Abhishek_Dalal_Intern_Report

STUDY OF LARGE DEFORMATION
ELASTO-PLASTIC CONTACT IMPACT
PROBLEMS USING ANSYS (APDL)
A Project Report Submitted
In Completion of the
Summer Internship Project
By
Abhishek Dalal
4th Year Student
DEPARTMENT OF MECHANICAL ENGINEERING
INDIAN INSTITUTE OF TECHNOLOGY (BHU) VARANASI
Under the Guidance of
Dr. Sachin Singh Gautam
Assistant Professor
DEPARTMENT OF MECHANICAL ENGINEERING
INDIAN INSTITUTE OF TECHNOLOGY GUWAHATI
July, 2016

ii
ACKNOWLEDGEMENT
Foremost, I would like to express my sincere gratitude to my guide Dr. Sachin Singh
Gautam for the continuous support during my summer internship study and research, for his
patience, motivation, enthusiasm, and immense knowledge. His guidance helped me in all the
time of internship. I could not have imagined having a better advisor and mentor for my summer
2016 internship. Dr. Gautam, it has been an honor to work with you.
I thank my fellow interns in IIT Guwahati: Ayush Jhalani, Shriram Ravindra Ashirgade
and Neeraj Upadhyay for the stimulating discussions, for the sleepless nights we were working
together, and for all the fun we have had in these past 9 weeks. Also I thank my friends in IIT
BHU: Krishna Pal, Harsh Rohan, Abhinav Roy, Garvit Sharma, Dontu Saikrishna and Amit
Sharma for their continuous support, motivation and healthy discussions on various topics.
Lastly, I would like to thank my family, my parents and sister for supporting me
throughout my life.

iii
Contents
ACKNOWLEDGEMENT .........................................................................................................ii
List of Figures ............................................................................................................................ v
Chapter 1 Introduction .......................................................................................................... 1
1.1 Review of Literature.................................................................................................... 1
1.1.1 Elasto-plastic Finite Element Analysis ................................................................ 3
1.1.2 Dynamic Finite Element Analysis........................................................................ 4
1.1.3 Contact Analysis .................................................................................................. 4
1.1.4 Objective Stress Measures ................................................................................... 8
1.2 Objective of the Present work ..................................................................................... 9
Chapter 2 Mathematical Modeling of Dynamic Elasto-Plastic Problem............................ 10
2.1 Introduction ............................................................................................................... 10
2.2 Kinematics of Finite Deformation............................................................................. 11
2.3 Stress Measures ......................................................................................................... 12
2.4 Strain Measures ......................................................................................................... 13
2.5 Theory of Plasticity ................................................................................................... 14
2.5.1 Yield Criterion.................................................................................................... 14
2.5.2 Strain Hardening Postulate................................................................................. 16
2.5.3 Plastic Flow Rule ............................................................................................... 17
Chapter 3 Results and Discussion....................................................................................... 18
3.1 Large Deformation, Dynamic, Elastic, Single Body Analysis .................................. 18
3.1.1 Dynamic Response of a Cantilever .................................................................... 18
3.2 Large Deformation Elastic Contact-Impact Analysis................................................ 21
3.2.1 Impact of Two Identical Bars............................................................................. 21
3.2.2 Oblique Impact of Two Infinite Blocks ............................................................. 25
3.3 Large Deformation Dynamic Elasto-Plastic Single Body Analysis.......................... 29
3.3.1 Dynamic Elasto-Plastic Analysis of a Cube....................................................... 29
3.3.2 Dynamic Elasto-Plastic Response of a fixed Square Plate ................................ 31
3.4 Large Deformation Elasto-Plastic Contact-Impact Analysis..................................... 34

iv
3.4.1 Elasto-Plastic Impact of a Sphere on a Plate...................................................... 34
Chapter 4 Conclusions and Results..................................................................................... 41
References ................................................................................................................................ 42

v
List of Figures
Figure 2.1 Yield surface in principal stress space.................................................................... 15
Figure 2.2 Multi-linear Isotropic Hardening............................................................................ 17
Figure 3.1 Cantilever: geometry and boundary conditions...................................................... 18
Figure 3.2 Dynamic response of the Cantilever at ∆t = 0.45 × 10-04
sec ................................. 20
Figure 3.3 Maximum Equivalent Stress response at ∆t = 0.45 × 10-04
sec .............................. 20
Figure 3.4 Impact of two identical bars.................................................................................... 21
Figure 3.5 Velocity at interface of bar A ................................................................................. 23
Figure 3.6 Velocity at interface of bar B.................................................................................. 23
Figure 3.7 Velocity at midpoint of bar A................................................................................. 24
Figure 3.8 Velocity at midpoint of bar B ................................................................................. 24
Figure 3.9 Oblique Impact of two Elastic Blocks .................................................................... 25
Figure 3.10 X-displacement of point P (frictional case) .......................................................... 26
Figure 3.11 X-displacement of point P (frictionless case)....................................................... 27
Figure 3.12 Z-displacement of point-P (frictional case) .......................................................... 27
Figure 3.13Z-displacement of point P (frictionless case) ........................................................ 28
Figure 3.14 Cube: geometry and boundary conditions ............................................................ 29
Figure 3.15 Elasto-plastic response of the cube....................................................................... 30
Figure 3.16 Plate: geometry and boundary conditions............................................................. 31
Figure 3.17 Maximum displacement response of the plate...................................................... 32
Figure 3.18 Maximum Equivalent Stress response of the plate............................................... 33
Figure 3.19 Equivalent Stress v/s Equivalent Plastic Strain .................................................... 33
Figure 3.20 Impact of Sphere on Plate..................................................................................... 34
Figure 3.21 Finite element plot of the sphere........................................................................... 35
Figure 3.22: Displacement of the sphere along X-direction for frictionless case.................... 36
Figure 3.23: Displacement of the sphere along Z-direction for frictionless case. ................... 37
Figure 3.24: Displacement of the sphere along X-direction for frictional case. ...................... 37
Figure 3.25: Displacement of the sphere along Z-direction for the frictional case.................. 38
Figure 3.26: Displacement of the sphere along X-direction (polycarbonate material) frictionless
case........................................................................................................................................... 38
Figure 3.27: Displacement of the sphere along Z-direction (polycarbonate material) frictionless
case........................................................................................................................................... 39

vi
Figure 3.28: Displacement of the sphere along X-direction (polycarbonate material) frictional
case........................................................................................................................................... 39
Figure 3.29: Displacement of the sphere along Z-direction (polycarbonate material) frictional
case........................................................................................................................................... 40

1
Chapter 1
Introduction
Impact phenomenon occur in many engineering situations. The basic characteristics of an
impact phenomenon are very brief duration, high force levels reached, rapid dissipation of
energy and large accelerations and decelerations present. In many cases, the bodies undergo
plastic deformations during impact. Sometimes frictional effects are also important during
impact. A rigorous analysis of the impact problem requires the consideration of all these
complex aspects. The analysis of the problems like the automobile crashes involves the use of
dynamics, elasto-plasticity and large deformation in a contact analysis scenario.
Closed form mathematical solutions are extremely difficult to develop even for simple
impact problems. As the impact scenario becomes more complex, as in the case of elasto-plastic
materials, arriving at a closed form solution is nearly impossible. This necessitates the use of
approximate numerical tools for the analysis, of which finite element analysis is arguably the
most versatile option.
1.1 Review of Literature
The classical incremental theory of elasto-plasticity [1] is generally taken to be the basis for
predicting elasto-plastic response. The constitutive equations for these materials is derived by
combining the stress-strain relations for elastic and plastic responses. The constitutive relation
for a plastic response requires three fundamental characteristics of the material: a yield
criterion, a flow rule, and a strain hardening relationship. The yield criterion demarcates the
elastic behavior from the plastic behavior. It specifies the state of multi-axial stress
corresponding to the start of the plastic flow. Various yield criteria have been proposed by
Tresca, von Mises etc. [2], where von Mises criterion is found to be the simplest and the most
accurate in terms of the predicted values. The flow rule relates the plastic strain rate to the
current stress. It is represented in terms of a plastic potential, analogous to the elastic potential,
and is termed as the associated flow rule if the potential is taken as the yield function. The
hardening rule specifies how the yield condition is modified during the plastic flow. For the

