CHAPTER THREE

FINITE ELEMENT FORMULATION

FOR TRUSS

Department of Mechanical Engineering

Debre Berhan University, Ethiopia

1
 CONTENTS:

3.1. Finite element formulation for spring

3.2. Finite element formulation for truss

3.2.1. Definition of truss
3.2.2. Finite element formulation for 1-D truss
3.2.3. Finite element formulation for 2-D truss
 3.1. Finite element formulation for spring

Hooks Law/Law of Elasticity

“ For relatively small deformation of an object, the displacement or size

of the deformation is directly proportional to the deforming force or load”

F 𝑭 ∝ ∆ : slope is constant
K
1 ❖ For the linear portion of the graph
𝐹=𝐾 ∗ ∆
∆
Where:
𝐹: External force exerted on the spring
𝑘 : Spring stiffness constant
∆ : Deformation due to external force

2
 Definition of the stiffness matrix and equation

k
• For an element, a stiffness matrix 𝑘 is the matrix that

𝑓 = 𝑘 𝑢 ……………………………….(1)

Where:
𝑓 : Local nodal force
𝑢 :Local nodal displacement
𝑘 : Element stiffness matrix/Local stiffness matrix , which relate the nodal
displacement 𝑢 to the nodal force 𝑓

3
❖ For a connection of series Spring elements
➢ Global stiffness matrix, 𝐾 𝐺

𝐺
𝐾 = 𝑘1 + 𝑘2 +….+ 𝑘𝑛

➢ The Global stiffness matrix equation

𝐺
𝐹 = 𝐾 𝑈 ………………………………….(2)

Where:
𝐹 : Global force
𝐺
𝐾 : Global stiffness matrix
𝑈 : global nodal force

4
 Derivation of the stiffness matrix and equation for a spring element
( A. using direct stiffness method )
o Consider a linear spring (one-dimensional) which is subjected to a tensile
force at both end.

✓ Obey Hooke's law

✓ Resist force only in the direction of the spring

1
k 2
𝐹1 𝐹2 +ve x
𝑈1 𝑈2

Where:
1 &2 : number of node 1 &2 respectively
𝐹1 & 𝐹2 : axial external force at node number 1& 2 respectively
𝑈1 & 𝑈2 : nodal displacement of node 1 &2
K: spring constant 5
 Degree of freedom (DOF)

➢ The number of displacement vector component at a node or element.

Example:
for one-Dimensional spring above
▪ 1 DOF at each node/per node (u1 or u2) or
▪ 2 DOF per element (u1 and u2)

6
 GENERAL STEPS

Step 1: Discretization and Element type selection

Element type selection

▪ One element
✓ 1-Dimensional 2 node element( u1 &u2)
➢ 2 DOF/element

Step 2: Derive the shape function or displacement function or

Interpolation function

✓ We will discuss briefly on chapter four

7
 Step 3: Define the stress/strain, and strain/Displacement Relation ship

o The tensile force (𝐹) produce a total elongation (deformation) ∆ of the

spring.
❖ Deformation of the spring:
∆ = 𝑢(𝐿)-𝑢(0)= 𝑢2 -𝑢1
➢ The total deformation is the difference of the nodal
displacement in the x-direction

o The 𝜎Τ𝜀 relation ship can be expressed in terms of the force/deformation

relationship
➢ Hooke’s law:
𝐹=𝑘∆
𝐹 = 𝑘(𝑢2 -𝑢1 )

8
 Step 4: Derive the local stiffness matrix/ element matrix and equations

A. Free body diagram of the spring element (element FBD)(using nodal

force)

1 𝑘 (1) 2
𝑓1 (1) 𝑓2 (1)
+ve x
𝑢1 (1) 1 𝑢2 (1)

Where:
✓ 𝑓1 (1) & 𝑓2 (1) : local nodal force (internal force) for element number one, at node
1 & 2 resp.
✓ 𝑢1 (1) & 𝑢2 (1) : local nodal displacement for element number one, at node 1 &2
resp
✓ 𝑘 (1) : local stiffness matrix for element number one
13
 B. Free body diagram for each node

Node #1 Node #2
(1) 1 1
𝑓1 𝐹1 𝑓2 (1)
𝐹2

C. Equilibrium Condition (Force equilibrium)

✓ If the spring is in equilibrium condition

σ 𝐹𝑥 =0

𝐹1 +𝐹2 =0

--≫ 𝐹1 =- 𝐹2 =-k(𝑢2 (1) - 𝑢1 (1) )

