < 1 by
assuming = 0 at = 0 provides a logarithmic law that is similar to the
expression used by the Cam-Clay model:
(25)
Friction Hardening
For most soils, the plot of deviatoric stress versus axial strain obtained in a
drained triaxial test can be approximated by a hyperbola (Figure 3). The CYSoil
model incorporates a friction-strain hardening law to capture this behavior.
FLAC3D 6.0
Figure 3: Deviatoric stress versus axial strain for a generic triaxial test.
The evolution parameter for the cap volumetric yielding envelope and that for
the shear yielding envelope are independent in the CYSoil model. By convention,
the mobilized plastic shear moduli, is defined as:
(26)
where is the mobilized friction angle and is plastic shear strain. Thus, the
ratio is the slope of the curve in a graph of mobilized stress ratio, ,
versus plastic shear strain . The expression for mobilized plastic shear modulus
is adopted from Byrne et al. (2003):
(27)
In this formula, is an internal constant (which, by default is set equal to the

user-provided mobilized friction angle under in-situ conditions), is the
ultimate friction angle, is the value of under in-situ conditions, is the
failure ratio used to assign a lower bound for . Note that < 1, and a typical
value used in many cases is = 0.9.
FLAC3D 6.0
For the CYSoil model, is expressed as
(28)
where is a user-defined calibration factor, which can depend on the value of

confinement (in triaxial test modeling) or in-situ confining stress (in field scale
modeling).
After substituting Equations (27) and (28) in Equation (26) and rearranging some
terms, we obtain:
(29)
The evolution law for mobilized friction, , in terms of plastic shear strain, ,
for the CYSoil model is obtained by integration of Equation (27), considering that
= 0 at :
(30)
Equation (30) is used in the built-in logic to calculate the mobilized friction from
the accumulated plastic shear strain.
Note that after solving Equation (30) for with , one obtains:
(31)
The mobilized plastic shear modulus can be expressed in terms of plastic shear
strain by substitution of Equation (31) for into Equation (27) and
performing some manipulations:
(32)
where
(33)
The graph of versus is represented in generic form in Figure 4.
FLAC3D 6.0
Figure 4: Generic form of plot versus .
It is assumed that the curve at very small strain (see Figure 4) is derived from
in-situ conditions (e.g., by using seismic shear wave velocity profiles drawn by
cross-hole or down-hole tests), and that is the plastic shear strain that
develops after initial conditions have been established. In this case, the initial
value of the evolution parameter should be left at zero (default value). The
code will automatically set the parameter equal to the initial value of
mobilized friction, friction mob, provided by the user. The implication is that the
graph in Figure 4 will be exercised starting at = 0 during the simulation.
Alternatively, the user can exercise the modulus degradation curve in a

simulation, starting with a value of that is larger than zero. This option
provides additional calibration flexibility. Note that when the initial non-zero
value of is provided by the user, the code automatically sets the internal
constant to zero. For example, Equation (31) can be used to derive a value of
plastic shear strain, consistent with the user-provided mobilized friction angle,
for input into the code.
FLAC3D 6.0
Dilation Hardening
A certain amount of irrecoverable volumetric strain, , is expected to take place

as a result of soil shearing. Also, under small (monotonic or cyclic) shear strains,
there is a tendency for the soil skeleton to contract due to grain rearrangements.
For larger shear strains, the soil skeleton may dilate if the soil is dense as a result
of grains riding over each other. A dilation strain-hardening table is used to
model this non-monotonic behavior. For the CYSoil model, the shear-hardening
flow rule has the form
(34)
where is the mobilized dilation angle. The evolution law for mobilized dilation
angle ψm is given by the following law, based on Rowe’s stress-dilatancy theory
(1962):
(35)
where is given in terms of by Equation (30), and is a constant.

According to the law of Equation (35), the material contracts for
and dilates for . In most cases, the constant volume stress ratio
can be expressed in terms of (known) ultimate values of friction, ,
and dilation, , as follows:
(36)
and and are ultimate (known) values of friction and dilation, respectively.
Dilatancy Cut-off
In addition, a dilatancy cut-off can be specified whereby the mobilized dilation

angle is set to zero when the void ratio, , exceeds a specified value, . The
values of initial (in-situ) void ratio, , and final void ratio, , are provided by
the user. The current value of void ratio is calculated by the code from the
known volumetric strain, (positive for extension), initial (in-situ) volumetric
strain, (usually zero), and initial void ratio, , using
(37)
so that = 0 if .
FLAC3D 6.0
Choice of Material Parameters
By default, the basic CYSoil model includes the following features and
functionalities, as described above:
• A frictional and cohesive strain-hardening Mohr-Coulomb shear envelope

together with: 1) a built-in friction-hardening law to reproduce the
hyperbolic stress-strain law behavior observed in drained triaxial tests;
and 2) a built-in dilation hardening/softening law to model shear-
induced volume changes whereby the soil contracts under small shear
strains and dilates under larger strains.
• An elliptic volumetric cap with a ratio of axes defined by a shape
parameter, which can be chosen to match a known value of lateral earth
pressure, in uniaxial compression, combined with a built-in cap
hardening law to capture the volumetric power-law behavior observed in
isotropic compaction tests.
General guidelines to select the properties for the basic model are provided
below.
Initialization of , and
The “hardening curve” and ratio, , of elastic bulk modulus to plastic bulk
modulus are volumetric properties that may be derived from the results of an
isotropic compression test. The laboratory curve, including a small unloading
excursion, is sketched in Figure 2. The parameter can be estimated from the
graph by taking the ratio of the two quantities and measured from the
unloading excursion (see Equation (17)). The two additional volumetric model
parameters and appearing in Equation (10) can be obtained by fitting this
equation to the laboratory curve (see sketch in Figure 2). To facilitate the curve
fitting process, a first estimate of can be calculated from Equation (9), using
a starting value of = 0.5 and the quantities and measured on the graph
close to the unloading point. Some iterations can then be taken in which and
are adjusted to obtain a better fit to the lab curve. In turn, the parameter
can be estimated (from known values of and ) using Equation (21).
Alternatively, the model parameters and can be estimated from published

values of parameters and from the plastic-hardening (PH) model
implementation in FLAC3D (or any equivalent mode, e.g., Hardening Soil model
implementation in Plaxis) using the following relationships.
At , Equation (20) gives:
FLAC3D 6.0
(38)
On the other hand, by definition:
(39)
we then obtain
(40)
The following relationship can be inferred from the definition of in an

oedometer setting.
(41)
where the relationship between and in Equation (21) has been used.
After substitution of Equation (40) into Equation (41) and some manipulations,
we obtain
(42)
Using , the multiplier is obtained as
(43)
Also, substitute of Equation (43) into Equation (40) gives
(44)
Equations (43) through (44) can be used as an approximation to relate the plastic
hardening model parameters to CYSoil model properties. Note, however, that
these two models differ in formulations; therefore, simulation results using the
two models cannot be expected to be identical.
Initialization of , , , and
These properties are the same as those in the Mohr-Coulomb model. The
properties of ultimate friction, cohesion, and dilation can be derived from
standard triaxial tests carried out at a minimum of two different confinements
FLAC3D 6.0
until the ultimate yield state has been achieved (see Figure 5). The tensile
strength can be determined from a tensile test carried out to sample failure.
Alternatively, the properties for standard soils can be found from the literature.
Figure 5: Shear stress versus normal stress (left) and volumetric strain versus
axial strain (right).
Initialization of the cap pressure
The cap pressure should be initialized to a value that is consistent with the in-
situ stress and takes into account volumetric soil over-consolidation, as
applicable. The initial cap pressure is derived from the yield criterion = 0 (see
Equation (7)):
(45)
where the over-consolidation ratio, , is an amplification factor larger or

equal to 1.
The code automatically assigns an initial value of the evolution parameter, ,

derived from Equation (24) that is consistent with the initial cap pressure.
Initialization of mobilized friction angle
FLAC3D 6.0
The mobilized friction angle is an important parameter that should be initialized

appropriately to be consistent with the in-situ stresses and accounting for shear
over-consolidation, as applicable. The following expression, consistent with
Equation (3), can be used to determine the mobilized friction angle for a
normally shear-consolidated material:
(46)
where and are the minimum and maximum principal effective stresses in-
situ, and is the cohesion. Equation (46) follows from the geometrical
consideration of Mohr circles presented in Figure 1. The initial state of an over-
consolidated soil is prescribed by specifying an initial value of the mobilized
friction property that is larger than the normally consolidated value .
Initialization of friction hardening parameter
By default, the friction hardening parameter is assumed to be zero under in-

situ conditions. Alternatively, the user can choose to exercise the modulus
degradation curve starting with a value of that is larger than zero (see
discussion on friction hardening). The following expression can be used to
initialize the value of plastic shear strain, consistent with the initial value of
mobilized friction, (see Equation (31)):
(47)
Alternative Features
User-defined hardening/softening laws
One flexible feature of the model is the capability to substitute alternative user-
defined hardening/softening laws for the built-in laws, which are communicated
to the model by means of tables. Up to five tables can be declared. Tables may
specify mobilized friction, dilation, and cohesion in terms of plastic shear strain
measure, ; tensile strength in terms of tensile plastic strain, ; and volumetric
cap pressure in terms of cap plastic volumetric strain, . When a table is
declared for a specific model property (e.g., friction, dilation, cohesion, tensile
strength, or cap-pressure), the associated user-defined law takes precedence
over the corresponding built-in law.
Duncan-Chang type model
FLAC3D 6.0
For geotechnical applications that do not require taking into account substantial
plastic volumetric strains, the volumetric cap is usually ignored. On the other
hand, it is still important to consider the dependency of the material elastic
moduli on the initial effective stresses. This type of problem is often analyzed
with the use of the Duncan and Chang hyperbolic soil model (1970).
To account for such behavior in the CYSoil model, the user should specify not to
use the volumetric cap by setting flag-cap = 0 (the default value). Also, the user
should assign to a (positive) zone-based measure of initial confinement (e.g.,
). The code will then use the value of in lieu of in Equations (19)
and (20) for bulk and shear moduli.
The emerging behavior of the model should be similar to that of a Duncan-Chang

model (in the case that dilation is fixed at zero, using flag-cap = 1), or that of the
CHSoil model in particular.
Constant friction angle
It is possible to use a constant friction angle as an alternative to the built-in

friction hardening logic in the CYSoil model. To select this option, the user needs
to set flag-shear = 1 (the default is 0) and the input value of the friction
property to the desired friction angle.
Dilation alternatives
The built-in dilation logic is used when flag-dilation = 0 (the default value).
There are two alternative options: 1) constant dilation ( flag-dilation = 1), in
which case the input value of the dilation property is used; and 2) Rowe’s
Dilation law (flag-dilation = 2), in which case is not internally derived from
and , but an input value that is set using friction-critical.
The implementation procedure for the CYSoil model follows the general
implementation of the double-yield model. At each timestep, new stresses are
computed using the current values of the model properties. In this process, an
elastic guess, , is first computed by adding to the old stress components,
increments calculated by application of Hooke’s law to the total strain increment
for the step. Principal stresses , and corresponding principal directions,
are calculated and ordered. If these stresses violate the yield criteria for shear,
volume (cap), and/or tension, corrections are applied to the elastic guess to give
FLAC3D 6.0
the new stress state. The stress tensor components in the system of reference
axes are then calculated from the principal values by assuming that the principal
directions have not been affected by the occurrence of a plastic correction.
The evolution parameters , , and for shear, cap, and tensile yielding,
respectively, are calculated independently by the code. The parameter for shear
yielding is the (accumulated) plastic shear strain, ; the plastic shear strain
increment is defined as
(48)
where and = 1,3 are the principal plastic shear

strain increments.
The evolution parameter for cap yielding is the modulus of plastic volumetric
strain, , and its increment is defined as
(49)
where = 1,3 are the principal plastic strain increments from yielding on the
cap.
The tensile hardening parameter measures the accumulated tensile plastic

strain; its increment is defined as
(50)
where is the increment of tensile plastic strain in the direction of the major
principal stress (recall that tensile stresses are positive).
The components of the plastic strain increment for each failure mode are
evaluated from expressions similar to those derived for the double-yield model.
Zone hardening increments are then calculated from the relevant (surface
averaged) plastic strain increments for all triangles involved in the zone. After
this, new zone properties (e.g., mobilized friction, cohesion, dilation, updated
cap pressure) are calculated and stored for use in the next step.
The hardening or softening lags one timestep behind the corresponding plastic
deformation. In an explicit code, this error is small because the steps are small.
Several simple tests are presented below to illustrate the CYSoil model features
and capabilities.
FLAC3D 6.0
cap-yield Model Properties
properties of the cap-yield (CYSoil) model.
cohesion f
ultimate cohesion, . The default is
0.0.
dilation f
ultimate dilation angle, . The
default is 0.0.
friction f
ultimate friction angle, . If the
specified value is less than 0.1
degree, it will be limited to 0.1
degree for numerical stability.
friction-mobilized f
mobilized friction angle, , which
should be initialized no less than the
value determined by (see this
equation) via a command or FISH
function.
pressure-cap f
cap pressure, , which can be
initialized via a command or a FISH
function.
pressure-initial f
initial mean effective stress, .
Required if pressure-cap is not
specified. It can be initialized via a
command or a FISH function.
FLAC3D 6.0
reference pressure, . A non-zero
value should be specified based on
the unit of stress/pressure adopted
in the model.
poisson f
current Poisson’s ratio, . The
default is 0.2.
shear-reference f
dimensionless elastic shear modulus
reference value, . A non-zero
value should be specified.
tension f
alpha f [a]
dimensionless cap yielding surface
parameter, . The default is 1.0.
beta f [a]
dimensionless calibration factor, .
The default is 1.0.
dilation-mobilized f [a]
current mobilized dilation angle, .
It is read-only, except for table-
dilation ǂ 0.
exponent f [a]
exponent for the pressure-
dependent elastic moduli, . The
default is 0.5, and the upper limit is
0.99.
failure-ratio f [a]
failure ratio, . The default is 0.9.
flag-brittle b [a]
default is false.
FLAC3D 6.0
flag-cap i [a]
= 0 (default), no cap. The elastic
moduli are functions of the mean
effective stress, and
pressure_initial should be specified.
= 1, with cap. The elastic moduli are

functions of the cap pressure and
pressure_cap should be specified via
a command or a FISH function, or
calculated internally otherwise, see
information on pressure_cap.
flag-dilation i [a]
= 0 (default), the built-in Rowe
dilation rule is used.
= 1, ≡ .
= 2, should be specified and is

calculated from this equation.
flag-shear i [a]
= 0 (default), the built-in shear
hardening law is used.
= 1, ≡ .
friction-0 f [a]
initial mobilized friction angle,
associated with zero plastic shear
strain. By default, = if is
initialized to 0.0; otherwise, = 0.0
if not specified.
friction-critical f [a]
constant . The default value is
calculated by this equation. Required
for specification only if flag-
dilation = 2.
over-consolidation-ratio f [a]
over-consolidation ratio, . The
default is 1.0.
FLAC3D 6.0
multiplier f [a]
multiplier on current plastic cap
modulus to give elastic bulk and
shear moduli, . The default is 5.0
with cap and 0.0 without cap.
shear-maximum f [a]
maximum (upper-bound limit) of
the elastic shear modulus, . The
default internally calculated value is
, where is the initial
when the model is first time set up.
shear-minimum f [a]
minimum (lower-bound limit) of the
elastic shear modulus, . The
default internally calculated value is
, where is the initial
when the model is first time set up.
strain-shear-plastic f [a]
accumulated plastic shear strain, .
By default, it is initialized internally.
strain-tensile-plastic f [a]
The default initial value is 0.0.
strain-volumetric-plastic f [a]
accumulated plastic volumetric
strain, . By default, it is initialized
internally.
table-pressure-cap i [a]
number of table relating cap
pressure to plastic volume strain.
The default is 0.
FLAC3D 6.0
number of table relating mobilized
dilation angle to plastic shear strain.
The default is 0.
friction angle to plastic shear strain.
The default is 0.
table-tension i [a]
number of table relating tensile
strength to plastic tensile strain. The
default is 0.
void-initial f [a]
initial void ratio, . The default is
1.0.
void-maximum f [a]
allowable maximum void ratio, .
The default is 999.0, a virtual value
to bypass the dilatancy cut-off.
bulk f [r]
current elastic bulk modulus
pressure-effective-cy f [r]
mean effective stress,
shear f [r]
current elastic shear modulus
stress-deviatoric-cy f [r]
deviatoric stress,
void f [r]
current void ratio,
young f [r]
Notes:
• One parameter between and must be specified.
FLAC3D 6.0
• The void related parameters, and , are for

dilatancy cut-off only. If dilatancy cut-off is not taken
into account, these parameters are not required for input
(default values are used).
at the same time.
Footnotes
FLAC3D 6.0
Simplified Cap-Yield (CHSoil) Model 327
Simplified Cap-Yield (CHSoil) Model

As discussed previously in the CYSoil model, when soils are subjected to
deviatoric loading, they usually exhibit a decrease in stiffness accompanied by
irreversible deformation. The well-known Duncan and Chang model (1970) is
commonly used to simulate this hyperbolic stress-strain behavior. The Duncan
and Chang model is relatively easy to use. However, as noted in in the CYSoil
model, this model has several drawbacks. In addition, hyperbolic relations that
rely on nonlinear elasticity are known to have significant limitations (Duncan et
al. 1980), including: 1) the relations are applicable prior to failure, but may
produce unrealistic behavior of soils at and after failure; 2) the hyperbolic
relations do not include volume change resulting from change in shear stress
(the implied dilation relies on the Poisson’s ratio), and thus may not be able to
predict deformations in dilatant soils accurately, such as dense sands under low
confining pressure; and 3) the relations predict an isotropic behavior in the
plane, which is not always realistic (soil strengths in triaxial compression and
extension usually differ in magnitude).
A simplified version of the CYSoil model, named the CHSoil model, is provided as
an alternative to the Duncan and Chang model that does not have the drawbacks
of this model. The CHSoil model is derived from the same strain hardening/
softening logic that exists in the CYSoil model, and therefore can provide a
realistic stress-strain relation at failure and post-failure. The CHSoil model
provides built-in features and does not have a volumetric cap.
The CHSoil model has three specific features:
1. a built-in friction hardening law that uses hyperbolic model

parameters as direct input;
2. a frictional Mohr-Coulomb shear envelope; and
3. two built-in dilation laws — one is based on Rowe’s stress

dilatancy theory (Rowe 1962), and the other is a user-defined
dilation hardening/softening law.
FLAC3D 6.0
The unloading behavior of the CHSoil model is elastic. Reloading is elastic up to

the outermost yield envelope reached previously. Note that in its present form,
the CHSoil model is not intended to simulate cyclic loading.
The elastic behavior of the CHSoil model is expressed using Hooke’s law. The
incremental expression of the law in terms of principal stress and strain has the
form
(1)
where , , and and are the tangent elastic

bulk and shear modulus, respectively.
In the CHSoil model, the elastic shear modulus depends on the initial value of
mean effective stress, . The mean effective stress is specified by the user,
using, for example, the expression
(2)
The variation of with is represented by the equation (see, for example,

Byrne et al. 2003)
(3)
The parameter is the reference pressure, and is the shear modulus

number ( is the value of the tangent elastic shear modulus at reference
pressure). is a constant modulus exponent ( ). Also, the tangent elastic
bulk modulus is described by the relation
(4)
The parameter is the bulk modulus number, the product ( is the

value of the tangent elastic bulk modulus at reference pressure), and is a
constant modulus exponent ( ).
FLAC3D 6.0
Some useful relations between , , Young’s modulus number, , and

Poisson’s ratio at reference pressure, , are listed for reference:
(5)
(6)
The user can specify either and or and as input properties for
the model. If and are specified, and are calculated internally
from Equation (5). and the resulting values are used with Equations (3) and (4).
Values of Poisson’s ratio are restricted to positive values smaller than 0.49.
Accordingly, upper and lower bounds for are specified internally as
(7)
As noted above, the initial mean effective pressure, , must be provided as an

input property for the model. (This value may be evaluated using FISH and stored
in the associated zone property offset.) The value of is used to calculate the
tangent elastic shear and bulk moduli, according to Equations (3) and (4). and
stay constant in this implementation; the moduli are not updated
automatically in terms of mean effective pressure.
Shear yielding is defined by a Mohr-Coulomb criterion. The yield envelope is

expressed as
(8)
where is the cohesion, is the mobilized friction angle, and, by definition,
(9)
The potential function is nonassociated and has the form
(10)
where
(11)
FLAC3D 6.0
and is the mobilized dilation angle. Several different laws are available in the
literature to characterize .
By default, can be given in terms of plastic shear strain via a user-defined

table of versus . The evolution parameter for shear yielding, , is defined
incrementally by
(12)
where = 1,3 are the principal, deviatoric, plastic shear-strain increments.
The model parameter flag-dilation must be set to zero to activate this option. If
flag-dilation = 0 and no table is provided, it is then assumed that dilation is
equal to the input value for ultimate dilation property set with dilation. Note
that dilation is not used if a dilation table is provided.
Two built-in dilation laws are also available. A law based on Rowe’s (1962)
dilatancy theory is used if flag-dilation = 1. This is the default setting.
Alternatively, a simple step function can be used in which is zero for ,
where is a constant specified by the user, and is equal to the ultimate
dilation value (set by dilation) for values of mobilized friction larger than .
The model property flag-dilation must be set to 2 to activate this option. The
additional constraint that cannot exceed (to prevent unwanted generation
of energy from taking place) is enforced internally by the code.
The tensile yield function is the same as that used for the CYSoil model, and is of
the form
(13)
The tensile strength, , is given in terms of the plastic tensile-strain measure,

, and input by means of a user-defined table. If no table is provided, it is
assumed that tensile strength is constant and equal to the input value of the
tensile strength property.
The evolution parameter for tensile yielding is the modulus of plastic tensile
strain, . The increment of plastic tensile strain is defined as
(14)
FLAC3D 6.0
where is the increment of tensile plastic strain in the direction of the major
principal stress (recall that tensile stresses are positive).
Friction Hardening
For most soils, the plot of deviatoric stress versus axial strain obtained in a
drained triaxial test can be approximated by a hyperbola. The CHSoil model
incorporates a friction strain-hardening law to capture this behavior. In this
formulation, the mobilized friction angle, , is given in terms of plastic shear
strain measure, , by means of the following differential law, similar to the one
implemented by Byrne et al. (2003) in their UBCSAND liquefaction model:
(15)
where the plastic shear modulus, , is given by
(16)
In this formula, is the elastic tangent shear modulus, is the ultimate

friction angle, and (the failure ratio) is a constant smaller than 1 (0.9 in most
cases) used to assign a lower bound for .
According to Equation (14), the elastic tangent shear modulus is a function of ,
(17)
After substitution of Equation (17) into Equation (16), the resulting expression in
Equation (15), and rearranging terms, we obtain
(18)
Using = 0 at = 0, integration of this equation gives
(19)
Solving for , we obtain
FLAC3D 6.0
(20)
This expression is used in the CHSoil model to calibrate mobilized friction in

terms of plastic shear strain. The use of this hardening law for modeling primary
loading in a triaxial test produces a hyperbolic-type curve of deviatoric stress
versus axial strain.
Over-Consolidation
The initial state of a normally consolidated soil or an over-consolidated soil is

prescribed by specifying an initial value to the friction property that is equal to or
larger than the normally consolidated value, . For normal consolidation,
Equations (15) and (19) give
(21)
The material behavior is considered to be elastic for stress points below the
current yield envelope.
The initial value of friction must be smaller than the ultimate value, . (An
upper bound equal to is set automatically by the model.) An initial value of the
evolution parameter, , that is consistent with the specified initial value of
mobilized friction angle is also assigned automatically.
Shear-Induced Compaction and Dilation
A certain amount of irrecoverable volumetric strain, , is expected to take place

as a result of soil shearing. Also, under small monotonic shear strains, there is a
tendency for the soil skeleton to contract due to grain rearrangements. For larger
shear strains, the soil skeleton may dilate if the soil is dense, as a result of grains
riding over each other. A dilation strain-hardening law is incorporated in the
logic to model this behavior.
The shear-hardening flow rule implemented in the CHSoil model has the form
(22)
where is the plastic volumetric strain increment, is the plastic shear

strain increment, and is the (mobilized) dilation angle.
FLAC3D 6.0
One possible way to characterize is to adopt an equation based on Rowe’s

stress-dilatancy theory (Rowe 1962). According to this theory, there is a
(constant-volume) friction angle, , below which the material contracts (i.e.,
for ), while for higher stress ratios (i.e., for ), the material
dilates. The equation has the form
(23)
where
(24)
and and are ultimate (known) values of friction and dilation, respectively.
Dilation is evaluated in terms of plastic shear strain based on the last two
equations and the assumed relation between and reported in Equation (19).
Rowe’s dilation law is selected by setting the material property flag-dilation = 1
(default value). A simple dilation law also can be used, in which = 0 for
, and for ; the option is activated by specifying flag-
dilation = 2. Finally, the user may choose to define an alternative dilation law by
creating an input table of dilation values versus plastic shear strain. The model
property flag-dilation must be set to zero to activate this option.
Correlation Between CHSoil Properties and Duncan and Chang Properties
A parallel can be drawn, for particular cases, between properties used in the
hyperbolic stress-strain relations reported in Duncan et al. (1980) for the Duncan
and Chang model and the properties of the CHSoil model. The correlations
depend on property input; they are listed in the following table.
Table 1: Property Correlation for Input of and
Model = No Dilation
Duncan and Chang & &
CHSoil & &
&
CHSoil & &
&
FLAC3D 6.0
Calibration to Triaxial Test Results on Nevada Sand
Results of drained triaxial compression tests at constant mean effective stress, ,

on Nevada sand are used to provide an example of the methodology that can be
used to calibrate the properties of the CHSoil model. The test results are listed in
data files “40-40.txt”, “40-80.txt”, and “40-160.txt” and correspond to values
of the mean stress at 40 kPa, 80 kPa, and 160 kPa, respectively. The relative
density of the sand for these tests is 40%. The test results (see the drained
triaxial compression tests for more details) consist of three sets of data:
1. deviatoric stress versus mean effective stress;
2. deviatoric stress versus axial strain (up to 25% strain); and
3. volumetric strain versus axial strain (up to 25% strain).
For this calibration exercise, we consider axial strains up to 5%, because strains
are not expected to develop above this level for the intended (static, drained)
application. The Nevada sand appears to be purely frictional in character. Also,
two main features characterize the data: 1) hyperbolic behavior is observed in the
plot of deviatoric stress versus axial strain; and 2) bilinear dilatant behavior is
exhibited by the plot of volumetric strain versus axial strain. The CHSoil model
capabilities, including friction hardening and variable dilation, are used to
simulate these features.
The elastic tangent shear and bulk moduli are functions of the initial mean
effective stress according to Equations (3) through (5). For friction hardening, we
use the built-in law, Equation (19). The dilation law is bilinear; its value is zero
below the stress ratio , and a constant, , above it:
(25)
The failure ratio, , is 0.99 for this exercise. Eight properties must be defined:
FLAC3D 6.0
Calibration of the model properties is done in two steps. First, using theoretical
considerations and values recorded in the literature, we derive a first estimate for
the property values. Second, we improve on the estimates by modeling the
triaxial experiments numerically and matching the results obtained in the
laboratory.
First Estimates — The value of ultimate friction is derived from a linear fit to plots
of deviatoric stress, , versus mean effective stress, , obtained from the
laboratory tests. The linear fit to maximum at given provides a value for the
maximum stress ratio, :
(26)
From the definition of the purely frictional Mohr-Coulomb criterion, we obtain

the relation
(27)
from which the ultimate friction angle, , can be derived. The estimated value
for = 40% sand is = 34°.
The value of ultimate dilation angle is derived from the slope, , of a linear fit to
the laboratory curves of minus volumetric strain versus axial strain. To relate to
, we use, as a first approximation, the expression for bilinear idealization of
triaxial stress results provided by Vermeer and de Borst (1984):
(28)
The first estimate for is 8.2° for = 40% sand.
To estimate the parameter for a given , we first derive the maximum value
of axial strain, , at which the volumetric strain is negligible from the laboratory
plots of volumetric strain versus axial strain at a given . From the knowledge of
axial strain, , we estimate from the laboratory plot of deviatoric stress versus
axial strain at that , and then the ratio . Three values of the ratio are
available (one at each ). The mean value of is used to calculate the
corresponding friction angle using Equation (27), where now is replaced by
, and by the average (see Equation (26)). The estimate is = 27.7° for =
40%.
FLAC3D 6.0
The first guess for is taken as 1200 for = 40% at atmospheric pressure;
this value is equivalent to that used in the triaxial numerical experiment for
dense sand with the cap-yield (CYSoil) model.
The value of for this exercise is arbitrarily selected as 0.35. Also, we select
. The elastic constant, , is estimated by matching the initial slopes
of versus axial strain curves obtained in similar triaxial tests (under constant
mean stress) performed numerically and in the laboratory. These estimates may
not be very accurate; more robust estimates for the elastic constants can be
obtained from laboratory results of small unloading-reloading excursions. Such
results were not available for this example.
cap-yield-simplified model properties
properties of the simplified Cap-Yield (CHSoil) model.
bulk-reference f
dimensionless bulk modulus
reference value,
cohesion f
ultimate cohesion,
dilation f
ultimate dilation angle,
friction f
ultimate friction angle,
friction-mobilized f
mobilized friction angle, , which
should be initialized no less than the
value determined by this equation
via a command or FISH function
poisson f
current Poisson’s ratio,
FLAC3D 6.0
pressure-initial f
initial mean effective pressure, ,
which should be initialized via a
command or FISH function
in the model.
shear-reference f
dimensionless shear modulus
reference value,
tension f
young-reference f
dimensionless Young’s modulus
reference value,
dilation-mobilized f [a]
current mobilized dilation angle, .
It is read-only, except for table-
dilation ǂ 0.
flag-brittle b [a]
default is false.
flag-dilation i [a]
= 0 for mobilized dilation angle, ,
equal to input value, dilation or a
function of plastic shear strain if
table is input with table-dilation
= 1 (default) for mobilized dilation

angle, , characterized by Rowe’s
stress-dilatancy theory
= 2 for mobilized dilation angle, =

0 if < , and = , if
FLAC3D 6.0
exponent-bulk f [a]
bulk modulus exponent, . The
default is 0.5.
exponent-shear f [a]
shear modulus exponent, . The
default is 0.5.
friction-0 f [a]
initial mobilized fiction angle,
associated with zero plastic shear
strain. By default, = if strain-
shear-plastic is initialized to 0;
otherwise, = 0 if not specified.
friction-critical f [a]
constant used in dilation law (flag-
dilation = 1 or 2),
strain-shear-plastic f [a]
By default, it is initialized internally.
strain-tensile-plastic f [a]
accumulated plastic tensile strain,
. The default initial value is 0.
failure-ratio f [a]
table-cohesion f [a]
number of table relating cohesion, ,
to plastic shear strain
dilation angle to plastic shear strain
table-tension f [a]
tensile strength to plastic tensile
strain
FLAC3D 6.0
bulk f [r]
shear f [r]
current elastic shear modulus,
young f [r]
Notes:
elasticity: 1) and , 2) and . When choosing
the latter, must be assigned in advance of .
at the same time.
Footnotes
FLAC3D 6.0
FLAC3D 6.0
Plastic-Hardening Model 341
Plastic-Hardening Model
The Plastic-Hardening (PH) model is a shear and volumetric hardening
constitutive model for the simulation of soil behavior. When subjected to
deviatoric loading (e.g., during a conventional drained triaxial test), soils usually
exhibit a decrease in stiffness, accompanied by irreversible deformation. In most
cases, the plot of deviatoric stress versus axial strain obtained in a drained
triaxial test may be approximated by a hyperbola. This feature was discussed by
Duncan and Chang (1970) in their well-known “hyperbolic-soil” model, which is
formulated as a non-linear elastic model. The PH model is formulated within the
framework of hardening plasticity (Schanz et al. 1999) allowing the removal of
the main drawbacks of the original non-linear elastic model formulation (e.g.,
detection of loading/unloading pattern, non-physical bulk modulus).
The main features of the PH model are:
1. hyperbolic stress-strain relationship during axial drained compression;
2. plastic strain in mobilizing friction (shear hardening);
3. plastic strain in primary compression (volumetric hardening);
4. stress-dependent elastic stiffness according to a power law;
5. elastic unloading/reloading compared to virgin loading;
6. memory of pre-consolidation stress; and
7. Mohr-Coulomb failure criterion.
The model is straightforward to calibrate using either conventional lab tests or

in-situ tests. It is well established for soil-structure interaction problems,
excavations, tunneling, and settlements analysis, among many other
applications.
FLAC3D 6.0
In this section, principal stress, , and strain, , i = 1,3, components are positive
in tension. Also, all stresses are effective by default if not explicitly stated. The
principal effective stresses are with the order , and is the most
compressive stress.
The PH model adopts hypo-elasticity for the description of elastic behavior,
(1)
(2)
where is the mean pressure defined as , is the volumetric elastic

strain defined as , and are the deviatoric
stress tensor and deviatoric elastic strain tensor, respectively. is defined as
if and if . and are the elastic bulk and shear moduli,
which can be derived from the elastic unloading-reloading Young’s modulus,
, and the unloading-reloading Poisson’s ratio, , are obtained using the
relations:
(3)
(4)
In the PH model, the Poisson’s ratio is assumed to be a constant with a typical

value of 0.2 (if otherwise not provided at input).
The Young’s modulus is a stress-dependent parameter:
(5)
where
(6)
and , , , , and are user-defined constant parameters. is the

reference unloading-reloading stiffness modulus at the reference pressure .
The current unloading-reloading stiffness modulus depends on the
maximum (minimum compressive) principal stress, , the cohesion, , and the
FLAC3D 6.0
ultimate friction angle, , as well as the power, . For clays, is usually close to
1. For sands, is usually between 0.4 and 0.9. The default cut-off factor is
0.1.
The PH model also employs an additional stiffness measure, , which defines

the initial slope of the hyperbolic stress-strain curve (see Equation (7) and
Figure 1). Parameter obeys the following power law:
(7)
Here is a material parameter, which could be estimated from a set of triaxial

compression tests with various cell stresses.
The shear yield function determining the onset and development of shear
hardening is defined as
(8)
where is a shear hardening parameter (one of the internal variables) defined

in (13), , , and is given as
(9)
where
(10)
The failure ratio has a value smaller than 1 (typically, = 0.9 is

used). Note that the ultimate deviatoric stress, , is defined as
(11)
is consistent with the Mohr-Coulomb failure criterion (Figure 1).
For a standard drained triaxial test, the connection between the axial (vertical
compressional) strain, , and deviatoric stress, , can be described by a
hyperbolic relation:
(12)
FLAC3D 6.0
Equation (12) is graphically represented in Figure 1 with the cut-off at .
Figure 1: Hyperbolic stress-strain relation in primary shear loading.
The shear hardening parameter is defined so that its incremental form is:
(13)
Due to the increase of , the shear yield surface will expand up to the ultimate
surface, which is defined by the conventional Mohr-Coulomb failure criterion. In
case the material experiences ultimate shear failure, the conventional Mohr-
Coulomb failure criterion applies.
The PH model uses the following flow rule between volumetric and shear plastic
strains:
(14)
FLAC3D 6.0
where is the mobilized dilation angle, which should be smaller or equal to

the user-defined ultimate dilation angle . The mobilized dilation angle is based
on Rowe’s dilation law (1962):
(15)
(16)
where the parameter is a contraction scale factor, with the allowable range of
0 to 0.25 and default value of 0.0 for most soils if the contraction is not
considered. The critical state friction angle is defined as
(17)
The mobilized friction is defined in terms of the current stress state
(18)
A non-associated flow rule consistent with the Mohr-Coulomb yield criterion is

used in the model. The shear potential function is defined as
(19)
where and .
In order to avoid over-dilatancy when the soil reaches its critical void state at
, the dilation angle needs a minor modification. One way proposed by Schanz
et al. (1999) is to set a cut-off rule, so that
(20)
Here we introduce a smoothing technique to avoid a sudden change of dilation

angle:
(21)
The dilation rules with cut-off, smoothing technique, and without dilation cut-
off are schematically compared in Figure 2.
FLAC3D 6.0
Figure 2: Volumetric strain curve for a standard triaxial compression test with
dilation cut-off and smoothing.
Volumetric Cap Criterion and Flow Rule
The volumetric (cap) yield function is defined as
(22)
where is a constant derived internally from other material parameters based on

a virtual (numerical) oedometer test, is a shear stress measure defined as
, and . The initial value of
hardening parameter , which denotes the preconsolidation pressure, can be
determined using the initial stress state and over-consolidation ratio, , so
that
(23)
FLAC3D 6.0
If has a large value, the model behaves as if there is no cap present. The
associated flow rule is adopted for volumetric hardening, which means that the
potential volumetric function is assumed to be the same as the volumetric yield
function (see Equation (24)).
The volumetric hardening parameter (one of the internal variables) is defined

incrementally as
(24)
where is the volumetric plastic strain increment.
Evolution of the hardening parameter, , is given by the relation:
(25)
where is a constant parameter that can be derived internally from other

material parameters. Instead of taking and as input material parameters,
another two parameters, and are specified as input. Parameter
denotes normal consolidation coefficient and stands for the tangent
oedometer stiffness at the reference pressure . If is not provided by the
user, it is taken as (default value). The PH model reserves the
flexibility for the user to input user-defined and instead of internal
calculation for and . A typical value for is between 0.6 and 2.0 for most
soils. Parameter is calculated as , where is a correction factor with
a typical value of , , , and
. Specifically, = 1 if at isotropic compression state.
The model checks for the tension failure condition. The tension failure and
potential functions are
(26)
where is the tension limit. By default, is zero and the user can provide a
value with an upper limit of . The model does not consider tension
hardening.
FLAC3D 6.0
Plastic Corrections
In the implementation of the PH model, elastic trial stresses, , are first

computed by adding to the old stresses, , elastic increments computed using
the total strain increment for the step (see Eq. (1.395)). During this step, the
moduli and mobilized dilation angle are assumed to be constants (for simplicity)
and calculated based on the old stress components. All stresses are assumed to be
effective.
The formulation for the trial stresses (denoted by a superscript ) in principal

axes are:
(27)
where , , and are the tangent elastic bulk and

shear modulus, respectively, and ( ) is the set of principal strain
increments.
The trial internal variables, and , take their values from the previous step. If
the hardening yield criteria (Equation (8) and/or (22)) are exceeded, both the
stress components and internal variables will be corrected as described below.
First, consider the situation when only shear hardening occurs, . The
plastic strain increment is oriented in the direction of the gradient of the
potential function in the principal stress space. Using Equation (19), we get
(28)
The shear hardening parameter increment is obtained from Equation (13),
(29)
Thus, the corrected (new) stress components become:
FLAC3D 6.0
(30)
and the corrected internal variable is
(31)
Using Equation (30) and definitions and from Equation (19), updated shear
stress and parameter (see Equation (9)) are calculated as
(32)
(33)
After substituting the corrected parameters , , into Equation (8), parameter

can be obtained by solving a nonlinear equation = 0 using an iteration
method. Note that in Equation (31) uses the current value of updated in each
iteration step, which implies that the correction algorithm is semi-implicit.
If the trial shear stress exceeds the Mohr-Coulomb shear failure surface, the
corrections used in Mohr-Coulomb model implementation apply.
Now consider the case when the elastic guess (Equation (22)) exceeds the
volumetric yield criterion, , and the volumetric hardening occurs. The
plastic strain increments are related to the gradient of the potential function in
the stress space as
(34)
Evolution of the hardening parameter is given by the relation:
(35)
In the above formulation, the plastic strain and volumetric hardening parameter
increment are related to the current stress measurement ( or ), so the
correction algorithm again uses a semi-implicit approach.
The corrected stress components are obtained by analogy with shear hardening,
FLAC3D 6.0
(36)
Noting that and , and using

