WELCOME

CLASS 05
STEPS OF FINITE ELEMENT METHOD (FEM)
 ELEMENT EQUATIONS
• Use of either Minimum Potential Energy Method and Weighted Residual Method will lead to equations
describing the behavior of an element, which are commonly expressed as;

𝑘 𝑞 = 𝑄
Where, 𝒌 = 𝐄𝐥𝐞𝐦𝐞𝐧𝐭 𝐩𝐫𝐨𝐩𝐞𝐫𝐭𝐲 𝐦𝐚𝐭𝐫𝐢𝐱, 𝒒 = vector of unknown at the element nodes, and 𝑸 = vector of
element nodal forcing parameters.
• Equation above is expressed in a general sense;

• For the specific problem of stress analysis, where 𝑘 = stiffness matrix, 𝑞 = vector of nodal displacement and
𝑄 = vector of nodal forces.
 STEP 4: DERIVE ELEMENT EQUATIONS
• By involving available laws and principles, we obtain equations governing the behavior of the element. The
equations are obtained in general terms and hence can be used for all elements in the discretized body. Two most
common methods are used to develop element equations:
a. The Energy Method b. The Residual Method

ENERGY METHODS
• This procedures are based on the idea of finding consistent states of bodies associated with stationary values of a
scalar quantity assumed by the loaded bodies.
𝐹
• In engineering, usually this quantity is a measure of energy or work 𝐹𝑥

Neutral
Maximum

Minimum

𝑥
 STEP 4: DERIVE ELEMENT EQUATIONS

STATIONARY VALUE
• In simple word the term stationary can imply a maximum,
minimum of saddle (a seat for a rider) point of a function 𝐹𝑥 .

• Under certain conditions, the function may simply assume a 𝐹

𝐹𝑥
minimum or a maximum value.

• To find the point of a stationary value, we equate the Maximum

Neutral
derivative of F is zero:
𝒅𝑭
=𝟎
𝒅𝒙
Minimum

𝑥
 STEP 4: DERIVE ELEMENT EQUATIONS
POTENTIAL ENERGY

• In case of stress-deformation analysis, the function F is often represented by one of the energy functions
stated previously.

• A simple column under the support condition, shown in this slide, is linear and elastic, and if it is in
equilibrium, it can be shown that the column will assume minimum potential energy. To comply with the
commonly used notation we denote energy by the symbol 𝜫𝒑 .

P
What is energy?
A
• The ability to do work or cause change.
• Work is the application of a force through a distance.
E
• Force can put matter into motion or stop it if it is already moving.
• Motion is a change in position of an object with time.
• To do work, energy is needed.
 STEP 4: DERIVE ELEMENT EQUATIONS
POTENTIAL ENERGY

• The potential energy is defined as the sum of the internal strain energy U and the potential of the external

loads W, the latter term denotes the capacity of load P to perform work through a deformation 𝝂 of the
column.

• Strain energy (U) is energy stored when a force is applied to a deformable object, causing it to change
shape. Strain energy can be calculated in terms of the force applied and the change in length

• In physics, potential energy is the energy held by an object because of its position relative to other objects,
stresses within itself.
Π𝑝 = U + 𝑊𝑝

• For an elastic solid in equilibrium, the expression U = W is a statement of the principle of real work: the
work of external forces on an elastic solid is completely stored as strain energy within the solid.
 STEP 4: DERIVE ELEMENT EQUATIONS
POTENTIAL ENERGY

Π𝑝 = U + 𝑊𝑝

• When we apply the principle of minimum potential energy, we essentially take the
derivative of Π𝑝 and equate it to zero. We assume that the load remains constant
while taking the derivative; then
• 𝛿Π𝑝 = 𝛿U + 𝛿𝑊𝑝 =0;
 STEP 4: DERIVE ELEMENT EQUATIONS

