Anda di halaman 1dari 55

Analysis of Mechanical Structures using Finite Element Method

3.4. Boundary conditions. Introducing supports

3.4.1. Boundary Conditions [17]

A key strength of the FEM is the ease and elegance with which it handles arbitrary
boundary and interface conditions. This power, however, has a down side. A big hurdle
faced by FEM newcomers is the understanding and proper handling of boundary
conditions. Below is a simple recipe for treating boundary conditions.

3.4.1.1. Essential and Natural B.C.

The key thing to remember is that boundary conditions (BCs) come in two basic flavors:
essential and natural.
 Essential BCs directly affect DOFs, and are imposed on the left-hand side vector
u.
 Natural BCs do not directly affect DOFs and are imposed on the right-hand side
vector f.
The basic recipe is:
Q1. If a boundary condition involves one or more degrees of freedom in a direct way,
it is essential. An example is a prescribed node displacement.
Q2. Otherwise it is natural.
The term “direct” is meant to exclude derivatives of the primary function, unless those
derivatives also appear as degrees of freedom, such as rotations in beams and plates.

3.4.1.2. B.C. in Structural Problems

Essential boundary conditions in mechanical problems involve displacements (but not


strain-type displacement derivatives). Support conditions for a building or bridge problem
furnish a particularly simple example. But there are more general boundary conditions that
occur in practice. Astructural engineer must be familiar with displacement B.C. of the
following types.
Ground or support constraints. Directly restraint the structure against rigid body
motions. Symmetry conditions. To impose symmetry or antisymmetry restraints
at certain points, lines or planes of structural symmetry. This allows the

100
Analysis of Mechanical Structures using Finite Element Method

discretization to proceed only over part of the structure with a consequent savings
in modeling effort and number of equations to be solved.
Ignorable freedoms. To suppress displacements that are irrelevant to the problem.
In classical dynamics these are called ignorable coordinates. Even experienced
users of finite element programs are sometimes baffled by this kind. An example
are rotational degrees of freedom normal to smooth shell surfaces.
Connection constraints. To provide connectivity to adjoining structures or
substructures, or to specify relations between degrees of freedom. Many
conditions of this type can be subsumed under the label multipoint constraints or
multifreedom constraints. These can be notoriously difficult to handle from a
numerical standpoint.

3.4.2. Introducing suports

The finite element structure analysis is subject to a set of boundary conditions which, in
the case of mechanical field problems, have to at least cancel possible kinematic
displacements under the action of the loads placed.
The removal of kinematic movements of with the analyzing finite elements structure is
done by the cancelling possible displacements (usually shifts) associated during
preprocessing with geometric entities and, after meshing, with nodes in accordance with
the coordinate system axes (global or local), preliminarily adopted.
Choosing the geometric entities, nodes and degrees of freedom with canceled
displacements is done while modeling so that after the analysis it results an increased
closeness to the real model. Also, in the case of boundary conditions modeling, advanced
programs can perform analyses with several sets (variants) of boundary conditions, so by
comparison, the most unfavorable state of structure strain can be identified.
The visualization using the analysis model of canceled displacements associated to the
points (nodes) in relation to the straight or circular axes of the coordinate system used, can
be made by combining the appropriate symbols corresponding to rotation and translation
displacement in Table 3, page 39.
Combining these symbols leads to the synthetic symbols shown in Table 4, page 40. These
are used for visualizing links to the point (node) level in relation to a tri-ortogonal
coordinate system with the designated axes (straight line or circular) 1, 2 and 3. In addition,
Table 4 shows the forces of reaction and the corresponding displacements, each associated
to a symbol.

101
Analysis of Mechanical Structures using Finite Element Method

The boundary conditions ensure complete fixing of the base structure with respect to the
base of the settlement in order to preventing the rigid body movements of the ensemble
structure. In general, the boundary conditions imposed certain values of degrees of
freedom in the restraint points of the structure. If the structure does not have enough
circumstances bearing a certain load case can generate shifts with infinite value.

3.4.2.1. The minimum support for a plane structure (2D)

Regardless of the distribution of external forces, the boundary conditions must cancel the
shift in relation to the two axes in two plane structure (x, y) and rotation about the axis
normal to the plane of the structure (the axis z) - Fig. 89.

Fig. 89 Minimal conditions for restraints

3.4.2.2. The minimum support for a volume structure (3D)

Regardless of the distribution of external forces, the minimum number is six and the
boundary conditions have to cancel the three shift in relation to the three axes (x, y, z) and
three rotations in relation to the axes correspond to those.
To fix the structure there are available numerous combinations of supports that ensures
that the structure rigid body movements are prevented. A possible example, see the figure
below (Fig. 90), where we used a hinged spherical bearing which block in point A the 3
possible translations of the body and three simple supports prevent rotation relative with:
the axis (z) - bearing B, axis (y) - bearing C and axis (x) - bearing D.

102
Analysis of Mechanical Structures using Finite Element Method

Fig. 90 Constraint of 3D structures

3.4.3. Restraints in CATIA

The following table shows the controls modeling of various geometric constraints
provided by the CATIA environment.

Clamp

Surface slider

Slider

Sliding pivot

Ball joint

Pivot

User – defined restraint

Isostatic restraint

103
Analysis of Mechanical Structures using Finite Element Method

3.5. Load modeling

Finite element modeling and analysis of mechanical structures consider loads as initial,
known data that are introduced in various ways so as to achieve a low rate of errors. Most
advanced programs that are based on FEM allow the model analysis with the consideration
of multiple sets (variants) of possible loads, in order to highlight the most unfavorable
states of strain.
The loads acting on a finite element modeled mechanical structure, usually in the form of
forces (moments), movements and temperatures, variation in space can be concentrated or
distributed, using the variation in time - static or dynamic.
Concentrated loads are considered to be acting on the finite element structure nodes. They
are actually in theory quasi-singularities, next to them resulting cvasiinfinite stresses.
However, in computer-aided design, loads can be used as concentrated forces when
studying displacement fields, strains and stresses in remote areas from their application
nodes.

Fig. 101 Different trends of loads

Distributed loads, with different laws of variation (constant linear, parabolic, etc.) may
interact externally - on a line or on a surface - or internally – in volume. Using these loads
allows modeling with reduced deviations from the real case, and thus, unlike the case of
using concentration tasks, the fields from the loading areas can be considered as reliable
for designing. The possibility of taking into consideration the distributed forces of inertial
type (linear, centrifugal and/ or mass) also leads to the increase in design accuracy and,
therefore, in the results obtained.
Loading is considered as being static when its value slowly increases from zero to the
nominal value (Fig. 101, a). The increase is so slow that inertial forces have extremely
low values and, therefore, they can be neglected for the finite element analysis.
Dynamic loads, variable in time, can present shock when large variations of intensity in
short periods of time (Fig. 101, b), regular (Fig. 101, c) or, generally, random (Fig. 101,
d) occur.

110
Analysis of Mechanical Structures using Finite Element Method

In some practical situations when the load values pertaining to the type of force cannot
be identified, but the displacement fields as a result of pretentioning montage or strain
restriction are partially or entirely known, as input data for the analysis of movements
can specify the structure nodes finite elements.

3.5.1. Loads in CATIA

The following table presents modeling commands load type forces, moments, temperature
and movement provided by the CATIA environment.

Pressure
Distributed Force
Moment
Bearing Load
Imported Force
Imported moment
Acceleration

Rotation Force

Line Force Density


Surface Force Density
Volume Force Density

Force Density
Enforced Displacements
Temperature Field

Temperature Field for Thermal Solution

Combined Loads

Assembled Loads

111
Analysis of Mechanical Structures using Finite Element Method

4.1. General considerations about solving phase2

While the pre-processing and post-processing phases of the finite element method are
interactive and time-consuming for the analyst, the solution is often a batch process, and
is demanding of computer resource.
The governing equations are assembled into matrix form and are solved numerically. The
assembly process depends not only on the type of analysis (e.g. static or dynamic), but
also on the model's element types and properties, material properties and boundary
conditions [47].
In the case of a linear static structural analysis, the assembled equation is of the form:

Kd = r, Eq. 15

where K is the system stiffness matrix, d is the nodal degree of freedom (dof) displacement
vector, and r is the applied nodal load vector. To appreciate this equation, one must begin
with the underlying elasticity theory. The strain-displacement relation may be introduced
into the stress-strain relation to express stress in terms of displacement.
Under the assumption of compatibility, the differential equations of equilibrium in concert
with the boundary conditions then determine a unique displacement field solution, which
in turn determines the strain and stress fields. The chances of directly solving these
equations are slim to none for anything but the most trivial geometries, hence the need for
approximate numerical techniques presents itself.
A finite element mesh is actually a displacement-nodal displacement relation, which,
through the element interpolation scheme, determines the displacement anywhere in an
element given the values of its nodal dof. Introducing this relation into the strain-
displacement relation, we may express strain in terms of the nodal displacement, element
interpolation scheme and differential operator matrix.
Recalling that the expression for the potential energy of an elastic body includes an
integral for strain energy stored (dependent upon the strain field) and integrals for work
done by external forces (dependent upon the displacement field), we can therefore express
system potential energy in terms of nodal displacement [47].

2
The content of this chapter (marked with [47]) was taken from the paper: Roensch, S. “The Finite
Element Method:A Four-Article Series" with the written consent of the author, whom I thank.

