3/23/2020

References

CE 6504 Finite Element Analysis (Text)

By: S.S. BHAVIKATTI
Finite Elements Method in Structures Concepts and Applications of Finite Element Analysis
(Part 1) By: Robert D. Cook, David S. Malkus and Michael E. Plesha
Finite Element Procedures in Engineering Analysis
By: K.-J. Bathe
The Finite Element Method
O.C. Zienkiewicz
AAiT An Introduction to the Finite Element Method
By: J. N. Reddy
March 2020
Bedilu Habte
AAiT – Civil Engineering – Bedilu Habte 3

Course Outline Internet References:

1. Introduction
FEM Primer Part 1 - 4
2. Preliminaries Mike Barton & S. D. Rajan
3. 1D (2-Node) Line Elements Arizona State University
Bar, Truss, Beam-elements, Shape functions http://enpub.fulton.asu.edu/structures/FEMPrimer-Part1.ppt
4. 2D Elements Introduction to Finite Element Methods
Department of Aerospace Engineering Sciences
Plane Stress and Plane Strain Problems
University of Colorado at Boulder
5. 3D Elements http://www.colorado.edu/engineering/CAS/courses.d/IFEM.d/Home.html
Tetrahedral, Hexahedral Elements Advanced Finite Element Methods
6. Plate Bending & Shells Department of Aerospace Engineering Sciences
University of Colorado at Boulder
7. Further Issues http://www.colorado.edu/engineering/CAS/courses.d/AFEM.d/Home.html
Modeling, Errors, Non-linearity
AAiT – Civil Engineering – Bedilu Habte 2 AAiT – Civil Engineering – Bedilu Habte 4

1
 3/23/2020

I Introduction What is FEM?

Structure is partitioned into FINITE
What is FEM?
ELEMENTS – that are joined to each other at
Finite element method is a numerical method
limited number of NODES
that generates approximate solutions to
engineering problems which are usually expressed
Behavior of an individual element can be
in terms of differential equations.
described with a simple set of equations
Used for stress analysis, heat transfer, fluid flow,
Assembling the element equations, to a large
electromagnetic etc.
set, is supposed to describe the behavior of the
whole structure.

AAiT – Civil Engineering – Bedilu Habte 5 AAiT – Civil Engineering – Bedilu Habte 7

What is FEM? Discretization Example

• Use of several materials within the same structure, Find the circumference of a circle with a unit
• complicated or discontinuous geometry, diameter – find the value of π.
• complicated loading, etc,
Approximation with that of regular polygons:
makes the closed form (analytical) solution of
structural problems very difficult.

One resorts to a numerical solution, the best of

which is the FEM.

AAiT – Civil Engineering – Bedilu Habte 6 AAiT – Civil Engineering – Bedilu Habte 8

2
 3/23/2020

Discretization Example Brief History

A formal mathematical theory for the FEM
Solution: started some 60 years ago
Let n be the number of sides of the The steps in FEA are very similar to the
inscribed or circumscribing polygon. method of the direct stiffness method in matrix
structural analysis
i) Inscribed polygon
 The term finite element was first used by Clough in
Perimeter p = n sin (180/n) 1960.
 The first book on the FEM by Zienkiewicz and Cheung
ii) Circumscribing polygon was published in 1967.
Perimeter p = n tan (180/n)  In the late 1960s and early 1970s, the FEM was applied
to a wide variety of engineering problems.
AAiT – Civil Engineering – Bedilu Habte 9 AAiT – Civil Engineering – Bedilu Habte 11

Discretization Example Brief History

Estimated vs. exact value of π = 3.1415926536
No. of sides Inscribed polygon Error  The Pioneers – 1950 to 1962; Clough,
3 2.5980762114 0.5435164422 Turner, Argyris, etc.; thought structural
4 2.8284271247 0.3131655288 elements as a device to transmit forces
8 3.0614674589 0.0801251947 (“force transducer”).
16 3.1214451523 0.0201475013
32 3.1365484905
 The Golden Age – 1962–1972;
0.0050441630
64 3.1403311570 0.0012614966
Zienkiewicz, Cheung, Martin, Carey etc.;
128 3.1412772509 0.0003154027 thought discrete elements approximate
1000 3.1415874859 0.0000051677 continuum models (displacement
10000 3.1415926019 0.0000000517 formulation).
100000 3.1415926531 0.0000000005
1000000 3.1415926536 0.0000000000