 # Force in each node:
Node #1 Node #2
𝑓1 (1) +( - 𝐹1 ) =0 𝑓2 (1) +( - 𝐹2 ) =0
⟹ 𝑓1 (1) = 𝐹1 = −𝑘 𝑢2 1
− 𝑢1 1
⟹ 𝑓2 (1) = 𝐹2 = 𝑘 𝑢2 1 − 𝑢1 1

𝑓1 (1) = 𝑘* 𝑢1 1 - 𝑘* 𝑢2 1 ……..(i) 𝑓2 (1) = −𝑘* 𝑢1 1 + 𝑘* 𝑢2 1 ……..(ii)

In matrix form: In matrix form:
(1) 𝑢1 1 (1) 𝑢1 1
𝑓1 = 𝑘 −𝑘 1
--------- (iii) 𝑓2 = −𝑘 𝑘 --------- (iv)
𝑢2 𝑢2 1

Combining equation (iii) and (iv)

𝑓1 (1) 𝑘 −𝑘 𝑢1 (1)
=
𝑓2 (1) −𝑘 𝑘 𝑢2 (1)

[𝑘](𝑒) → local element stiffness matrix

9
D. Compatibility condition
State that:
“ Displacement at all point is not a function of the path”
⟹ each node has one absolute displacement
✓ We used mostly for a shared node.

Example:
1 𝑘1 2 𝑘2 3

1 2
1 𝑘 (1) 2 2 𝑘 (2) 3
𝑓1 (1) 𝑓2 (1)𝑓2 (2) 𝑓3 (2)
𝑢1 (1) (1) 𝑢2 (2) 𝑢3 (2)
𝑢2

⟹ 𝑢2 (1) = 𝑢2 (2) =𝑢2

10
For a simplification:

✓ 𝑢1 (1) =𝑢1
✓ 𝑢3 (2) = 𝑢3

𝑓1 (1) 𝑘1 −𝑘1 𝑢1
𝑓2 (1)
=
−𝑘1 𝑘1 𝑢2 ….. For element #1

𝑓2 (2) 𝑘2 −𝑘2 𝑢2
𝑓3 (2)
=
−𝑘2 𝑘2 𝑢3 ….. For element #2

Therefore:

𝑓1 (1) 𝑘 −𝑘 𝑢1
= 𝑢2
𝑓2 (1) −𝑘 𝑘
11
Step 5: Assemble them & introduce the boundary condition
# for more than one element

𝐾 𝐺 = σ𝑁
𝑒=1 𝑘
(𝑒) and 𝐹 = σ𝑁
𝑒=1 𝑓
(𝑒)

Where:

✓ 𝑘 (𝑒) & 𝑓 (𝑒) are now the element stiffness matrix and force matrices
expressed in a global reference frame

𝐺
𝐹 = 𝐾 𝑈

𝐹 : Global nodal force matrix[ reaction force / external force(applied)]

𝐾 𝐺 : Global stiffness matrix
𝑈 : Global nodal displacement
12
 We can assemble using one of this two method:

a. Equilibrium method/ Newton third law

b. Super position method (Direct stiffness method)

a. Equilibrium method

Node #1 Node #2
𝑓1 (1) +( - 𝐹1 ) =0 −𝑓2 (1) +( 𝐹2 ) =0
⟹ 𝐹1 = 𝑓1 (1) = −𝑘(𝑢2 - 𝑢1 ) ⟹ 𝐹2 = 𝑓2 (1) = 𝑘(−𝑢1 + 𝑢2 )
Re arranging and in matrix form

𝐹1 𝑘 −𝑘 𝑢1
= 𝑢2
𝐹2 −𝑘 𝑘
b. Direct stiffness method

𝑢1 𝑢2 𝑢1 𝑢2
𝑢1 𝑢1
𝐹1 = 𝑘 −𝑘 𝑢2 𝐹2 = −𝑘 𝑘 𝑢2

in matrix form

𝐹1 𝑘 −𝑘 𝑢1
= 𝑢2
𝐹2 −𝑘 𝑘
 Example 1
For the following spring
Determine:
a. The deformation
b. The reaction force using DSM
k
P
L

Given:
K=200N/m
P=100 N
L=5m
Example 2
𝑘1 𝑘2
P
𝐹3

Given:
Determine:
𝑘1 =400N/m
𝑘2 = 600 N/m a. The nodal displacement
b. The reaction force at node #1
P=100 N c. Nodal force for each element
𝐹3 = 200 N

28
Example 3

Given:
Determine:
𝑘1 =1000N/mm
𝑘2 = 2000 N/mm a. The Global stiffness matrix
b. The displacement of node 3 &4
𝑘3 = 3000 N/mm c. The reaction force of node 1&2
P=5000 N d. Force in each spring (nodal force)
o Node 1 and 2 are fixed
Example 4

