University of the Philippines Diliman

College of Engineering
1
Institute of Civil Engineering

CE 28: Analytical and Computational

Methods in Civil Engineering III
Lecture 9: Introduction to Finite Element
Method
Prepared by:
Dominic M. Bautista
Instructor
Outline of Presentation 2

 Element Matrix for a Structural Axial Element

• Systems with horizontal element
• 2-D Truss system

 Element Matrix for Differential Equations

• Interpolating function
• Weak form of PDEs
• Weighted Residual Method
Finite Element Method 3

Derivation of Element Matrix for a Structural Axial Element

Example configurations of an axial element

Equations that govern the system (from mechanics):

𝑭 𝒆 𝒆 𝒆
𝒖𝟐 − 𝒖𝟏
𝑭𝟏 𝒆 + 𝑭𝟐 𝒆 = 𝟎 𝝈𝒆 = 𝒆 𝝈 𝒆 = 𝑬 𝒆 𝜺𝒆 𝒆
𝜺 =
𝑨 𝑳
Finite Element Method 4

Derivation of Element Matrix for a Structural Axial Element

 called the element stiffness matrix, Ke, that relates the element
internal force matrix, Fe, to the element nodal displacements, ue

From the definition of stress, 𝑭𝟏 𝒆 = 𝑨𝒆 𝝈𝒆

Hooke’s Law, 𝑭𝟏 𝒆 = 𝑨𝒆 𝑬𝒆 𝜺𝒆
𝒖 𝒆−𝒖 𝒆
𝟏 𝟐
From the definition of strain, 𝒆
𝑭𝟏 = 𝑨 𝑬𝒆 𝒆
𝑳𝒆
𝒖 𝒆−𝒖 𝒆
From equilibrium, 𝒆 𝒆
𝑭𝟐 = −𝑭𝟏 = 𝑨 𝑬 𝒆 𝒆 𝟐 𝟏
𝑳𝒆
Finite Element Method 5

Derivation of Element Matrix for a Structural Axial Element

𝒆−𝒖 𝒆 𝒖 𝒆−𝒖 𝒆
𝒖𝟏 𝟐 𝟐 𝟏
𝑭𝟏 𝒆 𝒆
=𝑨 𝑬 𝒆 𝑭𝟐 𝒆 𝒆
=𝑨 𝑬 𝒆
𝑳𝒆 𝑳𝒆

In matrix form, 𝑭𝟏 𝒆 𝑨𝒆 𝑬𝒆 𝟏 −𝟏 𝒖𝟏 𝒆
𝒆 =
𝑭𝟐 𝑳𝒆 −𝟏 𝟏 𝒖𝟐 𝒆

𝑭𝟏 𝒆 𝒖𝟏 𝒆 𝑨𝒆 𝑬𝒆 𝟏 −𝟏
𝑭𝒆 = 𝒅𝒆 = 𝑲𝒆 = 𝒆
𝑭𝟐 𝒆 𝒖𝟐 𝒆 𝑳 −𝟏 𝟏

element internal element nodal element stiffness

force vector displacements vector matrix
Finite Element Method 6

Example 9.1: Derive the element stiffness

matrix for each material given the following
properties:
Aal = 1500 mm2 Eal = 70 GPa
Ast = 900 mm2 Est = 200 GPa

𝟏𝟕𝟓 −𝟏𝟕𝟓
𝑲𝒂𝒍 = kN/mm
−𝟏𝟕𝟓 𝟏𝟕𝟓
𝟑𝟔𝟎 −𝟑𝟔𝟎
𝑲𝒔𝒕 = kN/mm
−𝟑𝟔𝟎 𝟑𝟔𝟎
 Finite Element Method 7

Derivation of Element Stiffness Matrix of an Inclined Axial Element

element internal element nodal
force vector displacements
vector
𝑭𝟏𝒙 𝒆 𝒖𝟏𝒙 𝒆
𝑭𝟏𝒚 𝒆 𝒖𝟏𝒚 𝒆
𝑭𝒆 = 𝒅𝒆 =
𝑭𝟐𝒙 𝒆 𝒖𝟐𝒙 𝒆
𝑭𝟐𝒚 𝒆 𝒖𝟐𝒚 𝒆

Rotation of coordinate axis will be

done such that the local coordinate
axis will be the axis of the element
Finite Element Method 8