• 𝑰𝒕 𝒄𝒂𝒏 𝒃𝒆 𝒔𝒉𝒐𝒘𝒏 𝒕𝒉𝒂𝒕, 𝛿 U = - 𝛿𝑊𝑝 ,

POTENTIAL ENERGY 𝛿 2 Π𝑝 = 𝛿 2 U - 𝛿 2 𝑊𝑝 > 0

𝛿Π𝑝
• = 0; • For example: Π𝑝 = Π𝑝 (𝑢1 , 𝑢2 , 𝑛3 ……………………………. 𝑢𝑚 ……….. 𝑢𝑛 )
𝛿𝑢1
𝛿Π𝑝
• = 0;
𝛿𝑢2
𝛿Π𝑝 • Where, 𝑢1 , 𝑢2 , 𝑛3 ……………………………. 𝑢𝑚 ……….. 𝑢𝑛 un are the total number of
• =0 unknowns (at the node). Where, n is the total number of unknown
𝛿𝑢3
…………………………
𝛿Π𝑝
• =0
𝛿𝑢𝑛
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS

EXACT SOLUTION
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS - PROCESS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS – SOLUTION EQUATION
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS – REARRANGING EXACT SOLUTION
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS - REARRANGING SOLUTIONS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS

𝟒𝑴𝟎 𝑳𝟐 𝝅𝒙
𝒀 = − 𝟑 𝐬𝐢𝐧
𝝅 𝑬𝑰 𝑳
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
• Figure below shows a schematic representation of Equation above. The shaded area in Fig denote error between
the approximate and exact solution 𝑢∗ − 𝑢 over the domain D. The residual R(x) over D is related to the error
𝑢∗ − 𝑢 and the sum or the integral of R(x) over D is minimized. R

R 𝐸𝑟𝑟𝑜𝑟: 𝑢∗ − 𝑢
𝑅 𝑥 = 𝐿𝑢 − 𝑓
u

𝑢∗
𝑅 𝑥 𝑊𝑖 𝑥 𝑑𝑥
𝐷 D
D
x
x
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS

• One of the two major alternatives for formulating the finite element method is the
method of weighted residual (MWR).

• A number of patterns are employed under the MWR, among which are collocation,
subdomain, least squares and Galerkin’s method.

• For many problems with certain mathematical characteristics, Galerkin’s method

yields results identical to those from variation procedures and is closely related to
them.
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
• Galerkin’s method has been the most commonly used residual method for finite element applications.

• The MWR is based on minimization of the residual left after an approximate or trial solution is
substituted into the differential equations governing a problem.

• As a simple illustration, let us consider the following differential equation:

𝑑 2 𝑢∗ 𝑑𝑢∗
2
− = 𝑥 1a
𝑑𝑥 𝑑𝑡

Where, 𝑢 ∗ is the unknown, 𝒙 is the coordinate, t is the time, and f(x) is the forcing function.
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
• In mathematical notation, previous equation is written as,

𝑑2 𝑑
• 𝐿𝑢∗ = 𝑓 where, 𝐿 = 𝑑𝑥 2
− 𝑑𝑡 is the differential operator. 1b

• We are seeking an approximate solution to previous equation and denote an approximate or trial function u for
𝒖∗ as
𝑛 2
• 𝑢 = 𝜑0 + 𝑖=1 𝛼𝑖 𝜑𝑖 = 𝜑0 + 𝛼1 𝜑1 + 𝛼2 𝜑2 + 𝛼3 𝜑3 + ⋯ … … … … … … … … … … … + 𝛼𝑛 𝜑𝑛

Here, 𝜑1 , 𝜑2 , 𝜑3 , … … … … . 𝜑𝑛 are known functions chosen in such a way as to satisfy the homogeneous boundary
conditions; 𝝋𝟎 is chosen to satisfy the essential, geometric or forced boundary conditions; and 𝛼𝑖 are parameters or
constants to be determined. Categories of boundary conditions are explained subsequently and discussed in
Chapter 3. Often, for convenience, u in above equation is written as
𝑛

𝑢= 𝛼𝑖 𝜑𝑖 3
𝑖=1
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
• In which 𝛼1 = 1 𝑎𝑛𝑑 𝜑1 = 𝜑0 . If the approximate solution u is substituted into equation 10 than we are left with
a residual

𝑅 𝑥 = 𝐿𝑢 − 𝑓 4
(which is zero if u= 𝑢 ∗ )

• In the method of weighted residuals, the aim is to find an approximate solution u for 𝒖∗ such that the residual
R(x) in above equation is made as small as possible or is minimized. In other words the error between the
approximate solution 𝒖∗ and the exact solution u is minimized. A number of schemes are possible to achieve the
aim of minimization of R(x). For the sake of completeness only brief and general statements for these methods
are given below:

