FEM Concepts

Level 1

Finite Element Analysis in Practice

This course is taught with a mixture of

theoretical information and applied finite
element analysis (FEA) using the
software. Concepts are illustrated with
simple, hands-on exercises.

Models are created that illustrate a broad

range of topics including theory, element
types,
analysis
types,
meshing
techniques, results evaluation, and more.

The purpose of this presentation is to expose some of

the fundamental aspects of FEA.

The expression FINITE ELEMENT is first

attributed to Clough.

An actual mechanical component is a continuous

elastic structure.

FEA divides (discretizes) the structure into small but

Finite, Well-defined, Elastic sub-structures.

FEA is a mathematical solution to

engineering problems where a physical
model is divided into discrete
components.
FEA models are defined by nodes and
elements (commonly called a mesh).
Basic engineering equations, such as
Hookes law, are solved at the nodes
and elements.
A matrix equation, including terms from
each element, is solved.

Predicts changes within the element (for

example, deformation and stress).
The results are plotted on the model using
colors to show the lowest and highest values.

Why Use FEA?

Provides a non-destructive means of testing products.

Faster prototyping for what if scenarios.

Design optimization.
Speed up time to market by shortening the design
cycle.

FEA requires engineering judgment. In the best case,

you should know the approximate answer before you
begin.

Proper selection of elements, materials, loads,

constraints, and analysis parameters comes from
experience.

Understand that the computer model never exactly

matches reality (its only an approximation).
The surest route to failure in FEA is to underestimate
the complexity of the technology.

FEA is a numerical procedure to simulate

loading conditions on a design &
determine its response to those loads.
Building blocks of a design in FEM is
known as Elements.
Elements have finite number of
unknowns, hence the name FEA.
As infinite number of unknowns of an
design is analyzed by finite number of
unknowns, smart approximation required.

Engineering problems mathematical

models of physical situations.
Mathematical Models DEs with set
of boundary & initial conditions.
DEs derived by applying
fundamental laws & principles to
system or control volume.
These governing equations represent
the balance of mass, force or energy.

System behavior is influenced by,

1) parameters regarding natural behavior.
Eg: viscosity, thermal conductivity,
modulus of elasticity, MI,
resistance etc
2) parameters that produce disturbance in
the system.
Eg: forces, moments, temperature
differences, PD, pressure diff. etc

Analytical solns gives exact solns

Homogenous part & Particular
part.
Natural behavior reflects in homogenous
part & Disturbance parameters reflects in
particular part.
Numerical Solns- when exact solns are not
possible due to complexities. Numerical
method approximates exact solutions only
at discrete points, called nodes

Preprocessing Phase
1) Create & discretize the solution
domain into finite elements.
2) Assume shape functions to represent
the physical behavior of an element.
3) Develop equations for an element.
4) Construct global stiffness matrix.
5) Apply boundary & initial conditions &
loading parameters.

Solution Phase
Solving a set of linear or nonlinear
algebraic equations to obtain nodal
results.

Post Processing phase

To evaluate and analyze the results.

Build/Mesh a Model
Define FEA Model
Element and Analysis Details
Define Loads and Constraints
Analyze Model (Solve)
Review Results and Create
Presentations

FEA Concepts

Key Terms:
Degrees of freedom(DOF).
Displacement matrix
Stiffness matrix
Force matrix

If the number of degrees of freedom

is finite, the model is called discrete,
and continuous otherwise.

The unknowns in a finite element problem are

referred to as degrees of freedom (DOF).
Degrees of freedom vary by element and analysis
type.