118
Analysis of Mechanical Structures using Finite Element Method

Applying the principle of minimum potential energy3, we may set the partial derivative of
potential energy with respect to the nodal dof vector to zero, resulting in: a summation of
element stiffness integrals, multiplied by the nodal displacement vector, equals a
summation of load integrals.
Each stiffness integral results in an element stiffness matrix, which sum to produce the
system stiffness matrix, and the summation of load integrals yields the applied load vector,
resulting in Kd = r. In practice, integration rules are applied to elements, loads appear in
the r vector, and nodal dof boundary conditions may appear in the d vector or may be
partitioned out of the equation [47].
Solution methods for finite element matrix equations are plentiful. In the case of the linear
static Kd = r, inverting K is computationally expensive and numerically unstable. A better
technique is Cholesky factorization, a form of Gauss elimination, and a minor variation
on the "LDU" factorization theme. The K matrix may be efficiently factored into LDU,
where L is lower triangular, D is diagonal, and U is upper triangular, resulting in LDU d
= r. Since L and D are easily inverted, and U is upper triangular, d may be determined by
back-substitution.
Another popular approach is the wavefront method, which assembles and reduces the
equations at the same time. Some of the best modern solution methods employ sparse
matrix techniques. Because node-to-node stiffnesses are non-zero only for nearby node
pairs, the stiffness matrix has a large number of zero entries. This can be exploited to
reduce solution time and storage by a factor of 10 or more. Improved solution methods are
continually being developed. The key point is that the analyst must understand the solution
technique being applied [47].
Dynamic analysis for too many analysts means normal modes. Knowledge of the natural
frequencies and mode shapes of a design may be enough in the case of a single-frequency
vibration of an existing product or prototype, with FEA being used to investigate the
effects of mass, stiffness and damping modifications. When investigating a future product,
or an existing design with multiple modes excited, forced response modeling should be
used to apply the expected transient or frequency environment to estimate the
displacement and even dynamic stress at each time step [47].
This discussion has assumed h-code elements, for which the order of the interpolation
polynomials is fixed. Another technique, p-code, increases the order iteratively until
convergence, with error estimates available after one analysis. Finally, the boundary
element method places elements only along the geometrical boundary. These techniques
have limitations, but expect to see more of them in the near future [47].

3
See Annex A.3
119
Analysis of Mechanical Structures using Finite Element Method

4.2. Modern design problem

 determining the behavior of a system under the effect of external actions (Fig.
113);

Fig. 113 System reaction

 which is the response (Fig. 114) of the system when subjected to external actions
(changes in the forces, temperatures and so on).

Fig. 114 Internal and external behavior of a system subject to a mechanical stress [44]

120
Analysis of Mechanical Structures using Finite Element Method

Each time an FE model is solved, it can help create a vast amount of results data. The
ability to process the data and quickly gain an understanding of the model behavior is
important for a fast analysis turnaround. The postprocessor should therefore allow full
control of results selection and include a robust and varied set of tools to manage and
display results, while at the same time facilitate easy comprehension of the data. Results
viewing becomes more complex with highly idealized models, so the postprocessing tools
should provide the ability to easily view appropriate results quantities on shell and beam
elements [50].

5.2. General considerations about the post-


processing phase

After a finite element model has been prepared and checked, boundary conditions have
been applied, and the model has been solved, it is time to investigate the results of the
analysis. This activity is known as the post-processing phase of the finite element method.
Post-processing begins with a thorough check for problems that may have occurred during
solution. Most solvers provide a log file, which should be searched for warnings or errors,
and which will also provide a quantitative measure of how well-behaved the numerical
procedures were during solution.
Next, reaction loads at restrained nodes should be summed and examined as a "sanity
check". Reaction loads that do not closely balance the applied load resultant for a linear
static analysis should cast doubt on the validity of other results. Error norms such as strain
energy density and stress deviation among adjacent elements might be looked at next, but
for h-code analyses these quantities are best used to target subsequent adaptive remeshing
[47].
Once the solution is verified to be free of numerical problems, the quantities of interest
may be examined. Many display options are available, the choice of which depends on the
mathematical form of the quantity as well as its physical meaning. For example, the
displacement of a solid linear brick element's node is a 3-component spatial vector, and
the model's overall displacement is often displayed by superposing the deformed shape
over the undeformed shape.
Dynamic viewing and animation capabilities aid greatly in obtaining an understanding of
the deformation pattern. Stresses, being tensor quantities, currently lack a good single
visualization technique, and thus derived stress quantities are extracted and displayed.
Principal stress vectors may be displayed as color-coded arrows, indicating both direction
and magnitude. The magnitude of principal stresses or of a scalar failure stress such as the

155
Analysis of Mechanical Structures using Finite Element Method

Von Mises stress may be displayed on the model as colored bands. When this type of
display is treated as a 3D object subjected to light sources, the resulting image is known
as a shaded image stress plot. Displacement magnitude may also be displayed by colored
bands, but this can lead to misinterpretation as a stress plot [47].
An area of post-processing that is rapidly gaining popularity is that of adaptive remeshing.
Error norms such as strain energy density are used to remesh the model, placing a denser
mesh in regions needing improvement and a coarser mesh in areas of overkill. Adaptivity
requires an associative link between the model and the underlying CAD geometry, and
works best if boundary conditions may be applied directly to the geometry, as well.
Adaptive remeshing is a recent demonstration of the iterative nature of h-code analysis
[47].
Optimization is another area enjoying recent advancement. Based on the values of various
results, the model is modified automatically in an attempt to satisfy certain performance
criteria and is solved again. The process iterates until some convergence criterion is met.
In its scalar form, optimization modifies beam cross-sectional properties, thin shell
thicknesses and/or material properties in an attempt to meet maximum stress constraints,
maximum deflection constraints, and/or vibrational frequency constraints.
Shape optimization is more complex, with the actual 3D model boundaries being
modified. This is best accomplished by using the driving dimensions as optimization
parameters, but mesh quality at each iteration can be a concern [47].
Another direction clearly visible in the finite element field is the integration of FEA
packages with so-called "mechanism" packages, which analyze motion and forces of
large-displacement multi-body systems.
A long-term goal would be real-time computation and display of displacements and
stresses in a multi-body system undergoing large displacement motion, with frictional
effects and fluid flow taken into account when necessary. It is difficult to estimate the
increase in computing power necessary to accomplish this feat, but 2 or 3 orders of
magnitude is probably close. Algorithms to integrate these fields of analysis may be
expected to follow the computing power increases [47].
In summary, the finite element method is a relatively recent discipline that has quickly
become a mature method, especially for structural and thermal analysis. The costs of
applying this technology to everyday design tasks have been dropping, while the
capabilities delivered by the method expand constantly. With education in the technique
and in the commercial software packages becoming more and more available, the question
has moved from "Why apply FEA?" to "Why not?". The method is fully capable of
delivering higher quality products in a shorter design cycle with a reduced chance of field
failure, provided it is applied by a capable analyst. It is also a valid indication of thorough

156
Analysis of Mechanical Structures using Finite Element Method

design practices, should an unexpected litigation crop up. The time is now for industry to
make greater use of this and other analysis techniques [47].

5.3. Results

The primary results in a finite element analysis are grid point displacements and rotations.
Element results such as stresses, strains, and strain energy density are derived from those
results. Other results include element forces, MPC forces, SPC forces, and grid point
forces. Results of a finite element analysis are post-processed using a graphical tool [4].

5.3.1. Displacements [4]

Displacements and rotations are computed in linear static, and frequency response
analyses. In addition, in frequency response velocities and acceleration are computed.
Eigenvectors are the primary result in a normal modes and buckling analyses. In a normal
modes analysis, they are normalized with respect to the mass matrix or with respect to the
maximum vector component. In a buckling analysis, the latter always applies.
Displacements, velocities, accelerations, and eigenvectors are grid point results. They are
plotted as a deformed structure, or as a contour on the undeformed structure. Some post-
processors, such as Ansys or Catia, also allow the animation of the displacements [4].

Fig. 139 Original and deformed displacement contour plot [10]

157
Analysis of Mechanical Structures using Finite Element Method

5.3.2. Stresses [4]

The stresses are secondary results in a static analysis.


Stresses near notches and other sharp corners, point loads and boundary conditions, and
rigid elements are often unreliable due to the singularities in these points. A mesh
refinement in such places can improve the stress prediction. A theoretically infinite stress
cannot be predicted by finite elements.
Stresses are primarily calculated at the Gauss integration points. These give the most
accurate prediction. However, element stresses, corner stresses, and grid point stresses
are provided.
Element stresses are calculated at the centroid of the element. They should only be post-
processed using an assign plot. Contouring of element stresses vastly underestimates the
extreme values due to the smearing across element boundaries.
The stresses of interest are usually found on the surface of a structure. Mesh refinement
will actually not just improve the stress prediction but also change the location of the point
of stress evaluation. Therefore, it is common practice to use a skin of thin membrane
elements in 3D modeling, or rod elements in 2D modeling, to evaluate the stresses on
element surfaces or edges, respectively. This method is accurate since it considers the
correct condition of a stress-free boundary if no load is applied to the boundary. The
method of skinning a model also has the advantage of much faster post-processing of solid
models because only the membrane skin needs to be displayed.
Besides assign plots, elements stresses can be viewed in tensor plots that can help in the
evaluation of the load path in a structure by evaluating the principal stress directions.
Corner stresses are computed by extrapolating the stresses from the Gauss points to the
element grid points. Corner stresses are plotted in a contour plot. Corner stresses for solid
elements are not available for normal modes analysis.
Grid point stresses are computed by averaging the corner stresses contributions of the
elements meeting in a grid point. The averaging does not consider the condition of a
stress-free boundary. Further, interfaces between different materials, where a stress jump
normally can be observed, are not considered correctly because of the smearing of the
stress. Grid point stresses are plotted in a contour plot. For first order elements, grid point
stresses do not provide higher accuracy over element stresses. For second order elements,
the stress prediction might improve by using grid point stress over element stresses,
considering the weaknesses mentioned above [4].
The next step is to determine whether a part will break by comparing the stress values
from the analysis results to the strength of the material. Every metal and most plastics have

158
Analysis of Mechanical Structures using Finite Element Method

what’s called a yield strength and an ultimate strength. If the stress within the part exceeds
the material yield strength, then the part will not return to it’s original shape when the load
is removed. Although the part is still in one piece, it’s going to remain bent, which
generally isn’t good. If the stress exceeds the ultimate strength, then the part will fracture
and break. Ideally, the whole aim of the analysis is to make sure the stresses within the
part remain below the yield strength of the material [10].