Derivation of Element Stiffness Matrix of an Inclined Axial Element

Relationship of the
displacements at node 1

In matrix form (with node 2),

𝒖𝟏𝒙 𝒍𝒆 𝒄𝒐𝒔φ𝒆 𝒔𝒊𝒏φ𝒆 𝒖𝟏𝒙 𝒆 In compact form,
𝟎 𝟎
𝒖𝟏𝒚 𝒍𝒆 −𝒔𝒊𝒏φ𝒆 𝒄𝒐𝒔φ𝒆 𝟎 𝟎 𝒖𝟏𝒚 𝒆 𝒅𝒍𝒆 = 𝑹𝒆 𝒅𝒆
=
𝒖𝟐𝒙 𝒍𝒆 𝟎 𝟎 𝒄𝒐𝒔φ𝒆 𝒔𝒊𝒏φ𝒆 𝒖𝟐𝒙 𝒆
[Re] is the rotation
𝒖𝟐𝒚 𝒍𝒆 𝟎 𝟎 −𝒔𝒊𝒏φ𝒆 𝒄𝒐𝒔φ𝒆 𝒖𝟐𝒚 𝒆
matrix
 Finite Element Method 9

Derivation of Element Stiffness Matrix of an Inclined Axial Element

𝒄𝒐𝒔φ𝒆 𝒔𝒊𝒏φ𝒆 𝟎 𝟎 NOTE: Re is an orthogonal
−𝒔𝒊𝒏φ𝒆 𝒄𝒐𝒔φ𝒆 𝟎 𝟎 matrix
𝑹𝒆 =
𝟎 𝟎 𝒄𝒐𝒔φ𝒆 𝒔𝒊𝒏φ𝒆
𝑹𝒆 −𝟏
= 𝑹𝒆 𝑻
𝟎 𝟎 −𝒔𝒊𝒏φ𝒆 𝒄𝒐𝒔φ𝒆
𝑹𝒆 𝑻 𝒅𝒍𝒆 = 𝑹𝒆 𝑻 𝑹𝒆 𝒅𝒆 = 𝒅𝒆
Similar to the nodal displacement matrix,

𝑭𝒍𝒆 = 𝑹𝒆 𝑭𝒆 𝑭𝒆 = 𝑹𝒆 𝑻
𝑭𝒍𝒆 Transformed element
matrix,
𝑭𝒆 = 𝑹𝒆 𝑻
𝑲𝒍𝒆 𝒅𝒍𝒆
𝑲𝒆 = 𝑹𝒆 𝑻
𝑲𝒍𝒆 𝑹𝒆
𝑭𝒆 = 𝑹𝒆 𝑻
𝑲𝒍𝒆 𝑹𝒆 𝒅𝒆
Finite Element Method 10

Derivation of Element Stiffness Matrix of an Inclined Axial Element

𝒄𝟐 𝒄𝒔 −𝒄𝟐 −𝒄𝒔
𝟐 𝟐
𝑲𝒆 = 𝒄𝒔 𝒔 −𝒄𝒔 −𝒔
−𝒄𝟐 −𝒄𝒔 𝒄𝟐 𝒄𝒔
−𝒄𝒔 −𝒔𝟐 𝒄𝒔 𝒔𝟐

NOTES: 𝒄 = 𝒄𝒐𝒔φ𝒆 𝒔 = 𝒔𝒊𝒏φ𝒆

angle of the element axis measured

φ𝒆 =
counterclockwise from the +x-axis
Finite Element Method 11

Example 9.2: Derive the element stiffness

matrix for each axial member shown.
Finite Element Method 12

Element Equations for Differential Equations

Strong Form Weak Form

Element
Equations
Interpolating/Shape functions
 Finite Element Method 13

Interpolating Functions
 Functions, usually polynomials, that estimates the variation of a
function between points (sometimes called shape functions)

