Finite element analysis was first developed for use in the aerospace and nuclear industries
where the safety of structures is critical. Today, the growth in usage of the method is
directly attributable to the rapid advances in computer technology in recent years. As a
result, commercial finite element packages exist that are capable of solving the most
sophisticated problems, not just in structural analysis, but for a wide range of phenomena
such as steady state and dynamic temperature distributions, fluid flow and manufacturing
processes such as injection molding and metal forming.
FEA consists of a computer model of a material or design that is loaded and analyzed for
specific results. It is used in new product design, and existing product
refinement. A company is able to verify that a proposed design will be
able to perform to the client's specifications prior to manufacturing or
construction. Modifying an existing product or structure is utilised to
qualify the product or structure for a new service condition. In case of
structural failure, FEA may be used to help determine the design
modifications to meet the new condition.
Within each of these modeling schemes, the programmer can insert numerous algorithms
(functions) which may make the system behave linearly or non-linearly. Linear systems
are far less complex and generally ignore many subtleties of model loading & behaviour.
Non-linear systems can account for more realistic behaviour such as plastic deformation,
changing loads etc. and is capable of testing a component all the way to failure.
Introduction
The most common application of FEA is the solution of stress related design
problems. As a result, all commercial packages have an extensive range of stress
analysis capabilities.
What is Stress ??
Stress can be described as a measurement of intensity of
force. As all engineers know, if this intensity increases
beyond a limit known as yield, the component's material will
undergo a permanent change in shape or may even be
subjected a to dramatic failure.
From a formal point of view, three conditions have to be met in any stress
analysis, equilibrium of forces (or stresses), compatibility of displacements and
satisfaction of the state of stress at continuum boundaries. These conditions,
which are usually described mathematically in good undergraduate strength of
material texts, are also applicable to non-linear analyses.
An alternative solution
An alternative solution may be obtained via the force matrix
method (otherwise known as the flexibility method). In the
previous description, the displacements were the unkown, and
solution is said to be obtained via the stiffness method. In the
force method, the forces are the nodal unknowns, while the
displacements are known. The solution is obtained for the
unknown forces via the flexibility matrix & the known
displacements. The stiffness method is more powerful & applicable than the
flexibility approach.
Non-Linear Analyses
In order to explain non-linearity in stress analyses, lets examine the nature of
linear solutions. Many assumptions are made in linear analyses, the two primary
ones being the stress/strain relationship & the deformation behaviour. The stress is
assumed to be directly proportional to strain and the structure deformations are
proportional to the loads. The second assumption is oftentimes mistaken to derive
from the first, a fishing rod is an example of a non-linear structure made of linear
material. A stress analysis problem is linear only if all conditions of
proportionality hold. If any one of them is violated, then we have a Non-Linear
problem.
Most real life structures, especially plastics, are non-linear, perhaps both in
structure and in material. Most plastic materials have a non-linear stress strain
relationship. The non-linearity arising from the nature of material is called
'Material Non-linearity'. Furthermore, thin walled plastic structures exhibit a non-
linear load-deflection relationship, which could arise even if the material were
linear (fishing rod). This kind is called geometric non-linearity.
In a nonlinear analysis, initial conditions at the start of each increment is the state
of the model at the end of the previous one. This dependency provides a
convenient method for following complex loading histories, such as a
manufacturing process. At each increment, the solver iterates for equilibrium
using a numerical technique such as the Newton-Raphson method. Due to the
iterative nature of the calculations, non-linear FEA is computationally expensive,
but reflects the real life conditions more accurately than linear analyses. The big
challenge is to provide a convergent solution at minimum cost (the minimum
number of increments).
See the 'non-linear' section of solution types for more details of such analyses & how
solutions are achieved. For details on applying loads & boundary conditions, see the
'improving results' and 'faster analyses' sections on the menu.
What is Vibration ??
Vibration usually becomes a concern when it's amplitudes grow large enough to
cause either excessive stress, or if it disturbs the people in, on or near the
vibrating object(s). As far as most structures are concerned, vibration will disturb
the people around the structure long before stress becomes an issue. There are
many items of equipment (balances, microscopes, cameras, transmission
equipment etc.) that are very sensitive to vibration.
Modal analyses are important in machines where there is likely to be cyclic out of
balance forces, such as in rotating machinery (engines, electric & penumatic
motors, generators, industrial equipment, etc.) and fluid flow
applications (due to alternating vortex shedding). The chief
aim of any vibration analysis is to ensure that the system is not
subject to a dangerous resonant condition during the range of
operation. A point to note is that although the response of the
system is time dependant, any excitation will be harmonic, and the solution may
be obtained using the eigenvalue approach. It is important to note that many
applications fall in a category beyond this range, and full dynamic analyses are
required.
The fundamental frequency is usually the one of most interest to design engineers,
as most systems are designed to operate below it. Oftentimes, an operating
frequency is higher than the fundamental, hence as the equipment speeds up or
slows down, it experiences a momentary 'shudder' period as it passes through the
resonance zone. There is a corresponding mode shape which describes the
displacement of the system due to the vibration.
Eignevalues are otherwise known as latent roots and characteristic values, the
square root of the eigenvalue is known as a natural frequency or resonant
frequency. There is also a number of terms used to describe mode shapes, they are
also known as eigenvectors, normal modes, characteristic vectors or latent
vectors. The first five modes of vibration for an aerofoil are given below:
There are also situations where the response of the structure to a particular forcing
excitation is required. This is usually found using a technique known as modal
superposition. The overall response is described in terms of a sum of modal
responses, with the contribution of a particular mode given by the proximity of
the forcing frequency to the natural frequency and the amount of damping present
in the system. The response is dominated by modes close to the excitation
frequency and therefore the modal series is often truncated to reduce computation.
Modal superposition methods can only be applied in applications with a harmonic
excitation, otherwise the response becomes non-linear & cannot be solved using
the eigenvalue extraction approach.
The results from a forced harmonic analysis can be used to determine whether the
displacement of a particular structure is within acceptable limits. By calculating
the stress induced by the vibration it is also possible to predict the fatigue life of a
particular component.
There are a large number of ways to determine eigenvalues & eigenvectors, the
best choice depends on the form of the equations being solved. The main methods
are the power, subspace, LR, QR, Givens, Householder & Lanczos methods. Each
method is usually well documented in advanced Engineering Math texts, except
for Lanczos, which is relatively new.
Modal-dynamic analyses are transient in nature. They give the response for the
model as a function of time where a cyclic (sinusoidal) load is applied to the
structure. Modal-dynamic analyses is also referred to as forced harmonic response
analysis. Complex displacements and phase angles are evaluated and deflections
& stresses may be calculated at specific times. This analysis type is formulated on
the principle of modal superposition, and so a natural frequency analysis must be
carried out first. The modal amplitudes are integrated through time & the response
is subsequently evaluated. This analysis solution must be linear in nature (in time
domain), as superposition & eigenvalue extraction techniques cannot be applied
to non-linear time domain applications.