DOF Type

Action

Application

Displacement

Force

Structural

Temperature

Heat Flow Rate

Thermal

Velocity

Fluid Flow Rate

Fluid

Voltage

Electromotive Force (EMF)

Electrostatic

Uy

Rot y
Rot x

Node

Ux

Uz

Rot z

Application
Problem

State (DOF) vector

u represents

Conjugate vector f
represents

Structures and solid

mechanics

Displacement

Mechanical force

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

A simplified view of the physical simulation

process, reproduced to illustrate modeling
terminology
FEM
Idealization
Physical
System

Mathematical
Model

Discrete
Model

CONTINUIFICATION
REALIZATION &
IDENTIFICATION

Solution

Discretization

Solution Error

Discretization+Solution Error

Modeling+Discretization+Solution Error

Discrete
Solution

Idealization passes from the physical

system to a mathematical model.
This is the most important step in
engineering practice, because it
cannot be canned. It must be done
by a human.
Models
Mathematical Models
Implicit vs. Explicit Modeling

The word model has the traditional

meaning of a scaled copy or representation
of an object.

A model is a symbolic device built to

simulate and predict aspects of behavior
of a system.

Mathematical modeling, or idealization, is a

process by which an engineer or scientist passes
from the actual physical system under study, to a
mathematical model of the system
The process is called idealization because the
mathematical model is necessarily an abstraction
of the physical reality

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.
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.
Analytical or Numerical?
Error Sources and Approximation
Other Discretization Methods

Analytical solutions, also called closed

form solutions, are more intellectually
satisfying, particularly if they apply to
a wide class of problems.
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

Next in order of importance is the

discretization
error.
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.

Element
Options

Selecting the type of element will

depend on the following:
Analysis type selected.
How you create your mesh.
Assumptions you can make about

geometric properties.

Line Elements: A line connecting 2 nodes (representing

beams, trusses, springs, actuators, pipes, and so on).

Area (2-D) Elements: YZ-planar elements that are each

triangular or quadrilateral (3 or 4 lines enclosing an
area). These elements can represent thin parts, crosssections of a specified thickness, or radial cross-sections
of axisymmetric parts.

Area (3-D Planar) Elements: Planar or nearly planar

elements in 3-D space. Each must be triangular or
quadrilateral and represents a thin part with a specified
thickness. The plane of the element is at the midplane
of the part it represents.

3-D Solid Elements: Must be enclosed volumes with 4,

5, or 6 triangular and/or quadrilateral faces and with 4,
5, 6, or 8 corner nodes.

1D:

2D:

3D:

A node is a coordinate location in space

where the DOF are defined. The DOF of this
point represent the possible response at this
point due to the loading of the structure.

An element is a mathematical relation that defines

how the DOF of a node relate to the next. These
elements can be lines (beams), areas (2-D or 3-D
plates) or solids (bricks and tetrahedra).

A node has a given set of DOF, which characterizes

the response. For structural analyses, these DOF
include translations and rotations in the three global
directions.
The type of element being used will also characterize
which type of DOF a node will have.
Some analysis types have only one DOF at a node. An
example is the temperature in a heat transfer analysis.
Heat flow rates are calculated from the nodal
temperatures and the material thermal conductivity (no
flow DOF).

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.
Nodes are usually located at the
corners or end points of elements.

 Elements can only transfer loads to one

another via common nodes.

No Communication
Between the Elements

Communication
Between the Elements

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.

The element degrees of freedom (DOF)

specify the state of the element.
DOFs are defined as the values (and
possibly derivatives) of a primary field
variable at connector node points.
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.

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.

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 .

For mechanical elements these are

fabrication properties which have
been integrated out from the element
dimensionality
Examples are cross sectional
properties of MoM elements such as
bars, beams and shafts, as well as
the thickness of a plate or shell
element.

The basic stress and strain

equations:

Basic equations do not require the use of a

computer to solve.
Computer-based analysis is needed when
complexity is added as follows:
Geometric complexity makes the elasticity

equation difficult or impossible to solve.

Variations in material properties exist
throughout the part.
Multiple load cases and complex or combined
loading exists.
Dynamics are of interest.