2
purpose of analysis, the nonlinear hardening relationship is usually modelled in terms of a
power law.
Unlike the case of small deformation of elastic materials, the governing equations of
elasto-plastic materials, which undergo large deformation, contain two types of nonlinearities:
(i) geometric nonlinearity associated with nonlinear strain displacement relations and, (ii)
material nonlinearity in the form of nonlinear stress-strain relations. For a contact problem,
there is an additional nonlinearity: Boundary nonlinearity due to the nonlinear boundary
conditions at the contact interface. An incremental formulation like the Updated Lagrangian
Formulation is the most convenient way to handle such nonlinearities unlike the Total
Lagrangian Formulation. The advantage of the former is that, if the increment size is
sufficiently small, linearized incremental equations can be used. However, when the overall
deformation is large, the number of increments becomes prohibitively large. In such a situation,
one has to resort to one of the various iterative schemes available for solving the nonlinear
incremental equations. The prominent among such schemes is the Newton-Raphson Method. In
this method, the stiffness matrix is updated in each iteration of the increment. Thus, we work
with the tangent modulus at that instant for each iteration. This method has two shortcomings.
The first being that there is chance of divergence if the slope of force-displacement curve
increases with the increment. Secondly, it takes more computational time as the stiffness matrix
has to be computed afresh in each iteration [3]. Thus, the Modified Newton-Raphson method in
which, the stiffness matrix is kept constant during the increment, is adopted.
The dynamic analysis is very basic to studying impact problems. There are various ways
of performing the time discretization in this category of problems of which, the direct
integration methods are prominent. These methods include the central difference method,
Houbolt method, Wilson-Ɵ method and Newmark method [4]. The first method is conditionally
stable one, while the rest of them are unconditionally stable. A study of the stability of Newmark
algorithm in nonlinear dynamics has been done by Hughes [5]. He has studied some potential
problems where the stability of the algorithm fails. He has suggested that it can happen in a
nonlinear problem if the discrete internal energy is negative. Analysis of accuracy and speed of
different algorithms is presented by Bathe [4]. It has been concluded that although central
difference method is beneficial in terms of speed and cost, the trapezoidal method of
Newmark’s algorithm is best in terms of stability and accuracy.

3
1.1.1 Elasto-plastic Finite Element Analysis
As stated earlier, the classical theory of elasto-plasticity [1] is generally taken to be the basis
for predicting the elasto-plastic response. Argyris et al. [6] presented one of the earliest
formulations and solution techniques for elasto-plastic problems. The effect of plasticity was
simulated by “initial loads” and the matrix method was used to discretize the linearized
governing equations. Two approaches, the direct incremental method and the iterative
incremental method, were discussed for obtaining the solution. The early applications of the
iterative incremental method [3] used the initial stiffness method. However, this approach is
not suited when the overall deformation is very large as it results in very slow convergence, or
in many cases, even divergence.
Nagtegaal and Jong [7] dealt with the computational aspects of elastic-plastic large
deformation analysis. A modified set of governing equations were formulated, which were
applicable for small strain and large rotational increments. Special attention was paid to the
integration of these equations in the deformation history. It is predicted that explicit methods,
like the tangent modulus procedure, are conditionally stable whereas the implicit schemes, such
as the mean normal method, are preferable. The main difficulty with the three dimensional
analysis is the vast computational memory that is required for carrying out the analysis. But,
with the advent of better machines and evolution of more efficient algorithms this problem has
been solved and more and more such analyses are being carried out.
Since the elasto-plastic constitutive relation is generally given in the differential form,
the constitutive function must be integrated to obtain the incremental stress. A major research
effort has been devoted in recent years to the development and critical assessments of stress
updating or integration schemes for the rate (incremental) equation of plasticity. The procedures
commonly used are the simple and robust Euler forward integration technique, the implicit
Euler backward integration technique and the radial return method. The trend has lately been
shifting towards the use of consistent schemes which improve the convergence characteristics
of the equilibrium iterations [8]. The recent works have clearly emphasized the use of implicit
integration methods [9] [10] [11] in view of their superior stability and convergence properties
[12]. A one step, fully implicit method of backward Euler integration scheme for large
deformation of beams has been developed by Gendy and Saleeb [13].
Bathe [4] has briefly reviewed the different approaches used for obtaining the solution
of general nonlinear problems. Total and updated Lagrangian descriptions of motion have been
used to formulate the incremental equations of motion. It has been noticed, for the problems

4
solved, that the two formulations give identical results, provided appropriate constitutive
relations are used. One of the areas of controversy in analyzing large deformation is the choice
of appropriate objective stress measures. Various objective stress measures have been discussed
by Bathe [4] and Chakrabarty [14]. These measures have been discussed in detail in section
1.1.4. The classical theory of elasto-plasticity assumes that the incremental strain tensor can be
additively decomposed into the elastic and plastic parts. While this may be true for small
incremental deformation, this assumption is no longer valid for large incremental strains and
rotations, unless appropriate strain measures are used [15] [16].
1.1.2 Dynamic Finite Element Analysis
Bathe et al [17] were probably the first to propose a large deformation dynamic finite element
formulation. Total and updated Lagrangian descriptions of motion were used to formulate the
incremental equations of motion. A detailed finite element analysis of nonlinear static and
dynamic response was also carried out by Mondkar and Powell [18]. A Lagrangian frame of
reference is used and a variational form of the incremental equations of motion for large
displacement, finite strain deformation is incorporated in the finite element program. It is found
that no specific time integration scheme is optimal for all types of nonlinear behavior. It is
inferred that for large load steps it is essential to use iterations to obtain accurate results.
Direct integration methods are the most commonly used methods for solving the
discretized and linearized governing equations of motion. Dynamic equilibrium including the
effect of damping forces is sought at discrete time instants instead of at all times. The advantage
is obvious: all solution techniques employed in static analysis can probably be used most
effectively in direct integration. The selection of any one direct integration method depends
upon considerations of accuracy, stability and cost. The importance of using equilibrium
iterations in dynamic nonlinear analysis was first realized by Bathe [4].
The Newmark implicit scheme has been employed to integrate the semi-discrete
equations of motion.
1.1.3 Contact Analysis
The analysis of contact has been a fascinating area since the initial days of the study of
mechanics itself. A contact system may contain only one or more than one contact body, and