Boundary Conditions
Theoretically, no boundary conditions are required. However, I do advise to apply
at least one to a model, as it will help the solver overcome any potential problems
due to rigid body motion (otherwise known as ill-conditioning). Just fix any point
on the model with a zero displacement, and the model will always solve.
If you intend applying boundary conditions, note that the results of eigenvalue
analyses are very sensitive to the way in which BC's are applied. Therefore,
always try to replicate the physical boundary conditions as closely as possible. If
you cannot achieve this for some reason, carry out a sensitivity analysis by
modifying the BCs slightly & comparing the results with your previous output.
This will give you an indication of the influence the BCs have on the results.
Introduction
If the excitation applied to a structure is impulsive rather than harmonic, many
modes contribute to the response and it becomes more appropriate to use direct
integration methods rather than modal analysis. There are a large number of
applications where transient analyses are necessary. Many structures are subject to
time varying loads such as impluse, blast, impact & seismic loadings. Transient
dynamic analysis determines the time-response history of a structure subjected to
a forced displacement function. The structure may behave linearly, or in some
cases, friction, plasticity, large deflections or gaps may produce nonlinear
behavior. Once the time response history is known, complete deflection and stress
information can be obtained for specific times.
Direct solutions are used to evaluate the response of a structure within a very
narrow frequency range of interest, and are usually used for models that subject to
high frequency impulses. The solution is purely transient, no frequency extraction
is carried out first.
The progression of the solution from one time step to the next is
achieved by implementing time integration techniques. Despite
many packages providing automatic time stepping estimates, the full response of
the structure may not be captured, and manual intervention will be required. If
there is a discontinuity in your automatically time stepped results, chances are
there is a spike in the response that is not being fully captured.
Implicit schemes, which are unconditionally stable, find the response at the end of
the time step in terms of the conditions at the end of the time step. In other words,
the calculation of the solution at time (t+∆t) is found by considering the response
at time (t+∆t). An important point to note is that the solution at each time step
involves matrix factorisation (evaluating the system of equations that comprise
the model), which is a computationally intensive process. Despite this
disadvantage, implicit schemes are often used, as the solution is inherently
reliable & robust. Implicit analyses allow much larger time steps than the others,
and so the solution can be obtained with fewer calculation increments. As implicit
schemes are always stable, the time step length is governed by considerations of
accuracy alone.
The Crank-Nicolson approach evaluates the next step of the solution by using the
prediction at the centre of the time step. As with the backward difference scheme,
this is an implicit solution which is conditionally stable (results in an oscillatory
solution if the critical time step for stability is exceeded). The central difference
method is more accurate than both the purely implicit or explicit techniques since
neither favours the response at the start or end of the time step.
Response spectrum analysis produces a set of results for each excitation load case
which is in the form of an envelope. All results are absolute values, each value
represents the maximum absolute value of displacement, moment, shear, etc. that
is likely to occur during the event which corresponds to the input response
spectrum.
Concepts associated with Dynamic Strucural Analyses
SHAKEDOWN ANALYSIS: If load intensities on a structure remain sufficiently
low, the response of the body is purely elastic (with the exception of stress
singularities). If the load intensities become sufficiently high, the instantaneous
load-carrying capacity of the structure becomes exhausted (unconstrained plastic
flow and damage evolution occurs) & collapses.
If the plastic strain increments in each load cycle are of the same
sign then, after a sufficient number of cycles, the total strains (and
therefore displacements) become so large that the structure
departs from its original form and becomes unserviceable. This
phenomenon is called incremental collapse or ratchetting.
If the strain increments change sign in every cycle, they tend to
cancel each other and total deformation remains small leading to
alternating plasticity. In this case, however, the material at the most stressed
points begins to fails due to low-cycle fatigue.
If, after some time plastic flow and damage evolution cease to develop further and
the accumulated dissipated energy in the whole structure remains bounded such
that the structure responds purely elastically to the applied variable loads, one
says that the structure shakes down
Introduction
Thermal analysis is used to determine the temperature distribution, heat
accumulation or dissipation, and other related thermal quantities in an object. The
nodal degrees of freedom (primary unknown data) are the temperatures. The
primary heat transfer mechanisms are conduction, convection and radiation. In
addition, less dominant phenomena such as change of phase (melting or freezing)
& internal heat generation can occur.
Conduction
Conduction is governed by Fourier's law, which is a differential equation
describing the rate of heat transfer as a function of temperature
gradient, material thermal capacitance & the rate of internal heat
generation. This law describes the temperature within the solid
body, but does not account for how heat will flow to & from the
component. In order to carry out analyses using a conduction
model alone, temperatures must be described as part of the
boundary condition description. Heat flows (otherwise known as heat flux) are
oftentimes specified along boundaries in addition to temperature BC's.
Radiation
Radiation type boundary conditions are applied if there is a significant
temperature difference between bodies in an enclosed space, or if there is a far
field heat source/sink (such as the sun or a very cold enviornment).
This heat transfer mechanism occurs exclusively at the surface and is a function
of the fourth power of the absolute temperatures (Kelvin), the emissivity of the
bodies & a value known as the Stefan-Boltzman constant. The emissivity is
dependent on surface properties such as the colour & finish. Radiation type
boundary conditions are highly non-linear due to the difference between fourth
order absolute temperatures. A further complication is due to incidents where the
surfaces of two adjacent radiating bodies are not flat and parallel to each other.
This case is overcome by introducing a shape factor (otherwise known as a view,
angle or interception factor) to the solution.
Convection
The convection heat transfer mechanism is due to the temperature gradient
between a fluid and a solid. This mechanism is complex as a boundary layer
usually exists within the fluid adjacent to the solid boundary. The heat flux is a
function of the temperature difference, ∆T, and a heat transfer coefficient, h.
The heat transfer coefficient is dependent on many factors such as fluid pressure,
velocity, density, specific heat (ratio of specific heats if the fluid is compressible),
viscosity & conductivity. It is also dependent on surface properties such as
roughness & geometry. Due to the extreme non-linear nature of convection type
phenomena, solutions are usually based on imperical relations such as log laws.
In order to implement convective heat transfer in FEA,
boundary conditions for specific cases have been
developed. Examples of which are Vertical Plate in
horizontal flows, flow over isothermal inclined flat plates,
flow through horizontal cylinders, flow over an inclined
surface, vertical enclosed space flow, flow in horizontal
tubes & ducts, generic convection as a function of temperature difference or
grashof & prandtl numbers, flow along a rotating disk, etc.. Each boundary
condition may have automatic implementation for each of the three flow types,
laminar, transition & turbulent.