The DOF components of each element

combine to form a matrix equation:
[K] {d} = {A}
[K] = element stiffness components
{d} = DOF results (unknown)
{A} = action value (e.g., force,

temperature)

To determine the displacement of a simple linear

spring under load, the relevant equation is:
{f} = [K] {d}

Known
where {f} = force vector
[K] = stiffness matrix
{d} = displacement vector

Unknown

For a more complex analysis, more terms are needed. This

is true in a dynamic analysis, which is defined by the
following equation:
{f} = [K] {d} + [c] {v} + [m] {a}
where...{f} = force vector
[K] = stiffness matrix
{d} = displacement vector
[c] = damping matrix
{v} = velocity vector
[m] = mass matrix
{a} = acceleration vector

FEA can be applied to a wide

variety of applications such as:
Dynamics
Nonlinear Materials
Heat Transfer
Fluid Flow
Electrostatics

Analysis
Options

The first decision in the FEA process

is to decide what type of analysis you
need to run.
The analysis type will dictate what
type of results you will obtain.
For example, if you need the
displacement of your part, then you
will need to run a structural analysis.

Linear
Linear static
Linear dynamics (incl. eight
analysis types + DDAM)
Nonlinear
Nonlinear static
MES
Modal w/ nonlinear materials
Riks (post buckling & collapse)
Thermal
Steady-state heat transfer
Transient heat transfer
Fluid Flow
Steady fluid flow
Unsteady fluid flow
Flow through porous media
Open channel flow

Electrostatic
Current and voltage
Field strength and voltage

Mass Transfer
Transient mass transfer

Multiphysics
Steady coupled fluid flow

and thermal
Transient coupled fluid flow
and thermal

Linear static
Small changes in stiffness.
No changes in loading direction.
Material remains in the linear elastic range.
Small deformation and strain.

Linear dynamics
Natural frequency (modal)
Natural frequency (modal) with load stiffening
Response spectrum
Random vibration
Frequency response
Transient stress (direct integration)
Transient stress (modal superposition)
Critical buckling load
Dynamic Design Analysis Method (DDAM)

Nonlinear/MES
Linear and nonlinear material models.
Large deformation and strain.
Failure due to:
Material yielding.
Local and structural buckling.
Permanent deformation residual stress.
Large-scale motion.

Nonlinear/MES
Surface-to-surface contact
Impact
Creep

Steady-state heat transfer

Steady-state conditions
Transient heat transfer
Time-varying conditions

Steady fluid flow

Unsteady fluid flow
Flow through porous media
Open channel flow

Electrostatic current and voltage

Electrostatic field strength and voltage

Transient mass transfer

Mass transfer refers to mass in transit due to
gradients in the concentration of species within a
mixture, and the transfer is due to random molecular
motion. A typical application example is chemical
species migrating through a membrane.

Meshing and
Modeling

For any region (3, 4, 5, 6 or 8-nodes), to be a valid

element, it:
Must consist of either three or four undivided
line
segments
(forming
triangular
or
quadrilateral faces).
If an element edge
consists of multiple line segments, the region is
invalid.
Must not have curved or arched sides.

The
exception to this is second order elements,
where midside nodes are located on the original
surface for curved, CAD-based geometry.

Proper Modeling Techniques

(continued)

Certain shapes can create elements which are not

recommended for FEA analysis. The following regions
will be eliminated, requiring adjustment of the
geometry or meshing parameters:
o Regions with any collinear or concave sides.
o Regions with excessive curvature relative to the
element size (causing non-flat, excessively warped
element faces).

Proper Modeling Techniques

(continued)

Valid and Invalid Regions

Meshing can be completed either by using automatic

mesh engines, semiautomatic structured meshing tools,
or by creating a mesh completely by hand.
Automatic mesh generation is usually completed on
CAD solid models or within outline sketches (for 2D
models).
Semiautomatic meshing tools and hand meshing
techniques are generally used for simple models where
a structured mesh is desirable.

There are two types of hand meshing

building from scratch and building from a
sketch or wireframe.
Building from scratch:
Draw the elements by hand one at a time,
line-by-line, to create a structured mesh.
Building from a sketch or wireframe:
Build a 2-D sketch or 3-D wireframe of the
model using construction objects and apply
a watertight surface mesh using the 2-D
mesh engine and/or structured meshing
tools. Use an automatic solid mesh engine
to
generate
internal
elements
for
3-D parts, or extrude planar meshes into
solids.