In FEM we express the approximate value of the solution on the

nodes using interpolating/shape functions in the form:
𝒏

ෝ = ෍ 𝑵𝒊 𝒖𝒊
𝒖
𝒊=𝟏

where 𝑢ො = approximate solution Ni = interpolating/shape function at node i

n = number of nodes ui = value of solution at node i
 Finite Element Method 14

For 1-D elements 𝒏

ෝ = ෍ 𝑵𝒊 𝒖𝒊 = 𝑵𝒊 𝒙 𝒖𝒊 + 𝑵𝒋 (𝒙)𝒖𝒋
𝒖
+x node i node j
𝒊=𝟏

xi L
In matrix 𝒖𝒊
ෝ = 𝑵𝒊
𝒖 𝑵𝒋
xj form, 𝒖𝒋

NOTE: xj - xi = L
Objective: at x = xi, u = ui Ni = 1, Nj = 0
and at x = xj, u = uj Ni = 0, Nj = 1
Linear model:
𝑵(𝒙) = 𝒄𝟏 + 𝒄𝟐 𝒙 Quadratic model: 𝑵 𝒙 = 𝒄𝟏 + 𝒄𝟐 𝒙 + 𝒄𝟑 𝒙𝟐
Finite Element Method 15

Example 9.3: Derive the interpolating functions for a 1-D

element using a linear model

+x node i node j Linear model: 𝑵(𝒙) = 𝒄𝟏 + 𝒄𝟐 𝒙

xi L at x = xi, u = ui Ni = 1, Nj = 0
xj at x = xj, u = uj Ni = 0, Nj = 1
NOTE: xj - xi = L

𝒙𝒋 − 𝒙 𝒙 − 𝒙𝒊
𝑵𝒊 = 𝑵𝒋 =
𝑳 𝑳
 Finite Element Method 16

For 2-D elements ෝ = 𝑵𝒊 𝒙, 𝒚 𝒖𝒊 + 𝑵𝒋 𝒙, 𝒚 𝒖𝒋 + 𝑵𝒌 (𝒙, 𝒚)𝒖𝒌

𝒖
node i 𝒖𝒊
(xi,yi) node k In matrix ෝ = 𝑵𝒊
𝒖 𝑵𝒋 𝑵𝒌 𝒖𝒋
(xk,yk) form, 𝒖𝒌
Objectives:
node j
(xj,yj) at x = xi and y = yi, u = ui Ni = 1, Nj = 0, Nk = 0
+y
at x = xj and y = yj, u = uj Ni = 0, Nj = 1, Nk = 0
+x
at x = xk and y = yk, u = uk Ni = 0, Nj = 0, Nk = 1

Linear model: 𝑵 𝒙 = 𝒄𝟏 + 𝒄𝟐 𝒙 + 𝒄𝟑 𝒚
 Finite Element Method 17

For 2-D elements Interpolating function:

𝟏
node i 𝑵𝒎 𝒆
𝒙, 𝒚 = 𝒂𝒎 + 𝒃 𝒎 𝒙 + 𝒄 𝒎 𝒚 m = i,j,k
𝟐𝑨 𝒆
(xi,yi) node k
where:
(xk,yk) 𝒂𝒊 = 𝒙𝒋 𝒚𝒌 − 𝒙𝒌 𝒚𝒋 𝒃𝒊 = 𝒚𝒋 − 𝒚𝒌 𝒄𝒊 = 𝒙𝒌 − 𝒙𝒋
𝒂𝒋 = 𝒙𝒌 𝒚𝒊 − 𝒙𝒊 𝒚𝒌 𝒃𝒋 = 𝒚𝒌 − 𝒚𝒊 𝒄𝒋 = 𝒙𝒊 − 𝒙𝒌
node j 𝒂𝒌 = 𝒙𝒊 𝒚𝒋 − 𝒙𝒋 𝒚𝒊 𝒃𝒌 = 𝒚𝒊 − 𝒚𝒋 𝒄𝒌 = 𝒙𝒋 − 𝒙𝒊
(xj,yj)
+y
+x 𝟏
𝟏 𝟏 𝟏
𝑨𝒆 = 𝒙𝒊 𝒙𝒋 𝒙𝒌
𝟐 𝒚 𝒚𝒋 𝒚𝒌
𝒊
Finite Element Method 18

Strong Form of a PDE

 consists of the governing equations and the boundary
conditions for a physical system

Weak Form of a PDE

 an integral form of the governing equations that are used to
derive the element equations

NOTE: The strong and weak form are equivalent equations but
expressed in different forms
Finite Element Method 19

Consider an axially loaded elastic bar (STRONG form)

The governing equation for this
analysis is the 2nd order ODE:

with boundary conditions:

Natural BC
where: u(x) = displacement function Essential BC
A(x) = cross-sectional area
𝑢ത = known displacement
b(x) = body force per unit length
𝑡=ҧ known stress
E = modulus of elasticity
Finite Element Method 20