5
contact impact may occur between two bodies or within a single body. The behavior of the
contact system is governed by three main groups of equations, i.e. the equation of motion,
constitutive equations and initial and boundary conditions. The boundary conditions can be
classified into two categories: prescribed boundary conditions and unknown boundary
conditions. The prescribed boundary conditions are deformation dependent. Among other
possible deformation dependent boundary conditions, contact-impact conditions are an
important class. Many Investigators tried to determine the stresses and displacements in two
elastic bodies coming in contact. In three dimensions, the problem of two contacting elastic
bodies was first formulated and solved by Hertz [19] under several restrictive assumptions.
Hertz assumed that the geometry of general curved surface in the vicinity of contact can be
described by quadratic terms only. Neglecting friction and assuming that the bodies in contact
deform as though they were elastic half spaces, Hertz used the Boussinesq solution to deduce
that the contact pressure distribution has to be elliptical to produce an elliptical area. For the
contact between two spheres, the contact area is circular. Research work based on Hertz’s
theory up to 1982 has been reviewed by Johnson [20]. Although Hertz evaluated the dimensions
of contact area and stresses at the contact surface, he presented only a speculative sketch of the
subsurface stresses. The stresses beneath the contact surface were analyzed later. For circular
contact area of radius a, it is observed that the maximum shear stress occur at a depth of 0.57a
[20].
With the development of computers, people resorted to numerical methods. Finite
element method has been the most widely used. In earlier studies, contacting bodies were
assumed to consist of linearly elastic materials and hence were assumed to undergo only small
deformations. The loading was assumed to be proportional. The node-to-node contact interface
model was used for discretizing the contacting bodies and the contact conditions (constraints)
involving nodal displacements and forces were applied by modifying the combined stiffness
matrix or the flexibility matrix of the contacting bodies. The investigators who used this model
are Ohte [21], Gaertner [22], Francavilla and Zienkiewicz [23], Sachdeva and Ramakrishnan
[24] etc.
Kalker and Randen [25], Hung and de Saxce [26] and Mahmoud et al. [27] used the
technique of mathematical programming for solving the frictionless small deformantion elastic
contact problem. They minimized the total strain energy to find the contact area and the contact
pressure distribution.
When the deformation is large, there is a possibility that a pair of nodes presently in
contact, may undergo a large relative displacement. If a node-to-node interface model is used,

6
re-meshing of the bodies become necessary for carrying out the subsequent analysis. To avoid
frequent re-meshing, one must use the node-to-segment interface model, in which one node of
a contacting body comes in contact with segment of the other. When a node-to-segment
interface model is used, the contact conditions acquire a complex form, when expressed in terms
of nodal variables. To impose these conditions, an additional set of finite element equations,
involving the contact stiffness matrix, needs to be developed using the principle of virtual work
of the contact forces. The methods which are normally used for implementing the contact
constraints are Lagrange multiplier method, penalty function method and transformation matrix
method. A survey of contact algorithms is given in Bohm [28] and Bourago [29].
Kikuchi and Oden [30] have provided the mathematical foundation for the Lagrange
multiplier method. The work of Bathe and Chaudhary [31] seems to be the first attempt to
develop an expression for contact stiffness matrix using this method. This was done for planar
and axisymmetric problems assuming the contact segment to be straight. Lagrange multiplier
was used to impose the contact constraints including friction forces based on Coulomb’s friction
law. Non-linear contact kinematics was used for developing a consistent contact matrix for two
dimensional problems by Wriggers and Simo [32] and by Parisch [33]. They used both the
Lagrange multiplier method and penalty function method. The transformation matrix method
for two dimensional problems was proposed by Chen and Yeh [34]. A formulation for non-
linear frictional model was proposed by Gallego and Anza [35] using perturbed Lagrangian
functional. All the above formulations were used for static large deformation elastic contact
problems. Some of these formulations were extended to elasto-dynamic contact problems [4]
[32] and dynamic elasto-plastic problems [36]. A literature survey of contact dynamics
modelling is presented by Gilardi and Sharf [37]. Sung and Kwak [38] and Bittencourt and
Creus [39] have presented the analysis of dynamic, large deformation contact problem
including the effects of friction. Master’s thesis by Werner [40] is an attempt to compare the
experimental results and the results predicted by a commercial finite element software for an
elasto-plastic impact problem.
Contact-impact algorithms, which are sometimes called as slideline algorithms, are the
computationally time-consuming parts of many explicit simulations of non-linear problems,
because they involve so many branches, and thus are not amenable to vectorization, which is
essential for speed on supercomputers. Belytschko and Neal [41] introduced the pinball
algorithm, which is a simplified slideline algorithm and is readily vectorized. Its major idea is
to embed pinballs in surface elements and to enforce the impenetrability condition only to these
pinballs. It can be implemented either by a Lagrange multiplier or penalty method. Malone and

7
Johnson [42] developed a parallel contact algorithm and implemented it in a non-linear dynamic
explicit finite element program to analyze the three-dimensional shell structures. The contact
algorithm accounted for initial contact, sliding and release through the use of parametric
representation of motion of points located on the surface of structure combined with a contact
surface representation, which approximates the actual surface by means of triangular search
planes. Brown et al. [43] described a general strategy for parallelizing solid mechanics
simulations. Such simulations often have several computationally intensive parts, including
finite element integration, detection of material contacts and particle interactions if smoothed
particle hydrodynamics is used to model highly deforming materials. Their strategy is to load-
balance each of the significant computations independently with whatever balancing technique
is most appropriate. The main advantage is that each computation can be scalably parallelized.
The drawback is the data exchange between the processors and the extra coding the must be
written to maintain multiple decompositions in a code.
The problems of undesirable oscillatory solutions and stability of time-integration
schemes have been tackled by quite a few people. Taylor and Papadopoulos [44] address the
problem of oscillatory solution along the contact interface. The standard second order implicit
integrators of the Newmark family, used in earlier works, were investigated by them. In their
work, control of these oscillations is attempted by introducing an artificial bulk viscosity for
the compressive waves in each body. Laursen and Chawla [45] proposed an energy conserving
algorithm to control the oscillations for the frictionless dynamic contact problems. It is shown
that a Lagrange multiplier enforcement of an appropriate contact rate constraint produces these
conservation properties. In particular, the ability of the formulation to produce accurate results,
where more conventional integration schemes fail, is emphasized by numerical simulations.
Armero and Petocz [46] presented a new dissipative time-stepping algorithm to improve the
stability of time integration scheme. They showed that although the traditional time-integration
schemes were stable for most of the linear problems, instability would creep in as soon as we
try to use them in non-linear cases. Given these considerations, they developed implicit time-
stepping algorithms for contact problems that possess unconditional (energy) stability in time
and lead stable enforcement of contact constraints.
Quite a few researchers have proposed new techniques. Ayari and Saouma [47]
developed a new formulation for contact problems using the concept of fictitious forces.
Contrary to the most existing models, this one involves very few matrix decompositions and
thus is computationally inexpensive. The model was applied to fracture mechanics based
analysis of a cracked dam under seismic excitation. They also showed that their model could

8
be applied to a vast majority of contact problems and was not only restricted to fracture
phenomenon. Simo and Laursen [48] extended the augmented Lagrangian formulation, which
has certain advantages over traditional Lagrangian formulation, to problems involving frictional
contact. Kane et al. [49] proposed a nonsmooth contact algorithm, which is capable of dealing
with multibody nonsmooth contact geometries for which neither the normals nor the gap
functions can be defined. Farahani et al [50] presented a solution method for contact/impact
problems, which is different from Lagrange multiplier and penalty methods. This method is
based on the stiffness matrix transformation and eliminating the normal degree of freedom of
contactor node. The method is absolutely general and can be used in static as well as dynamic
non-linear problems. In this case, the number of unknowns does not increase and the solution
scheme needs no modification as is the case with prior methods. Besides, the boundary
conditions are satisfied exactly.
1.1.4 Objective Stress Measures
In large deformation elasto-plastic problem, one comes across large rotations also besides large
displacements and strains. In updated Lagrangian formulation, this creates difficulty in updating
of stress components. Note that the constitutive equation gives the stress increment only due to
incremental straining of the material. If the Cauchy stress tensor is used as the stress measure,
then one cannot simply add the incremental stress to the Cauchy stress components at time t to
obtain their values at time t + ∆t. This is because, the change in Cauchy stress components also
occur due to the incremental rotation. Thus, the Cauchy stress tensor is not objective in the
sense that its components do not change even when the increment consists of pure rigid rotation.
The stress measures whose increment (or rate) depends only on the incremental deformation
(or rate of deformation) are called as objective stress (or stress rate) measures.
The most commonly used objective stress measure in the literature as well as in finite
element software on large deformation elasto-plastic problems is the Jaumann stress rate
measure. It is obtained by eliminating the rate of change of Cauchy stress component due to
rotation from the total rate. But it uses only the infinitesimal rotation tensor to represent the
incremental rotation. Therefore, when the incremental rotation is large, it is not expected to give
accurate results. Further, it is well known that the use of Jaumann stress rate measure predicts
an oscillatory solution when the incremental shear is large [51]. Other objective stress rate
measures like the Green-Naghdi rate and E-rate have been proposed [51] [52]. But they are
difficult to use. A simple stress updating procdure has been proposed by Varadhan [53] which