Information defining all parameters of each BC type must be input, this can
usually be carried out manually or via tables of data. An important point to note is
that very few packages have the capability to apply convective BC's to the level
described here. MSc/Thermal and SC03, the Rolls-Royce proprietary code, both
have extensive capabilities for applying convective type BC's. At least two other
FEA software vendors are currently considering implementing such capability
into their codes.
The usual procedure is to carry out a thermal analysis which evaluates the
temperature distribution. These temperatures can then be used to prime the
displacement analysis, and hence thermal deflections, strains & stresses can be
evaluated.
It is also possible to have fully coupled analyses where the temperatures &
displacements are a function of each other. This is most evident in analyses that
involve fluid flows, such as in a gas turbine or rocket. The heat transfer rates are
dependent on massflow rates, but massflow rates are a function of valve & seal
clearances (labrynth seals in gas turbines, nozzles in rockets). Therefore, we end
up with a scenario where clearances are a function of temperature and temperature
is a function of clearances. This type of problem can only be solved via a non-
linear and fully coupled solution.
One way of providing these temperatures is to carry out a coarse analysis on the
overall model and use the temperatures at relevant points to 'prime' the sub-
system model (otherwise known as a submodel).
FEA Packages with such facilities have the capability of writing out the
temperature-time values from the coarse model to a file. This file can be
subsequently used in the submodel, and so an accurate representation of the
thermal enviornment can be provided while studying the finer details contained
within the sub-system model. This approach is extremely rich as regards saving
on analysis times. Despite the high level of idealisation being implemented, it
provides very accurate & realistic modelling of the physical conditions being
investigated.
If a structure has one or more dimensions that are small relative to the others
(slender or thin-walled), and is subject to compressive loads, then a buckling
analysis may be necessary.
An important note is that the eigenvalue method does not take into account of any
initial imperfections in the structure and so the results rarely correspond with
practical tests. Eigenvalue solutions usually over estimate the buckling load and
give no information about the post-buckling state of the structure. Sudden
buckling simply does not occur in the real world.
Avoiding Instabilities
Any structure is most efficient when subjected to evenly distributed tensile or
compressive stress, such as occurring in cables, strings etc. Evidently, such modes
of loading makes the best use of the material, and its strength. On the other hand
bending (flexing) is the least efficient way of loading a structure. A high flexural
stiffness of the structure means high resistance to buckling. This is true even if the
load is entirely in-plane, since when buckling is imminent, the only stiffness that
counts is flexural.
When buckling occurs, symmetry of the part does not apply. There
is no symmetry of the buckled shape, although both the part, and
the loading may be symmetric. Correspondingly, when carring out
an FE buckling investigations, it is advisable to implement a
full 3D analysis of the structure under inspection.
By and large it is true that buckling usually occurs when compressive stress is
present. But what is not evident that compressive stress can prevail in un-expected
places. Shallow domes under internal pressure can develop local compressive
stress regions, and make it vulnerable to instabilities.
Introduction
Many kinds of electromagnetic phenomenon can be modeled from the
propagation of microwaves to the torque in an electric motor. Analysis of
electrostatic and magnetic fields passing through and around a structure provides
insight into the response, and hence a means for regulating these fields to attain
specific responses.
FEA can be used to analyse the linear electric or magnetic behaviour of devices.
Analyses typically involve the evaluation of magnetic,
electric and thermal fields. Further applications include the
analysis of shape-memory materials & piezoelectric effects.
An analysis can be static, harmonic or transient state in
nature. Due to the complexity of the practical applications of
the technique, it is not unusual to have magnetic, dielectric and thermal couplings
in a single model. Such complex analyses generally make realistic modelling an
ardouous task.
Application Areas
The application areas include, but is not limited to the design of:
Analysis Types
ELECTROSTATIC analyses involve the computation of
electric potential and fields in the absence of current. It
can also involve the calculation of capacitance between
conductors, and dielectric strength of insulators.
Introduction
Fluid flow problems arise in almost all industrial sectors: food processing, water
treatment, marine engineering, automotive, aerodynamics, and gas turbine design.
FEA facilitates the prediction of fluid flow, heat & mass transfer, and chemical
reactions (explosions) and related phenomena.
Solution Approach
The foundation of fluid dynamics is based on the Navier-Stokes equations, the set
of partial differential equations that describe fluid flow. In FEA, this equation is
rewritten as algebraic equations that relate the velocity, temperature, pressure, and
other variables, such as species concentrations. The resulting equations are then
solved numerically, yielding a complete picture of the flow.
The equations are solved iteratively using the method of weighted residuals. The
main method of solution is achieved via the Galerkin method, but others exist.
One such variant is the Petrov-Galerkin method, which is used to solve instances
of viscous, high Reynolds number flows.
For transient problems, a special time integration technique known as the semi-
explicit scheme is used in large analyses, as it is more economical than other
methods available.
Coupled Analyses
Due to the complex nature of the physical
processes being modelled, it is not unusal to
conduct coupled analyses as part of a design
program. Fluid-structural, fluid-thermal & fluid-
acoustic analyses are not uncommon.
Both steady-state & transient solutions are usually highly non-linear in nature.
Therefore they are computationally expensive to implement.
Introduction
Despite the fact that all physical phenomena are non-linear and time dependant to
some degree, linear static analyses remain the most useful and prolific form of FE
analyses carried out today. The reason for its widespread use is that linear
analyses are fast, oftentimes sufficiently representative of the physical phenomena
and very easy to carry out.
Linear static analyses are usually sufficient for situations where loads are known
and the instance at which peak stress occurs is obvious. When performing a linear
static stress analysis, the analyst applies static loads (forces, pressures or
prescribed displacements) to the model.
o Loads are all static. This means that the loads are applied to the structure
in a slow or steady fashion and in a way that makes them time independent
(are assumed to be constant for an infinite period of time).
Introduction
Every physical phenomena in the real world is nonlinear to some extent.
Engineers and scientists observe linear behaviour to be suitably representative for
many applications. However, there are many cases where the nonlinearity of the
problem cannot be ignored and so methods of obtaining nonlinear solutions have
to be adopted.
Due to the recent advances in computer hardware, more and more engineering
firms are turning to nonlinear finite element analysis to understand the behaviour
of their products. nonlinear analyses can be either transient or steady-state (static)
in nature, this document will outline the main points of nonlinear static analyses.
View the documents on dynamic or transient type analyses for more details on
time-dependent problems.
