Chapter 2 Axial and Torsional Elements

FINITE ELEMENT METHOD
(BDA 31003)
Chapter 2: Axial and Torsional Elements

What is Axial/Torsional Element
This type of element has only one (1) single degree of
freedom in each node. It could be axial or torsional
translation
A : cross section
E: elasticity
L: length of element
rotation
J : Shear modulus
G: Polar Inertia
L: length of element
Problems in Axial/Torsional Elements
Element 1 Element 2
1 2
1 2
Stiffness, Displacement and Force
1 k 2
f1 u1 u2 f2
Rearrange the equation in matrix

Elemental (local) Stiffness Matrix
The stiffness matrix of the element now can be expressed as
Now in general term, in every element consists of i and j nodes

Example Problem 3-1
a. Divide the structure into 6 element. Each segment

consists of 2 elements
b. Calculate the stiffness matrix of each element
Element Node 1 Node 2 L A E
1 1 2 La/2 Aa Ea
2 2 3 La/2 Aa Ea
3 3 4 Lb/2 Ab Eb
4 4 5 Lb/2 Ab Eb
5 5 6 Lc/2 Ac Ec
6 6 7 Lc/2 Ac Ec
Assemble Stiffness Matrices Globally
The assemble of stiffness matrices can be expressed

as simple as
However this is not a direct additional operation but

SELECTIVE
The order of [K} equals to the number of nodes multiply

by the number of degrees of freedom each node
Example Problem 3-2
Assemble the local/individual stiffness matrix in a global structural
stiffness matrix [K]
Arrange the structural stiffness matrix into a complete static matrix
equation [K]{u} = {f}
Element meshing
You want to divide into 3 elements , numbered e1, e2, e3 :

Assemble Elements
global static equation
1 e1 2 e2 3 e3 4
𝑢1 𝑢2 𝑢3 𝑢4
In the system there are 4 nodes and every node has 1 DOF, the
static equation will be:
Assemble Elements
stiffness matrix
Element e1
Assemble Elements
stiffness matrix
1 e1 2 e2 3 e3 4
𝑢1 𝑢2 𝑢3 𝑢4
Element e2
Assemble Elements
stiffness matrix
1 e1 2 e2 3 e3 4
𝑢1 𝑢2 𝑢3 𝑢4
Element e3
Assemble Elements
stiffness matrix – combine all elements
Assemble Elements
force vector
1 e1 2 e2 3 e3 4
𝑢1 𝑢2 𝐹2 𝑢3 𝑢4 𝐹4
Element e1
Assemble Elements
force vector
1 e1 2 e2 3 e3 4
𝑢1 𝑢2 𝐹2 𝑢3 𝑢4 𝐹4
Element e2
Force at node 2
has been defined by e1
Assemble Elements
force vector
Element e3
Assemble Elements
force vector – combine all elements
Constraints
I have now the static equation complete:
I can solve the unknown displacement vector {u} No.... WAIT a minute
Have you ever

considered this:
This end is fixed This end will move

not allowed to move at all at certain distance only,
no more than that
Oh.. No... ?????
Constraints

no more than that
In engineering, these are called CONSTRAINTS
In mathematics, some say BOUNDARY CONDITIONS
The nodes MUST follow the constraints !

Constraints

no more than that
Now you have to express in mathematics, see your FEM model :

Constraints
Before constraints :
After constraints :
The equations
must be modified
so that
Handling constraints
direct elimination method
Before introducing
constraints
After introducing
constraints
direct elimination method
After introducing
constraints
Then solve the simultaneous equation after 2 variables have been eliminated:
to find u2 and u3
Strain and Stress in Axial Element
Strain (ε) in an element occurs when the element is stretched or

compressed, i.e. Elongation (δ)
Stress (σ) in an element
Force (F) in an element

Solving Axial and Torsional Element Problem
1. Form element data table

2. Calculate individual [Ke] and force vector {F}
❖ (save them in global storage)
3. Assembly all element into global stiffness matrix
4. Apply constraints and loads
❖ by using either direct elimination method or penalty method
5. Solve {u} = [K]-1 {F}
6. Obtain displacement vector {u}
7. Calculate strain, 𝜀 𝑒 , stress, 𝜎 𝑒 and force in each element, 𝐹 𝑒 ,