Derive the global stiffness matrix and global stiffness matrix

equation for the below spring assemblage

ans
 Excercise

Given:
𝑘1 =1000N/mm Determine:
𝑘2 = 5000 N/mm
a. The Global stiffness matrix
𝑘3 = 500 N/mm b. The nodal displacement
c. The reaction force
P=400 N
d. Force in each element (nodal force)
Derivation of the stiffness matrix and equation for a spring element

( B. principle of Minimum potential energy Method)

❖ It is more adaptable to the determination of the element stiffness matrix and

equation for a complicated element such as:
✓ Plane stress/strain element
✓ Axisymmetric stress element
✓ Plate bending element
✓ 3-D solid stress element
❖ It is more applicable to Elastic Materials
❖ It is categorized as a “Variational Methods”
 Minimum potential energy Method

What is energy ?
✓ Defined as “ the capacity to do work”

Potential Energy?
✓ Defined as “ the energy stored on the body due to the work done on it
for changing its position, shape or configuration ”
✓ The amount of energy that exists due to the object position.

Types of potential energy:

✓ Gravitational potential energy
✓ Elastic potential energy
✓ Chemical energy & so on
Total potential energy(𝝅𝑷 ):
✓ Defined as the sum of the strain energy and the potential energy(𝑈) of the
external force (Ω)
𝜋𝑃 =𝑈+Ω …………………………………..(1)

Where :
✓ 𝜋𝑃 : total potential energy
✓ 𝑈:strain energy
✓ Ω :potential energy of the external forces
 Strain energy (𝑼 ):

✓ Is the capacity of the internal forces (or stress) to do work through

deformations(strain)
✓ Work done on the system

Potential energy of the external force(Ω):

✓ Is the capacity of forces such as body forces(self weight ),

surface traction forces(distributed force) and the applied nodal
forces to do work through deformation of the structure )
✓ Work done by the system
 Stiffness matrix and stiffness matrix equation for a spring
element

A. Spring subjected to gradually

B. Force/Deformation curve for
increasing force F
linear spring

We know that:
𝐹𝑚𝑎𝑥
spring constant ( K) = ………………….(3)
𝑋𝑚𝑎𝑥
Hooke's law:
𝐹
F=K*x= 𝑚𝑎𝑥 *x …………………………..(4)
𝑋𝑚𝑎𝑥
 Ԧ
External work done(𝑊𝑒 ) by 𝐹:

𝑥𝑚𝑎𝑥 𝑥 𝐹𝑚𝑎𝑥 .𝑥𝑚𝑎𝑥

𝑊𝑒 = ‫𝐹 ׬‬.𝑑𝑥 = ‫׬‬0 𝐹𝑚𝑎𝑥 𝑑𝑥 = ………..(5)
𝑥𝑚𝑎𝑥 2

✓ we note that F and dx are in the same direction when expressing

the second integral on the right side of Eq. (3).

❖ By the conservation of mechanical energy principle, the external

work due to the applied force F is transformed into the internal
strain energy (U) of the spring. This strain energy is then given
by:
𝐹 .𝑥
𝑊𝑒 =U= 𝑚𝑎𝑥 𝑚𝑎𝑥 ……………………….(6)
2
❖ Upon gradual reduction of the force to zero, the spring returns to its original
undeformed state. This return energy that is stored in the deformed elastic
spring is called internal strain energy (strain energy )

❖ Also :

𝐹𝑚𝑎𝑥 =k * 𝑥𝑚𝑎𝑥 …………………..(7)

Substitute equation (5) into (4), we can get :

𝐾 𝑥 2 𝑚𝑎𝑥
U= ……………………………………….(8)
2
Potential energy of the external force (Ω ):
✓ opposite in sign from the external work expression because the potential
energy of the external force is lost when the work is done by the external
force.
Ω= -𝐹𝑚𝑎𝑥 𝑥𝑚𝑎𝑥 ………………………………………………..(9)

Therefore, substituting eq (8) and eq(9) in to eq(1), yields the total potential
energy as:
1
𝜋𝑃 = k 𝑥 2 𝑚𝑎𝑥 - 𝐹𝑚𝑎𝑥 𝑥𝑚𝑎𝑥 ------------------------(10)
2
❖ In general for any deformation x of the spring corresponding to the force F, we
replace 𝒙𝒎𝒂𝒙 with x and 𝑭𝒎𝒂𝒙 with F and express 𝑼 and Ω as :
𝑥2
𝑈(𝑥) =k ………………………………………………..(9.a)
2

Ω(x) = -F. x ……………………………………………...(9.b)

