Beruflich Dokumente
Kultur Dokumente
METHODS
COURSE FILE
GEETHANJALI COLLEGE OF ENGINEERING AND TECHNOLOGY
Branch: ME Version No : 01
Distribution List :
2) Sign : 2) Sign :
2) Sign : 2) Sign :
3) Design : 3) Design :
4) Date : 4) Date :
2) Sign : 3) Date :
1. Contents
1.1. Cover page
1.2. contents
1.3. Syllabus copy
1.4. Vision of the department
1.5. Mission of the department
1.6. PEOs and POs
1.7. Course objectives and outcomes
1.8. Brief note on the importance of the course
1.9. Prerequisites
1.10. Instructional Learning outcomes
2.0 Course mapping with POs
2.1 Class Time Table
2.2 Individual Time Table
2.3 Lecture schedule with methodology being used
2.4 Detailed Notes
2.5 Additional Topics
2.6 University Question Papers
2.7 Question Bank
One dimensional problems: Finite Element modeling coordinates and shape functions. Stiffness equations
for axial bar element using potential energy approach. Assembly of global stiffness matrix and load
vector. Finite Element equations. Quadratic shape functions.
UNIT II
Analysis of Truss: Stiffness equations for a truss bar element oriented in 2D plane Methods of assembly
Plane truss element Space truss element Finite element analysis of trusses
Analysis of Beams: Element stiffness matrix for two noded, two degree of freedom per node beam
element and simple problems.
UNIT III
2D-problems , Finite element modeling of two dimensional stress analysis with CST, and treatment of
boundary conditions. Estimation of load vector, Stresses.
Finite element modeling of Axi symmetric solids subjected to Axisymmetric loading with triangular
elements. Two dimensional four nodded isoparametric elements
UNIT IV
Steady state Heat transfer Analysis: One dimensional analysis of Slab, fin and two dimensional analysis of
thin plate. Analysis of uniform shaft subjected to torsion.
UNIT –V
Dynamic analysis: Formulation of finite model, Element mass matrices, equations of Eigen values and
Eigen vectors for a stepped bar, truss.
Finite element Formulation of 3D problems in stress analysis, convergence requirements, mesh
generation, techniques such as semi automatic and fully Automatic use of software’s such as
ANSYS,NISA,NASTRAN
1.3 Vision of the Department:
To develop a world class program with excellence in teaching, learning and research
that would lead to growth, innovation and recognition
CO1) To understand the theory of elasticity including strain/displacement and Hooke’s law
relationships.
CO2) To analyze solid mechanics problems using classical methods and energy methods;
CO4) To solve for stresses and deflections of beams under unsymmetrical loading;
CO5) To analyze the maximum and minimum principal stresses using analytical and graphical
(mohr’s circle)methods.
CO7) To Understand the fundamental concepts of stress and strain and the relationship between both
through the strain-stress equations in order to solve problems for simple tridimensional elastic solids
Calculate and represent the stress diagrams in bars and simple structures.
CO8) To apply various failure criteria for general stress states at points
COURSE DESCRIPTION:
Finite Element Analysis (FEA) is a tool used for the evaluation of structures and systems, providing an
accurate prediction of a component's response subjected to thermal and structural loads. Structural
analyses include all types of steady or cyclic loads, mechanical or thermal. Thermal analyses include
convection, conduction, and radiation heat transfer, as well as various thermal transients and thermal
shocks.
FEA is used to analyze complex geometries, whereas very simple ones (for example, a beam) can be
analyzed using hand calculations. For a structure subjected to a load condition (thermal, mechanical,
vibratory, etc.) its response (deflection, stress, etc.) can be predicted and measured against acceptable
defined limits. In the most simplest terms, this is a factor of safety, which is the ratio of the stress in a
component, to the allowable stress of the material. If a factor of safety is too small, the possibility of
failure becomes unacceptably large; on the other hand, if the factor is unnecessarily large, the result is a
uneconomical or nonfunctional design. For the majority of structural and machine applications, factors
of safety are specified by design specifications or codes written by committees of experienced
engineers, such as the American Institute of Steel Construction (design & construction of structural steel
for buildings) and the American Concrete Institute (building codes requirements for reinforced
concrete).