Equation (36) we get
(37)
(38)
where . After substituting the corrected and into the

yield function , (Equation (22)), parameter can be obtained by solving the
nonlinear equation = 0 for the smallest root using an iterative method.
Finally,we consider the situation when the elastic trial stresses exceed both the
shear yield hardening surface and volumetric yield hardening surface. In this
case, the plastic strain increment is related to the partial derivatives of the
potential functions as follows:
(39)
Parameters can can be found by using an iterative method similar to that

adopted in the shear or volumetric hardening correction technique. If the trial
stress is out of both the volumetric hardening surface and Mohr-Coulomb failure
surface, the Mohr-Coulomb failure function should be used instead of the shear
hardening function.
In the initialization step, the initial stress, evolution parameters can and
strain increment are defined for the zone and are assumed to be constant during
this step. For simplicity, the stiffness moduli and dilation, which are dependent
on the current stress and evolution parameters, are also assumed to be constant
in the zone during this step. The trial elastic stresses and the possible stress
FLAC3D 6.0
corrections are defined at the sub-zone level. The trial stress increments obey
the linear elastic (Hooke’s) law. If the trial stresses violate the tension limit, the
tension failure procedure is called. The next step is to check whether the
tension-corrected stress is in the tension or compression side. If the corrected
stress is tensile, volumetric hardening will not apply. In either side, if the stress
is out of the shear failure or shear yield surface, the shear failure or hardening
procedure will be called, respectively. In the compression side, if the stress is out
of the volumetric yield surface, the volumetric hardening procedure will be
called. In particular, if the stress is also out of the shear failure or shear yield
surface, the mixed procedure with both the volumetric hardening and shear
failure/hardening corrections will be called. To ensure the averaged stress in the
zone level is within the tension limit, a second tension failure procedure is called
if any volumetric hardening or shear hardening/failure occurs. After all sub-
zones complete the stress check for tension and shear failure, shear and
volumetric hardening criteria, the internal variables, stiffness moduli and
dilation are updated based on the zone-averaged stresses.
Comparison of the PH model and CYSoil model
The PH and CYSoil models share many similar features. For example, both
models have stress-dependent moduli, both have primary shear and volumetric
hardening, both employ a hyperbolic relation for the deviatoric stress versus
axial strain during triaxial compression, the volumetric yield and potential
functions are the same in both models, etc. However, there are some differences
between the models as outlined below.
• For primary shear, the CYSoil model uses a Mohr-Coulomb type yield
function associated with the mobilized friction angle, while the PH model
uses the yield function described by Equation (19).
• The CYmodel uses the accumulated plastic shear strain as one of the
evolution parameters ( ), while the definition for differs from plastic
shear strain in the PH model (see Equation (13)).
• The CYSoil model moduli are dependent on the mean pressure, while the
PH model moduli are dependent on the maximum principal stress
(minimum compressive).
• The PH model introduces a modulus to determine the initial slope of
the hyperbolic curve during the triaxial compression, while the CYSoil
model does not have such modulus; this implies that these two models
may have different hyperbolic curves during the drained triaxial
compression, although the ultimate shear stresses are the same.
FLAC3D 6.0
• In the CYSoil model, the parameter is user-defined, while in the PH

model, this parameter can be internally calculated using other input
parameters.
• A constant ratio between the elastic and plastic volumetric strain is
assumed during isotropic compression at any pressure in the CYSoil
model, while the PH model does not adopt this assumption. As a result,
the hardening law provided by Equation (25) is used in the PH model.
• Most parameters of the PH model can be calibrated through the
conventional triaxial and oedometer compression tests (see the next
section on calibration), while calibration of the CYSoil model parameters
is more elaborate.
• In the CYSoil model, many parameters can be input through tables.
Currently, the PH model does not have this flexibility feature.
Calibration of Material Parameters Using Laboratory Data
This section presents information on calibration techniques for the PH model

material parameters using the conventional geotechnical laboratory tests. The
calibration example uses original test results obtained from the triaxial
compression tests on Monterey sand (Lade 1972). The triaxial test consisted of
loading the specimen followed by the unloading-reloading regimes. The original
test data are provided in Figure 3 and 4. The data are based on three sets of
triaxial compression tests with confining pressures of 1.2, 0.6, and 0.3 kgf/cm2.
The initial void ratios are 0.783, 0.786, and 0.781, respectively.
FLAC3D 6.0
Figure 3: Original vs. curves for the confining pressures of 1.2, 0.6, and 0.3
kgf/cm2 of the triaxial compression tests of Monterey sand (Lade 1972).
Figure 4: Original vs. curves for confining pressures of 1.2, 0.6, and 0.3 kgf/
cm2 of the triaxial compression tests of Monterey sand (Lade 1972).
Calibration of Friction Angle, , and Cohesion, :
FLAC3D 6.0
The calibration of these two parameter has no difference from the Mohr-
Coulomb model.
To estimate values of friction angle and cohesion from triaxial test data using
the relation vs. , follow the procedure outlined
below.
1. Plot deviatoric stress vs. using triaxial compression test lab data.
2. Use a trend line to fit the Mohr-Coulomb envelope.
3. The slope of the trend line is , which follows from an

alternative form of the Mohr-Coulomb failure criterion. This determines
friction angle .
4. The intercept of the trend line is , which determines

cohesion , as the friction angle is already known from the previous
step.
Alternatively, these two parameters can be estimated from triaxial test data
using the relation vs. by following the procedure
outlined below.
1. Plot deviatoric stress vs. using triaxial compression test lab data.
2. Use a trend line to fit the Mohr-Coulomb envelope.
3. The slope of the trend line is , which follows from an alternative form
of the Mohr-Coulomb failure criterion. This determines friction angle .
4. The intercept of the trend line is , which determines cohesion , as

the friction angle is already known from the previous step.
For the Monterey sand, the slope of the trend line of the Mohr-Coulomb
envelope using the relation vs. , presented in Figure 5, is 1.403. Based on this,
the friction angle is calculated as 34.65 degrees. The intercept is approximately
zero, which implies that the cohesion is zero.
FLAC3D 6.0
Figure 5: Determination of friction angle and cohesion from the triaxial

compression test data.
Calibration of , , and :
To estimate values of , , and from triaxial test data, follow the procedure
outlined below.
1. Plot the curve of vs. using the triaxial compression test data.
2. Use a trend line to fit the data. The slope of the line is , the intercept is
, based on the relation (derived from Equation
(12) and ).
Multiple sets of triaxial compression tests with various confining pressures can
be used to produce a set of pairs of ( , ). The final can be estimated as the
averaged one, and the pairs of ( , ) determine and based on Equation
(7).
Figure 6 plots the curves of vs. using triaxial compression test data of
the Monterey sand with confining pressures 1.2, 0.6, and 0.3 kgf/cm2. The
slopes of these lines are , and the intercepts are (as strain is given in
percentage).
FLAC3D 6.0
This figure determines three pairs of ( , ), summarized in Table 1. The

average value for is 0.957. Parameter needs no calibration and its value is
assumed to be 0.1 kgf/cm2. Finally, plotting parameters vs. ,
as shown in Figure 7, and using Equation (7) allows for determination of and
(remember that cohesion = 0 for this example). The slope of the trend line
in Figure 7 is = 0.707 and the intercept determines = (4.63) = 102.5
kgf/cm2.
Table 1: Determination of , , and
-0.3 0.9558 0.002459 406.7 212.3 5.358 1.099

-0.6 0.9678 0.001286 777.6 401.3 5.995 1.792
-1.2 0.9476 0.000930 1075.3 565.8 6.338 2.485
break
FLAC3D 6.0
Figure 6: Determination of and from three sets of triaxial compression

tests with three different confining stresses.
Figure 7: Determination of and from three sets of triaxial compression

tests with three different confining stresses.
FLAC3D 6.0
Calibration of :
After the calibration of parameter , it is straightforward to obtain from

Equation (5) using the unloading-reloading modulus obtained from the
original vs. data in a triaxial compression test. The final value of can be
averaged using data for different confining pressures. For the example of
Monterey sand, is determined to be 320.0 kgf/cm2. If the unloading-
reloading moduli are not available, a value in the range of can be used
for most soils. The PH model uses a default value of if no input is provided
for .
Calibration of Dilation Angle, :
The ultimate dilation angle can be estimated as , where is

the maximum slope ( ) of the vs. curve in the triaxial compression tests.
The slope is found as: . For
this example of the Monterey sand, the dilation angles are 6 ~ 7° (depending on
confining pressure). The values are summarized in Table 2.
Calibration of and :
Parameters and can be calibrated from the oedometer tests. The ultimate
value of from an oedometer test is . When , the tangent
modulus of the vs. curve is .
Calibration of Poisson’s Ratio, :
Poisson’s ratio can be estimated from the initial slope (usually a negative
value) of the vs. curve, so that . In the PH model, the
default value is 0.2. In this example for Monetary Sand, the Poisson’s ratio is
estimated to be 0.3.
Calibration of :
The material parameter is needed if the dilation smoothing technique

(Figure 2) is required. The value of can be determined by conducting
standard laboratory tests (ASTM D4254). If the standard ASTM D4254 laboratory
test data are not available, its value can be estimated through the trial-and-error
method based on the curves of the volumetric strain vs. axial strain of the triaxial
compression tests. In this example, its value is estimated as 0.803. The default
value of = 999.0 is used when it is not provided as input, which implies that
the dilation smoothing technique will not be activated.
FLAC3D 6.0
Calibration of Other Parameters:
Such material parameters as , , , , and are known initial

parameters and should be consistent with the initial conditions. Some
parameters can be estimated through the trial-and-error method.
A summary of all material properties determined for the Monterey sand is

provided in Table 2.
Table 2: Calibrated Material Parameters for Loose Monterey Sand
Parameters = -1.2 (kgf/cm2) = -0.6 (kgf/cm2) = -0.3 (kgf/cm2)
(kgf/cm2) 102.5
(kgf/cm2) 320.0
0.707
0.957
(kgf/cm2) 0.1
0.3
(degrees) 34.65
(kgf/cm2) 0.0
0.803
0.5
(degrees) 6.1 6.4 7.0
(kgf/cm2) -1.2 -0.6 -0.3
1.0 2.0 4.0
0.783 0.786 0.781
Using these parameters, the triaxial compression tests can be simulated by the
PH model. The plots presented in Figures 8 and 9 reveal a close match between
the simulated results and laboratory test data.
break
FLAC3D 6.0
Figure 8: Deviatoric stress vs. axial strain for consolidated drained triaxial
compression tests on fine Monterey sand.
Figure 9: Volumetric strain vs. axial strain for consolidated drained triaxial
compression tests on fine Monterey sand.
FLAC3D 6.0
plastic-hardening Model Properties
properties of the plastic hardening model.
cohesion f
cohesion, . The default and cut-off
(allowable minimum) value is
.
dilation f
ultimate dilation angle in degrees, .
The default is 0.0.
exponent f
exponent for elastic moduli, .
Allowable value is .
friction f
ultimate friction angle in degrees, .
The default and cut-off (allowable
minimum) value is 0.001 degrees.
stress-1-effective f
initial minimum principal effective
stress, . See the note*.
initial median principal effective
initial maximum principal effective
in the model. Most commonly used
and highly recommended reference
pressure is the standard atmospheric
FLAC3D 6.0
pressure, e.g., 100.0 if in kPa, 1.0e5

if in Pa. or other value compatible to
the adopted stress/pressure unit.
stiffness-50-reference f
secant stiffness, , at 50% of the
ultimate deviatoric stress, , when
in a triaxial test. A non-
zero value must be specified.
stiffness-oedometer-reference f [a]
tangent stiffness, , when
in an oedometer test. By
default, if is not
specified while has been
specified.
stiffness-ur-reference f [a]
unloading-reloading stiffness, ,
when in a triaxial test.
This value must be greater than
. By default, .
coefficient-normally-consolidation f [a]
normal consolidation coefficient,
. It is not allowed to be less than
, a range 0.5 to 0.7 is
common. The default is
.
constant-alpha f [a]
dimensionless parameter in the cap
yield function, . Initialized
internally. See the note* for an
alternative option.
factor-contraction f [a]
contraction factor, , allowable
range is 0.0 to 0.25. The default is
0.0.
FLAC3D 6.0
factor-cut f [a]
cut-off factor, . The default is
0.1.
flag-initialization f [a]
flag for initialization. If set to 1, the
internal variables will be
recalculated. The default is 0.
over-consolidation-ratio f [a]
over consolidation ratio, . For
normally consolidated soil, the value
should be input as 1.0. The default is
100.0, which denotes an “apparent”
non-cap shear plastic hardening
model.
poisson f [a]
Poisson’s ratio, . The default is 0.2.
stiffness-cap-hardening f [a]
hardening modulus for cap
hardening, . Initialized internally.
See the note* for an alternative
option.
failure-ratio f [a]
tension f [a]
void-initial f [a]
initial void ratio, . If the current
void ratio is not important, or you do
not wish to activate the dilation
smoothing technique, just use the
default value. The default is 1.0.
FLAC3D 6.0
void-maximum f [a]
allowable maximum void ratio, .
The default is 999.0, which is not
physical but it denotes that the
dilation smoothing technical will
nominally not activated.
bulk f [r]
current loading-unloading bulk
modulus,
shear f [r]
current unloading-reloading shear
modulus,
plastic-hardening-shear f [r]
accumulated shear plastic hardening
parameter,
plastic-hardening-volume f [r]
accumulated volumetric plastic
hardening parameter,
pressure-cap f [r]
cap pressure, . Determined by the
current stress, , and .
void f [r]
current void ratio,
Notes:
• The parameters , , must be specified via
commands or a FISH function the first time the PH model
is assigned because it is a stress-dependent model.
Negative values denote compressive stresses.
• Some parameter combinations may be rejected due to: 1)
one or more parameters are out of predetermined limits;
2) internal parameters cannot be correctly determined; or
3) numerical instability.
FLAC3D 6.0
• If both and are specified non-zero values by the

user, the PH model will use these user-defined
parameters and , and the internal calculation for
and will not be performed.
Footnotes
FLAC3D 6.0
FLAC3D 6.0
Swell Model 367
Swell Model
This model is based on the Mohr-Coulomb model with nonassociated shear and
associated tension flow rules. The difference is that the wetting-induced
deformations are taken into account by means of coupling wetting strains with
the model state prior to wetting. For model application and modeling procedures,
see Noorany et al. (1999), Pathak et al. (2003), Rodriguez-Ortiz et al. (2003), and
Pathak (2009). The yield and potential functions, plastic flow rules, and stress
corrections are identical to those of the Mohr-Coulomb model.
Wetting Strain and Stress Corrections
The wetting-induced strains can be expressed by the following logarithmic or

linear function of total compressive stress, , in the principal swelling
direction normalized by the atmospheric pressure, :
• logarithmic function
(1)
• linear function
(2)
In the above equations, , , , and are non-negative welling properties

determined from laboratory tests, and, by convention, compressive stresses and
strains are negative. Note that the wetting strain is assumed to be isotropic in the
lateral directions ( ). Stress is the total vertical stress component in
the local axes obtained at the equilibrium state prior to wetting and is not
modified in Equations (1) and (2) during calculations.
Corresponding wetting stresses in the principal swelling directions (defined by

) are then calculated based on the incremental form of Hooke’s Law:
FLAC3D 6.0
(3)
where , , and and are bulk and shear modulus,

respectively.
Finally, the global swelling stresses are obtained through resolution of the local
stresses in the principal swelling direction into the global axes. Global swelling
stresses are then added (in increments) to the global stress components over a
specified number of steps, . At each step, the resulting stress components are
examined for failure based on the Mohr-Coulomb failure criteria. Note that the
model should be cycled to equilibrium after steps are taken.
In the implementation of the Swell model, stresses corresponding to the elastic

guess for the step are first analyzed as described in the Mohr-Coulomb model.
The corresponding swelling stresses in the local axis (principal swelling
directions) are calculated using Equation (3) and then resolved into the global
axes. The swelling stresses in the global axes are added over a specified number
of steps, . The resulting stresses are then examined for failure based on the
Mohr-Coulomb failure criteria.
Upper bounds are defined for the swelling strains in and vertical to the swelling
plane based on the swelling measured under zero vertical stress. The code checks
the wetting strains at each step and sets values to the upper bounds once they
exceed the limits.
swell Model Properties
properties of the swell model.
bulk f
cohesion f
cohesion,
FLAC3D 6.0
Swell Model 369
constant-a-1 f
swelling property,
constant-a-3 f
swelling property,
constant-c-1 f
swelling property,
constant-c-3 f
swelling property,
constant-m-1 f
swelling property,
constant-m-3 f
swelling property,
dilation f
dip f
dip angle [degrees] of the local
swelling plane
dip-direction f
dip direction [degrees] of the local
swelling plane
friction f
normal v
normal direction of the local
swelling plane, ( , , )
normal-x f
to the local swelling plane,
normal-y f
direction to the local swelling plane,
FLAC3D 6.0
normal-z f
direction to the local swelling plane,
poisson f
Poisson’s ratio,
atmospheric pressure,
shear f
tension f
young f
Young’s modulus,
flag-swell i [a]
flag of swelling function, (1 -
logarithmic; 2 - linear). The default
is 1.
flag-brittle b [a]
default is false.
number-start i [a]
number of steps over which swelling
strains are introduced, . The
default is 1.
count-swell i [r]
count of step number after swelling
starts, must be reset to zero to start
a new swelling episode
FLAC3D 6.0
Swell Model 371
stress-local-vertical f [r]
local total stress vertical stress when
swelling starts; must be reset to zero
when soil swelling properties are
changed
stress-swell-xx f [r]
xx swelling stress component
stress-swell-yy f [r]
yy swelling stress component
stress-swell-zz f [r]
zz swelling stress component
stress-swell-xy f [r]
xy swelling stress component
stress-swell-xz f [r]
xz swelling stress component
stress-swell-yz f [r]
yz swelling stress component
Notes:
• The local swelling plane is a plane defined so that the
swelling is isotropic in this plane.
Footnotes
FLAC3D 6.0
FLAC3D 6.0
Mohr-Coulomb Tension Crack (MohrT) Model 373
Mohr-Coulomb Tension Crack (MohrT) Model

The Mohr-Coulomb model is not well suited for simulation of tensile failure
(fracturing). One non-physical behavior of the Mohr-Coulomb constitutive
model regarding the tensile failure is that the tensile plastic strain, , is
irreversible. Consequently, any reversal of the plastic strain that caused failure in
tension results in immediate generation of the compressive stresses. Typically,
the irreversible part of the tensile plastic strain in the geological materials is
relatively small (i.e., can be assumed to be zero). In other words, tensile fractures
need to close completely before compressive stresses are generated in the
direction perpendicular to the fracture. One consequence of this approximation in
the Mohr-Coulomb law is that whenever material fails in tension, it excessively
dilates, or increases in volume. This can be a problem, for example, in case of
dynamic response (e.g., earthquake or blasting) or deformation of the
overburden above the undercut that involves opening of the cracks as the
immediate roof collapses followed by closure of the cracks as the entire
overburden subsides. The constitutive model presented here is an extension of
the Mohr-Coulomb constitutive model that considers tensile plastic strains to be
reversible and prevents generation of compressive normal stresses
(perpendicular to cracks) before cracks close.
Model Description
The model assumes that a zone can have up to three mutually perpendicular
cracks. Each crack completely cuts throughout a zone. If there is no tensile
failure or all tensile cracks are closed, the model behaves the same as a perfectly
plastic Mohr-Coulomb constitutive model. If the tensile strength (which is
initially isotropic) is exceeded,
(1)
the following happens:
1. A crack is formed perpendicular to the tensile principal stress, . Model

parameter number-cracks for the number of cracks in the zone is set to
one.
FLAC3D 6.0
2. The information on the crack orientation (i.e., orientation of the unit

normal), is calculated and stored as model parameters, , and .
3. The tensile strength (which becomes anisotropic) perpendicular to the

crack is set to zero, , as a result of instantaneous softening.
4. The stresses are corrected (i.e., normal stress perpendicular to the crack is
set to zero) according to the tensile failure envelope and associated flow
rule (both defined as in the Mohr-Coulomb model), yielding the value of
parameter .
5. Crack opening, (model parameter strain-tension-plastic-1), is

incremented by , in each step.
In subsequent calculation steps, as long as is greater than zero, the stresses

are corrected as if material is yielding in tension in the direction perpendicular to
the crack and is calculated and added to . Also, the shear stresses on the
plane of the crack are set to zero (or stress state within the zone is such that the
traction on the crack plane is zero). After the crack closes (i.e., ), the model
behaves as if the crack does not exist, except that the tensile strength
perpendicular to the crack is zero.
If there is one crack in the zone, the second crack will be perpendicular to it. (The
orientation of the first crack is arbitrary, determined by the orientation of the
principal stresses when the crack forms.) Also, the tensile strength for the
second (and the third) crack is unaffected by the previous cracks in the zone.
There can be the maximum of three mutually perpendicular cracks within the
zone. Thus, when cracks one and two are formed, the orientation of the third
crack is predefined. Otherwise, the second and third cracks behave the same as
the first crack. The opening is tracked separately for each crack. Parameter
number-cracks provides the total number of cracks in the zone. The opening of
each crack is stored in . The orientations of the cracks are stored in
.
mohr-coulomb-tension Model Properties
properties of the Mohr-Coulomb tension model.
FLAC3D 6.0
Mohr-Coulomb Tension Crack (MohrT) Model 375
bulk f
cohesion f
cohesion,
dilation f
friction f
poisson f
Poisson’s ratio,
shear f
tension f
young f
Young’s modulus,
normal-1 v [r]
normal direction of the crack plane
2, ( , , )
normal-2 v [r]
2, ( , , )
normal-3 v [r]
3, ( , , )
number-cracks i [r]
number of crack sets developed
strain-tension-plastic-1 f [r]
accumulated plastic tensile strain in
the direction
the direction
FLAC3D 6.0
the direction
Notes:
Footnotes
FLAC3D 6.0
Model Tests and Examples 377
Model Tests and Examples
List of Examples
• Oedometer Test with Mohr-Coulomb Model
• Uniaxial Compressive Strength of a Jointed Material Sample
• Isotropic Consolidation Test with Double-Yield Model
• Isotropic Consolidation Test with Modified Cam-Clay Model
• Triaxial Compression Test with Hoek-Brown Model
• Triaxial Compression Test with Hoek-Brown-PAC Model
• Isotropic Compression Test with CYSoil Model
• Oedometer Test with CYSoil Model
• Drained Triaxial Test with CYSoil Model — Constant Dilation
• Drained Triaxial Test with CYSoil Model — Dilation Hardening
• Undrained Triaxial Test with CYSoil Model
• Drained Triaxial Compression Test with Simplified Cap-Yield (CHSoil)
Model
• Comparison between Mohr-Coulomb Model and Plastic-Hardening model
• Isotropic Compression Test with Plastic-Hardening Model
• Drained Triaxial Compression Test with Plastic-Hardening Model
• Undrained Triaxial Compression Test with Plastic-Hardening Model
• Oedometer Test with Plastic-Hardening Model
• Single Zone Swell Test
• Single Zone Loading-Unloading Test with MohrT Model
FLAC3D 6.0
FLAC3D 6.0
Oedometer Test with Mohr-Coulomb Model 379
Oedometer Test with Mohr-Coulomb Model

This example concerns the determination of stresses in a Mohr-Coulomb
material subjected to an oedometer test. In this experiment, two of the principal
stress components are equal and, during plastic flow, the stress point evolves
along a vertex of the Mohr-Coulomb criterion representation in the -plane (the
iso-pressure ( = constant) plane in the principal stress space). The purpose is
to validate the numerical technique adopted in FLAC3D to handle such a situation.
Note that FLAC3D uses no special techniques to deal with yield at vertex points of
the Mohr-Coulomb failure locus in the plane. Results of a numerical
experiment are presented and compared to an exact solution.
The boundary conditions for the oedometric test are sketched in Figure 1. They
correspond to the uniform strain rates
(1)
where v is the constant y-component of the velocity applied to the sample

, and L is the height of the sample.
Assuming zero initial stresses, the principal directions of stresses and strains are
those of the coordinate axes. For simplicity, we consider a sample of unit height
= 1.
FLAC3D 6.0
Figure 1: Boundary conditions for oedometer test.
In the elastic range, application of Hooke’s law gives, using at time :
(2)
where and .
To apply the Mohr-Coulomb failure criterion, we consider the yield functions
(3)
At the onset of yield, = 0 and, using Equations (2) and (3), we find
(4)
Hence, yielding will only take place provided 0.
FLAC3D 6.0
During plastic flow, the strain increments are composed of elastic and plastic
parts, and we have
(5)
Using the boundary conditions of Equation (1):
(6)
The flow rule for plastic flow along the edge of the Mohr-Coulomb criterion
corresponding to has the form (e.g., see Drescher 1991)
(7)
where and are the potential functions corresponding to and :
(8)
After partial differentiation, Equation (7) becomes
(9)
In further considering that by symmetry , we obtain
(10)
FLAC3D 6.0
The stress increments, derived from Hooke’s law, are given by the relations
(11)
where we have used the symmetry condition .
Substitution of Equation (6) in Equation (11) yields, using Equation (10):
(12)
The parameter can now be determined by expressing the condition that during
plastic flow, = 0. Using Equation (3), this condition takes the form
(13)
Substitution of Equation (12) in Equation (13) yields, after some manipulations,

the expression
(14)
where
(15)
The FLAC3D simulation is carried out using a single zone of unit dimensions.
Several properties are used in conjunction with the Mohr-Coulomb model:
bulk modulus 200 MPa

shear modulus 200 MPa
cohesion 1 MPa
friction 10°
dilation 10° and 0°
tension 5.67 MPa
FLAC3D 6.0
The velocity components are fixed in the x-, y-, and z-directions. A velocity of
magnitude 10-5 m/steps is applied to the top of the model in the negative
y-direction for a total of 1000 steps. The stress and displacement components in
the y-direction are monitored and compared to the analytic prediction obtained
from Equations (2), (4), and (12), using Equations (14) and (15). Two runs are
carried out with values of 10° and 0° for the dilation parameter. The match is
very good, as can be seen in Figure 2 and Figure 3, where numerical and analytic
solutions coincide at the precision of the plot resolution.
break
FLAC3D 6.0
Figure 2: Oedometric test—comparison of numerical and analytical predictions

for 10° dilation.
Figure 3: Oedometric test—comparison of numerical and analytical predictions

for 0° dilation.
FLAC3D 6.0
Data File
OedometerMohrCoulomb.f3dat
;---------------------------------------------------------------------
; oedometer test
; check plastic flow along an edge of the Mohr-Coulomb criterion
;---------------------------------------------------------------------
model new
model title "Oedometer test on Mohr-Coulomb sample"
zone property bulk 200 shear 200 co 1 friction 10 tension 5.67
[global vyv = -1.e-5]
call 'oedometerTheoretical'
zone gridpoint fix velocity
zone gridpoint initialize velocity-y @vyv range position-y 1.0
history interval 50
zone history displacement-y position 0 1 0
fish history @n_sy
fish history @a_sy
model save 'ini'
; --- dilation 10
zone property dil 10
@d_sigy
model step 1000
model save 'dil10'
; --- dilation 0
model restore 'ini'
zone property dil 0
@d_sigy
model step 1000
model save 'dil0'
FLAC3D 6.0
FLAC3D 6.0
Uniaxial Compressive Strength of a Jointed Material Sample 387
Uniaxial Compressive Strength of a Jointed

Material Sample
Problem Statement
The uniaxial compressive strength of a cylindrical sample of material is
evaluated numerically using the ubiquitous-joint model. This model takes into
consideration a direction of weakness (ubiquitous-joint) in a Mohr-Coulomb
material on which shear failure can be initiated. The compressive strength of the
sample is a function of the material and joint properties, as well as the angle, ,
formed by the direction of the compressive stress and its projection onto the
plane of weakness (see Figure 1).
In this example, the sample is selected as a cylinder with radius, a, and height, b,
such that = 4. The Mohr-Coulomb material has the following properties:
shear modulus ( ) 70 MPa

bulk modulus ( ) 100 MPa
cohesion ( ) 2 kPa
tension limit ( ) 2.4 kPa
The ubiquitous-joint properties include the following:
cohesion ( ) 2 kPa
tension limit ( ) 2.4 kPa
Analytical Prediction
As a definition, let
FLAC3D 6.0
in which is the weak-plane angle, as indicated in Figure 1. Before failure

occurs, the state of stress is homogeneous in the sample. Failure will be initiated
on the weak plane when, for 0,
provided the value of the compressive strength (tension positive) does not
violate the Mohr-Coulomb failure criterion,
in which
If this criterion is violated, or if 0, failure will occur in the matrix instead, on

planes inclined at an angle of with respect to the axis of symmetry of
the sample. See Jaeger and Cook (1979) for details.
Figure 1: Problem geometry.
FLAC3D 6.0
FLAC3D Model
For the numerical simulation, a cylinder with a radius of 1 m and height of 4 m is
selected. A system of reference axes with the x- and z-axes located in the base of
the cylinder and the y-axis pointing along the cylinder axis is selected. This
domain is discretized into 48 zones organized in a radial pattern (see Figure 2). A
uniform velocity is applied in the y-direction at both ends of the cylinder to
induce compression of the sample.
The effect of the variation of has been studied every five degrees from 0° to
90°. The input file uses a FISH function (solveAll) to calculate the compressive
strength at each value. A zone delete command is issued prior to each
calculation at a different . This deletes all zones to effectively reset the model.
The final vertical stress calculated with FISH function sigmav is added to a table at
the end of each run. This approach allows us to save the whole parametric
analysis in one file.
After initiation of the failure mechanism, the state of stress in the sample
becomes nonuniform. To better control the deformation of the system, the
magnitude of the velocities at the top and bottom of the sample are monitored
and adapted as a function of the unbalanced force value (servo-control). The
servo keyword used in the zone face apply command controls the magnitude of
the applied velocity. Cycling continues until an accumulated displacement of
4.5e-4 is reached (using the FISH function halt), indicating enough total strain to
cause failure at all angles.
FLAC3D 6.0
Figure 2: FLAC3D grid—uniaxial compressive strength test.
Results and Discussion

Figure 3 compares results of the FLAC3D runs with the analytical compressive
strength predictions. The match is satisfactory, with a relative error smaller than
2% for all values of .
FLAC3D 6.0
Figure 3: Compressive strength comparison.
Reference
Jaeger, J. C., and N. G. W. Cook. Fundamentals of Rock Mechanics, 3rd Ed. New
York: Chapman and Hall (1979).
Data Files
UniaxialStrengthJointed.f3dat
;---------------------------------------------------------------------
; compression test of cylindrical sample using
; ubiquitous joint model
;---------------------------------------------------------------------
model new
; Create the Extruder data, generated interactively
; and exported from the State Record
call 'geometry' ; Includes 'top' and 'bottom' group assignments
; Load FISH functions used to support -
; the strength calculation and the halt function
call 'fish-support.f3fis'
; The main function that
fish define solveAll
loop local beta (0,90,5) ; Check angles from 0-90 in 5 degree increments
FLAC3D 6.0
command
zone delete ; Remove all existing zones
zone generate from-extruder ; Generate new zones from extruder
zone cmodel assign ubiquitous-joint ; Assign model and properties
zone property bulk 1.e8 shea 7.e7 cohesion 2.e3
zone property friction 40. dilation 0. tension 2400.
zone property dip @beta dip-direction 0. joint-cohesion 1.e3
zone property joint-friction 30. joint-dilation 0. ...
joint-tension 2000.
zone face apply velocity-normal -5e-7 servo ...
ratio local minimum 1e-4 range group 'Top' or 'Bottom'
; Cycle till the target strain is reached
model solve fish-halt @halt
; Add results to table
table 'result' add (@beta,@sigmav)
end_command
end_loop
end
; Run all 18 cases
@solveAll
; Save the last state, and the accumulated table
model save 'final'
return
fish-support.f3fis
; Function that calculates final strength

fish define sigmav
local sum = 0.
loop foreach local gp gp.list
if gp.isgroup(gp,'bottom') then
sum = sum - gp.force.unbal.y(gp)
endif
endloop
sigmav = sum / 3
end
; Function that determines if solving should stop
fish define halt
global gpHalt = gp.near(0,0,0)
halt = gp.disp.y(gpHalt ) > 4.5e-4
end
FLAC3D 6.0
Isotropic Consolidation Test with Double-Yield Model 393
Isotropic Consolidation Test with Double-Yield

Model
To be sure that input parameters are reasonable, a single-element test is
performed with FLAC3D, exercising the double-yield model over stress paths
similar to those of the physical tests and plotting similar graphs.
As an illustration, the data file exercises the double-yield model for a material
that exhibits a response eleven times stiffer upon unloading than upon loading.
The loading tangent modulus observed in the physical test was constant
and equal to 10 MPa. The slope of unloading increments corresponded to a value
= 110 MPa. To define the volumetric properties of the numerical model, we
substitute those values in the equation on and find that = 10. As can be seen
from the the equation on , the hardening curve has a constant slope
corresponding to = 11 MPa. The hardening table is derived from this
result, assuming no over-consolidation.
Note that the input value for maximum bulk modulus, , must be higher than
(see the equation on ). The input shear modulus controls the ratio of
. In this example,
and the Poisson’s ratio is
Results of the numerical test are presented in the plot of minus vertical stress
versus minus vertical strain in Figure 1. The loading slope is 10 MPa, and the
unloading slope is eleven times stiffer, as expected.
FLAC3D 6.0
Figure 1: Single-element test in which unloading is eleven times stiffer than

loading.
The maximum elastic moduli and should be estimated for the

maximum pressure likely to be produced in the model. They should not be set
larger than this because FLAC3D does mass scaling (for a stable timestep) on the
basis of the moduli; setting them too high will give rise to sluggish response
(e.g., the model may be slow to converge to a steady-state solution). The elastic
moduli also act as a limit on plastic moduli.
If a material to be modeled has experienced some initial compaction (i.e., it is

over-consolidated), then may be set to this pre-consolidation pressure. In this
case, must also be set so that it is consistent with and the given table (use
zone property strain-volumetric-plastic to set ).
Data File
IsotropicCompressionDoubleYield.f3dat
; Isotropic Conpression Exercising with the double-yield model

model new
model title "Exercising the Double-Yield Model"
FLAC3D 6.0
Isotropic Consolidation Test with Double-Yield Model 395

zone cmodel assign double-yield
zone property bulk-maximum 1110e6 shear-maximum 507.7e6 ...
table-pressure-cap 1 multiplier 10
zone property density 1000 cohesion 1e10 tension 1e10
table 1 add (0,0) (1,1.1e7)
zone gridpoint initialize velocity-x -1e-6 range position-x 1
zone gridpoint initialize velocity-y -1e-6 range position-y 1
zone gridpoint initialize velocity-z -1e-6 range position-z 1
zone history stress-yy position (0.5,0.5,0.5)
zone history displacement-y position 0 1 0
model step 1000
zone gridpoint initialize velocity-x -0.1 multiply
zone gridpoint initialize velocity-y -0.1 multiply
zone gridpoint initialize velocity-z -0.1 multiply
model step 1600
FLAC3D 6.0
FLAC3D 6.0
Isotropic Consolidation Test with Modified Cam-Clay Model 397
Isotropic Consolidation Test with Modified Cam-

Clay Model
This simple one-zone example exercises the modified Cam-Clay model for a
normally consolidated material subjected to several load-unload excursions in an
isotropic compression test. The initial stresses is at isotropic stress so that =
= = -5 kPa. The material properties include the following:
maximum shear modulus ( ) 10 MPa

shear modulus ( ) 250 kPa
ratio-critical state ( ) 1.02
slope of normal consolidation line ( ) 0.2
slope of elastic swelling line ( ) 0.05
pressure reference ( ) 1 kPa
specific volume reference ( ) 3.32
pressure consolidation ( ) -5 kPa
The sample is assumed normally isotropic consolidation with = 1, so the

initial mean effective pressure and the preconsolidation pressure are the same.
Their values are specified using a simple FISH function:
fish define camclay_ini_p

loop foreach local p_z zone.list
local mean_p = -(zone.stress.xx(p_z) + zone.stress.yy(p_z) ...
+ zone.stress.zz(p_z))/3.0 - z.pp(p_z)
zone.prop(p_z,'pressure-effective') = mean_p
zone.prop(p_z,'pressure-preconsolidation') = mean_p
endloop
end
The results are shown in Figure 1, 2, and 3. Figure 1 plots the mean effective
pressure history. Figure 2 plots the relation vs. , which shows the normal
consolidation line and elastic swelling lines. In this example, the shear modulus
is explicitly specified as an input parameter; it is thus assumed constant, while
the bulk modulus is changing with the mean effective pressure. The bulk and
shear modulus histories are plotted in Figure 3.
FLAC3D 6.0
Figure 1: Pressure versus displacement.
Figure 2: Specific volume versus .
FLAC3D 6.0
Isotropic Consolidation Test with Modified Cam-Clay Model 399
Figure 3: Bulk and shear moduli versus displacement.
Data File
IsotropicConsolidationModifiedCamClay.f3dat
;------------------------------------------------------------
; Isotropic compression test on Modified-Cam-clay sample (drained)
;------------------------------------------------------------
model new
model title "Isotropic compression test for normally consolidated soil"

zone cmodel assign modified-cam-clay
; --- boundary and initial conditions ---
zone initialize stress xx -5. yy -5. zz -5.
; --- model properties ---
zone property shear 250. bulk-maximum 10000.
zone property ratio-critical-state 1.02 lambda 0.2 kappa 0.05
zone property pressure-reference 1. specific-volume-reference 3.32
;
call 'fishfunctions'
@camclay_ini_p
; --- histories ---
history interval 20
fish history @G
FLAC3D 6.0
fish history @K
fish history @p
fish history @lnp
fish history @v
zone history displacement-z position 0 0 1
; --- test ---
@loadingunloading
model save 'camiso'
return
FLAC3D 6.0
Triaxial Compression Test with Hoek-Brown Model 401
Triaxial Compression Test with Hoek-Brown

Model
The triaxial compression tests performed using the Hoek-Brown-PAC model are
repeated for the Hoek-Brown model, with the same material properties:
5
0.5
1.0 MPa
1.5 MPa
100 MPa
0.35
Two compression loading tests are performed:
at zero confining stress, = 0, and
at high confining stress, = 1.
For the zero confining stress case, we specify an associated flow rule; this is done
with the command:
zone property flag-dilation = -1
For the higher confining stress case, we need to specify a dilation angle that is
consistent with the limiting constant-volume stress, = 1.5, chosen for the
triaxial compression test for the Hoek-Brow-PAC model. We linearly interpolate
a value for dilation corresponding to the current confining stress level of = 1,
relative to a nonassociated zero dilation at = 1.5. The current dilation, , is
then taken to be a fraction of the current friction angle, , using the linear
interpolation:
FLAC3D 6.0
The following commands are used to apply the modified Hoek-Brown model for
this case:
zone property flag-dilation = 0.333
The FLAC3D results for the zero confining stress case are compared to the
analytical solution (see Triaxial Compression Test with Hoek-Brown-PAC Model)
in Figure 1 and Figure 2, and the results for the high confining stress case are
compared in Figure 3 and Figure 4.
break
FLAC3D 6.0
Figure 1: Triaxial compression test—stress versus axial strain ( = 0).
Figure 2: Triaxial compression test—lateral strain versus axial strain ( =

0).
FLAC3D 6.0
Figure 3: Triaxial compression test—stress versus axial strain ( = 1.0).

1.0).
FLAC3D 6.0
Data File
TriaxialCompressionHoekBrown.f3dat
; Triaxial tests on a Modified Hoek-Brown material,

model new
model title "Triaxial Test on a Hoek-Brown material"
call 'input_record'
@input
zone property density = 1.0 shear=@shear bulk=@bulk
zone property constant-sci=@sig_ci constant-mb=@mb ...
constant-s=@s constant-a=@a
@locptrs
zone face apply velocity-y @y_vel range position-y 1.0
zone face apply velocity-y [-y_vel] range position-y 0.0
fish history @record_variables
history interval 1000
fish history @eps_xx
fish history @eps_yy
fish history @sig_xx
fish history @sig_yy
fish history @sig_zz
model save 'ini'
; unconfined
zone property flag-dilation = -1
fish set @sig_conf = 0.0
zone face apply stress-xx = @sig_conf
zone face apply stress-zz = @sig_conf
zone initialize stress xx = @sig_conf yy = @sig_conf zz = @sig_conf
model step @cyc
model save 'conf0'
; confined
model restore 'ini'
zone property flag-dilation = 0.333
fish set @sig_conf = -1.0 ; negative is compression
model step @cyc
model save 'conf1'
FLAC3D 6.0
FLAC3D 6.0
Triaxial Compression Test with Hoek-Brown-PAC Model 407
Triaxial Compression Test with Hoek-Brown-

PAC Model
Triaxial compression tests are performed on models composed of Hoek-Brown-
PAC material in FLAC3D to verify the stress and strain paths that develop.
A single-zone model to simulate the triaxial loading tests is constructed in

FLAC3D. The triaxial load conditions are illustrated in Figure 1. Compression
loading tests are performed under two loading conditions: = 0 and 1.0 ( =
0 and 1.0 MPa).
The triaxial tests are performed on a sample of Hoek-Brown material with the
following properties:
5
0.5
1.0 MPa
1.5 MPa
100 MPa
0.35
The analytical solutions for stress and strain during compression loading are
presented by the plots shown in Figure 2 and Figure 3.
The FLAC3D results are compared to the analytical solutions in Figure 4 through
Figure 7. The solutions compare within 1%.
FLAC3D 6.0
Figure 1: Triaxial compression tests—loading conditions.
FLAC3D 6.0
Figure 2: Triaxial compression tests: a) Hoek-Brown failure envelope; b) stress-

strain plots.
FLAC3D 6.0
Figure 3: Triaxial compression tests: a) confining (lateral) strain versus axial

strain; b) volumetric strain versus axial strain.
FLAC3D 6.0
Figure 4: Triaxial compression test—stress versus axial strain ( = 0).

0).
FLAC3D 6.0
Figure 6: Triaxial compression test—stress versus axial strain ( = 1.0).