AAiT – Civil Engineering – Bedilu Habte 10 AAiT – Civil Engineering – Bedilu Habte 12

3
 3/23/2020

Brief History Brief History

 Consolidation – 1972 to mid 1980s;
Hughes, Bathe Argyris, etc.; variational
method, mixed formulation, error
estimation.
 Back to Basics – early 1980s to the
present; Elements are kept simple but
should provide answers of engineering
accuracy with relatively coarse meshes.

AAiT – Civil Engineering – Bedilu Habte 13 AAiT – Civil Engineering – Bedilu Habte 15

Brief History Proprietary Software

 The 1970s  advances in  ANSYS
mathematical treatments, including the  MSC/NASTRAN
development of new elements, and  ABAQUS
convergence studies.  ADINA
 Most commercial FEM software  ALGOR
packages originated in the 1970s and  NISA
1980s.  COSMOS/M
 STARDYNE
 The FEM is one of the most important  IMAGES-3D
developments in computational
methods to occur in the 20th century.
AAiT – Civil Engineering – Bedilu Habte 14 AAiT – Civil Engineering – Bedilu Habte 16

4
 3/23/2020

Typical Steps in FEA Process Common FEA Procedure for Structures

1. Discretize and Select Element Type 2 – 4. Derivation of Element Equations
2. Select a Displacement Function Derive the relationship between the unknown and
3. Define Strain/Displacement and given parameters at the nodes of the element.
Stress/Strain Relationships
e e e
4. Derive Element Stiffness Matrix & Eqs. f    k  u 
5. Assemble Equations and Introduce B.C.’s
6. Solve for the Unknown Degrees of 5a. Assembly
Freedom Assembling the global stiffness matrix from the
7. Solve for Element Stresses and Strains element stiffness matrices based on compatibility
8. Interpret the Results of displacements and equilibrium of forces.
For example:
AAiT – Civil Engineering – Bedilu Habte 17 AAiT – Civil Engineering – Bedilu Habte 19

Common FEA Procedure for Structures Common FEA Procedure for Structures
0. Idealization
The given structure needs to be idealized based on
engineering judgment. Identify the governing
equation. U i  u12  u12
1. Discretization Vi  v12  v12
The continuum system is disassembled into a
 Displacement of a node is always
number of small and manageable parts (finite the same for the adjoining elements
elements). and for the whole structure.

AAiT – Civil Engineering – Bedilu Habte 18 AAiT – Civil Engineering – Bedilu Habte 20

5
 3/23/2020

Common FEA Procedure for Structures Common FEA Procedure for Structures
i 1
F x  fx 2  f x 12 6. Solve for the primary unknowns
i 1
F y  fy 2  f y 12 1
 The sum of the forces on each element U    K  F 
of a particular node must balance the
external force at that node.
For the whole structure, this process results in the master 7/8. Compute other values of interest
stiffness equation: Secondary unknowns are determined using the
known nodal displacements. Result
 K '  U    F 
' '
interpretation.

AAiT – Civil Engineering – Bedilu Habte 21 AAiT – Civil Engineering – Bedilu Habte 23

Common FEA Procedure for Structures Types of FEA in Structures

5b. Introduce Boundary Conditions 1. Linear analysis: small deflection and elastic
After applying prescribed nodal displacements material properties.
(and known external forces) to the master stiffness 2. Non-linear analysis:
equation, the resulting equation becomes the • Material non-linearity: small deflection and
modified master stiffness equation: non-linear material properties.
• Geometric non-linearity: large deflection and
elastic material properties.
 K U    F  • Both material and geometric non-linearity.

