Sundararajan - Sujith - Spring 2014-2

University of Alberta
A rheological model for composite truck ground behavior

by
Sujith Sundararajan
A thesis submitted to the Faculty of Graduate Studies and Research in partial fulfillment
of the requirements for the degree of
Master of Science
in
Mining Engineering
Civil and Environmental Engineering
©Sujith Sundararajan
Spring 2014
Edmonton, Alberta
Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis
and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is
converted to, or otherwise made available in digital form, the University of Alberta will advise potential users
of the thesis of these terms.
The author reserves all other publications and other rights in association with the copyright in the thesis and,
except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or
otherwise reproduced in any material form whatsoever without the author's prior written permission.
This thesis is affectionately dedicated to my parents
SUNDARARAJAN
and
PRABHA
for their love, sacrifices, endless support and encouragement that has helped me
become what I am today
ABSTRACT
Mining haul trucks are operated on different ground conditions subjected to
adverse shock loading, creating the need to give more attention to the truck
components like suspension, tire, and the operating ground. Extensive
experimental research has been done on these components at scale.
In order to understand the deformation behavior and the overall relationship
between truck components and the operating ground for a full sized truck,
mathematical equations have been derived for truck operation on both rigid and
oil sand ground surfaces.
A rheological model is also proposed here that has the ability to predict the total
and individual strain of the components. A model consisting of suspension, tire,
and ground is created using MATLAB (Simulink). Finally, using the data
obtained from the three methods, graphs are plotted and comparisons made with
previous laboratory and field test results to validate the models and their
equations.
ACKNOWLEDGEMENT
I would like to thank my supervisor Dr. Tim Joseph for his guidance and support
throughout my time as a Master’s student in Mining Engineering. He was more
than a supervisor for me. Without his persistent support this research would have
not been possible. The discussions with him helped me to understand the real
picture. I would like to extend my thanks to Dr.Jozef Syzmanski for his valuable
inputs and feedback.
I am thankful for my great friends, who supported, encouraged, helped me in

editing the thesis, and provided valuable feedback throughout my graduate
studies, especially Sridhar Dasani, Swetha Parvathaneni, Santhanakrishnan
Ragupathy, Selvakumar Raju, Dineash Raju, Karthik Vaidhyanathan, Sumi, and
Yao Tang.
I would also like to thank Sanjith and Priya, my brother and sister-in-law, for their
continuous encouragement and support.
Special thanks to my office mates Zhihan Lin, Enjia Shi, Magreth, Niousha
Rahmani, Choushi Hu, and Ibrahim for their help. The support provided by them
in every aspect they could really helped me. The friendship with them will never
be forgotten.
I acknowledge the financial support provided by NSERC and the University of

Alberta. I offer my thanks to the Department of Civil and Environment
Engineering (UofA) for employing me at research/teaching assistant positions
during my graduate studies.
Finally, I thank all my family members for their support and encouragement.
Sujith
TABLE OF CONTENTS
CHAPTER 1 .......................................................................................................... 1
INTRODUCTION .................................................................................................. 1
1.1 Overview ....................................................................................................... 1
1.2. Purpose of research ...................................................................................... 2
1.3 Research approach ........................................................................................ 3
CHAPTER 2 .......................................................................................................... 6
LITERATURE REVIEW ....................................................................................... 6
2.1. Introduction .................................................................................................. 6
2.2. Fundamentals and stride towards the branch of rheology............................ 6
2.3. Important parameters in rheology, viscoelastic models and their studies.... 7
2.4. Complex fluid models and application of rheology in various fields ........ 16
2.5. Mathematical Parameters and suspension studies ..................................... 18
2.6. Ultra-class truck tire ................................................................................... 21
2.7. Ultra – class truck haul road ...................................................................... 25
CHAPTER 3 ........................................................................................................ 31
MATHEMATICAL MODEL FOR THE PERFORMANCE PREDICTION OF
TRUCK SUSPENSION AND TIRE ON A RIGID SURFACE .......................... 31
3.1. Introduction and Objective ........................................................................ 31
3.2. Basic assumptions ...................................................................................... 31
3.3. Suspension model parameters and its equations ........................................ 32
3.4. Truck tire model parameters and its equations .......................................... 38
3.5. Mathematical model equations .................................................................. 40
3.6. Mathematical model equation validation with field and laboratory results 42
3.6.1. Comparison with laboratory results .................................................... 43
3.6.2. Comparison with Field value (CAT 797B) ......................................... 47
3.7. Understanding the relationship between stroke length of suspension and
deformation of tire for full size truck ................................................................ 49
CHAPTER 4 ........................................................................................................ 51
MATHEMATICAL MODEL FOR PERFORMANCE PREDICTION OF
TRUCK SUSPENSION, TIRE, AND THE OIL SAND ...................................... 51
4.1. Introduction and objective ......................................................................... 51
4.2. Suspension model equations ...................................................................... 52
4.3. Tire model equations.................................................................................. 52
4.4. Oil sand model equations ........................................................................... 54
4.5. Oil sand mathematical model equations output ......................................... 57
4.5.1. Performance Output for CAT797B truck on oil sand ......................... 57
4.5.2. Performance output for CAT 797B truck on oil sand as per field
condition ....................................................................................................... 59
4.5.3. Understanding the relationship between stroke length of suspension
and deformation of tire and oil sand for full size truck…………...….……..61
4.5.4 Tire footprint area and ground stress comparison for CAT 797B
truck………………………………………………………………………...62
4.5.5. Performance prediction for 2.85m and 0.368m diameter tires………64
CHAPTER 5 ........................................................................................................ 67
RHEOLOGICAL MODEL FOR PERFORMANCE PREDICTION OF TRUCK
SUSPENSION AND TIRE ON A RIGID SURFACE ......................................... 67
5.1. Introduction and Objective ........................................................................ 67
5.2. Basic assumptions ...................................................................................... 67
5.3. Model layout and equations ....................................................................... 68
5.3.1. Rheological model for truck suspension and tire on a rigid surface ... 68
5.3.2. Truck Suspension model and related equations .................................. 69
5.3.3. Truck tire model and related equations ............................................... 75
5.4. Performance prediction ............................................................................. 76
5.4.1. Prediction for CAT 797B truck components ...................................... 76
5.4.2. Prediction for the previous laboratory result tires ............................... 80
5.5. Damping characteristics ............................................................................. 82
CHAPTER 6 ........................................................................................................ 85
RHEOLOGICAL MODEL FOR PERFORMANCE PREDICTION OF OIL
SAND, TRUCK SUSPENSION AND TIRE ON OIL SAND ............................. 85
6.1. Introduction and objective ......................................................................... 85
6.2. Rheological model for truck suspension and tire on oil sand .................... 85
6.2.1. Truck suspension and tire model parameters and equations ............... 86
6.2.2. Oil sand model parameters and equations .......................................... 87
6.3. Analytical output for the model ................................................................. 91
6.3.1. Analysis parameters ............................................................................ 91
6.3.2. Analysis Output to a CAT 797B truck ................................................ 92
CHAPTER 7…………………………………………………………………...100
MODELING OF TRUCK SUSPENSION, TIRE AND THE GROUND……...100
CHAPTER 8…………………………………………………………………...107
SUMMARY OF THESIS OUTCOMES AND CONCLUSIONS……………...107
CHAPTER 9…………………………………………………………………...112
FUTURE RECOMMENDATIONS……………………………………………112
LIST OF TABLES
Table.2.1. Material properties of tire tread and sidewall (From[13])……………22

Table.3.1. Model output and laboratory value comparison for 0.368m diameter
tire ......................................................................................................................... 45
Table.3.2. Model output and laboratory value comparison for 2.85m diameter tire
............................................................................................................................... 46
Table.3.3. Model output and field value comparison for 1-g load of CAT 797B. 48
Table.6.1. Parameters of oil sand rheological model elements ([60]) .................. 92
Table.7.1. Parameters for the Simulink model………………………………….101
Table.7.2. Simulink model output for different loads…………………………..106
Table.8.1. Comparison between physical, rheological, and Simulink model
outputs for CAT 797B…………………………………………………………..111
LIST OF FIGURES
Figure1.1 Flowchart showing the methodology and research approach to predict

the composite truck ground behavior……………………………………………...5
Figure 2.1Mechanical elements (a) Hooke Spring and (b) Newtonian Dashpot
(after [25])…………………………………………………………………………7
Figure.2.2. Mechanical models (a) Maxwell (b) Kelvin-Voigt (c) Standard linear
solid or Poynting - Thompson (d) Burgers (after [25])……………………………8
Figure.2.3. Strain response of Hooke and Kelvin model (after
[10])……………......................................................................................................9
Figure.2.4. Shear rate vs Viscosity for low - density polyethylene melt (after
[48])………………………………………………………………………………10
Figure.2.5. Shear strain vs shear stress for Newtonian and non-Newtonian fluids
(after [52])………………………………………………………………………..11
Figure 2.6. Oil sand Strain – time curve for 536kPa stress (after [60])………….15
Figure.2.7. Newtonian vs Shear thinning fluid (after [38])……………………...17
Figure.2.8: Oleo-pneumatic suspension………………………………………….19
Figure.2.9: Fixed and variable orifice (after [28])……………………………….19
Figure.2.10: Components of a radial and bias tires (after [19])………………….21
Figure.2.11. Load vs Footprint area of tire for various inflation pressures (after
[13])………………………………………………………………………………23
Figure.2.12. (Deformation.Diameter) vs Footprint area (after[60])……………..24
Figure. 2.13. Layout of Haul road for a 340ton truck (after [64])……………….25
Figure.2.14: Variation of oil sand stiffness under cyclic loading (after
[41])…………........................................................................................................26
Figure.2.15: Oil sand stiffness with season (after [44])………………………….27
Figure.2.16: Pressure stiffness with time for Oil sand (after [59])………………28
Figure.2.17: Stress bulb beneath the ground surface (after [60])………………...30
Figure.3.1. General Layout of components considered for the model…………...32
Figure.3.2. Haul truck positions and nomenclature……………………………...33
Figure.3.3. Pressure variation at Front and Rear struts for one cycle of CAT
797B……………………………………………………………………………...34
Figure.3.4. Recommended vs actual load distribution for CAT 797B…………..36
Figure.3.5. Available stroke length data for moving cycle of LR Strut…………37
Figure.3.6. Variation of available stroke length with pressure during moving cycle
for LR strut…................…………………………………………………………37
Figure.3.7. Variation in available stroke length with pressure during loading cycle
for LR Strut………………………………………………………………………38
Figure 3.8.Load vs Deformation for 0.368m diameter tire……………………....44
Figure 3.9.Load vs Footprint area for 0.368m diameter tire……………………..44
Figure 3.10.Load vs Deformation for 2.85m diameter tire…..…………………..45
Figure 3.11.Load vs Footprint area for 2.85m diameter tire……………………..46
Figure 3.12. Load vs Tire deformation for 55/80R63 tire at 600 kPa…...……….47
Figure 3.13. Load vs Tire footprint area for 55/80R63 tire at 600 kPa………….48
Figure 3.14.Available stroke length vs Tire deformation for CAT 797B at 600 kPa
……………………………………………………………………………………49
Figure.4.1. Oil sand deformation with different loading cycle…………………..58
Figure.4.2. Tire deformation with different loading cycle………….……………59
Figure.4.3. Tire deformation on oil sand with number of cycles for LR and RR
tires…………………………………………………………………………….....60
Figure.4.4. Comparison of total (tire and oil sand) deformation with increasing
number of cycles…………………………………………………………………60
Figure.4.5. Comparison between available stroke length and deformation of tire,
oil sand and total deformation at LR position……………………………………61
oil sand and total deformation at RR position…...……………………………….62
Figure.4.7. Ground stress distribution at the rear position of the loading area…..63
Figure.4.8. Tire footprint area with increasing number of cycles………….…….63
Figure.4.9. Oil sand deformation with load for 2.85m diameter tire…………….64
Figure.4.10. Tire deformation on oil sand with load for 2.85m diameter tire…...65
Figure.4.11. Oil sand deformation with load for 0.368m diameter tire………….65
Figure.4.12. Tire deformation on oil sand with load for 0.368m diameter tire….66
Figure.5.1. Rheological model for truck suspension and tire on a rigid surface...68
Figure.5.2. Rheological model for the truck suspension…………………………69
Figure.5.3. Relationship chart for strain value of the suspension………………..72
Figure.5.4 Truck tire element…………………………………………………….75
Figure.5.5. Suspension strain with time during loading cycle…………………...76
Figure.5.6. Variation of required force with available stroke length…………….77
Figure.5.7. Suspension strain with time during moving cycle for LR Strut……..78
Figure.5.8. Suspension strain with time during moving cycle for RR Strut……..78
Figure.5.9. Tire strain with time during moving cycle for LR Tire……………...79
Figure.5.10. Tire strain with time during moving cycle for RR Tire………..…...79
Figure.5.11. Strain for CAT 797B tire with g-loads……………………………..80
Figure.5.12. Strain for 2.85m diameter tire with g – loads………………………81
Figure.5.13. Strain for 0.368m diameter tire….………………………………….81
Figure.5.14. Variation of spring stiffness with pressure…………………………83
Figure.5.15. Variation of viscous damping coefficient with pressure…………...83
Figure.5.16. Variation of damping ratio with pressure…………………………..84
Figure.6.1. Rheological model for truck suspension, tire and oil sand………….85
Figure.6.2. Response of different rheological model under loading and unloading
(after [25])…..……………………………………………………………………88
Figure.6.3. Rheological model for oil sand (Burgers model)……………………89
Figure.6.4. Tire strain with stress on Oil sand…………………………………...93
Figure.6.5. Oil sand strain for 1sec loading duration during truck motion ...……94
Figure.6.6. Strain for 1sec loading on oil sand during first 10sec of a truck in
motion …………………………………………………………………………...94
Figure.6.7. Oil sand strain for 1sec loading period……………………..………..95
Figure.6.8. Oil sand total strain for different loading periods……………………96
Figure.6.9. Oil sand stress versus strain for different loading periods……...……96
Figure.6.10. Oil sand time dependant strain for different loading
periods….………………………………………………………………………...97
Figure.6.11. Oil sand permanent strain for different loading periods…………....98
Figure.6.12. Oil sand permanent strain modeling for 1 sec loading period……...98
Figure.6.13. Oil sand permanent strain modeling for different loading periods…99
Figure.7.1. Spring – Damper system for a truck…………….………………….100
Figure.7.2. Simulink model for truck suspension………………………………101
Figure.7.3. Output for CAT 797B rear suspension at 1g load………………….102
Figure.7.4. Simulink model for CAT 797B suspension and tire………………..102
Figure.7.5. Step input value (ground deformation)……………………………..104
Figure.7.6. Tire displacement at 1g load without step input………...………….105
Figure.7.7. Tire deformation at 1g load with step input ……………………….105
Figure.8.1. Research focus area and solutions………………………………….107
LIST OF SYMBOLS
" Inch
 Strain
 Strain rate
 Stress
 Isentropic gas constant
 Shear rate
 Relaxation time
 tr Deformation of tire on rigid surface
 ts Deformation of tire on oil sand
 os Deformation of oil sand
t Diameter of tire
$ Dollars
# Number
% Percentage
 Constant of proportionality
 Shear stress
 Bulk stress
 Viscosity
 Angular frequency
° Degree
 Damping ratio
LIST OF ABBREVIATIONS
AC Area of suspension cylinder
Atr Footprint area of tire for rigid surface
Aos Total area of oil sand

c Damping coefficient
cC Critical damping coefficient
C Celcius
CAT Caterpillar
D Depth of influence
d Ground deformation
E ps Post peak modulus of oil sand
E Young’s modulus
Fs Force on suspension
Fg Force on the ground
Ft Force on the tire
FOEM , F Damping force for full size OEM strut
FOEM , S Damping force for scale OEM strut
F Load or Force
F ft Force on the front tire
Frt Force on the rear tire

f Function
G Rigidity modulus
g Load level
Hz Hertz
kr Radial stiffness of tire
ks Stiffness of spring
kt Stiffness of tire
KOEM , F Full size OEM strut stiffness
KOEM , S Scale OEM strut stiffness
kN Kilo Newton
kPa Kilo Pascal
kp Oil sand pressure stiffness
LR Left rear
LF Left front
m Metres
Mpa Mega Pascal
mm Millimetres
min Minutes
ms Mass of the spring
NC Number of cycles
OEM Original equipment manufacturer
Ps Suspension pressure indicated by VIMS for every second
Pu Suspension pressure at unloaded condition
pi Inflation pressure of tire

Psi Pounds per square inch
Pas Pascal second
P Pressure
RR Right rear
RF Right front
R Gas constant
Sa Available stroke length
Su Stroke length at unloaded condition