Strong Form to Weak Form

We multiply an arbitrary weight function, w(x), to the governing ODE and

integrate over the entire domain
𝒍 𝒍
𝒅 𝒅𝒖
න 𝒘 𝑨𝑬 𝒅𝒙 + න 𝒘𝒃𝒅𝒙 = 𝟎
𝟎 𝒅𝒙 𝒅𝒙 𝟎

Applying integration by parts on the first integral,

𝒍 𝒍 𝒍
𝒅 𝒅𝒖 𝒅𝒘 𝒅𝒖
න 𝒘𝑨𝑬 − න 𝑨𝑬 𝒅𝒙 + න 𝒘𝒃𝒅𝒙 = 𝟎
𝟎 𝒅𝒙 𝒅𝒙 𝟎 𝒅𝒙 𝒅𝒙 𝟎
 Finite Element Method 21

Strong Form to Weak Form

Simplifying the first integral,
𝒍 𝒍
𝒅𝒖 𝒅𝒖 𝒅𝒘 𝒅𝒖
𝒘𝑨𝑬 − 𝒘𝑨𝑬 − න 𝑨𝑬 𝒅𝒙 + න 𝒘𝒃𝒅𝒙 = 𝟎
𝒅𝒙 𝒙=𝒍
𝒅𝒙 𝒙=𝟎 𝟎 𝒅𝒙 𝒅𝒙 𝟎

For essential boundaries, the value of the weight function is zero,

𝒍 𝒍
𝒅𝒖 𝒅𝒘 𝒅𝒖
− 𝒘𝑨𝑬 − න 𝑨𝑬 𝒅𝒙 + න 𝒘𝒃𝒅𝒙 = 𝟎
𝒅𝒙 𝒙=𝟎 𝟎 𝒅𝒙 𝒅𝒙 𝟎

Applying the first boundary condition,

𝒍 𝒍
𝒅𝒘 𝒅𝒖
𝒘𝑨𝒕ҧ 𝒙=𝟎 − න 𝑨𝑬 𝒅𝒙 + න 𝒘𝒃𝒅𝒙 = 𝟎
𝟎 𝒅𝒙 𝒅𝒙 𝟎
 Finite Element Method 22

Strong Form

𝒍 𝒍
𝒅𝒘 𝒅𝒖
Weak Form 𝒘𝑨𝒕ҧ 𝒙=𝟎 − න 𝑨𝑬 𝒅𝒙 + න 𝒘𝒃𝒅𝒙 = 𝟎
𝟎 𝒅𝒙 𝒅𝒙 𝟎

NOTE:
Due to the reduction of the order of the derivative of u, the solutions to the
weak form need not to be as “smooth” as the solutions of the strong form
(there is a reduction to the continuity requirements of the solution)

If the boundary condition at x = 0 is Direchlet, 𝒍

𝒅𝒘 𝒅𝒖 𝒍
− න 𝑨𝑬 𝒅𝒙 + න 𝒘𝒃𝒅𝒙 = 𝟎
𝟎 𝒅𝒙 𝒅𝒙 𝟎
Finite Element Method 23

Weak form of the Poisson Equation

𝝏𝟐 𝒖 𝝏𝟐 𝒖 With arbitrary boundary
+ 𝟐 = 𝒇(𝒙) or 𝛁𝟐 𝒖 = 𝒇(𝒙)
𝝏𝒙 𝟐 𝝏𝒚 conditions over its domain, D

