Finite Element Method (MSTR 515)

Module 1

Introduction
 Contents

Introduction Basic Theory of the FEM

Application of Finite Element Advantages and Disadvantages
Method /Analysis
Some Important Matrices and Equations
Summary of Finite Element History
Stiffness or displacement method
Steps in FEM
Theory of Elasticity
Types of Finite Elements
Boundary Conditions
Mesh
Strain-Displacement Relations
Beam Element
Temperature Effects (Initial Stress and Strain)
Plate and shell elements
Saint Venant’s Principle
Some Examples of FEM in Research

Department of Civil Engineering 2

Introduction

• The finite element method, Some Important Matrices and Equation, Elasticity
Theory,

• Strain-Displacement Relations, Stress-Strain Relations, Temperature effects.

• Initial Stress and Strain, Strain Transformation,

• Equilibrium, Compatibility, Constitutive Law, boundary Conditions

• Plane Stress, Plane Strain, Saint Venant’s Principle

Department of Civil Engineering 3

Introduction( Contd…)

• FEM is a generally applicable method for getting numerical solutions

• Problems of stress analysis, heat transfer, fluid flow, electric fields and other have
been solved by finite elements

• This course emphasizes stress analysis to structural mechanics

• FEM is a numerical procedure for solving a continuum mechanics problem with

an accuracy acceptable to engineers

Department of Civil Engineering 4

 Fig a): a plane structure of arbitrary shape ;

Fig b): a possible finite element model of the

structure

Fig c): a plane rectangular element, showing

nodel forces pi and qi. The dashed line
shows the deformation mode associated with
x- direction displacement of node 3

Imagine thar stresses and displacement of structure in Fig(a) must be found

Numerical answers are not found in any book.

Classical methods describes the problem with partial differential equations but yield no answers because the
geometry and loading are too complicated.

In practice, most problems are too complicated for a closed-form mathematical solutions.
A numerical solution is required, and the most versatile method that provides it is the Finite Element
Method
Department of Civil Engineering 5
Introduction( Contd…)

Fig. b) shows a finite element model. The quadrilateral and triangular

regions are finite element

Black dots are nodes where elements are connected to one another

A mesh is an arrangement of nodes and elements

This particular mesh shows triangular and quadrilateral elements with side
nodes and some with only corner nodes

Department of Civil Engineering 6

Application of Finite Element Method/Analysis

• Distribution of stresses and strain in structural elements

• Deflection analysis of beam other structures

• Dynamic Responses of Structure

• Vibration analysis

Department of Civil Engineering 7

 Summary of Finite Element History

• In 1943, Courant suggested piecewise polynomial interpolation over triangular

subregion as a way to get approximate numerical solutions
• He recognized this approach as a Rayligh-Ritz solution of variational problem,
and it is known as FEM today
• By 1953, engineers were writing stiffness equations in matrix notation and solving
equations with digital computers
• In 1956 M. J. Turner, R. W. Clough, H. C. Martin, and L. J. Topp published a
paper which established a broader definition of numerical analysis. The paper
centered on the "stiffness and deflection of complex structures
• The name “Finite Element” was coined in 1960.
• By 1963, the method was recognized as rigorously sound, and it became a
respectable area of study for academicians.

Department of Civil Engineering 8

Steps in FEM
• First step: differential equation of the problem is converted into an integral form.
These are two techniques to achieve this :
(i) Variational Technique and (ii) Weighted Residual Technique.
• Second step: the domain of the problem is divided into a number of parts, called
as elements. Division of the domain into elements is called a mesh.
• Third step: over a typical element, a suitable approximation is chosen for the
primary variable of the problem using interpolation functions (also called as shape
functions) and the unknown values of the primary variable at some pre-selected
points of the element, called as the nodes. Additional nodes are placed either on
the boundaries or in the interior. The values of the primary variable at the nodes
are called as the degrees of freedom.

Department of Civil Engineering 9

Steps in FEM (Contd…..)
Fourth step:
• The approximation for the primary variable is substituted into the integral form. If
the integral form is of variational type, it is minimized to get the algebraic
equations for the unknown nodal values of the primary variable. If the integral
form is of the weighted residual type, it is set to zero to obtain the algebraic
equations. In each case, the algebraic equations are obtained element wise first
(called as the element equations) and then they are assembled over all the elements
to obtain the algebraic equations for the whole domain (called as the global
equations).
Last step:
• The post-processing of the solution is done. That is, first the secondary variables
of the problem are calculated from the solution. Then, the nodal values of the
primary and secondary variables are used to construct their graphical variation
over the domain either in the form of graphs (for 1-D problems) or 2- D/3-D
contours as the case may be.
Department of Civil Engineering 10
Types of Finite Elements