AAiT – Civil Engineering – Bedilu Habte 22 AAiT – Civil Engineering – Bedilu Habte 24

6
 3/23/2020

Advantages of Finite Element Analysis Measures of Accuracy in FEA

Accuracy
• Models Bodies of Complex Shape
Error = |(Exact Solution)-(FEM Solution)|
• Can Handle General Loading/Boundary Conditions
Convergence
• Models Bodies Composed of Composite Materials
Limit of Error as:
• Model is Easily Refined for Improved Accuracy by
Number of Elements (h-convergence)
Varying Element Size and Type (Approximation or
Approximation Order (p-convergence)
Scheme)
Increases
• Time Dependent and Dynamic Effects Can Be Included Ideally, Error  0 as Number of Elements or
• Can Handle a Variety Nonlinear Effects Approximation Order is Higher

AAiT – Civil Engineering – Bedilu Habte 25 AAiT – Civil Engineering – Bedilu Habte 27

Common Sources of Error in FEA Numerical Methods

 Several approaches can be used to transform
• Domain Approximation the physical formulation of the problem to its
• Element Interpolation/Approximation finite element discrete analogue.
• Numerical Integration Errors  Ritz/ Galerkin methods – the physical
formulation of the problem is known as a
(Including Spatial and Time Integration)
differential equation.
• Computer Errors (Round-Off, Etc., )
 Variational formulation – the physical problem
can be formulated as minimization of a
functional.

AAiT – Civil Engineering – Bedilu Habte 26 AAiT – Civil Engineering – Bedilu Habte 28

7
 3/23/2020

Variational Method cont. Elasticity

 3D Stress block
y
xy
yz
xy
zy x
zx xz
• A mathematical model is a set of z
mathematical statements which attempts to
describe a given physical system.
AAiT – Civil Engineering – Bedilu Habte 29 AAiT – Civil Engineering – Bedilu Habte 31

Variational Method cont. Elasticity cont.

 Strong Form (SF): A system of ordinary or
partial differential equations in space and/or  Stress Equilibrium Equations
time, complemented by appropriate
boundary conditions.
 Weak Form (WF): A weighted integral
equation that “relaxes” the strong form into
a domain-averaging statement.
 Variational Form (VF): A functional whose
stationary conditions generate the weak
and strong forms.

AAiT – Civil Engineering – Bedilu Habte 30 AAiT – Civil Engineering – Bedilu Habte 32

8
 3/23/2020

Elasticity cont. Elasticity cont.

 3D Stress – Strain Relationships

AAiT – Civil Engineering – Bedilu Habte 33 AAiT – Civil Engineering – Bedilu Habte 35

Elasticity cont. Bar Elements

(1) The longitudinal dimension or axial
 Strain – Displacement
dimension is much larger that the
transverse dimension(s). The
intersection of a plane normal to the
longitudinal dimension and the bar
defines the cross sections.
(2) The bar resists an internal axial force
along its longitudinal dimension.

AAiT – Civil Engineering – Bedilu Habte 34 AAiT – Civil Engineering – Bedilu Habte 36

9
 3/23/2020

Bar Elements cont. Bar Elements cont.

• It must be in equilibrium.
• It must satisfy the elastic stress–strain
law (Hooke’s law)
• The displacement field must be
compatible.
• It must satisfy the strain–displacement
equation.

AAiT – Civil Engineering – Bedilu Habte 37 AAiT – Civil Engineering – Bedilu Habte 39

Bar Elements cont. Governing Equation

The governing differential equation of
the bar element is given by
d  du 
 AE q  0
dx  dx 
boundary conditions
u x0
0
 du 
 AE   P
 dx  x L

AAiT – Civil Engineering – Bedilu Habte 38 AAiT – Civil Engineering – Bedilu Habte 40

10
 3/23/2020

Approximate Solution
Total Potential Energy (TPE)
Admissible displacement function is
continuous over the length and satisfies  p  U W
any boundary condition:
dU   x  x  y  z d  x
Strain dU   d  x dV
x
Energy
 ε x 
(Internal U     xd
 x  dV
Principles of Minimum Potential Energy – Of all work)  0 
V
kinematically admissible displacement equations,
those corresponding to equilibrium extermize the 1
TPE. If the extremum is a minimum, the equilibrium
state is stable.
U 
2  
V
x  x dV