We multiply an arbitrary weight function, w(x,y), to the governing ODE and

integrate over the entire domain

න 𝒘𝛁𝟐 𝒖 𝒅𝑫 = න 𝒘𝒇 𝒅𝑫
𝑫 𝑫

Applying integration by parts on the first integral,

න 𝛁 𝒘𝛁𝒖 𝒅𝑫 − න 𝛁𝒘 𝛁𝒖 𝒅𝑫 = න 𝒘𝒇 𝒅𝑫
𝑫 𝑫 𝑫
Finite Element Method 24

Weak form of the Poisson Equation

න 𝛁 𝒘𝛁𝒖 𝒅𝑫 − න 𝛁𝒘 𝛁𝒖 𝒅𝑫 = න 𝒘𝒇 𝒅𝑫
𝑫 𝑫 𝑫

Applying Green’s Theorem on the first integral,

ර ෝ 𝒅𝑩 − න 𝛁𝒘 𝛁𝒖 𝒅𝑫 = න 𝒘𝒇 𝒅𝑫
𝒘𝛁𝒖 ° 𝒏
𝑩 𝑫 𝑫

This is the weak form of the Poisson Equation

For Laplace Equation, If the P.E. has Direchlet boundary conditions,

ර ෝ 𝒅𝑩 − න 𝛁𝒘 𝛁𝒖 𝒅𝑫 = 𝟎
𝒘𝛁𝒖 ° 𝒏 − න 𝛁𝒘 𝛁𝒖 𝒅𝑫 = න 𝒘𝒇 𝒅𝑫
𝑩 𝑫 𝑫 𝑫
Finite Element Method 25

Derivation of Element Matrix Equation

 The shape functions will be combined to the weak form of the
PDE to derive the element matrix equations

Weighted Residual Method

 the error, called the residual, between the approximate solution
and the exact solution is minimized over the domain

ෝ−𝒖=ε
𝒖 න 𝒘ε 𝒅𝑫 = 𝟎
𝑫
Finite Element Method 26

Element Matrix for the Axially Loaded Bar with Direchlet B.C.s
𝒍 𝒍
𝒅𝒘 𝒅𝒖
Weak Form of the ODE: − න 𝑨𝑬 𝒅𝒙 + න 𝒘𝒃𝒅𝒙 = 𝟎
𝟎 𝒅𝒙 𝒅𝒙 𝟎

In the weighted residual method, the weight function is the interpolating

function and the solution is the linear combination of
𝒏

𝒘𝒊 = 𝑵𝒊 ෝ = ෍ 𝑵𝒊 𝒖𝒊 = 𝑵𝟏 𝒙 𝒖𝟏 + 𝑵𝟐 (𝒙)𝒖𝟐
𝒖
𝒊=𝟏
For each element, 𝒙𝟐
𝒅𝑵 𝒅𝑵 𝒙𝟐
න 𝑨𝑬 𝒖 𝒅𝒙 = න 𝑵 𝒃𝒅𝒙
𝒙𝟏 𝒅𝒙 𝒅𝒙 𝒙𝟏
Finite Element Method 27

Element Matrix for the Axially Loaded Bar with Direchlet B.C.s

𝒅𝑵 −𝟏Τ𝑳
For a 2-node element = 𝟏 𝟏
𝒅𝒙 𝟏Τ𝑳 𝒅𝑵 𝒅𝑵 𝟐
− 𝟐
= 𝑳 𝑳
𝒙𝟐 − 𝒙 𝒙 − 𝒙𝟏 𝟏 𝟏
𝑵 = 𝒅𝑵 −𝟏 𝟏 𝒅𝒙 𝒅𝒙
𝑳 𝑳 = − 𝟐
𝒅𝒙 𝑳 𝑳 𝑳 𝑳𝟐
Since {u} is the solution and has constant values and assuming A& E are
also constants,
𝟏 𝟏
𝟐
− 𝟐 𝒙𝟐 𝒙𝟐
𝑨𝑬 𝑳 𝑳 න 𝒅𝒙 𝒖 = න 𝑵 𝒃𝒅𝒙
𝟏 𝟏 𝒙𝟏 𝒙𝟏
− 𝟐
𝑳 𝑳𝟐
Finite Element Method 28

Element Matrix for the Axially Loaded Bar with Direchlet B.C.s
Simplifying and letting the integral on the right to be the forces acting on
the nodes of the element,

𝑨𝒆 𝑬𝒆 𝟏 −𝟏 𝒖𝟏 𝒆 𝑭𝟏 𝒆
𝒆 =
𝑳𝒆 −𝟏 𝟏 𝒖𝟐 𝑭𝟐 𝒆

Which is the same element matrix that we derived using the principles of
mechanics
 Finite Element Method 29

