NONLINEAR ANALYSIS OF E STRUCTURES

A nonlinear analysis is one that considers the stress-strain behavior of materials and
geometric nonlinearity, directly discarding the principle of superposition. In concrete
structures, the linear-elastic behavior differs greatly from that observed experimentally,
especially when they are under high load levels. This is due to the intrinsic phenomena
of the materials that make up these structures (eg cracking of concrete, plasticization of
steel, etc.), their interaction and construction process. Due to the invalidity of the
superposition principle, the security formats applied in linear analyzes cannot be
directly applied in nonlinear ones. This has encouraged the development of safety
formats to be applied both in the analysis and in the non-linear design, also promoting
the study of the parameters that most affect the real ultimate response of the structure.
The main objective of this thesis is to develop a tool for the analysis of the reliability of
structures in a non-linear regime, which has the ability to adapt to any commercial
analysis model that allows working via the command line. This will be done with the
combination of an analysis model that allows taking into account both the non-linearity
due to the material and the geometry, plus a simplified reliability method, which focuses
on finding the reliability index through an approximation first order of the given limit
state equation. Another objective is to study the sensitivity and safety in the ultimate
response of concrete structures that have been designed to work in a non-linear range,
identifying the most relevant variables for the response, this through the approach of
several case studies.

.
Motivation and Numerical Methods The analysis of a structure consists of establishing
certain mechanical and physical hypotheses about its behavior, obtaining and solving
the corresponding mathematical model, and interpreting the results. If the established
hypotheses are such that the equations of the mathematical model are nonlinear, the
analysis is considered nonlinear. The validity of the results of a structural analysis
depends on the approximations of reality given by the due hypotheses and the precision
of the resolution of the corresponding mathematical model. Given this framework, a
series of real structures or structural problems that require the use of nonlinear models
are listed as motivation. Subsequently, some numerical methods are introduced that
allow, using computational tools, to carry out the nonlinear analysis of a structure.
Motivation and approach Motivation Nonlinear analysis of structures is an essential part
of the knowledge of structural engineers in disciplines such as: Civil (infrastructure),
Mechanical, Naval, Aerospace, Automotive, Biomechanics, etc. In some applications in
the Civil area, particularly in simple building structures, the effects resulting from the
non-linear behavior of the structures are contemplated by procedures codified in
structural design standards (see for example ACI 318-14, AISC 360-16 , Euro codes).
However, for structures that deviate from the assumptions assumed in the codified
procedures, the aforementioned standards describe and allow more detailed analyzes of
the effects resulting from the non-linear response of the structures. The need to carry out
these analyzes becomes essential for certain types of structures that work in a non-linear
regime under service loads. Some examples of these structures are: cable-stayed masts
(Sparling, 1995), suspended bridges (Larsen et al., 2000),
cable-stayed bridges (Wu et al., 2015; Madrazo-Aguirre et al., 2015), analysis of thin
plates or shells (Hunt and Wadee, 1998), slender and framed reticulated structures
(Morozov et al., 2011), covered with taut membranes (Bridgens and Birchall, 2012),
covers formed by cables (Feng et al., 2013).
Numerical Methods for Nonlinear Equations Determining the equilibrium
configurations of a structure with nonlinear behavior requires the ability to solve
systems of nonlinear equations. Therefore, nonlinear structure analysis procedures are
directly based on numerical methods for solving systems of nonlinear equations. In
several cases, practical applications of structure resolution have been the precursors of
what would later be numerical methods applied to various branches of Engineering,
such as the Finite Element Method (Zienkiewicz, 1972) (see (Criseld, 1996; Bathe ,
2014) for more detail). In order to avoid the simultaneous assimilation of numerical
procedures and structural concepts, it is decided to begin by presenting the most
important numerical methods in isolation in order to, in subsequent chapters, apply
them in the context of structural problems. The study and development of these
numerical methods is part of the area called Numerical Continuation, belonging to the
Applied Mathematics discipline (Doedel, 2014). The following sections present the
three main classes of numerical methods that are useful for solving systems of nonlinear
equations, described below: Incremental Methods, Iterative Methods, Arc Length
Methods.
Currently there are several general programs based on implementations of these
methods. Likewise, these programs can be applied to the study of structural stability
problems.
If the displacements are analyzed, the following considerations can be made:
Case 1.-The specific deformations and displacements are small. This is the case of the
analysis of linear structures where the displacements are small and the equilibrium is
analyzed without taking them into account.
Case 2.- The specific deformations are not small and the displacements are small. This
is the case of analysis of structures in an inelastic regime (Plastic calculation) where in
certain areas of the structure very important deformations are reached that translate into
the formation of plastic joints. , despite which the displacements of the structure remain
small and the equilibrium can continue to be analyzed without taking them into account,
this is, a physical nonlinearity.
Case 3.- The specific deformations are small and the displacements are not small, in this
case of nonlinear behavior of the structure due to geometric nonlinearity.
NON-LINEAR STATIC ANALYSIS.
Nonlinear static analysis assumes that the relationships between the loads and the
induced response are linear. For example, if you double the magnitude of the loads, the
response (displacements, strains, stresses, reaction forces, etc.) also doubles.