Selected lines or 1 object meshes may be extruded into

surfaces. Planar meshes (hand drawn or from structured
meshing tools) may be extruded into solids. Use one of
the following three commands with the Copy and
Join options enabled to do so:
Move or Copy
Rotate or Copy
Scale or Copy
(These three operations may be combined to
simultaneously move, rotate, and scale the selected
lines or mesh into a surface or solid structure.)

Loads and
Constraints

You will have to decide what type of loads and

constraints will properly define the engineering criteria
for the model.
In FEA, there are different types of loads and
constraints for each analysis type.
Applying the proper loads and constraints is one of the
most important factors in getting the correct answer.
Always double-check your model.

Introduction to Loads & Constraints

(continued)

There are multiple ways to apply different

loads and constraints to a model:

Nodal Loads and Constraints

Edge Loads and Constraints
Surface Loads and Constraints
Element Loads (this includes part-based or global

loads). An example is gravity or acceleration.

Other examples are axial preloads, end releases,
and neutral axis offsets applied to beam elements.

Displacements
Forces
Lumped Masses
Moments
Temperatures (thermal stress)
Voltages (piezoelectric materials)

Boundary Conditions: Prevent specified DOF from

undergoing translation or rotation in a specified
direction.

Boundary Elements: Act like a spring with a

specified stiffness along a specified direction.

Using Boundary Conditions to Model Symmetry

Along the line or plane of symmetry, boundary
conditions must be applied to represent the
symmetrical part, as follows:
Out-of-plane displacement = 0
Two in-plane rotations = 0

Line of
Symmetry

Using
Boundary
Conditions
to
Model
Antisymmetry (where the part is symmetrical but
the loads act in opposite directions)
Along the line or plane of antisymmetry,
boundary conditions must be applied to
represent the part, as follows:
Out-of-plane rotation = 0
Two in-plane displacements = 0

Line of
Antisymmetry

Proper boundary conditions are necessary for an

accurate analysis.
The global stiffness of the system must be modeled
correctly for any local behavior to be captured
correctly.

A key strength of the FEM is the ease

and elegance with which it handles
arbitrary boundary and interface
conditions
Essential and Natural B.C.
Boundary Conditions in Structural
Problems

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.
1. 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.
2. Otherwise it is natural.

A structural 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
anti-symmetry restraints at certain points, lines
or planes of structural symmetry.
Ignorable freedoms.-To suppress displacements
that are irrelevant to the problem
Connection constraints- To provide connectivity
to adjoining structures or substructures

Uniform or Hydrostatic Pressure and Traction

Applied to the faces of plate, composite, brick,
shell, and nonlinear membrane elements.
Applied to the edges of 2-D elements (though
selected as surfaces because the edges represent
surfaces).
Surface Force
Can specify magnitude and direction of a force that
will be evenly distributed over a given surface.

Variable Pressure or Traction

Define a function of the position (i.e., the X-, Yand/or Z-coordinates along the surface) that controls
the varying magnitude of the load acting on the
surface.

Gravity/Acceleration
Can
specify
gravitational
or
general
acceleration value and direction. You must
have a mass density defined for each part.
Centrifugal Loads
Specify center of rotation, angular velocity and
angular acceleration values.
Distributed Loads
Specify the magnitude and direction at each end
of beam elements.

Truss Elements

Truss elements are two-node members, which allow

arbitrary orientation in the X, Y, Z system.
The truss transmits axial force only and, in general, is a
three DOF element (i.e., three global translation
components at each end of the member). Trusses are
used to model structures such as towers, bridges, and
building framework (skeletons).

Guidelines for using truss elements:

The length of the element is much greater than the
width or depth (approximately 8-10 times or
more).
The connections to the rest of the model are as
ball joints that do not transfer moments.
The external forces are applied only at joints.

Objective: Construct and analyze a frame of truss

elements loaded with 2 nodal forces.

Geometry: Cross-sectional area = 1 in2.

Material: Aluminum (6061-T6).

Loads: Nodal forces as shown
in the image to the right.
Constraints:
Fully fixed at Point A.
Ty and Tz constrained at Point G.
The rest of the model will have Tz

constrained.