Element Matrix for the Laplace Equation with Triangular Elements

Weak Form of the PDE: ර ෝ 𝒅𝑩 − න 𝛁𝒘 𝛁𝒖 𝒅𝑫 = 𝟎

𝒘𝛁𝒖 ° 𝒏
𝑩 𝑫

𝒘𝒊 = 𝑵𝒊 ෝ = 𝑵𝟏 𝒙, 𝒚 𝒖𝟏 + 𝑵𝟐 𝒙, 𝒚 𝒖𝟐 + 𝑵𝟑 (𝒙, 𝒚)𝒖𝟑
𝒖

𝒆 𝟏
𝑵𝒎 𝒙, 𝒚 = 𝒆
𝒂𝒎 + 𝒃 𝒎 𝒙 + 𝒄 𝒎 𝒚 m = 1,2,3
𝟐𝑨
𝒂𝟏 = 𝒙𝟐 𝒚𝟑 − 𝒙𝟑 𝒚𝟐 𝒃𝟏 = 𝒚𝟐 − 𝒚𝟑 𝒄𝟏 = 𝒙𝟑 − 𝒙𝟐 𝟏 𝟏 𝟏
𝟏
𝒂𝟐 = 𝒙𝟑 𝒚𝟏 − 𝒙𝟏 𝒚𝟑 𝒃𝟐 = 𝒚𝟑 − 𝒚𝟏 𝒄𝟐 = 𝒙𝟏 − 𝒙𝟑 𝑨𝒆 = 𝒙𝟏 𝒙𝟐 𝒙𝟑
𝟐 𝒚 𝒚𝟐 𝒚𝟑
𝒂𝟑 = 𝒙𝟏 𝒚𝟐 − 𝒙𝟐 𝒚𝟏 𝒃𝟑 = 𝒚𝟏 − 𝒚𝟐 𝒄𝟑 = 𝒙𝟐 − 𝒙𝟏 𝟏
 Finite Element Method 30

Element Matrix for the Laplace Equation with Triangular Elements

ර ෝ 𝒅𝑩 − න 𝛁𝒘 𝛁𝒖 𝒅𝑫 = 𝟎
𝒘𝛁𝒖 ° 𝒏
𝑩 𝑫
𝝏𝑵𝒍
𝝏𝑵𝒎 𝝏𝑵𝒎 𝟏 𝟏 𝒃𝒍
𝛁𝒘 = = 𝒃𝒎 𝒄𝒎 𝝏𝒙
𝝏𝒙 𝝏𝒚 𝟐𝑨 𝒆 𝛁 𝒖 = 𝝏𝑵 𝒖 =
𝒍 𝟐𝑨𝒆 𝒄𝒍
𝝏𝒚 l = 1,2,3
m = 1,2,3
Substituting to the weak form, {u} is a constant vector

m = 1,2,3 𝟏
ෝ 𝒅𝑩
𝒃𝒎 𝒃𝒍 + 𝒄𝒎 𝒄𝒍 න 𝒅𝑫 𝒖 = ර 𝒘𝛁𝒖 ° 𝒏
l = 1,2,3 𝟒𝑨 𝟐
𝑫 𝑩
Finite Element Method 31

Element Matrix for the Laplace Equation with Triangular Elements

Assuming that the right hand side are the boundary fluxes vector, {Q}
𝟏
𝒃𝒎 𝒃𝒍 + 𝒄𝒎 𝒄𝒍 𝒖 = {𝑸} m = 1,2,3
𝟒𝑨
l = 1,2,3
Or in matrix form,
𝒃𝟏 𝒃𝟏 + 𝒄𝟏 𝒄𝟏 𝒃𝟏 𝒃𝟐 + 𝒄𝟏 𝒄𝟐 𝒃𝟏 𝒃𝟑 + 𝒄𝟏 𝒄𝟑 𝒖𝟏
𝟏
𝒃𝟐 𝒃𝟏 + 𝒄𝟐 𝒄𝟏 𝒃𝟐 𝒃𝟐 + 𝒄𝟐 𝒄𝟐 𝒃𝟐 𝒃𝟑 + 𝒄𝟐 𝒄𝟑 𝒖𝟐 = 𝑸
𝟒𝑨 𝒖𝟑
𝒃𝟑 𝒃𝟏 + 𝒄𝟑 𝒄𝟏 𝒃𝟑 𝒃𝟐 + 𝒄𝟑 𝒄𝟐 𝒃𝟑 𝒃𝟑 + 𝒄𝟑 𝒄𝟑
References 32

 Fish, J., and Belytschko, J. (2007). A First Course in Finite

Elements, 1st Ed, John Wiley & Sons, Ltd.

 Reddy, J., (1993). An Introduction to the Finite Element

Method, 2nd ed., McGraw-Hill, Inc.