Fig. 140 Stress Contour (units are MPa) [10]

In our suspension upright example, the stress contour shows a maximum stress of 325
MPa, which is above the material’s yield strength of 250 MPa, but below the tensile
strength of 345 MPa. Sothis means the part will bend under these loads, but won’t actually
break in two. Still, a bent upright is of no use to anyone so one possible change to make
to the part would be to increase the radius of the fillet in the high stress area to add some
extra material.

Fig. 141 Increased Fillet Radius [10]

159
Analysis of Mechanical Structures using Finite Element Method

And after rerunning the analysis, it’s clear that the maximum stress in the part has dropped
to a much more acceptable value of 120 MPa. Much better to have figured this out now,
rather than after having parts made and tested (not to mention cheaper and safer!).

Fig. 142 Results for Revised Design [10]

It’s clear that there are many benefits to using this type of simulation tool in engineering:
reduced costs; reduced design time; being able to assess a wide variety of designs; and
ending up with a stronger, lighter part.

5.3.3. Strains [4]

Strains are secondary results. They are calculated as elements strains. Remarks made
above on element stresses apply here too.

5.3.3.1. Strain Energy Densities

Strain energy densities are secondary results in static and normal modes analysis. They
are calculated as element strain energy densities. Remarks made above on element
stresses apply here too.

5.3.3.2. Forces

Element forces, MPC forces, SPC forces, and grid point forces are printed as tabulated
output.

160
2. GENERAL PROBLEM
DEFINITION

2.1. General considerations on the method

The computer-aided design of a mechanical system involves identifying the shape and size
of its constitutive elements by using advanced software in solid modeling, in the analysis
of physical processes, in the synthesis and/or optimization.
The mechanical system of a vehicle, installation, machinery, robot, aircraft, etc. can be
divided in assemblies which in turn are made of subassemblies and distinct constitutive
parts, called machine parts. Subassemblies can also be made up of other subassemblies or
other distinct constitutive parts. Both the assemblies and the subassemblies of the
mechanic system are standalone entities, useful for the structural study of the system and
for the technical optimization of the assemblage. The components of a mechanical system
(or subsystem) are in permanent, direct fixed interaction (removable or non-removable) or
direct moving interaction (without lubrication) or indirect (with lubrication). There are
many types of such connections, in terms of design and depending on functional and
technological necessities.
The practice of designing and building mechanincal systems is in permanent development,
constantly updading any performance achievements regarding the means, methods,
possibilities and technologies available. In terms of functionality, different mechanical
systems present certain machinery elements and/ or subassemblies which have identical
or quasiidentical functions. Gradually, well-known design algorithms, as well as
technologies specialized in executing and assemblying have been developed for these
elements or subassemblies, seldom called machine parts.
Analysis of Mechanical Structures using Finite Element Method

The emergence and booming development of numerical computer performance in the last
decades allowed to obtain advanced software which implies new possibilities of
modeling, analysis and synthesis of subassemblies, of elements and/ or machine parts.
Most of these advanced programs rely on numerical methods, including the best knownn
one, the Finite Element Method in the practice field of physical phenomena analysis.
Fig. 11 presents a general outline of the main activities for designing and implementing
mechanical products, especially aiming at the identification of a hierarchy of the programs
based on FEM.

COMPUTER AIDED ENGINEERING

COMPUTER AIDED DESIGN COMPUTER AIDED MANUFACTURING

Statically, kinematic and dynamic analysis Based FEM analysis of technological


of mechanical systems process

SOLID Design of technological process


Based FEM analysis of mechanical systems
MODELLING

MEF optimization and syntesis Product manufacturing

CAE / CAM interface Experimental modelling

Fig. 11 The CAD-CAM module interface

Solid modeling is the sum of the activities describing objects in terms of geometry and
physics or even of spatial domains in order to create a representation of them using the
numerical computer. The geometric shapes of the systemic studied elements can be
modeled with certain degrees of idealization based on the current design stage. In the
primary stages of complex mechanical systems, the degree of idealization related to
modeling elements is increased, with the main objective of analyzing and sinthetizing on
a main (functional) level, unlike the final stages, when modeling is made with as little
deviations as possible from the nominal shape and size, where the main objective relates
to the aspects of designing in detail.
The structural, static, kinematic and dynamic analysis of mechanical systems implies an
ongoing study of the correlations between parameters and the actual characteristics and
those imposed to the mechanical system, considering models with an increased degree of
idealization for the constitutive elements. In order to arrive at the required characteristics,

22
Analysis of Mechanical Structures using Finite Element Method

appropriate changes can be made, followed by reanalyzing or by synthesis and/ or


optimization models. Consequent to these operations result the main dimensional and
physical parameters of the elements and the subassemblies pertaining to the mechanical
system.
The finite element analysis of simple elements or subassemblies of a mechanical system,
using the results obtained in the previous stage, implies geometric remodeling and a
detailed specification of the shape and in the same manner, finite elements modeling with
increased accuracy. Some advanced programs based on FEM have special optimization
and shape synthesis modules. Such programs allow generating geometric shapes which
respect conditions of equal resistance, minimum volume or minumum mass. Furthermore,
finite elements are commonly used also for the dynamic study of mechanic systems with
elastic deformable elements.
The CAD-CAM module interface (Fig. 11) connects the engineering design software and
the changes in shape and size in order to adjust to the technological processes available.
The study of such processes (deforming, casting, diffusion, etc.) using finite element
analysis allows to determine the shape and dimensional parameters necessary for design
devices and for the establishment of the optimal technological regimes.
It is possible to rapidly achieve mechanical material products with high performance by
massive introduction of numerical computers with advanced programs for both
dimensional synthesis and implementation. The stages of computer-aided design and
manufacturing of technical systems modifies perpetually and is constantly being updated
in accordance with the progress in the field of modeling, analysis, as well as the
development of techical performance of computing systems.
In order to obtain professional products, modeling and theoretical analysis of real
phenomena can be done through two main directions: by studying theories on general
situations and practical studies through the analysis of concrete practical cases. Fig. 12
outlines the main steps listed in both directions. It also shows that the theoretical analyses
aims at the practical results is performed based on fundamental studies. General theoretical
analysis of real phenomena are based on computational theoretical models that are
assigned appropriate mathematical models.
The computational theoretical model (Fig. 12) is an approximation of the real model
which involves identifying the shape and size of the geometric domain and the physical
parameters known as the qualitative indication of unknown physical parameters. For
known geometrical and physical parameters, the variational functions and their limit
values are being established.
The analytical mathematical model (Fig. 12), associated with the computational model
made, is in most cases a system of differential and/ or integro-differential equations, with

23
Analysis of Mechanical Structures using Finite Element Method

complementary sets of appropriate initial and boundary conditions. In many cases it is


possible to describe mathematically physical phenomena, synthetized in the computational
model using a variational calculus through a functional. This description, often used due
to the simplicity of the methods and algorithms for solving the mathematical model,
developed in various forms, especially for mechanical engineering problems.

Fig. 12 Theoretical and practical approaches

To solve by the classical approach, following rough approximations regarding geometry,


initial and boundary conditions, and material properties applied on theoretical models of
mathematical calculus, we obtain simplified analytical mathematical models which can be
processed using the manual calculus, slide rule or the calculator.
For example, the calculus model of bendable mechanical structures, with the methods of
the theory of elasticity and strength of materials, we obtain specific mathematical models
leading to simple calculus relationships (Navier, Juravski, etc.) for different geometrical
fields (bars, plates, shells, tubes, discs, etc..) and specific physical conditions.
In order to increase the precision of the results obtained by classical methods (Fig. 13),
numerical methods through small, usually controllable approximations in respect to
geometry, boundary conditions and material properties, lead to numerical modeling that
can be solved only by numerical computer. The practice of numerical modeling which
involves the study of physical phenomena in continuous environments by splitting them
into smaller subdomains called finite elements, developed and became a business

24
Analysis of Mechanical Structures using Finite Element Method

performance programs (NASTRAN, ANSYS, ALGOR, COSMOS, CATIA, etc.) that


have pre and post-processors with advanced facilities of data input and processing.
Theoretical and practical studies applicable in the design of specific machine elements are
based on the techincal calculus model. Since advanced programs that are based on FEM
deal with analysis, preliminarily, sizing calculation is required (predimensioning) using,
in particular, traditional methods of strength of materials. In order to use advanced
software to analyze and optimize the shape of the machine elemet structure, predesigned
in both shape and size, it is necessary to complete one or more models of analysis adjusted
to the numerical method on which the program is based.