1.0).
FLAC3D 6.0
Data File
TriaxialCompressionHoekBrownPAC.f3dat
; Triaxial tests on a Hoek-Brown material

model new
model title "Triaxial Test on a Hoek-Brown material"
call 'input_record'
@input
zone cmodel assign hoek-brown-pac
zone property density=1.0 young=@young poisson=@poiss
zone property constant-sci=@sig_ci constant-mb=@mb constant-s=@s constant-a=@a
zone property stress-confining-prescribed=@sig3_cv
@locptrs
zone face apply velocity-y @y_vel range position-y 1
zone face apply velocity-y [-y_vel] range position-y 0
fish history @record_variables ;; 1
fish history @eps_xx
fish history @sig_xx
fish history @sig_yy
fish history @sig_zz
model save 'ini'
; unconfined
fish set @sig_conf = 0.0
model step @cyc
model save 'conf0'
; confined
model restore 'ini'
fish set @sig_conf = -1.0 ; negative is compression
model step @cyc
model save 'conf1'
FLAC3D 6.0
FLAC3D 6.0
Isotropic Compression Test with CYSoil Model 415
Isotropic Compression Test with CYSoil Model

Isotropic compression tests on dense, medium, and loose sand are simulated
using the CYSoil model. The initial stress state is isotropic in each zone; the
magnitude of the confining stress is = = = 10 kN/m2. The sand is
normally consolidated for the tests (the initial cap pressure is equal to 10 kN/m 2).
The CYSoil model properties for the tests are listed in Table 1. Any property that
is not listed in the table is taken as the default value of the CYSoil model (see the
complete properties). The base of the zones is fixed in the axial (y-) direction,
confining velocities of magnitude 10-6 m/step are applied at the top and lateral
sides of the zone for a total of 2500 steps. Five unloading/reloading excursions
are also included.
Table 1: CYSoil Model Properties
Parameter Dense Medium Loose

300 225 150
45 45 45
(kPa) 100 100 100
(kPa) 10 10 10
0.2 0.2 0.2
0.5 0.5 0.5
0.667 0.667 0.667
A plot of vertical (axial) stress versus axial strain for the test is shown for dense,
medium, and loose soil cases in Figure 1. The plot shows the power law and
stiffer behavior achieved by the denser soil, as expected from the model. A plot of
elastic bulk modulus versus axial strain for the test is shown in Figure 2. The
bulk modulus is seen to remain constant during unloading/reloading; also, the
value is higher for higher strain levels, consistent with the dependency of the
property on plastic deformation.
FLAC3D 6.0
Figure 1: Axial stress (in kN/m2) versus axial strain for dense, medium, and
loose sand.
Figure 2: Bulk modulus (in kN/m2) versus axial strain for dense, medium, and
loose sand.
FLAC3D 6.0
Isotropic Compression Test with CYSoil Model 417
Data File
IsotropicCompressionCYSoil.f3dat
model new
model title "Isotropic Compression Test using CYSoil Model"
[global _pc0 = 10.]
zone cmodel assign cap-yield
zone property density=1000 pressure-reference=100 poisson=0.2 multiplier=0.667
zone property friction=45.0 pressure-initial=@_pc0 ...
flag-cap=1 pressure-cap=@_pc0
zone cmodel assign null range id-list 2,4
zone property shear-reference 300 range id 5
zone initialize stress xx [-_pc0] yy [-_pc0] zz [-_pc0]
zone gridpoint initialize velocity-x -1e-6 range position-x 1
zone gridpoint initialize velocity-y -1e-6 ...
range union position-y 1 position-y 3 position-y 5
[global _z5 = zone.find(5)]
fish define _bulk_current1
global _bulk_current5 = zone.prop(_z5, 'bulk')
end
history interval 10
zone history stress-yy zoneid 5
zone history displacement-y position (0,5,0)
fish history @_bulk_current5
fish define trip
loop local i (1,5)
command
zone gridpoint initialize velocity-x -1e-6 range position-x=1.0
zone gridpoint initialize velocity-z -1e-6 range position-z=1.0
zone gridpoint initialize velocity-y -1e-6 range ...
union position-y=1 position-y=3 position-y=5
model step 300
zone gridpoint initialize velocity (-0.1,-0.1,-0.1) multiply
model step 3000
zone gridpoint initialize velocity (-1.0,-1.0,-1.0) multiply
model step 3000
end_command
FLAC3D 6.0
end_loop
end
model step 1000
@trip
return
FLAC3D 6.0
Oedometer Test with CYSoil Model 419
Oedometer Test with CYSoil Model

Oedometer test simulations are carried out using the CYSoil model for different
values of the parameter to evaluate the impact of the cap aspect ratio on
confining stress when yielding occurs on the cap. The values of considered for
the tests are 0.5, 1, 1.5, and 1020 (i.e., a value that is very large compared to 1, in
which case the behavior of the double-yield model is recovered).
We use the same setup and properties (apart from ) as in the isotropic
compression example, except in this case, the simulations are run in plane
strain, with fixed lateral boundaries to simulate oedometer test conditions.
Friction is assigned a large value to prevent shear yielding.
The stress ratio, , of confining stress to vertical stress, / , is plotted

versus axial strain for dense, medium, and loose sand in three figures: Figure 1
compares predictions for = 0.5 and = (double-yield model); Figure 2
compares predictions for = 1.0 and = (double-yield model); and Figure 3
shows plots for = 1.5 and = (double-yield model).
The results of the oedometer simulations show that, with the given properties, a
higher value of is achieved for the CYSoil model than the double-yield model.
And for all tests, the lower the aspect ratio of the cap, the higher the that is
achieved. Also, dense, medium, and loose sands converge to the same ultimate
value as deformation takes place, and they do so at a faster deformation rate
for the CYSoil model than the double-yield model.
FLAC3D 6.0
Figure 1: versus axial strain for dense, medium, and loose sand— = 0.5
(solid lines) and double-yield model (dashed lines).
FLAC3D 6.0
Oedometer Test with CYSoil Model 421
Data File
OedometerCYSoil.f3dat
model new
model title "Oedometer test - dense, medium, loose sand - CYSoil and DY"
;
zone group 'gcy' range position-x 0 1
zone group 'gdy' range position-x 2 3
zone group 'gnull' range union position-x 1 2 position-z 1 2 position-z 3 4
;
[global _pc0 = 10.]
;
zone cmodel assign cap-yield range group 'gcy'
zone property density=1000 pressure-reference=100. poisson=0.2 ...
multiplier=[2.0/3.0] flag-cap=1 flag-shear=1 friction=89.0 ...
pressure-cap=@_pc0 pressure-initial=@_pc0 range group 'gcy'
;
zone cmodel double-yield range group 'gdy'
zone property density=1000 bulk-maximum=1.e15 shear-maximum=0.75e15 ...
friction=89.0 pressure-cap=@_pc0 multiplier=[2.0/3.0] ...
range group 'gdy'
;
zone cmodel null range group 'gnull'
FLAC3D 6.0
;
call 'fish'
; double-yield
@cap_table(1, 300.)
zone property strain-volumetric-plastic=@evp0 table-pressure-cap=1 ...
range group 'gdy' position-z 4 5
@cap_table(2, 225.)
@cap_table(3, 150.)
; cap-yield-soil
zone property shear-reference 300. range group 'gcy' position-z 4 5
;
zone gridpoint initialize velocity-z -1.25e-7 ...
range union position-z 1 position-z 3 position-z 5
zone initialize stress xx [-_pc0] yy [-_pc0] zz [-_pc0]
;
@_hist_setup
fish history @_k0_d_cy
fish history @_k0_m_cy
fish history @_k0_l_cy
fish history @_k0_d_dy
fish history @_k0_m_dy
fish history @_k0_l_dy
model save 'ini'
; -- alpha = 0.5
zone property alpha=0.5 range group 'gcy'
model step 60000
model save 'al1'
; -- alpha = 1.0
model restore 'ini'
model step 60000
model save 'al2'
; -- alpha = 1.5
model restore 'ini'
model step 60000
model save 'al3'
return
FLAC3D 6.0
Drained Triaxial Test with CYSoil Model — Constant Dilation 423
Drained Triaxial Test with CYSoil Model —

Constant Dilation
Triaxial tests on dense, medium, and loose sand are simulated using the CYSoil
model. The model properties are listed in Table 1. This example does not consider
yielding on the cap; therefore, flag_cap is set to 0 (default value). In this case,
initial pressure, , is required; it is used by the code to calculate for bulk and
shear moduli. The dilation is constant; therefore, flag_dilation is set to 1, and a
value is assigned to the dilation property. The initial stress state is isotropic, with
mean pressure equal to 100 kPa. The cell pressure is kept constant during the
test, the base of the model is fixed in the axial ( y-) direction, and an axial velocity
of 10-6 m/step is applied at the top of the model. In addition, three unloading/
reloading excursions are performed.

300 225 150
40 35 30
10 5 0
0 or 100 kPa
(kPa) 100
(kPa) 100
0.2
0.5
0.9
0.35
A plot of deviatoric stress versus axial strain for the cohesionless simulation is
shown in Figure 1 for dense, medium, and loose sand cases. The plot shows the
hyperbolic behavior expected from the model, and the higher failure level
achieved by the denser soil. Three unloading/reloading excursions are also shown
FLAC3D 6.0
in the figure. The plot of volumetric strain versus axial strain for the
cohesionless simulation is shown in Figure 2. The volumetric behavior is
monotonic in the plot; the dilatant behavior of the dense sand is clearly shown.
The comparable results for the simulation with cohesion set to 100 kPa are
shown in Figure 3 and Figure 4. It is seen from Figure 3 that the initial stiffness
of cohesive sand is higher than that of the sand without cohesion. Similarly,
deviatoric stress at failure is higher for the cohesive material.
break
FLAC3D 6.0
Figure 1: (in kPa) versus axial strain for dense, medium, and loose
sand—constant dilation.
Figure 2: Volumetric strain versus axial strain for dense, medium, and loose
FLAC3D 6.0
sand—constant dilation (dilation = 100 kPa).
FLAC3D 6.0
Data File
DrainedTriaxialConstantDilationCYSoil.f3dat
; Drained Triaxial Tests - Constant Dilation

model title "Drained triaxial test - cap-yield-soil"
;
zone face skin
;
zone face apply stress-norm -100. range group 'East' or 'West'
zone face apply stress-norm -100. range group 'North' or 'South'
zone face apply velocity-z 0. range group 'Bottom'
zone face apply velocity-z -1e-6 range group 'Top'
;
zone property density 1000 pressure-reference=100. exponent=1.0 ...
beta=0.35 poisson 0.2
zone property flag-dilation=1 friction-mobilized=0
zone property shear-reference=@_Gref friction=@_friu ...
cohesion=@_coh dilation=@_dilu
zone property flag-cap=0 pressure-initial=100.
;
fish define _cy_q
local _z = zone.head
global _cy_q = zone.prop(_z, 'stress-deviatoric')
global _vsi = zone.strain.vol.inc(_z)
end
fish history @_cy_q
fish history @_vsi
history interval 50
;
fish define trip
loop local i (1,3)
command
model step 8000
zone face apply velocity-z 1e-7 range group 'Top'
mode step 6000
model step 6000
end_command
end_loop
end
;
model step 3000
@trip
model step 23000
;
history export 2 vs 1 reverse table 'deviatoricStress'
FLAC3D 6.0
history export 3 vs 1 reverse table 'volumetricStrain'

table 'deviatoricStress' ...
export [string.build('deviatoricStress%1', _IChoice)] truncate
table 'volumetricStrain' ...
export [string.build('volumetricStrain%1', _iChoice)] truncate
;
return
FLAC3D 6.0
Drained Triaxial Test with CYSoil Model — Dilation Hardening 429
Drained Triaxial Test with CYSoil Model —

Dilation Hardening
The simulations of drained triaxial test with constant dilation are repeated here
by with dilation hardening, this time also using a dilation hardening law as
described in the dilation hardening section of the CYSoil model. Initial dilation is
zero, and the ultimate values of dilation are 10, 5, and 0 degrees for dense,
medium, and loose sand, respectively. Test results for all three cases are written
to table files.
The simulation results of deviatoric stress and volumetric strain versus axial
strain are plotted in Figure 1 and Figure 2, respectively. The non-monotonic
volumetric behavior is apparent in the second plot: all soil types do compact
initially, and the denser soil is shown to dilate upon further shearing.
break
FLAC3D 6.0
sand—dilation hardening.
sand—dilation hardening.
FLAC3D 6.0
Drained Triaxial Test with CYSoil Model — Dilation Hardening 431
Data File
DrainedTriaxialDilationHardeningCYSoil.f3dat
; Drained Triaxial Tests - Dilation Hardening

model title "Drained triaxial test - cap-yield-soil"
;
zone face skin
;
;
zone property density 1000 pressure-reference=100. exponent=1.0 ...
beta=0.35 poisson 0.2
;
fish define _cy_q
global _vsi = zone.strain.vol.inc(_z)
end
fish history @_cy_q
fish history @_vsi
history interval 50
;
model step 50000
;
;
return
FLAC3D 6.0
FLAC3D 6.0
Undrained Triaxial Test with CYSoil Model 433
Undrained Triaxial Test with CYSoil Model

Tests similar to those in the drained triaxial test with dilation hardening are
repeated, but this time under undrained conditions. The model setup and
properties are the same as those used in the previous examples. In addition, the
groundwater flow configuration is selected, with flow turned off. Porosity is 0.3,
and the fluid bulk modulus is 2000 kPa for the runs.
The undrained simulation results for deviatoric stress and volumetric strain
versus axial strain are plotted in Figure 1 and Figure 2, respectively. A softening
stress-strain response is observed for the loose soil in Figure 1. Also, while the
excess pore pressure rises initially for all soils (as indicated in Figure 2), for the
medium and dense soils, it is shown to decrease upon further shearing as a result
of dilation taking place.
break
FLAC3D 6.0
sand—undrained triaxial tests.
Figure 2: Pore pressure versus axial strain for dense, medium, and loose
FLAC3D 6.0
Undrained Triaxial Test with CYSoil Model 435
Data File
UndrainedTriaxialCYSoil.f3dat
; Undrained Triaxial Tests - Dilation Hardening

model title "Undrained triaxial test - cap-yield-soil"
model config fluid
;
zone face skin
;
;
zone property density 1000 pressure-reference=100. ...
exponent=1.0 beta=0.35 poisson 0.2
;
zone fluid cmodel assign isotropic
zone fluid property porosity 0.3
zone gridpoint initialize fluid-tension -1e20
zone gridpoint initialize fluid-modulus 2e3
model fluid active off
;
fish define _cy_q
global _pp = zone.pp(_z)
end
fish history @_cy_q
fish history @_pp
history interval 50
;
model step 50000
;
history export 3 vs 1 reverse table 'porePressure'
table 'porePressure' ...
export [string.build('porePressure%1', _iChoice)] truncate
;
return
FLAC3D 6.0
FLAC3D 6.0
Drained Triaxial Compression Test with Simplified Cap-Yield (CHSoil) Model 437
Drained Triaxial Compression Test with

Simplified Cap-Yield (CHSoil) Model
Triaxial experiments are conducted numerically using the CHSoil model at the
three levels of constant mean stress, 40, 80, and 160 kPa, for the = 40% sand.
A servo-control is applied to maintain the mean stress constant during the
numerical experiments.
The estimates for model properties are used to conduct the numerical tests, and
the test results are compared to the available laboratory data (imported into
FLAC3D in tables). The properties are adjusted (see the section on calibration),
and the numerical experiment is repeated until a satisfactory curve fitting is
obtained.
The results of the curve fitting experiment are listed in Table 1.
Table 1: Calibration Results
40% 1800 0.35 34° 7.5° 28° 0.5 0.5
A comparison between numerical predictions using the calibrated properties and

laboratory results is shown in Figure 1 through Figure 3. Note that the soil-
mechanics convention for positive stress/strain is adopted in these plots.
(Dilation is negative.) The comparison is quite reasonable.
FLAC3D 6.0
Figure 1: Deviator stress versus mean stress for = 40%—comparison

between laboratory (line) and numerical (cross) results.
Figure 2: Deviatoric stress versus axial strain (in %) for = 40%—comparison

between laboratory (line) and numerical (cross) results.
FLAC3D 6.0
Drained Triaxial Compression Test with Simplified Cap-Yield (CHSoil) Model 439
Figure 3: Volumetric strain (in %) versus axial strain (in %) for =

40%—comparison between laboratory (line) and numerical (cross) results.
Data File
DrainedTriaxialCHSoil.f3dat
model new
[t ='Drained triaxial test at constant bmean pressure Dr=40 - cap-yield-soil']
model title [t]
;
zone cmodel assign cap-yield-simplified
zone property young-reference=1800 poisson=0.35 pressure-reference=100.0
zone property failure-ratio=0.99 friction=34.0 ...
exponent-bulk=0.5 exponent-shear=0.5
zone property density=1000. cohesion=0.0 dilation-mobilized=0.0 ...
friction-mobilized=0.0
zone property flag-dilation=2 friction-critical=28.0 dilation=7.5
zone property table-dilation 2 range id = 1
zone property pressure-initial= 40.0 range id = 1
zone property pressure-initial= 80.0 range id = 3
zone property pressure-initial=160.0 range id = 5
;
zone cmodel assign null range id-list = 2, 4
FLAC3D 6.0
;
zone initialize stress xx -40.0 yy -40.0 zz -40.0 range id = 1
;
zone gridpoint initialize velocity-x 0.25e-6 range position-x 1.0
zone gridpoint initialize velocity-y 0.25e-6 range position-y 1.0
zone gridpoint initialize velocity-z -0.5e-6 ...
range union position-z 1.0 position-z 3.0 position-z 5.0
;
[global z1 = zone.near(0.5,0.5,0.5)]
;
call 'servo'
;
fish history @q1
fish history @p1
fish history @eps_v1 ; vol. strain (%) dilation positive
fish history @eps_a1 ; axial strain (%)
fish history @q2
fish history @p2
fish history @q3
fish history @p3
model step 100000
model save 'chsoil-dtriax1'
FLAC3D 6.0
Comparison between Mohr-Coulomb Model and Plastic-Hardening model 441
Comparison between Mohr-Coulomb Model and

Plastic-Hardening model
This example compares the behavior of the Plastic-Hardening (PH) model and
Mohr-Coulomb (MC) model during the triaxial compression. Both models are
used in a one-zone triaxial compression test with a constant cell pressure of 100
kPa. The strength parameters (including friction angle, dilation angle, and
tension limit) are the same for both models. The stiffness of the PH model is
used as Young’s modulus for the MC model and is assumed to be three times
the value of . Material properties for this example are summarized in Table 1.
Material parameters not listed in the table are default values.
Table 1: Model Properties for the

Triaxial Compression Test
Parameters PH MC
(degrees) 30 30
(kPa) 0 0
(degrees) 10 10
(kPa) 2e4
(kPa) 2e4
0.2 0.2
(kPa) 6e4
0.6
0.9
(kPa) 100
(kPa) -100
Figure 1 shows a plot deviatoric stress versus axial strain for both PH and MC
models. It is easy to verify the following from the figure:
FLAC3D 6.0
1. The ultimate failure deviatoric stresses (200 kPa) are the same for both
models, as expected.
2. For the pre-failure curve, the PH and MC models are crossing at the half
of the failure stress (100 kPa), which is consistent with the concept of
stiffness.
3. The unloading stiffness in the MC model is the same as the loading

stiffness ( in the MC model, or in the PH model), while these
stiffnesses are different in the PH model.
Figure 1: Comparison of PH and MC models for a triaxial compression test.
Data Files
TriaxialCompressionPlasticHardening.f3dat
model new
;
zone cmodel assign plastic-hardening
zone property stiffness-50-reference=2.0e4 stiffness-ur-reference=6.0e4 ...
pressure-reference=100.0 exponent=0.6 poisson=0.2 ...
failure-ratio=0.9
FLAC3D 6.0
Comparison between Mohr-Coulomb Model and Plastic-Hardening model 443
zone property friction=30.0 dilation=10.0 cohesion=0.0

zone property stress-1-effective=-100.0 stress-2-effective=-100.0 ...
stress-3-effective=-100.0
;
zone gridpoint fix velocity-z
zone face apply stress-xx=-100.0 range union position-x 0 position-x 1
zone face apply stress-yy=-100.0 range union position-y 0 position-y 1
zone initialize stress xx -100.0 yy -100.0 zz -100.0
;
fish define hhhq_
local zp_ = zone.head
local gp_ = gp.find(8)
global hhhq_ = zone.stress.xx(zp_) - zone.stress.zz(zp_)
global hhha_ = -gp.disp.z(gp_)
end
history interval 10
fish history @hhhq_
fish history @hhha_
;
model step 10000
zone gridpoint initialize velocity-z 1e-6 range position-z 1
model step 1000
model step 10000
model step 1000
model step 10000
model step 1000
model step 20000
;
hist export 1 vs 2 table 'ph_qs_100'
table 'ph_qs_100' export 'ph_qs_100' truncate
;
model save 'ph100'
TriaxialCompressionMohrCoulomb.f3dat
model new
;
zone property young=2.0e4 poisson=0.2 friction=30 dilation=10 cohesion=0.0
;
zone gridpoint fix velocity-z
zone face apply stress-xx=-100.0 range union position-x 0 position-x 1
zone face apply stress-yy=-100.0 range union position-y 0 position-y 1
zone initialize stress xx -100.0 yy -100.0 zz -100.0
;
FLAC3D 6.0
fish define hhhq_

local zp_ = zone.head
local gp_ = gp.find(8)
global hhhq_ = zone.stress.xx(zp_) - zone.stress.zz(zp_)
global hhha_ = -gp.disp.z(gp_)
end
history interval 10
fish history @hhhq_
fish history @hhha_
;
model step 10000
model step 1000
model step 10000
model step 1000
model step 10000
model step 1000
model step 20000
;
hist export 1 vs 2 table 'mc_qs_100'
table 'mc_qs_100' export 'mc_qs_100' truncate
;
model save 'mc100'
FLAC3D 6.0
Isotropic Compression Test with Plastic-Hardening Model 445
Isotropic Compression Test with Plastic-

Hardening Model
This example presents the simulation of the isotropic compression test for dense,
medium, and loose sands using the Plastic-Hardening (PH) model. The initial
stress state is isotropic; the magnitude of the confining stress is 100 kPa. The
sands are normally consolidated with = 1. Material properties for this
example are summarized in Table 1. Material parameters not listed in the table
are default values.
The resulting plots of mean stress (represented by stress in y- direction) versus

axial strain are presented in Figure 1 and show the nonlinear behavior
(corresponding to the power law) of the cap hardening surface.
(kPa) 4.0e4 3.0e4 2.0e4
(kPa) 3.2e4 2.4e4 1.6e4
40 35 30
10 5 0
0.36 0.43 0.50
(kPa) 100 100 100
0.5 0.5 0.5
(kPa) -100 -100 -100
1.0 1.0 1.0
FLAC3D 6.0
Figure 1: Axial stress (in kPa) versus axial strain for dense, medium, and loose
sand.
Data File
IsotropicCompressionPlasticHardening.f3dat
model new
model title "Isotropic Compression Test"
;
zone property dens 1000 pressure-reference=100. exponent=0.5 ...
failure-ratio=0.9 poisson=0.2 o-c-r = 1.0
zone property stress-1-effective=-100. stress-2-effective=-100. ...
stress-3-effective=-100.
zone cmodel assign null range id-list=2, 4
; --- dense ---
zone property stiffness-50-reference=4e4 ...
stiffness-oedometer-reference=3.2e4 ...
friction=40 dilation=10 ...
coefficient-normally-consolidation=0.36 range id 5
; --- medium ---
friction=35 dilation= 5 ...
; --- loose ---
FLAC3D 6.0
Isotropic Compression Test with Plastic-Hardening Model 447

friction=30 dilation= 0 ...
;
zone gridpoint fix velocity-x -1e-6 range position-x 1.0
zone gridpoint fix velocity-z -1e-6 range position-z 1.0
zone gridpoint fix velocity-y -1e-6 ...
range union position-y 1.0 position-y 3.0 position-y 5.0
;
history interval 50
zone history displacement-y position (0 5 0)
;
fish def trip
loop i (1,5)
command
zone gridpoint fix velocity-x -2e-6 range position-x 1.0
zone gridpoint fix velocity-z -2e-6 range position-z 1.0
zone gridpoint fix velocity-y -2e-6 ...
range union position-y 1.0 position-y 3.0 position-y 5.0
model step 500
zone gridpoint initialize velocity -.15 -.15 -.15 mul
model step 3000
zone gridpoint initialize velocity -1.0 -1.0 -1.0 mul
model step 3000
end_command
end_loop
end
model step 1000
@trip
model save 'ph-iso'
return
FLAC3D 6.0
FLAC3D 6.0
Drained Triaxial Compression Test with Plastic-Hardening Model 449
Drained Triaxial Compression Test with Plastic-

Hardening Model
Triaxial tests on dense, medium, and loose sand are simulated using the Plastic-
Hardening (PH) model. The model properties are listed in Table 1. This example
does not consider yielding on the cap; therefore, the default value = 100 is
used. The initial stress state is isotropic, with mean pressure equal to 100 kPa.
The cell pressure is kept constant during the test, the base of the model is fixed
in the axial (y-) direction, and an axial velocity of 10-6 m/step is applied at the
top of the model. In addition, three unloading/reloading excursions are
performed.
Table 1: Plastic-Hardening Model Properties
(kPa) 4.0e4 3.0e4 2.0e4
40 35 30
10 5 0
0.36 0.43 0.50
(kPa) 0 or 20
(kPa) 100
0.5
(kPa) -100
100
1.02
A plot of deviatoric stress versus axial strain is shown in Figure 1, which reveals
the hyperbolic behavior. The unloading-reloading paths are also shown in the
figure. The plot of volumetric strain versus axial strain is shown in Figure 2. The
dilatancy of the denser sands is clearly represented. The smooth evolution of the
dilation angle when the void ratio is approaching the critical state occurs as a
result of the dilation smoothing technique implemented in the model logic.
FLAC3D 6.0
The comparable results for the simulation with cohesion set to 20 kPa are shown
in Figure 3 and Figure 4. It is seen from Figure 3 that the initial stiffness of
cohesive sand is higher than for the sand without cohesion. Similarly, deviatoric
stress at failure is higher for the cohesive material.
break
FLAC3D 6.0
FLAC3D 6.0
FLAC3D 6.0
Data File
DrainedTriaxialPlasticHardening.f3dat
; Drained Triaxial Tests - Constant Dilation

model title "Drained triaxial test - Plastic Hardening model"

zone face skin
exponent=0.5 void-maximum=1.02
zone property stiffness-50-reference=@E50ref friction=@fric ...
dilation=@dila cohesion=@coh ...
coefficient-normally-consolidation=@knc
zone face apply velocity-z 0 range group 'Bottom'
;
fish define ph_q
local z = zone.head
global ph_q = zone.stress.xx(z) - zone.stress.zz(z)
global vsi = zone.strain.vol.inc(z)
end
;
fish history @ph_q
fish history @vsi
;
fish define trip
loop local i (1,3)
command
model step 4000
zone face apply velocity-z 2e-7 range group 'Top'
model step 3000
model step 3000
end_command
end_loop
end
model step 1500
@trip
model step 30000
;
FLAC3D 6.0

;
return
FLAC3D 6.0
Undrained Triaxial Compression Test with Plastic-Hardening Model 455
Undrained Triaxial Compression Test with

Plastic-Hardening Model
Undrained triaxial tests on dense, medium, and loose sand are simulated using
the Plastic-Hardening (PH) model are repeated, but this time under undrained
conditions. This example does not consider yielding on the cap, therefore the
default value = 100 is used. The initial stress state is isotropic, with mean
pressure equal to 100 kPa. In this example, we realize the undrained condition by
setting a zero volumetric strain; therefore, the fluid logic does not need to be
configured. Triaxial tests on dense, medium, and loose sand are simulated using
the PH model. The model properties are listed in Table 1.
(kPa) 4.0e4 3.0e4 2.0e4
40 35 30
10 5 0
0.36 0.43 0.50
(kPa) 0
(kPa) 100
0.5
(kPa) -100
100
1.02
The undrained simulation results for deviatoric stress is plotted in Figure 1.
FLAC3D 6.0
Data File
UndrainedTriaxialPlasticHardening.f3dat
; Undrained Triaxial Tests

model title "Undrained triaxial test - PH"
;
;
;
exponent=0.5 void-maximum 1.02
zone property stiffness-50-reference=@E50ref ...
stiffness-ur-reference=[3*E50ref] ...
friction=@fric dilation=@dila ...
coefficient-normally-consolidation=@knc
;
zone gridpoint fix velocity-z -1e-6 range position-z 1
zone gridpoint fix velocity-x 5e-7 range position-x 1
zone gridpoint fix velocity-y 5e-7 range position-y 1
FLAC3D 6.0
Undrained Triaxial Compression Test with Plastic-Hardening Model 457
;
fish define ph_q
local z = zone.head
global ph_q = zone.stress.xx(z) - zone.stress.zz(z)
end
;
fish history @ph_q
;
model step 30000
;
;
return
FLAC3D 6.0
FLAC3D 6.0
Oedometer Test with Plastic-Hardening Model 459
Oedometer Test with Plastic-Hardening Model

This section presents the simulation of an oedometer test to reveal the capability
of the Plastic-Hardening (PH) model to reproduce the evolution of lateral stress
ratio . Three kinds of sands with friction angles of 30, 35, and 40
degrees are used in the simulations of this case (Case 1) and the default
consolidation coefficient is calculated as . The model properties are
listed in Table 1. The properties not listed in the table are using the default
values. Initially the model is in equilibrium with isotropic stress state in each
zone, = = = −1 kPa.
(kPa) 4.0e4 3.0e4 2.0e4
40 35 30
10 5 0
(Case 2) 0.7 0.6 0.5
(kPa) 100.0
0.5
(kPa) -1.0
1.0
The results of stress ratio evolution due to compression in the oedometer tests
are shown in Figure 1, and it is seen that they correctly reproduce the expected
evolution path. Figure 2 presents vertical oedometer pressure versus vertical
strain. The slope of the curves at the specified reference pressure of 100 kPa
reproduces the expected oedometer stiffness (as is not provided as part of
input, it is assigned to the same as by default; see the property section of the
PH model).
FLAC3D 6.0
The test is repeated in Case 2 using predefined values of = 0.7, 0.6, and 0.5,
respectively. The results for are plotted in Figure 3 and Figure 4 shows
variation of vertical oedometer pressure versus axial strain. Again, correct stress
paths and oedometer stiffness at the reference vertical pressure are successfully
reproduced.
break
FLAC3D 6.0
Figure 1: path calculated from the oedometer test with friction angles of 30,
35, and 40 degrees and default values.
Figure 2: Vertical pressure versus vertical strain from the oedometer test with
friction angles of 30, 35, and 40 degrees and default values.
FLAC3D 6.0
Figure 3: path calculated from the oedometer test with friction angles of 30,
35, and 40 degrees and specified values.
Figure 4: Vertical pressure versus vertical strain from the oedometer test with
friction angles of 30, 35, and 40 degrees and specified values.
FLAC3D 6.0
Data File
OedometerPlasticHardening.f3dat

zone prop density 1000 pressure-reference=100. exponent=0.5 ...
over-consolidation-ratio=1.0 factor-cut 0.01
zone prop stress-1-effective=-1 stress-2-effective=-1 ...
stress-3-effective=-1
zone prop stiffness-50-reference=4e4 friction=40. dilation=10. ...
coefficient-normally-consolidation=@_Knc1 range id 1
zone prop stiffness-50-reference=3e4 friction=35. dilation= 5. ...
zone prop stiffness-50-reference=2e4 friction=30. dilation= 0. ...
zone cmodel assign null range id-list 2 4
;
zone face apply velocity-z -1e-5 ...
range union position-z 1 position-z 3 position-z 5
zone initialize stress xx -1 yy -1 zz -1
;
[global _z15 = zone.near(0.5, 0.5, 4.5)]
fish def _k0_d
global _k0_l = 0.5*(z.s.xx(_z15)+z.s.yy(_z15)) / z.s.zz(_z15)
global _k0_m = 0.5*(z.s.xx(_z13)+z.s.yy(_z13)) / z.s.zz(_z13)
global _k0_d = 0.5*(z.s.xx(_z11)+z.s.yy(_z11)) / z.s.zz(_z11)
end
zone history displacement-z position 0 0 1
zone history stress-zz position 0.5 0.5 4.5
fish history @_k0_d
fish history @_k0_m
fish history @_k0_l
history interval 10
;
model step 1200
model save [string.build('oed%1', _ICase)]
return
FLAC3D 6.0
FLAC3D 6.0
Single Zone Swell Test 465
Single Zone Swell Test

In this example to test the swell model, a 2 m x 2 m single zone is generated, and
the following elastic material properties are assigned: density = 1000 kg/m 3; bulk
modulus = 100 MPa; and shear modulus = 30 MPa. With roller boundary
conditions enforced at the lateral and bottom boundaries, the model is settled
under the gravity loading (g = 10 m/s2). At this point, the vertical stress at
the middle of elevation in the zone is equal to -10 kPa. In the next step, the
displacements and velocities are reset in the model; the material model is
switched from elastic to swell, which has the same stiffness and high strength
(cohesion and tension). The swelling properties associated with the logarithmic
function are listed in Table 1.
Table 1: Swell Model Properties
Parameter Value
density 1000
gravity ( ) 10
(Pa) 1e8
(Pa) 3e7
(deg) 0
1e10
1e10
(0,0,1)
(Pa) 1e5
1.5335
-0.0187
200
Using the above parameters, an atmospheric pressure of 100.0 kPa, and

logarithmic function for strains (flag-swell = 1), the swelling strain in the
vertical direction obtained from the logarithmic function is 0.01523. Therefore,
FLAC3D 6.0
based on these analytical calculations, the zone swells upward by 0.03046 m.

Numerical simulations are done with the data file provided below. The property
is set to 200, which means the swelling stress will be introduced into the zone
over the next 200 steps. The model step and model solve command are used after
200 steps are taken to bring the model to equilibrium. Execution of the data file
reports 0.03046 m swelling at the top surface, which exactly matches the
analytical prediction.
The vertical displacement history at the zone top is plotted in Figure 1.
Figure 1: Vertical displacement at the zone top vs. steps.
Data File
SingleZoneSwell.f3dat
; signle zone swelling test

model new
model title "Single Zone test - Swell Model"
model gravity (0,0,-10)
;
zone create brick size 1 1 1 edge 2
zone face skin
FLAC3D 6.0
Single Zone Swell Test 467
zone property density 1000 bulk 1e8 shear 3e7

zone face apply velocity-normal 0 range group 'East' or 'West'
model solve ratio-local 1e-5
;
;
zone cmodel assign swell
zone property density 1000 bulk 1e8 shear 3e7 cohesion 1e10 tension 1e10
zone property normal (0,0,1) constant-a-1 1.5335 constant-c-1 -0.0187 ...
pressure-reference 1e5 number-start 200
history interval 10
model step 200
return
FLAC3D 6.0
FLAC3D 6.0
Single Zone Loading-Unloading Test with MohrT Model 469
Single Zone Loading-Unloading Test with MohrT

Model
This example compares the MohrT Model and the traditional Mohr-Coulomb
model with quasi-static loading-unloading of a single zone, subjected to
unidirectional/bidirectional strain.
Quasi-Static Loading-Unloading of a Single Zone — Unidirectional Strain
One zone with unit dimension is subjected to unidirectional strain, including

reversal of extension and compression. The material parameters are summarized
in Table 1. The zone has an initial isotropic stress of -1000 Pa. The strains in the
y-direction are zero. The velocities in the planes of y = 0 and z = 0 are fixed at
zero. The velocity in the plane of x = 1 is 1×10-8 m/s (extension). The zone is
strained for 500 steps; then, the velocity in the plane of x = 1 is changed to
-1×10-8 m/s (compression) and the zone is strained for 800 steps. The applied
strain history is summarized in Table 2. The data file is appended as
“SingleZoneMohrCoulombTensionCrack-1d.f3dat”.
Table 1: Model Properties
Material Parameter Value

(Pa) 2.5e8
(Pa) 1.5e8
30
10
(Pa) 1000
(Pa) 200
break
FLAC3D 6.0
Table 2: Unidirectional Strain History
Loading Direction Strain Range

Extension x-direction: 0 to 5e-6
Composition x-direction: 5e-6 to -3e-6
Figure 1 shows the stress versus strain in the zone in the x-direction during the
zone extension. As the strain is changing from zero to 5×10 -6, the stress is
changing from the initial -1000 Pa to the tension limit of 200 Pa and then
suddenly drops to zero because the crack occurs when the stress reaches the
tension limit. The stress in the loading direction remains zero during the further
extensional strain. When querying the zone parameters, the number of cracks
(number-cracks) is 1, and the unit normal direction for crack number 1 is (1,0,0).
Figure 2 shows the stress versus strain in the zone in the x-direction during the
compression phase of the test. As expected, the compressive stress is not
developed until the crack is closed. After the crack is closed, the stress-strain
path then is along the reversed direction of loading. When querying the zone
parameters, the number of cracks is still 1, and crack number 1 (unit normal
direction) is still (1,0,0). The overlapped loading and unloading paths in Figure 3
show that the model tensile cracking is fully reversible and that compressive
stresses do not develop until the crack is closed. Figure 3 shows the complete
stress versus strain path of loading and unloading.
It is interesting to compare the complete stress-strain path with the one (shown
in Figure 4) generated when the model is changed to a Mohr-Coulomb model but
with the same material parameters as in Table 1. Two major differences between
the MohrT and Mohr models are:
1. when one principal stress reaches the tension limit, it remains at the
tension limit with the Mohr-Coulomb model while it suddenly drops to
zero with the MohrT model; and
2. after reversal of the strain rate (from extension to compression), the

compressive stress is immediately generated with the Mohr-Coulomb
model while it is developed only after the crack is completely closed with
the MohrT model.
FLAC3D 6.0
Figure 1: Stress versus strain in the x-direction during the strain direction in
tension.
Figure 2: Stress versus strain in the x-direction during the reversed strain
direction.
FLAC3D 6.0
Figure 3: Stress versus strain in the x-direction during the complete strain
history.
Figure 4: Stress versus strain in the x-direction during the complete strain
history using the traditional Mohr-Coulomb Model.
FLAC3D 6.0
Quasi-Static Loading-Unloading of a Single Zone — Bidirectional Strain
One zone with unit dimension is subjected to bidirectional strain including

reversal of extension and compression. The same material parameters
summarized in Table 1 are used in this example. The zone has an initial isotropic
stress of -1000 Pa. The velocities in the plane of x = 0 and y = 0 are fixed to be
zero. The velocity in the plane of x = 1 is 1×10-8 m/s and the velocity in the plane
of y = 1 is 2×10-8 m/s (i.e., extension in both directions). The zone is strained for
500 steps; then, the velocity in the plane of x = 1 is changed to -1×10-8 m/s and
the velocity in the plane of y = 1 is changed to -2×10-8 m/s (i.e., compression in
both directions) and the zone is strained 800 steps. The applied strain history is
summarized in Table 3. The data file is appended as
“SingleZoneMohrCoulombTensionCrack-2d.f3dat”.
Figure 5 shows the stresses and versus strain during loading in the
extension direction. As the strain in the y-direction is changing from zero to
1×10-5, the stress is changing from the initial -1000 Pa to the tension limit of
200 Pa and then suddenly drops to zero because the crack occurs when the stress
reaches the tension limit. The stress remains zero during further extensional
strain. After the step when drops to zero, the number of cracks (number-
cracks) in the zone is set to one, and the unit normal direction for crack number
1 is (0,1,0).
It is interesting to find that the stress also has a sudden change when the
first crack is created; this is due to the Mohr-type failure model being based on
the maximum and minimum principal stresses, and when , which is the
maximum principal stress in this example, suddenly changes from 200 Pa to
zero, it has an impact on stress in other principal stresses.
The second crack with a normal direction (1,0,0) is developed when the stress
reaches 200 Pa. At the end of the loading path, the number of cracks ( number-
cracks) is two. Figure 6 shows stress–strain behavior during reversed
(compression) strain following the extensional strain. The crack in the x-plane is
closed first, and the crack closure in the y-plane occurs subsequently. As
illustrated in Figure 7, which shows the complete stress versus strain path of
loading and unloading, after the cracks are closed, the stress–strain path during
compression follows the path during extension.
FLAC3D 6.0
A comparison showing the complete stress versus strain path of loading and
unloading is plotted in Figure 8 when the model is changed to the traditional
Mohr-Coulomb model. This example also confirms the reversibility of the tensile
cracking in the MohrT model.
Table 3: Unidirectional Strain History
Loading Direction Strain Range

Extension x-direction: 0 to 5e-5
y-direction: 0 to 1e-4
Composition x-direction: 5e-5 to -3e-5
y-direction: 1e-4 to -6e-4
break
FLAC3D 6.0
Figure 5: Stress versus strain in the y-direction during the strain direction in
tension.
Figure 6: Stress versus strain in the y-direction during the reversed strain
direction.
FLAC3D 6.0
Figure 7: Stress versus strain in the y-direction during the complete strain
history.
Figure 8: Stress versus strain in the y-direction during the complete strain
history using the traditional Mohr-Coulomb Model.
FLAC3D 6.0
Data Files
SingleZoneMohrCoulombTensionCrack-1d.f3dat
model new
zone face skin
zone cmodel assign mohr-coulomb-tension
zone property bulk 2.5e8 shear 1.5e8 cohesion 1e3 friction 30 ...
dilation 10 tension 200
;
zone face apply stress-yy -1000 range group 'North' or 'South'
zone face apply stress-zz -1000 range group 'Bottom' or 'Top'
;
history interval 1
zone history stress-xx position (0.5,0.5,0.5)
zone history displacement-x position (1.0,1.0,1.0)
;
zone face apply velocity-x 0.0 range group 'West'
zone face apply velocity-x 1e-8 range group 'East'
model step 500
zone face apply velocity-x -1e-8 range group 'East'
model step 800
model save 'mohrt1d'
return
SingleZoneMohrCoulombTensionCrack-2d.f3dat
model new
zone face skin
zone cmodel mohr-coulomb-tension
zone property bulk 2.5e8 shear 1.5e8 cohesion 1e3 friction 30 ...
dilation 10 tension 200
;
zone face apply stress-zz -1000 range group 'Bottom' or 'Top'
;
history interval 1
zone history stress-xx position (0.5,0.5,0.5)
zone history stress-yy position (0.5,0.5,0.5)
zone history displacement-x position (1.0,1.0,1.0)
zone history displacement-y position (1.0,1.0,1.0)
;
zone face apply velocity-x 0.0 range group 'West'
zone face apply velocity-y 0.0 range group 'South'
zone face apply velocity-x 1e-8 range group 'East'
zone face apply velocity-y 2e-8 range group 'North'
model step 500
zone face apply velocity-x -1e-8 range group 'East'
zone face apply velocity-y -2e-8 range group 'North'
FLAC3D 6.0
model step 800

model save 'mohrt2d'
return
FLAC3D 6.0
References 479
References
Boukpeti, N. Modeling Static Liquefaction in Granular Deposits. Ph.D. Thesis,
University of Minnesota (2001).
Britto, A. M., and M. J. Gunn. Critical State Soil Mechanics via Finite Elements.
Chichester U.K.: Ellis Horwood Ltd. (1987).
Byrne, P. M., S. S. Park and M. Beaty. “Seismic Liquefaction: Centrifuge and