• Mathematically the idea of minimization can be expressed as

𝐷
𝑅 𝑥 𝑊𝑖 𝑥 𝑑𝑥 = 0; 𝑖 = 1,2, … … … … … … . . , 𝑛. 5
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS

• Where D denotes the domain of a structure or body under consideration. For the one-dimensional column

problem, the domain is simply the linear extend of the column. In above equation, 𝑊𝑖 denote weighting
functions. Various residual schemes such as collocation, subdomain, and Galerkin use different weighting
functions. For instance, in the case of the collocation method, 𝑊𝑖 =1. The Expression in above equation implies
that the weighted value of 𝑅 𝑥 over the domain of a structure vanishes.
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS

• As a simple illustration, let us consider the following second-order differential equation that governs the
problems of one-dimensional stress deformation in a column.

𝑑2 𝑣 ∗
• 𝑐 𝑑𝑦 2 =𝑓 6

• Where 𝑣 ∗ is the unknown (deformation), y is the coordinate axis, c denotes a material property, and f is the
forcing function. For a column, c=EA, f is the applied load, E is the modulus of elasticity and A is the cross-
sectional area. Assume that EA=1 and f=10 units of load; then above equation becomes

𝑑2 𝑣 ∗
• = 10 7
𝑑𝑦 2
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS

• An approximate solution for 𝑣 ∗ can be written as

• 𝑣 = 𝛼1 + 𝛼2 𝑦 + 𝛼3 𝑦 2 = 𝛼𝑖 𝜑𝑖 8

• Where, 𝜑1 =1, 𝜑2 =y, 𝜑3 =𝑦 2 ; This function should be chosen so as to satisfy the boundary conditions of the
problem. Then the residual R(y) is given by,

𝑑2 𝑣
• 𝑅 𝑦 =𝑑𝑦 2 − 10 9a
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS
𝐿
• According to the MWR, leads to 0
𝑅 𝑦 𝑊𝑖 𝑦 𝑑𝑦 = 0; 𝑖 = 1, 2, 3

𝐿 𝑑2 𝑣
or 0 𝑑𝑦 2
− 10 𝑊𝑖 𝑦 𝑑𝑦 = 0; 9b

𝐿 𝑑2 (𝛼𝑖 𝜑𝑖 )
or 0 𝑑𝑦 2
− 10 𝑊𝑖 𝑦 𝑑𝑦 = 0; 9c
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS

• Here, L is the length of the column which represents the domain D. Now, we can substitute for the second
derivative in above equation by differentiating 𝜗 in Eq. 8. The final result will yield three simultaneous equations
in 𝛼1 , 𝛼2 and 𝛼3 as
𝐿
𝑑 2 (𝛼𝑖 𝜑𝑖 )
− 10 𝑊1 𝑦 𝑑𝑦 = 0;
0 𝑑𝑦 2

𝐿
𝑑 2 (𝛼𝑖 𝜑𝑖 )
𝑊2 𝑦 𝑑𝑦 = 0;
0 𝑑𝑦 2

𝐿
𝑑 2 (𝛼𝑖 𝜑𝑖 )
− 10 𝑊3 𝑦 𝑑𝑦 = 0;
0 𝑑𝑦 2
 STEP 4: DERIVE ELEMENT EQUATIONS
METHOD OF WEIGHTED RESIDUALS

• Which are solved for 𝛼1 , 𝛼2 and 𝛼3 . When these values of𝛼1 , 𝛼2 and 𝛼3 are substituted into Eq. 8 we obtain the
approximate solution for 𝑣 ∗ .

ELEMENT EQUATIONS
• Use of either Minimum Potential Energy Method and Weighted Residual Method will lead to equations describing the
behavior of an element, which are commonly expressed as;

𝑘 𝑞 = 𝑄
Where, 𝑘 = Element property matrix, 𝑞 = vector of unknown at the element nodes, and 𝑄 = vector of element
nodal forcing parameters.

Equation above is expressed in a general sense; for the specific problem of stress analysis, where 𝑘 = stiffness matrix,
𝑞 = vector of nodal displacement and 𝑄 = vector of nodal forces.