9
accounts for the change in Cauchy stress components due to incremental rotation. In this
approach, the incremental stress is evaluated in a frame which rotates with the particle rather
than in a fixed frame. While transforming the stress components from the material frame to the
fixed at time t + ∆t, the transformation due to incremental rotation is taken into account.
1.2 Objective of the Present work
The finite element method in ANSYS APDL uses the Cauchy stress rate measure and
logarithmic strain rate measure for large deformation problems. The objective of this work is
to study the large deformation, elasto-plastic, contact-impact problems using ANSYS APDL.
Convergence of the program of this commercial FEM software is observed for each of the
specific problem. This is done for the following four cases:
1. Large Deformation, Dynamic, Elastic, Single Body Analysis
2. Large Deformation, Elastic, Contact-Impact Analysis
3. Large Deformation, Dynamic, Elasto-Plastic, Single Body Analysis
4. Large Deformation, Elasto-Plastic, Contact Analysis

10
Chapter 2
Mathematical Modeling of Dynamic Elasto-
Plastic Problem
In a dynamic elasto-plastic problem, the deformations and the strains in the body no longer
remain small and hence the response of the body is not proportional to the load. The response
of the body becomes non-linear. The non-linearities can be broadly classified into three:
geometric non-linearity, material non-linearity and boundary non-linearity. First two of these
non-linearities occur in the analysis of large deformation of elasto-plastic materials. The most
convenient formulation for such analysis is the Updated Lagrangian formulation with the
constitutive relations represented in an incremental manner. The third non-linearity occurs in
the analysis of contact-impact problems and this is attributed to the non-linear boundary
conditions at the contact surfaces.
2.1 Introduction
In the study of the deformation of a body subjected to external loading, often the original,
undeformed and unstressed state of the body is used for the formulation of its governing
equations. This method is known as the Lagrangian formulation. This formulation is convenient
for small deformation problems, where the deformed configuration does not deviate from the
original configuration significantly. Hence, the deformations are proportional to the loading and
can be described by an infinitesimal or linear strain tensor, for which the strain-displacement
relations are linear. On the other hand, for large deformation problems, one has to use a finite
strain measure, which is expressed by a non-linear strain-displacement relation. Furthermore,
the equations of motion, expressed in the original configuration, depend on the deformation.
This makes the governing equations highly complex and too difficult to solve. In such cases, it
is convenient to solve the problem in an incremental manner known as the Updated Lagrangian
Formulation. In this formulation, the state of the body is updated in increments using the
calculated incremental deformations and stresses. The response during ∆t is calculated based
on the configuration at time t, where ∆t represents the increment in time, called time step. Hence
all the relations are to be in an incremental form rather than the total form. This methodology

11
is particularly useful for elasto-plastic materials, since the constitutive relations for these
materials depend on the state of the body and therefore they are expressed in an incremental
form.
2.2 Kinematics of Finite Deformation
The relative deformation gradient tensor at time t +∆t is defined by:
t + ∆t
t
𝐹𝑖𝑗 =
𝜕 𝑡+ ∆𝑡
𝑥𝑖
𝜕 𝑡 𝑥𝑗
=
t + ∆t
t
𝑥𝑖,𝑗
(2.1)
where
t +∆t
𝑥𝑖 and
t
𝑥𝑗 denote the position vectors of the particle at time t and t + ∆t
respectively. When
t + ∆t
t
𝑭 is not singular, the polar decomposition theorem ( [54] [55]) allows
a decomposition of the form:
t + ∆t
t
𝐹𝑖𝑗 =
t + ∆t
t
𝑅𝑖𝑘
t + ∆t
t
𝑈 𝑘𝑗
(2.2)
Where
t + ∆t
t
𝑹𝑖𝑘 is an orthogonal tensor representing the material rotation and
t + ∆t
t
𝑼 𝑘𝑗 is a
positive definite symmetric tensor called the right stretch tensor. The right stretch tensor can be
diagonalized by the following transformation to obtain the principle stretches
t + ∆t
t
𝜆𝑖:
t + ∆t
t
𝑈𝑖𝑗
𝑃
=
t + ∆t
t
𝑄𝑖𝑘
t + ∆t
t
𝑈 𝑘𝑙
t + ∆t
t
𝑄𝑗𝑙
(2.3)
Where
t + ∆t
t
[𝑄] is orthogonal matrix and
t + ∆t
t
𝜆𝑖 =
t + ∆t
t
𝑈𝑖𝑖
𝑃
(𝑛𝑜 𝑠𝑢𝑚)
(2.4)
The tensor
t + ∆t
t
𝑹 and the transformation matrix,
t + ∆t
t
[𝑄] are used in the development of the
stress measure while the incremental logarithmic strain tensor is obtained from the
t + ∆t
t
𝜆𝑖.

12
2.3 Stress Measures
The large deformation problems are normally associated with rigid body rotations. Hence, the
stress measures used in these problems should be objective, which means that they should
vanish if the increment is a pure rigid body rotation. The most commonly used stress measure
known as Cauchy stress tensor is not objective. There are numerous objective stress measures,
of which some are discussed below.
One of the commonly used objective stress measure is the Second Piola-Kirchoff stress
tensor
t + ∆t
t
𝑆𝑖𝑗 which is related to Cauchy stress tensor
t +∆t
𝜎 𝑚𝑛 using the concept of
equivalent work between the two configurations [4]:
t + ∆t
t
𝑆𝑖𝑗 =
t
𝜌
t +∆t
𝜌
t
t + ∆t
𝑥𝑖,𝑚
t +∆t
𝜎 𝑚𝑛
t
t + ∆t
𝑥𝑗,𝑛
(2.5)
Here
t
𝜌 and
t +∆t
𝜌 represent the densities at time t and t + ∆t respectively,
t
t + ∆t
𝑥𝑖,𝑚 is the
derivative of the position vector at time t, viz.,
t
𝑥𝑖 with respect to
t +∆t
𝑥 𝑚 .
Another frequently used objective stress measure is Jaumann’s stress rate tensor 𝜎𝑖𝑗. It
is related to the Cauchy rate 𝜎𝑖𝑗̇ by the relation [4]:
t
𝜎𝑖𝑗Δ𝑡 = 𝜎𝑖𝑗̇ Δ𝑡 −
t
𝜎𝑖𝑘 (
t
Ω𝑗𝑘Δ𝑡) −
t
𝜎𝑗𝑘 (
t
Ω𝑖𝑘Δ𝑡)
(2.6)
t
Ω𝑗𝑘Δ𝑡 =
1
2
(
t
Δ𝑢𝑖,𝑗 −
t
Δ𝑢𝑗,𝑖)
(2.7)
Here,
t
Ω𝑗𝑘Δ𝑡 represents the incremental spin tensor and
t
Δ𝑢𝑖,𝑗 denotes the derivative of the
incremental displacement vector
t
Δ𝑢𝑖 with respect to
t
𝑥𝑗.
Dienes [51] proposed a stress rate 𝜎̂ based upon the rotation part of the deformation
gradient:
𝜎̂𝑖𝑗 = 𝜎̇𝑖 𝑗 − Ω𝑖𝑘 𝜎𝑘𝑗 + Ω 𝑘𝑗 𝜎𝑖𝑘 (2.8)
Where,