Pre-requisites
1.The trainees should have a basic knowledge of mathematics, engineering mechanics and mechanics of
solids.
2. It is assumed that the student has knowledge about basic calculus and differential equations.
3.It is also assumed that the student has some experience with Python (or is willing to learn)
Course Objectives of the Finite Element Methods
Course Goals
To learn basic principles and skills of finite element modeling and analysis.
To learn the theory and characteristics of finite elements that represent engineering structures.
To learn and apply finite element solutions to problems in mechanical engineering.
To develop the knowledge and skills needed to effectively evaluate finite element analyses
performed by others. *
Upon successful completion of this course, the student should be able to:
Students will learn advanced topics and techniques in finite element methods and how to implement and
apply these techniques to solve nonlinear systems of ordinary and partial differential equations.
Mapping on to Programme Educational Objectives and Programme Out Comes:
Course PO1 PO2 PO3 PO4 PO5 PO6 PO7 PO8 PO9 PO10 PO11
outcomes
CO1 √ √
CO2 √ √
CO3 √ √ √
CO4 √
CO5 √ √
CO6 √
CO7 √
CO8 √
CO9 √
CO10 √
k Graduates can participate and succeed in all competitive examinations and interviews.
PEO 1 Our graduates will apply their knowledge and skills to succeed in a computer
engineering career and/or obtain an advanced degree.
PEO 2 Our graduates will apply basic principles and practices of computing grounded
in mathematics and science to successfully complete hardware and/or software √
related engineering projects to meet customer business objectives and/or
productively engage in research.
PEO 3 Our graduates will function ethically and responsibly and will remain informed √
and involved as fully in their profession and in our society.
PEO 4 Our graduates will successfully function in multi-disciplinary teams.
PEO 5 Our graduates will communicate effectively both orally and in writing.
PEO2: Our graduates will apply basic principles and practices of computing grounded in
mathematics and science to successfully complete hardware and/or software related engineering projects
to meet customer business objectives and/or productively engage in research.
PEO3: Our graduates will function ethically and responsibly and will remain informed and involved
as fully in their profession and in our society.
PEO5: Our graduates will communicate effectively both orally and in writing.
Outcomes:
a. Graduates will demonstrate knowledge of mathematics, science and engineering applications
b. Graduates will demonstrate ability to identify, formulate and solve engineering problems
c. Graduates will demonstrate an ability to analyse, design, develop and execute the programs
efficiently and effectively
d. Graduates will demonstrate an ability to design a system, software products and components as per
requirements and specifications
f. Graduates will demonstrate working in groups and possess project management skills to develop
software projects.
i. Graduates will show the understanding of impact of engineering solutions on society and also be
aware of contemporary issues like global waste management , global warming technologies etc.
j. Graduates will develop confidence for self education and ability for life long learning.
k. Graduates can participate and succeed in all competitive examinations and interviews.
Teaching/Learning Methodology
A mixture of lectures, tutorial exercises, and case studies are used to deliver the various topics. Some of
these topics are covered in a problem-based format to enhance learning objectives. Others will be
covered through directed study in order to enhance the students’ ability of “learning to learn.” Some
case studies are used to integrate these topics and thereby demonstrate to students how the various
techniques are inter-related and how they can be applied to real problems in an industry.
LECTURE SCHEDULE:
Total Teaching aids Re
Sl Unit Reg./
No. of Topics to be covered Additional
used LCD. ma
No No. Periods OHP.BB rks
1 I 16 Introduction to FEM Regular OHP,BB
2 Basic concepts, historical background, Regular OHP,BB
3 application of FEM General description, Regular OHP,BB
4 comparison of FEM with other methods Basic Regular OHP,BB
equations of elasticity
5 Stress strain relation.
6 Strain displacement relations Regular BB
7 Rayleigh-Ritz method Regular
8 Weighted residual method Regular BB
9 One dimensional problems Regular OHP,BB
10 Stiffness equations for axial bar element using Regular BB
potential energy approach.
11 Virtual energy principle, Finite element analysis of Regular OHP,BB
uniform bar
12 Finite element analysis of stepped bar and tapered Regular BB
bar subjected to mechanical and thermal loads
13 Assembly of global stiffness matrix and load Regular BB
vector
14 Quadratic shape functions, properties of stiffness Regular OHP,BB
matrix.