AAiT – Civil Engineering – Bedilu Habte 41 AAiT – Civil Engineering

Jima – Bedilu Habte 43

Kinematically admissible Total Potential Energy

Displacement Functions
External work of loads
those that satisfy the single-valued nature of
displacements (compatibility) and the  p  U W
boundary conditions M

Usually Polynomials
W  
V
Xˆ b uˆ dV   Tˆx u dS   fˆi dˆi
S i 1
Continuous within element.
Inter-element compatibility. Prevent overlap TPE
or gaps. 1
p   x  AE  x dx  W
Allow for rigid body displacement and 2 L
constant strain.
AAiT – Civil Engineering – Bedilu Habte 42 AAiT – Civil Engineering
Jima – Bedilu Habte 44

11
 3/23/2020

Total Potential Energy Ritz-Method

1 Using the Ritz-method, approximate
2 displacement function is obtained by:
Assume arbitrary displacement
Strain energy 1   a1 f 1  a 2 f 2  ...  a n f n
  T  Adx
2 L Introduce this into the TPE functional
External energy Performing differentiation and integration to
Pu 2
obtain a function
1 Minimizing the resulting function
Total potential energy Π    T  Adx  Pu 2
2 L d
 0 for i  1 , 2 ,..., n
da i
AAiT – Civil Engineering
Jima – Bedilu Habte 45 AAiT – Civil Engineering – Bedilu Habte 47

Ritz-Method Ritz-Method
Consider the linear elastic one-dimensional
rod with the force shown below

• The potential energy of this system is:

2
1
2  du 
Π  2  EA   dx  2u 1
0
 dx 
AAiT – Civil Engineering – Bedilu Habte 46 AAiT – Civil Engineering – Bedilu Habte 48

12
 3/23/2020

Ritz-Method Galerkin-Method
Consider the polynomial function: For the one-dimensional rod considered in the
pervious example, the governing equation is:
To be kinematically admissible u must satisfy
the boundary conditions u = 0 at both (x =
0) and (x = 2) The Galerkin method aims at setting the
& residual relative to a weighting function Wi,
Thus: to zero. The weighting functions, Wi, are
chosen from the basis functions used for
constructing û (approximate displacement
function).
AAiT – Civil Engineering – Bedilu Habte 49 AAiT – Civil Engineering – Bedilu Habte 51

Ritz-Method Galerkin-Method
• TPE of this system becomes: Using the Galerkin-method, approximate
displacement function is obtained by:
• Minimizing the TPE: The governing DE is written in residual form
d  du 
RES   AE 
dx  dx 
Multiply this RES by weight function f and
• Thus, an approximate u is given by: integrate and equate to zero
Perform differentiation and integration to
obtain the approximate u
• Rayleigh-Ritz method assumes trial
functions over entire structure  L
f R dx  0

AAiT – Civil Engineering – Bedilu Habte 50 AAiT – Civil Engineering – Bedilu Habte 52

13
 3/23/2020

Galerkin-Method Galerkin-Method
Consider the rod shown below Equating the value in the bracket to zero and
performing the integral:
u1  3
4

u  0 . 75 2 x  x 2 
• In elasticity problems Galerkin’s method turns
out to be the principle of virtual work.

• Multiplying by Φ (weighting function) and

integrating gives (by parts):

AAiT – Civil Engineering – Bedilu Habte 53 AAiT – Civil Engineering – Bedilu Habte 55

Bar Element
Galerkin-Method
The function Φ is zero at (x = 0) and (x = 2)
and EA(du/dx) is the force in the rod, which Y   E
ŷ
equals 2 at (x = 1). Thus: x̂, û duˆ
2

dxˆ
L dˆ2 x , fˆ2 x
1 A xˆ  T  A
• Using the same polynomial function for u and
Φ and if u1 and Φ1 are the values at (x = 1): dˆ1 x , fˆ1 x


X
d  duˆ 
• Setting these and  AE   0
dxˆ  dxˆ 
E=A=1 in the integral:
AAiT – Civil Engineering – Bedilu Habte 54 AAiT – Civil Engineering – Bedilu Habte 56