13
Ω𝑖𝑗 = 𝑅̇ 𝑖𝑘 𝑅𝑗𝑘 (2.9)
Is the angular velocity. The stress rate 𝜎̂ is sometimes referred to as the Green-Naghdi rate.
2.4 Strain Measures
The linear of infinitesimal strain tensor, normally used in the small deformation problems, is
not a valid choice for the large deformation, elasto-plastic analysis. Green Lagrange Strain
tensor is one of the most commonly used strain measure for large deformation analysis.
The incremental Green-Lagrange strain tensor is a non-linear function of the
displacement [4]:
t
Δ𝑒𝑖𝑗 =
1
2
(
t
Δ𝑢𝑖,𝑗 +
t
Δuj,i +
t
Δuk,i
t
Δuk,j)
(2.10)
When the incremental deformation is small, the incremental linear strain tensor
t
Δεij =
1
2
(
t
Δui,j +
t
Δuj,i)
(2.11)
can be used as the measure of deformation.
These two strain measures occur in the virtual work expression at t + ∆t and its
transformation to time t respectively. The components of the Green Lagrange strain tensor are
invariant under rigid body rotation of the material unlike the small strain components. The
following are the disadvantages one would face using one of the above strain measures in a
constitutive law:
1. The solution obtained is dependent upon the size of the increment in the updated
Lagrangian formulation unless the increment size is sufficiently small.
2. The components of the strain tensors do not tend to infinite values when the principal
stretches tend to zero. Therefore a constitutive law which ensures that the appropriate
Cauchy stress components tend to negative infinity (as is physically realistic), even
though the strain components remain finite, should be used. This difficulty can be
avoided by using a strain measure whose components become minus infinity when the
principal stretches become zero.
The Logarithmic Strain measure introduced by Dienes [51] is free of the above disadvantages.
The components of the incremental logarithmic strain tensor, in the material frame, are defined
by:

14
t
Δϵij
𝐿
= ln (
t + ∆t
t
𝜆𝑖) 𝛿𝑖𝑗 (𝑛𝑜 𝑠𝑢𝑚 𝑜𝑣𝑒𝑟 𝑖)
(2.12)
Where the
t + ∆t
t
𝜆𝑖 are the principal stretches defined by equation (2.4). The components with
respect to any other frame can be obtained by the usual transformation law. The logarithmic
strain has the following additional advantage in elasto-plastic analysis. A loading test involving
elasto-plastic deformation followed by elastic unloading reveals that the slope of the elastic
unloading line is the same as that of the initial elastic loading line only when the true stress and
the logarithmic strain measures are used in a constitutive law.
In ANSYS, the plasticity constitutive models are applicable in both small-deformation and
large-deformation analyses. For small deformation, the formulation uses engineering stress and
strain. For large deformation (NLGEOM, ON), the constitutive models are formulated with the
Cauchy stress and logarithmic strain.
2.5 Theory of Plasticity
An Elasto-plastic material is characterized by the following:
1. Yield criterion
2. Strain hardening postulate
3. Flow rule
Baushchinger effect and the hysteresis loop are disregarded and the material is assumed to be
isotropic.
2.5.1 Yield Criterion
A law defining the limit of elastic behavior under any possible combination of stress is called
yield criterion. This law applies both to the loading directly from the annealed state and to
reloading of an element unloaded from a previous plastic state. The yield criterion is a scalar
function of the stress and internal variables and is given by the general function:
𝑓(𝜎, 𝜉) = 0 (2.13)
Where 𝜉 represents a set of history dependent scalar and tensor internal variables.

15
Equation (2.13) is a general function representing the specific form of the yield criterion for
each of the plasticity models. The function is a surface in stress space and an example, plotted
in principal stress space, as shown in this figure:
Figure 2.1 Yield surface in principal stress space
Stress states inside the yield surface are given by 𝑓(𝜎, 𝜉) < 0 and result in elastic deformation.
The material yields when the stress state reaches the yield surface and further loading causes
plastic deformation. Stresses outside the yield surface do not exist and the plastic strain and
shape of the yield surface evolve to maintain stresses either inside or on the yield surface.
The von Mises yield criterion which has been experimentally verified to predict the yield locus
quite accurately is used. Yielding occurs when the equivalent stress
t
𝜎𝑒𝑞 reaches the value of
the yield stress in uniaxial tension
t
𝜎 𝑦.
t
𝜎𝑒𝑞(𝜎𝑖𝑗) =
t
𝜎 𝑦
(2.14)
Where,
t
𝜎𝑒𝑞 = (
3
2
t
𝜎𝑖𝑗
′ t
𝜎𝑖𝑗
′
)
1/2 (2.15)
And
t
𝜎𝑖𝑗
′
is the deviatoric part of
t
𝜎𝑖𝑗.

16
2.5.2 Strain Hardening Postulate
The size and the shape of the yield surface in a strain hardening material depends upon the
complete history of plastic deformation since the previous annealing. A convenient
mathematical formulation for strain hardening is obtained by assuming that the yield surface
expands uniformly without changing its shape. This is known as isotropic hardening and the
relationship may be stated as
t
𝜎 𝑦 = 𝑓 (
t
𝜀 𝑒𝑞
𝑝
)
(2.16)
Where,
t
𝜀 𝑒𝑞
𝑝
= ∫ √
2
3
𝑑𝜀 𝑒𝑞
𝑝𝐿
𝑑𝜀 𝑒𝑞
𝑝𝐿
𝑝𝑎𝑡ℎ
(2.17)
Is the equivalent plastic strain determined from the path-line integration of an invariant of
t
𝑑𝜀 𝑒𝑞
𝑝𝐿
(plastic part of the incremental logarithmic strain tensor) and the function f is determined by the
relationship between the true stress and the plastic strain in uniaxial tension/compression. In
this project the following power law was used:
t
𝜎 𝑦 −
0
𝜎 𝑦 = 𝐾 (
t
𝜀 𝑒𝑞
𝑝
)
𝑛
(2.18)
Where
0
𝜎 𝑦 denotes the initial yield stress of the material.
In ANSYS, this law was used in the form of multi-linear isotropic hardening. The multi-linear
hardening behavior is described by a piece-wise linear stress-total strain curve, starting at the
origin and defined by sets of positive stress and strain values. The first stress-strain point
corresponds to the yield stress. Subsequent points define the elastic plastic response of the
material.

17
Figure 2.2 Multi-linear Isotropic Hardening
2.5.3 Plastic Flow Rule
The rule of plastic flow relates the current stress with the plastic strain increment. The plastic
potential is a function defining the ratios of the components of the incremental plastic strain.
The associated flow rule which assumes that the yield function and the plastic potential are the
same is used. It states that
t
𝑑𝜀𝑖𝑗
𝑝𝐿
= 𝑑𝜆
𝑑𝐹
𝑑
t
𝜎𝑖𝑗
(2.19)
where 𝑑𝜆 is a scalar to be found from the hardening relation. These flow rules are typically used
to model metals and give a plastic strain increment that is proportional to the stress increment.
If the plastic potential is not proportional to the yield surface, the model has a non-associated
flow rule, typically used to model soils and granular materials that plastically deform due to
internal frictional sliding. For non-associated flow rules, the plastic strain increment is not in
the same direction as the stress increment.