15 II 14 Stiffness equations for a truss bar element oriented Regular BB
in 2D plane
16 Finite element analysis of trusses Regular
17 Plane truss element Regular OHP,BB
18 Space truss element Regular OHP,BB
19 Methods of assembly Regular BB
20 problems Regular BB
21 Analysis of beams Regular OHP,BB
22 Hermite shape functions Regular BB
23 Element stiffness matrix Regular
24 Load vector Regular
25 problems Regular OHP,BB
26 III 16 2D-problems , CST, stiffness matrix and load Regular
vector
27 Iso parametric representation , shape functions, Regular OHP,BB
convergence requirements, problems
28 Two dimensional four noded isoparametric Regular
elements
29 Numerical integration Regular
30 Finite element modeling of Axisymmetric solids Regular OHP,BB
subjected to Axisymmetric loading with triangular
elements
31 3-D problems Regular LCD,OHP,BB
32 Tetrahedran element Regular OHP,BB
33 IV 16 Scalar field problems Regular BB
Micro Plan:
DETAILED NOTES:
Constant Strain Triangle
u ( x, y ) 1 2 x 3 y
v( x, y ) 1 2 x 3 y
where the ' s and 's are constants. These constants can be related to nodal displacements
for the triangular element:
Assume the
corners of the
triangle
(nodes) are
numbered
CCW, and have
coordinates
( x1 , y1 ) ,
etc. as shown.
At each node
(i=1,2,3),
assume the
nodal
displacements
are given by
(ui , vi ) .
We can now
write 6
"boundary
conditions" as
follows:
For u(x,y):
At node 1: u1 u ( x1 , y1 ) 1 2 x1 3 y1
At node 2: u2 u ( x2 , y2 ) 1 2 x2 3 y2
At node 3: u3 u ( x3 , y3 ) 1 2 x3 3 y3
For v(x,y):
At node 1: v1 v( x1 , y1 ) 1 2 x1 3 y1
At node 2: v2 v( x2 , y2 ) 1 2 x2 3 y2
At node 3: v3 v( x3 , y3 ) 1 2 x3 3 y3
We can now solve for the constants in terms of nodal displacements. Eqs. (0.2) can be written in
matrix form as
1 x1 y1 1 u1
1 x
2 y2 2 u2
1 x3 y3 3 u3
Solution is:
a1 x2 y3 x3 y2 , a2 x3 y1 x1 y3 , a3 x1 y2 x2 y1
b1 y2 y3 , b2 y3 y1, b3 y1 y2
c1 x3 x2 , c 2 x1 x3 , c3 x2 x1
(0.6)
and
1 x1 y1
2 A 1 x2 y2 2( area of triangle)
1 x3 y3
Note that the a's, b's and c's are constants and depend only upon the nodal coordinates (x,y) of the 3
corner nodes.
1
N i ( x, y ) (ai bi x ci y )
2A
A similar result is obtained for v(x,y):
3
v( x, y ) Ni vi
i 1
The quantitiesNi ( x, y ) are called shape functions. Note that the same shape functions apply
for both u ( x, y ) and v ( x, y ) .
We can now obtain the strains by substituting displacement functions (0.8) and (0.10) into strain
expressions Error! Reference source not found. to obtain:
u 3 Ni 3
b
xx ui i ui
x i 1 x i 1 2 A
v 3 N i 3
ci
yy vi vi
y i 1 y i 1 2 A
u v 3 Ni 3
Ni 3
ci 3
bi
xy ui vi ui vi
y x i 1 y i 1 x i 1 2 A i 1 2 A
(0.11)
The last 3 equations for strains can be put into matrix notation as:
u1
v
1
xx b1 0 b2 0 b3 0
1 u2
yy
0 c1 0 c2 0 c3
v2
2 A c b c b2 c3 b3
xy 1 1 2 u3
v3
(0.12)
Or, more compactly as (for any element "e"):
{ e } [ B e ]{q e }
b1 0 b2 0 b3 0
1
where
e
[B ] 0 c1 0 c2 0 c3
2A
c1 b1 c2 b2 c3 b3
u1
v
1
e u2
and {q }
v2
u3
v3
Since the terms in [ B e ] are constant for an element, the strains { e } are constant within an
element; hence the name "constant strain triangle" or CST.