Fig. 13 Classical and numerical analysis

25
Analysis of Mechanical Structures using Finite Element Method

This paper particularly seeks to explore the finite element analysis software for the
advanced design of machine elements (organs) and/ or mechanical system subassemblies.
Various finite element analyses presented are based on the definition of the problem in the
general context of the design and ends with the visualization, analysis and synthesis of the
results.
These analyses are performed in a CATIA environment, characterized by a high level of
integration of CAD and CAE modules, as it was shortly highlighted in the first chapter.
So, as of late, thanks to this integration and high level of communication of the human
operator with the programming environment, the design engineers can easily access CAE
modules (Computer Aided Engineering) of finite element analysis. This requires that
designers have knowledge of dealing with finite element analysis models and processing
results. This chapter presents the general problems related to finite element modeling
including the geometry, the material properties, the boundary conditions and commonly
encountered difficulties in various situations in practice.

2.2. FEM terminology1 [17]

The “degrees of freedom” term, as well as “stiffness matrix” and “force vector,” originated
in structural mechanics, the application for which FEM was invented. These names have
carried over to non-structural applications. Classical analytical mechanics is that invented
by Euler and Lagrange in the XVIII century and further developed by Hamilton, Jacobi
and Poincar´e as a systematic formulation of Newtonian mechanics. Its objects of attention
are models of mechanical systems ranging from material particles composed of
sufficiently large number of molecules, through airplanes, to the Solar System. The spatial
configuration of any such system is described by its degrees of freedom or DOF. These
are also called generalized coordinates. The terms state variables and primary variables
are also used, particularly in mathematically oriented treatments [17].
If the number of degrees of freedom is finite, the model is called discrete, and continuous
otherwise. Because FEM is a discretization method, the number of DOF of a FEM model
is necessarily finite. They are collected in a column vector called u. This vector is called
the DOF vector or state vector. The term nodal displacement vector for u is reserved to
mechanical applications. In analytical mechanics, each degree of freedom has a
corresponding “conjugate” or “dual” term, which represents a generalized force. In

1
The content of this chapter (marked with [17]) was taken from the paper Felippa, C.A.:
Introduction to Finite Element Methods, lecture notes, with the written consent of the author, whom
I thank.

26
Analysis of Mechanical Structures using Finite Element Method

variational mathematics this is called a duality pairing. In non-mechanical applications,


there is a similar set of conjugate quantities, which for want of a better term are also called
forces or forcing terms. They are the agents of change.
These forces are collected in a column vector called f. The inner product fT u has the
meaning of external energy or work. Energy is the capacity to do work. Thus energy and
work potentials are the same function (or functional), but with signs reversed. Just as in
the truss problem, the relation between u and f is assumed to be of linear and
homogeneous. The last assumption means that if u vanishes so does f. The relation is then
expressed by the master stiffness equations:

Ku = f. Eq. 1

K is universally called the stiffness matrix even in non-structural applications because no


consensus has emerged on different names [17].
The physical significance of the vectors u and f varies according to the application being
modeled, as illustrated in Table 1. If the relation between forces and displacements is
linear but not homogeneous, equation (Eq.1) generalizes to

Ku = fM + fI . Eq. 2

Here fI is the initial node force vector and fM is the vector of mechanical forces.

Table 1 Significance of u and f in Miscellaneous FEM Applications

Application Problem State (DOF) vector u Conjugate vector f


represents represents
Structures and solid Displacement Mechanical force
mechanics
Heat conduction Temperature Heat flux
Acoustic fluid Displacement potential Particle velocity
Potential flows Pressure Particle velocity
General flows Velocity Fluxes
Electrostatics Electric potential Charge density
Magnetostatics Magnetic potential Magnetic intensity

27
Analysis of Mechanical Structures using Finite Element Method

implements that model, or write the program yourself. This is explicit modeling. It requires
far more technical expertise, resources, experience and maturity than implicit modeling.
But for problems that fall out of the ordinary it could be the right thing to do.

Fig. 17 The Physical FEM. The physical system (left box) is the source of the simulation
process. The ideal mathematical model (should one go to the trouble of constructing it) is
inessential

In practice a combination of implicit and explicit modeling is common. The physical


problem to be simulated is broken down into subproblems. Those subproblems that are
conventional and fit available programs may be treated with implicit modeling, whereas
those that require special handling may only submit to explicit modeling.

2.4. The theoretical quasi-general model for finite


element analysis of a mechanical subassembly
element

Fig. 18 shows a theoretical quasi-general model of finite element analysis of a subset of


a mechanical element [40]. The continuous structure of this element, finite volume V and
surface S, is made of solid materials with different behavior (linear, nonlinear), described
by a specific laws. For the material or materials used, the values of density, the mechanical
properties (density, elasticity matrices, damping factor etc.) and thermal (thermal
expansion coefficient, specific heat, etc.) and the allowable resistances (traction,
compression, usually and shearing) are known.

31
Analysis of Mechanical Structures using Finite Element Method

Over the structure of the analyzing element act the following external forces: generalized
forces (Pi forces and / or moments, i = 1, 2 ... m) concentrated in points, generalized forces
distributed on a line (forces or moments q, on the Cp line), on an area (the forces and / or
p moments on the Sp surface) and in volume (mass forces fg, centrifugal t and / or the
inertial fj).
The structure of the analyzing element operates in a limited temperature range between T0
initial temperature and Tf final temperature. In addition, the structure may be under the
action of thermal fields of temperature (on line distribution, on the surface or in volume)
and/ or external thermal fluxes.
In the category of external loads are also included the elastic deformations required from
certain areas of the structure through known values of δi displacements. This leads to an
imposed shape of the deformed state of the structure area, which in Fig. 18, is synthesized
by line Ci.
Direct interactions between the analyzed element structure with the structures of other
elements of the mechanical subassembly can also be simulated using finite element
analysis. These interactions may be permanent (void displacements usually imposed by
boundary conditions) and / or temporary, also taking into account the friction (measured
by the values of friction coefficients, μ) of the materials in the interacting areas and initial
distances, δ0.
The possibility of finite element analysis of the quasi-general model shown is conditioned
by the existence of a set of imposed boundary conditions, usually synthesized by canceled
shifts corresponding to certain points of the geometric field problem (e.g. surface area of
Fig. 18). The solvability of finite element analysis model with loads and boundary
conditions imposed is provided by the lack of possibilities in the structure’s kinematic
movement.
Under the action of loads and imposed boundary conditions, the analyzed structure is
deformed and within it, there are distributed internal forces called stresses. From a
geometric perspective, displacement fields, the strains and stresses, are quantitatively
described using the following displacement vectors:
[d] = [u v w]T, Eq. 3

of the strains (Fig. 18, b),


[ε] = [εx εy εz γxy γyz γzx]T Eq. 4

and namely, of the stresses (Fig. 18,c),


[σ] = [σx σy σz σxy σyz σzx]T, Eq. 5

32
Analysis of Mechanical Structures using Finite Element Method

with the components connected to the tri-orthogonal straight coordinate system XYZ.

Fig. 18 Theoretical quasi-general model of finite element analysis

Study using advanced programs based on FEM quasi-general model presented above and
with given values for all input parameters, to constructively design the element to which
is associated, usually aims to determine the fields of displacement, strain, stress, thermal
and in addition, of parameters (force, displacement, temperature, etc.) from the connecting
areas with other elements. The values and variations of these parameters, the functioning
conditions and tolerated material characteristics make possible the evaluation of strength,
rigidity and thermal characteristics of the structure analysis.

33
Analysis of Mechanical Structures using Finite Element Method

In the analysis and synthesis of mechanical assemblies elements it is unlikely to encounter


an problem to which it is associated model described above. Usually, geometric shapes,
loads and boundary conditions imposed to practical elements of mechanical
subassemblies, are reduced to particular cases, leading to a simplified modeling, increased
accuracy and a faster computational process. Based on these considerations, in the case
advanced programs which are based on FEM, specific finite elements have been developed
and software modules that allow solving the problems making connections with different
possible particular cases.