Numerical Modeling,” in FLAC and Numerical Modeling in Geomechanics — 2003
(Proceedings of the 3rd International FLAC Symposium, Sudbury, Ontario, Canada,
October 2003), pp. 321-331. R. Brummer et al., eds. Lisse: A. A. Balkema (2003).
Carter, T. G., J. L. Carvalho and G. Swan. “Towards the Practical Application of

Ground Reaction Curves,” in Innovative Mine Design for the 21st Century
(Proceedings of the International Congress on Mine Design — Kingston, Ontario,
Canada — August 1993), pp. 151-171. W. F. Bawden and J. F. Archibald, eds.
Rotterdam: A. A. Balkema (1993).
Cundall, P., C. Carranza-Torres and R. Hart. “A New Constitutive Model Based on

the Hoek-Brown Criterion,” in FLAC and Numerical Modeling in Geomechanics —
2003 (Proceedings of the 3rd International FLAC Symposium, Sudbury, Ontario,
Canada, October 2003), pp. 17-25. R. Brummer et al., eds. Lisse: Balkema (2003).
Davis, R. O., and A. P. S. Selvadurai. Plasticity and Geomechanics. Cambridge

(2002).
Drescher, A. Analytical Methods in Bin-Load Analysis. Amsterdam: Elsevier (1991).
Duncan, J. M., et al. “Strength, Stress-Strain and Bulk Modulus Parameters for
Finite Element Analyses of Stresses and Movements in Soil Masses,” University
of California, Berkeley, College of Engineering, Report No. UCB/GT/80-01 (1980).
Duncan, J. M., and C. Y. Chang. “Nonlinear Analysis of Stress and Strain in Soils,”
Soil Mechanics, 96 (SM5), 1629-1653 (1970).
Hoek, E., and E. T. Brown. “Practical Estimates of Rock Mass Strength,” Int. J.
Rock Mech. Min. Sci., 34 (8), 1165-1186 (1998).
Hoek, E., and E. T. Brown. Underground Excavations in Rock. London: IMM (1980).
FLAC3D 6.0
Hoek, E., C. Carranza-Torres and B. Corkum. “Hoek-Brown Failure Criterion —

2002 Edition,” in Proceedings of NARMS-TAC 2002, 5th North American Rock
Mechanics Symposium and 17th Tunnelling Association of Canada Conference —
Toronto, Canada — July 7 to 10, 2002. Vol.~1., pp. 267-271. R. Hammah et al., eds.
Toronto: University of Toronto Press (2002).
Lekhnitskii, S. G. Theory of Elasticity of an Anisotropic Body. Moscow: Mir

Publishers (1981).
Pan, X. D., and J. A. Hudson. “A Simplified Three Dimensional Hoek-Brown Yield

Criterion,” in Rock Mechanics and Power Plants (Proceedings of the ISRM Symp.), pp.
95-103. M. Romana, ed. Rotterdam: A. A. Balkema (1988).
Roscoe K. H., and J. B. Burland. “On the Generalised Stress-Strain Behavior of

‘Wet Clay’,” in Engineering Plasticity, pp. 535-609. J. Heyman and F. A. Leckie,
eds. Cambridge: Cambridge University Press (1968).
Rowe, P. W. “The Stress-Dilatancy Relation for Static Equilibrium of an Assembly

of Particles in Contact,” Proc. Roy. Soc. A., 269, 500-527 (1962).
Shah, S. A Study of the Behaviour of Jointed Rock Masses. Ph.D. Thesis, University of
Toronto (1992).
Vermeer, P. A., and R. de Borst. “Non-Associated Plasticity for Soils, Concrete

and Rock,” Heron, 29 (3), 3-64 (1984).
Wood, D. M. Soil Behaviour and Critical State Soil Mechanics. Cambridge: Cambridge
FLAC3D 6.0
Fluid-Mechanical Interaction 481
Fluid-Mechanical Interaction
FLAC3D models the flow of fluid through a permeable solid, such as soil. The flow
modeling may be done by itself, independent of the usual mechanical calculation
of FLAC3D, or it may be done in parallel with the mechanical modeling, in order
to capture the effects of fluid/solid interaction. One type of fluid/solid interaction
is consolidation, in which the slow dissipation of pore pressure causes
displacements to occur in the soil. This type of behavior involves two mechanical
effects. First, changes in pore pressure cause changes in effective stress, which
affect the response of the solid. (For example, a reduction in effective stress may
induce plastic yield.) Second, the fluid in a zone reacts to mechanical volume
changes by a change in pore pressure.
The basic flow scheme handles both fully saturated flow and flow in which a
phreatic surface develops. In this case, pore pressures are zero above the phreatic
surface, and the air phase is considered to be passive. This logic is applicable to
coarse materials when capillary effects can be neglected. In order to represent
the evolution of an internal transition between saturated and unsaturated zones,
the flow in the unsaturated region must be modeled so that fluid may migrate
from one region to the other. A simple law that relates the apparent permeability
to the saturation is used. The transient behavior in the unsaturated region is only
approximate (due to the simple law used), but the steady-state phreatic surface
should be accurate (see Steady-State Fluid Flow with a Free Surface and
Spreading of a Groundwater Mound for examples).
1. Two fluid-transport laws corresponding to isotropic and

anisotropic permeability are available. The fluid-flow null model
is also provided to specify impermeable materials in the flow
domain.
2. Different zones may have different fluid-flow models (isotropic,

anisotropic or null) and properties.
3. Fluid pressure, flux, and leaky and impermeable boundary

conditions may be prescribed.
FLAC3D 6.0
482 FLAC3D Theory and Background • Fluid-Mechanical Interaction
4. Fluid sources (wells) may be inserted into the material as either

point sources or volume sources. These sources correspond to
either a prescribed inflow or outflow of fluid, and vary with time.
5. Both explicit and implicit fluid-flow solution algorithms are

available for fully saturated simulations. An explicit method of
solution is used for saturated/unsaturated flow.
6. Any of the mechanical and thermal models may be used with the
fluid-flow models. In coupled problems, the compressibility and
thermal expansion of the saturated material are allowed.
7. Coupling between fluid and mechanical calculations due to

deformable grains is provided through the Biot coefficient, .
8. Coupling to the thermal conduction calculation is provided

through the linear thermal expansion coefficient, ,
and the undrained thermal coefficient, .
9. The thermal-conduction fluid-flow logic is based on a linear

theory that assumes constant material properties and neglects
convection. Fluid and solid temperatures are locally equilibrated.
Nonlinear behavior can be specified by access to pore pressures
and material properties via FISH. Thermal coupling to the
conduction logic is addressed in Mathematical Description.
10. An advection model to take the transport of heat by convection

into account is also provided. This includes temperature-
dependent fluid density and thermal advection by the fluid. See
Advection for a description of this thermal-fluid coupling.
Fluid-flow and coupled undrained and drained calculations using the basic flow
scheme can be very slow when (1) the bulk modulus of the fluid is large
compared to the drained confined modulus, , (2) there is a large
contrast in permeability and/or porosity, or (3) there is a large variation in grid
size. A different numerical technique, saturated fast-flow logic, is provided to
speed the calculation for fully saturated, coupled fluid-mechanical simulations,
when the fluid can be considered as incompressible when compared to the drained
material compressibility. The logic is invoked with the zone fluid fastflow on
FLAC3D 6.0
command. The scheme and theoretical background is described in Fully

Saturated Fast Flow, and efficiency demonstrations and application of this logic
are presented in 1D Consolidation and Consolidation Settlement.
Dynamic pore-pressure generation and liquefaction due to cyclic loading can also
be modeled with FLAC3D Dynamic Pore-Pressure Generation contains the
documentation on this topic). FLAC3D does not represent capillary, electrical or
chemical forces between particles of a partially saturated material. However, it is
possible to introduce such forces by writing a FISH function that supplies the
appropriate internal stresses, based on the local saturation, porosity and/or any
other relevant variable. Similarly, the effect of variable fluid stiffness due to
dissolved air is not explicitly modeled, but a FISH function may be used to vary
the local fluid modulus as a function of pressure, time or any other quantity.
This section is divided into eight major parts:
1. The mathematical model description and the corresponding

numerical formulation for fluid flow and coupled fluid flow-
mechanical and fluid-flow thermal-conduction processes. (see
Mathematical Description and Numerical Formulation).
2. The calculation modes and associated commands for analyses

involving fluid flow. (link)
3. Material properties required for a fluid-flow analysis, and

includes the appropriate units for these properties. ((link))
4. Initial conditions, and a description of the different boundary

conditions and fluid sources and sinks that can be applied in a
FLAC3D model. (link)
5. The recommended procedures for solving both flow-only and

coupled-flow problems. This section also contains several
examples that illustrate the application of these procedures. We
recommend that you work through these examples before
attempting your own fluid analysis. (link)
6. Several verification problems that demonstrate the accuracy of

the fluid-flow logic in FLAC3D.
FLAC3D 6.0
7. Modeling techniques for specific fluid applications: Several topics

are covered: modeling solid weight, buoyancy forces and seepage
forces in a coupled analysis; relation between initialization of
pore pressures and deformation in a coupled analysis; and effect
of the Biot coefficient.
8. Finally, a summary of all the commands and FISH functions

related to fluid flow. A special zone-based pore pressure logic is
also included in this section.
The user is strongly encouraged to become familiar with the operation of FLAC3D
for simple mechanical problems before attempting to solve problems in which
flow and mechanical effects are both important. Coupled flow and mechanical
behavior are often very complicated, and require a good deal of insight to
interpret correctly. Before starting a big project, it is very important to spend
time experimenting with a small-grid version of the proposed simulation, to try
out various boundary conditions and modeling strategies. The time “wasted” on
these experiments will be amply repaid in terms of an overall reduction in both
staff time and execution time.
• FLAC3D Fluid-Thermal-Mechanical-Formulation — Mathematical

Description
◦ Conventions and Definitions
◦ Governing Differential Equations
◦ Fluid Flow Boundary and Initial Conditions in FLAC3D
• Numerical Formulation
◦ Saturated Fluid Flow
◦ Saturated/Unsaturated Flow
◦ Mechanical Timestep and Numerical Stability
◦ Total Stress Correction
◦ Fully Saturated Fast Flow
• Calculation Modes for Fluid-Mechanical Interaction
◦ Grid Not Configured for Fluid Flow
◦ Grid Configured for Fluid Flow
• Properties and Units for Fluid-Flow Analysis
◦ Permeability Coefficient
◦ Mass Density
◦ Fluid Moduli
◦ Porosity
◦ Saturation
FLAC3D 6.0
◦ Undrained Thermal Coefficient

◦ Fluid Tension Limit
• Fluid-Flow Boundary Conditions, Initial Conditions, Sources and Sinks
• Solving Flow-Only and Coupled-Flow Problems
◦ Time Scales
◦ Selection of a Modeling Approach for a Fully Coupled Analysis
◦ Fixed Pore Pressure (Used in Effective Stress Calculation)
◦ Flow-Only Calculation to Establish a Pore-Pressure Distribution
◦ No Flow — Mechanical Generation of Pore Pressure
◦ Coupled Flow and Mechanical Calculations
• Verification Examples
◦ Unsteady Groundwater Flow in a Confined Layer
◦ One-Dimensional Filling of a Porous Region
◦ Steady-State Fluid Flow with a Free Surface
◦ Spreading of a Groundwater Mound
◦ One-Dimensional Consolidation
◦ Consolidation Settlement at the Center of a Strip Load
◦ Transient Fluid Flow to a Well in a Shallow Confined Aquifer
◦ Pressuremeter Test
◦ Semi-confined Aquifer
• Verification of Concepts, and Modeling Techniques for Specific
Applications
◦ Solid Weight, Buoyancy and Seepage Forces
◦ Pore Pressure Initialization and Deformation
◦ Effect of the Biot Coefficient
• Input Instructions for Fluid-Flow Analysis
◦ Fluid Commands
◦ Fluid FISH Variables
◦ Zone-Based Pore Pressure
• References
FLAC3D 6.0
FLAC3D 6.0
FLAC3D Fluid-Thermal-Mechanical-Formulation — Mathematical Description 487
FLAC3D Fluid-Thermal-Mechanical-Formulation
— Mathematical Description
In this section, we consider fluid-mechanical coupling to the heat-conduction
logic (see fluid-thermal coupling via the convection logic). Most engineering
analyses involving simultaneous deformation and fluid/thermal diffusion
mechanisms are carried out using uncoupling techniques (e.g., in this section).
In addition to those calculation modes, FLAC3D provides for the option of coupled
fluid-thermal-mechanical analysis (i.e., in which the mechanical response of a
porous material can be studied under transient fluid flow and/or thermal
conditions). Although this section is mainly concerned with the modeling of
deformation-fluid diffusion problems, the general equations governing the
fluid-thermal-mechanical response in FLAC3D are presented here for
completeness.
The formulation of coupled deformation-diffusion processes in FLAC3D is done

within the framework of the quasi-static Biot theory, and can be applied to
problems involving single-phase Darcy flow in a porous medium. Various types
of fluids, including water and oil, can be represented with this model.
The variables involved in the description of fluid flow through porous media are
pore pressure, saturation and the three components of the specific discharge
vector. These variables are related through the fluid mass-balance equation,
Darcy’s law for fluid transport, a constitutive equation specifying the fluid
response to changes in pore pressure, saturation, volumetric strains and
temperature, and an equation of state relating pore pressure to saturation in the
unsaturated range. Pore pressure and temperature influences are involved in the
mechanical constitutive laws to complete the thermal fluid-flow mechanical
coupling.
Assuming the volumetric strain and temperature rates are known, substitution of
the mass balance equation into the fluid constitutive relation, using Darcy’s law,
yields a differential equation in terms of pore pressure and saturation that may
be solved for particular geometries, properties, boundary and initial conditions.
FLAC3D 6.0
Conventions and Definitions

As a notation convention, the symbol denotes component of the vector in
a Cartesian system of reference axes; is component of tensor . Also,
is used to represent the partial derivative of with respect to . (The symbol
denotes either a scalar variable or a vector component.)
The Einstein summation convention applies only on indices , and , which take
the values 1, 2, 3 for components that involve spatial dimensions. The indices
may take any values when used in matrix equations.
SI units are used to illustrate parameters and dimensions of variables. See

System of Units for conversions to other systems of units.
The following reference quantities are useful in the characterization of transient

fluid flow.
Characteristic length:
(1)
Fluid diffusivity:
(2)
where is mobility coefficient (FLAC3D permeability), is the Biot modulus, is

the Biot coefficient, , is the drained bulk modulus, and is the
shear modulus of the porous material.
The Biot coefficient takes into account the grain compressibility for the porous
material. If is equal to unity, the grains are considered to be incompressible
and the Biot modulus is equal to , where is fluid bulk modulus and
is porosity. The fluid diffusivity becomes
(3)
Note that for flow-only calculations (rigid material), the fluid diffusivity is
(4)
See Properties and Units for Fluid-Flow Analysis for additional discussion on the
relations between these properties.
FLAC3D 6.0
Characteristic time:
(5)
Governing Differential Equations

The differential equations describing the fluid-thermal-mechanical response of a
porous material, for heat transported by conduction in the material domain, are
as follows.
Transport Law
The fluid transport is described by Darcy’s law:
(6)
where is the specific discharge vector, is fluid pore pressure, is the tensor of
absolute mobility coefficients (FLAC3D permeability tensor) of the medium,
is the relative mobility coefficient, which is a function of fluid saturation, , is
the fluid density, and , = 1,3 are the three components of the gravity vector.
For future reference, the quantity (where is the modulus

of the gravity vector) is defined as the head, and as the pressure head.
Heat flow by conduction is described by Fourier’s law of heat transport:
(7)
where is the heat flux vector, is temperature and is the thermal

conductivity tensor. Note that temperature, flux, convective and adiabatic
boundary conditions are considered in the FLAC3D formulation.
Balance Laws
For small deformations, the fluid mass balance may be expressed as
(8)
where is the volumetric fluid source intensity in [1/sec], and is the variation
of fluid content or variation of fluid volume per unit volume of porous material
due to diffusive fluid mass transport, as introduced by Biot (1956).
The thermal energy balance is expressed as
FLAC3D 6.0
(9)
where is the heat stored per unit volume of porous material, and is the
volumetric heat source intensity.
The balance of momentum has the form
(10)
where is the bulk density, and and are the densities of the dry
matrix and the fluid, respectively, is porosity, and is saturation.
Constitutive Laws
Changes in the variation of fluid content are related to changes in pore pressure,
, saturation, , mechanical volumetric strains, , and temperature, . The
response equation for the pore fluid is formulated as
(11)
where is Biot modulus [N/m2], is the porosity, is Biot coefficient and is

the undrained thermal coefficient [1/°C], which takes into account the fluid and
grain thermal expansions.
In the FLAC3D formulation, the influence of capillary pressure is neglected and

the fluid pressure is taken as zero when saturation is less than one. The relative
fluid mobility is related to saturation by a cubic law, which is zero for zero
saturation and one for full saturation:
(12)
Fluid flow in the unsaturated zone is thus solely driven by gravity. While the
influence of gravity is not required to saturate an initially dry medium, gravity
drives the process of desaturation. In this case, some level of residual saturation
is present because the apparent permeability, , goes to zero as the
saturation approaches zero.
The thermal constitutive law is expressed as
(13)
FLAC3D 6.0
where , is the mass density of the medium, and is the specific

heat at constant volume.
Note that for nearly all solids and liquids, the specific heats at constant pressure
and constant volume are essentially equal. Consequently, and can be used
interchangeably.
The constitutive response for the porous solid has the form (see notation
conventions of Theoretical Background)
(14)
where the quantity on the left-hand side of the equation is Biot effective stress,
is the co-rotational stress rate, is the functional form of the constitutive
law, is a history parameter, is the Kronecker delta, and is the strain rate.
The thermal strain rates are expressed as (see Thermal Strains)
(15)
where is the coefficient of linear thermal expansion.
In particular, the elastic relations that relate effective stresses to strains are
(small strain)
(16)
where the superscript refers to the initial state, is the strain, and and
are the bulk and shear moduli of the drained elastic solid.
Biot and Terzaghi Effective Stress
The expression for Biot effective stress, , may be derived from

compliance principles (see Effect of the Biot Coefficient); as such, it characterizes
the deformability of the solid. On the other hand, Terzaghi effective stress,
, measures the stress level sustained by the solid matrix; it is still being
used to detect failure in a plastic material (or potential failure in an elastic solid).
Compatibility Equations
The relation between strain rate and velocity gradient is
(17)
FLAC3D 6.0
Fluid Flow Boundary and Initial Conditions in FLAC3D

Initial conditions correspond to known pressure and saturation. Those two values
must be consistent with the FLAC3D formulation: pore pressure must be zero if
saturation is less than one, and vice-versa. (By default, in FLAC3D, initial pore
pressure is zero and saturation is one.)
Four types of boundary conditions are considered, corresponding to (1) given

pore pressure, (2) given component of the specific discharge normal to the
boundary, (3) leaky boundaries, and (4) impermeable boundaries.
In FLAC3D, boundaries are impermeable by default.
A leaky boundary condition has the form
(18)
where is the component of the specific discharge normal to the boundary in

the direction of the exterior normal, is the leakage coefficient in [m3/N-sec],
is the pore pressure at the boundary surface, and is the pore pressure in the
region to or from which leakage is assumed to occur.
Additional combined boundary conditions, such as uniform pressure for given

total influx over a boundary segment, can be imposed using FISH.
Note that saturation cannot be imposed as a boundary condition (this is different

from FLAC). Also, to keep the model fully saturated, the fluid tension may be set
to a large negative number.
FLAC3D 6.0
Numerical Formulation 493
Numerical Formulation
Substitution of the fluid mass balance equation in the constitutive equation for
the pore fluid yields the expression for the fluid continuity equation:
(1)
In the FLAC3D numerical approach, the flow domain is discretized into brick-
shaped zones defined by eight nodes. (The same discretization is used for
mechanical and thermal calculations, when applicable.) Both pore pressure and
saturation are assumed to be nodal variables. Internally, each zone is subdivided
into tetrahedra (2 overlays are used by default), in which the pore pressure and
the saturation are assumed to vary linearly.
The numerical scheme relies on a finite difference nodal formulation of the fluid
continuity equation. The formulation can be paralleled to the mechanical
constant stress formulation (presented in Finite Difference Approximation to
Space Derivatives) that leads to the nodal form of Newton’s law. It is obtained by
substituting the pore pressure, specific discharge vector and pore-pressure
gradient for velocity vector, stress tensor and strain-rate tensors, respectively.
The resulting system of ordinary differential equations is solved using an explicit
mode of discretization in time (default mode). An implicit fluid flow calculation
scheme is also available. However, it is only applicable to saturated flow
simulations (flow in which the unit saturation is not allowed to change).
The principal results are summarized below. The formulation, as it applies to

saturated flow, is described first. This formulation is then extended to apply to
saturated/unsaturated flow simulations. For now we note that the saturated/
unsaturated algorithm is based on variable substitution and upstream weighting
techniques and conserves fluid flow.
Starting from a state of mechanical equilibrium, a coupled hydromechanical

static simulation in FLAC3D involves a series of steps. Each step includes one or
more flow steps (flow loop) followed by enough mechanical steps (mechanical
loop) to maintain quasi-static equilibrium.
FLAC3D 6.0
The increment of pore pressure due to fluid flow is evaluated in the flow loop;
the contribution from volumetric strain is evaluated in the mechanical loop as a
zone value which is then distributed to the nodes.
For the effective stress calculation, the total stress increment due to pore-
pressure change arising from mechanical volume strain is evaluated in the
mechanical loop, and that arising from fluid flow is evaluated in the flow loop.
Note that in FLAC3D, all material models are expressed in terms of Terzaghi
effective stresses (i.e., effective stresses are used to detect failure in plastic
materials). In this context, the pore-pressure field may originate from different
sources: a fluid flow analysis; a coupled fluid/mechanical simulation; or an
initialization with the zone gridpoint initialize pore-pressure or zone water
command.
Saturated Fluid Flow

In this section, as a first step towards the derivation of the more general FLAC3D
fluid flow algorithm, we consider flow in a medium which remains fully
saturated at all times.
Finite-Difference Approximation to Space Derivatives

By convention, the tetrahedron nodes are referred to locally by a number from 1
to 4, face is opposite node , and the superscript relates to the value of the
associated variable on face .
A linear pore-pressure variation and constant fluid density are assumed within a
tetrahedron. The pressure head gradient, expressed in terms of nodal values of
the pore pressure by application of the Gauss divergence theorem, may be
written as
(2)
where is the exterior unit vector normal to face , is the face surface area
and is the tetrahedron volume.
For numerical accuracy, the quantity , where corresponds to the

coordinates of one of the tetrahedron’s corners, may be substituted for in the
expression of the pressure head in Equation (2) without affecting the value of the
pressure head gradient. Accordingly, Equation (2) takes the form
FLAC3D 6.0
(3)
where the nodal quantity is defined as
(4)
Nodal Formulation of the Mass Balance Equation

For fully saturated flow ( is constant and equal to one), the fluid continuity
(Equation (1)) may be expressed as
(5)
where
(6)
is the equivalent of the instantaneous “body forces,” , used in the mechanical

node formulation, and
(7)
First consider a single tetrahedron. Using this analogy, the nodal discharge,
[m3/s], = 1,4, equivalent to the tetrahedron-specific discharge and the
volumetric source intensity,, may be expressed as (see Finite Difference
Approximation to Space Derivatives)
(8)
where
(9)
and
(10)
In principle, the nodal form of the mass balance equation is established by

requiring that, at each global node, the sum of equivalent nodal discharges ( )
from tetrahedra meeting at the node (averaged over two overlays) and nodal
contribution ( ) of applied boundary fluxes and sources be zero.
FLAC3D 6.0
The components of the tetrahedron-specific discharge vector in Equation (9) are

related to the pressure-head gradient by means of the transport law (see
equation). In turn, the components of the pressure-head gradient can be
expressed in terms of the tetrahedron nodal pore pressures, using Equation (3).
In order to save computer time, local matrices are assembled. A zone formulation
is adopted, whereby the sum of contributions Equation (9) from all tetrahedra
belonging to the zone and meeting at a node, , is formed and averaged over two
overlays. Local zone matrices that relate the eight nodal values to the
eight nodal pressure heads are assembled. Because these matrices are
symmetrical, 36 components are computed; these are saved at the beginning of
the computation and updated every ten steps in large-strain mode. By definition
of zone matrices, we have
(11)
where is the local vector of nodal pressure heads for the zone.
In turn, global nodal values are obtained by superposition of zone

contributions. Taking some liberty with the notation, we write for each global
node ,
(12)
where is the global matrix and is, in this context, the global vector of
nodal pressure heads.
Returning to our previous consideration, we write
(13)
where, for simplicity of notation, the sign is used to represent summation of

the contributions at global node of all zones meeting at that node. (A zone
contribution consists of contributions of tetrahedra involved in the zone,
averaged over two overlays.) Using Equation (8) for and Equation (8) for in
Equation (13), we obtain, after some manipulation,
(14)
where is a function of the nodal pore pressures defined in Equation (12)

together with Equation (4), is the known contribution of applied volume
sources, boundary fluxes and point sources, having the form
FLAC3D 6.0
(15)
and is the thermal-mechanical nodal discharge contribution defined as
(16)
Equation (14) is the nodal form of the mass-balance equation at node ; the
right-hand side term is referred to as the out-of-balance
discharge. This term is made up of two contributions: the out-of-balance flow
discharge, ; and the out-of-balance thermal-mechanical discharge,
. When the fluid is at rest (the fluid-flow calculation is turned off), the
out-of-balance flow discharge vanishes and pore-pressure changes are caused
by mechanical and/or thermal deformations only. For flow calculation only (no
thermal-mechanical coupling), the out-of-balance thermal-mechanical
discharge is equal to zero.
In FLAC3D, the Biot modulus is a nodal property, and we may write, using
Equation (10),
(17)
where
(18)
After substitution of Equation (16) in Equation (14), we write, rearranging terms,
(19)
where
(20)
An expression such as Equation (19) holds at each global node involved in the
discretization. Together, these expressions form a system of ordinary differential
equations that is solved in FLAC3D, for given , using either explicit or
implicit (saturated flow only) finite-difference schemes. The domain of
application of each scheme is discussed below.
FLAC3D 6.0
Explicit Finite-Difference Formulation

In the explicit formulation, the value of the quantity at a node is assumed
to vary linearly over a time interval . In principle, the derivative in the left-
hand side of Equation (19) is expressed using forward finite differences, and the
out-of-balance flow-discharge is evaluated at time . Starting with an initial
pore-pressure field, nodal pore pressures at incremental time values are
updated, provided the pore pressure is not fixed, using the expression
(21)
where
(22)
(23)
and
(24)
Numerical stability of the explicit scheme can only be achieved if the timestep
remains below a limiting value.
Stability Criterion
To derive the stability criterion for the flow calculation, we consider the situation
in which a node, , in an assembly of zones is given a pore-pressure
perturbation, , from a zero initial state. Using Equation (12), we obtain
(25)
If node belongs to a leaky boundary, we have
(26)
where is used to represent the pressure coefficient in the global leakage term
at node .
After one fluid-flow timestep, the new pore pressure at node is (see Equation
(21) through Equation (23))
FLAC3D 6.0
(27)
To prevent alternating signs of pore pressures as the computation is repeated for

successive , the coefficient of in the preceding relation must be positive.
Such a requirement implies that
(28)
The value of the poro-mechanical timestep used in FLAC3D is the minimum

nodal value of the right-hand side in Equation (28), multiplied by the safety
factor of 0.8.
To assess the influence of the parameters involved, it is useful to keep in mind

the following representation of the critical timestep. If is the smallest
tetrahedron characteristic length, we may write an expression of the form
(29)
where is the critical timestep, is the fluid diffusivity, and is a constant,

larger than unity, that depends on the medium geometrical discretization (e.g.,
= 6 for a regular discretization in cubes; see Karlekar and Desmond (1982) for a
thermal analogy).
The critical timestep in Equation (29) corresponds to a measure of the

characteristic time needed for the diffusion “front” to propagate throughout the
tetrahedron. To estimate the time needed for the front to propagate throughout a
particular domain, a similar expression can be used, provided is interpreted as
the characteristic length of the domain under consideration. Consider the case
where no leakage occurs and the properties are homogeneous. By taking the ratio
of characteristic times for the domain and the tetrahedron, we see that the
number of steps needed to model the propagation of the diffusion process
throughout that domain is proportional to the ratio of square characteristic
lengths for the domain and the tetrahedron. That number may be so large that
the use of the explicit method alone could become prohibitive. However, in many
groundwater problems, the advantage of this first-order method is that the
calculated timestep is usually small enough to follow nodal pore-pressure
fluctuations accurately.
FLAC3D 6.0
Implicit Finite-Difference Formulation

The requirement that should be restricted in size to ensure stability
sometimes results in an extremely small timestep (of the order of a fraction of a
second), especially when transient flow in layers of contrasting properties is
involved. The implicit formulation eliminates this restriction, but it involves
solving simultaneous equations at each timestep and, besides, it only applies to
saturated fluid-flow simulations.
The implicit formulation in FLAC3D uses the Crank-Nicolson method, in which

the pore pressure at a node is assumed to vary quadratically over the time
interval . In this method, the derivative in Equation (19) is
expressed using a central difference formulation corresponding to a half
timestep, while the out-of-balance flow discharge is evaluated by taking average
values at and . In this formulation we have
(30)
and
(31)
where
(32)
is given by Equation (24), and
(33)
From Equation (12), we may write
(34)
and
(35)
For no variation of the gravity term during the time interval , Equation (30)
can be written using Equation (4).
(36)
Using this last expression and Equation (34), Equation (35) becomes
FLAC3D 6.0
(37)
After substitution of Equation (37) into Equation (31), we obtain, using Equation
(35),
(38)
Finally, regrouping terms:
(39)
For simplicity of notation, we define the known matrix and vector as
(40)
and
(41)
With this convention, we may write Equation (39) in the form
(42)
One equation holds for each of the nodes involved in the grid, and the resulting
system is solved in FLAC3D using Jacobi’s iterative method. In this approach,
pore-pressure increments at iteration and node are calculated using the
recurrence relation
(43)
where Einstein’s summation convention applies to index only. Using the

Equation (40) of , Equation (43) takes the form
(44)
The initial approximation is chosen such that
(45)
(Note that in FLAC3D, pressure-dependent boundary conditions (contained in )

are updated in the implicit iterative procedure.)
FLAC3D 6.0
Convergence Criterion
In FLAC3D a minimum of 3 and a maximum of 500 iterations are considered, and
the criterion for detection of convergence has the form
(46)
The magnitude of the timestep must be selected in relation to both convergence

and accuracy of the implicit scheme. Although the Crank-Nicolson method is
stable for all positive values of (for no leakage), the convergence of Jacobi’s
method is not unconditionally guaranteed unless the matrix is strictly
diagonally dominant:
(47)
for (see Dahlquist and Bjorck 1974). According to the definitions of

( Equation (40)) and ( Equation (32)), this sufficient condition is always
fulfilled for sufficiently small values of . If convergence of Jacobi’s method is
not achieved, an error message is issued. It is then necessary to either reduce the
magnitude of the implicit timestep, or use the explicit method. (The explicit
timestep may be used as a lower bound.)
Also, although the implicit method is second-order accurate, some insight may
be needed in order to select the appropriate timestep. Indeed, its value must
remain small compared to the wavelength of any nodal pore-pressure
fluctuation. Typically, the explicit method is used earlier in the run or in its
perturbed stages, while the implicit method is preferred for the remainder of the
simulation. (Alternatively, the implicit method could be used with the explicit
timestep value for extra accuracy.)
Remember that the implicit approach should not be selected for simulations in
which changes of saturation are to be expected. In addition, computation time
and computer memory are two factors that must be taken into consideration
when selecting the implicit approach in FLAC3D. In the implicit method, a set of
equations requiring a minimum of three iterations must be solved at each
timestep. The amount of calculation required for one iterative step is
approximately equal to that needed for one timestep in the explicit scheme. Also,
intermediate values must be stored in the iterative procedure, requiring extra
FLAC3D 6.0
memory to be allocated in comparison with the explicit scheme. Those

inconveniences, however, can be more than offset by the much larger timestep
generally permitted by the implicit method, or by the gain in accuracy.
Saturated/Unsaturated Flow
For saturated fluid flow, the nodal volumetric flow rates in a zone are
related to the nodal pore pressures by Equation (11), which may be expressed
in matrix notation as
(48)
where the matrix is a known function of the zone geometry and the
saturated value of the mobility coefficient. This relation was derived by
application of the Gauss divergence theorem for pore-pressure gradient in the
tetrahedron, taking into account energy considerations, and using superposition.
It is extended to the case of unsaturated flow in coarse soils (constant air
pressure, no capillary pressure) by application of modifications:
a. The gravity term is multiplied by the average zone saturation, , to

account for partial zone filling.
b. The nodal flow rates are multiplied by the relative mobility, (see
equation), which is a function of the average saturation at the inflow
nodes for the zone, . This upstream weighting technique is applied to
prevent unaccounted flow in the filling process of an initially dry medium
(numerical dispersion caused by the nodal definition of saturation). The
relative permeability function has the same form as this equation:
(49)
c. Nodal inflow rates are scaled according to local saturation.
For saturated/unsaturated flow, the explicit finite difference nodal formulation of

the fluid continuity equation (1) takes the form (see Equation (21) through
Equation (23))
(50)
FLAC3D 6.0
where the subscript and superscript have been omitted for simplicity of
notation, and stands for tetrahedron volume. At a saturated node, = 1 and
Equation (50) becomes
(51)
At a partially saturated node, = 0 and Equation (50) may be expressed as
(52)
The approach leads to a technique of variable substitution whereby the saturation

variable is updated using Equation (52) at an unsaturated node (where the pore
pressure is equal to zero), and Equation (51) is used to increment the pore-
pressure variable at a saturated node (where saturation is constant). Those
relations are applied in such a way that the fluid volume balance is preserved
(i.e., Equation (50) is satisfied).
The explicit fluid timestep in FLAC3D is selected on the basis of a stability

criterion derived using the saturated fluid diffusivity ( is the largest
principal value of saturated mobility coefficient tensor, is Biot modulus). (See
Equation (28) and Equation (29).) The time scale associated with fluid storage is
usually much smaller than the characteristic time scale for phreatic storage.
While an estimate of the first one (two-dimensional flow) can be calculated
using , where is to be interpreted as average flow path length, a rough
estimate of the second may be obtained from the relation , where
is the average height of the medium associated with phreatic storage, and is
the related mean pore-pressure change.
Mechanical Timestep and Numerical Stability

The presence of the fluid will increase the apparent mechanical bulk modulus of
the medium which, in turn, affects the magnitude of the nodal mass in the
density scaling scheme used for numerical stability in FLAC3D. An upper bound
for this apparent modulus may be found by considering undrained, saturated
isothermal poro-mechanical conditions for which = 0, = 0 and = 0. The
incremental expression of the fluid constitutive law then takes the form
(53)
FLAC3D 6.0
Adopting the incremental form of the elastic law to describe the volumetric part
of the stress-strain constitutive relation in a time interval, we can write
(54)
Substitution of Equation (53) in Equation (54) gives
(55)
This apparent tangent bulk modulus, , is used in place of when

calculating nodal inertial mass (see this equation in the Theory and Background
section).
Also note that in FLAC3D, the dry density, , of the medium is considered in the
input. The saturated density, , is calculated internally using the
input value for fluid density, , porosity, , and saturation, . Body forces are
then adjusted accordingly in Newton’s equations of motion.
Total Stress Correction

In the FLAC3D formulation, nodal pore pressures are calculated first. The zone
pore pressures are then derived by volume averaging of the tetrahedra values
(obtained as arithmetic averages of nodal values).
The total stresses are then adjusted, consistent with the definition of Biot
effective stress, by adding an increment to account for the contribution from
pore pressure changes (see this equation). The stress increment is expressed
as
(56)
where
(57)
is the contribution from flow, and
(58)
is the contribution from thermal-mechanical coupling. Also, in those

expressions, the overline symbol stands for zone average in the sense mentioned
above.
FLAC3D 6.0
Fully Saturated Fast Flow

In the standard fluid-flow formulation in FLAC3D the fluid continuity equation
for fully saturated flow of a compressible fluid is given (based upon this equation
for isothermal conditions) as
(59)
where is the variation of fluid content (with respect to a reference state),

defined as the variation of fluid volume per unit volume of porous material, is
the volumetric strain, is Biot modulus, and is Biot coefficient. The fluid
mass balance for the standard formulation is given in Governing Differential
Equations by this equation, the balance of momentum by this equation, and the
material constitutive law has the form given by this equation.
The standard calculation procedure is applicable to cases of relatively high

material stiffness (compared to the fluid stiffness) and average porosity. The
performance of the standard scheme (in terms of solution speed and accuracy)
degrades when a large value for Biot modulus (compared to the solid matrix
stiffness) is used. A high value is generally characterized by the equation
(60)
or, if = 1, and a combination of high fluid bulk modulus and low porosity is
considered,
(61)
is fluid bulk modulus, is porosity, and and are drained elastic moduli of
the solid matrix.
There are several disadvantages to working with the standard scheme with high
values of Biot or fluid bulk modulus (i.e., when Equation (60), or Equation (61),
holds):
1. The stable explicit fluid timestep can be extremely small, because

it is inversely proportional to Biot modulus.
2. Mechanical convergence is very slow, as a consequence of the

additional stiffness, , due to the effect of the fluid on mass
scaling.
FLAC3D 6.0
3. The accuracy of the solution is poor because the standard scheme

is not well-adapted to solve the degenerate form of the pore
pressure equation, Equation (59).
The poor performance of the standard scheme develops because, when is very
large, Equation (59) degenerates to the expression
(62)
where the quantity
(63)
is defined as the unbalanced volumetric strain rate. The standard scheme is not
well-designed to solve this degenerate form of the pressure-based equation.
An Alternative Fast-Flow Algorithm

An alternative strain-based formulation (saturated fast-flow logic) is better
adapted for Equation (59) for high values of Biot modulus.[1]
To illustrate the principle of this scheme, consider, for simplicity, the case of a
material constitutive law with elastic volumetric behavior, expressed as
(64)
In this scheme, the quantity is evaluated from Equation (63) after

performing a flow step and a mechanical step. In general, the quantity will not be
negligible, and an iterative scheme is applied to reduce the value of the quantity
and enforce Equation (62) . First, the impact on Biot effective stress of the
“unbalanced” volumetric strain rate contribution is evaluated; this quantity,
, is interpreted as a pore pressure increment, , which is then used to
“correct” the (total) stress (see Equation (64) ), using
(65)
where
(66)
FLAC3D 6.0
The procedure is repeated in the course of a series of mechanical steps until the
unbalanced volumetric strain increment falls below a negligible value. Gridpoint
pore pressures are also updated to ensure proper feedback in the fluid-
mechanical calculation.
For stability and convergence of the scheme, the magnitude of the stress
correction increment cannot be applied in full at each calculation
step. Instead, the accumulated quantity is leaked gradually into the grid using a
“balloon” technique.[2] In this technique, an unbalanced volume, , is
calculated and stored in virtual containers (balloons) associated with each
gridpoint (or node). There are two contributions to the unbalanced volume at a
gridpoint: one from change of fluid content and one from volumetric strain
increment of zones meeting at the gridpoint. The content of the balloon is
updated as a result of volumetric deformation, after each mechanical step, and
unbalanced flow, after each fluid-flow step. (The variation of fluid content is
evaluated at gridpoints using the fluid mass balance equation, as in the standard
scheme.)
The unbalanced volumetric strain expression for the zone is calculated using
(67)
where is zone volume, and the unbalanced volume for the zone, , comes
from the contributions of the gridpoints associated with the zone. The pore
pressure increment used to correct normal stresses in a mechanical step (see
(65)) is calculated in proportion to the current unbalanced volumetric strain:
(68)
where is the relaxation coefficient. Also, the quantity is distributed

from one zone to balloons at gridpoints (using weighting functions) after each
mechanical step. Finally, the pore pressure at a gridpoint is updated using the
nodal form of (68):
(69)
where is the gridpoint volume.
The relaxation coefficient, , is a dimensionless number which must be small

enough to enforce convergence of the fast-flow scheme. The default value is
0.004. This value should be adequate to enforce numerical stability for most
FLAC3D 6.0
problems. It is possible to reset the value using the zone fluid fastflow-
relaxation command. (Additional information and recommendations for are
given in Section 1.4.1.1 of the Fluid-Mechanical Interaction volume of the FLAC
Version 7 manual.)
In an undrained simulation, the only contribution to unbalanced volumetric

strain in the balloon comes from mechanical volumetric changes. The volumetric
changes are converted gradually into pore pressure changes during the
mechanical iterations. The process is interrupted when the balloon is emptied.
In a fluid-mechanical simulation, the flow contribution to unbalanced volumetric

strain is added in the balloon during the flow calculation step, and enough
mechanical steps are taken afterwards to dissipate this contribution and that
from the mechanical volumetric changes.
During cycling using the saturated fast-flow logic, the average unbalanced volume
ratio, , and maximum unbalanced volume ratio, , are monitored to detect
convergence of the scheme:
(70)
(71)
where the summation and maximum are taken over all gridpoints in the model,
is the average bulk modulus of the zones meeting at the gridpoint, and is
the maximum value of the stress norm in the model. The balloon is considered
empty when both and fall below . This default value can be changed
with the model solve unbalanced-average and model solve unbalanced-maximum
commands.
The saturated fast-flow scheme is applicable only for the case of an

incompressible fluid. Otherwise, the standard scheme should be used (i.e., when
(60), or (61), does not apply). Therefore, the two schemes complement each
other.
When (60), or (61), does apply, the coupled, incompressible saturated fast-flow
scheme has several advantages over the standard scheme:
FLAC3D 6.0
1. Large fluid timesteps can be taken. The dependence of the stable

explicit timestep on the inverse of Biot modulus (or on porosity
and inverse of fluid bulk modulus) is removed in this scheme. The
stable explicit timestep is inversely proportional to in the
saturated fast-flow scheme. In fact, the Biot modulus, , in
Equation (59) is replaced by in the fast-flow scheme (see
(69) and (63)).
2. Convergence is much faster. There is no need to account for fluid

stiffness during the mass scaling calculation in the fast-flow
scheme. This allows much faster mechanical convergence.
3. There is better accuracy. The degenerate form of the coupled

fluid-mechanical equations, which apply for large values of Biot
modulus, are solved directly in the fast-flow scheme. This
provides a more accurate solution.
At present, the saturated fast-flow scheme cannot be used for several conditions:
1. unconfined fluid flow, in which a phreatic surface, or water table,

exists or develops;
2. models with interfaces;
3. models with attached grids; and
4. dynamic analysis.
Two verification problems for the saturated fast-flow logic are provided. See One-
Dimensional Consolidation and Consolidation Settlement at the Center of a Strip
Load.
Endnotes
[1] This scheme is derived from the saturated fast-flow scheme already available
in FLAC. See Section 1.4.1 of the Fluid-Mechanical Interaction volume of the
FLAC Version 7 manual for reference.
[2] See Section 1.4.1.1 of the Fluid-Mechanical Interaction volume of the FLAC
Version 7 manual for reference.
FLAC3D 6.0
Calculation Modes for Fluid-Mechanical Interaction 511
Calculation Modes for Fluid-Mechanical