8. Calculate Reactions, {R} = [K] {u} - {F}
Example Problem 3-3
E2=1, A2=2, L2=1,

E1=1, A1=1, L1=1,
F=10
Develop a finite element model to represent the problem

(2 Elements, 3 nodes)
Calculate the displacement in each node, Using Direct

Elimination Method.
Calculate every element [k] and {f}
Element Table
Element Node 1 Node 2 L A E AE/L

1 1 2 1 2 1 2
2 2 3 1 1 1 1
Element 1
Element 2
Assemble all elements to get [K] and {F}
You need to prepare a global storage for [K] and {F}
How to determine the order of global matrix:
Number of DOF in every node: DOF = 1

Number of nodes in the system: N=3
Order of Matrix: DOF x N = 3

Assemble all elements to get [K] and {F}
Elemental stiffness matrices for each element
Global stiffness matrix

Apply constraints
The static matrix equation

Displacements
penalty method
The idea in penalty method:

▪ The correct solutions at the constrained nodes are
obtained by introducing high value at the nodes so this
will ignore other value corresponding to the constraints.
▪ There is no elimination process and solving the equations
can be started immediately. This will be easier to program
and faster solution
penalty method
1 e1 2 e2 3 e3 4
𝑢1 = 𝛽1 𝑢2 𝑢3 𝑢4 = 𝛽4
Before introducing
constraints
After introducing constraints

Using high value penalty :
penalty method
Proof that penalty method can define the constraint
Since C is a huge value compared to other items the equations can be written
penalty method
1 e1 2 e2 3 e3 4
Another alternative approach
in imposing penalty constraints
𝑢1 = 𝛽1 𝑢2 𝑢3 𝑢4 = 𝛽4
Before introducing
constraints
After introducing
constraints
Using high value penalty :
Example Problem 3-4
The bar shown is fixed at the left end and the displacement at the right
end of the bar will be 0.35mm. Modulus of elasticity, E is 200 GPa.
Using Penalty method, Calculate:
(a) The deformation in every node, (b) The strain, ε in every element,
(c) The stress,σ in every element, (d) The elemental forces and
(e) The reaction forces
Element Table
Elemental stiffness matrices for each element
Global stiffness matrix

Apply constraints
Multiplying inverse matrix with the force vector

displacements
Strain, stress and element force
Elem node i node j L ui uj δ=ui-uj ε= δ/L σ =E×ε F=σ×A
1 1 2 0.15 0 0.00102647 410588.235
2 2 3 0.15 0.00102647 0.00130294 110588.235
3 3 4 0.2 0.00130294 0.00154869 110588.235
4 4 5 0.2 0.00154869 0.00035 -539411.765

Reactions {R} = [K] {u} - {F}
 R1   -410588.235 
   
   
 R 2   0 
   
 R3  =  0 N
   
 R4   0 
   
 R5   -539411.764 
   
Multi Point Constraints
If the displacements of nodes are related each other or depending on a

specific function, it is referred to as Multi Point Constraints.
c
3 4
b
a
2 1
1 2 5 u1
u2
u5
Additional Matrix in MPC
In order to obtained the correct equation
Additional matrix should be considered

Example Problem 3-6
Prepare a finite element model consisting of 6 elements

Every segment consists of 2 elements.
(use inner cross section for your model)
a. Calculate the deformation in every node

b. Calculate the strain (ε) in every element
c. Find the stress (σ) in every element
d. Find the elemental force.
e. Find the reaction force
Example Problem 3-7
Prepare a finite element model to investigate the displacement at point

A, B, C, D, E and F also the stresses of each element.
You have to provide of the following information

a. The deformation in every node
b. The strain (ε) in every element
c. The stress (σ) in every element
d. The elemental force.
e. The reaction force
f. Verify all results with FEM application software (LISA)
Loading
Loading can be generated by
● Nodal concentrated loading
● Distributed Body loading
● Thermal loading
Distributed Axial (Body) Loading
Thermal Loading
α : thermal expansion coefficient