14
 3/23/2020

Assumptions
 dˆ  dˆ 1x 
uˆ   2x  xˆ  dˆ 1x
 L 
The bar cannot resist shear forces.
That is: f̂1ŷ  f̂ 2 ŷ  0  xˆ xˆ   d̂1x 
û  1   
Effects of transverse displacements  L L  dˆ 2x 
are ignored.  d̂ 
û   N1 N 2   1x 
Hooke’s law applies. d̂ 2x 
That is: Where :
 x  E x xˆ xˆ
Shape functions: N1  1  and N 2 
L L

AAiT – Civil Engineering – Bedilu Habte 57 AAiT – Civil Engineering – Bedilu Habte 59

Select a Displacement Function Shape Functions

N1 and N2 are called Shape Functions or Interpolation
Assume a linear function. û Functions. They express the shape of the assumed
uˆ  a1  a2 xˆ displacements.
No. of coefficients = No. of DOF N1 =1 N2 =0 at node 1
N1 =0 N2 =1 at node 2
Written in matrix form:  a1 
uˆ  1 xˆ   N1 + N2 =1 at any point
a 2  1 1
N1
Expressed as function of d̂ 1x and d̂ 2x N2
uˆ (0)  dˆ1x  a1  a2 (0)  dˆ1x  a1
1 2
uˆ ( L)  dˆ2 x  a1  a2 ( L )  dˆ2 x  dˆ1x  a 2 L
L
AAiT – Civil Engineering – Bedilu Habte 58 AAiT – Civil Engineering
Jima – Bedilu Habte 60

15
 3/23/2020

Define Strain/Displacement and Stress/Strain Potential Energy Approach

Relationships
 dˆ  dˆ    U W
uˆ   2 x 1x  xˆ  dˆ1x p

 L  dU   x d  x dV
duˆ dˆ2 x  dˆ1x f  A
  1
dxˆ L  dˆ  dˆ2 x  U 
2   x  x dV
d  dˆ
ˆ fˆ1x  AE  1x 

V
  E  E 2 x 1x  L 
L M
 dˆ  dˆ1x  W    Xˆ b uˆ dV   Tˆx u dS   fˆix dˆix
fˆ2 x  AE  2 x 
 V S i 1
 L 
AAiT – Civil Engineering
Jima – Bedilu Habte 61 AAiT – Civil Engineering
Jima – Bedilu Habte 63

Derive the Element Stiffness Potential Energy Approach

Matrix and Equations 1
L
p  A x  x dxˆ  fˆ1 x dˆ1 x  fˆ2 x dˆ2 x   uˆ Tˆx dS   uˆ Xˆ b dv
 fˆ1x  AE  1  1  dˆ1x  2 0 S V
ˆ    
 f 2 x  L  1 1  dˆ2 x 
uˆ  [ N ] { dˆ }  1 1 ˆ
AE  1  1
 x  
 L L 
d
 kˆ 
L  1 1 
N
 K    kˆ (e)   N   1  xˆ xˆ 


e 1
 L L     B  dˆ
x 
N  dˆ 
Assemble Global Stiffness Matrix, apply BC and solve F     fˆ (e)  
dˆ   1 x  B    1
 L
1
the Master stiffness equation for the unknown
displacements: e 1  dˆ 2 x   L 
 K d   F 
AAiT – Civil Engineering
Jima – Bedilu Habte 62 AAiT – Civil Engineering – Bedilu Habte 64

16
 3/23/2020

Potential Energy Approach Potential Energy Approach

 x   D  x  U    dˆ B D B dˆ

*
T T T

[D] is the constitutive matrix  1

D   E  (elasticity property matrix) U   dˆ
*    1

dˆ2 x  L E 
1   dˆ1x 
 
L   dˆ2 x 
1x

 x   D Bdˆ
1
   L
 L 
E ˆ2
A
L U* 
L2

d1x  2dˆ1x dˆ2 x  dˆ22x 
 p    x  x dxˆ  fˆ1x dˆ1x  fˆ2 x dˆ2 x   uˆ Tˆx dS   uˆ Xˆ b dv
20 S V

 dˆ  fˆ  dˆ
T
fˆ  dˆ2 x fˆ2 x
1x 1x

AAiT – Civil Engineering – Bedilu Habte 65 AAiT – Civil Engineering – Bedilu Habte 67