18
Chapter 3
Results and Discussion
The objective of this report is to carry out the study of large deformation elasto-plastic
contact impact problems using ANSYS (APDL). This is done for the following four cases:
Brick 8-node element SOLID 185 was used to model the problems.
3.1 Large Deformation, Dynamic, Elastic, Single
Body Analysis
3.1.1 Dynamic Response of a Cantilever
Figure 3.1 Cantilever: geometry and boundary conditions

19
A benchmark problem solved by Bathe et al. [17] has been chosen for the large
deformation dynamic elastic single body analysis. Figure 3.1 shows the geometry and boundary
conditions of the cantilever analyzed. The geometric and material parameters used are:
 Length of the Cantilever L = 10 inches
 Height h = 1.0 inch
 Width b = 1.0 inch
 Young’s modulus E = 1.2 × 104
lb/in2
 Poisson’s ratio ν = 0.2
 Density ρ = 1.0 × 10-4
lb-sec2
/in4
Mesh and Time-step Convergence studies
A mesh and time-step convergence study is essential to ascertain that a proper mesh and
time-step are selected so that a correct solution is obtained with proper utilization of
computational resources. First a mesh convergence study is carried out to select an appropriate
mesh. As the problem is dynamic one, so the time-step convergence study is carried out next to
select a proper time step.
1. Mesh Convergence: Following are the discretizations used for the mesh convergence
study. The time step used is 1.35 × 10-4
sec and the convergence tolerance of 1% is used.
a. Discretization No. 1:
i. Number of elements: 80 (20×2×2)
ii. Degrees of freedom: 567
b. Discretization No. 2:
From the mesh convergence study, it is found that the second discretization gives more
accurate result. So, for final study it is selected.
2. Time-step Convergence: Following are the time steps considered for the time-step
convergence study. The discretization used is (40×4×4) and the convergence tolerance
used is 1%.
a. Time-step 1 = 1.35 × 10-04
sec
b. Time-step 2 = 0.45 × 10-04
sec
It is found that both the time steps give approximately same results. For the final study,
the second time-step is selected.

20
The result of the analysis is shown in Figure 3.2 and Figure 3.3.
Figure 3.2 Dynamic response of the Cantilever at ∆t = 0.45 × 10-04
sec
Figure 3.3 Maximum Equivalent Stress response at ∆t = 0.45 × 10-04
sec

21
3.2 Large Deformation Elastic Contact-Impact
Analysis.
3.2.1 Impact of Two Identical Bars
Figure 3.4 Impact of two identical bars
This problem is solved by Belytschko and Neal [41]. Figure 3.4 shows the initial
configuration of the bars along with the boundary conditions. One of the bars is given an initial
velocity so that it impacts with the second bar. The material properties are so chosen that the
wave speed in each of the bars is 10.0 m/sec. Since c =√
Ε
𝜌
, the material properties chosen are:

22
 Young’s modulus E = 10 N/m2
 Poisson’s ratio ν = 0.0
 Density ρ = 0.1 Kg/m3
The geometric properties of each bar are as follows:
 Length of the bar L = 10.0 m
 Thickness = 1.0 m
 Width = 1.0 m
The initial velocity of the impacting bar i.e. bar A is 1.0 m/sec. As Newmark’s scheme is used
to calculate the velocities and accelerations, it has to be made sure that the hitting body travels
with zero acceleration before the impact. In order to achieve this the initial gap between the two
bars has been kept at 0.1 m and the contact search length has been chosen as 0.11 m.
Mesh and Time-step Convergence Studies:
Mesh Convergence: Following are the discretizations used for the mesh convergence
study. The time-step used is 0.1 sec and the convergence tolerance of 1% is used.
a. Discretization No. 1:
b. Discretization No. 2:
The study shows that the second discretization gives better results. So, it is chosen for
further study.
2. Time-step Convergence: Following are the time-steps considered for the time-step
convergence study. The discretization used is (20×1×1) and the convergence tolerance
used is 1%.
a. Time-step 1 = 0.2 sec
b. Time-step 2 = 0.1 sec
The time-step convergence study shows that time-step 2 gives better result. So it is
chosen for further study.
Figure 3.5 and Figure 3.6 give the velocity time histories at the interface of first bar (Z = 10.1),
as well as the second bar (Z = 10.0). Figure 3.7 and Figure 3.8 give the velocity time histories
at the midpoint of bar A and bar B respectively.

23
Figure 3.5 Velocity at interface of bar A
Figure 3.6 Velocity at interface of bar B

24
Figure 3.7 Velocity at midpoint of bar A
Figure 3.8 Velocity at midpoint of bar B

25
3.2.2 Oblique Impact of Two Infinite Blocks
Figure 3.9 Oblique Impact of two Elastic Blocks
The second problem used for comparison is the oblique impact of two infinite elastic blocks
(plane strain problem) shown in Figure 3.9. This problem is reported in reference [46]. The top
block is given a uniform initial velocity, Vo = (-10, 0, -10) units (in the coordinate system shown
in figure). The geometric and material properties are:
 Hitting block
o Length = 5 units
o Height = 4 units
 Target block
o Length = 16 units
o Height = 6 units
 Young’s modulus (E) = 1000 units, for both the blocks.

26
 Poisson’s ratio (ν) = 0.0, for both the blocks
 Density (ρ) = 0.1 units, for both the blocks
Since the Poisson’s ratio is zero, the plane strain and the plane stress formulations will give the
same results. The problem is solved with only one element in Y-direction. To overcome the
numerical difficulty arising due to low stiffness in Y-direction, the displacement along this
direction is fixed at all nodes for both the bodies.
The following are the discretizations and the time-step considered:
 Discretization:
o Number of elements in the hitting block = 20 (5×1×4)
o Number of elements in the target block = 96 (16×1×6)
o Number of d.o.f in the hitting block = 180
o Number of d.o.f in the target block = 714
 Time-step = 0.01 units
There are two different cases. In the first case, there is no friction at the contact interface and in
the second case, friction coefficient (µ) between the hitting and the target bodies is 0.4.
The X and Z displacements at the point P are plotted against the time for both the
frictional and frictionless cases. Both the displacements and the time are measured from the
instant of contact between the two blocks.
Figure 3.10 X-displacement of point P (frictional case)

27
Figure 3.11 X-displacement of point P (frictionless case)
Figure 3.12 Z-displacement of point-P (frictional case)

28
Figure 3.13Z-displacement of point P (frictionless case)

29
3.3 Large Deformation Dynamic Elasto-Plastic
Single Body Analysis
3.3.1 Dynamic Elasto-Plastic Analysis of a Cube
Figure 3.14 Cube: geometry and boundary conditions
The problem of a cube subjected to uniaxial step load and undergoing large deformation has
been selected for single body elasto-plastic analysis. The geometry and boundary conditions
are shown in the Figure 3.14. The geometric and the material properties are:
 Side of the cube (L) = 30 mm
 Young’s modulus (E) = 1114.58 N/mm2
 Poisson’s ratio (ν) = 0.36
 Density (ρ) = 2000 × 10-09
Kg/mm3
 Hardening Coefficient (K) = 87.42 N/mm2
 Hardening exponent (n) = 0.909
 Initial yield stress (0
σy) = 44.58 N/mm2
The discretization used is:
 Number of elements = 1 (1×1×1)

30
 Number of degrees of freedom = 24
The convergence tolerance used is 1%. Time step used is 0.3 × 10-03
sec. The equivalent stress
is plotted against the equivalent plastic strain in the. The actual hardening law is:
t
𝜎𝑒𝑞 =
0
𝜎 𝑦 + 𝐾(𝜖 𝑒𝑞
𝑝
)
𝑛
Figure 3.15 Elasto-plastic response of the cube