Therefore by substituting eq(9.a) and (9.b) in to eq(1); we express the total

potential energy as :
𝑥2
𝜋𝑝 (𝑥) = k - F .x ………………………………………..(10)
2
Potential Energy:
✓ When potential energy is minimum, the body has a stable
equilibrium .
 Principle of Minimum potential energy
State that :
“of all the geometrically possible shapes that a body can
assume, the true one, corresponding to the satisfaction of the
stable equilibrium body, is identified by a minimum value of the
total potential energy”

𝜕𝜋 𝜕𝜋
=0 (i=1,2,3…n) or =0 …………….(2)
𝜕𝑑𝑖 𝜕 𝑑

Where
𝑑 = nodal displacement
 Example 1:
For the linear-elastic spring subjected to a force of 1000 lb shown in below
, evaluate the potential energy for various displacement values and show
that the minimum potential energy also corresponds to the equilibrium
position of the spring.
Example 2:

Derive the element stiffness matrix and element stiffness matrix

equation for below spring element. Using PMPEM

Figure : Linear spring subjected to nodal forces

Example 3
𝑘1 𝑘2
P
𝐹3

Given:
Determine:
𝑘1 =400N/m
𝑘2 = 600 N/m a. The nodal displacement
b. The reaction force at node #1
P=100 N c. Nodal force for each element
𝐹3 = 200 N

Using minimum potential energy method

28
Example 4

For the following spring assemblage

𝑘2
𝑘4
𝑘1 3 4 500 N
1 2 300 N

𝑘3
Given: Determine:

𝑘1 =1200N/mm a. The Global stiffness matrix

𝑘2 = 2000 N/mm b. The displacement
c. The reaction force
𝑘3 = 𝑘4 = 3500 N/mm Using Minimum potential energy method
 Exercise

Given:
Determine:
𝑘1 =1000N/mm
𝑘2 = 2000 N/mm a. The Global stiffness matrix
b. The displacement of node 3 &4
𝑘3 = 3000 N/mm c. The reaction force of node 1&2
P=5000 N d. Force in each spring (nodal force)
o Node 1 and 2 are fixed
 3.2. Finite element formulation for Truss/Bar

3.2.1. Definition of truss

✓ A truss is an engineering structure consisting of a straight members connected at

their ends by means of bolt, rivets, pins and welding.

✓ truss offers practical solutions to many structural problems in engineering such

as:
❖ power transmission
❖ Towers
❖ Bridges
❖ Roofs of buildings and so on
Truss have the following characteristics

1. All two force member (i.e. no bending)

✓ Members of a truss are generally considered to be two-force members.
➢ This term means that internal force act in equal and opposite
direction along the members.

Figure: a simply truss subjected to a load

2. Pinned joint( Good assumption even with bolt and welded)
[connection/joint type]
✓ The members are connected together by frictionless pin, bolt, rivelt’s, and
welding .

3. Loads applied at a joint

✓ All loads must be applied at a joints. This assumption is true for most
situations because trusses are designed in a manner such that the majority of
the loads are applied at a joints.
✓ Usually the weight of the members are negligible when compared those to the
applied load
❖ However, if the weight of the members are to be considered, then half of
the weight of each member is applied to the connecting joints.
 3.2.2. Finite element formulation of one- dimensional truss/Bar

Lets consider the deflection of a single members (uniform prismatic bar) when it is
subjected to force P
A,E
y P

L ∆L
x
Step 1: Discretization and element type selection
1 2 𝑓1 (1) 𝑓2 (1)
𝑢1 (1) 𝑢2 (1)

Step 2: Derive or select the shape function / displacement function/interpolation

function

❖ we will discuss briefly on chapter four !

Step 3: Define the stress/strain & strain/displacement relation ship

❖ The average normal strain(𝜀) of the member:

∆𝐿
𝜀 = ………………………………(1)
𝐿

❖ The average stress(𝜎) in the member:

𝐹
𝜎= ………………………………(2)
𝐴

❖ Over th elastic region, the stress and strain are related by Hooke’s law
𝜎= 𝐸 𝜀………………………………(3)
Combining equation(1),2, and 3
𝐸.∆𝐿
𝜎=
𝐿

𝐹 𝐸.∆𝐿
=
𝐴 𝐿

𝐸.𝐴
⇒ 𝐹=( )* ∆𝐿
𝐿

⇒ 𝐹=𝑘𝑒𝑞 * ∆𝐿 ------------------------------(4)
𝐸.𝐴
Where 𝑘𝑒𝑞 = =Equivalent stiffness of the bar ………………….(5)
𝐿

N.T: Equation(4) implies that a bar is acting like a spring!