Department of Civil Engineering 11

Mesh

• Mesh is your way of communicating

geometry to the solver,
• the accuracy of the solution is primarily
dependent on the quality of the mesh.
• The better the mesh looks, the more
accurate the solution is.
• A good-looking mesh should have well-
shaped elements, and the transition
between densities should be smooth and
gradual without skinny, distorted
elements. Mesh

Department of Civil Engineering 12

Beam Element

• Two types of beam elements

• Able to transmits moments (Beam Element)
• Not able to Transmit moments (bar or truss element)

Department of Civil Engineering 13

 Plate and shell elements

Plate and shell elements are used interchangeably and refer to surface-like elements used to represent thin-
walled structures.

A quadrilateral mesh is usually more accurate than a mesh of similar density based on triangles. Triangles are
acceptable in regions of gradual transitions.

Department of Civil Engineering 14

 Solid Elements

Tetrahedral (tet) mesh is the only

generally accepted means to fill a
volume, used as auto-mesh by many
FEA codes. 10-node Quadratic

Department of Civil Engineering 15

Basic Theory of the FEM
• In FEM, the actual structure is replaced by conceptual model which we can call as
finite element models

• What we analyze is in fact not the actual structure but the approximate finite
element models

• Requires lot of experience, judgement, intuitive thinking on the part of structural

analysis

• To master in FEM, analyze the real life problem and validate it

Department of Civil Engineering 16

Advantages and Disadvantages

Main advantage of FE analysis is that physical problems which were so far

intractable and complex for any closed bound solutions can be analyzed by this
method
• Can be efficiently applied to carter irregular geometry
• Can take care of any type of boundary
• Material anisotropy and inhomogeneity can be treated without much difficulty
• Any type of loading can be handled

Department of Civil Engineering 17

Advantages and Disadvantages (Contd..)

Disadvantages:
• There are many types of problems where some other method of analysis may
prove efficient than the FEM
• Another disadvantage of this method is cost involved in the solution of the
problem
• For vibration and stability problems in many cases the cost of analysis by FEM
may be prohibitive.
• Stress values may vary by 25-30% from fine mesh analysis to average mesh
analysis
• There are other trouble spots such as “aspect ratio” (ratio of longer to smaller
dimension of the elements) which may effect the final results. However, these are
easier to guard against than the problem of the proper mesh size and distribution

Department of Civil Engineering 18

Some Important Matrices and Equations
Degree of Freedom (DoF) as the displacement or rotation of a node
Fig. a): the DoF
Then, for an element with n DoF., we can write the equations {d} of a standard
beam element

Fig. (b) the

deformation mode
{d} = {1 0 0 0}
and required force
ki1
where di is the ith DoF and ri is the corresponding force or
moment applied to the element
kij are stiffness coefficient, if gathered into matrix
Fig.(c):the
deformation mode
where [k]: element stiffness matrix {d} = {0 1 0 0}
{d} : element nodal displacement vector ,and and required force
{r}: vector of element nodal loads ki2

Department of Civil Engineering 19

Stiffness or displacement method

• Displacement are the primary unknowns to computed .

• Stress is a secondary variable, computed from displacements.
• Most popular from the finite element method in structural mechanics
To describe the meaning of [k] ,

If all DoF are zero but the jth and if dj=1, then we see that {r} = {kij}, the jth
column of [k]

Department of Civil Engineering 20

 Stiffness or displacement method (Contd..)
• The interpretation given to columns of [k] also applies on the structural level

Fig. a) a structure that has three active DoF (u1,u2,u3). Its “finite element” are three linear springs of stiffness k1,
k2 and k3 , b) Nodal DoF and forces of typical element i

In the above figure, by assigning a unit displacement to every node in turn, each time writing
the necessary forces as a column in a 4-by-4 matrix, we find

Each dashed line in stiffness matrix, considering ui=1 and then

ui+1= 1 in above figure
The structure matrix can be built by adding element matrices
in an overlapping, we will see it in Module 2.

Node 4 in figure is fixed against motion, u4=0

Department of Civil Engineering 21

Stiffness or displacement method (Contd..)
• Thus the equations that relative to DoF are
Called as Structural or Global
Matrix instead of element
matrix

We can solve {D}, because

{d} for each element is
contained in {D}, deformation
of all elements are known.

From deformation we can compute stresses and the solution is complete

Department of Civil Engineering 22

 Theory of Elasticity

Figure shows a differential (not finite!)

element,
Fx and Fy have dimensions of force per unit
volume and can arise from gravity,
acceleration, a magnetic field and so on
Unit volume: dV= t dx dy
Fx and Fy produce total forces Fx dV and Fy
dV

Fig. Stresses and body forces per unit volume that act on the plane
differential element of constant thickness t.