Tangent Stiffness
The equilibrium path of a linear analysis is a straight line; when the load is
increased, the deflection increases correspondingly. With a non-linear analysis
however, the equilibrium path is non-linear, so the relationship between load and
displacements is not linear. An equilibrium path for a non-linear analysis is
shown. If we get the tangent to this curve (the slope), we have a property of the
system known as the tangent stiffness. Based on this, we observe that the tangent
stiffness is constant in a linear analysis, while it is variable in a non-
linear analysis.
o Increment 1, part 1
o The load applied is incremented (increased) to half the total
load (always strive for the first increment to be within the
linear range of the system being modelled, if there is a linear
portion to the equilibrium curve). The stiffness matrix is
formed using the initial conditions. This increment is indicated
by the yellow portion of the equilibrium curve.
o Increment 1, part 2
The linear stiffness matrix is referred to as the Tangent
stiffness matrix in a non-linear analysis (this was discussed in
the previous section). The tangent stiffness is used in the
second increment.
o Increment 2, part 1
The external load is incremented by a set amount, (to its final value in this
case) and the Newton-Raphson expression is formed.
o Increment 2, part 2
Using the external load vector and the stiffness matrix
(tangent stiffness) from the pevious increment, the current
iteration displacement is estimated by extrapolation via the
Newton-Raphson method.
o Increment 2, part 3
Once the first-guess displacement is determined, the internal forces
resisting the load are calculated. This value is subtracted from the external
load vector to determine a residual force.
o Increment 2, part 4
This residual is compared with a set tolerance, if it is
within the convergence tolerance, then the solution has
been determined. If not, the stiffness matrix is reevaluated
using the updated displacement and internal force vectors.
This is achieved using LU decomposition or any other appropriate
method, the updated tangent stiffness results.
o Increment 2, part 5
The internal load vector and the tangent stiffness matrix
from the pevious iteration are used to estimate a new
displacement. This is done by extrapolation via the
Newton-Raphson method.
o Increment 2, part 6
Using the second-guess displacement vector, the internal forces are again
determined and compared with the applied load. A residual exists, there is
still no convergence.
o Increment 2, part 7
Using the second-guess displacement and updated internal
force vectors, the the tangent stiffness matrix is updated,
again via LU decomposition of the sytem equations.
o Increment 2, part 8
o This procedure of guessing the displacement and updating
the internal force vector is continued until zero residual
force is obtained (within a tolerance). When this is
achieved, it indicates that the solution has converged and
that the current load and displacement vectors are final for
this increment.
o Increment 2, part 7
o The results are written to memory as those for the
increment and to file if required. When this is achieved the
current load increment has been completed. The final
equilibrium path for this increment is shown here. As can
be seen, three iterations were required to obtain convergence.
The next few sections will refer to this example to describe the full richness of
contemporary commercial finite element solvers. The explanations will not
encapsulate all the issues that commercial solvers deal with, it will merely give an
overview.
Both force and displacement control methods are based on incrementing one
variable and extrapolating the other based on the current state of the tangent
stiffness matrix. Therefore the increments are calculated by a prediction step, and
so are known as predictor or extrapolator type methods.
In addition, to the two methods outlined above, there is the strain correction
method, the secant method, the direct substitution method and quasi-Newton
Methods (such as BFGS and Inverse Broyden).
A force control method could have equally been used as the iteration control
parameter. Since the residual between the load vector applied and the current load
vector will differ, ther residual (difference between the two) measures the
departure from equilibrium. This is known as the residual convergence test.
Another criterion that can be used is the internal energy during each iteration.
This is is the work done by the residual forces through the incremental
displacements. Since this convergence criterion involves both the displacements
and forces, it is an attractive termination scheme.
On rare occasions, Newton iteration schemes will neither converge nor diverge,
but just oscillate in the iterative ether. To avoid excessive and unnecessary
computations, it is worth imposing a limit on the maximum number of iterations
per increment.
If a body we are modelling moves through space, we need to track its motion in
some way. The key difference between geometrically linear and nonlinear
structural analysis lies in its kinematics.
The total Lagrangian formulation is used when the equations are to be written
with respect to the original reference state only. The reference frame does not
change throughout the analysis. It is usually applicable to large deflection type
problems such as the one we were solving in our example.
The Eulerian method is used where the mesh is fixed in space and the material
flows through it. Therefore, Eulerian finite elements undergo no distortion due to
material motion. Despite this, the treatment of constitutive equations and updates
is complicated due to the convection of material through the elements. This type
of reference frame is most suitable for steady-state problems such as extrusion
and fluid mechanics problems.
A Final Word
Approach all nonlinear analyses with care, black-box codes are easily used and
frequently mis-used. Even codes that can solve difficult nonlinear problems will
not produce accurate results if modeled improperly (the solver can only answer
the questions posed, it is not magic). It is necessary to rigorously study each
computed solution, try to you understand what each solution tells you before
running another analysis.
FEA consists of a computer model of a material or design that is stressed and analyzed for specific results.
It is used in new product design, and existing product refinement. A company is able to verify a proposed
design will be able to perform to the client's specifications prior to manufacturing or construction.
Modifying an existing product or structure is utilized to qualify the product or structure for a new service
condition. In case of structural failure, FEA may be used to help determine the design modifications to meet
the new condition.
There are generally two types of analysis that are used in industry: 2-D modeling, and 3-D modeling. While
2-D modeling conserves simplicity and allows the analysis to be run on a relatively normal computer, it
tends to yield less accurate results. 3-D modeling, however, produces more accurate results while
sacrificing the ability to run on all but the fastest computers effectively. Within each of these modeling
schemes, the programmer can insert numerous algorithms (functions) which may make the system behave
linearly or non-linearly. Linear systems are far less complex and generally do not take into account plastic
deformation. Non-linear systems do account for plastic deformation, and many also are capable of testing a
material all the way to fracture.
A wide range of objective functions (variables within the system) are available for minimization or
maximization:
• Rod elements
• Beam elements
• Plate/Shell/Composite elements
• Shear panel
• Solid elements
• Spring elements
• Mass elements
• Rigid elements
• Viscous damping elements
Many FEA programs also are equipped with the capability to use multiple materials within the structure
such as:
FEA has become a solution to the task of predicting failure due to unknown stresses by showing problem
areas in a material and allowing designers to see all of the theoretical stresses within. This method of
product design and testing is far superior to the manufacturing costs which would accrue if each sample
was actually built and tested.
Introduction
The history of analytical approaches to studying fracture has been relatively short lived in
comparison with other methods. Only beginning around the turn of this century, there has
not been much time in which to do research and theoretical analysis of fracture, yet the
advances made have been significant, if not the backbone of much of our advancement in
this century. The work of such men as A. A. Griffith and G. R. Irwin are just a few of the
people who have helped to make these advances possible. Griffith's theory of brittle
fracture helps us to understand why brittle fracture occurs in a material. Likewise, as we
will later see, Irwin took Griffith's work and applied it to ductile materials, which is also
beneficial. Also work in the areas of fatigue and stress concentration have enabled us to
make more advances as far as the uses of specific materials are concerned.