Potential Energy Approach Potential Energy Approach

 p AL  E ˆ
p 
A
L

2 0
 x T  x  dxˆ  dˆ   P  uˆ  Tˆ dS   uˆ Xˆ dv
T
T
x
T
b
dˆ1x
 ˆ  ˆ
2  L2
 

2 d1 x  2 d 2 x   f1 x  0
S V

L
 p AL  E ˆ
p 
A ˆT T T
2 0

d B D B dˆ dxˆ 
dˆ2 x

2  L2
2d 2 x  2dˆ1x    fˆ 2x 0
T

 dˆ P  dˆ   N  Tˆ dS  dˆ  N  Xˆ dv
T T T T

AE  1 1  dˆ1x   fˆ1x 

x b
S V
   
AL ˆ L  1 1  dˆ2 x   fˆ2 x 
p    B D B dˆ  dˆ  fˆ 
2
d
T T T T

T T
 fˆ  {P}   N  Tˆ  dS   N  Xˆ dv
T
x
T
b
 k     B   D  B  dV  D   D
V
S V

AAiT – Civil Engineering – Bedilu Habte 66 AAiT – Civil Engineering – Bedilu Habte 68

17
 3/23/2020

Transformation Matrix (Local  Global) Quadratic Bar Element

Obtain displacement function for the one-dimensional
 dˆ1x   C S 0 0   d1x  quadratic element with three nodes shown below.
ˆ    
Let  d1 y   S C 0 0   d1 y 
C  cos θ ˆ   
d 2 x   0 0 C S  d 2 x 
S  sin θ dˆ2 y   0 0 S

C  d 2 y 
  A quadratic displacement function:

C S 0 0
 dˆ  T d  uˆ  a1  a2 xˆ  a3 xˆ 2
 S C 0 0   fˆ   T  f  No. of coefficients = No. of DOF  a1 
 
 T   Written in matrix form: uˆ  1 xˆ  
ˆx 2 a2 
 0 0 C S
  a3 
 0 0 S C
AAiT – Civil Engineering – Bedilu Habte 69 AAiT – Civil Engineering – Bedilu Habte 71

Element stiffness equation in local coordinate: uˆ L  d1 a2 L a3 L2

xˆ 
2 d1  a1  
 fˆ1 x  0  1 0   d1 x  2 4
1 uˆ xˆ 0  d2 
ˆ     d 2  a1
 f1 y  AE  0 0 0 0  d1 y 
ˆ     uˆ  d3 a2 L a3 L2
 f 2 x  L  1 0 1 0 d 2 x  xˆ 
L
2
d 3  a1  
  2 4
 fˆ2 y 
  0 
0 0 0 d 2 y   T   f   kˆ  T d
d 3  d1
uˆ  d 2 
xˆ 2
xˆ  2  d 3  2 d 2  d1  2
 f    T   kˆ T d
1 L L
 d1 
 T  1  T  T  xˆ 2 xˆ 2 4 xˆ 2 xˆ 2 xˆ 2   
   2 1  2   d2 
 L L L L L2   
Element stiffness matrix  f    T T  kˆ T d d3  d1 
 
  N1 N2 N3  d2 
in the global coordinate:  k   T T  kˆ T  d 
AAiT – Civil Engineering – Bedilu Habte 70 AAiT – Civil Engineering – Bedilu Habte
 3
72

18
 3/23/2020

Quadratic Shape Functions

Compatibility: The displacement approximation is continuous
N1 =1 N2 = N3 =0 at node 1 across element boundaries
N2 =1 N1 = N3 =0 at node 2
N3 =1 N1 = N2 =0 at node 3

N1 + N2 + N3 = 1 at any point

x1 x2 x3
El #1 El #2

Hence the displacement approximation is continuous across elements

AAiT – Civil Engineering – Bedilu Habte 73 AAiT – Civil Engineering – Bedilu Habte 75

Writing shape functions (without deriving):