Interaction
The calculation mode and commands required for a pore-pressure analysis
depend on whether the grid has been configured for fluid flow (i.e., whether the
model configure fluid command has been specified).
The flow calculation can be performed independent of, or coupled to, the
mechanical deformation calculation when model configure fluid is specified.
It is also possible to conduct simplified, uncoupled, fluid-mechanical

calculations with FLAC3D such as slope stability analyses, without configuring
the grid for fluid flow. This option, when applicable, provides faster solution
than the fluid flow configuration.
Grid Not Configured for Fluid Flow and Grid Configured for Fluid Flow reflect the
two possibilities. For convenience, all commands described below are
summarized in Input Instructions for Fluid-Flow Analysis, at the end of this
section.
Grid Not Configured for Fluid Flow

If the command model configure fluid has not been given, then a fluid-flow
analysis cannot be performed, but it is still possible to assign pore pressure at
gridpoints. In this calculation mode, pore pressures do not change, but failure,
which is controlled by the effective-stress state, may be induced when plastic
constitutive models are used.
A pore-pressure distribution can be specified at gridpoints, with the zone

gridpoint initialize pore-pressure command with a gradient, or with the zone
water plane (or zone water set) command.
If the zone water plane command is used, a hydrostatic pore-pressure

distribution is calculated automatically by the code, below the given water table
level. In this case, the fluid density (zone water density) and gravity (model
gravity) must also be specified. The fluid density value and water table location
FLAC3D 6.0
can be printed with the zone list fluid-density command, and the water table
can be plotted, if the water table is defined by zone water set or zone water plane
command.
In both cases, zone pore pressures are calculated by averaging from the gridpoint
values, and used to derive effective stresses for use in the constitutive models. In
this calculation mode, the fluid presence is not automatically accounted for in
the calculation of body forces: wet and dry medium densities must be assigned
by the user, below and above the water level, accordingly. The commands zone
gridpoint list pore-pressure and zone list pore-pressure print gridpoint and
zone pore pressures, respectively. plot add zone contour pore-pressure plots
contours of gridpoint pore pressures.
A simple example application of this calculation mode is given in the topic The
Effect of Water in the Problem Solving section.
Grid Configured for Fluid Flow

If the command model configure fluid is given, a transient fluid-flow analysis
can be performed, and change in pore pressures, as well as change in the
phreatic surface, can occur. Pore pressures are calculated at gridpoints, and zone
values are derived using averaging. Both effective-stress (static pore-pressure
distribution) and undrained calculations can be carried out in model configure
fluid mode. In addition, a fully coupled analysis can be performed, in which
changes in pore pressure generate deformation, and volumetric strain causes the
pore pressure to evolve.
If the grid is configured for fluid flow, dry densities must be assigned by the user
(both below and above the water level), because FLAC3D takes the fluid influence
into account in the calculation of body forces.
A fluid-flow model must be assigned to zones when running in model configure

fluid mode. Isotropic flow is prescribed with the zone fluid cmodel assign
isotropic command; anisotropic flow with the command zone fluid cmodel
assign anisotropic. An impermeable material is assigned to zones with the zone
fluid cmodel assign null command. Note that zones that are made null
mechanically are not automatically made null for fluid flow.
FLAC3D 6.0
Fluid properties are assigned to either zones or gridpoints. Zone fluid properties
include isotropic permeability, porosity, Biot coefficient, undrained thermal
coefficient and fluid density. Zone fluid properties are assigned with the zone
fluid property command, with the exception of fluid density, which is assigned
with the zone initialize fluid-density command. (Note that fluid density can
also be specified globally with the zone water density command.)
For isotropic flow, permeability is specified using the permeability property

keyword. For anisotropic flow, the three principal values of permeability are
specified using the property keywords permeability-1, permeability-2 and
permeability-3, and the orientation is defined using the keywords dip, dip-
direction, and rotation. The principal directions of permeability correspond to
, , , and form a right-handed system. The properties dip and dip-direction
are the dip angle and dip direction angle of the plane in which and are
defined. The dip angle is measured from the global -plane, positive down (in
negative -direction). The dip direction angle is the angle between the positive
-axis and the projection of the dip-direction vector on the -plane (positive
clockwise from -axis). The property rotation is the rotation angle between the
-axis and the dip vector in the - -plane (positive clockwise from dip-
direction vector). See Figure 1.
FLAC3D 6.0
Figure 1: Definitions for principal directions of permeability tensor.
Biot coefficient is assigned using the biot keyword; porosity is assigned using the
porosity keyword. If not specified, Biot coefficient = 1 and porosity = 0.5, by
default.
FLAC3D 6.0
Gridpoint fluid properties are assigned with the zone gridpoint initialize
command. These properties include fluid bulk modulus, Biot modulus, fluid
tension limit and saturation. Each of the gridpoint properties can also be given a
spatial variation.
Table 1 summarizes the ways the various properties can be specified. The fluid
properties are described further in Properties and Units for Fluid-Flow Analysis.
Table 1: Property specification methods
property keyword specific command

Permeability permeability in zones zone fluid property
(isotropic) permeability
Perm. princ. permeability-1 in zones zone fluid property

val. (aniso) permeability-1


Perm. princ. permeability-xx in zones zone fluid property

val. (aniso) permeability-xx
Perm. princ. permeability-xy in zones zone fluid property

val. (aniso) permeability-xy
Perm. princ. permeability-xz in zones zone fluid property

val. (aniso) permeability-xz
Perm. princ. permeability-yy in zones zone fluid property

val. (aniso) permeability-yy
Perm. princ. permeability-yz in zones zone fluid property

val. (aniso) permeability-yz
Perm. princ. permeability-zz in zones zone fluid property

val. (aniso) permeability-zz
Perm. princ. dip in zones zone fluid property dip

dir. (aniso)
Perm. princ. did-direction in zones zone fluid property dip-
dir. (aniso)
FLAC3D 6.0
property keyword specific command

direction
Perm. princ. rotation in zones zone fluid property rotation

dir. (aniso)
Porosity porosity in zones zone fluid property porosity
Biot biot in zones zone fluid property biot

coefficient
Undrained undrained- in zones zone fluid property undrained-
thermal thermal- thermal-coefficient
coeff. coefficient
Fluid fluid-modulus at zone gridpoint initialize

modulus gridpoints fluid-modulus
Biot modulus biot-modulus at zone gridpoint initialize

gridpoints biot-modulus
Saturation saturation at zone gridpoint initialize

gridpoints saturation
Fluid tension fluid-tension at zone gridpoint initialize

limit gridpoints fluid-tension
Fluid density fluid-density at zones zone initialize fluid-density
density globally zone water density
Note that fluid compressibility is defined in one of two ways in the model
configure fluid mode: 1) Biot coefficient and Biot modulus are specified; or 2)
fluid bulk modulus and porosity are specified. The first case accounts for the
compressibility of the solid grains (Biot coefficient is set to 1 for incompressible
grains). In the second case, solid grains are assumed to be incompressible (see
Biot Coefficient and Biot Modulus).
The zone properties can be printed with the zone fluid list property command,
and the gridpoint properties can be printed with the zone gridpoint list
command. Fluid density, along with the location of the water table (if specified)
can be printed with the zone water list command. The fluid-flow properties can
be plotted as a contour plot. For anisotropic flow, the global components of the
permeability tensor are available for plotting and printing, using the zone
FLAC3D 6.0
property keywords permeability-xx, permeability-yy, permeability-zz,

permeability-xy, permeability-xz and permeability-zz (please note that these
global components cannot be initialized directly).
An initial gridpoint pore-pressure distribution is assigned the same way for the
model configure fluid mode as for the non-model configure fluid mode (i.e.,
either with the zone gridpoint initialize pore-pressure command or the zone
water set or zone water plane command). Pore pressures can be fixed (and freed)
at selected gridpoints with the zone gridpoint fix pore-pressure (or zone
gridpoint free pore-pressure) command. Fluid sources and sinks can be applied
with the zone apply well command. See describes the various Fluid-Flow
Boundary and Initial Conditions.
The fluid-flow solution is controlled by the zone fluid active on and model solve
commands. Several keywords are available to help the solution process. For
example, zone fluid active on or off turns on or off the fluid flow calculation
mode. The application of these commands and keywords depends on the level of
coupling required in the fluid flow analysis. The section in Solving Flow-Only
and Coupled-Flow Problems describes the various coupling levels and the
recommended solution procedure. Example applications, ranging from flow-only
to coupled mechanical-flow calculations, are also described in this section.
Results from a fluid-flow analysis are provided in several forms. The commands
zone gridpoint list pore-pressure and zone list pore-pressure print gridpoint
and zone pore pressures, respectively. Histories of gridpoint and zone pore
pressures can be monitored with the command zone history pore-pressure and
specify source.gridpoint and source.zone, respectively. And for a transient
calculation, the pore pressure can be plotted versus real time by monitoring the
flow time with the model history fluid time-total command. plot add zone
contour pore-pressure plots contours of gridpoint pore pressures. plot add zone
contour saturation plots contours of saturation. The plot add flow command
plots specific discharge vectors. General information on the model configure
fluid calculation mode is printed with the model fluid list command. Several
fluid-flow variables can be accessed through FISH. These are listed in Input
Instructions for Fluid-Flow Analysis. There is one grid-related variable, gp.flow,
that can only be accessed through a FISH function; this corresponds to the net
inflow or outflow at a gridpoint. The summation of such flows along a boundary
where the pore pressure is fixed is useful because it can provide a value for the
total outflow or inflow for a system.
FLAC3D 6.0
FLAC3D 6.0
Properties and Units for Fluid-Flow Analysis 519
Properties and Units for Fluid-Flow Analysis

The properties that relate to fluid flow in FLAC3D are the permeability coefficient,
, the fluid mass density, , and either the Biot coefficient, , and Biot modulus,
(for flow through a material with compressible grains), or the fluid bulk
modulus, , and porosity, (for flow through a material with incompressible
grains only). The thermal-coupling parameters are the coefficient of drained
linear thermal expansion, (thermal-mechanical), and the undrained thermal
coefficient, (poro-thermal). These properties are examined below.
All thermal-poro-mechanical quantities must be given in a consistent set of

units. No conversions are performed by the program.
Permeability Coefficient
The isotropic permeability coefficient, (e.g., m2/(Pa/sec) in SI units), used in
FLAC3D is also referred to in the literature as the mobility coefficient. It is the
coefficient of the pressure term in Darcy’s law and is related to the hydraulic
conductivity, (e.g., m/s), by the expression
(1)
where is the gravitational acceleration.
The intrinsic permeability, (e.g., in m2), is related to :
(2)
where is the dynamic viscosity (e.g., units of N-sec/m2).
Equation (1) or eq:eqfluid-prop-k2 may be used to derive (required by FLAC3D)

from either , in velocity units, or , in [length]2 units (remember that must
end up with units [L3T/M] – e.g., in SI units this would be m3 sec/kg (or m2/Pa-
sec))). Using the values = 1.011 × 10-3 kg/(m-sec) for water at 20° C, = 9.81
× 103 Pa/m and 1 Darcy = 10-8 cm2, conversions may be derived to calculate in
SI units for water in FLAC3D:
FLAC3D 6.0
(in SI units) (in cm2) × 9.9 × 10-2
(in SI units) (in cm/sec) × 1.02 × 10-6
(in SI units) permeability in millidarcies × 9.8 × 10-13
If there is a variation of permeability across the grid, the timestep will be

controlled by the largest permeability (see this equation). For problems in which
steady state (but not transient behavior) is required, it may be beneficial to limit
the variations in permeability to improve convergence speed. For example, there
will probably be little difference in the final state between systems where there is
a 20:1 variation in permeability, compared to a 200:1 variation, when the flow is
moving entirely from one permeability region to the other.
The permeability coefficient (scalar or tensor) is a zone property specified using

the zone fluid property command (see Grid Configured for Fluid Flow).
Mass Density
Three different mass densities may be given as input to FLAC3D in different
circumstances: the dry density of the solid matrix, ; the saturated density of
the solid matrix, ; and the density of the fluid, . Note that densities are only
required if gravitational loading is specified.
If FLAC3D is configured for fluid flow (model configure fluid), then the dry
density of the solid material must be used. FLAC3D will compute the saturated
density of each element using the known density of the fluid, the porosity, , and
the saturation, : .
The only case when the saturated density is given as input is for an effective
stress calculation (static pore-pressure distribution) not carried out in model
configure fluid mode. The zone water command (or zone gridpoint initialize
pore-pressure command) specifies the location of the water table. The dry
density is specified for zones above the water table, and the saturated density for
zones below.
The solid density (dry or saturated) is given using the zone property density
command. The fluid density can be imposed globally using the zone water
density command, or it can be allowed to vary with position by using the zone
initialize fluid-density command. However, this capability of assigning spatial
FLAC3D 6.0
variation of fluid density must only be used with extreme caution because fluid
mass density is not advected in FLAC3D. All densities are zone variables in
FLAC3D, and are mass densities (e.g., kg/m3 in SI units).
Fluid Moduli
Biot Coefficient and Biot Modulus

The Biot coefficient, , is defined as the ratio of the fluid volume gained (or lost)
in a material element to the volume change of that element when the pore
pressure is changed. It can be determined in the same drained test as that used to
determine the drained bulk modulus, , of the material. Its range of variation is
between and 1, where is the porosity. In the particular case of an
incompressible solid constituent, = 1. This value is the default value adopted by
FLAC3D.
For an ideal porous material, the Biot coefficient is related to the bulk modulus of
the solid component :
(3)
The Biot modulus, , is defined as
(4)
where is the undrained bulk modulus of the material.
For an ideal porous material, the Biot modulus is related to the fluid bulk
modulus, :
(5)
where is the porosity. Thus, for an incompressible solid constituent ( = 1),
(6)
The calculation mode for compressible grains is turned on with the zone fluid
biot on command. The Biot coefficient is a zone property specified using the zone
fluid property biot command. The Biot modulus is a gridpoint variable specified
using the zone gridpoint initialize biot-modulus command.
FLAC3D 6.0
Fluid Bulk Modulus

In analyses where the grain compressibility can be neglected, the user has the
choice to either use the default value of Biot coefficient (i.e., = 1) and assign a
value equal to to Biot modulus, or give the fluid bulk modulus as input.
The bulk modulus, , is defined as
(7)
where is the change in pressure for a volumetric strain of .
When the fluid modulus is given as an input, the Biot modulus is computed
internally using Equation (7) for incompressible grains. In this calculation, the
porosity (a zone property) is evaluated at the nodes using nodal volume
averaging. he Biot coefficient is then set to 1 throughout the flow domain,
irrespective of any value given for that property. The fluid calculation for
incompressible grains is the default calculation in FLAC3D (i.e., zone fluid biot
off). The fluid modulus is a gridpoint variable specified using the zone gridpoint
initialize fluid-modulus command.
The “compressibility” of the fluid is the reciprocal of (i.e., ). For

example, for pure water at room temperature, 9
= 2 × 10 Pa, in SI units. In real
soils, pore water may contain some dissolved air or air bubbles, which
substantially reduce its apparent bulk modulus. For example, Chaney (1978)
reports that the fluid modulus can be reduced by an order of magnitude for an
air/water mixture at 99% saturation in compacted sand. For groundwater
problems, the bulk modulus of water may be different in different parts of the
grid, to account for the varying amounts of air present. A FISH function may be
used to change local moduli according to some law (e.g., the modulus may be
made proportional to pressure to represent a gas), but the user should be careful
not to make nonphysical assumptions.
Fluid Moduli and Convergence

If steady-state, fully saturated flow is required, the modulus (or ) is
unimportant to the numerical convergence process, because the response time of
the system and the timestep are both inversely proportional to (or ): the
same number of steps are necessary, independent of (or ). For systems
containing a phreatic surface, however, a low bulk modulus will speed
convergence to steady state, because the calculation for saturation change
FLAC3D 6.0
involves , not the product (or ) (see this equation and this
equation). Systems in which solid/fluid interaction is important are more
complicated to assess (some guidelines may be found in Time Scale). However, a
high value of (or ) compared to the mechanical will lead to slowly
converging solutions. In any case, from a numerical point of view, it is not
necessary to use values of (or ) that are larger than 20 times
(or ) in the simulation (see Time Scale, this equation, and 1D
Consolidation for an example).
Fluid Moduli for Drained and Undrained Analyses

In FLAC3D, whenever the fluid bulk modulus (or Biot modulus) is selected and
model configure fluid is specified, the drained bulk modulus must be specified
for the solid matrix. The apparent (undrained) bulk modulus of the solid matrix
will then be computed, and will evolve with time.
An undrained analysis may also be performed without specifying model configure

fluid. In this case, the undrained bulk modulus for the solid matrix should be
specified. The undrained bulk modulus is
(8)
For an incompressible solid constituent: = 1, , and this formula

becomes
(9)
Porosity
Porosity, , is a dimensionless number defined as the ratio of void volume to
total volume of an element. It is related to the void ratio, , by
(10)
The default value of , if not specified, is 0.5. should be given as a positive

number between 0 and 1, but small values (e.g., less than 0.2) should be used
with great caution because the apparent stiffness of the pore fluid is proportional
to . For low values of , the stiffness may become very large in comparison
to the stiffness of the solid material, causing the FLAC3D solution to take a very
long time to converge. Consider reducing , in this case (see Fluid Moduli and
Convergence and Time Scale for further guidance).
FLAC3D 6.0
Porosity is used by FLAC3D to calculate the saturated density of the medium and
evaluate Biot modulus in the case when the fluid bulk modulus is given as an
input. FLAC3D does not update porosity during the calculation cycle, since the
process is time-consuming and only the slope of the transient response is
affected. Porosity is a zone property specified using the zone fluid property
porosity command.
Saturation
Saturation, , is defined as the ratio of pore volume occupied by fluid to total pore
volume. In FLAC3D’s formulation, pore pressure is set to zero if the saturation at
any point is less than 1. The effect of dissolved and trapped air may be allowed by
reducing the local fluid modulus while keeping the saturation at 1 (i.e., we
imagine that there is an equivalent fluid present throughout the pore space).
Although no pore pressures are present in a partially saturated region, the
trapped fluid still has weight (i.e., body forces act), and the fluid moves under
the action of gravity (at a reduced apparent permeability — see this equation).
The initial saturation may be given by the user, but it is also updated during
FLAC3D’s calculation cycle as necessary to preserve the mass balance. Note that
in FLAC3D, saturation is not considered as an independent variable; it cannot be
fixed at any gridpoint.
Undrained Thermal Coefficient

The undrained thermal coefficient, , is defined as the pore-pressure variation
divided by 3 per unit temperature change in an undrained constrained test
(no deformation).
For an ideal porous material, this coefficient is related to the volumetric

thermal-expansion coefficients for the grains, , and fluid, , through the
formula
(11)
The undrained thermal coefficient is a zone property, specified using the zone
fluid property undrained-thermal-coefficient command.
FLAC3D 6.0
Fluid Tension Limit

In fine soils, the pore water may be able to sustain a significant tension. In
FLAC3D, negative pore pressures can develop, by default, up to a limit beyond
which desaturation will take place. The negative pore pressure limit is set using
the zone gridpoint initialize fluid-tension command, where the value is a
negative number (set to -1015, by default). Note that a negative pore pressure is
not the same as “tension” due to capillary, electrical or chemical forces. The
latter forces are best represented by an increased effective stress within the
constitutive model. Negative fluid pressures are unrelated to the fact that a
material is composed of grains – the pressures simply arise from the expansion
of a volume filled with fluid.
FLAC3D 6.0
FLAC3D 6.0
Fluid-Flow Boundary Conditions, Initial Conditions, Sources and Sinks 527
Fluid-Flow Boundary Conditions, Initial

Conditions, Sources and Sinks
Boundaries are impermeable by default; all gridpoints are initially “free” (i.e.,
the pore pressure at such gridpoints is free to vary according to the net inflow
and outflow from neighboring zones). This condition may be declared explicitly
by using the command zone gridpoint free pore-pressure at the appropriate
gridpoints. The opposite condition, zone gridpoint fix pore-pressure, may also
be set at any gridpoint. In general, fluid may enter or leave the grid at an
external boundary if the pore pressure is fixed. The effect of these two boundary
conditions is summarized:
1. Pore-pressure free — This is an impermeable boundary, and is the

default condition. There is no exchange between the grid and the outside
world. Pressure and saturation changes are computed according to this
equation or this equation, depending on the current value of saturation
and whether the fluid has “cavitated” (fallen below the tensile limit).
2. Pore-pressure fixed — This is a permeable boundary across which

fluid flows to and from the outside world. The saturation may only vary
if the set pressure is exactly zero. Otherwise, saturation is forced to be 1
(to conform with FLAC3D’s assumption that pore pressures can only exist
in a fully saturated material). Pore pressure cannot be fixed at a value
less than the tensile limit; FLAC3D will reset any such pressures to the
tensile limit.
As noted above, certain combinations of conditions are impossible (e.g., pore

pressure fixed at a value less than the tensile limit). FLAC3D “corrects” such
conditions before executing any calculation step. Pore pressure is fixed at some
pressure by using the zone gridpoint fix pore-pressure command. Alternatively,
the command zone face apply pore-pressure can be used for either external or
internal boundaries. It is important to note that the optional keyword interior
must be specified if the condition is to be applied at a non-surface node. The use
of zone face apply pore-pressure has the advantage that the pressure may be
controlled directly by a history (supplied by a FISH function).
FLAC3D 6.0
Fluid-flow boundary conditions can be applied to individual, or ranges of,

gridpoints, zone faces or zones via the zone face apply pore-pressure command.
The zone gridpoint fix well command causes a prescribed inflow or outflow to
be applied to a boundary gridpoint. If the interior keyword is added, the
condition can be applied to an interior gridpoint. The zone face apply discharge
and zone face apply leakage commands specify either a fluid flux or a leaky
boundary condition to faces of boundary zones. The zone apply well command
applies a volume rate of flow to zones within a specified range. All of these fluid-
flow boundary conditions except zone face apply leakage can refer to a history
(via the history keyword). See Input Instructions for Fluid-Flow Analysis for a
more detailed description and format for these commands.
Fixed-pressure gridpoints may act as a source or sinks. There is no explicit

command that can be used to measure the inflow or outflow at these points.
However, the FISH grid variable gp.flow records unbalanced nodal flows: a simple
function that allows inflows and outflows to be printed or plotted for any range
of gridpoints can be written.
Initial distributions of pore pressure, porosity, saturation and fluid-flow

properties may be specified with the zone gridpoint initialize or zone fluid
property command, as noted in Input Instructions for Fluid-Flow Analysis. If
gravity is also given (with the model gravity command), it is important that the
initial distributions are consistent with the gravitational gradient implied by the
value of gravity, the given density of water and the values of saturation and
porosity within the grid. If the initial distributions are inconsistent, then flow
may occur in all zones at the start of a run. When setting up a simulation, a few
steps should be taken to check for this possibility. Stress and Pore-Pressure
Initialization with a Phreatic Surface provides a detailed description of the
procedure for setting up initial conditions.
If a model contains interfaces, effective stresses will be initialized along the

interfaces (i.e., the presence of pore pressures will be accounted for within the
interface stresses when stresses are initialized in the grid). This occurs either
with or without model configure fluid specified. For example, the zone water
command (in non-model configure fluid mode) will include pore pressures
along the interface, because pore pressures are defined at gridpoints for
interpolation to interface nodes for this mode. Note that flow takes place,
FLAC3D 6.0
Fluid-Flow Boundary Conditions, Initial Conditions, Sources and Sinks 529
without resistance, from one surface to the other surface of an interface, if they
are in contact. Preferential flow along an interface (e.g., fracture flow) is not
computed.
FLAC3D 6.0
FLAC3D 6.0
Solving Flow-Only and Coupled-Flow Problems 531
Solving Flow-Only and Coupled-Flow Problems

FLAC3D has the ability to perform both flow-only and coupled fluid-mechanical
analyses. Coupled analysis may be performed with any of the built-in mechanical
models in FLAC3D.
Several modeling strategies are available to approach the coupled processes. One
of these assumes that pore pressures once assigned to the grid do not change;
this approach does not require any extra memory to be reserved for flow
calculation. The commands associated with this mode are discussed in Grid Not
Configured for Fluid Flow. Modeling strategies involving flow of fluid require
that the model configure fluid command be issued to configure the grid for fluid
analysis, and the model configure fluid command be issued for all zones in
which fluid flow can occur. Note that for coupled analysis, the fluid-flow model
is not made null automatically for zones that are made null mechanically. The
zone fluid cmodel assign null command must be given for fluid null zones. The
commands associated with model configure fluid mode are discussed in Grid
Configured for Fluid Flow.
The different modeling strategies for coupled analysis will be illustrated in the
following sections, the more elaborate requiring more computer memory and
time. As a general rule, the simplest possible option should be used, consistent
with the reproduction of the physical processes that are important to the
problem at hand. Recommended guidelines for selecting an approach based on
time scales are given in Selection of a Modeling Approach for a Fully Coupled
Analysis, and various modeling strategies are described in Fixed Pore Pressure to
Coupled Flow and Mechanical Calculations.
Time Scales
When planning a simulation involving fluid flow or coupled flow calculations
with FLAC3D, it is often useful to estimate the time scales associated with the
different processes involved. Knowledge of the problem time scales and
diffusivity help in the assessment of maximum grid extent, minimum zone size,
timestep magnitude and general feasibility. Also, if the time scales of the
different processes are very different, it may be possible to analyze the problem
FLAC3D 6.0
using a simplified (uncoupled) approach. (This approach for fully coupled

analyses is discussed in detail in Selection of a Modeling Approach for a Fully
Coupled Analysis.)
Time scales may be appreciated using the definitions of characteristic time given
below. These definitions, derived from dimensional analysis, enter the
expression of analytical continuous source solutions. They can be used to derive
approximate time scales for the FLAC3D analysis.
Characteristic time of the mechanical process
(1)
where is undrained bulk modulus, is shear modulus, is mass density, and

is characteristic length (i.e., the average dimension of the medium).
Characteristic time of the diffusion process
(2)
where is the characteristic length (i.e., the average length of the flow path
through the medium), and is the diffusivity defined as mobility coefficient
divided by storativity :
(3)
There are different forms of storativity that apply in FLAC3D depending on the
controlling process:
1. fluid storage
(4)
2. phreatic storage
(5)
3. elastic storage
FLAC3D 6.0
(6)
where is Biot modulus, is Biot coefficient ( = and = 1 for

incompressible grains), is drained bulk modulus, is shear modulus, is
fluid density, is gravity and is characteristic storage length (i.e., the average
height of the medium available for fluid storage).
For saturated flow-only calculations (rigid material), is fluid storage and is

the fluid (Equation (3) and (4)):
(7)
For unsaturated flow calculations, is phreatic storage and the diffusivity (from
Equation (3) and (5)) is
(8)
For a coupled, saturated, deformation-diffusion analysis with FLAC3D, is

elastic storage and is the true or generalized coefficient of consolidation, defined
from Equation (3) and (6) as
(9)
There are some noteworthy properties based on the preceding definitions:
1. Because explicit timesteps in FLAC3D correspond to the shortest

time needed for the information to propagate from one gridpoint
to the next, the magnitude of the timestep can be estimated using
the smallest zone size for in the formula for characteristic
time. It is important to note that the explicit fluid flow timestep
in FLAC3D is calculated using the fluid diffusivity (Equation (7)),
even in a coupled simulation. The timestep magnitude may thus
be estimated from the formula obtained after substitution of
Equation (7) in Equation (2), and using the smallest zone size
for :
(10)
FLAC3D 6.0
2. In a saturated fluid-flow problem, a reduced bulk modulus (i.e.,

or ) leads not only to an increased timestep, but also to an
increased time to reach steady state, so that the total number of
steps, , stays the same. This number may be estimated by
taking the ratio of the characteristic times for the model, , to the
critical timestep, , using Equation (2), (7) and (10), which gives
(11)
where and are characteristic length for the model and the
smallest zone.
3. In a partially saturated fluid-flow problem, adjustments can be

made to the fluid modulus ( or ) to speed convergence to
steady state. Be careful not to reduce (or ) so much that
numerical instability results. A condition for stability may be
derived from the requirement that the fluid storage (used in the
critical timestep evaluation) must remain smaller than the
phreatic storage over one zone height, :
(12)
where is an adjustment factor chosen equal to 0.3.
4. Using Equation (2), we see that, to avoid any boundary effects in

diffusion problems, the characteristic length, , of the model
must be larger than the dimension , where is the maximum
simulation time and is the controlling diffusivity. In turn, the
minimum simulation time is controlled by the relation
.
5. In a coupled flow problem, the true diffusivity is controlled by the

stiffness ratio (i.e., the stiffness of the fluid versus the
stiffness of the matrix):
(13)
With this definition for , Equation (14) may be expressed in two

forms:
FLAC3D 6.0
(14)
and
(15)
If is small (compared to 1), Equation (9) shows that FLAC3D’s

standard explicit timestep (see Equation (10)) can be considered
as representative of the system diffusivity. An order-of-
magnitude estimate for the number of steps needed to reach full
consolidation under saturated conditions, for instance, can be
calculated using Equation (11).
If is large (i.e., is large compared to (or

)), FLAC3D’s explicit timestep will be very
small, and the problem diffusivity will be controlled by the matrix
(see Equation (10) and (14)). The value for (or ) can be
reduced in order to increase the timestep and reach steady-state
computationally faster. Equation (14) indicates that if (or )
is reduced such that = 20, then the diffusivity should be within
5% of the diffusivity for infinite (or ) (see 1D Consolidation
for an example). The time scale is expected to be respected within
the same accuracy. Note that, in any case, should not be made
higher than the physical value of the fluid (2 × 109 Pa for water).
Selection of a Modeling Approach for a Fully Coupled

Analysis
A fully coupled quasi-static hydromechanical analysis with FLAC3D is often time-
consuming, and even sometimes unnecessary. There are numerous situations in
which some level of uncoupling can be performed to simplify the analysis and
speed the calculation. The following examples illustrate the implementation of
FLAC3D modeling approaches corresponding to different levels of fluid/
mechanical coupling. Three main factors can help in the selection of a particular
approach:
1. the ratio between simulation time scale and characteristic time of

the diffusion process;
FLAC3D 6.0
2. the nature of the imposed perturbation (fluid or mechanical) to

the coupled process; and
3. the ratio of the fluid to solid stiffness.
The expressions for characteristic time, (in Equation (2)), diffusivity, (in
Equation (9)), and the stiffness ratio, (in Equation (13)) can be used to
quantify these factors. These factors are considered in detail below, and a
recommended procedure to select a modeling approach based on these factors is
given in Recommended Procedure to Select a Modeling Approach.
Time Scale
We first consider the time scale factor by measuring time from the initiation of a
perturbation. We define as the required time scale of the analysis, and as the
characteristic time of the coupled diffusion process (estimate of time to reach
steady state, defined using Equation (2) and Equation (9)).
Short-term behavior
If is very short compared to the characteristic time, , of the coupled

diffusion process, the influence of fluid flow on the simulation results
will probably be negligible, and an undrained simulation can be
performed with FLAC3D (using either a “dry” or “wet” simulation — see
No Flow. The wet simulation is carried out using model configure fluid
and zone fluid active off). No real time will be involved in the
numerical simulation (i.e., ), but the pore pressure will change
due to volumetric straining if the fluid modulus ( or ) is given a
realistic value. The footing load simulation is an example of this
approach.
Long-term behavior
If , and drained behavior prevails at , then the pore-

pressure field can be uncoupled from the mechanical field. The steady-
state pore-pressure field can be determined using a fluid-only
simulation (zone fluid active on, zone mechanical active off) (the
diffusivity will not be representative), and the mechanical field can be
determined next by cycling the model to equilibrium in mechanical mode
with = 0 (zone mechanical active on, zone fluid active off). (Strictly
speaking, this engineering approach is only valid for an elastic material,
FLAC3D 6.0
because a plastic material is path-dependent.)
Another way to describe the time scale is in terms of undrained or drained

response. Undrained strictly means that there is no exchange of fluid with the
outside world (where outside world means outside the sample in a lab test, and
other elements in a numerical simulation or a field situation). Drained means that
there is a full exchange of fluid with the outside world, which implies that the
fluid pressure is able to dissipate everywhere. The words are typically associated
with short-term and long-term, respectively, because an undrained test usually
can be done quickly, while a drained test requires a long time for excess fluid
pressures to dissipate. In the field, short-term means that there is insignificant
migration of fluid, and long-term means that the pressure has stabilized (which
needs a long time).
Note that in simulations conducted outside of the fluid configuration, or within

the fluid configuration but with fluid turned off, the total stress adjustments due
to an imposed pore-pressure change (such as those resulting from the lowering
of the water table) will not be done internally by the code. The pore-pressure
increments may, however, be monitored (using a FISH function, for instance) and
used to decrement the total normal stresses before cycling to mechanical
equilibrium. The saturated and unsaturated mass densities will also need to be
adjusted if the water table has been moved within the grid, and the simulation is
conducted outside of the fluid configuration. (See Verification of Concepts, and
Modeling Techniques for Specific Applications for additional information on this
topic.)
Nature of Imposed Perturbation to the Coupled Process

The imposed perturbation to a fully coupled hydromechanical system can be due
to changes in either the fluid flow boundary condition or the mechanical
boundary condition. For example, transient fluid flow to a well located within a
confined aquifer is driven by the change in pore pressures at the well. The
consolidation of a saturated foundation as a result of the construction of a
highway embankment is controlled by the mechanical load applied by the weight
of the embankment. If the perturbation is due to change in pore pressures, it is
likely that the fluid flow process can be uncoupled from the mechanical process.
This is described in more detail below, and illustrated in Transient Fluid Flow to
a Well in a Shallow Confined Aquifer. If the perturbation is mechanically driven,
the level of uncoupling depends on the fluid versus solid stiffness ratio, as
described below.
FLAC3D 6.0
Stiffness Ratio
The relative stiffness, (see Equation (13)), has an important influence on the
modeling approach used to solve a hydromechanical problem:
Relatively stiff matrix ( )
If the matrix is very stiff (or the fluid highly compressible) and
is very small, the diffusion equation for the pore pressure can
be uncoupled, since the diffusivity is controlled by the fluid
(Detournay and Cheng 1993). The modeling technique will
depend on the driving mechanism (fluid or mechanical
perturbation):
1. In mechanically driven simulations, the pore pressure may be

assumed to remain constant. In an elastic simulation, the solid
behaves as if there were no fluid; in a plastic analysis, the
presence of the pore pressure may affect failure. This modeling
approach is adopted in slope stability analyses (see Fixed Pore
Pressure).
2. In pore pressure-driven elastic simulations (e.g., settlement

caused by fluid extraction), volumetric strains will not
significantly affect the pore-pressure field, and the flow
calculation can be performed independently (zone fluid active
on, zone mechanical active off). (In this case, the diffusivity will
be accurate because for , the generalized consolidation
coefficient in Equation (9) is comparable to the fluid diffusivity in
Equation (7).) In general, the pore-pressure changes will affect
the strains, and this effect can be studied by subsequently cycling
the model to equilibrium in mechanical mode (zone mechanical
active on, zone fluid active off). The fluid modulus ( or )
must be set to zero during mechanical cycling, to prevent
additional generation of pore pressure.
Relatively soft matrix ( )
If the matrix is very soft (or the fluid incompressible) and is

very large, then the system is coupled, with a diffusivity
governed by the matrix. The modeling approach will also depend
on the driving mechanism.
FLAC3D 6.0
1. In mechanically driven simulations, calculations can be time-

consuming. As discussed in note 5 of Time Scale, it may be
possible to reduce the value for (or ), such that = 20, and
not significantly affect the response.
2. In most practical cases of pore pressure-driven systems,

experience shows that the coupling between pore pressure and
mechanical fields is weak. If the medium is elastic, the numerical
simulation can be performed with the flow calculation in flow-
only mode (zone fluid active on, zone mechanical active off)
and then in mechanical-only mode (zone mechanical active on,
zone fluid active off with fluid modulus set to zero) to bring the
model to equilibrium.
It is important to note that, in order to preserve the true

diffusivity (and hence the characteristic time scale) of the
system, the fluid modulus (or ) must be adjusted to the
value
(16)
or
(17)
during the flow calculation (see Equation (9)), and to zero during the mechanical
calculation to prevent further adjustments by volumetric strains (Berchenko
1998).
Recommended Procedure to Select a Modeling Approach

It is recommended that the selection of a modeling approach for a fully coupled
analysis follow the procedure outlined in Table 1. First, determine the
characteristic time of the diffusion process for the specific problem conditions
and properties (see Time Scale), and compare this time to the actual time scale of
interest. Second, consider whether the perturbation to the system is primarily
pore-pressure driven or mechanically driven. Third, determine the ratio of the
stiffness of the fluid to the stiffness of the solid matrix. Table 1 indicates the
appropriate modeling approach based on the evaluation of these three factors.
FLAC3D 6.0
The table also indicates the required adjustment to the fluid modulus, (or
), for each case. Finally, the table lists several examples from the manual that
illustrate each modeling approach.
Table 1: Recommended procedure to select a modeling approach for a fully coupled

analysis
Modeling
Imposed Fluid vs Approach Adjust Fluid
Time Scale Process Solid & Main Bulk Modulus Examples (6)
Perturbation Stiffness Calculation ( or )
Commands
mechanical any Effective no fluid Curvature
or pore Stress (1) Slope Stability
pressure
(steady- with no
state fluid flow
analysis)
or
Effective = 0.0 (or
Stress (2) = 0.0)
model
configure
fluid
zone fluid
active off
zone
mechanical
active on
mechanical any Pore realistic value Footing on

or pore Pressure for (or ) Saturated Soil
pressure Generation - Adding Load
(3)
(undrained model Triaxial
analysis) configure Compression
fluid Test on a
FLAC3D 6.0
Modeling
Commands
Cam-Clay
Sample
zone fluid
active off
zone
mechanical
active on
in the pore any Uncoupled 1D

range of pressure Flow- Consolidation
Mechanical - Uncoupled
(4)
model
configure
fluid
step 1
zone fluid (or

active on
)
zone
mechanical
active off
step 2 = 0.0 (or

= 0.0)
zone fluid
active off
zone
mechanical
active on
FLAC3D 6.0
Modeling
Commands
in the mechanical any Coupled realistic value Footing on
range of Flow- for (or ) Saturated Soil
Mechanical - Coupled
(5)
model so that 1D
configure Consolidation
fluid - Coupled
zone fluid Consolidation

active on Settlement
zone Pressuremeter
mechanical Test
active on
Notes to Table 1:
1. The effective stress approach with no fluid flow is discussed in

Fixed Pore Pressure. In order to establish the initial conditions for
this effective stress analysis, use the zone water plane command,
the zone gridpoint initialize pore-pressure command or a FISH
function to establish steady-state pore pressures. Specify the
correct wet density to zones below the water table, and dry
density to zones above.
2. The effective stress approach with groundwater flow is discussed

in Flow-Only Calculation. In order to establish the initial
conditions for this effective stress analysis, use the zone
gridpoint initialize pore-pressure command or a FISH function
to establish steady-state pore pressures, or specify zone fluid
active on and zone mechanical active off and step to steady state
if the location of the phreatic surface is not known. Set (or
FLAC3D 6.0
) to a small value to speed convergence for a partially saturated

system. Note that (or ) should be greater than (or
) to satisfy numerical stability (see Equation (12)).
3. The pore-pressure generation approach is discussed in No Flow