SAE Society of Automotive Engineers
sec Seconds
t Time
T Absolute temperature
U/L Unloaded condition
VIMS Vital information management system
V Volume
wt Width of the tire
CHAPTER 1
INTRODUCTION
1.1 Overview
The truck is one of the most important components in the surface mining industry.
Repeated occurrence of high shock loading will result in premature failure of
truck components like suspension and the tire and could pose a danger to an
operator’s health and safety. The efficiency and utilization of the truck are also
decreased because of this resulting increased downtime. Production requirement
and global competition have driven global surface mining operations to move to
bigger equipment. Trucks have moved into the ultra-class category based on this
criteria. The two most commonly used trucks are CAT 797B and CAT 797F,
which have payloads of 380 tons and 400 tons. There has been an increase in
payload with years and this could increase further in the future resulting in a huge
load acting on the ground. Therefore, it is necessary to develop an understanding
of suspension-tire-ground interactions which can be used for the performance
prediction of various trucks operated on hard and soft surfaces.
The knowledge of deformation behavior and the relationship between the

components is still not evident. Until now researchers have focussed on
performance of suspension, tire and the ground separately and under small scale,
which do not provide any relationship between these components. This motivated
the research further to understand the overall behavior of the truck from the
suspension to the ground under full scale.
The performance of each of the components are interrelated and depend on the
conditions of each other. Even a small change in the performance of any one of
the components would make the truck deviate from its normal behavior. Ground
conditions are a critical factor behind this and especially oil sand deforms
providing an undulated surface for the truck to be operated on. Hence
1
Chapter 1 - Introduction
understanding of the composite behavior of the truck over varying load conditions
is essential to rid a solution for the problems faced by truck operators.
1.2. Purpose of research
Nowadays, most of the trucks in the mining industry are equipped with the VIMS
payload system. In a sample field study conducted by Sarma (2009) on CAT
797B truck, it was found that when the truck started its hauling cycle, the
available stroke length was only 0.032m (1.27”). Also after 500 operating hours,
the available stroke length of the suspension dropped from 0.152m (6”) to 0.124m
(4.9”) [61]. This difference in available stroke length is high and the loading
effects on components of a truck will be very large, resulting in premature failure
of truck components. Hence, the study on suspension cylinder to predict
performance is vital for the truck to eliminate the detrimental effect of varying
ground conditions.
Tires, on the other hand are subjected to high g-level, causing unexpected failures
due to sidewall bulge and impacts [13]. The cost of a single tire is over $130,000
and companies experience shortage of tires [14]. Over time the rim and the tires
experience cyclic loading at high g loads resulting in crack and premature failure.
So understanding the performance of the tire for varying load conditions helps in
ensuring that they are operated with minimum damage for a longer period of time
and in turn maximizing the tire life.
The in-pit ground, especially oil sand, deteriorates within a few cycles resulting in
uneven surfaces. Oil sand exhibits two main charateristics, deformation and
rutting. The behavior of oil sand under these two categories are relatively well
unknown and correlating these effects to the action of the hauler will help the
operators to operate the truck with less downtime. Also the strength of oil sand is
very poor to bear a huge amount of load from the truck. Due to these, the driving
becomes very difficult causing a huge force to be transferred to the truck
2
components, which leads to major fatigue, equipment instability, and improper

payload distribution [11] [40] [41].
All these factors highlight the need for evaluating the suspension, tire and the
ground together and finding an effective solution for understanding the behavior
and relationship between the respective components. It is necessary to provide a
better way for prediction and understanding of the complexity of composite truck-
ground behavior. The mathematical equations and rheological models proposed in
this research will help the individuals concerned with truck and road maintenance
to monitor and improve either the ground condition or the components on the
truck.
1.3 Research approach
Dynamic behavior of truck components is a subject of on-going interest to the

mining industries and so are the practical models to interpret, characterize,
predict, and interrelate the various phenomena such as available stroke,
deformation, damping, and load impact.
This research work begins with the derivation of mathematical equations using
various fundamentals of physics related to materials. Basic thermodynamics of
gases, general stiffness equations of tires involving parameters like inflation
pressure, overall diameter and width were used to find equations relating the
stroke length of the suspension and the deformation of the tire. The work on the
interaction of tires on oil sand by Sharma (2009) is also included when truck is
considered to be operated on oil sand.
Finally with all the derived equations, comparisons have been made with previous
data obtained for the following on rigid and oil sand surfaces
3
 Field data for a CAT 797B with 3.84m diameter tires (55/80R63) [60]
 Experimental data for a 0.368m diameter labscale tire [60]
 Experimental data for a 2.85m diameter tire (30 00R51) [13]
Next, a rheological model is proposed here that predicts the total deformation and
the deformation of individual truck components and the ground. Basic mechanical
models like Burgers, Kelvin-Voigt, and Maxwell have been used. The normal
stress-strain Hooke equation and the Newtons fluid equation are used to derive an
expression for individual component models. Data measured up by Sharma (2009)
for oil sand parameter of Burgers model are used and extended here by including
suspension and tires in the picture. Finally, comparisons were made to the
mathematical equations derived earlier.
Finally a model consisting of suspension, tire, and ground was modelled using
MATLAB (SIMULINK) in order to understand the behavior and deformation
characteristics of each component. Plots were produced to better understand the
response of the Simulink model for various conditions. With the results obtained
from the Simulink model, comparisons were then made with the results of the
mathematical and rheological methods.
The mathematical and rheological methods used here to analyze the behavior of
components provides good information to understand the relationship between the
components of the truck and also on the increasing problems of shock loading due
to bad road conditions. Using the models in this paper, graphical analyses were
conducted to easily correlate the functionality of each component for various
loading conditions.
This work enables engineers and operators concerned with truck performance to
better understand the impact of high loads and the behavior of the truck
components to the ground. This will also help operations teams in industry to
4
perform on-time maintenance of the equipment, and maintain the quality of haul
roads for a safe work environment, cost reductions and improvement in reliability.
Figure. 1.1. Flowchart showing the methodology and research approach to predict
the composite truck ground behavior
5
CHAPTER 2
LITERATURE REVIEW
2.1. Introduction
This chapter provides an understanding of previous work to better understand the

scope of the research. A truck has a complex behavior while loaded and in motion
and which depends on the performance of components like suspension and tires.
The effects of these components are further related to the ground conditions.
Firstly, a rheological method for predicting the deformation of both the truck
components and ground has been examined. Field studies had been conducted by
researchers in an attempt to realize and improve the performance of components
like suspension, tires and ground involved in truck – ground behavior. Also,
experimental results were obtained under large and small scale. Finally, these are
reviewed and where applicable, incorporated into this study.
2.2. Fundamentals and stride towards the branch of rheology
“Rheology is the study of materials deformation and flow properties” [10]. In

general, for a beginner, it is a discipline known as a mixture of more than one
branch. The formal explanation of rheology was accepted in the year 1929. Since
it is quite a new science, its prominence came only during the latter half of the
20th century. It is now recognized worldwide and many countries now have
national societies of rheology [10].
The most important aspect of rheology is the time dependent behavior of

materials. Rheology is essentially concerned with materials that show a
combination of elastic, viscous and plastic behavior. Two well-known British
scientists, Isaac Newton and Robert Hooke, set the borders of the present
knowledge of rheology. Their work set the frontiers of “classical elasticity” and
“fluid dynamics” [65].
6
Chapter 2 –Literature review
Weber (1835) performed an experiment on silk threads and it showed a non –

perfect elastic behavior. He noted that during application of load, it showed an
instantaneous elongation followed by an additional extension with time.
Conversely it showed an instantaneous reduction in length followed by an
additional steady decrease in length till it reached its original length. This
response matched with the behavior of a viscoelastic solid [71]. Thus, materials
having a solid-like behavior which cannot be defined by Hooke’s law alone and
similarity to flow like a liquid were termed as “viscoelasticity” [65].
Scientists then realized that there are also materials which do not come under the
category of pure solid or pure liquid. Thus “Rheology” is focussed on the study of
the behavior of materials that fall under the two parameters termed as,
“viscoelasticity” and “non-Newtonian fluid mechanics” [70]. Also, the terms
“elasticity” and “viscosity” are very important in rheology because materials
display either of the property or mixture of both at once.
2.3. Important parameters in rheology, viscoelastic models and their studies
The most influential period in rheology was the introduction of mechanical

models. Viscoelastic materials are basically modelled using two elements, the
spring and the dashpot (Figure.2.1).
Figure.2.1. Mechanical elements (a) Hooke Spring and (b) Newtonian Dashpot
(after [25])
There are basically four main mechanical models (Figure.2.2) and each of these is
made up of spring and dashpot elements in parallel or series combination. The
7
constitutive equations for these models are usually derived based on the force,
extension, stress, strain and time dependancy of materials [71].
Figure.2.2. Mechanical models (a) Maxwell (b) Kelvin-Voigt (c) Standard linear
solid or Poynting - Thompson (d) Burgers (after [25])
Whatever the type of model used in rheology, it is mainly interested with four
main sub-areas [10]
i) Rheometry: Measurement of rheological properties
ii) Constitutive equations
iii) Measurement of flow behavior in complex geometries
iv) Calculation of behavior in complex flows
In terms of springs in mechanical models, it is purely an elastic component with

the rheological property such as Young’s Modulus. Hooke (1678) presented a
one-dimensional law that relates stress over two material factors: the modulus of
elasticity (E) and axial strain (ɛ) [1] [2], which can be generalized that stress is
proportional to strain and is given by,
  E [2.1]
8
Barnes (1999) indicated the difference between a pure elastic material (Hooke
model) and Kelvin model (Figure.2.3). He noted that when a sudden stress is
applied, the final strain value of the elastic material reached immediately whereas
the strain took a time curve since it also has a viscous element [10].
Figure.2.3. Strain response of Hooke and Kelvin model (after [10])
In terms of a damper in mechanical models, viscosity is the main parameter and is

a “property of liquid, which can be stated as the resistance to fluid flow” [62]. The
viscosity of a viscoelastic material does not remain constant and varies according
to the parameters affecting it. Barnes (1999) noted that the following is the main
cause for a change in viscosity.
i) Change in Shear rate and time of shearing
ii) Change in Pressure
iii) Change in Temperature
The effect of shear rate on viscosity for a particular temperature was found to
reduce for a low-density polyethylene melt [48]. He also gave the sketch which
shows the variation (Figure.2.4).
9
Figure.2.4. Shear rate vs Viscosity for low-density polyethylene melt (after,

[48])
Similarly Lewis (1987) noted that the magnitude of change in viscosity is 2% per
degree C change in temperature. An experiment was conducted on liquids to
verify the effect of temperature on viscosity [62] and it was found that the
relationship followed the “Arrhenius equation” (Equation. 2.2)
 Ea 
   exp   [2.2]
 RT 
Where,  is the viscosity (Pas),  is a constant (Pas), Ea is the activation energy

(Kcal g-mole), R is the gas constant (kcal/g-mole °K), and T is the absolute
temperature (°K).
The important difference between “Non-Newtonian fluids” and “Viscoelasticity”

that forms the scope of rheology was described by Barnes (1999). He stated that
“All Viscoelastic liquids are non-Newtonian but not all non-Newtonian liquids are
Viscoelastic” [10]. Nguyen (2012) studied the flow behavior of various Non-
Newtonian fluids. The flow characteristics of these fluids vary in different ways
10
from Newtonian fluids. He studied the characteristics of different Non-Newtonian

fluids and presented a graph (Figure.2.5) to represent the variation [52].
Figure.2.5. Shear strain vs shear stress for Newtonian and Non-Newtonian fluids
(after [52])
Further, Non-Newtonian fluids are classified into time – dependent and time-
independent. The materials described as per figure. 2.5 are time-independent
fluids. For a “Newtonian fluid”, the viscosity is a constant. Viscosity decreases
with increase in shear rate for a “Pseudoplastic fluid” and it increases for a
“Dilatant fluid”. “Bingham plastic” has a yield stress and the for the materials to
start flowing, the applied stress should be more than the yield stress. “Yield –
Pseudoplastic” has both yield stress and non-linear flow. Time-dependent fluids
are thixotropic and rheopectic fluids. The vast majority of fluids are time-
independent [52].
The linearity of many materials under rheology that show characteristics in-
between that of a pure solid and pure liquid are not known until now since it can
provide either a linear or non-linear behavior [48]. A solid which shows a
viscoelastic behavior is known as “viscoelastic solid” and a liquid which shows a
11
viscoelastic behavior is known as “viscoelastic liquid”. Regarding linear models,

Kelvin (1865) came with a linear rheological equation for a “viscoelastic solid”
(equation. 2.3) and Maxwell for a “viscoelastic liquid” (equation 2.4) [71].
d
  G   [2.3]
dt
where, G is a material constant
d d
   [2.4]
dt dt
where,  is the time constant,  is the viscosity (Pas),  is the strain
Later on, Jeffrey (1929) extended the Maxwell equation and came up with a
constitutive equation in one dimension that could relate to modern science
(equation 2.5) [39].
 u  2u
      [2.5]
t x xt
where,  is the elastic viscosity (Pas),  is the relaxation time (sec),  is the shear
stress,  is the Newtonian viscosity (Pas).
He used the one-dimensional equation to solve some of the problems related to

earth’s crust [2]. Boltzmann (1874) provided an integral constitutive equation
(equation 2.6) which can describe both solid like and liquid like behavior [53].
t
d  
  t   E  t    d [2.6]
0
d
12
Where,  is the stress,  is the strain, E is the relaxation function,  is the

variable of integration
Provenzano et al (2002) stated that for relaxation function is not dependant on

stress and strain as per Boltzmann’s equation. So he used two non-linear
viscoelastic theories known as “Schapery’s single integral non-linear theory” and
“Modified superposition principle” to study and identify the best theory to predict
the behavior of ligaments. He found the later to provide better results in predicting
the elastic and viscous non-linearity [53].
Various research has been done on the applicability, merits and demerits of
viscoelastic models. Roylance (2001), stated that there are three commonly used
“Viscoelastic tensile test”. They are creep, relaxation, and dynamic (sinusoidal)
loading. These three tests exhibit a time – dependant behavior [55]. Apart from
creep and relaxation, viscoelastic materials also exhibit “hysteresis” [29].
Hackley and Ferraris (2001) discussed about various rheological terminologies

and even gave the equation for the sinusoidal loading (equation.2.7 and 2.8) of
stress and strain, which is a property of linear viscoelasticity. Both creep and
relaxation can be considered as Time-Dependant behavior [35].
i) Creep: Materials response of strain with time to a constant stress loading
ii) Relaxation: Materials response of stress with time to a constant strain
   0 cos t [2.7]
   0 cos t    [2.8]
where,  is the angular frequency,  is the phase angle, t is the time
Roylance (2001) found the use of the sinusoidal loading method in polymers,
which provides a suitable value for its “short time” response. He studied the linear
viscoelastic and creep compliance response of polymers and polymer – matrix
13
composites and noted that upon loading, initial strain occurs similar to elastic
deformation (glassy compliance - Cg) and later on it reaches the equilibrium
position (rubbery compliance - Cr). He also stated that the accuracy of both creep
and relaxation of time equal to seconds or less is very low and is appropriate only
for studying material reaction from minutes to days [55].
Both de Haan and Sluimer (2001) and Robert et al (2012) say that both Kelvin –
Voigt and Maxwell models are not good in predicting the time – dependant
behavior (i.e. creep and relaxation). Dey and Basudhar (2010) noted that when
analyzed the creep and relaxation behavior for a viscoelastic soil, the Maxwell
model and Kelvin-Voigt model were not able to predict the time-dependant
behavior properly.
To overcome the demerits of these two models, De Haan and Sluimer (2001)
studied the time – dependant behavior of building materials using a standard
linear solid model. They found that this model was able to handle damping
situations for building materials very well and a direct relation could be set for
time-dependant behavior like creep and relaxation [23].
Attila et al (2004) used another three parameter model, the Poynting-Thomson

model to define the stresses and the displacement that arise in a foil tube with time
for a temperature range of -20°C and 45°C. They identified the Poynting-
Thomson model as the best model in describing creep and relaxation
characteristics. They used separate models and equations for both creep (equation
2.9) and relaxation (equation 2.10) by changing the location of the mechanical
element, and derived the rheological parameters using a least square and finite
element method [6].
 E1  E2 E 
   1  [2.9]
E1  E2 E1  E2 E1  E2
14
  E1 
   E1   1   [2.10]
E2  E2 
Where,  is the viscosity,  and  are the stress and strain rates, E1 and E2 are the
Young’s modulus of springs 1 and 2,  and  are the stress and strain
Dey and Basudhar (2010) did a study of viscoelastic soils and showed how the
“Burgers model” is more effective than the other mechanical models (Maxwell,
Kelvin-Voigt, and Poynting-Thompson) in predicting the time-dependant
behavior of “saturated viscoelastic soil” and described the “Burgers model” as the
most effective model [25].
Later, Sharma (2009) used the “Burgers model” to find a stiffness value for oil
sand through interpretation of the deformation – time curve. The strain time curve
for oil sand at 536 kPa was plotted (Figure.2.6) and the slope of the straight line
drawn asymptotic to the strain-time curve at constant load to determine the value
of the viscosity of the second damper. Similarly the values of the other parameters
were found using the model [60].
Figure.2.6. Oil sand Strain – time curve for 536kPa stress (after [60])
15
2.4. Complex fluid models and application of rheology in various fields
Larson (1999) noted that materials with both properties of elasticity and viscosity
can also be regarded as “complex fluids”, which comes under the category of
rheology. Some examples of complex fluids are suspensions, foods, shampoo,
fresh cement and concrete, agricultural products, toothpaste etc [48]. Most of the
fluids discovered are non-Newtonian with only a few being Newtonian like water
and some oils. Hence non-linearity is important in describing the complex fluids
which are the complex fluid models [63]. Various types of model and their use
were explained by Hackley and Ferraris (2001). They noted that the models can
be regarded as two, three or four parameter models based on a number of
unknown parameters [35].
i) “Power – law model” - Shear-thinning and Shear – thickening fluids
ii) “Herschel-Beckley”, “Bingham Plastic” and “Casson” – Viscoplastic
fluids exhibiting a yield response
iii) “Cross” and “Carreau-Yasuda” – Pseudoplastic flow
iv) “Ellis” and “Meter” – Pseudoplastic material exhibiting a power law
relationship
The simplest of all the models is the “Power – law” model [63]. Balhoff (2005)
studied various models and modelled the non-Newtonian flow of packed beds at
pore scale and described the Power law relationship (equation. 2.11) [7].
  0 n [2.11]
Where,  is the shear stress (Pa), 0 is the viscosity (Pas),  is the shear rate
(1/sec), n is the power law index
“If the value of n is less than unity it is a Shear – thinning fluid and if it greater
than unity it is a Shear – thickening fluid” [7]. Hou et al (2007) studied the shear
16
– thinning effect on a fluid viscous damper. He tested the effect on Silicone oil
and found that the shear rate of the oil was high enough to cause a shear thinning
effect. He found the shear rate value to be 24000sec-1 and concluded that this
value of Silicone oil is very high and it confirms that it is a non-Newtonian liquid
[38].
Figure.2.7. Newtonian vs Shear thinning fluid (after [38])
A comparison of different models was done on toothpaste [63]. It was found that
for the behavior of toothpaste, the “Herschel-Beckley” model was found to be the
best fit. Larson (1999) studied the rheological properties of a glassy liquid like
zinc alkali and found that the property is dominated by the modulus and a very
long relaxation time [48].
Banfill (2006) studied the rheology of fresh cement and concrete to understand
the performance with practical situations. He found the behavior very close with
Bingham model and was able to understand the link between rheology and
technology [8].
After the formal naming of rheology in 1929, a lot of new things were developed
in rheology and the application of it improved vastly. Differential integral
equations related to experimental advances were made, different advance fluids
like memory fluids were found out, and finally computational rheology came into
the picture with the ability to perform various simulations. Today the scope of
17
rheology is wider and has applications in various fields like polymers, suspension,
synthetic-fiber, plastic-processing industries, pharmaceutical and food industries,
biotechnology, chemical processing industries [26].
2.5. Mathematical Parameters and suspension studies
In the mining industry, there has been a rise in unplanned downtime due to
increase in equipment capacity and production requirements [27]. Components
like suspension, tire, and ground conditions are the main reason behind this and
research has been done for each component separately. Suspension systems can
improve the characteristics of a vehicle like safety and ride comfort. A lot of
research has been carried out to meet the requirement of better ride quality and
safety of the operator [5] [17] [72]. The important parameters of the suspension
are that it provides better handling by making the contact between the tire and
road stable, and also reacting to load variations resulting in ride comfort [74].
El-Sayed (2003), in his research, discussed the history, types and characteristics
of suspension systems. His main focus was investigating the suspension system
for a large ultra-class truck such as the Caterpillar 797B. He recorded suspension
data from the field and found that “topping-Up” seemed as the main reason for
failure of the system. There was only a clearance of 3.18 cm at the rear strut
location when a trip cycle started. Due to this, the main concerns were truck
reliability and spinal injury to the operator [28]. The majority of the weight during
truck hauling was distributed at the rear suspension [18].
El-Sayed (2003) conducted a thermodynamic and fluid flow analysis and studied
some parameters on the current shock absorber, known as a “simple shock
absorber” or “oleo-pneumatic suspension”. He was the first to make an attempt to
understand the effect of a variable orifice in suspension design [28]. “Oleo-
pneumatic suspensions” are the most effectively designed suspensions for aircraft
18
and its characteristics are affected even with a small deviation from the normal
operating parameters [36].
Figure.2.8: Oleo-pneumatic suspension