Work of Griffith
A. A. Griffith started his work in around the 1920s. At this time, it was accepted that the
theoretical strength of a material was taken to be E/10, where E is Young's Modulus for
the particular material. He was only considering elastic, brittle materials, in which no
plastic deformation took place. However, it was observed that the true values of critical
strength was as much as 1000 times less than this predicted value, and Griffith wished to
investigate this discrepancy. He discovered that there were many microscopic cracks in
every material which were present at all times. He hypothesized that these small cracks
actually lowered the overall strength of the material because as a load is applied to these
cracks, stress concentration is experienced. This stress concentration magnifies the
stresses at the crack tip, and these cracks will grow much more quickly, thus causing the
material to fracture long before it ever reaches its theoretical strength. It should be noted
that Griffith believed that, at the crack tips, the value of stress actually reached the
theoretical maximum, but the overall average of the stress was lowered. It should also be
noted that this phenomenon of stress concentration is not only relegated to microscopic
cracks in a material. Any void in the material (holes that have been machined or drilled
out), corners, or hollow areas in the internal area of the material also cause stress
concentration to occur, and most times, fracture will begin in one of these areas simply
because of this phenomenon. (fig 22.3 Reed-Hill)
From this work with stress concentration and working with elastic, brittle materials,
Griffith formulated his own theory of brittle fracture, using elastic strain energy concepts.
His theory described the behavior of crack propagation of an elliptical nature by
considering energy methods. The equation basically states that when a crack is able to
propagate enough to fracture a material, that the gain in the surface energy is equal to the
loss of strain energy, and is considered to be the primary equation to describe brittle
fracture.
Work of Irwin
Griffith's work was significant, however it did not include ductile materials in its
consideration. Another man, G. R. Irwin, in the 1950s, began to see how the theory would
apply to ductile materials. He determined that there was also a certain energy from plastic
deformation that had to be added to the strain energy originally considered by Griffith in
order for the theory to work for ductile materials as well, creating what is known as the
strain energy release rate.
Stress Intensity
The term stress intensity is not to be confused with stress concentration work done by
Griffith. The stress concentration is how the stress is amplified at a crack tip, whereas the
stress intensity is used to describe the distribution of stress around a particular flaw. This
term is used when investigating modes of fracture (link to part about K1c), in particular,
mode I fracture, which is the most common. This term is used when computing the plane
stresses and strains which exist in front of a moving crack. This value is dependent upon
many things and is different for each material. Among the things which it depends on is
the applied stress, the size and placement of the crack, as well as the geometry of the
specimen.
Fatigue
Introduction
In selecting materials for a given application one must have an idea of the final geometry
and the dimension of the part. Under certain environmental conditions and given loadings
the part must be able to function properly. One way to make this decision is by
comparing the failure criterion to a critical load factor. Fracture is a very complex
process that involves the nucleation and growth of micro and macro voids or cracks,
mechanisms of dislocations, flip bands, and propagation of microcracks, and the
geometry of the material. There has been no one set theory "set in stone" to handle all of
these factors in fracture. However there are many proposed theories used to understand
the complex nature of fracture in the material. One such class of theories involved energy
concepts. In order to understand the complex nature of fracture in materials, one must
understand the nature and character of initial cracking. This is only possible if we know
the distribution of internal stressed in the body, but also the stress needed to initiate
fracture and the length of the crack as shown in Figure 1.
One such model used to demonstrate the propagation of a crack in a brittle material is
called the elastic strain energy model.
E: modulus of elasticity
γs = specific surface energy
a = one half the length of an internal crack
In today's material world many materials also experience some plastic deformation
during fracture during fracture. Therefore the crack extension involves more than just an
increase in surface energy. γp represents a plastic deformation energy associated with
crack extension. γs + γp can be substituted into the above equation to model materials that
undergo some plastic deformation.
G: the strain energy release rate.
γp: plastic deformation energy associated with crack extension.
γs: the specific surface energy
Please note that crack propagation can only occurs when it exceeds the critical value of
G.
Example Problem
If the specific surface energy for polmethyl acrylate is 36.5 ergs/ cm2 and its
corresponding modulus of elasticity is 2.38 GPa. Compute the critical stress required for
propagation of an internal crack length is 0.03m.
Solution
σc= 60,719 Pa
The process of fracture consists of crack initiation and crack propagation. The condition
necessary for crack initiation is if the crack-like cut is able to propagate. If there was a
perfectly elastic body with a slit already present. If the slit is to propagate thereby
increasing its surface, the slit will need a certain amount of energy. This energy is called
the energy of fracture. With the formation of a new surface the strain in the corresponding
area will be reduced which results in the release of corresponding elastic energy from the
body.
δτ: the energy of fracture necessary for the formation of a new fracture surface
area.
G :the energy released into the crack tip per unit area of the crack (rate of elastic
strain energy release).
δ: the crack growth increment.
Energy Release Rate
The energy release rate often denoted by G is the amount of energy, per unit length along
the crack edge, that is supplied by the elastic energy in the body and by the loading
system in creating the new fracture surface area. In terms of the stress intensity factor
there is relationship called the Irwin relationship. Note that there our two models for the
stress intensity factor one for plane stress and plane strain.
(Plane Strain)
G: the energy release rate.
ν: Poisson's Ratio.
K: the stress intensity factor.
E: the modulus of elasticity.
G=K2/E (Plane Stress)
G: the energy release rate.
K: the stress intensity factor.
E: the modulus of elasticity.
The total energy release rate in combined mode cracking can be obtained by adding the
energies from the different modes (Figure 4.),
Fig. 4 Three modes of crack surface displacements Mode I (opening or tensile mode),
Mode II (sliding mode), and Mode III (tearing mode).
(From Parton V.Z. Fracture Mechanics from Theory to Practice Pg. 66 Figure 47, Gordon
and Breach Science Publishers.)
These models by Irwin started the foundation of linear elastic fracture mechanics
(LEFM). This discipline of fracture mechanics characterizes the state of material loading
over a volume of sufficient size that the fracture strength of many engineering materials
can be given in terms of the critical (maximum) stress intensity factor, KIC.
Fig. 5 A cracked body with a force (F) and (a) is the crack length.
(From Portela A., Dual Boundary Element Analysis of Crack Growth, Pg. 26 Figure 2.4,
Computational Mechanics Publications.)
Shown in the figure is a cracked body with a force being exerted on it and the
propagation of a crack (Figure 5.).
The crack resistance is a function of the plastic behavior of the material at the crack tip
and of its fracture characteristics. This particular property is dependent upon strain rate.
At the tip of a crack moving at high velocity the strain rates are very high, and it must be
expected that the material behaves in a more brittle manner the higher the crack speed.