Step 4: Derive the local stiffness matrix and equation (using DSM)

1 2 (1)
𝑘𝑒𝑞
𝑓1 +ve X
𝑓2 (1)
𝑢1 (1) 𝑢2 (1)

⟹Element stiffness matrix for bar

𝐸𝐴 𝐸𝐴
𝑘𝑒𝑞 −𝑘𝑒𝑞 𝐿 𝐿
𝑘= = 𝐸𝐴 𝐸𝐴
−𝑘𝑒𝑞 𝑘𝑒𝑞
𝐿 𝐿

𝐸𝐴 1 −1
Or 𝑘 = ∗
𝐿 −1 1

⟹ The element stiffness matrix equation:

𝑓1 (1) 𝑘𝑒𝑞 −𝑘𝑒𝑞 𝑢1

𝑓2 (1)
=
−𝑘𝑒𝑞 𝑘𝑒𝑞 𝑢2 …………………….(6)
 Substituting equation 5 into 6:

𝑓1 (1) 𝐸𝐴 1 −1 𝑢1
(1)
== ∗ * 𝑢 ………………………….(7)
𝑓2 𝐿 −1 1 2

Step 5: Assemble the element stiffness matrix equation to obtain the global stiffness matrix
equation, and introduce the boundary condition
𝐸𝐴 𝐸𝐴
𝐹1 𝐿 𝐿 𝑢1
= * 𝑢 ………………………………….….(8)
𝐹2 𝐸𝐴 𝐸𝐴 2
𝐿 𝐿

⇒ we know that, 𝐹2 =P(external applied force at node 2) and 𝐹1 is the reaction

force at node 1
➢ Introducing the boundary condition to equation (8)
✓ i.e. 𝑢1 =0
=0

𝐸𝐴 𝐸𝐴
𝐹1 𝐿 𝐿 𝑢1 =0
= 𝐸𝐴 𝐸𝐴 * 𝑢
𝑃 2
𝐿 𝐿

𝐸𝐴
⇒ 𝑃 = * 𝑢2 ………………………………..(9)
𝐿

Step 6: Solve the unknown Degree of freedom(or Generalized Displacement)

𝑃∗𝐿
⇒ 𝑢2 = ………………………………………..(10)
𝐸∗𝐴
Step 7: Solve for the element stress, strain, reaction force and internal nodal force

⟹ 𝜎1 =𝐸 *𝜀1
Where
✓ 𝜎1 = stress for element number 1
(𝑢 −𝑢 )
✓ 𝜀1 = strain for element number 1, 𝜀1 = 2 1
[

(𝑢2 −𝑢1 ) 𝑃∗𝐿

𝜎1 = 𝐸 *𝜀1 = 𝐸 * ; where 𝑢2 =
[ 𝐸∗𝐴

𝑃
⇒ 𝜎1 =
𝐴

Step 8: Interpret the results

Example #1
For the following bar:
Aluminum Given
Steel

x 𝐸𝑎𝑙 =70 Gpa

P 𝐸𝑠𝑡𝑒𝑒𝑙 =200 Gpa
P=20KN
1m 1m
𝐴𝑎𝑙 =2*10−4 𝑚2
𝐴𝑠𝑡𝑒𝑒𝑙 = 4*10−4 𝑚2

Determine:
A. The global stiffness matrix and displacement
B. The reaction force
C. The strain in each elements &
D. The internal(nodal) force in each nodes(Exercise)
 Exercise

1. A bar assemblage is loaded with force P at one end and constrained at the other
end, as shown on the figure below
E,2A E,A
E,A
1 2 3 4 P x

Determine
A. The displacement at node 2 and 3
B. The stress in the bar assembly
C. The reaction force
2. For the following bar assemblage, determine the nodal displacements, the force
in each elements, & the reaction forces

1 2 3
5 KN
1m 1m

𝐸=210 Gpa
A= 4*10−4 𝑚2
 Individual assignment

Drive the transformation of vectors in two-dimensional(Rotation matrix)

 3.2.3. Finite element formulation of two- dimensional truss

❖ In many problems, it is convenient to introduce both the local (𝑥 ′ ,𝑦 ′ ) and

global (𝑥, 𝑦) coordinate (reference ) system.

❖ Local coordinate’s are always chosen to represent the individual elements

conveniently.

❖ Global coordinate’s are chosen to be convenient for the whole structure.

❖ In general, two frame of references are required to describe the truss

structure:
A. Global coordinate system(fixed) (𝑥, 𝑦)
B. Local coordinate system (movable) (𝑥 ′ ,𝑦 ′ )
❖ The global coordinate system will be used to define:

1. Node location and element angle(𝜽)

⇒ to represent the location of each joints(nodes) & to keep track of the
orientation of each members(elements), using angle such as 𝜃

2. Constraints (i.e. Boundary condition) & external loads

⇒ to apply the constraints & the applied load in terms of their respective
global components.