1. Piston rod, 2. Nitrogen chamber, 3. Cylinder, 4 and 7. Opening,
5. Oil chamber, 6. Ball check valve, 8. Fixed Orifice (after [61])
Comparisons were made between these two types of suspension using the
modeling software “MATLAB (Simulink)”. He found that the variable orifice
suspension showed a better performance than the fixed orifice suspension by
varying the damping force with load conditions [28].
Figure.2.9: Fixed and variable orifice (after [28])

19
Later on, Santos (2007) focussed on reducing the shock loading effect caused due
to “topping-up” by introducing a modified suspension model. He performed
laboratory tests on a scaled version of the OEM and modified struts. The test was
carried out for compression stroke and with hydraulic oil only. To collect the
signals from sensors, a data acquisition system was used. Processed via a
LabView system [57].
Even though Santos (2007) showed the effectiveness of a modified strut, there
were certain exceptions. The velocity had to be greater than 45mm/s to
significantly affect the damping force. The modified damper showed improved
performance only at 75mm of stroke. The computer model was valid only at
higher frequencies [57]. Finally the most important thing is that the test was done
with only hydraulic oil, which is not the case in an actual mining truck
suspension, where both oil and gas are used.
Finally, Soni (2009) continued the work ignored by El-Sayed (2003) and Santos
(2007). He performed the test by including Nitrogen gas in the suspension and
found the effects due to it. A scaling approach was carried out to predict the
performance of the full size suspension. Four parameters were obtained such as
stroke, displacement, internal pressures, and damping force with respect to
loading frequency and initial charging pressure [61].
 KOEM , F  FOEM , F
   [2.12]
 KOEM , S Closure FOEM , S
Where, KOEM , F and KOEM , S are the stiffness of full size and scaled original
equipment manufacturers strut, FOEM , F and FOEM ,S are the force on the full size
and scaled original equipment manufacturers strut
20
2.6. Ultra-class truck tires
The development of pneumatic tires began in 1845 [68]. The construction work
and details of its parts appeared only in 1950 [51]. They also stated that study of a
tires individual component's fatigue behavior is very expensive and sometimes
requires the destruction of the entire tire. 73% of the tires produced usually are for
passenger cars. These are classified basically into two types: bias and radial. The
latter is mostly used in industries [69]. The characteristics of a pneumatic tire is
still not fully understood and studies are going on to predict the exact behavior
[47].
Figure.2.10: Components of a radial and bias tires (after [19])
Literatures only exists for passenger and highway tires. But, these do not provide
sufficient information to model or predict the performance of ultra-class “off-
road” tires used by the mining industry [22] [66] [67]. Various large scale
experiments and analysis are performed by tire manufacturers, but they are very
reluctant to provide that information to the public because of competition.
21
However, Bolster (2007), obtained information from the ‘Society of Automotive

Engineers (SAE)’ about tires and rims. Several site visits in BC and Fort
McMurray were made and even manufacturers, service providers, and end users
of rims and tires were approached [13].
He obtained the material properties of a tire sidewall and tread from Goodyear
(Table.2.1). Even though the purpose of these components of the tire are the same,
the properties varied. Mainly the Young’s modulus value which is the parameter
that is used to find out the deformation for various loads also varied [13].
Material Property Tire Tread Tire Sidewall

Elastic modulus (N/ m2) 3.75 x 106 2.9 x 106
Poissons ratio 0.49 0.49
Shear modulus (N/m2) 2.9 x 106 2.9 x 106
Thermal expansion coefficient 6.7 x 10-4 6.7 x 10-4
Density (kg/m3) 979 979
Thermal Conductivity (W/mk) 0.14 0.14
Tensile strength (N/m2) 1.4 x 107 1.4 x 107
Yield strength (N/m2) 9.2 x 106 9.2 x 106
Table.2.1. Material properties of Tire tread and sidewall (From [13])
Bolster (2007) performed tests on 30.00R51 series rims and tire at the I.F.
Morrison Laboratory, University of Alberta. A 170 ton hauler model was used for
this test. Stress and strain data were collected for various rim and tire positions
and at various g-loads. It showed that the value of load is concentrated more at the
base contact (180 Deg). Predictions were made for load values higher than 1.4g
due to laboratory loading restrictions [13].
22
Figure.2.11: Load vs Footprint Area of tire for various inflation pressures (after
[13])
Lin (2007) studied the strengths of all tire parts and discussed how the modulus
varies for different tire manufacturers. He also stated the importance of inflation
pressure. With the measurement of pressure changes in many large size tires, he
came up with a formula known as “Ryne’s regression formula”, which has a great
influence for pneumatic tire design [51].
Kasprzak et al (2006) described that the shape, size, and contact footprint of the
tire is influenced by the inflation pressure and it should be considered as a
parameter of the tire rather than as an operating condition [45].
Sharma (2010) conducted a tire flexure test on a rigid surface, sand, and oil sand.
He used a 0.368m diameter tire and made comparisons with the field and
experimental values of 2.85m, 3.84m, and 0.368m tire sizes at 1g – 1.4g load.
Both static and cyclic tests were conducted for different inflation pressures [60].
23
Figure.2.12: (Deformation. Diameter) vs Footprint area (after [60])
Figure.2.12 shows that the result is a common linear line connecting the lines of
all the tire sizes. He obtained an empirical equation (Equation 2.13) for the tire
footprint area of a rigid surface and proved that the equation could be valid for
when compared to the field value.
A  1.35 [2.13]
Where,  ,  , and A are the deformation, diameter, and footprint area of the tire
Sharma (2010) also noted that the value of 1.35 in equation 2.13 changes for oil
sand with 11% bitumen content. He described that variability as C, which has a
value between 1.66 to 1.11 based on the number of cycles. Finally, he obtained a
relationship for deformation of a tire on oil sand (Equation.2.14) [60].
 1

1  2 4 Ft  2
 ts    os    os 
CE 
[2.14]
2  
 
24
Where,  os and  ts are the deformation of oil sand and tire on oil sand, E and Ft
are the Young’s modulus and force on the tire, C is the oil sand model constant
2.7. Ultra – class truck haul road
Haul roads are built basically with three materials. Sand for the sub-base, pit run
for the base, and crushed gravel for the surface layer [73]. There has been an
improved application of crushed limestone for the surface layer, which has shown
better stiffness and rolling resistance than the crushed gravel [37].
Figure. 2.13. Layout of Haul road for a 340ton truck (after [64])
Haul roads also play a vital role in mining operations similar to trucks and
shovels. Tannant and Regensburg (2001) predicted the deflections at the surface
for various truck sizes (170ton – 340ton) and found that for a 340ton truck the
deflections was around 8mm [64].
Most mine haul roads have problems like Potholes, rutting, and settlement. These
generate impact forces, which are then transferred to the frame and suspension of
the truck through the tires. Impact forces due to poorly maintained haul roads
reduces the life of the tire, causes metal to metal contact in suspension, shortens
truck life, increases maintenance costs and safety issues. Hence for today’s
mining industries operating trucks with higher truck loads and speeds, it is very
25
important to maintain the road surface condition in perfect condition to avoid all
the consequences [64].
Concrete and Asphalt are the materials that can offer better traction and rolling
resistance than crushed gravel, but it is more costly and mining companies are not
ready to invest so much. The largest trucks are used in the oil sand fields, which
handles a large amount of materials every year. The worlds largest deposit of oil
sand is in Alberta, Canada. Nowadays in the oil sand fields, for permanent and
temporary roads, oil sand is used as a construction material. [4].
Mining, construction, and manufacturing companies are mainly concerned with

bituminous oil sand since it creates mobility problems for large haul trucks and
shovels. The main reason for this difficulty is the presence of the bitumen content
ranging between 8% to 15% by volume. Usually the deterioration rate is higher
during summer than winter [3]. Joseph et al (2003) found that the stability of the
truck and shovel worsens with just few cycles of operation on a soft-ground [41].
Grozic (1999) stated that as the number of cycles increases, the stiffness of almost
all the Geotechnical materials becomes less and this is truly applicable for oil
sands [33]. Joseph (2002) showed that as the number of loading cycle on the truck
increases, the oil sand softens and in turn decreases the stiffness [40].
Figure.2.14: Variation of oil sand stiffness under cyclic loading (after [41])
26
Joseph (2002) also stated that the oil sand stiffness is a function of deformation
regardless of temperature, grade, and Geotechnical properties. He proposed an
empirical equation (2.15) to predict the stiffness and the relationship between
other parameters [43].
B
 kD  1   d
    C  [2.15]
 F     D
Where, k is the ground stiffness, F is the load, D is the depth of influence, d is

the ground deformation, B and C are empirical constants.
Figure.2.15: Oil sand stiffness with season (after [44])
Sharif – Abadi (2006) conducted an experiment on oil sand and came up with an
empirical equation to predict the total deformation (equation. 2.16) [59].
     
t   B     [2.16]
 At   106.7  NC  
0.3
With A  0.0317  7.318 , B  0.0002  0.036
27
Where,  t is the total deformation of the ground,  is the stress, t is the time, NC
is the number of cycles, A and B are empirical constants
Sharif – Abadi (2006) performed a plate load test in the laboratory and to
eliminate the effects of varying footprint, he used pressure instead of force, which
is the pressure per unit deformation, known as “Pressure stiffness”. He also found
out that for oil sand, regardless of the number of cycles, loading frequency,
relaxation interval between cycles, the “Pressure stiffness” converged to a
constant value of 8KPa/mm (Figure. 2.16) [59].
Figure.2.16: Pressure stiffness with time for Oil sand (after [59])
Anochie - Boateng et al (2010) conducted experiments on three oil sands with

bitumen content of 8.5%, 13.3% and 14.5% by volume and found from the results
that the behavior of oil sand and its modulus is affected by bitumen content,
temperature, and applied stress. He found the dynamic modulus as the best
parameter that can take into account of all the characteristic that affect the
behavior of oil sand. Comparing with various model outputs and the accuracy of
it, he proposed an adjustable model for practical use in the field [4].
28
E   204 1.712 wb1.882T 1930 [2.17]
Where, E  is the dynamic modulus,  is the bulk stress (  d  3 3 ), T is the
temperature, wb is the bitumen content,  d is the cyclic stress, and  3 is the

constant confining stress
On the other hand, regarding permanent deformation of soils, Barksdale (1972),

and Lekarp et al (2000) found moisture content and loading parameters as the
most important parameters [9] [49]. Anochie - Boateng (2007) believed that this
permanent deformation may be due to two main parameters known as “Sinkage”
and “Rutting” [3]. Sinkage is measured when the truck is being loaded and rutting
is measured at a point when the truck makes a number of passes [56].
Joseph (2005) described “oil sand as elastic-plastic material and proposed a

constitutive model (equation. 2.18)”[42]. But this model is applicable only for
static loading condition. He also stated that oil sand shows viscous behavior also
along with elastic and plastic properties [42].
E ps  1.37 31143 [2.18]
Where, E ps is the Elastic modulus,  3 is the confining pressure, and 1 is the axial
strain
Li and Chalaturnyk (2005) proposed few empirical models to describe the

behavior of oil sand (equation. 2.19 and equation. 2.20) [50].
0.5
 
E  950 Pa  3  [2.19]
 Pa 
E  343 30.875 [2.20]
29
Where, E is the Elastic modulus,  3 is the confining pressure, and Pa is the

atmospheric pressure
The depth of influence (stress level) beneath the ground surface was studied by
Joseph (2002) and he described that the depth upto which the stress influence
depends on the footprint area of the tire (equation. 2.21) [40].
D 3 A [2.21]
Where, D is the depth of influence, and A is the footprint area of the tire
Further, the parameters of the stress bulb; including size, shape, and magnitude in
the layers of a haul road depends on the inflation pressure and size of the tire [64].
It was also found that pore pressure are not a major concern beneath the surface
since it is dissolved mostly on the surface of the ground and do not pose any
influence while calculating the effective stress [59].
Figure.2.17: Stress bulb beneath the ground surface (after [60])

30
CHAPTER 3
MATHEMATICAL MODEL FOR THE PERFORMANCE PREDICTION
OF TRUCK SUSPENSION AND TIRE ON A RIGID SURFACE
3.1. Introduction and Objective
The main aim of the mathematical model is to provide a means of understanding

the performance and characteristics of truck components like suspension, tire and
the ground without complicated mathematics. This chapter consists of the
mathematical model for the truck components on a rigid surface. Mathematical
equations are the fundamentals for finding the required parameters without any
simulation or analysis. These equations are simple and can be derived based on
basic mathematics, which most of the people concerned with truck operation and
maintenance can understand.
The important truck components like suspension and the tire are considered here
and each component has their applicable mathematical equation based on the
properties of the respective component. These parts work together and the overall
performance of each part depends on the condition of the other parts. This means
they are interrelated and dependant on each other for the overall success and
failure. Obtaining the performance relationship between them is interesting since
each part has a separate functions and characteristics.
3.2. Basic assumptions
Some assumptions have been made such that they do not affect the parameters
and accuracy of the model and their outputs.
i) The force throughout the system is the same (i.e. Force on the suspension =
Force on the tire = Force on the ground)
ii) The Nitrogen gas inside the suspension cylinder behaves isentropically (i.e.
there is no change in Temperature) such that PV  = a constant
31
Chapter 3 – Mathematical model for the performance prediction of truck suspension and tire on a
rigid surface
iii) The deflection of a rigid surface is very small when compared to the
deflection of the tire and suspension, hence it can be neglected and the
performance of the rigid surface due to load conditions are ignored
Figure.3.1. General Layout of components considered for the model
3.3. Suspension model parameters and its equations
Most mining trucks are equipped with electronic devices that can display field
data directly on the dashboard, which the operator can easily view for reference
and safety. Vital information management system (VIMS) is one system used in
mining trucks and which provides suspension pressure data at 1Hz. This data was
used as the reference point for this research; and all equations and parameters
were derived and calculated with the help of this data.
32
rigid surface
The force passing throughout the system is constant from suspension to ground
(Equation.3.1) and any variation per second depends on the pressure response of
the suspension. The area of the cylinder is constant and hence only the available
length of the suspension (the stroke) varies per second.
Fs  Ft  Fg  Ps  AC [3.1]
Where, Fs , Ft , and Fg are the Force on the suspension, tire, and the ground (kN),
AC is the area of suspension cylinder (m2), Ps is the suspension pressure indicated

by VIMS per second (kPa).
Figure.3.2. Haul truck positions and nomenclature
Different nomenclatures have been considered for each position of the truck and
used in this research (Figure.3.2). Both the left front (LF) and right front (RF)
position has a single tire and suspension respectively, but the left rear (LR) and
right rear (RR) positions have two tires and a suspension. The force on the LF and
RF tires match the force acting on the LF and RF suspension respectively. But the
truck rear has four tires connected to two suspensions and hence the total force
33
rigid surface
acting on the LR and RR tires is always split into half from the total force acting
on the LR and RR suspensions respectively.
Fs
Frt  [3.2]
2
Where, Frt is the force on a rear tire (kN)
An example of suspension pressure variation derived from the VIMS system for
CAT 797B haul truck during field operation is given in Figure.3.3 for one cycle;
that includes truck loading, travel, and dumping. The data provides some sense
that the load was not equally distributed throughout the body of the truck. LR strut
had the highest suspension pressure at all times. This shows that the truck was
loaded more on one side, i.e. the left. This variation results in damage to the truck
components through load precession, which could cause operator injury,
suspension damage, tire failure, and frame breakage. Any of the above impacts
constitutes a reduction in life of the truck, which in turn affects the reliability,
increases downtime and decreases the production.
30
L F Strut LR Strut
25
RF Strut RR Strut
Pressure (MPa)
20
15
10
0
0 100 200 300 400 500 600 700 800
Time (sec)
Figure.3.3. Pressure variation at Front and Rear struts for one cycle of CAT 797B
34
rigid surface
It is important to understand that the pressure of the cylinder is proportional to the

g-level impacts on the truck. Road conditions and improper load on the truck
causes the suspension try and operate outside the total available stroke length.
This phenomenon is known as “Bottoming out and topping - up” as described by
El-Sayed (2003) and Santos (2007).
Frequent “Bottoming out” damages the inner parts of suspension due to metal to
metal contact. Also the impact force on the tire and the ground is magnified. As
such, it is necessary to verify the variation in available stroke length of the
suspension. Equation [3.3] is the isentropic equation which relates pressure and
the available stroke length [61].
1
P  
S a   u   Su [3.3]
 Ps 
Where, S a is the available stroke length (m), Su is the stroke length at the
unloaded condition (m), Pu is the suspension pressure at the unloaded condition
(kPa), Ps is the suspension pressure per second (kPa),  is the Isentropic gas
constant which is 1.39 for Nitrogen
The performance and the force deformation curve for the suspension depends on
degree of inflation, area of the cylinder and the initial available stroke length.
These parameters were included into this model. Any suspension is based on the
design requirements. The gas inside the suspension acts as a spring and the
hydraulic oil provides the damping force. The ratio of liquid and gas should be
maintained for appropriate performance of both the suspension and overall the
truck. The available stroke length of the suspension provides the fundamental for
the prediction of the varying suspension parameters, bad road conditions, and
improper load distribution.
35
rigid surface
A full analysis completed for the unloaded and loaded conditions showed that the
load on the truck was not distributed in balance and varied from the actual load
distribution data recommended by the truck supplier (Figure.3.4).
Recommended Load Actual Load distribution

distribution
Rear Front
Front loaded
Rear empty
empty 54.88%
loaded 59.21%
43.5%
66.7%
Rear Front Rear