We can now evaluate the internal strain energy U. Substituting (0.13) into
Error! Reference source not found. gives:
Ue 1 {q e }T [ B e ]T [ D e ][ B e ]{q e }dV
2 V
= 12 {q e }T [ B e ]T [ D e ][ B e ]dV {q e }
V
(0.16)
The quantity in parentheses can be identified as the element stiffness matrix [k e ] and (0.16) can
be written as:
U e 12 {q e }T [k e ]{q e }
[k e ] t e Ae [ B e ]T [ D e ][ B e ]
Note that the element stiffness matrix [k e ] is a 6x6 matrix, i.e., we have a 6 degree-of-freedom
(dof) element.
Note that the general form for the strain energy (0.17) can be written in index notation also:
6 6
e e T e e e e e
U 12 {q } [k ]{q } 1
2 ij qi q j
k
i 1 j 1
(0.20)
The stiffness matrix for the CST defined by (0.19) can be written in sub-matrix notation as:
Dij are material properties ( E , ) defined by Error! Reference source not found. and
where the
the b and c are geometry parameters (x-y coordinates of nodes) defined by (0.6).
i i
e L
V [u ( s ) p cos v( s ) p sin ]tds
0
6
V {q } {F } Fie qie
e e T e
i 1
1 pt e L cos
2
e
1 pt L sin
2
e
1 pt L cos
e
where {F } 2
1 pt e L sin
2
0
0
The matrix {F} represents the equivalent generalized nodal force vector due to pressure load on
boundary 1-2, i.e., we have replaced the pressure p on boundary 1-2 by the nodal forces {F} at nodes 1
and 2.
Note that the total force due to p on boundary 1-2 is (ptL) and divides equally between nodes 1 and 2.
will be a reaction force on the element node. Call these reaction forces {S}.
S4
S3 S2
2 S1
S6 1
S5
3
Si reactions from adjacent elemen
Error! Reference source not found. and noting that U e and V e are functions of nodal
displacements qie , i 1,..., 6 , we have
6
(U e V e )
(U e V e ) 0 qie
i 1 qie
Since qi 0 , then
(U e V e )
0 for i 1, 2,...,6
qie
Substituting U (0.20) and V [(0.23) and (0.25)] into (0.27) gives the equilibrium equation for any
element.
[k e ]{qe } {F e } {S e }
N el N el 6 6 N el
U str U e
e T e
1 {q } [ k ]{q }
2
1e
e
2 ij i j
e1 i 1 j 1
e e
k q q
e1 e 1
(0.29)
and
N el N el N el
Vstr V e {q e }T {F e } {q e }T {S e }
e1 e1
e 1
N el 6 N el 6
e e e e
Fi qi Si qi
e1 i 1 e1 i 1
(0.30)
The principle of minimum potential energy for the structure requires that
M
(U str Vstr )
(U str Vstr ) 0 qi
i 1 qi
(0.31)
{q} contains the M degrees of freedom for the structure (NOT dof for each element). For
where
N el N el N el
1 {q
2
e T e e
} [k ]{q } {q } {F } e T e
e T
{q } {S }
e
for i 1, 2,..., M
(0.33)
Problem: The energy terms for each element are in terms of the element dof, but in order to obtain
the equations of equilibrium for the structure (above equation), we have to take the partial derivatives
with respect to the global structural dof. In order to complete the above, the element degrees of
freedom {q e } must be written in terms of the M global structural degrees of freedom {q}. For
any element, we can write a transformation between element local and global dof (called the local-
global transformation):
{q e } [T e ] {q}
(6 x1) (6 xM ) ( Mx1)
The transformation will be nothing more then 1's and 0's. As an example, suppose we have the
following element and structural node numbering:
q2 q4 q6
1 2 3 4 1 q1 2 q3 3 q
1 5 9 q10
5 2 6 6 7 10 8 p 5 q9 6 7
3 7 11 q18
4 8 12 q17
y y
9 10 11 12 9 10 11
Consider element 7. Suppose we place element node 1 at global node 6.