2.5. Types of solvable problems using finite elements


analysis

M.E.F. - Approximate solving method using a computer for a wide range of engineering
problems:
o Equilibrum problems – determining unknown, time-independent parameters, for
a steady state (linear or nonlinear static analysis, heat transfer analysis, the fluid
flow or the magnetic field distribution).
o Custom values problems - determining certain critical values of the physical
parameters, time invariables, equilibrium configurations and given boundary
conditions (analysis natural frequency analysis, flexure, laminar flow regimes,
resonance...).
o Propagation problems - unknown time-dependent parameters - the study of
transient regimes (dynamic analysis of elastic and inelastic structures, heat
transfer, unsteady flow).
Formulating an engineering problems involves:
 Identifying the type of problem;
 Identifying the working hypotheses adopted (the geometry in the problem’s field,
the material properties, variation field of main sizes, the functioning mode);
 Identifying the initial and boundary conditions.
In Table 2 are presented the main types of solvable problems with advanced programs that
are based on FEM, depending various criteria of quasi-general model customization. In
practice, these problems can be encountered separately or in combination, following
several criteria simultaneously customization.
Many practical applications with materializing in mechanical parts also include heat
transfer processes and so in order to design is necessary to know the specific fields through

34
Analysis of Mechanical Structures using Finite Element Method

thermal analysis. A part of the results of these tests, along with other types of loads can be
considered for the analysis of mechanical components and/ or, sometimes, system
subassemblies.
Table 2 Types of FEA problems
Quasi-general
Customization criterion Type of problem (analysis)
problem
Thermal
Type of fields
Mechanic
Unidimensional
Bidimensional
Type of domain
Tridimensional
Advanced
Combined
software
analysis based Static
Dependancy on time variables
on FEM of Dynamic
elements and Current
mechanical
Type of obtained values Custom vectors
systems
Stability
Liniar
Dependancy between parameters
Nonliniar
Possibility of considering No connections
connections With connections

The components of the mechanical systems have various three-dimensional (3D) shapes.
In many practical cases, the shapes of elements are or may be considere two-dimensional
(2D) - with one dimension much smaller than the other two - or one dimensional (1D) -
where one dimension is much larger than the other two . For the finite element analysis of
mechanical system assemblies with advanced programs based on FEM, without the
detailed consideration of direct interactions between parts, often components can be of
different forms (one-dimensional , two-dimensional and / or three-dimensional) . Starting
from the possible forms of the domain of the studied element, finite element analysis can
be three-dimensional, two-dimensional, one-dimensional or combined.
In terms of time-dependent loads, solving finite element model associated with the
mechanical system element is called static analysis - without considering time as a variable
- or dynamic analysis – with time-dependent unknown variables.

35
Analysis of Mechanical Structures using Finite Element Method

The study of the elements structure of mechanical systems with advanced software based
on FEM which leads to the determination of the field displacements, strains, stresses and
thermal as a result of the loads and the imposed boundary and the normal limit conditions,
is considered normal analysis. In addition, using the same types of programs, in case of
loading problems and abnormal boundary and limit conditions, limit states which may
occur during operation can also be analyzed. In this sense, it is very common in the practice
of design the stability analysis and also the analysis of vectors and their forms that lead to
the underlying causes of critical load flexure and, respectively, the custom frequencies
and the corresponding geometric configurations.
In terms of load-displacement dependencies and stress-strain can highlight the following
types of analysis : linear, geometrically nonlinear, physically nonlinear (material) or
geometrico-physically nonlinear. The first type of analysis is appropriate in cases of
structures with small displacements when loads remain invariable during deformation and
movement direction and can summarized, following a proportionality factor.
The second type of analysis corresponds to mechanical structures with large
displacements, disproportionate with the loads, with the variable directions during
deformation . The third type of analysis, unlike the first two, deals with the nonlinear,
elastic and plastic behavior of a material, by means of suitable stress-strain characteristics.
The last analysis is the general possible case when the two dependencies load-
displacement and stress-strain are nonlinear.
Most finite element analyses of several element (subassembly) structures usually do not
take into account the interactions between them, through the modeling of specific
connecting phenomena, considered "frozen" by the continuity of the whole finite element
structure at a nodal level. Starting from the importance of design processes (displacements,
strains, stresses and frictions) from the connecting local areas, in the last years, there have
been defined and implemented in many programs (including CATIA) specific connecting
elements (translation coupling, rotation, roto-translation, rigid or elastic) that take into
account the relative movements and contact elasticities that allow analysis links.

2.6. The model for analysis [40]

In order to efficiently and accurately simulate the behavior of mechanical systems or of


subassembly elements with advanced programs that are based on FEM, a specific analysis
model must be made. Finite element modeling, in many cases for analysis with the scope
of designing, as a consequence of some features related to the geometric configuration, of
the material behavior and physical phenomena, does not involve the consideration of the
whole structure.

36
Analysis of Mechanical Structures using Finite Element Method

Minimizing the analysis models without sacrificing the accuracy of monitored parameters
can be achieved by customizing the geometric configuration of inferior shapes (a 3D
structure to a 2D or an 1D one, a 2D structure to 1D), by considering for modeling the area
from the domain of the problem where the variations of unknown physical parameters are
significant, and/ or by using symmetry.

Fig. 19 Complete symmetry

By customizing the geometric configuration of the structure to be analyzed, the conformity


degree with the model reality decreases differently, in relation with both the values and
variations of known and/ or unknown parameters. For example, for the 3D structure of the
cylindrical piece, can be analyzed by reducing it to a 1D model, the accuracy of loading
parameters at input and unknown parameters of the support and concentration area
(dimensional jumps and keyways) decreases.
Through this customization, the finite element model size (the number of nodes) decreases
considerably and considering the simplicity of the model in conjunction with the results
obtained, it is seldom considered to be effective for checking the shafts of the standard
broadcast. In the case of special transmissions for increased accuracy checks, a complex
model (3D) analysis of the entire structure of the tree is made.

Fig. 20 Antisymmetry

37
Analysis of Mechanical Structures using Finite Element Method

The structures of elements of mechanical systems to be analyzed with finite elements may
have geometric, material, loading and/ or boundary conditions to a plane, two orthogonal
planes, three orthogonal planes or even multiple plan symmetry.
For the purpose of creating a mechanical structure model analysis, taking into account the
symmetry, it is necessary for the model to have a common symmetry regarding the
geometry, the material properties and the imposed boundary conditions. In terms of
loading, it is possible to create models as a result of complete symmetry (Fig. 19, b) when
the loading has the same symmetry of geometry and boundary conditions (Fig. 19, a), or
models (Fig. 20, b) generated by geometric and boundary conditions symmetry, and
loading antisymmetry (Fig. 20, a).
Setting boundary conditions which take into account in the case of modelings which
consider symmetries is being made by monitoring the accurate simulation of deformation
processes from the initial structure. For example, the bar-like structure of Fig. 19,a, a full
symmetry with a YZ parallel plane, the displacement after X is being canceled (TX) and
rotations after the X and Y axes (RX, RY) or for the same structure (Fig. 20, a) but with
an anti-symmetric load, the X and Y translations (TX, TY) are canceled.
The analysis of an asymmetric loading model shown in Fig. 21, a, in the case of the
geometric symmetry structure and linear behavior, can be done by solving the model
analysis associated to the half of the geometric domain for two sets of loads and boundary
conditions corresponding to complete symmetry (Fig. 21, b) and the load antisymmetry
(Fig. 21, c). The final state corresponding to the initial structure is obtained by summing
the results for the two sets.

Fig. 21 Asymmetric loading

38
Analysis of Mechanical Structures using Finite Element Method

Table 3. Mobilities of the joints


No. of
Canceled Displacement No. of free
Symbol canceled
displacements directions mobilities
mobilities

3 3

2 4
Straight line or
Rotation
circular line
2 4

1 5

3 3

2 4
Straight line
2 4

1 5

Translation
3 3

2 4

Circular line
2 4

1
5

39
Analysis of Mechanical Structures using Finite Element Method

Table 4. Reactions and displacements in joints


Symbol
Coordinate Free
Straight Circular Introduced reactions
system Displacements
axes axes

F1/F2/F3/M1/M2/M3 -

F2/F3/M1/M2/M3 T1

F1/F2/M1/M2/M3 T3

F2/M1/M2/M3 T1/T3

F1/F2/F3/M1/M2 R3

F2/F3/M1/M2 T1/R3

F1/F2/M1/M2 T3/R3

F2/M1/M2 T1/T3/R3

F1/F2/F3/M1/M3 R2

F2/F3/M1/M3 T1/R2

F1/F2/M1/M3 T3/R2

F2/M1/M3 T1/T3/R2

F1/F2/F3/M2 R1/R3

40
Analysis of Mechanical Structures using Finite Element Method

Table 4 (cont.)

F2/F3/M2 T1/R1/R3

F1/F2/M2 T3/R1/R3

F2/M2 T1/T3/R1/R3

In the case of symmetry of the structure to be analyzed using a plane, two planes or three
planes (Fig. 22, a), the analysis model is reduced to half, a quarter or, respectively, an
eighth (Fig. 22, b) of the geometric domain.

Fig. 22 Different types of simmetry

Full symmetry to an axis, in a random case nonreductive to the axial-symmetric one (Fig.
23, a), involves the shaping of an angular sector (Fig. 23, b) or when the problem is of an
axial-symmetrical nature, it leads to a plane model determined by the axial semisection
by structure.
The problem of creating the optimal finite element analysis model is complex depending
on the type of physical phenomena, the aimed requirements and performance of the
program used

41
Analysis of Mechanical Structures using Finite Element Method

Fig. 23 Full simmetry to an axis

2.7. Samples of analysis models

The following are a few models for analysis: one-dimensional model – bars structure (Fig.
24), two-dimensional model – surface (Fig. 25), three-dimensional model – volumes (Fig.
26), model for thermic analysis (Fig. 27).

Fig. 24 Bars structure model for analysis (support beam)

Fig. 25 Surface model for analysis (membrane)

42
Analysis of Mechanical Structures using Finite Element Method

Fig. 26 Volume model for analysis (reducer shaft)

Fig. 27 Model for thermic analysis (engine valve)

2.8. The general procedure of FEA

In general, there are three phases (Fig. 28) in any computer-aided engineering task:
 Pre-processing – defining the model and environmental factors to be applied to
it (typically a finite element model, but facet, voxel and thin sheet methods are
also used).
 Analysis solver (usually performed on high powered computers).
 Post-processing of results (using visualization tools).
This cycle is iterated, often many times, either manually or with the use of commercial
optimization software.