Front empty loaded empty
loaded 56.5% 45.12% 40.79%
33.3%
Figure.3.4. Recommended vs actual load distribution for CAT 797B
From this analysis it was found that the rear end of the truck was loaded less than
the recommended value. But when the truck was in motion, the pressure at the
rear end increased due to motion fluctuations (See Figure.3.3). The available
stroke length at the rear suspension was only 0.152m when compared to the front
suspension which had an available stroke length of 0.343m. From the data, the
maximum pressure experienced by the front strut was 13200 kPa which developed
an available stroke length of 0.088m. Compared to rear struts where the minimum
available stroke length was 0.026m. Given the common status it was decided to
perform the analysis only on the rear suspension and its tires.
A sample field data set from a CAT 797B hauler was analyzed using equation
[3.3] with the corresponding stroke lengths determined. Figure 3.5 provides the
calculated values for the available stroke lengths with time.
36
rigid surface
0.120
Available stroke length (m) 0.100
0.080
0.060
0.040
0.020
0.000
0 100 200 300 400 500 600
Time (sec)
Figure.3.5. Available stroke length data for moving cycle of LR Strut
Figure 3.6 provides a variation of stroke length with the pressure of LR Strut
during moving cycle under load condition. The stroke length varied throughout
the truck moving cycle and the minimum value reached was 0.028m. The change
in the stroke length by pressure for a CAT 797B truck essentially follows an
exponential function.
0.120
Available stroke length (m)
0.100
0.080
0.060
Sa = 35.425Ps-0.716
0.040
0.020
0.000
0 5000 10000 15000 20000 25000 30000
Pressure (kPa)
Figure.3.6. Variation of available stroke length with pressure during movie cycle
for LR Strut
37
rigid surface
0.180
0.160
Available stroke length at
Avaialble stroke length (m) 0.140 unloaded condition
0.120
0.100
0.080
0.060
0.040
0.020
0.000
2020 4148 6976 10746 13574 14140
Pressure (kPa)
Figure.3.7. Variation in available stroke length with pressure during loading cycle
for LR Strut
Figure.3.7 shows that as per the field data, the initial pressure and available stroke
length of the truck at unloaded condition were 2020kPa and 0.152m, where the
pressure and stroke length varied based on the load acting. The truck after being
loaded had an available stroke of only 0.033m before motion. This clearly
indicated an opportunity for bottoming out to occur. Similarly Soni (2009) did a
comparison and found that a truck had only 0.032m to start with. It seems that the
degree of remaining closure for most trucks operated in the field after being
loaded is low, which makes the suspensions susceptible to serious damage. This
does not look good with regard to the overall truck structural performance with
even a small change in the condition of the road affecting the suspension and
other components of the truck.
3.4. Truck tire model parameters and its equations
Given a procedure to evaluate the performance of the suspension, it is necessary

to consider the tire parameters. Truck tire performance is dependent on the loads
applied. Both the suspension and the ground conditions are very important in
38
rigid surface
defining the performance of a tire. If the truck tire is exposed to high g-level, it
will fail sooner resulting in lower availability and utilization of the truck.
Most tires used in the mining industry are radial tires. From the literature review
the following points were noted for pneumatic tires highlighting several important
parameters to be considered in any tire evaluation for input into a model.
i) It is clear from the tires literature that only a few researchers have considered
incorporating pressure into models [45] [47] [51]. It is important that the
shape of the tire should be always maintained through appropriate inflation
pressure. Poor inflation pressure results in improper load distribution and
causes high impact forces on the tire as well as other components of the truck.
Regular checking of inflation pressure is a must since deflection of the tire
varies according to this.
ii) A pneumatic tire contains several components. The tread and sidewall are
made up of different rubber. Young’s modulus is the property which defines
the strength of these components and this is also different for both the tread
and the sidewall even though they perform the same function. It is difficult to
obtain the Young’s modulus value for different tires and hence to avoid any
difficulty in obtaining the appropriate deflection value for a given loads,
stiffness was considered instead of Young’s modulus. Young’s modulus is
more related to the elastic limit of a material but stiffness is related to the
interionic distance of the material. Overall stiffness value gives an indication
of the overall tire performance.
Lin (2007) provided a stiffness equation for a pneumatic tire, which is known as
“Ryne’s regression equation (equation. 3.4)”. This equation was considered in the
model to predict the change in deformation due to inflation pressure [51].
kr  2.68 pi wt  t  33.1 [3.4]

39
rigid surface
Where, kr , pi , wt , and t are the radial stiffness (kN/m), inflation pressure (kPa),
width (m), and diameter of the tire (m)
Rubber acts as an elastic component similar to an elastic spring and the general
stiffness equation for any spring type material depends on the force applied and
the experienced deformation (Equation.3.5 and 3.6) [24].
Ft
kr  [3.5]
 tr
Ft
 tr  [3.6]
kr
Where,  tr is the deformation of the tire on a rigid surface (m), Ft is the force on
the tire (kN), kr is the radial stiffness of the tire (kN/m)
3.5. Mathematical model equations
To predict the overall relationship between the suspension and the tire, equations
from sections.3.3 and 3.4 were used to derive the final equations. Using equation
(3.1) and (3.4) in equation (3.6) provides the deformation of the tire with respect
to pressure of the suspension and inflation pressure of the tire (Equation.3.7)
Ps  AC
 tr  [3.7]
2  68 pi wt  t  33.1
Where, Ps is the suspension pressure (kPa), AC is the area of suspension cylinder
(m2), pi , wt , and t are the inflation pressure (kPa), width (m), and diameter of
the tire (m),  tr is the deformation of tire on a rigid surface (m)
40
rigid surface
The objective here is to understand the relationship between the available stroke
length and deformation of the tire for variable inflation pressure on a rigid
surface. Hence the final equation providing such a relationship was derived
considering equation (3.3) and equation (3.7).
Pu  AC
 tr , f  
[3.8]
 Sa 

   2.68 pi wt  t  33 1
 Su 

Pu  AC
 tr ,r  
[3.9]
S 

2  a   2.68 pi wt  t  33 1
 Su 

Where, S a is the available stroke length (m), Su is the stroke length at unloaded
condition (m),  tr , f and  tr ,r are the deformation of a front and rear tire on rigid
surface (m), Pu is the suspension pressure at unloaded condition (kPa)
It should be noted that the equation (3.8) is applicable for the front tires and
equation (3.9) for rear tires on a rigid surface. The denominator for the rear end
equation is multiplied by 2 since there are two tires for one suspension and the
pressure acting on the rear suspension is equally divided and transferred to the
two tires.
The deformation of a tire as per the derived equation (Equation.3.8 and 3.9)
depend only on one variable, i.e. the available stroke length. All other parameters
are constant and this relationship gives the performance of the tire with respect to
the suspension performance. Conversely, the deformation of the tire is inversely
proportional to the available stroke length since as the available stroke length
decreases, the deformation of the tire increases. In this scenario, under field
41
rigid surface
conditions, the deformation of the tire and available stroke length react opposite to
each other.
As the tire deforms as per the loading condition, it creates a footprint on the
ground. The footprint area of the tire is an important factor that should be
considered in any model since the contact between the tire and the ground gives
an idea of the tire load withstanding capability and inflation pressure accuracy. It
also allows for the prediction of the total surface of the ground exposed to the
applied force, which helps in calculating the stresses experienced by the ground.
As seen in the literature review, Sharma (2009) compared various tire sizes to
obtain an equation for the footprint area of the tire (equation. 3.10), which is used
in the model. It basically depends on the deformation of the tire and its diameter
[60].
Atr  1.35 trt [3.10]
Where, Atr is the footprint area of tire on rigid surface (m2),  tr is the deformation
of tire on a rigid surface (m), t is the diameter of tire (m)
3.6. Mathematical model equation validation with field and laboratory

results
It is always better to compare the derived equations with field and laboratory
results since it can prove the accuracy of the equations determined for various
scenarios. Sharma (2009) performed laboratory test on 0.368m diameter tire and
obtained the field tire deformation and footprint area for 1g load (1040kN) of a
full size truck (CAT 797B). Bolster (2007) performed laboratory tests on 2.85m
diameter truck tire and obtained deformation characteristics for various loads and
inflation pressures. To check the validity, accuracy and applicability of the
42
rigid surface
derived equation under full scale loading for different truck sizes, comparisons
were made with these field and laboratory results obtained by the previous
researchers.
3.6.1. Comparison with laboratory results
Initial comparisons were made with tires used by Sharma (2009) and Bolster
(2007). They performed laboratory tests and obtained results for various loads and
inflation pressures. The dimension of the 0.368m diameter tire was obtained from
a data sheet given by Sharma (2009). Bolster (2007) provided the dimensions of
the tires used as 30R0051 with an external diameter of 2.85m and width of
0.762m (Bolster, 2007). The first parameter (30) in the designation of the tire
model gives the width of the tire in inches. So, converting that to the required unit
gives the exact width of the tire.
Both Sharma and Bolster used a ram for the application of force to the tires tested.
As such, it was difficult to find out the exact parameters like the isentropic
constant of gas used (a constant value = 1.39), the initial and final pressure, and
the initial and final available stroke length. Hence in this study the force value
was used instead of pressure, stroke length and area of the cylinder from tire
deformation equation (3.8). Equation (3.9) was used for loads applied to two tires
and equation (3.8) was used here since both researchers performed laboratory
tests on a single tire and the force applied is then directly transferred to that tire.
43
rigid surface
0.040
Tire deformation (m) 0.035
0.030
0.025
138 kPa
0.020 165 kPa
0.015 193 kPa
0.010 207 kPa
0.005
0.000
1.5 2 2.5 3 3.5 4
Load (kN)
Figure 3.8.Load vs Deformation for 0.368m diameter tire
0.020
0.018
Tire footprint area (m2)
0.016
0.014
0.012 138 kPa
0.010 165 kPa
0.008
193 kPa
0.006
207 kPa
0.004
0.002
0.000
1.5 2 2.5 3 3.5 4
Load (kN)
Figure 3.9.Load vs Footprint area for 0.368m diameter tire
Figure 3.8 and 3.9 provides the data for the variation of deformation and footprint
area with load for a 0.368m diameter tire. The increase in both parameters is
linear since the only variation when using equation (3.7) is the force and the
stiffness is a constant for a given pressure. The comparisons between the model
and laboratory values are provided in Table.3.1.
44
rigid surface
Inflation Tire deformation (m) Tire footprint area (m2)

Load
pressure Laboratory Laboratory
(kN) Model Model
(kPa) results results
2.0 0.018 0.019 0.009 0.012
138 2.5 0.022 0.023 0.011 0.014
(20psi) 3.0 0.026 0.028 0.013 0.016
3.5 0.031 0.033 0.015 0.018
2.0 0.015 0.017 0.008 0.010
165 2.5 0.019 0.021 0.010 0.013
(24psi) 3.0 0.023 0.025 0.011 0.014
3.5 0.027 0.029 0.013 0.016
Table.3.1. Model output and laboratory value comparison for 0.368m diameter
tire
There are deviations in tire deformation and footprint area comparison. This
deviation is due to the values compared are low in magnitude, which resulted in
some measurement error. A better understanding could have been made if the
initial available stroke length of the suspension was known.
0.300 1.5 g
1.4 g
1.3 g
0.250 1.2 g
Tire deformation (m)
1g 1.1 g
Loaded
0.200
0.150 552 kPa

621 kPa
0.100 689 kPa
0.050
0.000
421.4 457.14 500 535.71 578.57 614.29
Load (kN)
Figure 3.10.Load vs Deformation for 2.85m diameter tire
45
rigid surface
1.200
1.5 g
1.4 g
Tire footprint area (m2) 1.3 g
1.000 1.2 g
1g 1.1 g
Loaded
0.800
552 kPa
0.600
621 kPa
0.400 689 kPa
0.200
0.000
421.4 457.14 500 535.71 578.57 614.29
Load (kN)
Figure 3.11.Load vs Footprint area for 2.85m diameter tire
Figures 3.10 and 3.11 provide data for the variation of deformation and footprint
area with load for the 2.85m diameter tire. The increase in both parameters is
effectively linear since the only variation when using equation (3.7) is the force,
and the stiffness is constant for a given pressure. The comparisons between the
model and laboratory values are provided in Table.3.2.
Inflation Tire deformation (m) Tire footprint area (m2)

Load
pressure Laboratory Laboratory
(kN) Model Model
(kPa) results results
182.14 0.084 0.111 0.322 0.502
421.43 0.193 0.200 0.744 0.796
552 457.14 0.210 0.211 0.807 0.849
(80psi) 500 0.230 0.221 0.883 0.889
535.71 0.246 0.234 0.946 0.947
578.57 0.266 0.245 1.022 0.964
Table.3.2. Model output and laboratory value comparison for 2.85m diameter tire
46
rigid surface
Again there is a minor deviation between the results of the mathematical model
and laboratory values. The deviation in comparison to tire deformation might be
due to measurement error due to the low measured value. In the case of the tire
footprint area, the deviation is also low. But when the standard footprint area
equation (3.10) is used for the laboratory values, the footprint area does not match
with the corresponding deformation values. This may explain the deviation in
footprint area between laboratory and model results.
3.6.2. Comparison with Field value (CAT 797B)
Sharma (2009) performed a field test and found the values for the deformation of
the CAT 797B truck with 55/80R63 truck tire size. He found that the deformation
of the rear tire at 1g load (1040kN) and 600 kPa inflation pressure was 0.27m
resulting in 1.29m2 of the contact footprint area (Sharma, 2009).
0.450
1.4g 1.5g
0.400 0.416
1.3g
1.2g 0.388
0.350 1.1g 0.360

1.0g 0.333
0.300 0.305
0.250 0.277
0.200
0.150
0.100
0.050
0.000
1040 1144 1248 1352 1456 1560
Load (kN)
Figure 3.12. Load vs Tire deformation for 55/80R63 tire at 600 kPa
Comparison with laboratory results showed that the mathematical model equation
is valid (Equation.3.7 to 3.10). Further validation by comparison with field values
47
rigid surface
would enhance their accuracy. Equations (3.9) and (3.10) were used for this
comparison. Equation (3.9) was used since the comparison is made for a rear tire,
where the pressure from the suspension is divided equally between the two tires.
2.50
1.5g
1.3g 1.4g
2.16
2.00 1.2g
1.1g 2.01
1g 1.87
1.72
1.50 1.58
1.44
1.00
0.50
0.00
1040 1144 1248 1352 1456 1560
Load (kN)
Figure 3.13. Load vs Tire footprint area for 55/80R63 tire at 600 kPa
Both deformation and the footprint area for the tire were obtained. Table.3.4
provides a comparison of values obtained from the field; the predicted value by
Sharma (2009), and the model output for 1-g load (1040kN).
Tire diameter Inflation Tire Tire

Description (m) and pressure deformation footprint
designation (kPa) (m) area (m2)
3.84
Field value 600 0.270 1.30
(55/80R63)
Predicted value 3.84
600 0.270 1.40
(Sharma, 2009) (55/80R63)
3.84
Model output 600 0.277 1.44
(55/80R63)
Table.3.3. Model output and field value comparison for 1-g load of CAT 797B
48
rigid surface
It can be seen that the values from the mathematical model equations matches
with the field values. It was not possible to do a comparison for the values above
1g due to lack of data. Further validation is required for the derived equations
above this load. Hence from the comparison of both the field and laboratory
results it can be said that the mathematical model equations appear valid and may
be used to predict results close to actual field data, extended to full scale truck
size. Also it is clear that the mathematical equations model can predict
deformation and footprint for any size tire, for small to larger loads.
3.7. Understanding the relationship between stroke length of suspension and

deformation of tire for full size truck
All the above comparisons were made considering the action of force on a tire
without considering the pressure and available stroke length of the suspension.
The results obtained in all the comparisons provided a linear relationship. But in
order to have a better understanding, the available stroke length of the suspension
was also included here for the comparison (i.e. using equation. 3.9).
0.350
0.300
0.250
0.200
δtr = 0.5034e-24.68Sa
R² = 0.9594 Available Stroke
0.150
length at Unloaded
0.100 condition
0.050
0.000
0.000 0.050 0.100 0.150
Figure 3.14.Available stroke length vs tire deformation for CAT 797B at 600 kPa
49
rigid surface
There are no field or laboratory results for the comparison between suspension
and tire performance for a full sized ultra-class truck. Hence it was decided to
study the relationship for a CAT 797B truck since the suspension pressure data
and characteristics of both the suspension and the tire were known. Equation (3.9)
was used for this comparison. The stroke length variation and tire deformation
follows an exponential form (Figure.3.14).
This shows that as the available stroke length increases or decreases, the
deformation does not increase or decrease linearly, since more than one parameter
is involved to affect the relationship. There are many parameters and each
parameter is important to provide the overall relationship between the available
stroke length and the deformation of the tire. Those parameters are the ones given
in equations (3.8) and (3.9). Similarly, these equations can be used for different
types of trucks used in the mining industry, where some of the parameters of the
suspension and the tire are known.
50
CHAPTER 4
MATHEMATICAL MODEL FOR PERFORMANCE PREDICTION OF
TRUCK SUSPENSION, TIRE, AND THE OIL SAND
4.1. Introduction and objective
Oil sand exhibits a total different behavior from a rigid surface. From the
literature review it is clear that sinkage and rutting are the main characteristics
that occur in oil sand during truck loading and movement. By amount of bitumen
content present in the oil sand, the truck tire will sink into the ground
proportionally. Also the oil sand due to its poor load withstanding capability
deflects under the action of truck loads. A method to discern the sinkage of oil
sand with 11% bitumen content and the corresponding deformation of a tire on oil
sand under varying load conditions has been addressed in this chapter.
In chapter 3 the mathematical model equations and results for truck component’s
performance on rigid surface were provided. This chapter deals with the model
equations and their results for the performance prediction relative to oil sand and
the truck component’s on oil sand. Since laboratory tests are already done by
Sharma (2009) on oil sand as deformation for various loads, it was decided not to
perform the tests again as the same results would be obtained.
The main focus of this chapter was to provide a method to discern the relationship
and behavior of a composite truck - ground model including the truck suspension,
tire and the oil sand. Also prediction using the mathematical model without the
influence of laboratory test results is a main goal of this research. The equations
obtained by Sharma (2009) and pressure stiffness predicted by Sharif-Abadi
(2006) for 11% bitumen oil sand are considered in this model and modified as per
the objective of this research.
51
Chapter 4 – Mathematical model for the performance prediction of truck suspension, tire, and the
oil sand
4.2. Suspension model equations
The suspension parameter accounting for the ground condition is dependent only
on reaction force due to the ground. Characteristics are not dependant on the
bitumen content, moisture content of oil sand or atmospheric pressure and
temperature. Hence the equation for the change in stroke length of the suspension
is the same as used in chapter 3.
4.3. Tire model equations
Since the tire comes into contact with the oil sand which has lower stiffness than a
rigid surface, the equations related to the tire will change from that on the rigid
surface.
Most of the tests performed to predict the deformation of oil sand was conducted
with a plate. But in field, only tire comes into contact and there is a difference
between the reaction caused on oil sand due to tire and normal cylindrical plate.
Sharma (2009) performed cyclic tests with a 0.368m diameter tire on oil sand of
11% bitumen and room temperature. After comparing with several data from the
laboratory results he found out the relationship for the deformation of the tire on
oil sand (Equation.4.1) (Sharma, 2009). He was the first to perform the test with a
tire on oil sand and hence it was decided to use equation (4.1) and modify it to
obtain the relationship exactly as per the field operation.
 1

1  2 4 Ft  2 
 ts    os   
CE 
[4.1]
2  
os
 
52
oil sand
Where,  ts is the deformation of tire on oil sand (m),  os is the deformation of oil
sand (m), Ft is the force on the tire (MPa), E is the Young’s Modulus of the tire
(MPa), C is the oil sand constant
As discussed previously since Young’s modulus of tires varies for different

components, it was decided to use the equations in chapter 3 to predict it. As
Young’s modulus is a constant value, it does not vary with load. Also since
Sharma (2009) obtained the deformation and footprint area for a 1g load for
55/80R63 tires from the field, it was possible for him to evaluate a stiffness
equivalent for Young’s modulus. In the mathematical model here that problem is
eliminated, as measurement of a field value is not necessary.
The procedure for determining Young’s modulus for a tire is as follows. First the
deformation and footprint area of a tire on a rigid surface is calculated for a 1g
load using equation (3.2) and equation (3.3). Finally the obtained values are used
in equation (4.4) to predict a Young’s modulus value for the tire. Since the tire
effectively acts as an elastic spring, the Hooke equation is applicable and is used
here (Sharma, 2009). A similar method can be used in predicting the Young’s
modulus value for different tires with the condition that the inflation pressure and
dimensions of the tire are known.