3. Solutions(i.e. Displacements)
⇒ to represent the solutions, that is the displacement of each joints in global
direction
❖ The local coordinate system will be used to define the individual
elements behavior and values:

1. Internal (nodal) forces

2. Element displacement &
3. Element stresses

⇒ The local coordinate system is needs to describe the two force members
behavior of individual members (elements)
 Transformation of vectors in two-Dimension [ Rotational Matrix]

✓ We want that to relate the components of this vector in one coordinate

system to components in another
❖ Assume ‘𝑑’ is not coincident with either the local or the global axis.
⟹ we want to relate the global displacement to the local one
𝑦
𝑦 ′
𝑑
𝑥′

𝑦
𝜃 𝑥
𝑥

✓ There are two axes above. The prime axes are rotate with respect to the
original x-y systems
 position of point ‘d’

# From global coordinate system # From local coordinate system

𝑑 = (x,y) 𝑑 = (𝑥 ′ ,𝑦 ′ )

❖ Relate the (𝑥 ′ ,𝑦 ′ ) & (x,y) : How?

⇒ The secret is constructing the cut triangle

𝑦
𝑦′

𝑦
𝐼𝐼
𝜃 𝐼
𝑥
𝑥 ′ =𝑥 𝑐𝑜𝑠𝜃 + 𝑦 sin 𝜃
𝑦 ′ =𝑦 cos 𝜃 − 𝑥 𝑠𝑖𝑛𝜃 …………………..(a)

Re- arranging:
𝑥 ′ =𝑥 𝑐𝑜𝑠𝜃 + 𝑦 sin 𝜃
𝑦 ′ = -𝑥 𝑠𝑖𝑛𝜃+ 𝑦𝑐𝑜𝑠𝜃 ……………………..(b)

Equation (b ) in matrix form :

𝑥′ 𝑐𝑜𝑠𝜃 sin 𝜃 𝑥
=
𝑦 ′ −𝑠𝑖𝑛𝜃 𝑐𝑜𝑠𝜃 𝑦

Transformation matrix 𝑇

𝑥′ 𝑥 𝑥 −1 𝑥′
⇒ ′ = 𝑇 𝑦 or = 𝑇
𝑦 𝑦 𝑦′
A. Global stiffness matrix for arbitrary oriented in the 2-D plane
y 𝑣2 ,𝑓2𝑦

2 𝜃
𝑢2 ,𝑓2𝑥
𝑣1 ,𝑓1𝑦

1 𝜃 𝑢 ,𝑓
1 1𝑥

Note that:

✓ The lateral displacement (𝑣 ′1 and 𝑣 ′ 2 ) does not contribute to the stretch of the
bar, within the linear theory.

=> 𝑣 ′1 = 𝑣 ′ 2 =𝑓 ′1𝑦 =𝑓 ′ 2𝑦 =0
where

x,y : Global coordinate system in x axis and y axis respectively

𝑥 ′ ,𝑦 ′ : local coordinate system in 𝑥 ′ and 𝑦 ′ axis respectively
𝑢′1 , 𝑢′ 2 : local coordinate nodal displacement at node 1 and 2 in 𝑥 ′ axis respectively
𝑓 ′1𝑥 , 𝑓 ′ 2𝑥 : local coordinate nodal force at node 1 and 2 in 𝑥 ′ axis respectively
𝑣 ′1 , 𝑣 ′ 2 : local coordinate nodal displacement at node 1 and 2 in 𝑦 ′ respectively
𝑓 ′1𝑦 , 𝑓 ′ 2𝑦 :local coordinate nodal force at node 1 and 2 in 𝑦 ′ respectively

𝑢1 , 𝑢2 : global coordinate nodal displacement at node 1 and 2 in x axis respectively

𝑓1𝑥 , 𝑓2𝑥 : global coordinate nodal force at node 1 and 2 in x axis respectively
𝑣1 , 𝑣2 :global coordinate nodal displacement at node 1 and 2 in y axis respectively
𝑓1𝑦 , 𝑓2𝑦 : global coordinate nodal force at node 1 and 2 in y axis respectively
 Derivation of the local stiffness matrix and equation
(in the local coordinate system)
y 𝑓2𝑦 𝑦′
𝑓1𝑦 2 𝑓2𝑥
1 A,E 2
𝑥′
𝑓1𝑥 𝑢′1 , 𝑓 ′1𝑥 L 𝑢′ 2 , 𝑓 ′ 2𝑥
1
x