So far we have assumed that the crack resistive force is independent of crack length. This
is true only for crack under plain strain. For plane stress, the crack resistance varies with
amount of crack growth. When a particular specimen is loaded, the crack starts
propagating, a further increase of the stress is required to maintain crack growth,
although the crack is longer it can withstand a higher stress. During stable crack growth
the energy release rate is equal to the crack resistive force. As can be seen in Figure 6. the
growth rate of the crack increases with an increase in crack size. For example in a ductile
material the energy required for crack growth is the same as the amount of work for
formation of a new plastic zone at the tip of the advancing crack, plus the work required
for initiation, growth and coalescence of microvoids.
Fig. 6 A graph of the increase of growth rate with crack size.
(From Broek D., Elementary Engineering Fracture Mechanics, Pg. 145 Figure 6.2,
Kluwer Academic Publishers Group.)
In a plastic, cylindrical test specimen deforms in a way that the degrees of deformation is
nearly constant along the whole cross section. Research has shown that deformation of
geometrically similar specimens of different diameters made from similar material is the
same in any phase of the tensile test and not only the reductions of area but the contour
lines in the vicinity of necking. The absolute value of the stress is independent of the size
is the strain rate is constant.
A proportion law of L. Gillemot states that to the same deformation of two geometrically
similar specimens from similar material, similar specific energy is necessary if the cross
head speed is proportional with the diameters of the specimen.
U1 and U2: cross head speeds applied during the tensile test.
d1 and d2: diameters of the specimen.
In order to learn more about the complex nature of fracture in materials impact testing
conditions were established. The conditions that were judged the most relative to the
potential for fracture are:
1. deformation at low temperatures
2. a high strain rate (rate of deformation)
3. a triaxial stress state
Two tests called the Charpy and Izod tests are used to measure the impact energy (also
known as notch toughness). These tests are important, because one can obtain
information to model the behavior of actual structures so that the laboratory test results
can be used to predict service performance under different environments.
With the Charpy V-notch (CVN) technique, the specimen is in the shape of a bar of
square cross section with a V notch. The load is applied as an impact below from a
weighted pendulum hammer that is released from a position h. The pendulum with a
knife edge strikes and fractures the specimen at the notch. The pendulum continues its
swing, rising to a maximum height h', which is lower than h. The energy necessary to
fracture the test piece is directly calculated from the difference in initial and final heights
of the swinging pendulum (Figure 7.). The impact energy (toughness) from the Charpy
test is related to the area under the total stress-strain curve. The difference in the Charpy
and the Izod techniques is in the way that the specimens are supported in the apparatus
machine.
One can expect that materials with large values of strength and ductility to have large
impact fracture energies. One has to also note that the impact data are very sensitive to
test conditions such as temperature, specimen size, and notch configuration as can be
seen in Figure 8.
Fig. 7 Illustration of Charpy and Izod Impact Tests.
(From Callister W.D. Jr., Materials Science and Engineering : An Introduction , Pg. 198
Figure 8.13, John Wiley and Sons, Inc.)
Fig. 8 A graph of the temperature dependence on the Charpy V-notch impact energy
(curve A) and percent shear fracture (curve B).
(From Callister W.D. Jr., Materials Science and Engineering : An Introduction , Pg. 199
Figure 8.14, John Wiley and Sons, Inc.)
References
Fatigue
by: Shawn M. Kelly
Overview
• Cyclic Stresses
• S-N Curve
• Crack Initiation and Propagation
• Propagation Rate
• Factors That Affect Fatigue Life
• Example Problems
• References
The concept of fatigue is very simple, when a motion is repeated, the object that is doing
the work becomes weak. For example, when you run, your leg and other muscles of your
body become weak, not always to the point where you can't move them anymore, but
there is a noticeable decrease in quality output. This same principle is seen in materials.
Fatigue occurs when a material is subject to alternating stresses, over a long period of
time. Examples of where Fatigue may occur are: springs, turbine blades, airplane wings,
bridges and bones.
This page will cover the topics included in Materials Science and Engineering, and
Introduction by Callister, as well as other information that may be helpful to the student
in an introductory materials science class.
Cyclic Stresses
There are three common ways in which stresses may be applied: axial, torsional, and
flexural. Examples of these are seen in Fig. 1.
Figure 1
Visual examples of axial stress, torsional stress, and flexural stress.
There are also three stress cycles with which loads may be applied to the sample. The
simplest being the reversed stress cycle . This is merely a sine wave where the maximum
stress and minimum stress differ by a negative sign. An example of this type of stress
cycle would be in an axle, where every half turn or half period as in the case of the sine
wave, the stress on a point would be reversed. The most common type of cycle found in
engineering applications is where the maximum stress (σmax)and minimum stress (σmin)
are asymmetric (the curve is a sine wave) not equal and opposite. This type of stress cycle
is called repeated stress cycle. A final type of cycle mode is where stress and frequency
vary randomly. An example of this would be automobile shocks, where the frequency
magnitude of imperfections in the road will produce varying minimum and maximum
stresses.
Figure 2
A S-N Plot for an aluminum alloy
The significance of the fatigue limit is that if the material is loaded below this stress, then
it will not fail, regardless of the number of times it is loaded. Material such as aluminum,
copper and magnesium do not show a fatigue limit, therefor they will fail at any stress
and number of cycles. Other important terms are fatigue strength and fatigue life. The
stress at which failure occurs for a given number of cycles is the fatigue strength. The
number of cycles required for a material to fail at a certain stress in fatigue life.
1. Crack Initiation: The initial crack occurs in this stage. The crack may be caused
by surface scratches caused by handling, or tooling of the material; threads ( as in
a screw or bolt); slip bands or dislocations intersecting the surface as a result of
previous cyclic loading or work hardening.
2. Crack Propagation: The crack continues to grow during this stage as a result of
continuously applied stresses
3. Failure: Failure occurs when the material that has not been affected by the crack
cannot withstand the applied stress. This stage happens very quickly.
Figure 3
A diagram showing location of the three steps in a fatigue fracture
under axial stress
One can determine that a material failed by fatigue by examining the fracture sight. A
fatigue fracture will have two distinct regions; One being smooth or burnished as a result
of the rubbing of the bottom and top of the crack( steps 1 & 2 ); The second is granular,
due to the rapid failure of the material. These visual clues may be seen in Fig. 4:
Figure 4
A diagram showing the surface of a fatigue fracture. Notice that the rough surface
indicates brittle failure, while the smooth surface represents crack propagation
Figure 6
An example of the striations found in fatigue fracture. Each striation is thought to be the
advancement of the crack. There may be thousands of striations in a beachmark
Propagation Rate
The rate at which a crack grows has considerable importance in determining the life of a
material. The propagation of a crack occurs during the second step of fatigue failure. As a
crack begins to propagate, the size of the crack also begins to grow. The rate at which the
crack continues to grow depends on the stress level applied. The rate at which a crack
grows can be seen mathematically in equation 8.16 in Callister by:
Eq. 1
The variables A and m are properties of the material, da is the change in crack length, and
dN is the change in the number of cycles. K is the change in the stress intensity factor or
by equation 8.17(a & b):
Eq. 2
Rearrangement and integration of Eq. 1 gives us the relation of the number of cycles of
failure, Nf, to the size of the initial flaw length, ao, and the critical crack length, ac, and
Eq. 2:
Eq. 4
The Mean stress has the affect that as the mean stress is increased, fatigue life decreases.