43
Analysis of Mechanical Structures using Finite Element Method

Fig. 28 Analysis procedure based on FEM [30]

44
Analysis of Mechanical Structures using Finite Element Method

analysis. The next step in the pre processing stage is to apply loads and boundary
conditions to the model.
Loads are usually defined as forces acting on a certain point, but can also be torques,
pressures, temperatures, or even a velocity or acceleration such as gravity. Boundary
conditions are constraints that define how and where the part is held or bolted on, and are
required to stop the part flying off into space when a force is applied. They basically tell
the software which nodes aren’t allowed to move during the analysis.
Once the model’s been meshed, materials defined and loads and boundary conditions
applied, you’ve now got a pre processed FEA model ready for solving [10].

3.1. Geometry modeling

3.1.1. General considerations about geometry modeling [40]

The components of a mechanical system, elastic deformable or elastic-plastic solids, have


a variety of geometric shapes. In terms of the geometric shape of mechanical elements can
be one-dimensional (1D), two-dimensional (2D), three-dimensional (3D) or combined.
The one-dimensional elements are objects wich having two dimensions smaller than the
third. In practice, these elements come together, usually called as bars, beams, flexible,
axles, shafts etc.
Depending on the shape of the axis, the bars can be straight (see Fig. 32, a, b) or curved
(Fig. 32, c); from the point of view of the required applications, it can be traction-
compression rods and / or bars required in bending and, after such changes in cross-section
along the axis, the bars may be constant orbars variable section.
The bars can be divided as follows:
(i) depending on the form of the axis: straight bars (Fig. 32, a, b) or curved
(Fig. 32, c);
(ii) from the point of view of applications: bars required tensile-compressive and
/ or flexural bars required;
(iii) according to the cross-section along the axis: bar with a constant or variable
section.
Flexible elements - wires, bands, belts (Fig. 32, d) – wich have reduced bending stiffness,
made of metallic or non-metallic type are considered having bar type geometry. Axles and
straight shafts, as 1D-dimensional car elements, that support other elements in rotating

47
Analysis of Mechanical Structures using Finite Element Method

machine are applied to bending and shear, respectively, torsion, bending, shear and tensile-
compressive.
Geometrically, the 1D-elements are described using two elements: the axial curve and the
cross-sectional shape and size.

Fig. 32 One-dimensional elements

Two-dimensional elements, as well as the constituent bodies having a much smaller size
than the other two (Fig. 33,a), is found in practice that the tiles, membranes, coatings,
dishes, etc.
Geometrically, these elements are described by defining the (i) shape and size of the
median surface and (ii) width (Fig. 33,b).
Depending on the median surface, the plates can be flat or curved plates with single or
double curvature. Commonly, the thickness of the plats used in practice is usually
constant. For the small values of the thickness, the plats are called membranes.

Fig. 33 Two-dimensional elements

48
Analysis of Mechanical Structures using Finite Element Method

Three-dimensional elements, also referred to as solid bodies or blocks, have three


dimensions about the same size and can not be reduced in any of the forms described
above (Fig. 34,a).
The combined elements have geometry domain composed by two or more parties who
framed the structures in the groups above presented. For example, Fig. 34,b shows the
structure of a body composed of two sub-domains, two-dimensional and three-
dimensional.
The concept of mechanical subassembly is primary, has general character, and is used to
identify and study the smallest components or parts thereof. Analytical calculation
methods of mechanical assemblies patterns associated to the elements are the same
methods used in the strength of materials, characterized by increased levels of idealization
and simplicity, in terms of the final design relationships.

Fig. 34 Three-dimensional elements

In the field of design, to analyze the components of a mechanical system (trusses, frames,
platforms, tanks, machinery or equipment, etc..) those can be considered as one-
dimensional geometric shapes, two-dimensional or three-dimensional, often cases lead to
practical combinations thereof.
The modeled element or the subassembly in order to study based on FEM may be referred
structure. In computer aided design, a structure described oneself using basic geometric
entities: point, line, area and / or volume, relative to one or more previously defined
coordinate systems. In addition, structures of analysis models, in particular mechanical
systems, may include the idealized elements (typically, the rigid) provided by the used
application software.
In order to geometric define the domains and subdomains of a structure, to prepare the
finite element analysis model, you can use a Cartesian coordinate system (Fig. 35,a),
cylindrical (Fig. 35, b) and / or spherical (Fig. 35, c). Each of these coordinate systems,
depending of the geometric configuration of the structure domain, can be global reference

49
Analysis of Mechanical Structures using Finite Element Method

system, to which the whole area of the problem, and a local reference system associated
with each of a sub-domain thereof.

Fig. 35 Coordinate systems

In the example of Fig. 36, the Cartesian coordinate system is a global reference system,
and the third cylindrical coordinate system are local system of reference.
Most FEM based application software have preprocessors that contain modules for
geometric modeling. In addition, these programs are able to import geometric models from
other programs mainly specialized in geometric modeling (AUTOCAD, EUCLID,
ProEngineering etc.)

Fig. 36 Local and global coordinate systems

Geometric modeling of the structure domain using advanced programs is performed using
basic geometric entities like lines, surface and / or volume that can be divided forming
finite element sets. In order to identify, for finite element modeling, basic geometric
entities have the following benchmarks: vertices, edges, and faces.

50
Analysis of Mechanical Structures using Finite Element Method

The structure domains and subdomains are generated by assembling the elementary
entities using the commands from software library used. The following are common
primary elementary entities: lines through points, arcs, circles, helix, involute, conic
(ellipse, parabola, hyperbola) for unidimensional domanin; surfaces by dots or lines for
surface domains; volume by points, lines or areas for volume domains.
Computer-aided geometric modeling of complex domains, based on elementary entities
primary, secondary implies elementary entities which are obtained by additional
operations (intersection, copy, offset, extrusion and so on).
Computer-aided geometric modeling of complex domains, based on primary elementary
entities, implies the secondary elementary entities which are obtained by additional
operations (intersection, copy, offset, extrusion and so on).

3.1.2. Basic concepts about 3D modeling

In 3D computer graphics, 3D modeling is the process of developing a mathematical


representation of any three-dimensional surface of object (either inanimate or living) via
specialized software. The product is called a 3D model. It can be displayed as a two-
dimensional image through a process called 3D rendering or used in a computer simulation
of physical phenomena. The model can also be physically created using 3D printing
devices.
Models may be created automatically or manually. The manual modeling process of
preparing geometric data for 3D computer graphics is similar to plastic arts such as
sculpting. Recently, new concepts in 3D modeling have started to emerge. Recently, a new
technology departing from the traditional techniques starts to emerge, such as Curve
Controlled Modeling that emphasizes the modeling of the movement of a 3D object
instead of the traditional modeling of the static shape [42].
3D models represent a 3D object using a collection of points in 3D space, connected by
various geometric entities such as triangles, lines, curved surfaces, etc. Being a collection
of data (points and other information), 3D models can be created by hand, algorithmically
(procedural modeling), or scanned. 3D models are widely used anywhere in 3D graphics.
Actually, their use predates the widespread use of 3D graphics on personal computers.
Many computer games used pre-rendered images of 3D models as sprites before
computers could render them in real-time [42].
Almost all 3D models can be divided into two categories.
 Solid - These models define the volume of the object they represent (like a rock).
These are more realistic, but more difficult to build. Solid models are mostly used
51
Analysis of Mechanical Structures using Finite Element Method

3.2. Assign material properties

From the point of view of the internal structure and implicitly the physical properties,
materials can be homogeneous, with the same structure and physical properties in all
points, or inhomogeneous, when this condition is not met. For homogeneous materials,
physical parameters are constant related to spatial geometric variables.
Depending on the type of variation of physical properties in the vicinity of each point,
there are isotropic materials, with properties independent of the direction to which it
relates, or anisotropic, with properties dependent to the direction by which it is considered.
The properties of isotropy and homogeneity are not mutually conditioned. A solid body
can be both homogeneous and isotropic, homogeneous and anisotropic, isotropic and
inhomogeneous or even inhomogeneous and anisotropic [40].