— Mechanical Generation of Pore Pressure. In order to establish
the initial conditions for the pore-pressure generation analysis,
use the zone gridpoint initialize pore-pressure command or a
FISH function to establish steady-state flow, or specify zone fluid
active off and zone mechanical active on and step to steady state
if the location of the phreatic surface is not known. Set (or
) to a small value to speed convergence for a partially saturated
system. Note that (or ) should be greater than (or
) to satisfy numerical stability (see Equation (12)).
4. The uncoupled fluid-mechanical approach is described in

Selection of a Modeling Approach for a Fully Coupled Analysis.
This approach is recommended for pore pressure-driven systems,
and should be used carefully if . Note that the adjusted
value for (or ) during the flow-only step should satisfy
Equation (16) so that the coupled diffusivity will be correct.
5. The fully coupled approach is discussed in Coupled Flow and

Mechanical Calculations. Note that for , if (or ) is
adjusted to reduce = 20, the time response will be close
(typically within 5%) to that for infinite (or ).
The saturated fast-flow option for fully coupled analysis (zone

fluid fastflow on) is discussed in Fully Saturated Fast Flow. Note
that this method can only be applied for fully saturated flow. ther
limitations are discussed in Fully Saturated Fast Flow.
6. Examples are given in Verification Examples on Fluid. Example

Applications (E.A.) and Verification Problems (V.P.) that
demonstrate the various methods are in the Example Problems
and Verification Problems.
FLAC3D 6.0
Fixed Pore Pressure (Used in Effective Stress Calculation)

In some calculations, the pore-pressure distribution is important only because it
is used in the computation of effective stress at all points in the system. For
example, in modeling slope stability, we may be given a fixed water table. To
represent this system with FLAC3D, it is sufficient to specify a pore-pressure
distribution that is unaffected by mechanical deformations that may occur.
Because no change in pore pressures is involved, we do not need to configure the
grid for groundwater flow.
The zone water set command can be used to specify an initial hydrostatic pore-
pressure distribution below a given fixed phreatic surface. The water density
must be provided (with the zone water density command), and appropriate dry
and saturated material densities supplied by the user above and below the water
table, respectively. Alternatively, the zone gridpoint initialize pore-pressure
command or a FISH function can be used to generate the required static pore-
pressure distribution.
The pore-pressure distribution corresponds to an initial state for which there is

no strain. It remains constant and is unaffected by mechanical deformation.
Fluid flow does not take place. The influence of this pore-pressure distribution is
on failure in material for which yield depends on mean effective stress. For an
example application, see Curvature Slope Stability.
Flow-Only Calculation to Establish a Pore-Pressure

Distribution
Flow-only calculations may be performed to determine the flow and pressure
distribution in some system independent of any mechanical effects. For example,
it may be necessary to evaluate the groundwater changes that result from the
digging of a drainage ditch or the activation of a pumping well. In other
instances, an initial pore-pressure distribution may be needed for a coupled
calculation. In both cases, FLAC3D may be run in the fluid-flow mode without
any mechanical calculation being done. Mechanical calculations may or may not
be done subsequently.
The first step in the command procedure for a flow-only calculation is to issue a
model configure fluid command so that extra memory can be assigned for the
fluid-flow calculation. The mechanical calculation should be inhibited with the
zone mechanical active off command. Then a choice must be made between the
explicit and implicit fluid-flow solution algorithm. By default, the explicit
FLAC3D 6.0
algorithm will be selected, but the implicit mode of calculation may be activated
(and deactivated) at any stage of the calculation using the zone fluid implicit on
or zone fluid implicit off command. Note however, that the implicit calculation
mode only applies to problems in which the medium remains fully saturated (i.e.,
the saturation remains constant and equal to one); it will give wrong results if
desaturation occurs in the simulation. When using the implicit mode of
calculation, always make sure that desaturation has not taken place (e.g., by
plotting saturation contours).
In the explicit mode, the fluid-flow timestep will be calculated automatically, but
a smaller timestep can be selected using the model fluid timestep command. The
magnitude of the timestep must be specified by the user in the implicit mode.
This is done by issuing a model fluid timestep command. For saturated flow, it is
often more efficient to use the implicit solution mode when contrasting
permeabilities exist.
The fluid-flow model and properties must be specified for all zones in which
fluid flow may occur. Initial and boundary conditions are assigned to complete
the fluid-flow problem setup. The fluid-flow domain in a fluid-only or fluid-
mechanical simulation is defined by the assembly of zones with a non-null fluid
flow model. Flux boundary conditions, for instance, can be assigned by
specifying the zone face apply flux command with the range keyword to
correspond to the boundaries of that domain. (Remember that zones that are
nulled mechanically are not automatically nulled for fluid flow.)
The model step command may be specified to execute a given number of fluid-
flow steps. To stop the calculation automatically when a particular fluid-flow
time is reached, a model solve command may be issued. Alternatively, the model
solve fluid time-total or model solve fluid cycles setting may be specified to
prescribe a fluid-flow time limit or maximum number of steps; then a model
solve command may be given. A steady-state flow condition can also be
calculated by using the model solve command with the ratio keyword to specify
the limiting unbalanced fluid-flow ratio defining the steady flow state.
If the computed pore-pressure distribution is then to be used in a mechanical

calculation where pore pressure can be assumed to remain constant, the zone
fluid active off and zone mechanical active on commands should be given. The
Biot coefficient or fluid bulk modulus should also be set to zero to prevent extra
pore pressures from being generated by mechanical deformation.
FLAC3D 6.0
Steady-State Fluid Flow with a Free Surface presents an example flow-only

analysis in which unconfined flow through a vertical embankment is calculated.
The analysis produces the steady-state phreatic surface. Note that in this
example, the fluid bulk modulus is set to a low value to enable the phreatic
surface to develop quickly. This can be done when the time taken to reach steady
state is unimportant. Note, however, that there is a lower limit for fluid modulus
in order to avoid numerical instability (see Equation (12)).
No Flow — Mechanical Generation of Pore Pressure

The undrained (short-term) response may be analyzed in FLAC3D using both dry
and wet approaches. In a dry simulation, the generation of pore pressure under
volumetric strain is not simulated directly. Instead, its effect on mechanical
deformation is taken into account by assigning the undrained bulk value,
, to the material bulk modulus, , in the FLAC3D simulation. In
this case, two different techniques can be applied to detect failure in a Mohr-
Coulomb material. In the first one, the constant pore pressures are initialized
using the zone water or zone gridpoint initialize pore-pressure command, and
undrained cohesion and friction are given as input. In the second, the material is
assigned a zero friction and a cohesion equal to the undrained shear strength, .
The first technique applies to problems where changes in pore pressure are small
compared to the initial values. The second technique strictly applies to plane-
strain problems with Skempton coefficient equal to one ( ). Note
that a dry simulation can be carried out whether or not the model configure
fluid command has been issued. However, if the command has been used, the
fluid bulk modulus ( or ) must be set to zero to prevent additional
generation of pore pressure.
In a wet simulation, the short-term response of a coupled system is analyzed in

the fluid configuration of FLAC3D. In this case, drained values should be used for
the material bulk modulus, friction and cohesion. If the zone fluid active off
command is given, and the Biot modulus (or fluid modulus) is given a realistic
value, then pore pressure will be generated as a result of mechanical
deformations. For example, the “instantaneous” pore pressures produced by a
footing load can be computed in this way. If the fluid bulk modulus ( or ) is
much greater than the solid bulk modulus, convergence will be slow; it may be
possible to reduce the fluid bulk modulus without significantly affecting the
behavior. (See Note 5 in Table 1.) The data file Footing-InstantPP.f3dat illustrates
pore pressure buildup produced by a footing load on an elastic/plastic material
FLAC3D 6.0
contained in a box. The corresponding project file, “Footing.f3prj,’’ is located in

the “datafiles\Fluid\Footing” folder. The left boundary of the box is a line of
symmetry, and the pore pressure is fixed at zero along the top surface to prevent
pore-pressure generation there. By default, the porosity is 0.5; permeability is
not needed, since flow is not calculated. Note that the pore pressures are fixed at
zero at gridpoints along the top of the grid. This is done because at the next stage
of this model, a coupled, drained analysis in which drainage will be allowed at
the ground surface will be performed (see Coupled Flow and Mechanical
Calculations). The zero pore-pressure condition is set now to provide the
compatible pore-pressure distribution for the second stage.
As a large amount of plastic flow occurs during loading, the normal stress is
applied gradually by using the FISH function ramp to supply a linearly varying
multiplier to the zone apply command. Figure 1 shows pore-pressure contours
and vectors representing the applied forces. It is important to realize that the
plastic flow will occur in reality over a very short period of time (on the order of
seconds); the word “flow” here is misleading since, compared to fluid flow, it
occurs instantaneously. Hence, the undrained analysis (with zone fluid active
off) is realistic.
The saturated fast-flow logic can be used for this example because the material
is fully saturated, and the stiffness of the fluid is significantly greater than that
of the solid matrix. It is only necessary to include the zone fluid fastflow on
command in Footing-InstantPP.f3dat to perform a saturated fast-flow calculation.
Calculational stepping stops, and equilibrium is achieved, when the unbalanced
volumes, and , and the unbalanced force ratio are smaller than the
predefined limits. The instantaneous pore-pressure generation is nearly the
same (within approximately 5%) as that obtained by the basic flow scheme, as
illustrated in Figure 2. The runtime for the saturated fast-flow scheme is roughly
three times faster than that for the basic flow scheme.
Footing-InstantPP.f3dat:
model new
model title 'Instantaneous pore pressures generated under an applied load'
model configure fluid
point 2 0 1 0 point 3 0 0 10
; --- mechanical model ---
zone property bulk 5e7 shear 3e7 fric 25 cohesion 1e5 tens 1e10
FLAC3D 6.0

; --- apply load slowly ---
fish define ramp
ramp = math.min(1.0,float(global.step)/1000.0)
end
zone face apply stress-normal = -0.3e6 fish @ramp ...
range position-x 0 3 position-z 10
; --- fluid flow model ---
; --- pore pressure fixed at zero at the surface ---
zone gridpoint fix pore-pressure 0 range position-z 10
; --- settings ---
; --- histories ---
zone history pore-pressure position (2,1,9)
; --- timing ---
fish define time0
t0 = time.clock/100.0
end
fish define runtime
runtime = time.clock/100.0 - t0
end
@time0
model solve ratio 1e-5
list @runtime
model save 'load'
return
FLAC3D 6.0
Figure 1: Instantaneous pore pressures generated under an applied load.
Figure 2: Instantaneous pore pressures generated under an applied load — with

zone fluid fastflow on.
FLAC3D 6.0
Coupled Flow and Mechanical Calculations

By default, FLAC3D will do a coupled fluid-flow and mechanical calculation if the
grid is configured for fluid, and if the Biot modulus, or fluid bulk modulus, and
permeability are set to realistic values. The full fluid-mechanical coupling in
FLAC3D occurs in two directions: pore-pressure changes cause volumetric strains
that influence the stresses; in turn, the pore pressure is affected by the straining
that takes place.
The relative time scales associated with consolidation and mechanical loading
should be appreciated. Mechanical effects occur almost instantaneously (on the
order of seconds, or fractions of seconds). However, fluid flow is a long process:
the dissipation associated with consolidation takes place over hours, days or
weeks.
Relative time scales may be estimated by considering the ratios of characteristic

times for the coupled and undrained processes. The characteristic time
associated with the undrained mechanical process is found by using the
saturated mass density for and the undrained bulk modulus, , as defined in
this equation (or this equation for incompressible grains) for in Equation (1).
The ratio of the diffusion characteristic time (Equation (2) and (9)) and the
undrained mechanical characteristic time is thus
(18)
In most cases, is approximately 1010 Pa, but the mobility coefficient may
differ by several orders of magnitude; typical values are
10-19m2/Pa-sec for granite;
10-17m2/Pa-sec for limestone;
10-15m2/Pa-sec for sandstone;
10-13m2/Pa-sec for clay; and
10-7m2/Pa-sec for sand.
For rock and soil, is of the order of 103kg/m3, while is approximately

1010 Pa. Using those orders of magnitude in Equation (18), we see that the ratio of
fluid-to-mechanical time scales may vary between for sand, 106 for clay,
FLAC3D 6.0
108 for sandstone, 1010 for limestone and 1012 for granite. If we exclude
materials with mobility coefficients larger than that of clay, we see that this ratio
remains very large, even for small values of .
In practice, mechanical effects can then be assumed to occur instantaneously

when compared to diffusion effects. This is also the approach adopted in FLAC3D,
where no time is associated with any of the mechanical steps taken together with
the fluid-flow steps. The use of the dynamic option in FLAC3D may be
considered, to study the fluid-mechanical interaction in materials such as sand,
where mechanical and fluid time scales are comparable.
In most modeling situations, the initial mechanical conditions correspond to a

state of equilibrium that must first be achieved before the coupled analysis is
started. Typically, at small fluid-flow time (compared to — see Time Scale), a
certain number of mechanical steps must be taken for each fluid step to reach
quasi-static equilibrium. At larger fluid-flow time, if the system approaches
steady-state flow, several fluid-flow timesteps may be taken without
significantly disturbing the mechanical state of the medium. A corresponding
numerical simulation may be controlled manually by alternating between flow-
only (zone fluid active on and zone mechanical active off) and mechanical-only
(zone fluid active off and zone mechanical active on) modes. The model step
command can be used to perform calculations for both the flow-only and
mechanical-only processes.
As an alternative approach, the tedious task of switching between flow-only and

mechanical-only modes may be avoided by using the model solve command in
combination with appropriate settings; model solve mechanical unbalanced-
maximum (or model solve mechanical ratio-maximum) will set a limit to the
maximum out-of-balance force (or force ratio), under which quasi-static
mechanical equilibrium will be assumed. model mechanical substep n auto
declares the mechanical module as the slave that must perform n sub-cycles (or
fewer, if equilibrium is detected — see previous setting) for each step taken by
the master. model fluid substep m implicitly declares the flow module as the
master. (The keyword auto is omitted for the master process.) If for each fluid-
flow timestep the mechanical module needs only one sub-step to reach
equilibrium, then the number of fluid-flow sub-steps is doubled, but never
exceeds the number m (by default, m = 1). The system reverts to the original
setting of sub-steps whenever the succession of one mechanical timestep for
FLAC3D 6.0
each fluid-flow group of sub-steps is broken. If the option time-total is specified

in the model solve fluid command, the computation will proceed until the fluid-
flow time given by the model solve time-total parameter is reached.
In a third approach, the model step command may be used while both mechanical
and fluid-flow modules are on. In this option, one mechanical step will be taken
for each fluid-flow step. Here, fluid-flow steps are assumed to be so small that
one mechanical step is enough. Input Instructions should be consulted for a
complete list of available command options for a coupled analysis.
To illustrate a fully coupled analysis, we continue the footing simulation (done in

No Flow — Mechanical Generation of Pore Pressure), using the model solve
command with appropriate settings to calculate the consolidation beneath the
footing. The data file is Footing-Coupled.f3dat:
Footing-Coupled.f3dat:
model restore 'load'

; --- turn on fluid flow model ---
zone fluid property permeability 1e-12
model fluid active on
; --- set mechanical limits ---
model mechanical substep 100
model mechanical slave on
model fluid substep 10
; --- histories ---
model history fluid time-total
zone history displacement-z position 0,1,10
zone history pore-pressure position 2,1,9
; --- solve to 3,000,000 sec ---
@time0
model solve fluid time-total 3.0e6 or mechanical ratio 1e-4
list @runtime
model save 'age_3e6'
return
The screen printout should be watched during the calculation process – eight
variables are updated on the screen after the model solve command is issued:
1. the current step number;
FLAC3D 6.0
2. the total number of sub-cycles taken by the master process (fluid flow)
for the most recent step;
3. the total number of sub-cycles taken by the slave process (mechanical)

for the most recent step;
4. the process that is currently active;
5. the current sub-cycle number;
6. the current maximum unbalanced force or force ratio;
7. the total time for the master (fluid flow) process; and
8. the current timestep.
The unbalanced force ratio tolerance is 10-4 for this problem; enough mechanical
steps are taken to keep the unbalanced force ratio below this tolerance. However,
the limit to mechanical steps, defined by model mechanical substep, is set to 100
in this example. If the actual number of mechanical steps taken is always equal
to the set value of model mechanical substep, then something must be wrong.
Either the force limit or model mechanical substep has been set too low, or the
system is unstable and cannot reach equilibrium. The quality of the solution
depends on the force tolerance: a small tolerance will give a smooth, accurate
response, but the run will be slow; a large tolerance will give a quick answer, but
it will be noisy.
The characteristic diffusion time for this coupled analysis is evaluated from
Equation (2), using Equation (9) for the diffusivity and a value of = 10 m
corresponding to the model height. Using the property values in this example,
is estimated at 1,200,000 seconds. Full consolidation is expected to be reached
within this time scale; the numerical simulation is carried out for a total of
3,000,000 seconds.
Figure 3 shows the time histories to 3,000,000 seconds of displacements under

the footing load. In this simulation, pore pressures remain fixed at zero on the
ground surface; hence, the excess fluid escapes upward. The leveling off of the
histories indicates that full consolidation has been reached.
FLAC3D 6.0
Figure 3: Consolidation response — time histories of footing displacements.
The coupled simulation can also be run using the saturated fast-flow scheme
because the foundation material is fully saturated. In this case the zone fluid
fastflow on command is invoked. When the fluid flow is turned on, the coupled
calculation will continue using the saturated fast-flow scheme. The results are
similar to the results for the basic flow case, as indicated by Figure 4 (compare to
Figure 3). The unbalanced force ratio tolerance is set to 10-6 for this simulation,
in order to provide a smoother displacement history. Even with this lower
tolerance, the saturated fast-flow run is approximately 10 times faster than the
basic flow simulation.
FLAC3D 6.0
Figure 4: Consolidation response — time histories of footing displacements —

with zone fluid fastflow on.
Finally, a comparison is made with an uncoupled drained analysis. First, a flow

only calculation is made, then a mechanical only calculation is performed, with
fluid bulk modulus set to zero. The data file for the uncoupled run is Footing-
Uncoupled.f3dat. The displacement histories for the uncoupled drained
simulation are plotted in Figure 5. The final settlement match within 3% of the
basic flow results.
Footing-Uncoupled.f3dat:
model restore 'load'

; --- turn on fluid flow model ---
zone fluid property permeability 1e-12
; --- histories ---
FLAC3D 6.0

; --- solve to 2,000,000 sec ---
; --- uncoupled ---
model mechanical active off
model solve fluid time-total 3.0e6
model mechanical active on
zone gridpoint initialize fluid-modulus 0
model solve
model save 'age_3e6_uncoup'
return
Figure 5: Consolidation response — footing displacements for uncoupled

drained analysis.
If a sudden change of loading or mechanical boundary condition is applied in a

coupled problem, it is important to allow the undrained (short-term) response to
develop before allowing flow to take place. In other words, FLAC3D should be run
to equilibrium under zone fluid active off conditions following the imposed
mechanical change. The model solve logic can then be used (with zone fluid
active on) to compute the subsequent coupled flow/mechanical response. If
FLAC3D 6.0
changes in fluid boundary conditions occur physically at the same time as

mechanical changes, then the same sequence should be followed (i.e., mechanical
changes | equilibrium | fluid changes | coupled solution).
Another example of fully coupled behavior is the time-dependent swelling that

takes place following the excavation of a trench in saturated soil. In this case,
negative pore pressures build up immediately after the trench is excavated; the
subsequent swelling is caused by the gradual influx of water into the region of
negative pressures. We model the system in two stages: in the first, we allow
mechanical equilibrium to occur, without flow; then we allow flow, using the
model solve command to maintain quasi-static equilibrium during the
consolidation process. The fluid tension is initialized to a large negative number
to prevent desaturation. The data file used is Swelling.f3dat. The corresponding
project file is “Swelling.f3prj,” located in the “datafiles\Fluid\Swelling” folder.
The trench is excavated in the left-hand part of a flat soil deposit that is initially
fully saturated and in equilibrium under gravity. The material is elastic in this
case, but it could equally well have been a cohesive material, such as clay. In this
run, we assume impermeable conditions for the free surfaces. Figure 6 shows the
displacement vectors that accumulate during the time that flow is occurring; the
trench is seen at the left-hand side of the model. Figure 7 shows the time history
of pore pressure near the crest of the trench; note that there is an initial negative
excursion in pressure arising from the instantaneous expansion of the soil
toward the trench. Figure 8 shows histories of horizontal and vertical
displacement at the crest. The characteristic time for this problem, evaluated
using the model length of 40 m for , is approximately 5 × 108 seconds (based
on Equation (2) and (9)); the numerical simulation is carried out to that time.
Swelling.f3dat:
model new
model title "Maintaining equilibrium under time-dependent swelling conditions"
model config fluid
zone cmodel null range position-x 0,2 position-z 2,8
zone fluid property permeability 1e-14 poros 0.5
FLAC3D 6.0
zone initialize fluid-density 1000

zone fluid cmodel assign null range position-x 0,2 position-z 2,8
; initial and boundary conditions
zone gridpoint fix velocity-x range position-x 0
zone gridpoint fix velocity-x range position-x 40
model gravity 0,0,-10
zone initialize stress-xx -1.6e5 grad 0,0,20000
zone initialize stress-yy -1.6e5 grad 0,0,20000
zone initialize stress-zz -1.6e5 grad 0,0,20000
zone gridpoint initialize pore-pressure 8.0e4 grad 0,0,-10000
; --- settings ---
;
model history mechanical unbalanced-maximum
model solve
model save 'swell1'
;
;fish history @ gp pp 3,0,7
zone history pore-pressure position 4.5,0.5,6.5
zone history displacement-x position 2,0,8
zone gridpoint fix pore-pressure range position-x 40
model solve fluid time-total 5e8 or mechanical unbalanced-maximum 50
model save 'swell2'
return
break
FLAC3D 6.0
Figure 6: Swelling displacements near a trench with impermeable surfaces.
Figure 7: History of pore pressure behind the trench face.
FLAC3D 6.0
Figure 8: Displacement histories at the trench crest — vertical (top) and

horizontal (bottom)
FLAC3D 6.0
Verification Examples 561
Verification Examples
Several verification problems that illustrate the fluid-flow modeling capabilities
in FLAC3D are presented. The data files for these examples are contained in the
“\Datafiles\Fluid” directory.
• Unsteady Groundwater Flow in a Confined Layer

• One-Dimensional Filling of a Porous Region
• Steady-State Fluid Flow with a Free Surface
• Spreading of a Groundwater Mound
• One-Dimensional Consolidation
• Consolidation Settlement at the Center of a Strip Load
• Transient Fluid Flow to a Well in a Shallow Confined Aquifer
• Pressuremeter Test
• Semi-confined Aquifer
FLAC3D 6.0
FLAC3D 6.0
Unsteady Groundwater Flow in a Confined Layer 563
Unsteady Groundwater Flow in a Confined Layer

A long embankment of width = 100 m rests on a shallow saturated layer of soil.
The width ( ) of the embankment is large in comparison to the layer thickness,
and its permeability is negligible when compared to the permeability, = 10-12
m2/(Pa sec), of the soil. The Biot modulus for the soil is measured to be = 10
GPa. Initial steady-state conditions are reached in the homogeneous layer. The
purpose is to study the pore-pressure change in the layer as the water level is
raised instantaneously upstream by an amount = 2 m. This corresponds to a
pore-pressure rise of (with the water density = 1000 kg/m3 and
acceleration of gravity = 10 m/s2) at the upstream side of the embankment.
Figure 1 shows the geometry of the problem:
Figure 1: Confined flow in a soil layer.
FLAC3D 6.0
The flow in the layer may be assumed to be one-dimensional. The model has
width . The excess pore pressure, , initially zero, is raised suddenly to the
value at one end of the model. The corresponding analytical solution has the
form
where the -axis is running along the embankment width and has its origin at
the upstream side, , , , and .
In the FLAC3D model, the layer is defined as a column of 25 zones. The excess
pore pressure is fixed at the value of 2 × 104 Pa at the face located at = 0, and at
zero at the face located at = 100 m. The model grid is shown in Figure 2.
The analytical solution is programmed as a FISH function for direct comparison

to the numerical results at selected fluid-flow times corresponding to = 0.05,
0.1, 0.2 and 1.0. The analytical and numerical pore-pressure results for these
times are stored in tables.
Figure 2: FLAC3D grid for fluid flow in a confined soil layer.
FLAC3D 6.0
UnsteadyGroundwaterFlowConfinedLayer-Explicit.f3dat contains the FLAC3D

data file for this problem, using the explicit formulation to obtain the solution.
UnsteadyGroundwaterFlowConfinedLayer-Implicit.f3dat contains the data file
using the implicit formulation. The comparison of analytical and numerical
excess pore pressures at four fluid-flow times for the explicit solution is shown
in Figure 3, and for the implicit solution in Figure 4. Normalized excess pore
pressure ( ) is plotted versus normalized distance ( ) in the two figures,
where Tables 2, 4, and 6 contain the analytical solution for excess pore
pressures, and Tables 1, 3, and 5 contain the FLAC3D solutions. The four flow
times are 5 × 104, 105, 2 × 105, and 106 seconds for both explicit and implicit
solutions. Steady-state conditions are reached by the last time considered. For
both solution formulations, the difference between analytical and numerical pore
pressures at steady state is less than 0.1%.
break
FLAC3D 6.0
Figure 3: Comparison of excess pore pressures for the explicit-solution

algorithm (analytical values = lines; numerical values = crosses).
Figure 4: Comparison of excess pore pressures for the implicit-solution

algorithm (analytical values = lines; numerical values = crosses).
FLAC3D 6.0
Data File
UnsteadyGroundwaterFlowConfinedLayer-Explicit.f3dat
model new
model title 'Unsteady groundwater flow in a confined layer: Explicit Method'
@constants
zone create brick size 1 1 25 point 1 (10 0 0) point 2 (0 10 0) ...
point 3 (0 0 @length)
zone fluid biot on
zone fluid property permeability @c_cond
zone gridpoint initialize biot @c_biom
zone face apply pore-pressure @dp1 range position-z 0
zone face apply pore-pressure 0 range position-z 100
; --- settings ---
;
model solve fluid time-total 5e4
zone gridpoint list fluid range position-x -0 position-y 0
@num_sol
@ana_sol
@num_sol
@ana_sol
@num_sol
@ana_sol
@num_sol
@ana_sol
model save 'confinedlayer-explicit'
UnsteadyGroundwaterFlowConfinedLayer-Implicit.f3dat
model new
model title 'Unsteady groundwater flow in a confined layer: Implicit Method'
@constants
zone create brick size 1 1 25 point 1 (10 0 0) point 2 (0 10 0) ...
point 3 (0 0 @length)
zone fluid biot on
zone fluid property permeability @c_cond
FLAC3D 6.0
zone gridpoint initialize biot @c_biom

zone face apply pore-pressure @dp1 range position-z 0
zone face apply pore-pressure 0 range position-z 100
; --- settings ---
zone fluid implicit on
model fluid timestep fix 1e3
;
zone gridpoint list fluid range position-x 0 position-y 0
@num_sol
@ana_sol
@num_sol
@ana_sol
@num_sol
@ana_sol
@num_sol
@ana_sol
model save 'confinedlayer-implicit'
FLAC3D 6.0
One-Dimensional Filling of a Porous Region 569
One-Dimensional Filling of a Porous Region

In this problem, flow is driven through an initially dry porous layer of large
lateral extent under a constant pressure, , applied at the base. The transient
location of the filling front is compared to an exact sharp-front solution for the
cases with and without gravity.
Voller et al. (1996) give an analytic solution for this problem under the
assumptions of a sharp-front, rigid-porous matrix and incompressible
Newtonian fluid. In their solution, the flow is governed by Darcy’s law, and there
is a constant atmospheric pressure in the air ahead of the free surface.
Let the -axis of reference be oriented in the direction of flow, with the origin at
the base of the layer. The solution for the front location, , may be expressed in
terms of two dimensionless variables: and ; and a
dimensionless parameter, , defined using the expressions ,
and . In these equations, is porosity, is intrinsic permeability
(product of mobility coefficient, , and dynamic viscosity, ), is fluid density,
and is gravity.
When gravity is ignored, the solution has the form
(1)
For filling under gravity, the front location is given by
(2)
Equation (2) may be shown to converge to the no-gravity solution when goes
to zero.
The numerical solution to the filling problem is presented in dimensionless form.

To derive this solution, scaled properties are used in the simulations: = 1, =
0.25, = 0.5; and for gravity flow, = 1 and = 1. Further, using = 4 in the
preceding definitions, the characteristic parameters for the simulation are = 2,
= 1, = 1.
FLAC3D 6.0
To simulate incompressible flow, the bulk modulus of the fluid, , is assigned a

value that is large compared to the pore-pressure variations in the simulation (
= 100). For both cases, the grid corresponds to a column of 25 zones, 0.625
units high and 0.025 units wide. The initial value of pore pressure and saturation
is zero. The pore pressure is fixed at , and the saturation is given a value of 1 at
the base of the model. The simulation is conducted for a total of 0.25 units of
time ( = 0.125). A FISH function, flacfront, captures the times at which nodes
reach a saturation of 1% and 99%. The analytic sharp-front solution is evaluated
by another FISH function, solution. The data file for the case without gravity is
listed in “1DFillingPorousRegion-NoGravity.f3dat”; the data file for the case with
gravity is listed in “1DFillingPorousRegion-Gravity.f3dat”.
The results with and without gravity are presented in Figure 1 and Figure 2,
respectively. As seen in these figures, the sharp-front solution is bounded above
and below by the 99% and 1% saturation fronts. In fact, the vertical distance
between these fronts corresponds directly to the grid size in the direction of
propagation of the filling front. (The saturation at a node can only start to
increase when the pore pressure at the node below it becomes positive, and thus
full saturation is reached there.) This distance can be reduced by increasing the
number of zones in the column height. The evolution of nodal pore pressure with
time follows a stepwise pattern, more pronounced as the fluid is less
compressible. This behavior occurs because a node must be fully saturated before
its pore pressure can increase. One way to reduce this effect without changing
the grid size is to introduce flow in the unsaturated region (capillary pressure) in
the fluid flow formulation.
break
FLAC3D 6.0
Figure 1: Location of filling front ( vs ) — no gravity (analytical solution =

solid line; numerical values = dotted lines).
Figure 2: Location of filling front ( vs ) — with gravity (analytical solution =

solid line; numerical values = dotted lines).
FLAC3D 6.0
Data Files
1DFillingPorousRegion-NoGravity.f3dat
model new
model title "One-dimensional filling - no gravity"
@setup
zone create brick size 1 1 25 point 1 (0.025,0,0) point 2 (0,0.025,0) ...
point 3 (0,0,0.625)
zone fluid property permeability @c_perm porosity @c_poro biot 0. ;
zone fluid biot on
zone gridpoint initialize biot 100
zone gridpoint initialize saturation 0.0
zone face apply pore-pressure @c_p0 range position-z 0
; --- settings ---
@solution1
table 1 label 'Analytical solution'
table 2 label ' 1% saturation front'
table 3 label '99% saturation front'
; --- test ---
fish set @lzf = .025 @uzf = .025
model solve fluid time-total 0.25
model save 'asat1'
return
1DFillingPorousRegion-Gravity.f3dat
model new
model title "One-dimensional filling - with gravity"
@setup
point 3 (0,0,0.625)
zone fluid property permeability @c_perm poros @c_poro biot 0.
zone fluid biot on
zone gridpoint initialize biot 100
zone face apply pore-pressure @c_p0 range position-z 0
; --- settings ---
zone initialize fluid-density @c_den
model gravity 0 0 @c_grav
FLAC3D 6.0

@solution2
table 1 label 'Analytical solution'
table 2 label ' 1% saturation front'
table 3 label '99% saturation front'
; --- test ---
fish set @lzf = .025 @uzf = .025
model save 'asat2'
return
FLAC3D 6.0
FLAC3D 6.0
Steady-State Fluid Flow with a Free Surface 575
Steady-State Fluid Flow with a Free Surface

This example is the classical problem of steady-state seepage flow through a
homogeneous embankment with vertical slopes exposed to different water levels
and resting on an impermeable base. The total discharge, , and the length of
seepage face, , are compared to the exact solutions. Figure 1 shows the geometry
and boundary conditions of the problem.
Figure 1: Problem geometry and boundary conditions.
The fluid is homogeneous, flow is governed by Darcy’s law, and it is assumed

that the pores of the soil beneath the phreatic surface are completely filled with
water and the pores above it are completely filled with air. The width of the dam
is , the head and tail water elevations above the impervious base are and ,
respectively.
FLAC3D 6.0
The exact solution for the total discharge through a dam section of unit thickness
was shown by Charny (Harr 1991) to be given by Dupuit’s formula:
(1)
where is mobility coefficient, is water density and is gravity. The length, ,

of the seepage face (elevation of the free surface on the downstream face of the
dam above ) was obtained by Polubarinova-Kochina (1962), and is given in
Figure 2 as a function of the characteristic dimensions of the problem.
Figure 2: Seepage face solution after Polubarinova-Kochina (1962).
The FLAC3D simulation is conducted for a particular set of parameters:
L = 9m
h1 = 6 m
h2 = 1.2 m
Several material properties are used:
permeability ( ) 10-10 (m/sec)(Pa/m)
FLAC3D 6.0
porosity ( ) 0.3
water density ( ) 1000 kg/m3
water bulk modulus ( ) 1000 Pa
soil dry density ( ) 2000 kg/m3
gravity ( ) 10 m/sec2
Two cases corresponding to two different initial conditions have been studied:
CASE 1: The water level in the embankment is initially at = = 1.2 m.

The upstream level is raised to = .
CASE 2: The water level is initially at = = 6 m. The downstream level is

lowered to = .
The data file for Case 1 is listed in “FreeSurfaceSteadyStateFluidFlow-

Case1.f3dat”, and in “FreeSurfaceSteadyStateFluidFlow-Case2.f3dat” for Case 2.
The grid and boundary conditions are the same for both cases. The grid contains
30 zones in width, 20 zones in height, and 1 zone in thickness. The boundary
conditions correspond to a static pore-pressure distribution up to level on the
upstream face, and up to on the downstream face, zero pore pressure from
level to on the downstream face, and to no-flow conditions across the
remaining boundaries. The differences between the two cases are the initial pore
pressure and saturation distributions. In Case 1, saturation and pore pressure are
zero above . Below that level, the saturation is 1 and the pressure is hydrostatic.
In Case 2, saturation is 1 for all gridpoints, and the pore pressure inside the mesh
follows a gravitational gradient. The numerical simulation is carried out until
steady-state conditions are detected.
To speed the calculation to steady state, the water bulk modulus is given a small
value ( = 103 Pa) compatible with free surface stability. (The criterion used is
, where is the maximum vertical zone dimension in the vicinity
of the phreatic surface, as discussed in Time Scales.)
The final flow pattern is similar for both initial conditions (see Figure 3 and
Figure 4). The numerical value of seepage length is defined as the distance on the
downstream face of the dam, between the tail water elevation and the point
where the magnitude of the flow vector vanishes. The analytical value of seepage
length is determined from Figure 2. For this particular problem, = 0.2,
= 1.5, and the value of is thus 0.1. As seen in the figures, the numerical value
FLAC3D 6.0
of seepage length compares well with the analytical solution sketched there as a
bold line. A FISH function, qflac, is used to determine the discharge, , per unit
thickness of the dam: the steady-state numerical value is 1.914 × 10 -6m2/s for
Case 1; and 1.912 × 10-6m2/s for Case 2. The values are close to the analytic value
of 1.920 × 10-6m2/s, as determined from Equation (1) for this particular problem.
break
FLAC3D 6.0
Figure 3: Steady-state flow vectors and seepage face solution—Case 1.
Figure 4: Steady-state flow vectors and seepage face solution—Case 2.
FLAC3D 6.0
Case 1 is repeated to test fluid flow across two sub-grids that are connected by
using either the zone attach command or the zone interface create command.
“FreeSurfaceSteadyStateFluidFlow-Case1-Attach.f3dat” lists the commands for
flow across an attached grid, and “FreeSurfaceSteadyStateFluidFlow-
Case1-Interface.f3dat” lists the commands for flow across an interface. Both
model results are nearly identical to the original Case 1; the values for discharge
for both the Attach and Interface models are 1.924 × 10-6m2/s and 1.913 ×
10-6m2/s, respectively. Compare Figure 5 and Figure 6 to Figure 3. Note that if
maximum-edge is specified with the zone interface element command, then the
interface will act as an impermeable boundary. Remove the semicolon from the
line of zone.interface.1.element.maximum-edge to observe this response.
break
FLAC3D 6.0
Figure 5: Steady-state flow vectors for attached grid—Case 1.
Figure 6: Steady-state flow vectors for grid with an interface—Case 1.
FLAC3D 6.0
Reference
Harr, M. E. Groundwater and Seepage. Dover (1991).
Data Files
FreeSurfaceSteadyStateFluidFlow-Case1.f3dat
model new
model title "Steady state flow through a vertical embankment-case 1"
point 3 (0,0,6.0)
zone fluid property permeability 1e-10 porosity 0.3
zone fluid biot off
zone gridpoint initialize saturation 1.0 ...
range union position-x 0 position-z 0.0 1.19
zone gridpoint initialize pore-pressure 1.2e4 grad (0,0,-1e4) ...
range position-z 0.0 1.19
zone face apply pore-pressure 6.0e4 grad 0 0 -1e4 ...
range position-x 0.0
range position-x 9.0 position-z 0.0 1.19
zone face apply pore-pressure 0.0 ...
; Add particle tracks
zone fluid track create line begin (0,0.1,0.5) end (0,0.1,5.5) ...
segment 12 group "set1"
zone fluid track active on
; --- settings ---
zone initialize fluid-density 1e3
zone gridpoint initialize fluid-tension 0.0
; --- test ---
model solve ratio 1.e-3
@show_seepage_face
list @qflac @qsol
model save 'ch2a'
return
FreeSurfaceSteadyStateFluidFlow-Case2.f3dat
model new
FLAC3D 6.0
[t="Steady state flow through a vertical embankment-case 2"

model title [t]
point 3 (0,0,6.0)
zone fluid biot off
range position (0.01,0,5.99) (6.0,0.15,6.0)
range position (9.0,0,0) (9.0,0.15,1.19)
range position (9.0,0,1.2) (9.0,0.15,6.0)
; --- settings ---
; --- test ---
@show_seepage_face
list @qflac @qsol
model save 'ch2b'
FreeSurfaceSteadyStateFluidFlow-Case1-Attach.f3dat
model new
[t="Steady state flow through a vertical embankment-Case 1-ATTACHed grid"]
model title [t]
zone create brick size 10 1 10 point 0 (0,0,0) point 1 (4.5,0,0) ...
point 2 (0,0.15,0) point 3 (0,0,6)
zone create brick size 15 1 20 point 0 (4.5,0,0) point 1 (9,0,0) ...
point 2 (4.5,0.15,0) point 3 (4.5,0,6) ...
merge off
zone attach by-face
zone fluid biot on
zone gridpoint initialize biot 3.333e3
FLAC3D 6.0

; --- settings ---
; --- test ---
@show_seepage_face
list @qflac @qsol
model save 'attfl'
FreeSurfaceSteadyStateFluidFlow-Case1-Interface.f3dat
model new
[t = "Steady state flow through a vertical embankment - case 1 - INTERFACE"]
model title [t]
zone create brick size 15 1 10 point 0 (0,0,0) point 1 (4.5,0,0) ...
point 2 (0,0.15,0) point 3 (0,0,6) ...
group 'Left'
zone create brick size 15 1 20 point 0 (5.5,0,0) point 1 (10,0,0) ...
point 2 (5.5,0.15,0) point 3 (5.5,0,6) ...
group 'Right'
; --- interface ---
zone interface 1 create by-face range position-x 5.5
stiffness-normal 2e9 friction 15 tension 1e10
; zone interface 1 element maximum-edge 1 ; <==will make interface impermeable
zone gridpoint initialize position-x -1.0 add range group 'right'
zone fluid property permeability 1e-10 poros 0.3
zone fluid biot on
zone gridpoint initialize biot 3.333e3
FLAC3D 6.0

zone property bulk 2 shear 1
model cycle 0 ; <--- to initialize weighting factors
; --- settings ---
; --- test ---
@show_seepage_face
list @qflac @qsol
model save 'intfl'
FLAC3D 6.0
FLAC3D 6.0
Spreading of a Groundwater Mound 587
Spreading of a Groundwater Mound

This problem studies the transient evolution of a groundwater mound within a
porous medium. The mound spreads out and flows along an impervious base
under the influence of gravity. It is assumed that the fluid is incompressible, and
that the water is contained initially in a cylindrical region with radius and
height . The water saturation within the mound is equal to one, and Darcy’s
law is applicable. The mound elevation is compared to an analytic solution as the
elevation evolves with time.
Kochina et al. (1983) have derived the solution for the height, , of the mound.
The solution, as given by Barenblatt (1987), assumes a hydrostatic pore-pressure
distribution within the mound. In the case of zero residual saturation, it may be
expressed in the form
(1)
where , , , and the characteristic time is given as

with ( is the mobility coefficient, is water
density, is gravity and is porosity).
This solution applies to long time scales, when the influence of the details of the
initial mound geometry have disappeared.
The results are presented in dimensionless form: the scaled geometrical

parameters = 1, = 1 are used, and the scaled water properties = 0.5 × 10-4,
= 103, and = 0.5 are prescribed in the numerical simulation. To model
incompressible flow, the bulk modulus of the fluid, , is given a value that is
large compared to the pore-pressure variations in the simulation ( = 2 × 105);
the value = 10 is used for gravity. The FLAC3D grid corresponds to a quarter
cylinder and contains 2000 zones (see Figure 1). The radius of the model is 2, and
its height is 1 unit. The initial saturation is 1 within the mound (radius = 1 unit,
height = 1 unit), and zero outside. The initial pore-pressure distribution within
the mound is hydrostatic. All boundaries are impermeable by default. As time
goes on, the mound spreads out under its own weight. The simulation is
conducted for a total dimensionless time value of = 0.85, with intermediate
results at = 0.35, 0.45 and 0.65.
FLAC3D 6.0
Figure 1: FLAC3D grid and initial state of saturated column.
Saturation contours are sketched in Figure 2 to Figure 5. The analytic prediction

for the mound height is calculated by a set of FISH functions and plotted for
comparison in the figures (bold line).
Figure 2 corresponds to = 0.35; at that time, the initial shape of the mound still
persists, and a comparison with the analytic solution is probably not yet
appropriate. For larger times (see Figure 3 to Figure 5), the spreading of the
groundwater mound described by Equation (1) is captured by the numerical
solution with reasonable accuracy. The numerical estimate lags behind the
analytical prediction; the discrepancy may be explained by the occurrence of
residual saturation in the numerical solution and by the coarse discretization
used in the simulation.
FLAC3D 6.0
Figure 2: Saturation contours and analytical mound elevation at = 0.35.
FLAC3D 6.0
FLAC3D 6.0
The sketch of flow vectors and head contours in Figure 6 corresponds to = 0.85.
(The lack of smoothness in the contour plot is caused by the head jump across
the phreatic surface.) Two regions with water flowing predominantly downward
in the core of the mound and outward in its periphery can be seen in the figure.
(The vertical bold line is drawn at the location where the time derivative of the
analytical mound elevation vanishes: = 0.)
Figure 6: Head contours and analytical mound elevation at = 0.85
References
Barenblatt, G. Dimensional Analysis. Gordon and Breach Science Publishers (1987).
Kochina, I., N. Mikhailov and M. Filinov. “Groundwater Mound Damping,” Int. J.