In general stresses are functions of the coordinates

σxx is the rate of change of σx wrt x and σxx dx is the change of σx over a distance dx
Equilibrium of forces in the x direction is
−𝜎𝑥𝑡𝑑𝑦 − 𝜏𝑥𝑦𝑡𝑑𝑥 + 𝜎𝑥 + 𝜎𝑥, 𝑥𝑑𝑥 𝑡𝑑𝑦 + 𝜏𝑥𝑦 + 𝜏𝑥𝑦, 𝑦𝑑𝑦 𝑡𝑑𝑥 + 𝐹𝑥𝑡𝑑𝑥𝑑𝑦 = 0

Department of Civil Engineering 23

Theory of Elasticity ( Contd..)
• Similarly there is a corresponding y-direction equilibrium equation.
• In three dimensions, the body force vector is
{F} ={Fx Fy Fz}

And the differential equations of equilibrium are

𝜎𝑥 , 𝑥 + 𝜏𝑥𝑦, 𝑦 + 𝜏𝑧𝑥 , 𝑧 + 𝐹𝑥 = 0
𝜏𝑥𝑦, 𝑥 + 𝜎𝑦, 𝑦 + 𝜏𝑦𝑧, 𝑧 + 𝐹𝑦 = 0
𝜏𝑧𝑥 , 𝑥 + 𝜏𝑦𝑧, 𝑦 + 𝜎𝑧, 𝑧 + 𝐹𝑧 = 0

When an elastic body is deformed, no cracks appear in stretching, no kinks appear in bending, and no
part overlaps another.
Stated more elegantly, this is compatibility condition: the displacement field is continuous and
singled-valued

Department of Civil Engineering 24

Boundary Conditions
Apply to stress and displacement, and are part of the problem
definition
In the figure shown,
• the left edge does not move (u=v=0 on x=0);
where the pressure acts, 𝜎𝑦 =-p and 𝜏𝑥𝑦 =0
• on the right edge, 𝜎𝑥 = 𝜏𝑥𝑦 =0; and so on
Where displacements are prescribed, stresses are unknown and
assume values implicitly dictated by the solution.
Similarly, where stresses are prescribed, displacements are unknown

Department of Civil Engineering 25

Boundary Conditions ( Contd..)
• If a stress or displacement field satisfied equilibrium, compatibility, and boundary
conditions, then a solution has been found
• If the load versus response behavior is linear, then the solution is unique
• If the elements are based on displacement fields, so that nodal displacements are
the primary unknowns, then the compatibility conditions is satisfied within
elements
• Suitably chosen fields also provide compatibility between elements and satisfy
displacement boundary conditions
• Equilibrium equations and boundary conditions on stress are satisfied
approximately
• The approximation improves as more DoF are used, barring computational
difficulties, the exact solution is achieved in the limit of an infinitely redefined
mesh

Department of Civil Engineering 26

 Strain-Displacement Relations
• Relation between displacement and strain is essential in element formulation

• In Figure, a general strain field converts configuration 012 to configuration

0’1’2’

• u and v are functions of the coordinate

• u,x dx are small in comparison with u and v

• Normal strain is the ratio of change in length to the original length

A similar analysis yields the y-direction normal strain

as

Department of Civil Engineering 27

Strain-Displacement Relations (Contd..)

Shear Strain is defined as the amount of change in a right angle. Because displacement increments are small,
𝛽1 ~ tan 𝛽1 and 𝛽2 ~ tan 𝛽2 , so the engineering shear strain is

Strain-displacement relations can be stated in

matrix-operator form

Department of Civil Engineering 28

 Stress -Strain Relations
The stress-strain relation is abbreviated as 𝛜 = 𝐂 {𝛔} or as 𝛔 = 𝐄 𝛜 (in one dimension)

(Express Hooke’s Law. For real solids the “law” is an approximation limited to small strain)
-
Where [C] is a matrix of material , [E] is a matrix of material stiffness and [E]= 𝐂

The stress vector {𝛔} and the strain vector 𝛜 are, respectively

Nine independent coefficients in 𝐄 , 𝐄 is symmetric,

if the material is orthotropic. If axes x,y, and z coincides
with principal directions of the material, we can write

A material is orthotropic if its mechanical or thermal properties

are unique and independent in three mutually perpendicular
directions. Examples of orthotropic materials are wood, many
crystals, and rolled metals.