E [4.2]

 Ft 
 
E  Atr 
[4.3]
  tr 
 
 t 
Ft  t
E [4.4]
 tr  Atr
53
oil sand
Where,  and  are the stress (MPa) and the strain, Ft and t are the force on the
tire (kN) and diameter of the tire (m),  tr and Atr are the deformation (m) and
footprint area of tire on rigid surface (m2)
Replacing equation (4.4) in equation (4.1) provides a modified equation for

prediction of deformation of a tire on oil sand.
 1 2

1  4 Ft 
 ts    os2     os  [4.5]
2  CEg1 
 
 
Where, Eg1 is the Young’s modulus of the tire predicted at 1g load (MPa)
4.4. Oil sand model equations
An array of models predicts the dynamic modulus and resilient modulus of oil
sand. But none of the models from the literature prove to predict field values
when a tire acts on oil sand, and each model is applicable only for a particular
condition, with each model having their own constants that could be determined
only with the help of performing experiments in the laboratory.
Hence with all these difficulties, to be able to predict the deformation value of oil
sand with just the mathematical equations, it was resolved to use the pressure
stiffness developed by Sharif – Abadi (2006) and the equations given by Sharma
(2009) modified it as per the research objective here. Sharif – Abadi (2006)
performed instantaneous loading cyclic tests on oil sand at 200, 400, 600, and 800
kPa.
From the test results he predicted that regardless of the duration of loading and
relaxation, the pressure stiffness value converged to 5.5kPa/mm. Using the
54
oil sand
pressure stiffness, the problem of varying footprint area of the tire with cyclic
loading is eliminated. Also the deformation of the oil sand depends on the total
area of oil sand exposed to the load, which cannot be predicted with the current
available mathematical equations. Hence pressure stiffness is a parameter that can
used to effectively ignore the effects of varying footprint area and in turn
predicting the deformation of oil sand.
The number of loading cycles ranged from 15 – 20 cycles for every full load.
With just few loading cycles, the oil sand does not reach a maximum stiffness
value, hence the pressure stiffness of 5.5kPa/mm is considered as the standard
value in this analysis. Pressure stiffness is the ratio of total pressure to the total
deformation (Equation.4.6) (Sharif – Abadi, 2006).
Pos
kp  [4.6]
 os
 Fg 
 
kp   Aos 

Fg
[4.7]
 os Aos   os
Where, k p is the oil sand pressure stiffness (kPa/m), Fg is the force on the oil sand
ground (kN), Aos and  os are the footprint area (m2) and deformation of oil sand
(m)
It should be noted that since the tire sinks into the oil sand, the total area of oil
sand under load depends on both the deformation of oil sand and deformation of
the tire which inherently changes the contact footprint area. Equation (4.8) has an
oil sand deformation term and a constant C. This constant clearly gives the
differentiation between the deformation and footprint area of the rigid and oil
sand grounds. The value of C after performing tests on oil sand for 80 cycles was
55
oil sand
obtained by Sharma (2009) and it varies from 1.66 to 1.11 from the first cycle to
the 80th cycle (Sharma, 2009).
Aos  C  os   ts   t [4.8]
Where, Aos is the area of oil sand (m2),  os is the deformation of oil sand (m),  ts is
the deformation of the tire on oil sand (m), t is the diameter of the tire (m)
The modified equation is obtained by substituting equations (4.5) and (4.8) in
equation (4.7). Since the force on the ground is equal to the force on the tire, the
term Fg is replaced by Ft in the final equation.
Ft
kp  [4.9]
C   os   ts   t   os
Now bringing the deformation parameter on one side, the equation (4.9) becomes,
Ft
 os   ts    os  [4.10]
k p  C  t
Substituting equation (4.1) in equation (4.10)
   1
 
      2  t  2        
1 4 F Ft
 os
2  os
CE  os
  os
k p  C  t
[4.11]
     
Expanding equation (4.11),
Ft  E
 os  [4.12]
 k  C     k p  C  E  t 
2
p t
2
56
oil sand
Where,  os is the deformation of oil sand (m), k p is the pressure stiffness of oil
sand (kPa/m), Ft is the force on the tire (kN), E is the Young’s modulus of the
tire (kPa), t is the diameter of the tire (m), C is the oil sand model constant
The pressure stiffness of oil sand, and Young’s modulus and diameter of the tire
are a constant. The variables related to the deformation of oil sand are the force on
the tire and the oil sand model constant. Hence equation (4.12) is the modified
equation that can predict the deformation of oil sand with bitumen content of 11%
under cyclic loading activity of the truck.
4.5. Oil sand mathematical model equations output
The process in obtaining the mathematical model equations for the oil sand is
totally different from the process used for the rigid surface. In the same way the
procedure in obtaining the output for oil sand and the truck components on oil
sand using the model equations are also different.
i) Calculate the Young’s modulus of the tire on a rigid surface for 1g load
(Equation.4.4)
ii) Substitute the Young’s modulus value in equation (4.12) and find out the
deformation of oil sand for every loading cycles
iii) Substitute the obtained value from step 1 and step 2 into equation (4.5) and
calculate the deformation of tire on oil sand for every loading cycle
4.5.1. Performance Output for CAT797B truck on oil sand
CAT 797B pressure data, the tire parameters, and the pressure stiffness value
were used in the derived equations to obtain the deformation output for the
suspension, tire, and the oil sand. Three different loading cycle methods were
57
oil sand
considered for the analysis to predict the response of the derived equations for
various load conditions.
A first loading cycle consists of applying a constant 1g load. The second loading
cycles consist of applying a constant increasing load of 52kN. Finally the last
loading cycle consists of applying a load as per the field data where the load
varies with time.
0.160
Constant 1g load
0.140
Oil sand deformation (m)
Constant increasing load

0.120
Actual field load
0.100
0.080
0.060
0.040
0.020
0.000
0 5 10 15 20 25
Number of cycles
Figure.4.1. Oil sand deformation with different loading cycle
It can be seen that the oil sand cyclically deformed differently for each of the
methods (Figure.4.1). For a constant 1g applied load, the oil sand deformed
gradually and for an increasing load (52kN) the oil sand deformed rapidly. But the
third cyclic loading method allowed the oil sand to deform at a faster pace than
the other two methods. As the load increase was higher by number of cycles, this
allowed the oil sand to compact further with fewer numbers of cycles.
58
oil sand
0.300 Constant 1g load

Constant increasing load
0.250
Actual field load
0.200
0.150
0.100
0.050
0.000
0 5 10 15 20 25
Number of cylces
Figure.4.2. Tire deformation with different loading cycle
Tire also deformed in the same manner as oil sand (Figure.4.2). The rate of
deformation of both oil sand and tire decreased gradually with increasing number
of cycles. This would follow for any truck acting on oil sand as the ground would
become stiffer and increase its ability to withstand load with gradually decreasing
deformation.
4.5.2. Performance Output for CAT797B truck on oil sand as per field
condition
As such, from chapter 3 it was evident that the tire rear was loaded more than the
front (figure.3.4). To understand the effect due to this high loading at the rear,
comparisons were done between the LR and RR position of the same truck to
evaluate numerous parameters like total deformation (tire and oil sand), tire
deformation, oil sand deformation, footprint area, and the available stroke length
(for suspension). Figure.4.3 compares tire deformation for the LR vs RR tire sets
during a loading cycle.
59
oil sand
0.250
0.200
0.150
LR Tire
0.100 RR Tire
0.050
0.000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Number of cycles
Figure.4.3. Tire Deformation on oil sand with number of cycles for LR and RR
Tire
The deviation in the total deformation (oil sand + tire) with cyclic loading was
also studied. Figure.4.3 clearly shows that the total deformation at the LR position
is higher than the RR position. The deformation of oil sand and the tire at LR
position was higher than at RR position because of varying load conditions and
ground pressure, and this made the total deformation also to be more on LR
position.
0.350
LR δt = 0.0125n + 0.1326
Total deformation (m)
0.300
RR R² = 0.9735
0.250
0.200
δt = 0.0059n + 0.1367
0.150
R² = 0.9837
0.100
0.050
0.000
0 2 4 6 8 10 12 14
Number of cycles
Figure.4.4. Comparison of total (tire and oil sand) deformation with increasing
number of cycles
60
oil sand
Clearly the LR strut had larger forces acting due to a higher payload and this
caused the difference in the total deformation for both sides. In an ideal scenario
this should not be the case, since the load should be equally distributed. This
shows that the truck was not loaded properly and while travelling it created an
unequal deformed running surface. The problem generated here is for the same
truck and also for the next truck. Even if the next truck has a balanced payload
distribution, because of the evenness that follows along the same road in the
running surface causes the balanced truck to move reflecting the ground
deformation.
4.5.3. Understanding the relationship between stroke length of suspension

and deformation of tire and oil sand for full size truck
A study was made to find the relationship between the available stroke length and
the deformation of the tire, oil sand and the overall total deformation (tire and oil
sand) (Figure.4.5 and Figure.4.6) using the mathematical model equations.
0.450
LR Tire
0.400
Oil sand
0.350
Total
δ = -0.123ln(Sa) - 0.1704
Deformation (m)
0.300 R² = 0.9793
0.250 δ = -0.188ln(Sa) - 0.2612
R² = 0.9793
0.200
0.150
0.100
δ = -0.065ln(Sa) - 0.0908
0.050
R² = 0.9793
0.000
0.000 0.020 0.040 0.060 0.080 0.100 0.120 0.140
oil sand and total deformation at LR position
61
oil sand
Since the LR strut had the highest pressure, the total available stroke length at the
end of the loading cycle was only 0.032m when compared to RR value of 0.062m
as seen in figures (4.5) and (4.6). This could also lead to metal to metal contact in
the suspension resulting in suspension failure, LR tire damage and were an
indicator of unbalanced load. At the end of the loading cycle, the oil sand under
RR position deformed to 0.087m. But under LR position it deformed to 0.141m,
which is 1.62 times the deformation under RR position. This would create an
undulated ground condition causing the truck to sink more towards the LR. The
final deformation of the LR tire (0.265m) is 1.61 times the deformation of the RR
tire (0.164m) and this difference is very high.
0.300
0.250
δ = -0.14ln(Sa) - 0.1433
R² = 0.9906
Deformation (m)
0.200
RR Tire
0.150 Oil sand
δ = -0.091ln(Sa) - 0.0935
R² = 0.9906 Total
0.100
δ = -0.049ln(Sa) - 0.0498
0.050
R² = 0.9906
0.000
0.000 0.050 0.100 0.150

oil sand and total deformation at RR position
4.5.4. Tire footprint area and ground stress comparison for CAT 797B truck
Footprint area of a tire and the ground stress variation for various loading
conditions of LR and RR positions were compared here (Figure.4.7 and 4.8). The
stress followed close to a linear pattern. As the number of cycles increased, the
stress on the ground also increased. The stress at the LR was more than that at RR
62
oil sand
because of higher load. The footprint area of oil sand at the LR is also more than
at the RR.
At the end of the loading cycle, the LR tire has footprint area of 2.31m 2 when
compared to the RR, which has only footprint area of 1.49m2. The higher load
factor is the reason for more stress and contact area at the LR.
700
LR position
600
RR position
Ground stress (kPa)
500
400
300
200
100
0
0 2 4 6 8 10 12 14
Number of cycles
Figure.4.7. Ground stress distribution at the rear position of the loading area
2.000
LR Tire
1.600
RR Tire
1.200
0.800
0.400
0.000
0 2 4 6 8 10 12 14
Number of cycles
Figure.4.8. Tire footprint area with increasing number of cycles
63
oil sand
It could be seen that parameters such as deformation of tire and oil sand, footprint
area of oil sand, and ground stress distribution can be found out using the derived
equations but the most important parameter is the sinkage (deformation of oil
sand under loading cycle) of the truck on oil sand deformable ground. This is very
important because as the truck is loaded, it gives an idea of how the oil sand
deforms and in turn how it affects the stability of the truck. No comparison could
be made for the final output because there are no field values available, but further
comparisons using the model equations for various trucks and for lower and
higher g loads are required.
4.5.5 Performance prediction for 2.85m and 0.368m diameter tires
Tire and oil sand deformation was evaluated for the tires used by Bolster (2007)
and Sharma (2009), but without including an available stroke length for
comparison, since the pressure data and suspension parameter were not known.
The values of load and inflation pressure were considered similar for the tests
performed by Bolster (2007) and Sharma (2009). Since the tires were different in
size for each researcher, the load and the inflation pressures varied.
0.120
0.100
0.080
552 kPa
0.060
628 kPa
0.040 689 kPa
0.020
0.000
421.4 457.14 500 535.71 578.57 614.29
Load (kN)
Figure.4.9. Oil sand deformation with load for 2.85m diameter tire
64
oil sand
Figures (4.9) and (4.10) provide the oil sand and tire deformation for a 2.85m
diameter tire. Analysis was done for three tire inflation pressures (552, 628, and
689kPa). As expected the oil sand deformation was higher for tire with higher
inflation pressure. This is because a stiffer tire creates a smaller footprint area and
resulting in a higher stress on the ground in turn yielding a higher oil sand
deformation. The opposite applies for tires of lower stiffness which create a larger
footprint area and lower stress on the ground a lower oil sand deformation. Tire
deformation is higher for softer tires and lower for stiffer tires. This is true as a
tire with higher stiffness can better withstand load.
0.200
0.180
0.160
0.140
0.120
552 kPa
0.100
628 kPa
0.080
689 kPa
0.060
0.040
0.020
0.000
421.4 457.14 500 535.71 578.57 614.29
Load (kN)
Figure.4.10. Tire deformation on oil sand with load for 2.85m diameter tire
0.035
0.030
0.025
138 kPa
0.020
165 kPa
0.015 193 kPa
0.010 207 kPa
0.005
0.000
0 1 2 3 4 5
Load (kN)
Figure.4.11. Oil sand deformation with load for 0.368m diameter tire
65
oil sand
0.020
0.018
0.016
0.014
0.012
138 kPa
0.010
0.008 165 kPa
0.006 193 kPa

0.004 207 kPa
0.002
0.000
1.5 2 2.5 3 3.5 4 4.5
Load (kN)
Figure.4.12. Tire deformation on oil sand with load for 0.368m diameter tire
Figures (4.11) and (4.12) illustrate the oil sand and tire deformation by varying
load for a 0.368m diameter tire. The deformation behavior for this tire on oil sand
was the same as for the 2.85m diameter tire.
66
CHAPTER 5
RHEOLOGICAL MODEL FOR PERFORMANCE PREDICTION OF
TRUCK SUSPENSION AND TIRE ON A RIGID SURFACE
5.1. Introduction and Objective
Rheological models are similar to mathematical models but provide a different

mode of approaching a problem in a more theoretical manner. The mathematics
involved here is not difficult and as such makes the rheological model approach
an easy tool to understand. The main aim of this model was to predict the
performance of the truck components and the ground through simple viscoelastic
models and to convey an alternative and mode of understanding the overall
relationship.
Based on properties behavior, a model has been assigned to each of the

components in the system like suspension, tire and ground. This chapter consists
of only the model for suspension and tire since the ground is essentially
considered rigid, with no deflection. Although the responses that can be predicted
using these models are limited, the main advantage comes in predicting the
deformation behavior with an acceptable approximation to actual behaviour. This
field of study provides results close to field values maintaining the physics of the
issue. Connecting the individual parts and obtaining the performance relationship
between them is interesting since each part has separate functions and
characteristics.
5.2. Basic assumptions
i) The deformation of the rigid surface is lower compared to a tire and

suspension, hence no model has been assigned for a rigid surface
ii) The force throughout the system is same
67
Chapter 5 – Rheological model for the performance prediction of truck suspension and tire on a
rigid surface
5.3. Model layout and equations
Establishing the relationship between the stress and deformation is the key task in
rheology. The models described in the literature review are used matched to the
corresponding components of the truck based on behavior. Only a few elements
are included in the overall model. This is because the mathematics become more
complex with more elements, regardless with just few elements the prediction is
good.
5.3.1. Rheological model for truck suspension and tire on a rigid surface
Both the suspension and tire model combined together give a rheological model
for truck suspension and tire on a rigid surface (Figure.5.1). The suspension
model is viscoelastic because of the presence of both viscous and elastic element.
But the tire is merely an elastic element. The force applied on each model is used
to predict the total and individual component strain values.
Figure.5.1. Rheological model for truck suspension and tire on a rigid surface
68
rigid surface
The suspension exhibits no permanent deformation as spring stiffness is higher

than the stiffness of the damper (viscosity). Due to this the spring is able to pull
the hydraulic oil back from the annulus in retraction mode into the damper with
no permanent deformation. The total strain of the combined model is a
combination of viscoelastic and elastic model strains (Equation.5.1). Even though
the force is same on each model, the stress developed is not the same because of
different contact areas. This model can be used to predict the total and individual
strain for various loading conditions. Since the system is a continuous moving
system the effect of creep and relaxation are not a major concern and as such are
not considered for prediction here.
 rt   st   tt [5.1]
Where,  rt is the total strain on a rigid surface,  st is total strain of the suspension,
 tt is total strain of the tire
5.3.2. Truck Suspension model and related equations
For a rheological model, the two main truck suspension elements, the gas spring
and the hydraulic oil damper are considered. The shape of the suspension is not
important for the model. Based on the corresponding behavior, Kelvin – Voigt
model has been used for the suspension (Figure.5.2).
Figure.5.2. Rheological model for the truck suspension

69
rigid surface
The spring in the model acts for the gas in the suspension. Meanwhile the damper
in the model acts for the hydraulic oil in the suspension. Both these constitute to
the overall performance of the suspension.
In the suspension, the gas takes the initial load and total strain occurs
(Equation.5.2) with respect to the time due to the damping force provided by the
damper. The strain value increases gradually and when the maximum strain is
reached, it remains as a constant for the remaining time until the stress on the
suspension changes. The total stress is the sum of stress acting on the spring and
the damper (Equation.5.3) [10].
 st   ss   sd [5.2]
 st   ss   sd [5.3]
Where,  st is the total strain of the suspension,  ss is the suspension spring strain,
 sd is the suspension damper strain,  st is the total stress on the suspension

(MPa),  ss is the suspension spring stress (MPa),  sd is the suspension damper
stress (MPa)
An important point to be noted is that the strain of the suspension spring is

directly proportional to the Elastic modulus (Equation.5.4) [58].
Ess
 ss  [5.4]
 ss
Where, Ess ,  ss , and  ss are the Young’s modulus (MPa), strain, and stress of the
suspension spring (MPa)
70
rigid surface
The gas shows a different behavior than the spring, hence the Elastic modulus
property is not applicable for the gas in the suspension. Even considering the
stiffness of the spring is not an appropriate method for rheological models.
The stiffness is not a constant and it varies with the available stroke length. The
lesser the available stroke length, the more the stiffness and more the available
stroke length, the lesser the stiffness. But this is not the case of spring used in the
model, which usually has a constant stiffness and Elastic modulus. Hence it was
decided to use the general isentropic gas law to predict the strain value of the
suspension.
For an isentropic suspension with constant temperature, the product of volume

and pressure at any given situation is a constant (Equation. 5.5) [57].
PV   C [5.5]
Where, P , V ,  are the pressure (kPa), volume (m3), and Isentropic constant of
the gas
The process is continuous and there are n pressure variations in the suspension
due to load changes (Equation.5.6). Also the area of the suspension cylinder is a
constant, which gives the variation of pressure with respect to only the available
stroke length of the suspension (Equation.5.7) [57].
  