Engng. Sci., 21, 413-421 (1983).
Data File
GroundwaterMound.f3dat
model new
FLAC3D 6.0

; --- geometry ---
zone create cylinder size 20 10 10 point 1 (0 2 0) point 2 (0 0 1) ...
point 3 (2 0 0)
zone fluid property permeability 0.5e-4 porosity 0.5 biot 1
zone fluid biot on
zone gridpoint initialize biot 1e5
; if fluid modulus is specified:
; set fl biot off
; ini fmodulus 0.5e5
zone gridpoint initialize fluid-tension 0
; --- initial conditions ---
model range create 'mound' cylinder end-1 0 0 -1 end-2 0 0 2 radius 1.01
zone gridpoint initialize saturation 1.0 range named-range mound
zone gridpoint initialize pore-pressure 1e4 grad 0 0 -1e4 ...
range named-range mound
; --- settings ---
model save 'mound0'
; --- test ---
model title "Spreading of a groundwater mound (t = 0.35)"
model save 'mound1'
model save 'mound2'
model save 'mound3'
model save 'mound4'
return
FLAC3D 6.0
One-Dimensional Consolidation 593
One-Dimensional Consolidation
A saturated layer of soil of thickness = 20 m (shown in Figure 1) and large
horizontal extent rests on a rigid impervious base. A constant surface load, =
5
10 Pa, is applied on the layer under undrained conditions. The soil matrix is
homogeneous and behaves elastically; the isotropic Darcy’s transport law
applies. The applied pressure is initially carried by the fluid, but as time goes on
the fluid drains through the layer surface, transferring the load to the soil
matrix. The solution to this one-dimensional consolidation problem may be
expressed in the framework of Biot theory (see Detournay and Cheng 1993).
Figure 1: One-dimensional consolidation.
The diffusion equation for the pore pressure, , has the form
(1)
where:
= ;
= = the storage coefficient;
FLAC3D 6.0
= ;
= is the Biot moduls; and
= is the Biot coefficient.
The boundary conditions are = and = 0 at =H( is the step

function), and = 0 and = 0 at = 0.
Because the stress is constant, Equation (1) reduces to
(2)
with boundary conditions = 0 at , and = 0 at = 0.
The initial value, , for the pore pressure induced from loading of the layer may
be derived from the fluid constitutive law ( ) by considering
undrained conditions ( = 0) and using the one-dimensional mechanical
constitutive law ( ) to express strain in terms of stresses. It is
given by
(3)
The solution for the pore pressure is
(4)
where:
= ;
= ;
= ; and
= .
The vertical displacement, , is found by considering the equilibrium equation

= 0 together with the mechanical constitutive equation .
By expressing as , we obtain upon integration, taking Equation (4) and
the boundary conditions into account,
(5)
FLAC3D 6.0
here is defined as .
The following properties are prescribed for this example:
dry bulk modulus, 5 × 108 Pa

dry shear modulus, 2 × 108 Pa
Biot modulus, 4 × 109 Pa
Biot coefficient, 1.0
permeability, 10-10 m2/Pa-sec
The FLAC3D model grid for this problem is a column of 20 zones of unit
dimensions lined up along the z-axis. (See Figure 2.) The base of the column is
fixed, and lateral displacements are restricted in the x- and y-directions. A
mechanical pressure, , is applied at the top of the column.
At first, flow is prevented and the model is stepped to equilibrium to establish

the initial undrained conditions. At the end of this stage, the stress has the
value , in equilibrium with the applied pressure, and the pore pressure has
the initial undrained value (see Figure 3). Drainage is then allowed
by setting the pore pressure to zero at the top of the column. The FLAC3D model
is cycled to a flow time of 5000 seconds, which is approximately the magnitude
of the characteristic time, , for this problem.
The analytical solution for the pore pressure, , and vertical displacement, , at
the column mid-height are evaluated using FISH functions, and compared to the
numerical solution as time proceeds. The results are plotted versus fluid-flow
time in Figure 3 and Figure 4. The transfer of pore pressure to effective stress is
illustrated in Figure 5, where the evolutions of normalized total stress, ,
effective stress, , and pore pressure, , with fluid-flow time are
presented. The FLAC3D data file is listed in “1DConsolidation-Coupled.f3dat”.
break
FLAC3D 6.0
Figure 2: FLAC3D grid for one-dimensional consolidation.
Figure 3: Comparison between analytical and numerical values of pore pressure

in a 1D consolidation test — = 4 × 109 Pa.
FLAC3D 6.0
Figure 4: Comparison between analytical and numerical values for vertical

displacement in a 1D consolidation test — = 4 × 109 Pa.
Figure 5: Evolution of pore pressure, total and effective stresses in a 1D

consolidation test — = 4 × 109 Pa.
FLAC3D 6.0
Note that as the stiffness ratio increases, the number of

zones should increase to keep the error small. This effect is caused by the
numerical technique used to evaluate pore-pressure changes caused by
volumetric strain in model configure fluid. Also, as the ratio increases, the
timestep will decrease. However, to keep the error small and the timestep
sufficiently large without significantly affecting the solution, it is acceptable to
limit the value of Biot modulus (or ) used in the simulation to a value of
approximately twenty times (see the fifth item in Solving Flow-
Only and Coupled-Flow Problems). To demonstrate the validity of this numerical
approach, the analytical solutions for pore pressure and displacement
corresponding to = 1.5 × 1010 ( 20) and = 1.5 × 1015 ( 20 × 105) are
compared in Figure 6 and Figure 7. As seen in those plots, the responses are
indeed similar (maximum relative error less than 5%).
break
FLAC3D 6.0
Figure 6: Comparison between analytical pore-pressure solutions for two large

values of M in a 1D consolidation test.
Figure 7: Comparison between analytical vertical displacement solutions for

two large values of M in a 1D consolidation test.
FLAC3D 6.0
For values of the stiffness ratio, , much larger than 20, the convergence of the
basic algorithm for coupled fluid flow will be very slow (i.e., numerous
mechanical steps will be necessary to bring the system to mechanical equilibrium
after each flow step). This is illustrated by repeating teh couple test with Biot
modulus = 4 × 1010 Pa (10 times larger than in the former numerical
calculation, see 1DConsolidation-Coupled10M.f3dat). In the former calculation
with = 4 × 109 Pa ( = 5.2), the coupled flow calculation ran in approximately
4 seconds (on a 1.6 GHz i7 computer). For the calculation with = 4 × 1010 Pa (
= 52), the calculation time is approximately 40 seconds (ten times slower).
The comparison of the pore pressure and displacement history results to the
analytical solutions for = 4 × 1010 Pa are shown in Figure 8 and Figure 9,
respectively. The maximum relative error is less than 4%.
break
FLAC3D 6.0
Figure 8: Comparison between analytical and numerical values of pore pressure

in a 1D consolidation test - = 4 × 1010 Pa.

displacement in a 1D consolidation test - = 4 × 1010 Pa.
FLAC3D 6.0
The solution time can be reduced by using the uncoupling technique described in
Stiffness Ratio (the applicability of the uncoupling technique to this problem
follows from the irrotational character of the solution). In this test
(1DConsolidation-Uncoupled.f3dat), the numerical simulation is repeated for a
Biot modulus of 4 × 1010 Pa using the uncoupling technique. In this test, the
initial undrained conditions are obtained using the undrained bulk value for the
material and setting Biot modulus to zero. The pore pressure, which is not
updated in this calculation mode, is then initialized at the value . The rest of
the simulation is carried out in a series of ten time increments to enable a
recording of the variables history. In each increment, the fluid flow calculation is
performed for a time interval of 500 seconds. During that stage, a scaled value
for Biot modulus is used in order to preserve the coupled system true diffusivity
(see this equation). Next, Biot modulus is set to zero to prevent additional
generation of pore pressure, and the system is run to mechanical equilibrium
using the drained value of the bulk modulus. Fluid and mechanical calculations
are repeated until the total simulation time reaches 5000 sec. The results of this
simulation (which does not involve any direct calculation of pore-pressure
change due to volumetric straining) are presented in Figure 10 to Figure 12. The
calculation time for this uncoupled simulation is approximately 4 seconds. The
maximum relative error is less than 1%.
break
FLAC3D 6.0
Figure 10: Comparison between analytical and numerical values of pore

pressure in a 1D consolidation test - = 4 × 1010 Pa.
displacement in a 1D consolidation test - = 4 × 1010 Pa.
FLAC3D 6.0
Figure 12: Evolution of pore pressure, total and effective stresses in a 1D

consolidation test - = 4 × 1010 Pa.
This particular example involves fully saturated coupled flow. In such cases,
solution times can be reduced by using the saturated fast-flow algorithm
described in Fully Saturated Fast Flow. The algorithm is selected by using the
zone fluid fastflow on command. If 1DConsolidation-Coupled.f3dat with =4×
1010 Pa ( = 52) is repeated with this command, the runtime is now
approximately 12 seconds (see 1DConsolidation-FastFlow.f3dat). As increases,
the use of the fast-flow algorithm becomes more beneficial.
The results of the fast-flow calculation are shown in Figure 13 and Figure 14. The
maximum relative error is less than 2%.
break
FLAC3D 6.0
Figure 13: Comparison between analytical and numerical values of pore

pressure in a 1D consolidation test.
displacement in a 1D consolidation test.
FLAC3D 6.0
Reference
Detournay, E., and A. H. D. Cheng. “Fundamentals of Poroelasticity,” in
Comprehensive Rock Engineering, Vol. 2, pp. 113-171. J. Hudson et al., eds. London:
Pergamon Press (1993).
Data File
1DConsolidation-Coupled.f3dat
model new
model title "One-dimensional consolidation (coupled)"
@setup(4.0e9)
; --- model geometry ---
zone create brick size 1 1 [int(hh)]
zone face skin
zone fluid biot on
zone property bulk @c_bulk sh @c_shear
zone face apply velocity-x 0.0 range group 'West' or 'East'
zone face apply velocity-y 0.0 range group 'North' or 'South'
zone face apply velocity-z 0.0 range group 'Bottom'
zone initialize stress-zz 0.
zone face apply stress-zz @sig0 range group 'Top'
zone fluid property permeability @c_perm biot @c_biotc
zone gridpoint initialize biot @c_biotm
zone gridpoint initialize pore-pressure 0
; --- first establish undrained response ---
; --- histories ---
fish history @pp10
fish history @c_szz
fish history @c_eszz
fish history @c_uz
fish history @ft
; --- drained response ---
zone face apply pore-pressure 0 range group 'Top'
model solve fluid time-total 500 or mechanical ratio 1e-4
FLAC3D 6.0

history export 1 vs 5 table 1
table 1 label 'Pore Pressure'
table 2 label 'Vertical Displacement'
table 3 label 'Total Stress'
table 4 label 'Effective Stress'
model save '1dcons_coup'
return
1DConsolidation-Coupled10M.f3dat
model new
model title "One-dimensional consolidation (coupled)"
@setup(4.0e10)
zone face skin
zone fluid biot on
; --- histories ---
fish history @pp10
fish history @c_szz
fish history @c_uz
fish history @ft
FLAC3D 6.0

model save '1dcons_coup10M'
return
1DConsolidation-Uncoupled.f3dat
model new
model title "One-dimensional consolidation (uncoupled)"
@setup(4.0e10)
zone face skin
zone property bulk @c_bulk shear @c_shear
zone fluid biot on
model fluid active off mech on
zone property bulk @c_bulku
zone gridpoint initialize biot 0.0
model solve
zone gridpoint initialize pore-pressure @p0
FLAC3D 6.0

fish define my_solve
loop local ii (1,10)
global c_age = 500.0*ii
command
zone property bulk @c_bulk
zone gridpoint initialize biot @c_biotma
model fluid active on mech off
model solve fluid time-total @c_age
zone gridpoint initialize biot 0.0
end_command
table.y(1,ii) = pp10
table.x(1,ii) = ft
table.y(2,ii) = c_uz
table.x(2,ii) = ft
table.y(3,ii) = c_szz
table.x(3,ii) = ft
table.y(4,ii) = c_eszz
table.x(4,ii) = ft
end_loop
table.label(1) = 'Pore Pressure'
table.label(2) = 'Vertical Displacement'
table.label(3) = 'Total Stress'
table.label(4) = 'Effective Stress'
end
@my_solve
model save '1dcons_uncoup'
return
1DConsolidation-FastFlow.f3dat
model new
model title "One-dimensional consolidation (coupled - fast flow)"
@setup(4.0e10)
zone face skin
; set fluid biot on
FLAC3D 6.0

zone fluid property permeability @c_perm
; --- fast flow setting ---
zone fluid fastflow on
; --- histories ---
fish history @pp10
fish history @c_szz
fish history @c_uz
fish history @ft
model step 0
model fluid timestep fix 1
model fluid timestep auto
model save '1dcons_ff'
return
FLAC3D 6.0
Consolidation Settlement at the Center of a Strip Load 611
Consolidation Settlement at the Center of a Strip

Load
The prediction of the time settlement of foundations on soft saturated soil is a
problem of wide practical interest in soil mechanics. The fluid can in general be
considered as incompressible compared to the soil matrix for the simulations,
and thus this type of problem is ideal for testing the saturated fast-flow scheme in
FLAC3D, described in Fully Saturated Fast Flow. This verification problem
involves the consolidation settlement produced by a uniform strip loading on a
semi-infinite elastic medium, which is free to drain at the surface.
The two-dimensional poro-elastic problem of a semi-infinite stratum loaded on

the plane = 10 by a uniform normal pressure, , along the strip has
been analyzed by McNamee and Gibson (1960). The fluid is assumed to be
incompressible in their solution. The case in which the pore pressure vanishes at
= 10 (i.e., a completely permeable surface) is considered. The consolidation
surface settlement for the special case of zero Poisson’s ratio is given by the
closed form solution
(1)
where
(2)
(3)
and are the nondimensional total vertical displacement and undrained

vertical displacement, respectively, at = 0. is the nondimensional horizontal
distance, and is the nondimensional shear modulus. These quantities are
defined:
(4)
where is the vertical displacement, and is the shear modulus.
FLAC3D 6.0
erf is the error function and is the exponential integral. is a dimensionless

time parameter, given as
(5)
where is the scaled diffusivity, and is time. For zero Poisson’s ratio, is
defined as
(6)
where is the mobility coefficient.
For the numerical simulation with FLAC3D, = 1, = 0.5, = 1, and = 1. The

problem conditions are simulated under plane-strain conditions, and advantage
is taken of the problem geometry half-symmetry.
The FLAC3D model grid used for this analysis is shown in Figure 1. The grid has
dimensions of 10 by 1 by 10, and a total of 1287 zones. The origin of the reference
axes is located at the strip centerline, the z-axis is pointing up, the -axis is
oriented to the right, and the y-axis is pointing into the plane of analysis. The
isotropic elastic material model is used with the shear modulus equal to 0.5 and
Poisson’s ratio equal to 0. Roller boundaries are specified along the symmetry
and bottom boundaries. The right boundary is fixed in all three directions.
Displacements are fixed in the y-direction to simulate plane-strain conditions.
The simulation is run using the fluid configuration (model configure fluid) with
an initial saturation of 1 throughout the model. Pore pressure is fixed at zero at
the surface of the model ( = 10) to model free draining conditions.
FLAC3D 6.0
Figure 1: FLAC3D grid.
A uniform mechanical pressure of = 1 is applied along the strip location to

simulate the strip loading. The fastflow option (zone fluid fastflow on) is
selected. The undrained response is calculated first (with model fluid active off
and the model stepped to equilibrium). Displacements are then reset to zero, and
a coupled fluid-mechanical simulation is performed (using model fluid active on
and model mechanical active on) for a period of 8 time units.
The predicted time history of the surface settlement at the center of the strip is
compared to the analytical value in Figure 2. The difference between results over
the course of the simulation is less than 1%.
FLAC3D 6.0
Figure 2: Consolidation settlement ( ) at the center of the strip versus

log( ) — comparison between FLAC3D and analytical solutions.
The surface consolidation settlement calculated by FLAC3D at the end of the

simulation ( = 8) is compared to the analytical solution in Figure 3. The results
are in good agreement. The discrepancy observed away from the strip location is
attributed to boundary effects. This discrepancy can be reduced by increasing the
model lateral and vertical extents.
FLAC3D 6.0
Figure 3: Surface settlement profile at = 8 — comparison between FLAC3D and

analytical solutions.
The predicted excess pore pressure at = 0, = 0, = 9.5, scaled by the

undrained pore pressure value, is plotted versus log( ) in Figure 4. The FLAC3D
results are compared to the reference solution obtained by Schiffman et al.
(1969) and reported by Burghinoli et al. (2001).
An interesting observation is that pore pressure rises initially above the initial
undrained value before decaying to zero. This behavior is typical of the Mandel-
Cryer effect (Cryer 1963). The peak pressure is slightly underestimated by the
FLAC3D solution, but a fair agreement is obtained at later times.
Displacement contours at = 8 are shown in Figure 5.

break
FLAC3D 6.0
Figure 4: Scaled excess pore pressure at = -0.5 versus log( ) — comparison

between FLAC3D and reference solution given by Schiffman et al. (1969).
Figure 5: Displacement contours at = 8.
FLAC3D 6.0
References
Burghinoli, A., S. Miliziano and F. M. Soccodato. “Effectiveness of the fast-flow
algorithm: 2D consolidation benchmark and tunneling application,” FLAC and
Numerical Modeling in Geomechanics — 2001 (Proceedings of the 2nd International
FLAC Symposium, Lyon, France, October 29 - 31, 2001), pp. 345 - 352. D. Billaux et
al., eds. Swets & Zeitlinger (2001).
Cryer, C. W. “A Comparison of the Three-Dimensional Consolidation Theories of

Biot and Terzaghi,” Quart. J. Mech. and Appl. Math., XVI, 4, 401-412 (1963).
McNamee, J., and R. E. Gibson. “Plane strain and axially symmetric problems of
the consolidation of a semi-infinite clay stratum,” Quart. J. Mech. and Appl. Math.
XIII, Pt. 2, (1960).
Schiffman, R. L., A. Chen and J. C. Jordan. “An Analysis of Consolidation

Theories,” J. Soil Mech. and Found. Div., ASCE, 95(SM1), 285-312 (1969).
break
FLAC3D 6.0
Data File
ConsolidationSettlement.f3dat
model new
model title "Strip load on semi-infinite elastic medium"
call 'geometry'
zone generate from-building-blocks
zone face skin
zone gridpoint group 'output' range position (0,0,10) (4,0,10)
;@setup
zone property bulk=[1.0/3.0] shear=0.5
zone face apply velocity-x 0 range group 'West'
zone face apply velocity-z 0 range group 'Bottom'
zone face apply velocity-y 0 range group 'North' or 'South'
zone face apply velocity (0,0,0) range group 'East'
; --- apply load slowly ---
fish define ramp
ramp = math.min(1.0,float(global.step)/1000.0)
end
zone face apply stress-normal -1.0 fish @ramp range position-x 0 1 group 'Top'
zone fluid property permeability=1.0 porosity 0.3
; --- pore pressure fixed at zero at the surface ---
; --- settings ---
zone fluid fastflow on
model step 0
; --- undrained response ---
model save 'undrained'
; --- consolidation settlement ---
model step 0
;
[global gpnt0 = gp.near(0,0,10)]
[global gpnt1 = gp.near(0,0,9.5)]
[global u0 = gp.pp(gpnt1)]
fish define zd0
zd0 = gp.disp.z(gpnt0)
FLAC3D 6.0
global uu0 = gp.pp(gpnt1) / u0

global tau = zone.fluid.time.total*1.0 ; <-- cv = 1.0
global cons = ana(0,tau,0.5) ; <-- sh = 0.5
end
fish history @zd0
fish history @cons
fish history @uu0
fish history @tau
;
model fluid timestep fix 1e-4
history interval 50
model solve fluid time-total 0.1 or mechanical ratio 1e-5 ; t = 0.1
model fluid timestep auto
history interval 20
model solve fluid time-total 8.0 or mechanical ratio 1e-5 ; t = 8.0
; --- settlement ---
fish define css
loop foreach local pnt gp.list
if gp.group(pnt) = 'output' then
local x = gp.pos.x(pnt)
table(1,x) = ana(x,tau,0.5)
table(11,x) = -gp.disp.z(pnt)
endif
endloop
end
@css
; --- pressure ---
call 'u-nu0p0.dat' ; table 50
model save 'strip8'
return
FLAC3D 6.0
FLAC3D 6.0
Transient Fluid Flow to a Well in a Shallow Confined Aquifer 621
Transient Fluid Flow to a Well in a Shallow

Confined Aquifer
A shallow confined aquifer of large horizontal extent is characterized by a
uniform initial pore pressure, , and initial isotropic stress, . A well, fully
penetrating the aquifer, is producing water at a constant rate, , per unit depth
from time, . The elastic porous medium is homogeneous and isotropic, and
the flow of groundwater is governed by Darcy’s law. Transient effects are linked
to the compressibility of water and the soil matrix. In this problem, the effect of
pore-pressure changes are small compared to the overburden, and the vertical
stress in the aquifer may be assumed to remain constant with time. Also,
horizontal strains are neglected compared to the vertical ones. The conditions of
fluid flow to the well are illustrated schematically in Figure 1. The numerical
solution to this problem is presented using both coupled and uncoupled
modeling approaches.
Figure 1: Flow to a well in a shallow confined aquifer.
A cylindrical system of coordinates is chosen with the z-axis pointing upward in

the direction of the well axis. Substitution of the transport law in the fluid mass-
balance equation gives, taking into consideration that = 0,
FLAC3D 6.0
(1)
where is the homogeneous permeability coefficient, is the Biot modulus, and

is the Biot coefficient.
Partial differentiation, with respect to time, of the elastic constitutive relation

yields, for constant ,
(2)
where .
Using this last equation to express in terms of in Equation (1), we obtain,

after some manipulations,
(3)
where is the diffusion coefficient, is the storage

coefficient, and, with the problem being axisymmetric and not dependent on ,
the Laplacian of may be expressed as
(4)
The solution to this differential equation with boundary conditions
(5)
(6)
is due to Theis (1935). It has the form
(7)
where . The dimensionless variable is given by
(8)
and is the exponential integral, defined as
(9)
FLAC3D 6.0
The vertical displacement may be obtained by integration of the equilibrium

equation = 0 after expressing in terms of by means of the
mechanical constitutive equation and substituting for . This yields,
after substitution of the boundary condition, and using Equation (7),
(10)
where and .
The stresses are derived from the mechanical constitutive equations and
Equation (7) for . They have the form
(11)
where .
The FLAC3D grid for this problem corresponds to a nine-degree wedge in a

hollow cylinder of unit height. The axis and radius of the well correspond to
cylinder axis and radius, respectively. The cylinder outer radius is selected as 100
m to model the far boundary of the flow domain. Figure 2 shows the FLAC3D grid.
Figure 2: FLAC3D grid for a well in a shallow aquifer.
FLAC3D 6.0
A total of 31 zones are used, lined up, and graded in the radial direction. The
displacements are fixed in the radial and tangential directions, and in the vertical
direction at the cylinder base. A vertical pressure of magnitude is applied at
the top of the model.
The properties for this example are defined:
dry bulk modulus ( ) 118 MPa

dry shear modulus ( ) 71 MPa
water bulk modulus ( ) 2 GPa
Biot coefficient ( ) 1.0
porosity ( ) 0.4
permeability ( ) 2.98 × 10-8
The initial pore pressure is 147 kPa, and the initial isotropic stress is -147 kPa.
Because the Biot coefficient is equal to one (incompressible soil grains), the Biot
modulus is equal to the ratio between water bulk modulus and porosity (in this
case, = 5 GPa). The well-pumping rate per unit aquifer thickness is 2.21 10 -3
m2/s, and the well radius, , is selected as 1 m.
Stresses and pore pressures are initialized to the values given above. The well
flow-rate is modeled as a surface flux of magnitude applied to the well
radius .
The coupled problem is solved (model fluid active on and model mechanical
active on) using the explicit solution algorithm. The maximum out-of-balance
mechanical force is limited to 10.0, the maximum number of mechanical sub-
steps in the coupled fluid-mechanical calculation step is limited to 1000, and the
mechanical process is the “slave” module to the master fluid-flow process. This
is accomplished with the commands

By specifying these commands, the out-of-balance mechanical force will be kept

to a small value while the fluid-flow calculation proceeds.
This example is pore-pressure driven, and the value for the stiffness ratio, , is
approximately 23 for the specified fluid bulk modulus of 2 GPa. Thus, the flow
calculation may be uncoupled from the mechanical calculation, and the approach
FLAC3D 6.0
discussed in Stiffness Ratio may be applied. The fluid modulus during the flow-
only step is defined by Equation (7) in order to preserve the diffusivity of the
system. During the mechanical-only step, the fluid modulus is set to zero to
prevent further adjustments by volumetric strains. The following commands are
applied for the uncoupled calculation for a 4-second flow time:

zone gridpoint initialize fluid-modulus = @uwb
model solve time-total 4.
zone gridpoint initialize fluid-modulus = 0.0
model solve unbalanced-maximum 10.
The FISH variable uwb is the adjusted fluid modulus calculated by Equation (7).
Note that, if conditions are such that , it is not necessary to adjust the
fluid modulus during the flow calculation because the diffusivity will be accurate.
The analytical solutions for pore pressure, stresses, and vertical displacement are
programmed as a FISH function. The exponential integral function used in the
analytical solutions is programmed as a separate FISH function. Analytical and
numerical values are stored in tables. The results are then compared in graphical
form. The pore-pressure comparison at selected times is presented in Figure 3
for the coupled solution, and Figure 4 for the uncoupled solution. The vertical
displacement values and stresses at 32 seconds are processed by the FISH
function well_32 and are illustrated in Figure 5 and Figure 7 for the coupled
solution, and in Figure 6 and Figure 8 for the uncoupled solution.
The results for both the coupled and uncoupled solutions are essentially identical
and compare well with the analytical solution. The uncoupled solution is reached
much more quickly than the coupled solution. Note that the coupled calculation
requires more than 500,000 steps, while the uncoupled calculation requires
approximately 16,000 steps.
FLAC3D 6.0
Figure 3: Pore-pressure distribution at 4, 8, 16, and 32 seconds—coupled

solution (analytical values = lines; numerical values = symbols).
Figure 4: Pore-pressure distribution at 4, 8, 16, and 32 seconds—uncoupled

solution (analytical values = lines; numerical values = symbols).
FLAC3D 6.0
Figure 5: Radial, tangential, and vertical stress distributions at 32 seconds —

coupled solution (analytical values = lines; numerical values = symbols).
Figure 6: Radial, tangential, and vertical stress distributions at 32 seconds —

uncoupled solution (analytical values = lines; numerical values = symbols).
FLAC3D 6.0
Figure 7: Radial displacement distribution at 32 seconds—coupled solution

(analytical values = line; numerical values = symbols).
Figure 8: Radial displacement distribution at 32 seconds—uncoupled solution

(analytical values = line; numerical values = symbols).
FLAC3D 6.0
Reference
Theis, C. V. “The Relation between the Lowering of the Piezometric Surface and
the Rate and Duration of Discharge of a Well Using Groundwater Storage,” Trans.
Am. Geophys. Union, 10, 519-524 (1935).
Data File
WellConfinedAquifer.f3dat
model new
model title 'Transient fluid flow to a well in a shallow confined aquifer'
@constants
; --- model geometry (hollow cylinder - 9 degree wedge) ---
zone create brick point 0 (1,0,0) point 1 (100,0,0) ...
point 2 (0.987688,0.156434,0) point 3 (1,0,1) ...
point 4 (98.7688,15.6434,0) ...
point 5 (0.987688,0.156434,1) ...
point 6 (100,0,1) point 7 (98.7688,15.6434,1) ...
size (31,1,1) ratio (1.1,1,1)
zone face skin
;model range create 'in' cyl end-1 (0,0,-100) end-2 (0,0,100) radius @rin
;model range create 'out' cyl end-1 (0,0,-100) end-2 (0,0,100) radius 99. not
zone gridpoint group 'xline1' range position (1,0,0) (100,0,0)
zone gridpoint group 'xline2' range position (5,0,1) (100,0,1)
zone property bulk @c_k shear @c_g
zone face apply velocity-x 0.0
zone face apply velocity-y 0.0
zone initialize stress xx @sig0 yy @sig0 zz @sig0
zone fluid property permeability @c_kw porosity @c_poro
zone gridpoint initialize fluid-modulus @w_bulk
zone gridpoint initialize pore-pressure @p0
zone face apply discharge @qin range group 'West'
; --- setting ---
model save 'well-ini'
; --- pumping (coupled analysis) ---
model solve fluid time-total 4. or mechanical unbalanced-maximum 10.
FLAC3D 6.0
@sol_pp(1,2)
@sol_pp(3,4)
@sol_pp(5,6)
@sol_pp(7,8)
model save 'well-cpl'
; --- pumping (uncoupled analysis) ---
model restore 'well-ini'
model solve fluid time-total 4.
model solve unbalanced-maximum 10.
@sol_pp(1,2)
model solve unbalanced-maximum 10.0
@sol_pp(3,4)
@sol_pp(5,6)
@sol_pp(7,8)
model save 'well-ucpl'
return
FLAC3D 6.0
Pressuremeter Test 631
Pressuremeter Test
The pressuremeter test is used to determine in-situ mechanical properties of
soils (Wood 1990). A long, rubber membrane is expanded against the walls of a
vertical borehole (Figure 1). The pressure inside the membrane is constant.
Radial displacements of the borehole wall are measured as a function of the
pressure. The soil deforms in plane strain in the plane normal to the borehole
and sufficiently distant from the ends of the membrane.
The borehole (radius = 0.03 m) is drilled in homogeneous, isotropic soil. The

soil is assumed to behave as a linearly elastic material saturated with
groundwater. Several mechanical properties of the soil and groundwater are
assumed in this problem:
shear modulus ( ) 11.1 MPa

bulk modulus ( ) 33.3 MPa
porosity ( ) 0.48
soil permeability ( ) 1.02 × 10-14

bulk modulus of water ( ) 500 MPa
FLAC3D 6.0
Figure 1: Cylindrical cavity expansion in pressuremeter test.
The initial state of (total) stress in the soil is isotropic: = -327.87

kPa, while the initial pore pressure is = 147.0 kPa.
The soil is allowed to consolidate for 300 seconds after the drilling of the
borehole. After the consolidation, the pressure inside the rubber membrane is
increased from zero to 1 MPa during 600 seconds. (The rubber membrane
prevents further drainage of the groundwater into the borehole.)
The analytical solution for two-dimensional consolidation of a borehole in an

elastic medium was obtained by Detournay and Cheng (1988). The drilling of the
borehole is simulated by removing the stress acting on the inner boundary of the
borehole and setting the pore pressure to zero at time = 0. Since the initial
stress is isotropic, the loading conditions can be decomposed into two modes: (1)
mode 1, an isotropic stress; and (2) mode 2, an initial pore-pressure distribution.
The boundary conditions at the wall of the borehole for each loading mode can be
expressed:
1. mode 1
FLAC3D 6.0
2. mode 2
The stresses and displacements due to mode 1 loading are described by the
classical Lamé solution. Because the volumetric strain computed from the Lamé
solution is zero throughout the domain, the mode 1 loading does not generate
pore pressure, and deformation takes place instantaneously. The evolution of the
pore-pressure field due to mode 2 loading is governed by a homogeneous
diffusion equation. The deformation and stress fields can be calculated from the
pore-pressure field. The problem is solved in the Laplace transform domain, and
the solutions are transformed back to the time domain using the numerical
inversion method developed by Stehfest (1970). The complete solution is
described by Detournay and Cheng (1988). The analytical solutions are calculated
and imported into FLAC3D tables for comparison to the numerical results.
Because the problem is axisymmetric, it is simulated using a row of 61 zones

forming a truncated wedge of 9° angle. The geometry of the model and boundary
conditions are illustrated in Figure 2. The FLAC3D grid is shown in Figure 3. The
far-field boundary is at radius = 1.50 m and can be considered at infinity if
radius is scaled to the radius of the borehole ( = 0.03 m). (The length resulting
from the diffusivity of the model and time of the simulation are also much
smaller than .)
FLAC3D 6.0
Figure 2: Domain of FLAC3D simulation.
Figure 3: FLAC3D grid.
The simulation is conducted in the steps corresponding to the actual operations

in the field test:
FLAC3D 6.0
(1) The total stress on the contour of the borehole is reduced to zero, simulating
drilling of the borehole. Although the total stress (in the FLAC3D model) is
reduced to zero in steps, the change in real flow time is instantaneous (i.e., the
model undergoes undrained deformation). (The fluid flow calculation is turned
off: model fluid active off.) The model is iterated to reach mechanical
equilibrium.
(2) The pressure boundary condition at the contour of the borehole is set to zero.
The model consolidates for 300 seconds (model fluid active on), resulting in the
drainage of the groundwater into the borehole.
(3) The contour of the borehole is defined as impervious (due to installation of

the rubber membrane), and the pressure boundary condition is applied on the
contour of the borehole: 1.0 MPa in 100 increments at each 6 seconds. That is, the
soil consolidates under the applied load for 6 seconds before the next load
increment (0.01 MPa) is applied.
The profiles of the normalized pore pressure, , normalized radial stress,

, and tangential stress, (as a function of the normalized radius ),
after 300 seconds of consolidation, calculated from FLAC3D and using the closed-
form solution from Detournay and Cheng (1988), are shown in Figure 4 and
Figure 5. Agreement between the curves is very good.
The profiles of the normalized pore pressures, and the radial and tangential
stresses after 600 seconds of pressurization of the borehole are shown in Figure
6 and Figure 7.
FLAC3D 6.0
Figure 4: Pore-pressure profiles—300 seconds consolidation.
Figure 5: Profiles of radial and tangential normal stresses—300 seconds

consolidation.
FLAC3D 6.0
Figure 6: Pore-pressure profiles—600 seconds consolidation.

consolidation.
FLAC3D 6.0
The pressure variations in the pressuremeter test are often such that nonlinear,
plastic deformations are induced in a soil. Therefore, the same problem is
simulated using a Mohr-Coulomb model for plastic deformation of the soil.
Three parameters are assumed in the simulation:

cohesion ( ) 26 kN/m2
The profiles of the normalized pore pressures, and the normalized radial and
tangential stresses after 600 seconds of pressurization of the borehole in a
Mohr-Coulomb material, are shown in Figure 8 and Figure 9.
The numerical solution for a linearly elastic material is generated using the data
file listed in “Pressuremeter-Elastic.f3dat”. The input data file for a Mohr-
Coulomb material is the same, except that (1) the material model is declared a
Mohr-Coulomb material (zone cmodel assign mohr-coulomb ), and (2) the
corresponding material properties are added (zone property bulk 3.33e7 shear
1.11e7 friction 30 dilation 10 cohesion 26000 ). The data file “fishFunctions”
generates the tables with the profiles of the normalized pore pressure and the
normalized stresses (see “Pressuremeter-MC.f3dat”). The data file “preana”
contains tables in which the analytical elastic solutions for pore pressure, radial
and tangential stress profiles after 300 seconds of pressurization are stored.
break
FLAC3D 6.0
Figure 8: Pore-pressure profile—600 seconds consolidation (Mohr-Coulomb

material).

consolidation (Mohr-Coulomb material).
FLAC3D 6.0
References
Detournay, E., and A. H. D. Cheng. “Poroelastic Response of a Borehole in a Non-
Hydrostatic Stress Field,” Int. J. Rock Mech. Sci. & Geomech. Abstr., 25(3), 171-182
(1988).
Stehfest, H. “Numerical Inversion of Laplace Transforms,” Communic. Ass.

Comput. Mach., 13, 47-49 (1970).
Data Files
Pressuremeter-Elastic.f3dat
; pressuremeter test in poro-elastic material, (a) Elastic model

model new
model title 'Pressuremeter - elastic model'
call 'fishFunctions' suppress
[global time0 = time.clock]
zone create brick ...
point 0 0.03 0 0 point 1 1.5 0 0 ...
point 2 0.03 0 -0.01 point 4 1.5 0 -0.01 ...
point 3 2.96307e-2 0.46930e-2 0 point 6 1.48153 0.23465 0 ...
point 5 2.96307e-2 0.46930e-2 -0.01 point 7 1.48153 0.23465 -0.01 ...
size 61 1 1 ratio 1.1 1 1
zone face skin
; --- mechanical model, elastic ---
zone property bulk 3.33e7 shear 1.11e7
; --- gw model ---
zone fluid cmodel isotropic
zone fluid biot on
zone fluid property permeability 1.02e-14
zone gridpoint initialize biot [5.0e8/0.48]
; -- boundary conditions for all but 'West' stress
zone face apply stress-normal -327870 range group 'East'
zone gridpoint fix pore-pressure range group 'West'
zone face apply velocity-normal 0 range group 'North'
zone face apply velocity-normal 0 range group 'South'
zone face apply velocity-normal 0 range group 'Top'
; --- excavate: reduction of total pressure to zero in steps ---
[global factor = 0.99]
FLAC3D 6.0
fish define slow_excav

if zone.mech.ratio < 1e-5 then
factor = math.max(0, factor - 0.0001)
endif
slow_excav = factor
end
zone face apply stress-normal -327870 fish @slow_excav range group 'West'
model save 'pre1'
; --- let water flow out for 300 s ---
zone gridpoint initialize pore-pressure 0 range group 'West'
model save 'pre2'
; --- apply pressure inside the borehole ---
zone gridpoint free pore-pressure range group 'West'
fish define charge
charge = math.max(0.0,fluid.time.total-300.0) / 600.0
end
zone face apply stress-normal -1e6 fish @charge range group 'West'
[global time = (time.clock-time0)/100.0]
list @time
model save 'pre3'
return
break
FLAC3D 6.0
Pressuremeter-MC.f3dat
; pressuremeter test in poro-elastic material, (b) Mohr-Coulomb model

model new
model title 'Pressuremeter - Mohr-Coulomb model'
call 'fishFunctions' suppress
[global time0=time.clock]
point 0 0.03 0 0 point 1 1.5 0 0 ...
point 2 0.03 0 -0.01 point 4 1.5 0 -0.01 ...
point 3 2.96307e-2 0.46930e-2 0 point 6 1.48153 0.23465 0 ...
point 5 2.96307e-2 0.46930e-2 -0.01 point 7 1.48153 0.23465 -0.01 ...
size 61 1 1 ratio 1.1 1 1
zone face skin
; --- mechanical model, Mohr-Coulomb ---
zone property bulk 3.33e7 shear 1.11e7 friction 30 dilation 10 cohesion 26000
; --- gw model ---
zone fluid biot on
zone fluid property permeability 1.02e-14
zone gridpoint initialize biot [5.0e8/4.8]
; -- boundary conditions for all but 'West' stress
zone face apply stress-normal -327870 range group 'East'
zone gridpoint fix pore-pressure range group 'West'
zone face apply velocity-normal 0 range group 'North'
zone face apply velocity-normal 0 range group 'South'
zone face apply velocity-normal 0 range group 'Top'
; --- excavate: reduction of total pressure to zero in steps ---
[global factor=0.99]
fish define slow_excav
if zone.mech.ratio<1e-5 then
factor = math.max(0,factor - 0.0001)
endif
slow_excav = factor
end
zone face apply stress-normal -327870 fish @slow_excav range group 'West'
model save 'prm1'
; --- let water flow out for 300 s ---
zone gridpoint initialize pore-pressure 0 range group 'West'
model mechanical slave
model save 'prm2'
FLAC3D 6.0
; --- apply pressure inside the borehole ---

zone gridpoint free pore-pressure range group 'West'
fish define charge
charge = math.max(0.0,fluid.time.total-300.0) / 600.0
end
zone face apply stress-normal -1e6 fish @charge range group 'West'
[global time = (time.clock-time0)/100.0]
list @time
@maket
table 1 label 'num. pp t=600s'
table 2 label 'num. sigr t=600s'
table 3 label 'num. sigt t=600s'
table 4 label 'num. sigz t=600s'
model save 'prm3'
return
FLAC3D 6.0
FLAC3D 6.0
Semi-confined Aquifer 645
Semi-confined Aquifer
Fluid leakage into a shallow semi-confined aquifer can be modeled with FLAC3D
using the zone face apply leakage command. This is demonstrated for the
example defined by the sketch in Figure 1. The aquifer has a length , height ,
and rests on an impermeable base. Fluid flow obeys Darcy’s law; the mobility
coefficient is homogeneous and isotropic. The semi-permeable top layer has
permeability , and thickness . The effect of gravity is neglected in this
example. Fluid pressure at the top of the leaky layer is constant and equal to .
The lateral fluid-flow conditions correspond to a constant pressure at the left
boundary, and at the right.
The objective is to determine the steady-state pore-pressure profile and total

leakage into the aquifer. The general solution of pore pressure for a shallow
semi-confined aquifer has the form (see Strack 1989)
(1)
where is the seepage factor, which has the dimension of length and is defined
as ; and are constants determined from the pressure
boundary conditions.
Figure 1: Shallow semi-confined aquifer.
The boundary conditions for this problem are
at
FLAC3D 6.0
at
The pore pressure solution is
(2)
where
(3)
(4)
The steady state discharge over the height of the aquifer is obtained from Darcy’s
law:
(5)
After differentiation of Equation (2) with respect to , and substitution into

Equation (5), we obtain
(6)
The total amount of leakage into the aquifer is, by continuity of flow, equal to the
difference between the discharge leaving at = and that entering at = 0.
Using Equation (6), we obtain, after some manipulation,
(7)
Equation (6) and Equation (7) are used for comparison to the FLAC3D solution.
The FLAC3D data file for this problem is listed in “SemiConfinedAquifer.f3dat”.