This occurs because the stress applies is greater.
I mentioned previously that scratches and other imperfections on the surface will cause a
decrease in the life of a material. Therefore making an effort to reduce these
imperfections by reducing sharp corners, eliminating unnecessary drilling and stamping,
shot peening, and most of all careful fabrication and handling of the material.
Another Surface treatment is called case hardening, which increases surface hardness and
fatigue life. This is achieved by exposing the component to a carbon-rich atmosphere at
high temperatures. Carbon diffuses into the material filling interstisties and other
vacancies in the material, up to 1 mm in depth.
Figure 7
A case hardened steel gear. Notice the effect of diffusion of Carbon into the material
produces a "case" around the gear.
Example problems
1. Consider a flat plate of some metal alloy that is to be exposed to repeated tensile-
compressive cycling in which the mean stress is 25 MPa. If: ao = 0.25 mm, ac =
5.0 mm, m = 4.0, A = 5 * 10-15, Y = 2.0, and Nf =3.2 * 105 cycles .
Find: Estimate the maximum tensile stress to yield the fatigue life prescribed
Eq. 3
Comments or Questions? Email Shawn Kelly.
References
1
Beer, Ferdinand P, and E. Russell Johnston, Jr. Mechanics of Materials. 2nd ed. New
York: McGraw-Hill, Inc. 1992. Images: Fig. 2.54(b), Fig. 3.8(b), Fig. 4.19
Reed-Hill, Robert E, and Reza Abbaschian. Physical Metallurgy Principles. 3rd ed.
Boston: PWS Publishing Company, 1994.
Callister, William D Jr. Materials Science and Engineering, an Introduction. 3rd ed. New
York: John Wiley & Sons, Inc., 1994.
Chris Meyer, Jireh Yue, Jared Mutter, Ron Halahan, and Matt Gordon
Stress Concentration
The fracture of a material is dependent upon the forces that exist between the atoms.
Because of the forces that exist between the atoms, there is a theoretical strength that is
typically estimated to be one-tenth of the elastic modulus of the material. However, the
experimentally measured fracture strengths of materials are found to be 10 to 1000 times
below this theoretical value. The discrepancy is explained to exist because of the
presence of small flaws or cracks found either on the surface or within the material.
These flaws cause the stress surrounding the flaw to be amplified where the
magnification is dependent upon the orientation and geometry of the flaw. Looking at fig.
1, one can see a stress profile across a cross section containing an internal, elliptically-
shaped crack. One can see that the stress is at a maximum at the crack tip and decreased
to the nominal applied stress with increasing distance away from the crack. The stress is
concentrated around the crack tip or flaw developing the concept of stress concentration.
Stress raisers are defined as the flaws having the ability to amplify an applied stress in
the locale.
Fig. 1: (a) The geometry of surface and internal cracks. (b) Schematic stress profile along
the line X-X' in (a), demonstrating stress amplification at crack tip positions.
The magnitude of the nominal applied tensile stress is σ o; the radius of the curvature of
the crack tip is ρ; and a represents the length of a surface crack, or half the length of an
internal crack.
The ratio of the maximum stress and the nominal applied tensile stress is denoted as the
stress concentration factor, Kt, where Kt can be calculated by Equation 2. The stress
concentration factor is a simple measure of the degree to which an external stress is
amplified at the tip of a small crack.
Stress raisers are typically more destructive in brittle materials. Ductile materials have the
ability to plastically deform in the region surrounding the stress raisers which in turn
evenly distributes the stress load around the flaw. The maximum stress concentration
factor results in a value less than that found for the theoretical value. Since brittle
materials cannot plastically deform, the stress raisers will create the theoretical stress
concentration situation.
Reference:
Stress Intensity
The Liberty Bell (Philadelphia, PA)
Jefferson Kim, MSE 2094, Term Project (Edited R.D. Kriz 3-5-00)
Stress Intensity Factor, K, is used in fracture mechanics to more accurately predict the
stress state ("stress intensity") near the tip of a crack caused by a remote load or residual
stresses. When this stress state becomes critical a small crack grows ("extends") and the
material fails. The load at which this failure occurs is referred to as the fracture strength.
The experimental fracture strength of solid materials is 10 to 1000 times below the
theoretical strength values, where tiny internal and external surface cracks create higher
stresses near these cracks, hence lowering the theoretical value of strength. The large
crack seen in the picture of the Liberty Bell was the result of small cracks and internal
residual stresses not known at the time. The original, "as fabricated" cracks were very
small and hard to see with naked eyes, and according to Hertzberg, during the war against
the British, the bell was polished whenever they saw a crack on the surface. Hardly a
solution based on what we understand today. Unlike "stress concentration", Stress
Intentsity, K, as the name implies, is a parameter that amplifies the magnitude of the
applied stress that includes the geometrical parameter Y (load type). These load types are
categorized as Mode-I, -II, or -III. The Mode-I stress intensity factor, KIc is the most often
used engineering design parameter in fracture mechanics and hence must be understood if
we are to design fracture tolerant materials used in bridges, buildings, aircraft, or even
bells. Polishing just won't do if we detect a crack. Typically for most materials if a crack
can be seen it is very close to the critical stress state predicted by the "Stress Intensity
Factor".
Generally there are three modes to describe different crack surface displacement in
Fig.8.3 (Hertzberg, p321). Mode I is opening or tensile mode where the crack surfaces
move directly apart. Mode II is sliding or in-plane shear mode where the crack surfaces
slide over one another in a direction perpendicular to the leading edge of the crack. Mode
III is tearing and antiplane shear mode where the crack surfaces move relative to one
another and parallel to the leading edge of the crack. Mode I is the most common load
type encountered in engineering design and will be explained here in more detail.
The value of the stress intensity factor, K, is a function of the applied stress, the size and
the position of the crack as well as the geometry of the solid piece where the cracks are
detected, Fig.8.5 (Hertxberg, p323). The tensile stress in X and Y directions, and the
shear stress in the X-Y plane can calculated in terms of K and position can be written as:
Mode-I
Mode-I
Fracture Toughness, KIc
Engineers are mostly worried about the brittle fracture because the brittle fractures bring
most devastating accidents and happen rapidly, and usually the brittle fractures take place
when the applied stress increases such that the stress state at the crack tip reaches a
critical value. The fracture toughness can be defined in terms of the stress intensity factor,
K, but at a critical stress state. as:
where Y is a dimensionless parameter that depends on both the specimen and crack
geometry in Fig8.11(Callister, p193), and the greek symbol "Sigma" is an applied stress
and "a" is crack length. Generally, for the elliptical shaped crack,the equation is modified
to include the geometry of the crack with three different Y's.