u u  PV
PV s s  PV
n n [5.6]
 
 Pu   Vs   Pu   Vn 
     and      [5.7]
 Ps   Vu   Pn   Vu 
 
 Pu   AC  Sa   Sa 
     [5.8]
 Ps   AC  Su   Su 
71
rigid surface
Re- writing equation (5.6)
1
 Sa   Pu  
   [5.9]
 Su   Ps 
Where, Pu is the suspension pressure at unloaded condition (kPa), Ps is the
instantaneous pressure (kPa), Pn is the pressure at nth cycle (kPa), Vu is the volume
of gas at unloaded condition (m3), Vs is the volume of gas at loaded condition
(m3), Vn is the volume of gas at nth cycle (m3), Ac is the area of the suspension
cylinder (m2), S a is the available stroke length (m), Su is the available stroke
length at unloaded condition (m),  is the Isentropic constant
The pressure in the suspension at the unloaded condition is kept as the reference
point to derive a ratio of available stroke length to initial stroke length
(Equation.5.10) equivalent to Equation (5.9). The strain of the suspension is a
function of the reference point pressure, isentropic constant, and the current
pressure indicated by the VIMS (Figure.5.3).
Figure.5.3. Relationship chart for strain value of the suspension
72
rigid surface
The focus of this model was to determine the strain value. Equation (5.9) provides
the ratio value but the difference of the obtained value with unity gives the strain
value; hence equation (5.10) is used to predict the strain value for varying load
conditions.
1 1
Sa     P 
=  ss  f  ,    1   u   1   u  [5.10]
Su  t   Ps 
Where, S a is the available stroke length (m), Su is the stroke length at unloaded
condition (m),  ss is the effective suspension spring strain,  u is the stress at
unloaded condition (MPa),  t is the stress at time t (MPa), Pu is the suspension
pressure at unloaded condition (MPa), and Ps is the instantaneous suspension

pressure (MPa)
The stress in equation (5.9) is replaced by pressure since the pressure and the
stress on the suspension are the same. The instantaneous pressure and the pressure
at the unloaded condition are used to evaluate the total strain value of the
suspension.
 sd
 sd  f  ,   [5.11]
sd
The strain rate of the damper is a function of both the stress and the viscosity of
the hydraulic oil (Equation.5.11) [58]. To obtain the strain experienced by the
damper, equation (5.11) was integrated. The truck suspension stress values vary
with time and the variation is recorded per second. The applied stress per second
is considered here for the prediction of damper strain (Equation.5.15).
73
rigid surface
d sd  sd
 sd   [5.12]
dt sd
 
d sd   sd  dt [5.13]
  sd 
Considering the initial strain as zero and time between zero and t, and integrating
correspondingly, we obtain the final strain equation [46].

 sd t
0d sd  sd 0dt [5.14]
  sd 
 sd   t [5.15]
  sd 
Where,  sd ,  sd , and  sd are the strain, stress (MPa), and viscosity of the
suspension damper (MPas), t is the time (sec)
Equation (5.15) clearly shows that the strain experienced by the damper increases
with respect to time. For a given time period the stress is considered a constant (as
the data acquisition is 1Hz) and based on the viscosity of the hydraulic oil, the
strain increases with time until it reaches its maximum corresponding to a given
stress. The stress in the suspension system varies per second and the strain value
depends on the stress. Hence once a maximum strain is reached, there is no
increase or decrease in strain unless the stress is changed. The viscosity of the
hydraulic oil is always constant since there is no effective temperature change
inside the suspension for any adjacent records within a period of time.
The total strain experienced by the Kelvin – Voigt model may be discerned using
either the strain equation for the spring or the strain equation for the damper.
Since the gas spring undergoes an immediate deformation resulting in strain
74
rigid surface
increase or decrease, the equation yielding suspension gas spring strain is used to
obtain a more appropriate value.
5.3.3. Truck tire model and related equations
A truck tire has a linear load - deformation relationship similar to that of a spring.
Rubber acts as an elastic material and hence it was decided to use the concept of
an elastic spring for the tire. There might be some minor permanent deformation
in the tire but it is negligible, hence no damper is included in the model. The tire
takes the load and deforms immediately. Also when the load is removed, the tire
regains its original shape regardless of time delay (the recovery is immediate).
Figure.5.4 Truck tire element
The load from the suspension is directly transferred to the tire and the strain varies
accordingly. The process is different from the suspension gas spring and hydraulic
damper where the strain and recovery depends on time. But here, both the strain
and recovery are immediate (Equation.5.19).
Ft
 tr  [5.16]
kr
 tr
 ts  [5.17]
t
 tr   ts  t [5.18]
75
rigid surface
Ft
 ts  f  F , k ,    [5.19]
kr  t
Where, Ft is the force on the tire (kN), k r is the radial stiffness of the tire (kN/m),
t is the diameter of the tire (m)
5.4. Performance prediction
The pressure data from the VIMS system of a CAT 797B truck was used along
with the rheological model equations to predict the performance for actual full
size truck. Tire comparisons were made for the tires used by Sharma (2009) and
Bolster (2007).
5.4.1. Prediction for CAT 797B truck components
Due to high loading for the rear suspension and tire sets as seen in the analysis
from chapter 3 (See Figure.3.2), both the LR and RR of the truck are considered
here for equation validation and comparison.
90
LR Strut RR Strut
80
70
60
Strain (%)
50
40
30
20
10
0
0 20 40 60 80 100 120 140 160
Time (sec)
Figure.5.5. Suspension strain with time during loading cycle
76
rigid surface
The LR strut experienced a higher strain than the RR strut (Figure.5.5). The
maximum strain on the LR strut was 20% greater than the maximum strain on the
RR strut. As time increased the strain increased in a non-linear fashion.
As a general rule, the mining industry assumes an available stroke length of

0.025m for safe operation, at which status the maximum strain should be 83.4%.
In Figure.5.5 it can be seen that the LR strut is close to the minimum safe
operational strain value. The remaining strain percentage could be easily closed
during an active cycle due to of bad road conditions.
An important aspect to be noted is that the force required for the suspension to
reach an available stroke length of 0.025m from 0.152m happens quickly, but the
amount of force required to move from 0.025m to 0.003m increases rapidly, with
smaller increment of strain. This was clearly indicated by the behavior of the gas
calculated using equation (5.9). The variation is provided in figure (5.6).
70000
98.3% strain
60000
50000
Force (kN)
40000
30000
20000
83.4% strain
10000
0
0.025 0.020 0.015 0.010 0.005 0.003
Figure.5.6. Variation of required force with available stroke length
77
rigid surface
Figures (5.7) and Figure (5.8) provide the strain variation for the suspension when
the truck moved to a dumping area. The maximum strain of the LR strut is
83.11%, whereas for RR strut it is 78.71%. The variation in RR strut shows that
the minimum and maximum strain values reached are 3% and 78.71%. A truck
experienced this difference in strain levels during moving cycle clearly highlights
the bad road conditions.
90
80
70
60
Strain (%)
50
40
30
20
10
0
0 20 40 60 80 100 120 140 160
Time (sec)
Figure.5.7. Suspension strain with time during moving cycle for LR Strut
90
80
70
60
Strain (%)
50
40
30
20
10
0
0 20 40 60 80 100 120 140 160
Time (sec)
Figure.5.8. Suspension strain with time during moving cycle for RR Strut
78
rigid surface
Figures (5.9) and (5.10) show the strain of LR and RR tire for a CAT 797B truck.
At 1-g load the strain for a tire is 7% which defines the correct applicable pressure
for any OTR tire. It may be seen that the LR tire has its strain value above the 1-g
load strain. This is just for one cycle which has around 150 seconds. When
considering the number of cycles the truck has to go around per month, the
number of larger strains experienced by the truck tire will be more.
10 Strain for 1–g load of

9 CAT 797B truck
LR Tire
8
7
Strain (%)
6
5
4
3
2
1
0
0 20 40 60 80 100 120 140 160
Time (sec)
Figure.5.9. Tire strain with time during moving cycle for LR Tire
7
Strain for 1–g load of RR Strut
6
CAT 797B truck
5
4
Strain (%)
0
0 20 40 60 80 100 120 140 160
Time (sec)
Figure.5.10. Tire strain with time during moving cycle for RR Tire
79
rigid surface
Since the truck experiences a huge amount of load in the field, the difference in
strain for various inflation pressure and g-loads were studied (Figure.5.11). The
strain increases linearly with the load. The lower inflated tire has the highest
strain percentage because of less stiffness and the less ability to withstand load
than the higher inflated tire
12
1.5g
1.4g
1.3g
10
1.2g
1.0g 1.1g
Strain (%)
599 kPa
6
621 kPa
4
0
1040 1144 1248 1352 1456 1560
Load (kN)
Figure.5.11. Strain for CAT 797B tire with g-loads
5.4.2. Prediction for the previous laboratory result tires
The strain - load relationship for 2.85m and 0.368m diameter tires is linear
(Figure.5.12 and 5.13). Similar to all tires the strain is more for the lower inflated
tire
At 1.5g load, the strain experienced by the 2.85m diameter tire is 9.4%. The load
withstanding capability of this tire is good. Since there is no data available for
comparison above 1.5g, the strain value above 1.5g was not predicted. More
comparisons could be done in the future to predict the capability of the derived
equations above 1.5g.
80
rigid surface
10
1.4g 1.5g
1.3g
1.2g
8
1g 1.1g
Loaded
6
Strain (%)
552 kPa
621 kPa
4
689 kPa
0
421.4 457.14 500 535.71 578.57 614.29
Load (kN)
Figure.5.12. Strain for 2.85m diameter tire with g – loads
12
10
8
Strain (%)
138 kPa
6 165 kPa
193 kPa
4
207 kPa
0
1.5 2 2.5 3 3.5 4
Load (kN)
Figure.5.13. Strain for 0.368m diameter tire
Similarly for maximum load, the strain experienced by the 0.368m diameter tire is
9.6%. Both tires investigated show a good resistance to load and the prediction
done here matches with the laboratory results and shows that the equations are
valid. Even the comparison done for the full size truck (CAT 797B) tire provided
81
rigid surface
results closer to the field value. More comparisons can be done to predict the
accuracy of the equations for different types of truck.
5.5. Damping characteristics
Since the suspension model consists of spring and the dashpot, it is easy to predict
the damping properties through the conventional spring damper method. The
damping ratio decides the behavior of the suspension and provides an estimate of
how the suspension provides the damping with various load condition. It gives a
prediction of the damping level of the system corresponding to critical damping.
Equations (5.20 to 5.24) provide an outline of the relationship of various
parameters involved in finding out the damping properties [61].
kss
0  [5.20]
ms
Where, 0 is the undamped natural frequency (Hz), k ss is the stiffness of the
suspension spring (N/m), ms is the mass on the spring (kg)
d  0 1   2 [5.21]
d  2f [5.22]
Where, d is the damped natural frequency (Hz),  is the damping ratio, f is the
frequency of vibration (Hz)
c
  [5.23]
cc
c
  [5.24]
2 ms  kss
82
rigid surface
Where, c is the damping coefficient (Ns/m2), cc is the critical damping coefficient
The spring stiffness usually is a constant. But since the gas acts as a spring in the
suspension, the stiffness here is not a constant. The frequency of vibration is
considered as unity since the pressure data is recorded per second and assumption
is made that there is only one vibration per second.
25000000
Spring stiffness (N/m)
20000000
k = 673.01P + 3E+06
R² = 0.9969
15000000
10000000
5000000
0
0 5000 10000 15000 20000 25000
Pressure (kPa)
Figure.5.14. Variation of spring stiffness with pressure
3500000
3000000
Viscous damping coefficient
2500000
c = 110.51P + 453431
2000000 R² = 0.9997
(Ns/m2)
1500000
1000000
500000
0
0 5000 10000 15000 20000 25000
Pressure (kPa)
Figure.5.15. Variation of viscous damping coefficient with pressure
83
rigid surface
The stiffness of the spring increases linearly with pressure (Figure.5.14). This
generally means that the spring force is more when the pressure is more. The
property of the gas is the reason for this behavior. The viscous damping
coefficient is a property of the damper and is responsible for providing the
damping force. This also increases linearly with load (Figure.5.15). This property
generally is also a constant but for this type of spring damping system, it becomes
a variable since the stiffness and the mass on the spring varies with time.
1.000
0.800
Damping ratio
0.600
ξ = 2.2729P-0.118
R² = 0.9779
0.400
0.200
0.000
0 5000 10000 15000 20000 25000
Pressure (kPa)
Figure.5.16. Variation of damping ratio with pressure
Figure.5.16 provides the damping ratio of the suspension. The maximum damping
ratio is 0.93 and the minimum is 0.70. This type of damping is known as “Under
damped condition”, which means that if the damping ratio is in the range 0 to 1
then the system regains its original position with less vibration and within
seconds. Also the systems velocity and amplitude with force decreases
exponentially with time [31]. For the suspension of CAT 797B, the retraction time
is within this range. The vibration is also less and attaining the state
corresponding to a given pressure is not affected. All the characteristics predicted
give a good approximation of the suspension behavior.
84
CHAPTER 6
RHEOLOGICAL MODEL FOR PERFORMANCE PREDICTION OF OIL
SAND, TRUCK SUSPENSION AND TIRE ON OIL SAND
6.1. Introduction and objective
In chapter 5, the model for suspension and tire performance prediction relative to
a rigid surface was considered. Here, the model is studied for a truck operating on
oil sand. Oil sand has two main running surface characteristics relative to truck
motion; sinkage and rutting. The method to predict sinkage in oil sand with a 11%
bitumen content was provided in chapter 4. The main objective of this chapter is
to provide a method to predict the rutting characteristics in oil sand under
different loading conditions.
6.2. Rheological model for truck suspension and tire on oil sand
Figure.6.1. Rheological model for truck suspension, tire and oil sand
85
Chapter 6 – Rheological model for the performance prediction of oil sand, truck suspension and
tire on oil sand
Similar to the models in chapter 5, only the spring and dashpot are assigned to
represent the behavior of suspension, tire, and oil sand for predicting the overall
behavior of a truck ground composite. The models of each component are
connected and their corresponding equations are used to predict the strain
characteristics. The only difference between the overall models in chapter 5 and
chapter 6 is that the oil sand model is included here since it is far softer than a
rigid surface, and strain will develop for varying load. The total strain of the
model is the sum of strains of the individual models (Equation.6.1).
 ost   st   tt   gt [6.1]
Where,  ost is the total strain of the rheological model,  st is the total strain of the
suspension model,  tt is the total strain of the tire model, and  gt is the total
strain of the oil sand ground model
6.2.1. Truck suspension and tire model parameters and equations
The model for the suspension is the same as that used in the chapter 5. This is
because the suspension used for trucks on any ground surface is the same. The
behavior of the suspension is thus similar to the Kelvin – Voigt model, with the
only difference developing for varying ground conditions. The tire model is also
the same because it is considered here to behave only elastically, even if it is
operated on different surfaces. Hence the final strain equation for the suspension
is the same. The same suspension parameters are considered and so the strain
characteristics will not change from that investigated in chapter 5.
However, when a tire is loaded on flexible oil sand, it deflects under the action of
load and that gives rise to a different strain behavior from that experienced on a
rigid bearing surface. Also the footprint area of the tire on oil sand varies from
86
tire on oil sand
that on the rigid surface, hence the general Hooke equation for a spring is used
(Equation 6.2) [58].
 ts
 st  [6.2]
Ets
Where,  st and  ts are the total strain and stress on tire spring (MPa), Ets is the
Young’s modulus of the effective tire spring (MPa)
6.2.2. Oil sand model parameters and equations
There are different rheological models like Kelvin – Voigt, Maxwell, and Burgers
models. When a saturated viscoelastic medium such as oil sand is subjected to
loading and unloading conditions, an appropriate rheological model should be
able to predict the exact behavior. During loading, oil sand exhibits instantaneous
strain followed by time dependant strain including permanent strain. Similarly
during unloading, there will be elastic strain recovery followed by time dependant
strain recovery. In this chapter the analysis accounts for the vehicle in motion.
Behavior under loading and unloading conditions illustrated in figure.6.2, gives an

idea of the response of the models. In this work here, the response shown by a
model was matched to actual oil sand response. However, these are the drawbacks
to three models, which do not match the response of the material
 Maxwell model – When unloaded, there is an instantaneous recovery but there

is no time dependant recovery
 Kelvin – Voigt model – During loading there is no time – independent strain
and when unloading there is no permanent deformation
 Standard linear solid model –Fails to provide permanent deformation under
loading conditions.
87
tire on oil sand
Figure.6.2. Response of different rheological model under loading and unloading

conditions (after [25])
These differences are accounted for in the Burgers model (Figure.6.2 and 6.3).
When a truck is moving, a load is applied on the ground which is accounted for
immediately in the spring of model 2 (Figure.6.3). This is followed by an increase
in strain with time in model 1. Finally there will be a permanent deformation that
is recorded by the damper in model 2 and is irreversible (Figure.6.3).
Of the four models described above, the ability of the Burgers model to predict
the behavior of the oil sand made it more appropriate option to use for oil sand.
This model consists of both the Kelvin – Voigt (Ground model 1) and Maxwell
model (Ground model 2) (Figure.6.3). The overall behavior of these two models
was used to discern the strain characteristics of the oil sand.
88
tire on oil sand
Figure.6.3. Rheological model for oil sand (Burgers model)
Equations (6.3) to (6.6) provide the equations relative to model 1 [10]. The total
strain on both the spring and the damper are the same, but the stress is the sum of
the stresses on the spring and the damper.
 tg   gs   gd
1 1 1
[6.3]
 gs
 gs  1
[6.4]
1
Egs1
  gd1 
 gd   t [6.5]
1   gd 
 1 
 tg   gs   gd
1 1 1
[6.6]
Where,  tg1 is the total strain for oil sand model 1,  gs1 and  gd1 are the strains in
the model 1 spring and damper,  gs1 and  gd1 are the stresses for model 1 spring
89
tire on oil sand
and damper (MPa), Egs1 is Young’s modulus for model 1 spring (Mpa),  gd1 is the
viscosity of model 1 damper (MPas), and t is time (sec)
Equations (6.7) to (6.10) provide the equations for model 2 [10]. The total strain is
the sum of the strains on both the spring and the damper. But the total stress in
this case is the same as the stress for the spring and the damper.
 tg   gs   gd
2 2 2
[6.7]
 gs
 gs  2
[6.8]
2
Egs2
  gd2 
 gd   t [6.9]
2   gd 
 2 
 tg   gs   gd
2 2 2
[6.10]
The total strain of the oil sand model (equation.6.12) is the sum of the strain of the
ground model 1 (equation.6.3) and ground model 2 (equation.6.5). The stress
experienced by each model throughout is the same because of the influence of
same tire footprint area at any instantaneous time influence of the trucks motion
(Equation.6.13).
 tos   tg   tg
1 2
[6.11]
  gs1    gs2    gd2 