The local equation:

𝑓 ′1𝑥 = -𝑘𝑒𝑞 (𝑢′1 - 𝑢′ 2 )

𝑓 ′ 2𝑥 = -𝑘𝑒𝑞 (𝑢′ 2 - 𝑢′1 )
In matrix form:

𝑓 ′1𝑥 1 −1 𝑢′1 𝐸𝐴
′ = 𝑘𝑒𝑞 ′ ; 𝑘𝑒𝑞 =
𝑓 2𝑥 −1 1 𝑢2 𝐿

𝑘′ 𝑒

𝐸𝐴 1 −1
𝑘′ 𝑒 =
𝐿 −1 1

where 𝑘 ′ 𝑒 :Local element stiffness matrix[ from local coordinate]

Therefore:

𝑓′ 𝑒 = 𝑘′ 𝑒 𝑢′ 𝑒 …………………………………….(1)

Where:

𝑓 ′ 𝑒 = local coordinate nodal force

𝑘 ′ 𝑒 =local coordinate stiffness matrix
𝑢′ 𝑒 =local coordinate nodal displacement
 Relate local coordinate nodal force with the global coordinate nodal
force
Using transformation matrix:

𝑓 ′1𝑥 =𝑓1𝑥 𝑐𝑜𝑠𝜃 + 𝑓1𝑦 sin 𝜃

𝑓 ′ 2𝑥 =𝑓2𝑥 𝑐𝑜𝑠𝜃 + 𝑓2𝑦 sin 𝜃

In matrix form:
𝑓1𝑥
𝑓 ′1𝑥 cos 𝜃 sin 𝜃 0 0 𝑓1𝑦
=
𝑓 ′ 2𝑥 0 0 cos 𝜃 sin 𝜃 𝑓2𝑥
𝑓2𝑦

𝑓′ 𝑒 =𝑇 𝑓 𝑒 ………………………………..(2)
Where :

𝑓 ′ 𝑒 :local coordinate nodal force

𝑓 𝑒 : global coordinate nodal force
 𝑇 : transformation matrix

cos 𝜃 sin 𝜃 0 0
=
0 0 cos 𝜃 sin 𝜃

Note that:

Forces in local coordinate: Forces in global coordinate:

𝑓 ′1𝑥 𝑓1𝑥
𝑓1𝑦
𝑓 ′ 2𝑥
𝑓2𝑥
𝑓2𝑦
Equation(2) can be rewritten as:

𝑓 𝑒 =𝑇 −1 𝑓′ 𝑒
N.T:

▪ For the orthogonal coordinate system; the transformation matrix has the
characteristics of :

𝑇 −1 = 𝑇 𝑇 …………………………..(i)

▪ If 𝑘 ′ 𝑒 is symmetric, then 𝑘 𝑒 remain symmetric

cosθ 0
⸫ T −1 = sin θ 0 ……………………………….(3)
0 cosθ
0 sinθ
Relate the local nodal displacement with the global nodal displacement

y 𝑣2

𝑢2
𝑣1 2

𝑢1
1
x
Note that:
Local coordinate nodal displacement Global coordinate nodal displacement
′
𝑢1
𝑢1
𝑣1
𝑢′ 2
𝑢2
2 DOF /element 𝑣2
4 DOF /element
Using transformation matrix:

𝑢′1 =𝑢1 𝑐𝑜𝑠𝜃 + 𝑣1 sin 𝜃

𝑢′ 2 =𝑢2 𝑐𝑜𝑠𝜃 + 𝑣2 sin 𝜃

In matrix form:
𝑢1
𝑢′1 cos 𝜃 sin 𝜃 0 0 𝑣1
𝑢′ 2
=
0 0 cos 𝜃 sin 𝜃 𝑢2 …….…………………….(a)
𝑣2

𝑢′ 𝑒 =𝑇 𝑒 𝑢 𝑒 ……………………………………..(4)
Where :

𝑢′ 𝑒 :local coordinate nodal displacement

𝑢 𝑒 : global coordinate nodal displacement
𝑇 𝑒 : transformation matrix
 Substitute equation(4) and (2) into equation (1)

𝑇 𝑒 𝑓 𝑒 = 𝑘′ 𝑒 𝑇 𝑒 𝑢 𝑒 ………………………………… (5)

Inorder to get the final equation : 𝒇 = 𝑲 𝒖 ; multiply equation (5) both side by
𝑇 −1
𝐼
−1 𝑒 𝑒 −1
𝑇 𝑇 𝑓 = 𝑇 𝑘′ 𝑒
𝑇 𝑒
𝑢 𝑒