The analytical solution is programmed in FISH as part of the data file. The
FLAC3D model is a 20 zone by 2 zone mesh with a constant pore pressure of =
20 kPa applied at the left boundary, = 0, and a constant pore pressure of = 10
kPa applied at the right boundary, = 20 m. A leaky aquifer boundary condition
is applied along the top boundary of the model, = 1 m, using the zone face
apply leakage command. The pore pressure at the top is = 1.8 kPa, and the
leakage coefficient (see this equation), is evaluated to be = 2.98 × 10-9 m3
/(N sec), based on the properties of the leaky layer. The properties for this
FLAC3D 6.0
Semi-confined Aquifer 647
problem are listed in the setup function. The fluid-flow calculation mode is
turned on, the mechanical calculation mode is turned off, and the simulation is
run until steady-state flow is reached.
The FISH function checkit compares the amount of leakage calculated by FLAC3D
to the solution of (7) at steady-state flow. The difference is printed (in a FISH
dialog message) to be 0.03%. The analytical and numerical pore pressure profiles
recorded along the base of the model, from = 0 to = 20, are compared in
Figure 2.
Figure 2: Pore pressure profile.
Reference
Strack, O. D. L. Groundwater Mechanics. New Jersey: Prentice Hall (1989).
pagebreak
FLAC3D 6.0
Data File
SemiConfinedAquifer.f3dat
; file for 'Semiconfined aquifer'

model new
model title 'Semiconfined aquifer'
call 'parameters'
@setup
point 0 (0,0,0) point 1 (@c_l,0,0) ...
point 2 (0,1,0) point 3 (0,0,@c_h) ...
size 20 2 2
zone fluid property porosity=@c_n permeability=@c_k
zone gridpoint initialize fluid-modulus=@c_wb
zone face apply leakage @c_ps @c_cs range position-z @zm @zp ;c_h
zone gridpoint fix pore-pressure @c_p0 range position-x 0
zone gridpoint fix pore-pressure @c_p1 range position-x @c_l
; --- fluid flow solution ---
model solve fluid time-total 1
return
FLAC3D 6.0
Verification of Concepts, and Modeling Techniques for Specific Applications 649
Verification of Concepts, and Modeling

Techniques for Specific Applications
Solid Weight, Buoyancy and Seepage Forces

When fluid flows through a porous medium there are, following Terzaghi (1943)
and Taylor (1948), three forces acting per unit volume on the solid matrix: the
solid weight, the buoyancy, and the drag or seepage force (also see Bear 1972).
These forces are automatically taken into account in the FLAC3D formulation.
This may be shown as follows.
In FLAC3D, equilibrium is expressed using total stress:
(1)
where is the undrained (saturated bulk) density, and is the gravitational

vector. (Note that Einstein notation convention for summation over repeated
indices applies to the equations in this section.) Undrained density may be
expressed in terms of drained density, , and fluid density, , using the
expression
(2)
where is porosity, and is saturation. The definition of effective stress is
(3)
Substitution of Equation (2) and Equation (3) in Equation (1) gives, after some
manipulations,
(4)
where we have introduced fluid unit weight, , and piezometric head, , as
(5)
(6)
FLAC3D 6.0
and is the gravitational magnitude.
In Equation (4), the term can be associated with solid weight, with
buoyancy, and with seepage force (drag).
A Simple Example Illustrating Solid Weight, Buoyancy and Seepage

Forces
A simple example is given below to illustrate the contribution of these individual
terms in the context of FLAC3D methodology. We consider a layer of soil of large
lateral extent and thickness, = 10 m, resting on a rigid base. The layer is
elastic, the drained bulk modulus, , is 100 MPa, and the shear modulus, , is 30
MPa. The density of the dry soil, , is 500 kg/m3. The porosity, , is uniform
with a value of 0.5. The mobility coefficient, , is 10-8 m2/(Pa-sec). The fluid
bulk modulus, , is 2 GPa, and gravity is set to 10 m/sec2.
Initially, the water table is at the bottom of the layer, and the layer is in
equilibrium under gravity. We study the heave of the layer when the water level
is raised, and also the heave or settlement under a vertical head gradient.
The problem is one dimensional; the FLAC3D model is a mesh composed of a

single column of 10 zones in the -direction. The axes origin is at the bottom of
the model. The mechanical boundary conditions correspond to roller boundaries
at the base and sides of the model. The fluid-flow boundary conditions are
described for the individual cases below.
This example is run using the groundwater configuration ( model configure

fluid). The coupled groundwater-mechanical calculations are performed using
the basic fluid-flow scheme. The calculation times are quite short for this small
model; for larger models, uncoupled modeling can be applied to speed the
calculation.
For reference, in a comparison of FLAC3D results to the analytical solutions, the

one-dimensional incremental stress-strain relation for this problem condition is
(7)
where is the Biot coefficient (set equal to 1 for this simulation), is the
drained bulk modulus, is the shear modulus, and is the vertical strain.
FLAC3D 6.0
Solid Weight — We first consider equilibrium of the dry layer. The dry density of
the material is assigned, and the saturation is initialized to zero (the default
value for saturation is 1 in model configure fluid mode). The flow calculation is
turned off, and the mechanical calculation is on. The value of fluid bulk modulus
is set to zero to prevent any generation of pore pressure under volumetric
straining for this stage. The model is cycled to equilibrium. By integration of
Equation (1) applied to the dry medium, we obtain
(8)
Vertical stress at the end of the FLAC3D simulation is plotted versus elevation in
Figure 1. The values match those obtained for equilibrium under gravity of the
dry medium (Equation (8)), as expected.
Figure 1: Vertical stress versus elevation — dry layer.
The vertical displacement at the top of the model is found from the equation.
(9)
FLAC3D 6.0
The calculated value from FLAC3D matches the analytical value at this stage
(-1.786 × 10-3 m). Note that the equilibrium ratio limit (model solve mechanical
ratio) is reduced to 10-6 to provide this level of accuracy for this example.
Buoyancy — We continue this example by raising the water table to the top of the
model. We reset the displacements to zero, and assign the fluid properties listed
above. The pore pressure is fixed at zero at the top of the model, and the
saturation is initialized to 1 throughout the grid. (Note that a fluid-flow
calculation to steady state is faster if the state starts from an initial saturation 1
instead of a zero saturation.) Fluid-flow and mechanical modes are both on for
this calculation stage, and a coupled calculation is performed to reach steady
state. The code uses the saturated density for this calculation, as determined
(internally) from Equation (2). By integration of Equation (1) for the saturated
medium, we obtain
(10)
The comparison of total vertical stress profile from FLAC3D to that from Equation
(10) is shown in Figure 2:
Figure 2: Vertical stress versus elevation — saturated layer.
FLAC3D 6.0
At steady state, the pore pressure is hydrostatic:
(11)
Contours of pore pressure at steady state are shown in Figure 3.
Figure 3: Pore pressure contours at steady state — saturated layer.
The vertical displacement induced by raising the water table is now upwards. The
amount of heave is calculated starting from Equation (7). We write this equation
in the form
(12)
For this example, = 0. After substitution of Equation (8), Equation (10) and
Equation (11) into Equation (12), we obtain, after some manipulation,
(13)
The term is the buoyant density. Substitution of Equation (2) for

undrained density in Equation (13) produces
FLAC3D 6.0
(14)
Finally, after integration between 0 and H, we obtain
(15)
and, for , this gives
(16)
The final numerical displacement at the top of the model compares well with the
analytical value (+1.786 × 10-3 m). The displacements, induced upward, are
plotted in Figure 4:
Figure 4: Heave of the layer at steady state — saturated layer.
Additional Rise in Water Table — We continue from this stage, and model the effect
of an additional rise in the water level on the layer. This time, the water table is
raised to 20 m above the top of the model. The corresponding hydrostatic
pressure is , where is 20 m and = 0.2 MPa.
FLAC3D 6.0
We reset displacements to zero and apply a pressure of 0.2 MPa at the top of the
model. A fluid pore pressure is applied (with zone gridpoint fix pore-pressure
2e5), as is a mechanical pressure (with zone face apply stress-normal), along the
top boundary. We now perform the coupled calculation again for an additional
500 seconds of fluid-flow time. No further movement of the model is calculated.
This is because the absolute increase in is balanced by the increase in pore
pressure, and the Biot coefficient is set to 1. Thus, no displacement is produced.
At the end of this stage, the hydrostatic pore pressure is given by
(17)
where is the fluid pressure at the base of the layer, and is the pressure at
the top. For this case, = 0.3 MPa, and = 0.2 MPa.
Seepage Force (Upwards Flow) — We now study the scenario in which the base of
the layer is in contact with a high-permeability over-pressured aquifer. The
pressure in the aquifer is 0.5 MPa. We continue from the previous stage, reset
displacements to zero, and apply a pore pressure of 0.5 MPa at the base ( zone
gridpoint fix pore-pressure 5e5). The coupled mechanical-flow calculation is
performed until steady state is reached. The plot of displacement vectors at this
stage, shown in Figure 5, indicates heave as a result of the upwards flow.
FLAC3D 6.0
Figure 5: Heave of the layer at steady state — seepage force from over-
pressured aquifer.
The analytical solution for the heave can be calculated from Equation (7). There
is no change in total stress, and so the term drops out. Also, the Biot
coefficient is equal to unity. Thus we can write
(18)
where is given by Equation (17), and is the steady-state pore pressure

distribution at the end of this stage. This pore pressure is given by
(19)
where = 0.5 MPa. Substitution of Equation (19) and Equation (17) into
Equation (18), and further integration produces
(20)
For = , we obtain
FLAC3D 6.0
(21)
The FLAC3D result for surface heave compares directly to this result ( = 7.143 ×
10-3 m).
Seepage Force (Downwards Flow) — The seepage force case is repeated for the
scenario in which the base of the layer is in contact with a high-permeability
under-pressured aquifer. This time, a pressure value of = 0.1 MPa is specified
at the base. The displacements are reset, and the coupled calculation is made.
The layer settles in this case, which can be seen from the displacement vector
plot in Figure 6. The analytical value for the displacement may be derived from
Equation (21) after replacing for . The FLAC3D settlement compares well
with the analytical settlement of = -7.143 × 10-3 m.
Figure 6: Settlement of the layer at steady state — seepage force from under-
pressured aquifer.
The corresponding project file, “WeightBuoyancySeepage.f3prj,” is located in the

“datafiles\fluid\WeightBuoyancySeepage” folder. Below is the datafile.
FLAC3D 6.0
WeightBuoyancySeepage.f3dat:
;; Solid weight, buoyancy and seepage forces

model new
model title 'Buoyancy - fully coupled'
@setup
zone property bulk @m_bu shear @m_sh
zone initialize density @m_d
; --- (column is dry) ---
zone gridpoint initialize saturation 0
; --- gravity ---
model gravity 0 0 @mgrav
; --- histories ---
fish history @num_dis
fish history @ana_dis1
; --- initial equilibrium
model fluid mechanical on
model solve
list @num_dis @ana_dis1
model save 'ini'
; --- Fully coupled ---
model restore 'ini'
model title 'Buoyancy - fully coupled'
history delete
; --- histories ---
; --- add water ---
zone initialize fluid-density=@w_d
zone gridpoint initialize fluid-modulus=2e8
zone gridpoint initialize fluid-tension=-1e10
zone fluid property porosity=@m_n permeability 1e-8
; --- static equilibrium ---
FLAC3D 6.0

; --- we can run this simulation coupled, using ---
model save 'ex1a'
; --- Uncoupled ---
model restore 'ini'
model title 'Buoyancy - uncoupled'
history delete
; --- add water ---
zone initialize fluid-density=@w_d
zone gridpoint initialize fluid-modulus=2e8
zone gridpoint initialize fluid-tension=-1e10
zone fluid property porosity=@m_n permeability 1e-8
; --- we can run this simulation uncoupled, using ---
; --- histories ---

model save 'ex1b'
; --- Sink it deeper, under 20m of water ---
model restore 'ex1b'
model title 'Sink it deeper, under 20m of water'
history delete
; --- fluid boundary conditions ---
zone gridpoint fix pore-pressure 2e5 range position-z 10
; --- apply pressure of water ---
zone face apply stress-normal -2e5 range position-z 10
; --- histories ---
FLAC3D 6.0

model save 'ex1c'
; --- Seepage force: flush up ---
model restore 'ex1c'
model title 'Seepage force: flush up'
history delete
; --- flush fluid up ---
; --- histories ---
model save 'ex1e'
; --- Seepage force: flush down ---
model restore 'ex1c'
model title 'Seepage force: flush down'
history delete
; --- flush fluid up ---
; --- histories ---
model save 'ex1f'
return
FLAC3D 6.0
Pore Pressure Initialization and Deformation

In FLAC3D, when equilibrium stresses are initialized in the model with the zone
initialize command (following, for example, the procedure as described in
Stresses with Gradients — Uniform Material and mechanical steps are taken, no
stress increment is calculated by the code, and thus no displacement is
generated, because the model is in equilibrium with the stress boundary
conditions and applied loads. In other words, FLAC3D does not calculate the
deformations associated with installation of equilibrium stresses, when using the
zone initialize stress command. If the stress state is installed in this way at
the beginning of a run, the initial stress state is taken as the reference state for
displacements. (This is worth noting because the logic may be different from
other codes, in which the zero stress state is taken as reference for calculation of
displacements.)
The situation regarding pore pressure in a fluid-mechanical simulation is

similar: by default, if pore pressures are initialized with the zone gridpoint
initialize pore-pressure command, are in equilibrium with the fluid boundary
conditions and hydraulic loading, and fluid steps are taken, then by default no
increment of pore pressure will be generated by the code. If the zone gridpoint
initialize pore-pressure command is issued at the beginning of the run, this
initial state is taken as the reference state for pore pressure. There will be no
stress change (and, if mechanical steps are taken, no displacement) as a result of
the pore pressure initialization, because no increment of pore pressure is
calculated by the code. In other words, by default, FLAC3D does not calculate the
deformation associated with installation of equilibrium pore pressures when
using the zone gridpoint initialize pore-pressure command.
This only applies, by default, to equilibrium pore pressures established using the
zone gridpoint initialize pore-pressure command, the zone water command or
a FISH function to initialize pore pressures. Pore pressure change that is
calculated by FLAC3D on the other hand, will always generate stress change; if the
system is brought out of equilibrium by the stress change and mechanical steps
are taken, then deformations will be generated, if conditions allow.
If the deformation associated with a new distribution of pore pressure, assigned

using the zone gridpoint initialize pore-pressure command, the zone water
command or a FISH function to initialize pore pressures, needs to be calculated,
then a special stress-correction technique needs to be used. This technique
consists of 1) subtracting from the total normal stresses, the increment of pore
FLAC3D 6.0
pressure in the zones affected by the change, multiplied by Biot coefficient (the
pore pressure increment in a zone is calculated by averaging nodal values), 2)
adjusting the saturation to zero in the dry region, and to one in the region filled
with fluid, 3) adjusting the input material density to the bulk value, above and
below the phreatic surface (in case it exists), and the simulation is run using
model configure fluid, and 4) cycling the model to mechanical equilibrium. A
simple example illustrating the heave of a soil layer using this technique is
presented below.
Heave of a Soil Layer

For the example, we consider a layer of soil of large lateral extent, and thickness
= 10 meters, resting on a rigid base. The layer is elastic, the drained bulk
modulus, , is 100 MPa, and the shear modulus, , is 30 MPa. The bulk density
of the dry soil, , is 1800 kg/m3, and the density of water, , is 1000 kg/m3. The
porosity, , is uniform; the value is 0.5. And gravity is set to 10 m/sec 2.
Initially, the water table is at the bottom of the layer, and the layer is in
equilibrium under gravity. We evaluate the heave of the layer when the water
level is raised to the soil surface. This simple problem is similar to the one
analyzed in teh section on Solid Weight, Buoyancy and Seepage Forces. The
difference is, here, we do not require the code to find the new pore-pressure
distribution. The simulation can be carried out with or without using the
groundwater configuration (model configure fluid). We consider both cases.
The grid for this example contains 20 zones: 10 in the -direction, 2 in the
-direction, and 1 in the -direction. The origin of axes is at the bottom of the
model. The mechanical boundary conditions correspond to roller boundaries at
the base and lateral sides of the model.
We first consider equilibrium of the dry layer. We initialize the stresses, using
the zone initialize stress-xx, zone initialize stress-yy and zone initialize
stress-zz commands, using a value of 0.5714 (equal to ( )/( ))
for the coefficient of earth pressure at rest, .
There are two competing effects on deformation associated with raising the
water level: first, the increase of pore pressure will generate heave of the layer;
and second, the increase in soil bulk density due to the presence of the water in
the pores will induce settlement. To model the combined effects on deformation,
of a rise in water level up to the soil surface, we proceed as follows.
FLAC3D 6.0
If we do not use model configure fluid, we specify a hydrostatic pore-pressure

distribution corresponding to the new water level by using either the zone
gridpoint initialize pore-pressure command or the zone water command, and
we specify a wet bulk density for the soil beneath the new water level. We apply
the stress correction described under point 1) above, using the commands
zone initialize stress-xx -1e5 add grad 0 0 1e4

zone initialize stress-yy -1e5 add grad 0 0 1e4
zone initialize stress-zz -1e5 add grad 0 0 1e4
Finally, we cycle the model to static equilibrium.
If we do use model configure fluid, we again specify a hydrostatic pore-pressure

distribution corresponding to the new water level using the zone gridpoint
initialize pore-pressure or zone water command. The saturation is initialized to
1 below the water level, and to zero above. However, we do not update the soil
density to account for the presence of water beneath the new water level (this is
done automatically by the code). We again perform a stress correction, as
described earlier. Finally, we model fluid active off and cycle the model to static
equilibrium.
The final response is identical for both cases. The plot of displacement vectors in
Figure 7 indicates that the rise of the water table has induced a heave of the soil
layer. The surface heave, , can be evaluated analytically using Equation (22),
(22)
where . The theoretical value for is 1.786 × 10-3 meters. The

theoretical and numerical values compare well. The data files for the simulations
are listed in SoilLayerHeave-NoFluid.f3dat and SoilLayerHeave-withFluid.f3dat.
FLAC3D 6.0
Figure 7: Heave of a soil layer.
SoilLayerHeave-NoFluid.f3dat:
model new
model title 'Raising the water table - not in configure fluid'
@setup
point 2 (0,1,0) point 3 (0,0,10) ...
size 2 1 10 ratio 1 1 1
zone property bulk @m_bu sh @m_sh
; --- column is dry ---
zone property density @m_d
; --- gravity ---
; --- histories ---
history interval 10
fish history @ana_dis
; --- initial equilibrium ---
FLAC3D 6.0
zone initialize stress-zz -1.8e5 grad 0 0 1.8e4

zone initialize stress-xx -1.029e5 grad 0 0 1.029e4
zone initialize stress-yy -1.029e5 grad 0 0 1.029e4
model save 'ini'
; --- raise water level ---
; (note: when not in configure FLUID, water density is assigned using:)
zone water density @w_d
; (we can do it this way ...)
zone gridpoint initialize pore-pressure 1e5 grad 0 0 -1e4
; (or this way ...)
;water table origin 0 0 _H normal 0 0 -1
; (total stress adjustement)
; --- use wet density below water table ---
zone property density @m_rho
model save 'ncfl'
return
SoilLayerHeave-withFluid.f3dat:
model new
model title 'Raising the water table - configure fluid'
@setup
point 2 (0,1,0) point 3 (0,0,10) ...
size 2 1 10 ratio 1 1 1
zone property bulk @m_bu sh @m_sh
; --- column is dry ---
; (initialize sat at 0)
zone property density @m_d
; --- gravity ---
; --- histories ---
history interval 10
fish history @ana_dis
; --- initial equilibrium ---
FLAC3D 6.0
zone initialize stress-zz -1.8e5 grad 0 0 1.8e4

zone initialize stress-xx -1.029e5 grad 0 0 1.029e4
zone initialize stress-yy -1.029e5 grad 0 0 1.029e4
;
model solve
model save 'ini'
; --- raise water level ---
; (initialize sat at 1 below the water level)
; (note: in configure FLUID, water density is assigned using:)
zone initialize fluid-density @w_d
; (we can do it this way ...)
zone gridpoint initialize pore-pressure 1e5 grad 0 0 -1e4
; (or this way ...)
; (note: we can use water table command in configure fluid,
; this is different from FLAC)
;water table origin 0 0 _H normal 0 0 -1
; (total stress adjustement)
; --- Note: no need to specify wet density below water table ---
model solve
model save 'cfl'
return
Effect of the Biot Coefficient

The Biot coefficient, , relates the compressibility of the grains to that of the
drained bulk material:
(23)
where is the drained bulk modulus of the matrix, and is the bulk modulus
of the grains (see Detournay and Cheng 1993, for reference).
For soils, matrix compliance is usually much higher than grain compliance (i.e.,
), and it is a valid approximation to assume that the Biot
coefficient is equal to 1.
FLAC3D 6.0
For porous rocks, however, matrix and rock compliances are most often of the
same order of magnitude and, as a result, the Biot coefficient may be almost
zero. Consider, for example, a sample of porous elastic rock. The pores are
saturated with fluid at a pressure, , and a total external pressure, , is applied
around the periphery (i.e., on the outside of an impermeable sleeve). The
problem can be analyzed by superposition of two stress states: state a, in which
fluid pressure and external pressure are both equal to ; and state b, in which
pore pressure is zero, and the external pressure is (see Figure 8).
Figure 8: Decomposition of stresses acting on a porous, elastic rock.
The stress-strain relation for state a may be expressed as
(24)
For state b (there is no fluid), we can write
(25)
The total strain is given by superposition of the strain in state a and in state b:
(26)
After substitution of from Equation (24), and from Equation (26), we obtain
(27)
After some manipulations, the stress-strain equation takes the form
(28)
FLAC3D 6.0
Clearly, then, in the framework of Biot theory, a zero Biot coefficient implies that
the elastic stress-strain law becomes independent of pore pressure. Of course, in
general, porous rocks do not behave elastically, and pore pressure has an effect
on failure. Also, if fluid flow in rocks occurs mainly in fractures, Biot theory may
not be applicable. Nonetheless, there are numerous instances where the small
value of the Biot coefficient may help explain why pore pressure has little effect
on deformation for solid, porous (i.e., unfractured) rocks. (For example, the
effect on surface settlement of the raising or lowering of the water table in a
solid porous rock mass may be unnoticeable.)
Note that the preceding discussion addresses only one of the effects of grain
compressibility. The Biot coefficient also enters the fluid constitutive law, which
relates change of fluid content to volumetric strain.
The logic for grain compressibility, as developed in the framework of Biot theory,
is provided in FLAC3D. Simple verification examples are described below to
illustrate the logic.
For reference, in the examples below, in the special case = 0, the principal
stress-strain relations have the form`
(29)
(30)
(31)
where is the variation of fluid content per unit volume of porous media, and
is the incremental volumetric strain.
Undrained Oedometer Test

An undrained oedometer test is conducted on a saturated poro-elastic sample.
The Biot modulus is 1 MPa, and the Biot coefficient is 0.3 for this test. The model
is a single square zone of unit dimensions with roller boundaries on the sides
and bottom. A constant velocity, , is applied to the top. Fluid flow is turned off,
and the simulation is run for 10 calculational steps. The zone fluid biot on
command is given to select the Biot modulus and coefficient rather than the fluid
bulk modulus.
Because the sample is laterally confined, = = 0 and . For

undrained conditions, = 0. The analytical value for pore pressure (see
Equation (31)) is then
FLAC3D 6.0
(32)
The analytical stresses are obtained by substituting pore pressure and strain
components into Equation (29) and Equation (30):
(33)
(34)
After 10 calculation steps, . The agreement between analytical and

numerical values for pore pressure and stresses is checked with the FISH function
checkit. The analytical and numerical results are identical. The corresponding
project file, “UndrainedOedometer.f3prj”, is located in the “datafiles\Fluid\
UndrainedOedometer” folder. Below is the datafile:
UndrainedOedometer.f3dat:
; Undrained Oedometer Test

model new
model title 'Undrained Oedometer test'
point 2 (0,1,0) point 3 (0,0,1) size 1 1 1
zone property bulk 1e8 shear 0.3e8
zone fluid biot on
zone fluid property porosity=0.5 biot=0.3
zone gridpoint initialize velocity-z=-1e-3 range position-z 1
; --- undrained compression ---
zone gridpoint initialize biot=1e8
model step 10
return
Pore Pressure Generation in a Confined Sample

The effect of pore pressure generation is shown for the case of a confined sample
in an impermeable sleeve. The sample geometry and properties are the same as
in the previous example, Undrained Oedometer Test. Roller boundaries are set on
all four sides of the model. The boundaries are also impermeable (by default).
FLAC3D 6.0
Fluid flow is turned on, and a volumetric water source with a unit flow rate is
applied to the model to raise the pore pressure. The simulation is run for 10 fluid
flow steps.
At the end of the simulation, . The grid is fully constrained, hence

. The analytical value for pore pressure is found,
from Equation (31), to be
(35)
The analytical stresses are then derived from Equation (29) and Equation (30) to
be
(36)
Numerical and analytical values for pore pressure and stresses are compared
with the FISH function checkit, and the results are identical. The corresponding
project file, “PorePressureGenerationConfined.f3prj” is located in the “datafiles\
Fluid\PorePressureGenerationConfined” folder. Below is the data file:
PorePressureGenerationconfined.f3dat:
; file for 'Pore pressure generation in a confined sample'

model new
model title 'Pore pressure generation in a confined sample'
point 2 (0,1,0) point 3 (0,0,1) size 1 1 1
zone property bulk 1e8 shear 0.3e8
zone fluid biot on
zone fluid property porosity=0.5 biot=0.3 permeability=1e-10
zone gridpoint initialize biot=1e8
; --- water source ---
zone apply well 1.0
model step 10
return
FLAC3D 6.0
Pore Pressure Generation in an Infinite Layer

This example is similar to the preceding confined sample, except that the top
boundary is not constrained. All boundaries are impermeable, and a volumetric
water source with unit flow is applied to raise the pore pressure. Both mechanical
and fluid-flow calculations are turned on, and the simulation is run for 10 fluid-
flow steps. Note that 200 mechanical sub-steps are taken every fluid step in
order to keep the system in quasi-static equilibrium state.
For this example, , and . Using these

conditions in Equation (30), we obtain
(37)
After substituting Equation (37) for in Equation (31) and solving for , we
find
(38)
Analytical expressions for pore pressure and stress can now be derived from
Equation (29) and Equation (31):
(39)
(40)
Numerical and analytical values for vertical displacement, pore pressure and
stresses are compared with the FISH function checkit, and the results are
identical. The corresponding project file, “PorePressureGenerationInfinite.f3prj”
is located in the “datafiles\Fluid\PorePressureGenerationInfinite” folder. Below
is the data file:
PorePressureGenerationinfinite.f3dat:
; file for 'Pore pressure generation in an infinite layer'

model new
model title 'Pore pressure generation in an infinite layer'
point 2 (0,1,0) point 3 (0,0,1) size 1 1 1
zone property bulk 2.0 shear 1.0
FLAC3D 6.0
zone fluid biot on

zone fluid property porosity=0.5 biot=0.3 permeability=1.0
zone gridpoint initialize biot=1
zone gridpoint fix velocity-x
; --- water source ---
zone apply well 1.0
model solve fluid time-total 1.6 or mechanical ratio 1e-10
;
model save 'ppgen-infinite'
;
return
FLAC3D 6.0
Input Instructions for Fluid-Flow Analysis 673
Input Instructions for Fluid-Flow Analysis
Fluid Commands
The following commands are provided to run fluid problems. There is no fluid
command. The commands used for fluid analysis take the form of fluid-related
keywords on the commands used for standard mechanical analysis. The
command model configure fluid must be the first fluid command given before
any other commands for fluid analysis are issued. model configure fluid can be
specified at any stage of a FLAC3D analysis.
Table 1: Fluid Commands
geometry generate from-faces fluid
model history fluid
model results interval fluid
model solve fluid
model fluid
zone apply well
zone apply-remove well
zone face apply discharge
zone face apply leakage
zone face apply pore-pressure
zone face apply-remove discharge
zone face apply-remove leakage
zone face apply table fluid
zone face group fluid
zone gridpoint fix pore-pressure
FLAC3D 6.0
zone gridpoint fix well
zone gridpoint free pore-pressure
zone gridpoint free well
zone gridpoint initialize pore-pressure
zone history property-fluid
zone list profile property-fluid
zone relax excavate table fluid
zone relax modify table fluid
zone results model-fluid
zone fluid
zone fluid zone-based-pp
zone water
Fluid FISH Variables

The following scalar variables are available in a FISH function to assist with fluid
analysis.
Table 2: Fluid FISH

Functions
gp.biot.modulus
gp.pp
gp.pp.fix
gp.flow
gp.fluid.modulus
gp.fluid.tension
gp.sat
zone.flux
zone.fluid.model
zone.fluid.prop
FLAC3D 6.0
zone.fluid.ratio
zone.fluid.stress.normal
zone.fluid.time.total
zone.fluid.timestep
Zone-Based Pore Pressure

A feature that allows you to set zone based pore pressure using FISH has been
added to FLAC3D. This allows for the rapid import of pore pressures from an
external source. The stress increment resulting from this zone-based pore
pressure change is also incorporated. This feature has been implemented strictly
for calculating mechanical changes to a model that result from changes in pore
pressure in a zone. Fluid flow should not to be active in FLAC3D when this zone-
based pore pressure feature is active.
This feature assumes that FLAC3D is configured for fluid calculations (model
configure fluid) and that fluid flow calculations are set to off (zone fluid active
off).
Zone-based pore pressures are stored in a “fluid-zone” which is enabled by the

model configure fluid command. The stress change generated by a change in
zone-based pore pressure is determined by the following equations:
(1)
where is the Biot coefficient, a material property of a fluid model (the default is
1.0).
The command to signal the use of zone-based pore pressure to FLAC3D is zone
fluid zone-based-pp on. When this command is used, the FISH function zone.pp
will allow you to retrieve and assign pore pressure to a zone. Note that when zone
fluid zone-based-pp on, FLAC3D keeps track of the last two pore pressures that
were assigned to this zone with the zone.pp FISH function. This allows the
computation of a pore pressure change ( ). Each time a pore
FLAC3D 6.0
pressure is assigned to a zone with the zone.pp FISH function, the previously
stored pore pressure becomes and the newly assigned pore pressure becomes
.
If zone fluid zone-based-pp is initially set to off, FLAC3D will use gridpoint-
based pore pressures in its calculations (this is FLAC3D’s default mode). If a zone
fluid zone-based-pp on command is encountered, it automatically initializes the
zone-based pore pressure (both and ) with the average gridpoint-based
pore pressures. This method also establishes the baseline pore pressures in the
zones. If zone fluid zone-based-pp was on and a zone fluid zone-based-pp off
command is issued, then the zone-based pore pressures are not used to initialize
the gridpoint-based pore pressures with the zone-based pore-pressures.
The commands zone gridpoint fix pore-pressure and zone gridpoint initialize
pore-pressure do not alter the zone-based pore pressures. They only work with
gridpoint-related quantities. Plotting “Zone -> Color by Contour -> Value: Pore
Pressure” shows only gridpoint-related quantities. Instead, the zone-based pore
pressure can be stored in zone.extra and then can be plotted through “Zone ->
Color by Contour -> Extra; Source Zone.”
The following is one modeling sequence demonstrating the use of this feature.
1. Create your model geometry (zones) — remembering model configure

fluid.
2. Assign mechanical model and properties to zones.
3. Assign fluid model and Biot coefficient (1 by default) to zones.
4. zone fluid active is set to off; zone mechanical active is set to on and
zone fluid zone-based-pp is set to on.
5. Assign a baseline pore pressure to each zone. Use the zone.pp function to
do this.
6. Assign updated pore pressure to each zone. Use the zone.pp function to do
this.
7. Solve to mechanical equilibrium.
FLAC3D 6.0
A simple example case is provided in “\datafiles\Fluid\ZoneBasedPorePressure”;

the main datafile is listed below.
ZoneBasedPorePressure.f3dat:
model new
;
zone create brick size 2 2 2 point 1 (1 0 0) point 2 (0 1 0) point 3 (0 0 1)
; mechanical model
zone property density 1.0 bulk 400 shear 300
; fluid flow model
zone fluid property biot 1.0
;
zone fluid zone-based-pp on
;
;
; read baseline pp from file basetp.dat and store in zones
@_baseline
; read updated pp and temp from file updatetp.dat and store in zones
@_update
;
model solve
; store zone-based pp to zone extra
@_pt2extra
;
model save 'zonebased'
return
FLAC3D 6.0
FLAC3D 6.0
References 679
References
Barenblatt, G. Dimensional Analysis. Gordon and Breach Science Publishers (1987).
Bear, J. Dynamics of Fluids in Porous Media. New York: Dover (1972).
Berchenko, I. Thermal Loading of a Saturated Rock Mass: Field Experiment and

Modeling Using Thermoporoelastic Singular Solutions. Ph.D. Thesis, University of
Minnesota (1998).
Biot, M. A. “General Solutions of the Equations of Elasticity and Consolidation for

a Porous Material,” J. Appl. Mech., Trans. ASME, 78, 91-96 (1956).
Burghinoli, A., S. Miliziano and F. M. Soccodato. “Effectiveness of the fast-flow

algorithm: 2D consolidation benchmark and tunneling application,” FLAC and
Numerical Modeling in Geomechanics — 2001 (Proceedings of the 2nd International
FLAC Symposium, Lyon, France, October 29 - 31, 2001), pp. 345 - 352. D. Billaux et
al., eds. Swets & Zeitlinger (2001).
Carslaw, H. S., and J. C. Jaeger. Conduction of Heat in Solids, Second Edition. Oxford
at the Calderon Press (1959).
Chaney, R. C. “Saturation Effects on the Cyclic Strength of Sands,” in Proceedings:

ASCE Geotechnical Engineering Division Specialty Conference (Pasadena, California,
June 19-21, 1978), Vol. 1, pp. 342-358 (1978).
Crank, J. The Mathematics of Diffusion, 2nd Ed. Oxford: Oxford University Press
(1975).
Cryer, C. W. “A Comparison of the Three-Dimensional Consolidation Theories of

Biot and Terzaghi,” Quart. J. Mech. and Appl. Math., XVI, 4, 401-412 (1963).
Dahlquist, G., and A. Bjorck. Numerical Methods. Prentice Hall (1974).
Detournay, E., and A. H. D. Cheng. “Fundamentals of Poroelasticity,” in

Comprehensive Rock Engineering, Vol. 2, pp. 113-171. J. Hudson et al., eds. London:
Pergamon Press (1993).
Detournay, E., and A. H. D. Cheng. “Poroelastic Response of a Borehole in a Non-

Hydrostatic Stress Field,” Int. J. Rock Mech. Sci. & Geomech. Abstr., 25(3), 171-182
(1988).
FLAC3D 6.0
Harr, M. E. Groundwater and Seepage. Dover (1991).
Karlekar, B. V., and R. M. Desmond. Heat Transfer, 2nd Ed. St. Paul: West
Publishing Co. (1982).
Kochina, I., N. Mikhailov and M. Filinov. “Groundwater Mound Damping,” Int. J.

Engng. Sci., 21, 413-421 (1983).
McNamee, J., and R. E. Gibson. “Plane strain and axially symmetric problems of
the consolidation of a semi-infinite clay stratum,” Quart. J. Mech. and Appl. Math.
XIII, Pt. 2, (1960).
Polubarinova-Kochina, P. Y. Theory of Groundwater Movement. Princeton:

Princeton University Press (1962).
Schiffman, R. L., A. Chen and J. C. Jordan. “An Analysis of Consolidation

Theories,” J. Soil Mech. and Found. Div., ASCE, 95(SM1), 285-312 (1969).
Stehfest, H. “Numerical Inversion of Laplace Transforms,” Communic. Ass.

Comput. Mach., 13, 47-49 (1970).
Strack, O. D. L. Groundwater Mechanics. New Jersey: Prentice Hall (1989).
Taylor, D. W. Fundamentals of Soil Mechanics. New York: John Wiley (1948).
Terzaghi, K. Theoretical Soil Mechanics. New York: John Wiley (1943).
Theis, C. V. “The Relation between the Lowering of the Piezometric Surface and
the Rate and Duration of Discharge of a Well Using Groundwater Storage,” Trans.
Am. Geophys. Union, 10, 519-524 (1935).
Voller, V., S. Peng and Y. Chen. “Numerical Solution of Transient, Free Surface
Problems in Porous Media,” Int. J. Numer. Meth. Engng., pp. 2889-2906 (1996).
FLAC3D 6.0

Flac3d 600 Theory

Uploaded by

Copyright:

Available Formats

Flac3d 600 Theory

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Flac3d 600 Theory

Uploaded by

Copyright:

Available Formats

What is the purpose of this document?

What is the purpose of this document?

What advice is given for printing this document?

What advice is given for printing this document?

FLAC3D 6.

© 2017 Itasca Consulting Group, Inc.

FLAC3D Theory and Background: Table of

THEORETICAL BACKGROUND ………………………………………………………………………… 3

Interfaces and Changing Interacting Objects in Small Strain ………………… 83

FACTOR OF SAFETY ……………………………………………………………………………………… 87

CONSTITUTIVE MODELS ……………………………………………………………………………… 155

Undrained Triaxial Test with CYSoil Model ………………………………………… 433

FLUID-MECHANCIAL INTERACTION …………………………………………………………… 481

One-Dimensional Consolidation ………………………………………………………… 593

FLAC3D Theory and Background

The theoretical formulation for FLAC3D is described in the first section,

A Factor of Safety calculation, based upon the “strength reduction method,” is

The fourth section provides a description of Constitutive Models in FLAC3D. The

The last section describes the implementation of Fluid-Mechanical Interaction in

▪ Nodal mixed discretization on strain

Formulation of a 3D Explicit Finite Difference

Mathematical Model Description

An important aspect of the model is the inclusion of the equations of motion,

As a notation convention, a bold letter designates a vector or tensor, depending

By definition, tension and extension are positive.

Rate of Strain and Rate of Rotation

Equations of Motion and Equilibrium

Boundary and Initial Conditions

in which is the corotational stress-rate tensor, [H] is a given function, and κ

1. finite difference approach (First-order space and time derivatives

2. discrete-model approach (The continuous medium is replaced by

3. dynamic-solution approach (The inertial terms in the equations

Finite Difference Approximation to Space Derivatives

By application of the Gauss divergence theorem to the tetrahedron, we can write

For a constant strain-rate tetrahedron, the velocity field is linear, and [ n] is

where the superscript l relates to the value at node l.

Substitution of Equation (11) in (10) yields, reorganizing terms by node

If we replace vi with 1 in Equation (9) we obtain, by application of the divergence

Using this last relation, and dividing Equation (12) by V, we get

and the components of the strain-rate tensor may be expressed as

Nodal Formulation of the Equations of Motion

Fixing time, t, we consider an equivalent static problem governed at any instant

with body forces defined as (Equation (5))

In the framework of the finite difference approximation adopted here, the

Following the notation conventions (a superscript refers to the nodal value of a

while the internal work rate is given by

Using Equation (15), we can write, for a constant strain-rate tetrahedron,

The stress tensor is symmetric and, defining a vector T l with components

while E I may be expressed as

where N n (with n = 1,4) are linear functions of the form

From Equation (29) and (30), we may write

Also, substitution of Equation (26) for δvi in Equation (25) gives

For static equilibrium of the tetrahedron in the framework of the equivalent

and Equation (34) may be written as

and the out-of-balance force [F] <l> is given by

Explicit Finite Difference Approximation to Time Derivatives

Nodal displacements are calculated in the code from the relation

Constitutive Equations in Incremental Form

where is the corotational stress increment, and Hij* is a given function.