PROPERTIES OF THE SHAPE FUNCTIONS
The Kronecker delta property (the shape function at any node
The shape function at any node has a value of 1 at that node and has value of 1 at that node and a value of zero at all other nodes)
a value of zero at ALL other nodes.
1 1
1 1
x 2 -x x -x 1
x 2 -x x-x1 N 1 (x )  N 2 (x ) 
N 1 (x)  N 2 (x)  x 2  x1
x 2  x1 x 2  x1
x 2  x1
x1 x2 x
x1 x2 x El #1
x2 - x Node at which N1 is 0
N1 (x)  x 2 -x 
x 2  x1 N 1 (x ) 
x 2 -x 1  Notice that the length of the element = x2-x1
x 2 - x1
Check  N1 (x  x1 )  1 x 1 -x  x -x 1 
x 2  x1 N 2 (x )  
x 1 -x 2   x 2 -x 1  The denominator is
x -x
and N1 (x  x 2 )  2 2  0
the numerator evaluated at
x 2  x1 the node itself
AAiT – Civil Engineering – Bedilu Habte 74 AAiT – Civil Engineering – Bedilu Habte 76

19
 3/23/2020

3-Node bar element: varying quadratically inside the bar STRAIN and STRESS WITHIN EACH ELEMENT

N1 (x) N 2 (x) N 3 (x)

From equation (1), the displacement within each element

1
 (x )= N d

The strain in the bar d

ε
dx
x1 x2 x3 x Hence
 dN 
N1 (x) 
x 2 - x x 3 - x  u(x)  N1 (x)d1x  N 2 (x)d 2x  N 3 (x)d 3x ε dBd (2)
x 2 - x1 x 3 - x1   dx 

N 2 (x) 
x1 - x x 3 - x  This is a quadratic finite element in The matrix B is known as the “strain-displacement matrix”
x1 - x 2 x 3 - x 2  1D and it has three nodes and three
x1 - x x 2 - x  dN 
N 3 (x)  associated shape functions per element. B  
x1 - x 3 x 2 - x 3   dx 

AAiT – Civil Engineering – Bedilu Habte 77 AAiT – Civil Engineering – Bedilu Habte 79

Writing shape functions for higher DOFs:

For a linear finite element
 x 2 -x x -x 1 
N  N 1 (x ) N 2 (x )    
 x 2  x1 x 2  x1 
Hence
 -1 1  1
B    1 1
 x 2  x1 x 2  x1  x 2  x1

 -1 1   d 1x 
ε  B d    d 
 2  x1
x x 2  x1   2x 
d 2x -d 1x

x 2  x1
For a linear bar element, strain is a constant within the element.

AAiT – Civil Engineering – Bedilu Habte 78 AAiT – Civil Engineering – Bedilu Habte 80

20
 3/23/2020

2 Node Linear Element

Displacement is linear L
 (x )  a 0  a 1x d 2x
x1 x x2
-1  1
d 2x -d 1x x 
d 1x x El #1 ε= 1
x 2  x1
x1 x2
N1 x2-x 1-
Strain is constant N1 = ----------- N1 = -----------
x2-x1 2
d u
The stress in the bar   E ε= E 1
d x x-x1 +1
N2 = ---------- N2 = ---------
The stiffness matrix x2-x1 2
Inside the element,   EB d N2
the approximate
T
stress is k     B  E B  d V
V

AAiT – Civil Engineering – Bedilu Habte 81 AAiT – Civil Engineering – Bedilu Habte 83

Natural Coordinates 3 Node Quadratic Element

L 1 3 x 2 - x x3 - x
2 N1  
x1 x3 x2 x 2  x1 x 3  x1
O P 2 0
1 -1 1 1-   -   1
x1       2
2 1 2
 
3
x2 = x1 + L 1 2 x -x x -x
N2  1  3
Mapped on the following Natural Coordinate x1  x 2 x 3  x 2
2 -1 -  -  1
3      2
 2 1 2
 
P 1 2
-1 1 x - x x2 - x
N3  1 
 x1  x 3 x 2  x 3
3
1 -1-  1 - 
2

1

1

 1  2 
AAiT – Civil Engineering – Bedilu Habte 82 AAiT – Civil Engineering – Bedilu Habte 84

21
 3/23/2020

Beam Theory – Terminology Beam Theory – Kinematics

A general beam is a bar-like member designed to
resist a combination of loading actions such as
biaxial bending, transverse shears, axial stretching
or compression, and possibly torsion.
If the beam is subject primarily to bending and axial
forces, it is called a beam-column.
A beam is straight if its longitudinal axis is straight. It
is prismatic if its cross section is constant.
A spatial beam supports transverse loads that can
act on arbitrary directions along the cross section.
A plane beam resists primarily transverse loading on
a preferred longitudinal plane.