 tos     t [6.12]
 Egs   Egs    gd 
 1   2   2 
 osm   gs   gs   gd
1 2 2
[6.13]
Where,  osm is the overall stress on the model (Mpa),  tos is the total strain
experienced by the model,  tg2 is the total strain of the model 2,  gs2 and  gd2 are
90
tire on oil sand
the strains of the model 2 spring and damper,  gs2 and  gd2 are the stresses of the
model 2 spring and damper (MPa), Egs2 is Young’s modulus of model 2 spring
(MPa),  gd2 is the viscosity of model 2 damper (MPas), and t is the time (sec)
The time taken to reach an equilibrium value for a particular load depends on the
load applied. When a constant load is applied the total increase in strain follows
the relationship (Equation.6.14) [25].
  Egs 
 
 osm      osm     osm  t
 1 t
  gd 
 tos  t   1 e  1
[6.14]
Egs1    Egs    gd 
   2   2 
Similarly when the load is removed, the recovery strain follows the relationship
(Equation.6.15) [25]
   Egs 
1     Egs1 tu 
 t
 osm e  1 e  gd1  
   
 ros gd1 
[6.15]
Egs1   
   
Where,  ros is the oil sand recovery strain, tu is the time after unloading (sec), and
t is the time at which stress is removed (sec)
6.3. Analytical output for the model
6.3.1. Analysis parameters
The response of the suspension is not described here as the same parameters are
used as in chapter 5. Due to the varying oil sand ground conditions, only the
91
tire on oil sand
response of a truck tire and the oil sand is predicted. Analysis was performed here
based on the loading and footprint of a CAT 797B truck.
The important parameters for prediction using a spring damper system are
Young’s modulus and viscosity. For oil sand with 11% bitumen content, Sharma
(2009) found these parameters (Table.6.1) which have been used here for
prediction of the strain characteristics of oil sand.
Ground
Element Parameter Value
model
Spring Young’s modulus (MPa) 7.3
1
Damper Viscosity (MPas) 200 – 400
Spring Young’s modulus (MPa) 7.6
2
Damper Viscosity (MPas) 4000 - 12000
Table.6.1. Parameters of oil sand rheological model elements [60]
For analysis purpose, an average viscosity is considered for both the models.
Hence for model 1 and model 2 dampers, viscosities of 300MPas and 8000MPas
are used.
6.3.2. Analytical output to a CAT 797B truck
VIMS data from a CAT 797B truck was used to predict the forces applied on the
tire and oil sand. Although the speed of the truck varied, pressure data was
available at 1Hz collection. Hence the assumption was made that for every second
the truck will come into contact with a new ground surface. The tire footprint area
was obtained using the equations in chapter 5, and then the stress was calculated
based on the force and the footprint area.
92
tire on oil sand
Figure.6.4 shows the tire strain acting on oil sand with stress. A tire acting on oil
sand creates a larger footprint area than acting on rigid surface. This permits the
tire to deform less.
9
8
ε = 0.0089σ
7
R² = 1
6
Tire strain (%)
5
4
3
2
1
0
0 100 200 300 400 500 600 700 800 900
Stress (kPa)
Figure.6.4. Tire strain with stress on Oil sand
To determine the strain characteristics of oil sand, two conditions were considered
reflecting the actual field loading conditions. The first condition considered that
the stress was applied on the ground for just one second, due to motion of the
truck. Figure.6.5 illustrates the oil sand strain when loaded for 1sec. The
maximum strain value reached during a moving cycle was 11.28%. This showed
that even when the duration of loading is small, the oil sand deforms greatly
creating an undulated surface.
93
tire on oil sand
12
10
8
Strain (%)
0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Time (sec)
Figure.6.5. Oil sand strain for 1sec loading duration during truck motion
12
10
8
Strain (%)
0
0 2 4 6 8 10
Time (sec)
Figure.6.6. Strain for 1sec loading on oil sand during first 10sec of a truck in
motion
The strain experienced by the oil sand during a 10 seconds period (Figure.6.6)
clearly highlights the poor load carrying capability of the oil sand, where the
strain varies highly.
94
tire on oil sand
The overall strain of the oil sand when loaded may be evaluated in three parts
viscoelastic medium. Upon loading, an immediate elastic strain occurs which will
be followed by a time dependant strain. There will be a permanent deformation
which depends on the amount of time the load has been applied. Figure.6.7 shows
the percentage strain experienced by oil sand for the different strain components
when loaded for 1second. The strain value of the immediate elastic strain is
greater since it is independent of time and the total strain due to it occurs
immediately. Both the permanent strain and time dependant strain, since the time
of loading is just 1 second, is very low. Another criterion was considered based on
the available data. When a truck suddenly stops at a location during the hauling
cycle, a constant load is applied to the ground for that amount of time. The
durations considered in this analysis were 10, 20, and 30 seconds (Figure.6.8).
12
10
8
Strain (%)
Strain with time
6
Immediate elastic
strain
4
Permanent strain
0
0 20 40 60 80 100
Time (sec)
Figure.6.7. Oil sand strain for 1sec loading period
95
tire on oil sand
18
16
14
12
Strain (%)
10 10sec loading
8 20sec loading
6 30sec loading
4
2
0
0 10 20 30 40 50
Time (sec)
Figure.6.8. Oil sand total strain for different loading periods
The total strain of oil sand for 30 second duration of constant loading was as
expected higher (Figure.6.8). Figure.6.9 provides the variation of strain with stress
for various stationary truck – ground loading times, where clearly the stress for
the higher duration creates the largest strain.
20
18
10sec loading
16
20sec loading
14
30sec loading
12
Strain (%)
10
8
6
4
2
0
0 200 400 600 800 1000
Stress (kPa)
Figure.6.9. Oil sand stress versus strain for different loading periods
96
tire on oil sand
To evaluate the difference type of strains experienced by oil sand, study was
carried out for different loading durations (10, 20, and 30 seconds). The ground
model 1 (Kelvin – Voigt) in figure.6.3 provides the time dependant strain,
whereas the damper in ground model 2 (Maxwell) provides the permanent strain
(Figure.6.3). The variation in immediate elastic strain of spring in ground model 2
was not studied since the strain is independent of time.
Figures.6.10 and 6.11 illustrates the effect of loading duration on strain with time.
Both time dependant (Figure.6.10) and permanent strains (Figure.6.11) are
dependent on time and thus the loading duration clearly affects the total oil sand
strain.
The strain increase which is dependent on loading duration causes the undulated
ground conditions. As the number of truck passes increases along the same path,
the permanent strain increases. Poor load carrying capability of oil sand is
evident. Other road materials which withstand a constant stress for long durations
do not deform like oil sand.
5
Strain (%)
4
10sec loading
3 20sec loading
30sec loading
2
0
0 10 20 30 40 50
Time (sec)
Figure.6.10. Oil sand time dependant strain for different loading periods
97
tire on oil sand
0.3500
Permanent strain (%) 0.3000

10sec loading
0.2500
20sec loading
0.2000 30sec loading
0.1500
0.1000
0.0500
0.0000
0 200 400 600 800 1000
Stress (kPa)
Figure.6.11. Oil sand permanent strain for different loading periods
The behavior of oil sand material for an applied stress of 857kPa commensurate
with a 797B hauler for three loading durations was studied. Figure.6.12 provides
the strain characteristics for a loading duration of 1second and figure.6.13
provides the strain characteristics for loading durations of 30 and 120 seconds.
Figure.6.12. Oil sand permanent strain modeling for 1 sec loading period
98
tire on oil sand
Figure.6.13. Oil sand permanent strain modeling for different loading periods
The immediate elastic strain in these three cases is the same since it is not
dependant on time and the strain occurs immediately. The total strain is however
affected by both the time dependant strain and the permanent strain. As the
loading duration increases, the time dependant strain increases rapidly. For
loadings durations from 30 to 120 second, the increase in strain with time is
almost double. This is due to properties of oil sand, the smaller duration for the
truck at rest on oil sand, and gives rise to better serving quality for the road
surface. The duration of loading and the stress applied are important factors for
the performance prediction of oil sand. As the number of cycles increases, the
permanent strain of oil sand increases.
99
CHAPTER 7
COMPOSITE MODELING OF TRUCK SUSPENSION, TIRE, AND OIL
SAND
A simulation study was performed in Matlab/Simulink to predict the deformation

of the suspension and the tire. Variation in these parameters with ground
deformation was also studied. Since only few ouputs were required using the
simulation model, simulink was used because of its simplicity to model the
combined suspension, tire and oil sand ground conditions.
Springs and dampers were the two main elements considered in this analysis. The
gas in the suspension was considered as a spring and the hydraulic oil considered
to be a liquid damper. The tire model was considered as an elastic spring
(Figure.7.1).
Figure.7.1. Spring – Damper system for a truck
X1, X2, and X3 are the relative datums of suspension, tire and the ground. For the
suspension and tire systems, simulation using the spring and dashpots provided
100
Chapter 7 – Composite modeling of the truck suspension, tire, and oil sand
values that closely matched field values. Two models were created to analyze the
response of the system. The first model was created for the suspension alone,
where the deformation of the suspension and the time taken to achieve
equilibrium was simulated (Figure.7.2). The parameters related to the suspension
model were obtained for 1g from mathematical model used in chapter 5.
Parameters (at 1g load) Value Unit

Force (F) 2080 kN
Stiffness of spring (k1) 16855339 N/m
Viscous damping coefficient (b1) 2682194 Ns/m2
Stiffness of the tire (k2) 3751237 N/m
Mass (m) 212023 kg
Table.7.1. Parameters for the Simulink model
Figure.7.2. Simulink model for truck suspension
101
Figure.7.3 shows the output for a CAT 797B truck at 1g load obtained using the
Simulink suspension model (Figure.7.2). The suspension deforms to a value of
0.123m and then stabilizes at that point within fraction of a second. This
deformation remains the same until the force on the suspension is removed.
Figure.7.3. Output for CAT 797B rear suspension at 1g load
Figure.7.4. Simulink model for CAT 797B suspension and tire

102
To simulate results for a model that consists of both the suspension and a tire, a
different approach was used (Figure.7.4). Newtons second law of motion was
considered to be the basis to this model. Forces acting on the model were used
and the deformation of both the suspension and a tire were obtained.
The Simulink model was analyzed using a force to create a displacement of tire
and suspension. Since this is a system which attains equilibrium and also which is
considered to return back to its initial position once the load is removed, the net
force acting on it should be zero (equation.7.1). A step input models the oil sand
ground and the deformation of it is modeled based on the mathematical model
output for different loads from chapter 4. The basic equations used in this model
are given below
F  0 [7.1]
d 2 x1
m  F  k1  x1  x2   b1  x1  x2  [7.2]
dt 2
d 2 x2
m  k1  x1  x2   b1  x1  x2   k2  w  x2  [7.3]
dt 2
Where, F is the force (kN), x1 and x2 are the suspension and tire positions (m),
k1 and k 2 are the stiffness of suspension and the tire (kN/m), b1 is the viscous
damping coeffiecient (Ns/m2), w is the step input parameter, m is the mass (kg)
When a force is acting on the system, the opposing force would be from the
combined suspension spring and the damper and the tire spring. The force acting
on the suspension and damper are given by the second and third terms
( k1  x1  x2  and b1  x1  x2  ) in equation 7.2. Similarly the force due to the tire is
given by the third term ( k2  w  x2  ) in equation 7.3. The step input here is
modeled using a sinusoidal estimation with time limits based on an evaluation of
103
the actual field data. The relationship used produces a road profile similar to a
bump or potholes based on the input values (equation.7.4) [2].
w  t   a 1  cos8 t  2 for 0  5  t  0  75 [7.4]
and w  t   0 otherwise [7.5]
Where, w  t  is the step input with respect to time, a is a constant parameter, t is
the time (sec)
Figure.7.5. Step input value (ground deformation)
Figure.7.5 and provides a simplified ouput for the ground deformation model as a
step input. Here the ground is modeled to be deformed by 0.1m, from which the
deformation of a tire is obtained (Figure.7.7).
Figures.7.6 and 7.7 provide the deformation of a tire with no ground deformation
(rigid surface) and with ground deformation. The deformation of the tire increases
as the tire passes over a bump (Figure.7.7). It clearly shows that the deformation
of tire varied with respect to a step input. If the step input is positive, the
104
deformation of tire increases, whereas if it is negative, the deformation of the tire

decreases.
Figure.7.6. Tire deformation at 1g load without step input
Figure.7.7. Tire deformation at 1g load with step input
The figures also show that the displacement after a particular period of time
reaches equilibrium. The time taken by the tire to reach equilibrium is higher
105
(Figure.7.6 and 7.7). This is because in the system design, the tire does not have
any damping property to attain equilibrium and the damping from suspension is
considered as a overall damping coeffiecient value for the system (Figure.7.4).
The damping is not sufficient to permit the tire to attain equilibrium within
fraction of seconds. This is due to the total stiffness increase in the overall model
(Figure.7.4) compared to a sole model for the suspension (Figure.7.2).
The ouput provides a tire deformation of 0.554m. But for a set of dual rear tires,
there are two parameters (stiffness) that oppose the force, whereas in the model
only one (stiffness) is considered. Hence the deformation of a single tire at the
rear should be half of the total deformation (i.e. 0.277m). Ouputs were provided
for various input parameters to validate the Simulink model (Table.7.2).
Force (N) Suspension deformation (m) Tire deformation (m)

290000 0.033 0.039
650000 0.085 0.087
1060000 0.105 0.140
1500000 0.110 0.198
2080000 (1g) 0.123 0.278
Table.7.2. Simulink model output for different loads
The model created in this chapter can be used for comparison of various truck
models and sizes. Since the response of the gas in the suspension is different from
a spring, the stiffness and viscous damping coeffiecients are different. Hence
before running the Simulink model, these parameters are important to verify to
obtain accurate results.
106
CHAPTER 8
SUMMARY OF THESIS OUTCOMES AND CONCLUSION
This thesis had two main focusses. The first was to predict a method that can
determine the deformation and strain of the suspension and tire for a truck by
varying inflation pressures and operating conditions. The second was a prediction
of sinkage, rutting, and permanent deformation of oil sand. Other important
parameter such as damping ratio for a suspension was also discussed during the
course of the research.
Figure.8.1. Research focus area and solutions
107
Chapter 8 – Summary of thesis outcomes and conclusion
With the help of the mathematical and rheological models used here, some of the
problems related to current research were eliminated. Figure.8.1 provides a
summary of the problems and solutions provided through this research.
The mathematical equations and rheological models were created for performance
prediction of oil sand, as well as suspension and tire on a rigid and oil sand
ground surfaces. The thermodynamics of the suspension, and parameters related
to the tire were used to predict the overall relationship between the deformation
of a suspension and tire on a rigid surface (Equation.8.1 and 8.2). Inclusion of the
radial stiffness of a tire was a key point here since performance of any size tire
and for any inflation pressure can be discerned. Also difficulty in discerning
Young’s modulus for various tires is eliminated.
Pu  AC
 tr , f  
[8.1]
 Sa 

   2.68 pi wt  t  33 1
 Su 

Pu  AC
 tr ,r  
[8.2]
S 

2  a   2.68 pi wt  t  33 1
 Su 

Equation (8.1) is for front tires because there is only one tire per suspension and
equation (8.2) is for rear tires since there are two tires configured to a single
suspension. Comparisons done with previous field and laboratory results proved
that the above two equations are valid. It showed that the deformation of the tire
follows an exponential relationship with respect to the available stroke length of
the suspension.
For performance of a truck on an oil sand surface, the criteria was different, where
the deformation of oil sand also effects the overall relationship. Sharma (2009)
provided the equation for deformation of a tire on oil sand. A pressure stiffness
108
for the oil sand was included to be able to eliminate the problem due to varying
performance oil sand, after the earlier work of Sharif – Abadi (2006).
Both the pressure stiffness and the deformation of a tire on oil sand were
combined and the final equation for the deformation of oil sand was obtained
(equation.8.3). The output using this equation showed that the response of oil
sand depends on the amount of load, number of loading cycles, and the type of
loading (increasing cyclic, varying, constant). The sinkage, ground deformation
characteristics of the oil sand can be found out using this equation.
Ft  E
 os  [8.3]
 k  C     k p  C  E  t 
2
p t
2
In predicting the strain value for the suspension, tire, and oil sand, rheological
models were used. Equations related to Kelvin – Voigt, Elastic spring, Maxwell,
and Burgers model were used in the research. Equation (8.4) provides the total
strain value for the truck on a rigid surface.
  ss   
E t
 F 
 1  e  sd     t 

 tr   ss [8.4]
 Es    kr  t 
 b  
Equation (8.5) provides the total strain value for the truck on oil sand. This
equation can be used to find the rutting ground deformation characteristics of the
oil sand. Oil sand experiences permanent strain because of its viscoelastic nature.
The difficulty to find this parameter is now eliminated. Rheological model term 3
in equation (8.5) can be used to predict this value. From the output it showed that
the permanent deformation depends on the load applied, duration of the load, and
the loading cycle.
109
  Egs 
 
 osm      osm     osm  t
 1 t
  gd 
 tos  t   1 e  1
[8.5]
Egs1    Egs    gd 
   2   2 
The recovery rate of oil sand was also found (equation.8.6) where the rate of
recovery was more for smaller loads and durations, whereas for higher loads and
durations, the rate of recovery and the amount of recovery was less.
  Egs1 t    Egs1 tu 

 osm e  gd1   1 e   gd1  
 ros  [8.6]
Egs1   
   
Finally simulation was done using Matlab Simulink software. Two models were
created in the simulation. The first model was only for the suspension, which was
used to predict the time taken by the suspension to attain a steady state.
The second model was used to simulate the entire model consisting of suspension,
tire, and the ground. The ouput from the first model showed that the suspension
attained steady state within seconds. The output from the second model gave the
tire deformation value for various loads.
A step input and a sinusoidal wave were provided for the road profile. To validate
each of the models used in this research, comparisons were made for various
loads (Table.8.1). The most important factor in this research is the comparison of
outputs of all the three models used (Mathematical, Rheological, and Simulink).
The comparison was made for a CAT 797B truck suspension, and tires
considering an inflation pressure of 600kPa and diameter of 3.84m (Table.8.1).
110
Suspension Deformation (m) Tire deformation (m)

Force
(kN) Mathematical Rheological Mathematical Rheological
Simulink Simulink
model model model model
290 0.033 0.033 0.033 0.039 0.039 0.039