However, Y factor is 1.0 for the plate of infinite width and 1.1 for a plate of semi-infinite
width. When the thickness of specimen is very large with respect to the crack length, the
stress intensity factor for Mode I is often called the plane strain "Fracture Toughness".
This modification of stress intensity into a plane strain fracture toughness parameter can
be approximated by a relationship that includes specimen geometry, and yield strength.
Hence the specimen thickness is shown to be the most significant parameter that controls
the transition of fracture toughness from "plane stress" to "plane strain", see Fig. 8.12
(Callister, p194).
The plain strain fracture toughness for Mode I, KIc is also a function of many other
factors such as temperature, strain-rate and microstructure. Hence KIc is unique for a
particular material and is a fundamental material property so it is a very important
consideration for material selection and design.
KIc, stress, and Y factor are important variables for engineers to design and to determine
the safety of machinery, and often the size of the cracks is a very important factor to
make decisions such that the maximum allowable size of the crack can be written as
Also because KIc is unique for a particular material, engineers can use this variable for
selecting appropriate materials for a range of different applications. From the table in
Appendix B engineers can also decide how much load and stress can be allowed for a
particular specimen geometry. This critical information helps engineers to optimize the
design and the safety on the operations and to prevent or minimize possible accidents.
For example, in aircraft components, there are a lot of rivet holes and small cracks which
bring Y calibration factor high up to the critical stress. What engineers do is measure the
length of cracks to calculate the maximum cracks length and to compare with safety
measurement. They can also make a hole at the tip of cracks, which brings down Y
calibration factor and the also the the stress concentration. Additionally, engineers clean
the fracture surfaces to prevent further damages. Not only does cleaning lower the Y
calibration, but it also helps to protect the surface from undesireable chemical reactions.
Various cleaning methods are described in the table below.
Conclusion
Stress intensity and fracture toughness are critically important fracture mechanics
parameters used by materials engineers and designers. We saw that there are a lot of
factors that determine fracture of a material. KIc is an unique material property, that is
used by engineers to design and manufacture products for durability and safe operation.
Callister, W. Materials Science and Engieering. John Wiley and Sons, New York1994.
The most efficient method of learning is by example. Therefore, I would like to present to
you a simple FEA problem: the case of a three-member truss. The method of solution to
this problem should demonstrate the basic concepts of FEA which are present in any
analysis.
Before introducing specific quantities for our example, let's first take a look at our
structure:
The overall objective of our analysis will be to determine the displacements of the truss
members given the load P.
The first thing we must do is choose our elements. For our situation this is easy: each
truss member should be one element. Further division would accomplish nothing, since
each truss member can only support axial loads.
The matrix kij is called the " stiffness matrix." It is the matrix which defines the geometric
and material properties of the bar. Stiffness matrices are a fundamental part of FEA.
These matrices always define inherent properties of the system being studied. For the
system at hand, we need to determine the stiffness matrix. The way we will go about
doing this may seem a little strange at first, but try to follow the reasoning as it does
make sense. Let's begin by assuming u1 = 1 and u2 = u3 = u4 = 0. Then our matrix takes
the form:
Each force Fi is equal to kj1. Now, recall from mechanics of materials that the
displacement of a rod is given by u = FL/AE. With displacement u1 = 1, force 1 is F1 =
AE/L. To maintain equilibrium, we must also have a force F3 = -AE/L:
Since our Fi's equal our ki1's, we have:
It important to remember that our element can support only axial loads. Therefore,
displacements u2 and u4 can not give rise to stresses in the bar since these displacements
are perpendicular to the axis of the bar. Thus, the stiffness coefficients of these
displacements must be zero: ki2 = ki4 = 0. Finally, a displacement u3 = 1 will result in
forces just opposite to those from u1 = 1, so ki3 = -ki1. Our stiffness matrix is:
It must be emphasized that the stiffness matrix just derived is only valid for bars parallel
to the x-axis. Through a similar derivation it can be shown that the stiffness matrix for
any bar oriented at an angle "theta" to the x-axis is:
where c = cos"theta" and s = sin"theta". Note that when "theta" = 0, this stiffness matrix
reduces to the one we derived for a horizontal bar.
Now knowing the stiffness matrix for any axially loaded bar, we can apply it to a real
situation with specific quantities. Consider the following truss:
The displacements and external forces are:
Note the symbols we are using: R is an external force on the truss; F is an internal force
resulting from the stresses imposed on the structure during a displacement. Knowing the
orientations of each element, we can set up matrices for them. Using "theta" = 90 degrees
for element 1, "theta" = 135 degrees for element 2, and "theta" = 0 degrees for element 3
we obtain the following matrices:
Element 1:
Element 2:
Element 3:
We can now generate a set of equilibrium equations for each node. Consider the
following figure:
The nodal forces (resulting from element displacements) must be equal and opposite to
the externally applied forces. Note that we have all forces drawn in positive x and y
directions. Thus, for equilibrium at node 1:
We want to solve for R1 and R2. Obtaining the nodal forces F2(element3), F2(element2), F1(element3),
and F1(element2) from our previously determined matrices we get:
R1 = AE/L ( 3u1/2 - u2/2 - u3 - u5/2 + u6/2 )
R4 = AE/L ( u4 - u6 )
Now recall what we are trying to do here: given a load P, we want to solve for the
displacements at each node. Observing that node 2 is pinned and that node 3 is on a
roller, the displacements u3, u4, and u5 must equal 0. These values are quite important
because without them we wouldn't be able to solve the problem. As a matter of fact,
values such as these are always needed in finite element analyses; they are known as "
boundary conditions." Next, we must state the reactions which are known from our
particular loading. We can see from the truss that R1 = 0, R2 = -P, and R6 = 0. Entering the
known displacements and reactions into our matrix we get:
We can now finish our problem by solving this matrix for u1, u2, and u6:
u1 = - PL/AE
u2 = -4PL/AE
u6 = -PL/AE
This application of FEA to a simple three-member truss shows in general how the method
works. Most applications to engineering problems, however, are much more complex.
Such analyses require large numbers of elements and nodes in order to accurately
represent the physical system being studied. These analyses inevitably require the
application of a computer.
For Virginia Tech engineering students interested in FEA, one undergraduate course is
available: ESM 4734 - An Introduction to the Finite Element Method. In this course,
students study the theory and application of FEA to problems in various fields of
engineering and applied sciences. The pre-requisite for the course is: ESM 2074 -
Computational Methods.
References
Finite Element Primer, Bruce Irons and Nigel Shrive, John Wiley & Sons, Inc., 1983.