3.2.1. Modeling the mechanical behavior of materials

Solid materials which after deformation (change in shape) under the action of loads return
to the original shape and size are called elastic and those which do not comply with this
condition are called plastics. In the case of mechanical system elements, usually the
behavior of the material is elastic, or, sometimes, elasto-plastic. This occurred even at
mechanical system elements that, in order to obtain the final form, technological processes
based on plastic deformations cold or hot were used.
To determine the stress, deformation, thermal and displacement fields of a constant solid,
deformable under the action of external loads (Fig. 18,a), it is necessary to know the laws
of mechanical behavior of these materials.
The behavior of elastic materials can be linear, when the stress-strain dependencies are
linear, or nonlinear, when these dependencies are linear. Assuming linearity, for any
material, the variation of stress vector components (see Eq.5) between the initial state as
a reference (a random point P, Fig. 18,a) and the final state (point P ') is described as a
linear combination of the deformation vector components (see Eq.4), the relationship

[δ] = [E] [ε], Eq. 6

in which,

70
Analysis of Mechanical Structures using Finite Element Method

 E xxxx E xxyy E xxzz E xxyz E xxxz E xxxy 


E E yyyy E zzyy E yyyz E yyxz E yyxy 
 yyxx
E E zzyy E zzzz E zzyz E zzxz E zzxy 
E    yyxy  Eq. 7
 E yzxx E yzyy E yzzz E yzyz E yzxz E yzxy 
 E xzxx E xzyy E xzzz E xzyz E xzxz E xzxy 
 
 E xyxx E xyyy E xyzz E xyyz E xyxz E xyxy 

is the material elasticity matrix. Therefore, in theoretical cases, the stress and symmetric
strain tensors (tangential stresses and angular deformations apply to the relations: τij = τji
and γki = γik), the material elasticity matrix contains 36 elements, called constants
(modules) of elasticity. These constants characterize the material response to tri-
orthogonal axial loads (traction, compression and torsion in both directions).
Considering the elasticity symmetry, the matrix (Eq. 7) takes the following shape
 E xxxx E xxyy E xxzz E xxyz E xxxz E xxxy 
 E yyyy E zzyy E yyyz E yyxz E yyxy 

 E zzzz E zzyz E zzxz E zzxy 
E    
Eq. 8
 E yzyz E yzxz E yzxy 
 Simetric E xzxz E xzxy 
 
 E xyxy 

Consequently, for the linear elastic material it’s showing 21 virtually spring constants.
The materials consisting of parallel fibers embedded in a constitutive homogeneous mass
is characterized by the symmetry to a normal plane in the direction of the fibers. In this
case, the number of independent elastic constants is reduced to 13 and if OX is parallel
with the material fibers and the symmetry plane is OZY, the elasticity matrix (Eq. 8)
becomes
 E xxxx E xxyy E xxzz E xxyz 0 0 
 E yyyy E zzyy E yyyz 0 0 
 
 E zzzz E zzyz 0 0 
E    
Eq. 9
 E yzyz 0 0 
 Simetric E xzxz E xzxy 
 
 E xyxy 

71
Analysis of Mechanical Structures using Finite Element Method

In the case of materials with two orthogonal symmetry planes, called orthotropic or
orthogonally anisotropic, describing the linear elastic behavior of a material is made with
9 independent elastic coefficients, such as the non-zero elements of the elasticity matrix
 E xxxx E xxyy E xxzz 0 0 0 
 E yyyy E zzyy 0 0 0 
 
 E zzzz 0 0 0 
E    
Eq. 10

 E yzyz 0 0 
 Simetric E xzxz 0 
 
 E xyxy 

In this case, implicitly, there is a third plane of elastic symmetry, orthogonal with the first
two. The elastic constants corresponding to the attached tri-orthogonal coordinate straight
system directions are called main elasticity constants. The determination of these elasticity
constants is achieved by experimental tests on test tubes, taking into account the
hypothesis of the inexistence of "coupling" between normal and specific shear stresses
corresponding to the orthotropic axes.
Isotropic materials are characterized by some invariable types of behavior for every
defining direction and, therefore, the elasticity matrix, with the same form, regardless of
the direction of the reference axes of the coordinate system adopted in the practice of
design has the following configuration:
  2   0 0 0
   2  0  0 0

   2 0 0 0
E   
Eq. 11

 2 0 0
 Simetric 2 0 
 
 2

with the two independent elasticity coefficients defined by the relationships:


E
 Eq. 12
1   1  2 
E
 Eq. 13
21   

where E, the longitudinal elasticity module and ν, the transverse contraction coefficient,
are the technical elasticity constants of the material.

72
Analysis of Mechanical Structures using Finite Element Method

3.2.2. Modeling thermal behavior of materials

In many practical situations, the elements of mechanical systems are working at different
temperatures of those considered as normal, as a consequence of the existence of internal
or external heat sources. As a result, in the structures of these elements occur thermal stress
and strains that overlap with the mechanical ones which appear as consequence of requests
during functioning..
To consider the finite element analysis of thermal effects, it is necessary to know the
thermal characteristics of the structure materials to be analyzed. Usually, for
thermomechanical analyses of structures it is requisite that for the materials from which
they are made, to know the specific heat and the conductivity coefficients and thermal
expansion related to the directions of a tri-orthogonal straight coordinate system.
The specific heat is the characteristic parameter which quantifies the heat storage capacity
of the material and it is measured in J / kg K.
The coefficient of thermal conductivity, measured in W/mK is a material constant variable
with temperature. This variation, generally, is quasi-linear and can be described with the
relationship [40]
  0 1  bT  Eq. 14

where λ0 is the coefficient of thermal conductivity at the temperature of 0° C and b - a


experimentally determined constant.
The linear thermal expansion coefficient , expressed in K-1, quantifies the distortion
ability of the material under the action of the thermal fields, which also varies with
temperature.

73
Analysis of Mechanical Structures using Finite Element Method

3.3. Meshing (discretization) – types of finite


elements 3D, 2D, 1D

3.3.1. Discretization [17]

Mathematical modeling is a simplifying step. But models of physical systems are not
necessarily simple to solve. They often involve coupled partial differential equations in
space and time subject to boundary and/or interface conditions. Such models have an
infinite number of degrees of freedom.

3.3.1.1. Analytical or Numerical

At this point one faces the choice of going for analytical or numerical solutions. Analytical
solutions, also called “closed form solutions,” are more intellectually satisfying,
particularly if they apply to a wide class of problems, so that particular instances may be
obtained by substituting the values of free parameters. Unfortunately they tend to be
restricted to regular geometries and simple boundary conditions. Moreover some closed-
form solutions, expressed for example as inverses of integral transforms, may have to be
anyway numerically evaluated to be useful.
Most problems faced by the engineer either do not yield to analytical treatment or doing
so would require a disproportionate amount of effort. The practical way out is numerical
simulation. Here is where finite element methods enter the scene.
To make numerical simulations practical it is necessary to reduce the number of degrees
of freedom to a finite number. The reduction is called discretization. The product of the
discretization process is the discrete model. For complex engineering systems this model
is the product of a multilevel decomposition.
Discretization can proceed in space dimensions as well as in the time dimension. Because
the present book deals with static problems, we need not consider the time dimension and
are free to focus on spatial discretization.

3.3.1.2. Error Sources and Approximation

Fig. 16 conveys graphically that each simulation step introduces a source of error. In
engineering practice modeling errors are by far the most important. But they are difficult
and expensive to evaluate, because model validation requires access to and comparison

74
Analysis of Mechanical Structures using Finite Element Method

with experimental results. These may be either scarce, or unavailable in the case of a new
product in the design stage.
Next in order of importance is the discretization error. Even if solution errors are ignored
— and usually they can—the computed solution of the discrete model is in general only
an approximation in some sense to the exact solution of the mathematical model. A
quantitative measurement of this discrepancy is called the discretization error. The
characterization and study of this error is addressed by a branch of numerical mathematics
called approximation theory.
Intuitively one might suspect that the accuracy of the discrete model solution would
improve as the number of degrees of freedom is increased, and that the discretization error
goes to zero as that number goes to infinity. This loosely worded statement describes the
convergence requirement of discrete approximations. One of the key goals of
approximation theory is to make the statement as precise as it can be expected from a
branch of mathematics

3.3.1.3. Other Discretization Methods

The most popular discretization techniques in structural mechanics are finite element
methods and boundary element methods. The finite element method (FEM) is by far the
most widely used.
The boundary element method (BEM) has gained in popularity for special types of
problems, particularly those involving infinite domains, but remains a distant second, and
seems to have reached its natural limits.
In non-structural application areas such as fluid mechanics and electromagnetics, the finite
element method is gradually making up ground but faces stiff competition from both the
classical and energybased finite difference methods.
Finite difference and finite volume methods are particularly well entrenched in
computational fluid dynamics spanning moderate to high Reynolds numbers.

3.3.2. The Finite Elements

The finite element method (FEM) is the dominant discretization technique in structural
mechanics. The FEM can be interpreted from either a physical or mathematical viewpoint.
The basic concept in the physical FEM is the subdivision of the mathematical model into
disjoint (non-overlapping) components of simple geometry called finite elements or
elements for short. The response of each element is expressed in terms of a finite number
75
Analysis of Mechanical Structures using Finite Element Method

of degrees of freedom characterized as the value of an unknown function, or functions, at


a set of nodal points. The response of the mathematical model is then considered to be
approximated by that of the discrete model obtained by connecting or assembling the
collection of all elements.
The disconnection-assembly concept occurs naturally when examining many artificial and
natural systems. For example, it is easy to visualize an engine, bridge, building, airplane,
or skeleton as fabricated from simpler components.
Unlike finite difference models, finite elements do not overlap in space. In the
mathematical interpretation of the FEM, this property goes by the name disjoint support
or local support.

3.3.3. The meshing procedure

Just like members in the truss example, one can take finite elements of any kind one at a
time. Their local properties can be developed by considering them in isolation, as
individual entities. This is the key to the modular programming of element libraries.
In the Direct Stiffness Method, elements are isolated by the disconnection and localization
steps. The procedure involves the separation of elements from their neighbors by
disconnecting the nodes, followed by referral of the element to a convenient local
coordinate system. After that we can consider generic elements: a bar element, a beam
element, and so on. From the standpoint of the computer implementation, it means that
you can write one subroutine or module that constructs, by suitable parametrization, all
elements of one type, instead of writing a new one for each element instance.
Following is a summary of the data associated with an individual finite element. This data
is used in finite element programs to carry out element level calculations.
The numerical methods (including FEM) of physical phenomena analysis occurring in
continuous geometric domains involve replacing them with idealized domains
(approximate) assemblies of smaller domains, in the case of FEM called finite elements.
The borders of finite elements consist of points (nodes), straight lines or curves (nodal
lines) and/ or flat or random planes (nodal surfaces).
The choosing operation of the number of nodes and the type of line or surface while
respecting continuities on a nodal level, with the scope of finite element modeling
geometric domains, is called meshing. Fig. 75 shows, for example, the two-dimensional,
D, meshed in triangular finite elements with straight side lines (Fig. 75, a) and curved
lines sides (Fig. 75,b). In the first case, that of finite elements with straight side lines, the
meshing error can be reduced by increasing the number of nodes and so, implicitly, the

76
Analysis of Mechanical Structures using Finite Element Method

number of finite elements. In the second case, the meshing error can be decreased also
thanks to the curved borders of the finite elements.