x x
q2e q6e q12 q14
q1e 3 q5e q11 7 q13
1 6
7 q22
7 q4e
q3e q21
2 11
Element nodes Structural nodes
and local dofs and global dofs
We see that for element 7, there is a correspondence between the 6 element local dofs at element
nodes 1, 2 and 3, and the 6 structural global dofs at nodes 6, 11and 7. We see that local (element)
node 1 corresponds to global node 6, local (element) node 2 corresponds to global node 11, and local
node 3 corresponds to global node 7. We can write this local to global transformation
{q e } [T e ]{q} as:
global node #
1 2 3 4 5 6 7 8 9 10 11 12
1 0
[0] [0] [0] [0] [0] 0 1 [0] [0] [0] [0] [0] [0
e 7
q1,2
7 1 0
q3,4 [0] [0] [0] [0] [0] [0] [0] [0] [0] [0] [0
7 0 1
q
5,6 1 0
[0] [0] [0] [0] [0] [0] 0 1 [0] [0] [0] [0] [0
Each [0] is a (2x2). The above says that for element 7, local (element) node 1 corresponds to global
node 6, i.e., local dofs 1,2 correspond to global dofs 11,12; local node 2 corresponds to global node 11,
i.e., local dofs 3,4 correspond to global dofs 21,22, etc.
Now transform U e and V e from local to global dof by substituting (0.34) into (0.29) and (0.30) to
obtain
Now we can define the following element matrices in global dof (instead of local element dof):
[ K ge ] [T e ]T [k e ] [T e ]
( MxM ) ( Mx 6) (6 x 6) (6 xM )
{Fge } [T e ]T {F e }
( Mx1) ( Mx 6) (6 x1)
{S ge } [T e ]T {S e }
To see what an element stiffness and force matrix written in global dof looks like, consider element 7
1
2
6 7 7 7
[ K g7 ] k11 k13 k12 {Fg7 }
7 7 7 7
k31 k33 k32
8
10
11 7 7 7
k21 k23 k22
12
N el N el N el
M M
U str U e
1 {q}
2
T
[k ge ]{q} e
12 k gij qi q j
e 1 i 1 j 1
e1 e 1
(0.37)
N el N el N el
Vstr V e {q}T {Fge } {q}T {S ge }
e1 e 1
e1
N el M N el M
Fgie qi e
S gi qi
e 1 i 1 e1 i 1
(0.38)
for i 1, 2,..., M
(0.39)
N el N el N el
[k ge ]{q} {Fge } {S ge } {0}
e 1 e1 e1
(0.40)
or
N el e N el
e
N el
e
[k g ] {q} {Fg } {S g }
e1 e1 e 1
When all the element contributions have been summed, we simply write
A single element by itself is useless. We must determine the equilibrium equations for an assemblage
of elements that comprise the entire structures.
Consider the following structure (only a few elements are taken to simplify the discussion) with a
uniformly pressure p on the right boundary and fixed on the left boundary (assume a constant
thickness t).
We note that the left side is fixed (nodes 1, 5 and 9). Hence, displacement boundary conditions will
require that
For each element, we can construct a table called the element connectivity that specifies which
structural (global) nodes are connected by an element. Hence, for the problem above, we have the
following element connectivity table:
1 1 5 2
2 5 6 2
3 5 10 6
4 5 9 10
5 2 6 3
6 6 7 3
7 6 11 7
8 6 10 11
9 3 7 4
10 7 8 4
11 7 12 8
12 7 11 12
Note that for the CST, element nodes MUST be given as CCW. Element node 1 can be attached with
any global node of the element.
Note that if we are careful in numbering the nodes and choosing the element connectivity in a
"systematic" manner, there will be a pattern to the element connectivity table (see above). An
automatic mesh generator, like the one in FEMAP, tries to follow this pattern.
Note that the global node numbers for the structure are somewhat arbitrary, i.e., we could number
them in any order. However, it will turn out that there are optimum ways to number nodes (for a
given structure and mesh) in order to reduce the bandwidth of the structural stiffness matrix [K] - this
saves time solving the equations. For the mesh above, it would be optimum to number downward and
left-to-right, as opposed to left-to-right and downward. We'll discuss that later. Likewise, the element
numbering is arbitrary, but again there may be optimum approaches. An automatic mesh generator
tries to do the numbering in an optimum fashion.