Department of Civil Engineering 29

 Stress -Strain Relations (Contd..)
In two dimensions, the problems are modeled as plane stress and plane strain Consider a body that lies in the xy plane.
Then by definition, 𝜏𝑦𝑧 = 𝜏𝑧𝑥 = 𝛾𝑦𝑧 = 𝛾𝑧𝑥 = 0. If the problem is plane stress, then 𝜎𝑧 = 0 and 𝜖𝑧 need not enter the
solution process Thus for isotropic material [E]= 𝐂 - , then
• In continuum mechanics, a material is said to be under plane
stress if the stress vector is zero across a particular plane

• Thin plates subject to forces in their plane only

• No force in the z-direction and no variation of any forces in

z-direction 𝜎𝑧 = 𝜏𝑦𝑧 = 𝜏𝑧𝑥 = 0.

The constitutive law reduces to

Department of Civil Engineering 30

Stress -Strain Relations (Contd..)
• If the problem is the plane strain, then 𝜖𝑧 = 0 and 𝜎𝑧 need not enter the solution process. By discarding
rows 3,5 and 6 from the 6-by-6 [E]. We find, for an isotropic material

• Long body subjected to significantly lateral forces but very little

longitudinal forces
• Pipes, long strip footings, retaining walls, gravity dams, tunnels etc
• The displacement in longitudinal direction (z-direction) is zero in typical
strip
Hence, the plane strain problem reduces is

Department of Civil Engineering 31

Stress -Strain Relations (Contd..)
• Poission’s ratio ν is little affected by temperature
• Modulus E is affected more: for stainless steel E decreases about 20% when the
temperature rises from 00C to 4500C
• Barring plastic flow, elastic properties are almost independent of stress
• For example, an increase in hydrostatic pressure from 0 MPa to 350 MPa increase
the moduli of steel and aluminum 0.8% and 2.6 % respectively (F. Birch, 1938)
• When strain rates are high, as in wave propagation, modulus E is higher than its
static value. The difference is negligible for common metals but appreciable for
rubberlike materials
• If test data are lacking, as in often the case with anisotropic materials, we can only
estimate the elastic constants. Even when the constants are known, anisotropy has
an adverse effect on the accuracy of finte element solutions

Department of Civil Engineering 32

Temperature Effects (Initial Stress and Strain)
With additional stress and strain 𝛔 = 𝐄 𝜖 − 𝜖0 + 𝜎0

Where 𝜖0 and 𝜎0 are vectors of initial strains and initial stresses, respectively

As examples, 𝜖0 might describe moisture-induced swelling and 𝜎0 might describe stresses produced
by heating. Alternatively, both effects can be placed in 𝜖0 , or 𝜖0 and 𝜎0 can be viewed as different
ways to express the same thing. For example, free expansion of an orthotropic material with principal
axes xyz produces the strain
𝜖0 = 𝛼𝑥 𝑇 𝛼𝑦 𝑇 𝛼𝑧 𝑇 0 0 0
Where T is temperature above an arbitrary reference temperature and the α’s are coefficients of thermal
expansion

If boundary conditions permit unrestrained expansion and contraction, the material is homogenous, and
the temperature field T= T(x y z) is a linear function x,y, and z, then all stresses are zero in isotropic or
rectilinearly orthotropic bodies.

Department of Civil Engineering 33

Saint Venant’s Principle
There is an approximations in defining boundary conditions to represent a
support-structure interface.

For example, consider a cantilever beam, free at one end and attached to a column with rivets at the
other end.
• whether the riveted joint is totally rigid or partially rigid,
• and as to whether each point on the cross section at the fixed end is specified to have the same
boundary conditions.

Saint Venant considered the effect of different approximations on the solution to the total problem.

It states that as long as the different approximations are statically equivalent, the resulting solutions
will be valid provided we focus on regions sufficiently far away from the support.

That is, the solutions may significantly differ only within the immediate vicinity of the support .

Department of Civil Engineering 34

Some Examples of FEM in Research

3D model of Deformed shape of URM

URM building building after analysis
3D model of Tunnel Excavation
Source:https://www.researchgate.net/publication/322977327_Seismic_Performance_of_
Traditional_Unreinforced_Masonry_Buildings_in_Nepal
with support installation
Source:http://rmlab.ku.edu.np/research
/research-area/

Department of Civil Engineering 35

Some Examples of
FEM in Research ( Contd..)

Source:https://www.researchgate.net/publication/310257509_COMPARATIVE_NUMERICAL_ST
UDY_OF_2D_AND_3D_FINITE_ELEMENT_MODELING_OF_HYDROPOWER_TUNNEL_CASE_STUD
Y_FROM_LESSER_HIMALAYAN_REGION_OF_NEPAL

Department of Civil Engineering 36

https://www.sciencedirect.com/science/article/abs/pii/S014102961300521X

Department of Civil Engineering 37