All real structures behave nonlinearly in one way or another at some level of loading. In
some cases, linear analysis may be appropriate. In many others, the linear solution can
produce erroneous results because the assumptions on which it is based are violated.
Nonlinearity can be caused by material behavior, large displacements, and contact
conditions.

Use a nonlinear study to solve a linear problem. The results may be slightly different
due to different procedures.

In nonlinear static analysis, dynamic effects such as inertial and damping forces are not
considered. Processing a nonlinear study differs from processing a static study in the
following ways:

WHEN TO USE NONLINEAR ANALYSIS

Linear analysis is based on static and linearity superpositions and is therefore valid as
long as those assumptions are valid. When one (or more) of these assumptions fails,
linear analysis produces incorrect predictions and nonlinear analysis must be used to
model the nonlinearities.

The linearity assumption is true if:

All materials in the model comply with Hook's Law , which says that stress is directly
proportional to strain. Some materials exhibit this behavior only if the unit strains are
small. As strains increase, stress-strain relationships become nonlinear. Other materials
show nonlinear behavior even when strains are small. A material model is a
mathematical simulation of the behavior of a material. A material is said to be linear if
its stress-strain relationships are linear. Linear analysis can be used to analyze models
with linear materials assuming there are no other types of nonlinearities. Linear
materials can be isotropic, orthotropic or anisotropic. Whenever a model material
exhibits nonlinear stress-strain behavior under the specified load, nonlinear analysis
should be used. Nonlinear analysis offers many types of material models.

The induced displacements are small enough to ignore the change in stiffness caused by
loading. Nonlinear analysis offers a great deformation option when defining the material
properties of a solid component or shell. Matrix stiffness calculations can be repeated at
each solution step. The user controls the frequency of repetition of the matrix stiffness
calculation.

The boundary conditions do not vary during the application of the loads. The loads must
be constant in magnitude, direction and distribution. They should not change while the
model is deformed. For example, contact problems are nonlinear in nature because
boundary conditions change upon contact with the load. However, linear analysis offers
an approximate solution for contact problems where the large deformation effect is
considered.

TYPES OF NON-LINEARITIES.
Nonlinear material.
Geometric nonlinearity due to large displacements and small deformations.
Geometric nonlinearity due to large displacements and deformations.
Contour nonlinearity.
Consideration of the construction process
NON-STRUCTURAL LINEARITIES.
The main sources of structural nonlinearities encountered in practical applications
are the following:

GEOMETRIC NON-LINEARITIES

In nonlinear finite element analysis, a major source of nonlinearities is due to the

effect of large displacements on the overall geometric configuration of structures.
Structures that are being subjected to large displacements can have significant
changes in their geometry due to load-induced deformations that can cause the
structure to respond nonlinearly in a sharpening and/or softening mode.

For example, cable-type structures. They usually present a stiffening behavior as the
applied loads increase, while the arches could experience softening followed by a
sharpening, a behavior known as snap-through buckling.

.
NON-LINEARITIES OF THE MATERIAL

Another important source of nonlinearities derives from the nonlinear relationship

between stress and strain that has been recognized in various structural behaviors.
Several factors can cause material behavior to be nonlinear. The dependence of the
stress-strain relationship of the material on the loading history (as in the case of
plasticity problems), the duration of loading (as in the case of creep analysis), and
the temperature (as in the case of case of thermoplasticity) are some of these factors.
This class of nonlinearities, known as material nonlinearities, can be devised to
simulate such effects that are relevant to different applications through the use of
constitutive relationships.

Limiting connections between beams and columns during earthquakes is one of the
applications where material nonlinearities can occur.

NON-LINEARITIES OF CONTACT.

A special class of nonlinear problems is related to the changing nature of the

boundary conditions of the structures involved in the analysis during movement.
This situation arises in the analysis of contact problems.

Structure buckling , gear tooth contacts, splice problems, threaded connections, and
impact solids are several examples that necessitate evaluation of contact contours.
Evaluation of contact contours (nodes, lines or surfaces) can be achieved using
separation (contact) elements between nodes on adjacent contours.
SOLUTION PROCEDURE FOR NON-LINEAR PROBLEMS.

In nonlinear problems, the stiffness of the structure, the applied loads and/or the
layout conditions can be affected by the induced displacements. The equilibrium of
the structure must be established in a deformed shape that is unknown and must be
assumed. At each equilibrium state along the equilibrium path, the resulting set of
equations is nonlinear. Therefore, a direct solution is not possible and an iterative
approach is necessary.