AAiT – Civil Engineering – Bedilu Habte 85 AAiT – Civil Engineering – Bedilu Habte 87

Mathematical Models Beam Theory – Terminology

One-dimensional mathematical models of structural
beams are constructed on the basis of beam
theories.
The simplest and best known models for straight,
prismatic beams are based on the:
 Bernoulli-Euler theory (also called classical beam theory or
engineering beam theory)
 Timoshenko beam theory

AAiT – Civil Engineering – Bedilu Habte 86 AAiT – Civil Engineering – Bedilu Habte 88

22
 3/23/2020

Bernoulli-Euler Model Bernoulli-Euler Beam Theory

AAiT – Civil Engineering – Bedilu Habte 89 AAiT – Civil Engineering – Bedilu Habte 91

Mathematical Model Iso-P Shape Function

Total Potential Energy of Beam Members

AAiT – Civil Engineering – Bedilu Habte 90 AAiT – Civil Engineering – Bedilu Habte 92

23
 3/23/2020

Iso-P Shape Function Element Stiffness

AAiT – Civil Engineering – Bedilu Habte 93 AAiT – Civil Engineering – Bedilu Habte 95

Element Equations Loading

AAiT – Civil Engineering – Bedilu Habte 94 AAiT – Civil Engineering – Bedilu Habte 96

24
 3/23/2020

General 3D Frame Element Example 1 – Analyse the plane truss

AAiT – Civil Engineering – Bedilu Habte 97 AAiT – Civil Engineering – Bedilu Habte 99

Stiffness of a 3D Frame Element Example 1

 EAx EA 
 L 0 0 0 0 0  x 0 0 0 0 0 
L
 12 EI z 6EI z 12 EI z 6 EI z 
 0 0 0 0 0  0 0 0 
 L3 L2 L3 L2 
 12 EI y 6 EI y 12 EI y 6 EI y 
 0 0
L3
0 
L2
0 0 0 
L3
0 
L2
0 
 GI x GI x 
 0 0 0 0 0 0 0 0  0 0 
 L L 
 6 EI y 4 EI y 6 EI y 2 EI y 
 0 0 
L2
0
L
0 0 0
L2
0
L
0 
 6 EI z 4 EI z 6 EI z 2 EI z 
 0 0 0 0 0  0 0 0 
[ Ki j ]   L2 L L2 L 
 EAx 0 0 0 0 0
EAx
0 0 0 0 0 
 L L 
 12 EI z 6 EI z 12 EI z 6EI z 
 0  0 0 0  0 0 0 0  2 
 L3 L2 L3 L 
 0 12 EI y 6 EI y 12 EI y 6 EI y
0  0 0 0 0 0 0 
 L 3
L 2
L 3
L2 
 GI x GI x 
 0 0 0  0 0 0 0 0 0 0 
 L L 
 0 6 EI y 2 EI y 6 EI y 4 EI y
0  0 0 0 0 0 0 
 L 2
L L 2
L 
 6 EI z 2 EI z 6EI z 4 EI z 
 0 0 0 0 0 0 0 0 
 L2 L L2 L 

3/23/2020 98
AAiT – Civil Engineering – Bedilu Habte 98 AAiT – Civil Engineering – Bedilu Habte 100

25
 3/23/2020

Example 2 – Analyze the Plane Frame Example 2 – Boundary Condition & Solve

AAiT – Civil Engineering – Bedilu Habte 101 AAiT – Civil Engineering – Bedilu Habte 103

Example 2 – Element and node numbering

AAiT – Civil Engineering – Bedilu Habte 102 AAiT – Civil Engineering – Bedilu Habte 104

26