Example Problem 3-8
Prepare a finite element model consisting of 3 elements

To represent each segment. (use inner cross section for your model)
a. Calculate the deformation in every segment, at A1, A2, A3 and A4

b. Calculate the strain (ε) in every element
c. Find the stress (σ) in every element
d. Find the elemental force.
e. find the reaction force
Torsional Element
𝐸
𝐽𝐺 𝐺=
𝑘= 2 1+𝜈
𝐿 Shear Modulus
𝜃𝑖 𝜃𝑗
𝐽𝐺 1 −1
Inertia polar or Torsional constant
𝑘𝑒 = 𝜃𝑖 (for circular cross section) 𝜋𝑅4
𝐿 −1 1 𝐽=
𝜃𝑗 2
Summary
𝐴𝐸 1 −1
Axial Load Element stiffness matrix 𝑘 =
𝐿 −1 1
𝐽𝐺 1 −1
Torsional Load Element stiffness matrix 𝑘 =
𝐿 −1 1
Static matrix equation [k] {u} = {F}
Element deflection 𝛿 𝑒 = 𝑢𝑗𝑒 − 𝑢𝑖𝑒
Element strain 𝜀 𝑒 = 𝛿 𝑒 Τ𝐿𝑒
Element stress 𝜎 𝑒 = 𝐸𝑒 𝜀 𝑒
Element force 𝐹 𝑒 = 𝜎 𝑒 𝐴𝑒
Nodal Reaction {R} = [k] {u} - {F}

Thank you

Chapter 2 Axial and Torsional Elements

Uploaded by

Copyright:

Available Formats

Chapter 2 Axial and Torsional Elements

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 2 Axial and Torsional Elements

Uploaded by

Copyright:

Available Formats

What are axial/torsional elements?

What are axial/torsional elements?

How are elemental stiffness matrices derived?

How are elemental stiffness matrices derived?

FINITE ELEMENT METHOD

Chapter 2: Axial and Torsional Elements

Rearrange the equation in matrix

The stiffness matrix of the element now can be expressed as

Now in general term, in every element consists of i and j nodes

a. Divide the structure into 6 element. Each segment

The assemble of stiffness matrices can be expressed

However this is not a direct additional operation but

The order of [K} equals to the number of nodes multiply

You want to divide into 3 elements , numbered e1, e2, e3 :

Have you ever

This end is fixed This end will move

This end is fixed This end will move

In engineering, these are called CONSTRAINTS

In mathematics, some say BOUNDARY CONDITIONS

The nodes MUST follow the constraints !

This end is fixed This end will move

Now you have to express in mathematics, see your FEM model :

Strain (ε) in an element occurs when the element is stretched or

Stress (σ) in an element

Force (F) in an element

1. Form element data table

7. Calculate strain, 𝜀 𝑒 , stress, 𝜎 𝑒 and force in each element, 𝐹 𝑒 ,

E2=1, A2=2, L2=1,

Develop a finite element model to represent the problem

Calculate the displacement in each node, Using Direct

Element Node 1 Node 2 L A E AE/L

You need to prepare a global storage for [K] and {F}

How to determine the order of global matrix:

Number of DOF in every node: DOF = 1

Order of Matrix: DOF x N = 3

Elemental stiffness matrices for each element

Global stiffness matrix

The static matrix equation

The idea in penalty method:

After introducing constraints

The static matrix equation

Multiplying inverse matrix with the force vector

Strain, stress and element force

Elem node i node j L ui uj δ=ui-uj ε= δ/L σ =E×ε F=σ×A

1 1 2 0.15 0 0.00102647 410588.235

2 2 3 0.15 0.00102647 0.00130294 110588.235

3 3 4 0.2 0.00130294 0.00154869 110588.235

4 4 5 0.2 0.00154869 0.00035 -539411.765

If the displacements of nodes are related each other or depending on a

In order to obtained the correct equation

Additional matrix should be considered

Prepare a finite element model consisting of 6 elements

a. Calculate the deformation in every node

Prepare a finite element model to investigate the displacement at point

You have to provide of the following information

Loading can be generated by

● Nodal concentrated loading

● Distributed Body loading

α : thermal expansion coefficient

Prepare a finite element model consisting of 3 elements