450 0.064 0.065 0.064 0.062 0.062 0.060
650 0.085 0.086 0.084 0.087 0.088 0.087
850 0.096 0.097 0.096 0.113 0.114 0.113
1060 0.105 0.105 0.105 0.141 0.142 0.141
1400 0.114 0.114 0.114 0.186 0.187 0.185
1500 0.116 0.116 0.115 0.198 0.200 0.198
1700 0.118 0.119 0.118 0.228 0.231 0.228
2080
0.123 0.123 0.123 0.277 0.277 0.277
(1g)
2450 0.125 0.126 0.126 0.325 0.330 0.325
Table.8.1. Comparison between mathematical, rheological, and Simulink model
outputs for CAT 797B
The comparisons show that all the models provide values close to each other and
even matching with the field values at 1g load. The ability of these models can
now be visualized and the equations and the simulation models used here may
now be utilized for various truck sizes under different operating conditions, such
that performance can be predicted. Another important factor is the ability of these
models and equations to eliminate the drawbacks in the current research area,
providing a simple method to predict the various performance characteristics of
suspension, tire, and oil sand.
111
CHAPTER 9
FUTURE RECOMMENDATIONS
The mathematical equations for the relationship between deformation of tire and
suspension available stroke length were calculated for a CAT 797B truck.
However comparisons with field data and various other trucks operated on rigid
surfaces need further validation. Hence field studies should be done for various
trucks and comparisons made.
Oil sand surfaces with deformation characteristics were considered in this

research. But some of the mines are now employing limestone as the surface
material for the haul roads. This material behaves plastically and there could be
some deformation which might effect the performance of the truck. Hence studies
could be done on this material and the overall relationship between suspension,
tire, and the ground be predicted.
Room temperature condition were considered for the oil sand in the model and
with a bitumen content of 11%. But in field the bitumen content varies from 8% -
15% and temperature varies between -30°C to +30°C. Hence suitable laboratory
set up could be done to predict the deformation and footprint area equations for
these varying operating conditions.
The wear rate of the tire is another important factor. If the road condition is very
bad it might create more wear on the tire. This results in reduction of the overall
diameter of the tire and in turn reduces the stiffness. Hence the effect of this
parameter could be included to predict the stiffness of the tire as the operating
hours increases. This will definitely help to find out the change in strength of the
tire with time providing a better relationship between truck components and the
ground.
112
Chapter 9 – Future recommendations
A truck encountering a haul road corner will create the forces acting on the tire
deviate from the central axis due to slip resulting in transfer of force on a footprint
area. In this case the deformation at that area will be more than the expected. Slip
angle and the cornering stiffness are the parameters behind this behavior. This
analysis could also be done through dynamic analysis on a truck.
The value of Young’s modulus and viscosity of oil sand with 11% bitumen
content was used in the rheological model to predict the deformation
characteristics. This value was obtained from Sharma (2009) by performing
laboratory tests and with the help of a Burgers model. A similar method could be
performed for oil sand with various bitumen content and using the value obtained
and the equations from this research, deformation characteristics could be
predicted for different operating conditions.
The effect of rolling resistance was not included in determining the rutting
potential of oil sand. After finding out the Youngs modulus and viscosity of
various oil sand materials, these values could be used in a rheological model
along with the effect of rolling resistance to predict the rutting potential.
113
REFERENCES
[1] About Rheology. (n.d.). Retrieved July 10, 2013, from

http://www.iq.usp.br/mralcant/About_Rheo.html
[2] Agharkakli, A., Chavan, U., & Phrithran, S. (2012). Simulation and analysis of
passive and active suspension system using quarter car model for non -
uniform road profile. International Journal of Engineering Research and
Applications, 2(5), pp 900 - 906.
[3] Anochie - Boateng, J. K. (2007). Advanced testing and characterization of

transportation soils and bituminous sands. PhD Thesis. University of Illinois,
Illinois.
[4] Anochie - Boateng, J. K., Tutumluer, E., & Carpenter, S. H. (2010). Case
study: Dynamic modulus characterization of naturally occuring bituminous
sands for sustainable pavement applications. International Journal of
Pavement research and technology, 3(6), pp 286-294.
[5] Appleyard, M., & Wellstead, P. E. (1995). Active suspensions : some

background. IEE Proc-Control theory Appl, 142(2), pp 123-128.
[6] Attila, C., Zoltan, D., & Laszlo, D. (2004). Changing the rheological features
of AG-BAG type packaging foil tubes depending on temperature.
[7] Balhoff, M. (2005). Modeling the flow of Non - Newtonian fluids in packed
beds at the pore scale. PhD Thesis. Louisiana State University, Louisiana.
[8] Banfill, P. F. (2006). Rheology of fresh cement and concrete. pp 61-130.
[9] Barksdale, R. D. (1972). Laboratory evaluation of rutting in base course

114
materials. 3rd International conference on structural design of asphalt
pavements, (pp. 161-174).
[10] Barnes, H. A., Hutton, J. F., & Walters, K. (1999). Introduction to Rheology
(2nd ed.). Elsevier Science Publications Ltd.
[11] Berezan, J. J., Joseph, T. G., & del Valle, V. D. (2004). Monitoring whole
body vibration effects on ultra-class haulers. CIM Bulletin, 97, 1082, pp 1-4.
[12] Blanchard, E. D. (2003). On the control aspects of semiactive suspensions for

automobile applications. MSc. Thesis. Virginia Polytechnic Institute and
State University, Virginia.
[13] Bolster, M. J. (2007). Tire rim interactions for ultra-class truck performance.
MSc Thesis. University of Alberta, Edmonton.
[14] Bridgestone, F. (2008). Earthmover tire catalogue and base price list, Off-
The Road tires. pp 16-18.
[15] Cameron, R., Mahood, R., Lewko, R., & Skitmore, J. (1996a. 1995). Haul
road investigation of problem areas for 240 - Ton heavy haulers driving on
170 - Ton haul road design at Syncrude Canada Limited. Internal report ,
Syncrude Canada Limited.
[16] Cameron, R., Mahood, R., Lewko, R., & Skitmore, J. (1996b. 1996). Haul
road design, construction and monitoring procedures for 240 - Ton heavy
haulers at Syncrude canada Limited. Internal report, Syncrude Canada
Limited .
[17] Cao, J., Liu, H., Li, P., & Brown, D. J. (2008). State of the art in vehicle
active suspension adaptive control systems based on intelligent
115
methodologies. IEEE Transactions on intelligent transportation systems,
9(3), pp 392-405.
[18] Caterpillar. (2002). Caterpillar service manual for 793C truck, manual form
SENR 1540, specifications, SENR 1455-04.
[19] Caterpillar Inc. (2004). Caterpillar performance handbook.(35th Edition),5-9.
[20] Choudhury, S. F., & Sarkar, M. R. (2012). An approach on performance

comparison between automotive passive suspension and active suspension
system (PID Controller) using Matlab/Simulink. Journal of Theoretical and
Applied Information Technology, 43(2), pp 295 - 300.
[21] Creed, B., Kahawatte, N., & Varnhagen, S. (2010). Development of a full car
vehicle dynamics model for use in the design of an active suspension control
systems.
[22] Cunagin, W. D., & Grubbs, A. B. (1984). Automated acquisition of trucl tire
pressure data. Transportation research record 1123, Transporation research
board, pp 112 - 121.
[23] De Haan, Y. M., & Sluimer, G. M. (2001). Standard linear solid model for
dynamic and time dependant behavior of building materials. 46(1), pp 49-76.
[24] Dekker, M. (1999). Nonlinear and variable stiffness systems.
[25] Dey, A., & Basudhar, P. K. (2010). Applicability of Burger model in

predicting the response of Viscoelastic soil beds.
[26] Doraisamy, D. (2001). The origins of rheology : A short historical excursion.
116
[27] Doucet, R. (2001). 21st Century maintenance management. CIM Bulletin, (p.
93).
[28] El-Sayed, M. (2003). Suspension Systems in Large Mining Trucks. MEng

Dissertation. University of Alberta, Edmonton.
[29] Findley, W. N., Lai, J. S., & Onaran, K. (1989). Creep and relaxation of
nonlinear viscoelastic materials.
[30] Gao, W., Zhang, N., & Du, H. (2007). A half - car model for dynamic
analysis of vehicles with random parameters. 5th Australian congress on
applied mechanics. Brisbane.
[31] Gavin, H. P. (2010). Vibrations of single degree of freedom systems.
[32] Goodman, R. E. (1989). Introduction to rock mechanics (2nd ed.). New

York: John Wiley and Sons.
[33] Grozic, J. H. (1999). The behavior of loose gassy sand and its susceptibility
to liquefaction. PhD Thesis. University of Alberta, Edmonton.
[34] Guglielmino, E., Sireteann, T., Stammers, C. W., Ghita, G., & Giuclea, M.
(2008). Semi active suspension control. Springer - Verlag London Limited.
[35] Hackley, V. A., & Ferraris, C. F. (2001). Guide to Rheological nomenclature

: Measurements in ceramic particulate systems.
[36] Hadekel, R. (1940). A method of estimating the performance of Oleo-

Pneumatic struts. The Aircraft Engineer, 7(174).
[37] Haul roads. (n.d.). Retrieved August 15, 2013, from Hammerstone
117
corporation:
http://www.hammerstonecorp.com/files/37.HammerStoneHaulRoadsWeb.pdf
[38] Hou, C.-Y., Hsu, D.-S., Lee, Y.-F., Chen, H.-Y., & Lee, J.-D. (2007). Shear -
thinning effects in annular - orifice viscous fluid dampers. Journal of
Chinese Institute of Engineers, 30(2), pp 275-287.
[39] Joseph, D. D. (1986). Historical perspectives on the elasticity of liquids.

Journal of Non - Newtonian fluid mechanics, 19, pp 237 - 249.
[40] Joseph, T. G. (2002). OsEIP: the oil sands - equipment interactions program.
CIM Bulletin, 95(1064), pp 58-61.
[41] Joseph, T. G. (2003). Large mobile equipment operating on soft ground. 18th
International Mining Conference and Exhibition of Turkey, (pp. 143-147).
[42] Joseph, T. G. (2005). Mathematical, static and inferred dynamic loaded

properties of oil sand. Progress report, Phases II and III (Final report), CAT/0405.
[43] Joseph, T. G., & Hansen, G. (2002). Oil sand reaction to cable shovel
motion. CIM Bulletin, 95(1064), pp 62-64.
[44] Joseph, T. G., Sharif - Abadi, A. D., & Shi, N. (2003). A broken material
approach to modeling oil sand under dynamic load. In Proceedings of CAMI
Conference. Calgary.
[45] Kasprzak, E. M., Lewis, K. E., & Milliken, D. L. (2006). Inflation pressure
effects in the Nondimensional tire model.
[46] Kelly. (n.d.). Solid Mechanics Part 1. Lecture.
118
[47] Kizhakkethara, I. (1995). Non - linear static analysis of aircraft tire subjected
to inflation pressure and ground contact loads using finite element analysis.
MSc Thesis . University of New Orleans, New Orleans.
[48] Larson, R. G. (1999). The structure and rheology of complex fluids.

NewYork: Oxford University Press.
[49] Lekarp, F., Isacsson, U., & Dawson, A. (2000). Permanent strain response of
unbound aggregates. Journal of Transportation Engineering, 126(1), pp 76-
83.
[50] Li, P., & Chalaturnyk, R. J. (2005). Geomechanical model of oil sands. SPE
International thermal operations and heavy oil symposium, (pp. B453 -
B457). Calgary.
[51] Lin, Z. (2007). Geometry and mechanics of pneumatic tires.
[52] Nguyen, Q.-H., & Nguyen, N.-D. (2012). Incompressible Non-Newtonian

fluid flows. In Y. Gan (Ed.), Continuum Mechanics - Progress in
fundamentals and Engineering applications (pp. 47 - 72). Intech Publication.
[53] Provenzano, P. P., Lakes, R. S., Corr, D. T., & Vanderby Jr, R. (2002).
Application of nonlinear Viscoelastic models to describe ligament behavior.
pp 45-57.
[54] Robert, W., Funnell, J., Maftoon, N., & Decreamer, W. F. (2012). Mechanics
and modeling for the middle ear.
[55] Roylance, D. (2001). Engineering Viscoelasticity.
119
[56] Saarilahti, M. (2002). Soil Interaction Model : Manual 1. Seltra -
Documentation of the computer programme for calculating of the
trafficability of terrain and mobility of forest tractors, University of Helsinki.
[57] Santos, R. F. (2007). A Rapid Dampening Suspension for Ultra-Class

Haulers. MSc Thesis. University of Alberta, Edmonton.
[58] Schiessel, H., Metzler, R., Blumen, A., & Nonnen, T. M. (1995). Generalized
viscoelastic models : Their fractional equations with solutions. IOP
Publishing Limited.
[59] Sharif-Abadi, A. D. (2006). Cyclic performance of soft ground. PhD Thesis.

University of Alberta, Edmonton.
[60] Sharma, A. (2009). Scale tire - oil sand interactions. MSc Thesis. University
of Alberta, Edmonton.
[61] Soni, R. K. (2009). Scaled rapid dampening suspension evaluations. MSc

Thesis. University of Alberta, Edmonton.
[62] Stroshine, R. (1998). Mathematical properties of agricultural materials and

food products. Purdue University, West Lafayette.
[63] TA Instruments : Rheology application note. (n.d.). Retrieved July 29, 2013,
from http://www.tainstruments.com/library_download.aspx?File=RN9.pdf
[64] Tannant, D. D., & Regensburg, B. (2001). Guidelines for Mine haul road
design.
[65] Tanner, R. I., & Walters, K. (1998). Rheology:An historical perspective (Vol.
7). Elsevier Publication Ltd.
120
[66] Tielking, J. T. (1994). Force transmissibility of heavy truck tires. Tire
Science and Technology: 22, 1, pp 60-74.
[67] Tielking, J. T., & Abraham, M. A. (1990). Measurement of truck tire

footprint pressures. Transportation research record : 1435, Transportation
research board, pp 92-99, Washington, DC.
[68] Tomkins, E. (1981). The history of the pneumatic tyre. London: Eastland
Press.
[69] Van der Steen, R. (2007). Tyre/road friction modeling. Literature survey,
Eindhoren University of Technology, Eindhoren.
[70] Vermant, J. (n.d.). Rheology and structure of complex fluids. Retrieved July
10, 2013, from http://www.eu-softcomp.net/FILES/IFFFS_REO-1.pdf
[71] Walters, K. (n.d.). History of Rheology. Retrieved July 10, 2013, from
http://www.eolss.net/Sample-Chapters/C06/E6-197-01.pdf
[72] Williams, R. A. (1994). Electronically controlled automotive suspensions.

Computing and control Engineering Journal, 5(3), pp 143-148.
[73] Wills, D. S. (1989). Guidelines for monitoring mine haul road construction
at Syncrude canada limited. Internal report, Syncrude Canada Limited.
[74] Xue, X. D., Cheng, K. E., Zhang, Z., Lin, J. K., Wang, D. H., Bao, Y. J., et
al. (2011). Study of art of automotive active suspensions. 4th International
conference on Power electronics systems and applications, (pp. 360 - 366).
121
[75] Ziegenmeyer, J. D. (2007). Estimation of disturbance inputs to a tire coupled
quarter - car suspension test rig. MSc. Thesis. Virginia Polytechnic Institute
and State University, Virginia.
122
APPENDIX
Appendix – A
A.1. Dimensions of CAT 797B rear suspension
123
Appendix – B
B.1. Mathematical model output for ϕ 2.85m tire on a rigid surface
B.1.1. Result for various inflation pressures
1.400
1.200
1.000
0.800
552kPa
0.600
621kPa
0.400 689kPa
0.200
0.000
0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350
Tire Deformation (m)
B.2.1. Result for various inflation pressures
0.020
0.018
0.016
0.014
0.012 138kPa
0.010 165kPa
0.008 193kPa
0.006
207kPa
0.004
0.002
0.000
0.000 0.010 0.020 0.030 0.040
124
B.3.1. Result at inflation pressure of 599 kPa
2.50
599kPa
2.00
1.50
1.00
0.50
0.00
0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450
B.3.2. Result at inflation pressure of 620 kPa
2.50
620kPa
2.00
1.50
1.00
0.50
0.00
0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450
125
B.4. Mathematical model output for ϕ 3.84m tire on oil sand
B.4.1. Deformation of tire and oil sand with suspension pressure
0.3000
0.2500
Tire
0.2000
Deformation (m)
0.1500
Oil sand
0.1000
0.0500
0.0000
13574
10369
10558
10746
11312
13008
13951
14140
14328
14517
14705
15271
16779
17533
2639
2828
3016
4148
4336
4525
6410
6787
6976
7164
7541
9992
Suspension pressure (kPa)
126

Sundararajan - Sujith - Spring 2014-2

Uploaded by

Copyright:

Available Formats

Sundararajan - Sujith - Spring 2014-2

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

Sundararajan - Sujith - Spring 2014-2

Uploaded by

Copyright:

Available Formats

University of Alberta

A rheological model for composite truck ground behavior

Civil and Environmental Engineering

Mining haul trucks are operated on different ground conditions subjected to

components like suspension, tire, and the operating ground. Extensive

experimental research has been done on these components at scale.

In order to understand the deformation behavior and the overall relationship

oil sand ground surfaces.

and individual strain of the components. A model consisting of suspension, tire,

I am thankful for my great friends, who supported, encouraged, helped me in

I acknowledge the financial support provided by NSERC and the University of

Table.2.1. Material properties of tire tread and sidewall (From[13])……………22

Figure1.1 Flowchart showing the methodology and research approach to predict

 ts Deformation of tire on oil sand

 os Deformation of oil sand

AC Area of suspension cylinder

Atr Footprint area of tire for rigid surface

Aos Total area of oil sand

Fg Force on the ground

Ft Force on the tire

FOEM , F Damping force for full size OEM strut

FOEM , S Damping force for scale OEM strut

Frt Force on the rear tire

KOEM , F Full size OEM strut stiffness

KOEM , S Scale OEM strut stiffness

Pu Suspension pressure at unloaded condition

pi Inflation pressure of tire

Su Stroke length at unloaded condition

The knowledge of deformation behavior and the relationship between the

1.2. Purpose of research

components, which leads to major fatigue, equipment instability, and improper

1.3 Research approach

Dynamic behavior of truck components is a subject of on-going interest to the

This chapter provides an understanding of previous work to better understand the

2.2. Fundamentals and stride towards the branch of rheology

“Rheology is the study of materials deformation and flow properties” [10]. In

The most important aspect of rheology is the time dependent behavior of

Weber (1835) performed an experiment on silk threads and it showed a non –

2.3. Important parameters in rheology, viscoelastic models and their studies

The most influential period in rheology was the introduction of mechanical

In terms of springs in mechanical models, it is purely an elastic component with

Figure.2.3. Strain response of Hooke and Kelvin model (after [10])

In terms of a damper in mechanical models, viscosity is the main parameter and is

Figure.2.4. Shear rate vs Viscosity for low-density polyethylene melt (after,

Where,  is the viscosity (Pas),  is a constant (Pas), Ea is the activation energy

The important difference between “Non-Newtonian fluids” and “Viscoelasticity”

from Newtonian fluids. He studied the characteristics of different Non-Newtonian

viscoelastic behavior is known as “viscoelastic liquid”. Regarding linear models,

where, G is a material constant

where,  is the time constant,  is the viscosity (Pas),  is the strain

He used the one-dimensional equation to solve some of the problems related to

Where,  is the stress,  is the strain, E is the relaxation function,  is the

Provenzano et al (2002) stated that for relaxation function is not dependant on

Hackley and Ferraris (2001) discussed about various rheological terminologies

   0 cos t    [2.8]

where,  is the angular frequency,  is the phase angle, t is the time

Attila et al (2004) used another three parameter model, the Poynting-Thomson

2.4. Complex fluid models and application of rheology in various fields

(1/sec), n is the power law index