Beam
Elements

Beam elements are slender structural

members that offer resistance to axial
forces, bending, and torsion under
applied loads.
Beams are found in building frames,
transmission towers, and bridges.
A beam differs from a truss in that a
beam resists moments (twisting and
bending) at the connections.

Beam Elements

(continued)

Guidelines for using beam elements:

The length of the element is much greater than the width or

depth.
The cross-section of the element is constant. (Note: A variable

cross-section wizard is available for approximating tapered

beam spans via a series of multiple elements, each having a
constant cross-section that differs from the adjacent beam
elements.)
The element is able to transfer moments, though end releases

are available to simulate pinned or slotted connections.

The element is able to handle a load distributed along its

length.

Objective: Determine the

maximum deflection of the
beam due to its own weight.
Geometry: W10 x 100
cross-section.
Material: Steel (AISI 4130).
Loads: Gravity in the -Y
direction.
Constraints:
Far end has constraints against all DOF except Rz.
Near end has constraints against all DOF except Tx and Rz.

2-D
Elements

Two-dimensional elements, with three or

four corner nodes, that are formulated in
the Y-Z plane. They are used to model
any part or assembly that can be
accurately represented using a 2-D
outline or cross-section and for which 2D loading and constraint assumptions are
valid (examples to follow).

2-D Geometry Types

Axisymmetric: For parts that are represented by
cross-sections revolved about an axis (such as a
bearing, shaft seal, or hydraulic cylinder).
Plane strain: A cross-section of an object of semiinfinite length, having no deflection normal to the
cross-section (such as a large dam).
Plane stress: Objects of finite thickness with no
stress normal to the cross-section (such as a plate
under in-plane loads).

Objective: Determine the

hoop stress at the inner
radius of the cylinder from
the applied pressure load.
Material: Steel (AISI
4130).
Loads: Uniform internal
pressure of 10,000 psi.
Constraints: The bottom
surface will have Tz
constrainted.

Plate/Shell
Elements

Plate/shell elements have three or four corner

nodes are planar (or nearly planar), and are
formulated in 3-D space. These elements are
used to model and analyze thin objects such as
pressure vessels and automotive body parts.
A numerical thickness is assigned to the
elements but the model geometry itself has no
thickness.
Stresses are assumed to vary linearly through
the thickness.

Guidelines for using plate/shell elements:

The thickness is small in relation to the length and
width (about 1/10).
Good for small displacements and rotations.
Elements remain planar (i.e., no warpage).
No rotation about the direction normal to the
element.

Planar (or nearly planar) elements, with

three or four corner nodes, formulated in
three-dimensional space.

Used to model "fabric-like" objects such

as tents or cots, or structures such as the
roof of a sports stadium.

Represent thin solids of a specified

thickness that exhibit no stress normal to
the thickness.

There are two types of composite

elementsthin and thick. Thick
composites have a core layer that is
typically much thicker than the other
laminae.

Each element can have multiple

laminae, each with different material
properties and fiber orientations.

Multiple failure criteria are available.

As with plates and membranes, the

thickness, in this case for each
lamina, is defined numerically (not
graphically).

Objective: Determine the maximum stress in the

plate from the applied pressure load. Use the
4 Point structured meshing tool to make the plate.

Geometry: The plate is 10 x 5 x 0.25.

Constraints:

Material: Steel (AISI 4130).

Loads: Uniform pressure
of 50 psi.

The two long edges will have

Ty and Tz constraints.

One short edge will have Tx

and Tz constraints.

Brick Elements

Brick elements are four-, five-,

six- or eight-node elements
formulated in three-dimensional
space.
Brick elements should be used
for solids with irregular shapes
and when the stress through the
thickness of a part is important.

Objective: Using brick

elements, determine the
maximum bending stress
in the beam from the
applied load.

Material: Steel (4130).

Constraints:

Loads: 10,000 pounds distributed

across the free end.

The center horizontal row of nodes (running in the Y-direction)

at the fixed end will be fully constrained.

The remaining nodes at the fixed end will have only the Tx
constraint applied.