31
3.3.2 Dynamic Elasto-Plastic Response of a fixed Square Plate
Figure 3.16 Plate: geometry and boundary conditions
The impact response of a square plate fixed at all its edges and characterized by the elasto-
plastic behavior is obtained. Figure 3.16 shows the geometry and boundary conditions of the
plate. Only one quarter of the plate has been modelled due to symmetry along X and Y axes.
All specified quantities refer to the quarter plate model. The geometric and material data for the
plate are:
 Side of the plate (L) = 125 mm
 Thickness of the plate (h) = 25 mm
 Young’s modulus (E) = 1114.58 N/mm2
 Hardening coefficient (K) = 87.42 N/mm2
 Initial yield stress (0
σy) = 44.58 N/mm2
 Density of the plate material (ρ) = 2000 × 10-09
Kg/mm3
Mesh and Time-step Convergence Study
The mesh and time-step used are given below:

32
1. Mesh Convergence: The time-step used is ∆t = 0.3 × 10-03
sec and the convergence
tolerance used is 1%. The discretizations used are:
a. Discretization 1:
i. Number of elements = 200 (10×10×2)
ii. Number of degrees of freedom = 1089
b. Discretization 2:
i. Number of elements = 1600 (20×20×4)
ii. Number of degrees of freedom = 6615
From the mesh convergence study, it is found that the second discretization gives better
results. So it is chosen for the final study.
2. Time Convergence: The time-steps used are stated below. The mesh used is (20×20×4)
and the convergence tolerance used is 1%.
a. Time-step 1: 0.6 × 10-03
sec
b. Time-step 2: 0.3 × 10-03
sec
It is found that both the time-steps give nearly the same result. The second time step is
used for the study. Maximum displacement ratio for the plate is plotted over time as
shown in Figure 3.17. Figure 3.18 shows the maximum equivalent stress response of
the plate. Relationship between equivalent stress and equivalent plastic strain was also
analyzed and plotted as shown in Figure 3.19.
Figure 3.17 Maximum displacement response of the plate

33
Figure 3.18 Maximum Equivalent Stress response of the plate
Figure 3.19 Equivalent Stress v/s Equivalent Plastic Strain

34
3.4 Large Deformation Elasto-Plastic Contact-
Impact Analysis
3.4.1 Elasto-Plastic Impact of a Sphere on a Plate
Figure 3.20 Impact of Sphere on Plate
The problem of an impact of a sphere with a plate is selected for elasto-plastic contact-impact
analysis (see Figure 3.20). The sphere is given an initial velocity, V = (10, 0, -10), in the
coordinate system shown in figure. The plate is fixed along the edges. The coefficient of friction
used is µ = 0.2.
The geometric details of the sphere and plate are shown in Figure 3.20. The material
properties assumed for both the bodies are as follows:
 Young’s modulus (E) = 4 × 105
units

35
 Density (σ) = 4 units
 Initial yield stress ((σy)0) = 5000 units
 Hardening Coefficient (K) = 40000 units
Time step used for the analysis is 0.002 units. The mesh details of the sphere and plate are as
follows:
 Number of elements for plate = 1024 (16×16×4)
 Number of degrees of freedom for sphere = 2991
 Number of degrees of freedom for plate = 4335
Figure 3.21 Finite element plot of the sphere
The X and Z displacements of the bottom node of the sphere are shown in the Figure 3.22 and
Figure 3.23 for the frictionless case and in the Figure 3.24 and Figure 3.25 for the frictional
case.
The analysis was repeated for sphere and ball made of polycarbonate material with
similar geometry and mesh details. The material properties of this material are given below:

36
 Young’s modulus (E) = 1.11458 × 109
N/m2
 Density (ρ) = 2000 kg/m3
 Initial yield stress = 4.458 × 107
N/m2
 Hardening Coefficient (K) = 8.742 × 107
N/m2
The velocity given to the sphere is (10,0,-10) m/s, in the coordinate system shown in Figure
3.20 and the coefficient of friction at the contact interface is 0.3, for the case of analysis with
friction. The time step used in this analysis is 0.0003 seconds. The X and Z displacements of
the bottom node of the sphere for the frictionless case are shown in the Figure 3.26 and Figure
3.27. The results for the frictional case are shown in the Figure 3.28 and Figure 3.29.
Figure 3.22: Displacement of the sphere along X-direction for frictionless case.

37
Figure 3.23: Displacement of the sphere along Z-direction for frictionless case.
Figure 3.24: Displacement of the sphere along X-direction for frictional case.

38
Figure 3.25: Displacement of the sphere along Z-direction for the frictional case.
Figure 3.26: Displacement of the sphere along X-direction (polycarbonate material)
frictionless case.

39
Figure 3.27: Displacement of the sphere along Z-direction (polycarbonate material)
frictionless case.
Figure 3.28: Displacement of the sphere along X-direction (polycarbonate material) frictional
case.

40
Figure 3.29: Displacement of the sphere along Z-direction (polycarbonate material) frictional
case.

41
Chapter 4
Conclusions and Results
The objective of this work is to study the large deformation, elasto-plastic, contact-impact
problems using ANSYS APDL. Convergence of the program of this commercial FEM software
is observed for each of the specific problem. This is done for the following four cases:
The results are obtained for nodal displacement, nodal velocities, nodal accelerations and
maximum stress. All the results were plotted in time domain.

42
References
[1] R. Hill, The mathematical theory of plasticity, Oxford University Press, 1950.
[2] L. E. Malvern, Introduction to the Mechanics of a Continuous Medium, Englewood Cliffs,
New Jersy: Prentice Hall Inc., 1969.
[3] O. C. Zienkiewicz, X. K. Li and S. Nakazawa, "Dynamic transient analysis by a mixed,
iterative method," International journal for numerical methods in engineering, vol. 23,
no. 7, pp. 1343-1353, 1986.
[4] K.-J. Bathe, Finite element procedures, New Delhi: Prentice Hall of India, 1996.
[5] T. J. Hughes, "A note on the stability of Newmark's algorithm in nonlinear structural
dynamics.," International Journal for Numerical Methods in Engineering, vol. 11, no. 2,
pp. 383-386, 1977.
[6] J. H. Argyris, "Elasto-plastic matrix displacement analysis of three-dimensional
continua," ROYAL AERONAUTICAL SOCIETY JOURNAL, vol. 69, pp. 633-636, 1965.
[7] J. C. Nagtegaal and J. E. De Jong, "Some computational aspects of elastic‐plastic large
strain analysis," International Journal for Numerical Methods in Engineering, vol. 17,
no. 1, pp. 15-41, 1981.
[8] M. A. Crisfield, Non-linear Finite Element Analysis of Solids and Structures, vol. 1,
Chichester: John Wiley and Sons, 1994.
[9] R. H. Dodds, "Numerical techniques for plasticity computations in finite element
analysis," Computers & Structures, vol. 26, no. 5, pp. 767-779, 1987.
[10] P. Jetteur, "Implicit integration algorithm for elastoplasticity in plane stress analysis,"
Engineering Computations, vol. 3, no. 3, pp. 251-253, 1986.
[11] S. Caddemi and J. B. Martin, "Convergence of the Newton‐Raphson algorithm in elastic‐
plastic incremental analysis," International journal for numerical methods in engineering,
vol. 31, no. 1, pp. 177-191, 1991.