Several strategies have been developed to perform nonlinear analysis. Unlike linear
problems, it is extremely difficult, if not impossible, to implement a single strategy
with general validity for all problems. Very often, the problem at hand forces the
analyst to try different solution procedures or to select a certain procedure to obtain
the correct solution successfully (for example, "snap-through" buckling problems,
which is the violent loss of geometric stability , in frames and shells require
deformation-controlled loading strategies such as displacement-based and arc-length
controls rather than force-controlled loading).

For these reasons, it is strictly necessary that a computer program used for linear
analyzes has several alternative algorithms to cover a wide spectrum of nonlinear
applications. Such techniques lead to increased flexibility and the analysis can obtain
greater reliability and efficiency for the solution of a particular problem.

TIME CURVE CONCEPT.

For nonlinear static analysis, loads are applied in incremental steps through the use
of "time" curves. A time curve indicates how a load or constraint changes during
solution steps. For nonlinear dynamic analysis and nonlinear static analysis with
time-dependent material properties (e.g., creep), "time" represents the actual time
associated with the application of the loads.
The choice of the size of the "time" step depends on several factors such as the level
of nonlinearities of the problems and the solution procedure. A computer program
should be equipped with an adaptive auto-stepping algorithm to facilitate the analysis
and reduce the cost of the solution.

NON-STRUCTURAL LINEARITIES.

Structural nonlinearities The main sources of structural nonlinearities encountered in

practical applications are the following:
Geometric nonlinearities In nonlinear finite element analysis, a major source of
nonlinearities is due to the effect of large displacements on the overall geometric
configuration of structures. Structures that are being subjected to large displacements
can have significant changes in their geometry due to load-induced deformations that
can cause the structure to respond nonlinearly in a stiffening and/or softening mode. For
example, cable-type structures generally show sharpening behavior with increasing
applied loads while the arch may experience softening followed by sharpening, a
behavior called snap-through buckling. Cable-type structures Snap-through buckling
(violent loss of geometric stability) Material non-linearities.
Another important source of nonlinearities derives from the nonlinear relationship
between stress and strain that has been recognized in various structural behaviors.
Several factors can cause material behavior to be nonlinear. The dependence of the
stress-strain relationship of the material on the loading history (as in the case of
plasticity problems), the duration of loading (as in the case of creep analysis), and the
temperature (as in the case of case of thermoplasticity) are some of these factors. This
class of nonlinearities, known as material nonlinearities, can be devised to simulate such
effects that are relevant to different applications through the use of constitutive
relations. The boundary of beam-column connections during earthquakes is one of the
applications where material nonlinearities can occur. Beam-Column Connections
Contact Nonlinearities A special class of nonlinear problems is related to the changing
nature of the boundary conditions of the structures involved in the analysis during
movement. This situation arises in the analysis of contact problems. - Structural non-
linearities. Structure buckling, gear tooth contacts, splicing problems, threaded
connections, and impact solids are several examples that necessitate evaluation of
contact contours. Evaluation of contact contours (nodes, lines or surfaces) can be
achieved using separation (contact) elements between nodes on adjacent contours

DIFFERENCE BETWEEN LINEAR ANALYSIS AND NON-LINEAR

ANALYSIS

A linear analysis is one where the forces are directly proportional to the
displacements and as its name indicates there is a linear relationship between these
two variables. We could use an analogy with a straight line, the curve that describes
this behavior is a straight line. . We can be in the elastic range (Hook's law is
fulfilled) or we can also be in the inelastic range, but linearity is preserved. You
obtain the equilibrium equations from the undeformed geometry (in its initial state)
and for this reason they are called first order.
On the other hand, a non-linear analysis no longer exists. This proportionality and
depending on the problem or case to be analyzed, the relationships are no longer a
straight line and can be polynomial, trigonometric, exponential, radical, hyperbolic
relationships, etc. This is reflected in the stiffness matrix.

In reality, all structures behave in a non-linear way but engineers simplify and
idealize the problem and almost always do linear analyses.

You obtain the equilibrium equations in nonlinear analysis from an initial deformed
geometry that will be your starting point and from this moment by giving increments
of load you obtain another deformed geometry which will serve as the basis for the
next one...and so on until reaching the desired level of precision is therefore called
second order.

Among many nonlinearities, the most common are:

Geometric non-linearity, material non-linearity, delta p effects (upper case) delta p

effects (lower case) the effects of semi-rigid connections etc.

As it is an iterative process, it requires a lot of time and capacity because computer

programs that use many algorithms are required to find the solution depending on
what is needed. This is pushover. There are algorithms such as Newton-Rapshon,
Runge Kutta, which are methods that try to follow nonlinearity, but at the same time
linearizing the problem.