Fig. 75 Meshing elements

Continuous development of advanced software that are based on FEM was achieved by
diversifying the types of finite elements in association with various physical phenomena
analyzed. Although FEM is a general method for solving differential or integro-
differential equations of governing different physico-technical phenomena with various
initial and boundary conditions, from reasons of productivity and economy, finite
elements with an increased degree of generality have not been developed but were defined
and modeled finite elements customized for different domains (1D, 2D, 3D), low levels
of approximation (linear and parabolic algebraic functions), materials (linear, nonlinear)
and types of problems (static, dynamic, and so on).
Generally, in terms of types of geometric domains to model, finite elements can be one-
dimensional for the linear type of geometric domain, two-dimensional, for surfaces; three-
dimensional, for volumes. These finite elements in terms of approximation function may
be linear, straight edge lines, or non-linear, with curved edges. Typically, nonlinear finite
elements by the type of polynomial approximation used, there may be quadratic parabolics
(of second order), with one intermediate node, or cubic (third order) with two intermediate
nodes.

77
Analysis of Mechanical Structures using Finite Element Method

3.3.4. Types of Finite Elements [17]

3.3.4.1. Element Dimensionality

Elements can have intrinsic dimensionality of one, two or three space dimensions. There
are also special elements with zero dimensionality, such as lumped springs or point
masses. The intrinsic dimensionality can be expanded as necessary by use of kinematic
transformations. For example a 1D element such as a bar, spar or beam may be used to
build a model in 2D or 3D space.

3.3.4.2. Element Nodes

Each element possesses a set of distinguishing points called nodal points or nodes for
short. Nodes serve a dual purpose: definition of element geometry, and home for degrees
of freedom. When a distinction is necessary we call the former geometric nodes and the
latter connection nodes. For most elements studied here, geometric and connector nodes
coalesce.
Nodes are usually located at the corners or end points of elements, as illustrated in (Table
5). In the so-called refined or higher-order elements nodes are also placed on sides or
faces, as well as possibly the interior of the element.

Fig. 76 Mixed finite elements

From the point of view of geometrical or physics features required by analysis questions,
the finite element modeled and implemented in a libraries of specific software
applications, are different.

78
Analysis of Mechanical Structures using Finite Element Method

In some elements geometric and connection nodes may be at different locations. Some
elements have purely geometric nodes, also called orientation nodes to complete the
definition of certain geometric attributes.
Table 5 Main finite elements

Type Geometric Approximation polynomial type


shape
Linear Quadratic Cubic

1-D Element Line


–line (Spring,
truss, beam,
pipe, etc.)

2-D Element Triangular


- plane
(Membrane,
plate, shell,
etc.)
Quadrilateral

3-D Element Tetrahedron


– solid (3-D
fields -
temperature,
displacement,
stress, flow Pentahedron
velocity)

Hexahedron

In order to achieve meshing with various degrees of approximation of a geometric area,


some of the FEM based software applications used mixed finite element in terms of the
approximation (nodal lines straight and curves lines). Thus, Fig. 76 shows the plan area

79
Analysis of Mechanical Structures using Finite Element Method

of the mesh of finite elements and non-linear; to achieve continuity of the structure of
finite elements between the nodal areas discretized at the two types of finite elements is
interposed an intermediate mixed discretized by finite elements having common sides of
the same numbers of nodes.

3.3.4.3. Element Geometry

The geometry of the element is defined by the placement of the geometric nodal points.
Most elements used in practice have fairly simple geometries. In one-dimension, elements
are usually straight lines or curved segments. In two dimensions they are of triangular or
quadrilateral shape. In three dimensions the most common shapes are tetrahedra,
pentahedra (also called wedges or prisms), and hexahedra (also called cuboids or
“bricks”).

3.3.4.4. Element Degrees of Freedom

The element degrees of freedom (DOF) specify the state of the element. They also function
as “handles” through which adjacent elements are connected. DOFs are defined as the
values (and possibly derivatives) of a primary field variable at connector node points. Here
we simply note that the key factor is the way in which the primary variable appears in the
mathematical model. For mechanical elements, the primary variable is the displacement
field and the DOF for many (but not all) elements are the displacement components at the
nodes.

3.3.4.5. Nodal Forces

There is always a set of nodal forces in a one-to-one correspondence with degrees of


freedom. In mechanical elements the correspondence is established through energy
arguments.

3.3.4.6. Element Constitutive Properties

For a mechanical element these are relations that specify the material behavior. For
example, in a linear elastic bar element it is sufficient to specify the elastic modulus E and
the thermal coefficient of expansion α.

3.3.4.7. Element Fabrication Properties

For mechanical elements these are fabrication properties which have been integrated out
from the element dimensionality. Examples are cross sectional properties of MoM
80
Analysis of Mechanical Structures using Finite Element Method

elements such as bars, beams and shafts, as well as the thickness of a plate or shell element.
For computer implementation the foregoing data sets are organized into data structures.
These are used by element generation modules to compute element stiffness relations in
the local system.
Table 6 Finite element types
Geometry Specifics Input parameters Output
type characteristics parameters
1D Straight or curved bar Sectional parameter Nodal
subjected to traction- (area) displacements
compression The set of material Elemental
parameters stresses
Straight or curved bar Sectional parameter Efforts
subjected to traction- (area and inertia Reaction forces
compression, shearing, moments)
torsion and bending The set of material
parameters
2D Flat plate in a flat state of Thickness Nodal
stresses The set of material displacements
parameters Elemental
Flat plate in a flat state of Thickness stresses in
strains different plans
The set of material
parameters Nodal forces
Reaction forces
Spatial plate Thickness
The set of material
parameters
Shell Thickness
The set of material
parameters
Layered plate Thickness
The set of material
parameters
Spatial structure in The set of material Nodal
asymmetric stress state parameters displacements
3D Spatial structure The set of material Elemental
parameters stresses
Nodal forces
Reaction forces
In (Table 5) illustrates the main finite elements contained by commercial advanced
analysis programs with FEM, indicating the main input and output parameters. As a result,
81
Analysis of Mechanical Structures using Finite Element Method

specific finite element are found, imposed by the type of geometric domain (bars, plates,
membranes and massive spatial structures), the resistance of structures (bars resistant to
traction-compression or traction-compression, shearing, torsion and bending), the stress
and deformation states (surfaces in a plane state of stresses deformations or axisymmetric
spatial structures), the type of input/ output parameters and the internal configuration of
the material (massive, laminated structures) [40].
In addition to these types of finite elements that have a structural nature, most advanced
programs also have specialized unstructural elements (Rigid, Spring, etc.) for modeling
mechanical bond problems and increased degree of idealization modeling (mass
concentrated rigid spring, spring-damper etc.), and for total or part modeling of structures,
when accuracy and cost are appropriate. Table 6 shows the main unstructural element
groups, contained in the groups mentioned above, which are frequently encountered in
commercial software for the analysis of the mechanical fields.

Table 7 Input parameters of finite elements


Type of finite Output
Characteristics Input parameters
element parameters
Nodal
Masses, inertial moments
Inertial, mass Mononodal displacements
and inertial matrixes
Nodal forces
Degrees of liberty with Nodal
Rigid Multinodal canceled relative displacements
displacements Nodal forces
Arc Binodal Linear and torsional rigidity Nodal
displacements
Rigidity and dampering Nodal forces
Damper-arc Binodal
constant
Internal forces

Inertial or mass elements shapes the structure of an element of a mechanical system or


parts of it by reducing it to a material point to which it is assigned mass properties and/ or
inertia equivalent properties.
Simplified modeling of mechanical structure areas with increased rigidity can be made
with rigid finite elements that introduce one or more nodes with invariable relative
positions as degrees of freedom defined above, to another node. In contrast to this
possibility of modeling of mechanical system elements or parts of them can be replaced
with elements that reduce their structure down to two material points (nodes) connected
82
Analysis of Mechanical Structures using Finite Element Method

by an arc (arc finite element) or through an damper-arc system (finite element of a damper-
arc type) which are inserted as input data accordin to the corresponding characteristics.

3.3.5. Classification of Mechanical Elements [17]

The following classification of finite elements in structural mechanics is loosely based on


the “closeness” of the element with respect to the original physical structure. It is given
here because it clarifies points that recur in subsequent sections, as well as providing
insight into advanced modeling techniques such as hierarchical breakdown and global-
local analysis.

3.3.5.1. Primitive Structural Elements

These resemble fabricated structural components. They are often drawn as such; see (Fig.
77). The qualifier primitive distinguishes them from macroelements, which is another
element class described below. Primitive means that they are not decomposable into
simpler elements.

Fig. 77 Examples of primitive structural elements

83