43
[12] M. Ortiz and Egor Paul Popov, "Accuracy and stability of integration algorithms for
elastoplastic constitutive relations," International Journal for Numerical Methods in
Engineering, vol. 21, no. 9, pp. 1561-1576, 1985.
[13] A. S. Gendy and A. F. Saleeb, "Mixed Finite Element Modeling for the Dynamics of
Beam Assemblages Undergoing Large Overall Motions in Space," International Journal
of Computational Engineering Science, vol. 2, no. 2, pp. 309-338, 2001.
[14] J. Chakrabarty, Theory of Plasticity, New York: McGraw Hill, 1981.
[15] E. H. Lee, "Elastic-plastic deformation at finite strains," Journal of Applied Mechanics,
vol. 36, no. 1, pp. 1-6, 1969.
[16] I. Doghri, Mechanics of Deformable Solids, Springer, 2000.
[17] K. Bathe, E. Ramm and E. L. Wilson, "Finite element formulations for large deformation
dynamic analysis," International Journal for Numerical Methods in Engineering, vol. 9,
no. 2, pp. 353-386, 1975.
[18] D. P. Mondkar and G. H. Powell, "Finite element analysis of non‐linear static and
dynamic response," International journal for numerical methods in engineering, vol. 11,
no. 3, pp. 499-520, 1977.
[19] H. Hertz, "On the contact of elastic solids," J. reine angew. Math, vol. 92, no. 110, pp.
156-171, 1881.
[20] K. L. Johnson, "One hundred years of Hertz contact," Proceedings of the Institution of
Mechanical Engineers, vol. 196, no. 1, pp. 363-378, 1982.
[21] S. Ohte, "Finite element analysis of elastic contact problems," Bulletin of JSME, vol. 16,
no. 95, pp. 797-804, 1973.
[22] R. Gaertner, "Investigation of plane elastic contact allowing for friction," Computers &
Structures, vol. 7, no. 1, pp. 59-63, 1977.
[23] A. Francavilla and O. C. Zienkiewicz, "A note on numerical computation of elastic
contact problems," International Journal for Numerical Methods in Engineering, vol. 9,
no. 4, pp. 913-924, 1975.

44
[24] T. D. Sachdeva and C. V. Ramakrishnan, "A finite element solution for the two‐
dimensional elastic contact problems with friction," International Journal for Numerical
Methods in Engineering, vol. 17, no. 8, pp. 1257-1271, 1981.
[25] J. J. Kalker and Y. Van Randen, "A minimum principle for frictionless elastic contact
with application to non-Hertzian half-space contact problems," Journal of engineering
mathematics, vol. 6, no. 2, pp. 193-206, 1972.
[26] N. D. Hung and G. de Saxcé, "Frictionless contact of elastic bodies by finite element
method and mathematical programming technique," Computers & Structures, vol. 11, no.
1, pp. 55-67, 1980.
[27] F. F. Mahmoud, A. K. Al-Saffar and A. M. El-Hadi, "Solution of the non-conformal
unbonded contact problems by the incremental convex programming method,"
Computers & Structures, vol. 39, no. 1, pp. 1-8, 1991.
[28] J. Böhm, "A comparison of different contact algorithms with applications," Computers &
Structures, vol. 26, no. 1, pp. 207-221, 1987.
[29] N. G. Bourago, "A survey on contact algorithms," in International Workshop on Grid
Generation and Industrial Application, Moscow, 2002.
[30] N. Kikuchi and J. T. Oden, Contact problems in elasticity: a study of variational
inequalities and finite element methods, Philadelphia: SIAM, 1988.
[31] K. Bathe and Anil Chaudhary, "A solution method for planar and axisymmetric contact
problems," International Journal for Numerical Methods in Engineering, vol. 21, no. 1,
pp. 65-88, 1985.
[32] P. Wriggers and J. C. Simo, "A note on tangent stiffness for fully nonlinear contact
problems," Communications in Applied Numerical Methods, vol. 1, no. 5, pp. 199-203,
1985.
[33] H. Parisch, "A consistent tangent stiffness matrix for three‐dimensional non‐linear contact
analysis," International Journal for Numerical Methods in Engineering, vol. 28, no. 8,
pp. 1803-1812, 1989.
[34] W. Chen and J. Yeh, "Finite element analysis of finite deformation contact problems with
friction," Computers & structures, vol. 29, no. 3, pp. 423-436, 1988.

45
[35] F. J. Gallego and J. J. Anza, "A mixed finite element model for the elastic contact
problem," International Journal for Numerical Methods in Engineering, vol. 28, no. 6,
pp. 1249-1264, 1989.
[36] Z. H. Zhong, Finite element procedures for contact-impact problems, Oxford university
press, 1993.
[37] G. Gilardi and I. Sharf, "Literature survey of contact dynamics modelling," Mechanism
and machine theory, vol. 37, no. 10, pp. 1213-1239, 2002.
[38] J. H. Sung and B. M. Kwak, "Large displacement dynamic analysis with frictional contact
by linear complementarity formulation," Computers & structures, vol. 80, no. 11, pp. 977-
988, 2002.
[39] E. Bittencourt and G. J. Creus, "Finite element analysis of three-dimensional contact and
impact in large deformation problems," Computers & structures, vol. 69, no. 2, pp. 219-
234, 1998.
[40] T. C. Werner, "Elasto-Plastic Impact of a Cantilever Beam Using Non-Linear Finite
Elements and Event Simulation," 1998.
[41] T. Belytschko and M. O. Neal, "Contact‐impact by the pinball algorithm with penalty and
Lagrangian methods," International Journal for Numerical Methods in Engineering, vol.
31, no. 3, pp. 547-572, 1991.
[42] J. G. Malone and N. L. Johnson, "A parallel finite element contact/impact algorithm for
non‐linear explicit transient analysis: Part I—The search algorithm and contact
mechanics," International journal for numerical methods in engineering , vol. 37, no. 4,
pp. 559-590, 1994.
[43] K. e. a. Brown, "Parallel strategies for crash and impact simulations," Computer methods
in applied mechanics and engineering, vol. 184, no. 2, pp. 375-390, 2000.
[44] R. L. Taylor and P. Papadopoulos, "On a finite element method for dynamic
contact/impact problems," International Journal for Numerical Methods in Engineering,
vol. 36, no. 12, pp. 2123-2140, 1993.

46
[45] T. A. Laursen and V. Chawla, "Design of energy conserving algorithms for frictionless
dynamic contact problems," International Journal for Numerical Methods in
Engineering, vol. 40, no. 5, pp. 863-886, 1997.
[46] F. Armero and E. Petőcz, "A new dissipative time-stepping algorithm for frictional
contact problems: formulation and analysis," Computer methods in applied mechanics
and engineering, vol. 179, no. 1, pp. 151-178, 1999.
[47] M. L. Ayari and V. E. Saouma, "Static and dynamic contact/impact problems using
fictitious forces," International journal for numerical methods in engineering, vol. 32, no.
3, pp. 623-643, 1991.
[48] J. C. Simo and T. A. Laursen, "An augmented Lagrangian treatment of contact problems
involving friction," Computers & Structures, vol. 42, no. 1, pp. 97-116, 1992.
[49] C. e. a. Kane, "Finite element analysis of nonsmooth contact," Computer methods in
applied mechanics and engineering, vol. 180, no. 1, pp. 1-26, 1999.
[50] K. Farahani, M. Mofid and A. Vafai, "A solution method for general contact–impact
problems," Computer methods in applied mechanics and engineering, vol. 187, no. 1, pp.
69-77, 2000.
[51] J. K. Dienes, "On the analysis of rotation and stress rate in deforming bodies," Acta
mechanica, vol. 32, no. 4, pp. 217-232, 1979.
[52] D. R. Metzger and R. N. Dubey, "Objective tensor rates and frame indifferent constitutive
models," Mechanics research communications, vol. 13, no. 2, pp. 91-96, 1986.
[53] S. N. Varadhan, "Dynamic large deformation elasto-plastic analysis of continua, Master's
Thesis," 1997.
[54] W. Jaunzemis, Continuum mechanics, New York: The MacMillian Company, 1967.
[55] L. A. Segel, Mathematics Applied to Continuum Mechanics, New York: MacMillian
Publishing Company, 1977.

Abhishek_Dalal_Intern_Report

More Related Content

What's hot (19)

Viewers also liked (11)

Similar to Abhishek_Dalal_Intern_Report (20)

Abhishek_Dalal_Intern_Report