𝑒 −1
𝑓 =𝑇 𝑘′ 𝑒
𝑇 𝑒
𝑢 𝑒
; 𝑇 −1
= 𝑇 𝑇

𝑓 𝑒 =𝑇 𝑇 𝑘′ 𝑒 𝑇 𝑒 𝑢 𝑒

⸫ 𝑓 𝑒 = 𝑲 𝑒 𝑢 𝑒

Where:
𝑲 𝑒 = Global element stiffness matrix
= 𝑇 𝑇 𝑘′ 𝑒 𝑇 𝑒
 𝑲 𝑒= 𝑇 𝑇
𝑘′ 𝑒
𝑇 𝑒

𝑐𝑜𝑠𝜃 0
𝐸𝐴 1 −1 cos 𝜃 sin 𝜃 0 0
= 𝑠𝑖𝑛𝜃 0 * *
0 𝑐𝑜𝑠𝜃 𝐿 −1 1 0 0 cos 𝜃 sin 𝜃
0 𝑠𝑖𝑛𝜃

After multiplication of the above matrix, we can get:

𝑐𝑜𝑠 2 𝜃 𝑐𝑜𝑠𝜃𝑠𝑖𝑛𝜃 −𝑐𝑜𝑠 2 𝜃 −𝑐𝑜𝑠𝜃𝑠𝑖𝑛𝜃

𝑲 𝑒 = * 𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃
𝐸𝐴 𝑠𝑖𝑛2 𝜃 −𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃 −𝑠𝑖𝑛2 𝜃
𝐿 −𝑐𝑜𝑠 2 𝜃 −𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃 𝑐𝑜𝑠 2 𝜃 𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃
−𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃 −𝑠𝑖𝑛2 𝜃 𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃 𝑠𝑖𝑛2 𝜃

𝑒
N.T: 𝑲 : have a symmetrical property
Or, simply

𝑐2 𝑐𝑠 −𝑐 2 −𝑐𝑠
𝑲 = * 𝑐𝑠2
𝑒 𝐸𝐴 𝑠 2 −𝑐𝑠 −𝑠 2 ……………………(6)
𝐿 −𝑐 −𝑐𝑠 𝑐 2 𝑐𝑠
−𝑐𝑠 −𝑠 2 𝑐𝑠 𝑠2

❖ Where
✓ c=𝑐𝑜𝑠𝜃
✓ s=𝑠𝑖𝑛𝜃
✓ 𝑐 2 = 𝑐𝑜𝑠 2 𝜃
✓ 𝑐𝑠=𝑐𝑜𝑠𝜃𝑠𝑖𝑛𝜃
✓ 𝑠 2 =𝑠𝑖𝑛2 𝜃
Therefore, the global stiffness matrix and equation is :

𝑓1𝑥 𝑐𝑜𝑠 2 𝜃 𝑐𝑜𝑠𝜃𝑠𝑖𝑛𝜃 −𝑐𝑜𝑠 2 𝜃 −𝑐𝑜𝑠𝜃𝑠𝑖𝑛𝜃 𝑢1

𝑓1𝑦 𝐸 𝐴 𝑠𝑖𝑛2 𝜃 −𝑠𝑖𝑛2 𝜃 𝑣1
= * 𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃 −𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃
𝑢2
𝑓2𝑥 𝐿 −𝑐𝑜𝑠 2 𝜃 −𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃 𝑐𝑜𝑠 2 𝜃 𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃
𝑓2𝑦 −𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃 −𝑠𝑖𝑛2 𝜃 𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜃 𝑠𝑖𝑛2 𝜃 𝑣2

Global stiffness matrix

Global nodal displacement

Global nodal force for for each element
each element
 Example

For the truss structures shown:

y
2 (1) 3

(2) 0.4 KN
1 x

Given
✓ Element (1) & (2) are made from aluminum (Al)
✓ Element (3) made from steel (st)
✓ 𝜃=30°
✓ 𝐴𝑎𝑙 =200 𝑚𝑚2 , 𝐴𝑠𝑡 =100 𝑚𝑚2
✓ 𝐸𝑎𝑙 =69𝐾𝑁/ 𝑚𝑚2 , 𝐸𝑠𝑡 =207𝐾𝑁/ 𝑚𝑚2
✓ 𝐿(1) =260mm, 𝐿(2) =150mm, 𝐿(3)=300mm
Determine

1. The global stiffness matrix(assembled)

2. Displacement of joint(node) 2 and 3 [global displacement]
3. Stress, strain and internal forces in each member
4. Local nodal displacement for each element (Exercise)
 NEXT

CHAPTER FOUR
FINITE ELEMENT FORMULATION FOR BEAM