Note that for this structure, we have 12 global nodes. There are 2 degrees of freedom (dof) at each
node (u and v). Hence, the structure has 24 dof and the structural stiffness matrix [K] will be (24x24).
The structural equilibrium equations can be written as:
[ K ] {q} {Q} {S }
(24 x 24) (24 x1) (24 x1)
Lets see how each element contributes to global matrices. Take element 1 to start with. Note that we
can use sub-matrix notation to divide the element matrices as following. Use a superscript of 1 on the
k terms to indicate element 1.
k111 1 1 F11
k12 k13
(2 x 2) (2 x1)
1 1
[k ] k21 k122
1 1
k23 , 1
{F } F2
(6 x 6) 1 1 1 1
k
31 k 32 k33 F3
We now look at element 1 and note that element node numbers 1, 2, 3 correspond to global node
numbers 1, 5, 2 (from the drawing of the mesh, or from the element connectivity table). We can
indicate this information on the stiffness and force matrices as follows:
1 5 2
k111 1 1 F11
k12 k13
1 (2 x 2) 1 (2 x1)
1
1 1 1
1
[k ] 5 k21 k22 k23 , 1
{F } 5 F2
(6 x 6) 1 1 1 (6 x1) 1
k
2 31 k32 k33 2 F3
Hence, we see that element 1 contributes stiffness and forces to global nodes 1, 5 and 2. Placing these
contributions into the global stiffness matrix gives:
Element 1 only
K 1 2 3 4 5 6 7 8 9 10 11 12 {q}
1 1 1 1
k11 k13 k12 q1,2
2 1 1 1
k31 k33 k32 q3,4
3
q5,6
4
q7,8
5 1 1
k21 k23 k122 q9,10
6
10
11
6 11 7
7
6 k11 7
k12 7
k13 F17
7 7 7 7
11 21 k k22 k23 F
2
7 7 7 7
7 k31 k32 k33
F3
12
q23,24
** remember, each block is a (2x2) sub-matrix
Element 7 only
K 1 2 3 4 5 6 7 8 9 10 11 12 {q}
1
q1,2
2
q3,4
3
q5,6
4
q7,8
5
q9,10
6
10
11
12
q23,24
** remember, each block is a (2x2) sub-matrix
Note that the distributed pressure load p is applied only to the right boundary of elements 10 and 11.
Hence {F} for all elements except 10 and 11 will be zero. For elements 10 and 11, we will have
0 0
0
7 0 7
1 ptL48 1 ptL
2 2 812
{F 10 } 8
11
{F } 12
0 0
1 ptL48 1 ptL812
2 4 2 8
0 0
K 1 2 3 4 5 6 7 8 9 10 11 12 {q}
1 X X X
q1,2
2 X X X X X
q3,4
3 X X X X X
q5,6
4 X X X X
q7,8
5 X X X X X X
q9,10
6 X X X X X X X
7 X X X X X X X
8 X X X X
9 X X X
10 X X X X X
11 X X X X X
12 X X X X
q23,24
X means that one or more elements have contributed here
1 2 5
9 10 11
[ K 22 ] [k33 ] [k33 ] [k11 ]
each sub-matrix is (2x2)
The global node 2-6 coupling term [ K 26 ] will have contributions from elements 2 and 5 since
only these elements share the boundary between nodes 2 and 6:
2 5
[ K 26 ] [ k32 ] [ k12 ].
Global node 6 will have stiffness contributions from elements 2, 3, 5, 6, 7, and 8:
2 3 5 6 7 8
[ K 66 ] [k22 ] [k33 ] [k22 ] [k11 ] [k11 ] [k11 ].
Question? What happened to the reactions {S} for each element? Why don't they show up in the
structural stiffness matrix?
Simple. It is equilibrium. Recall that when we make a free-body, in this case take a single finite
element as the free-body, we will have equal and opposite reactions where the cut is made though the
body. Consider elements 1 and 2 below:
S41
S31 S12 S 42
2 S11 S32
1 1 1 2
2 S6 1 S62 S22
2
S51 1
3 3 S52
At the boundary between elements 1 and 2, the reactions are equal and opposite. Hence, we add
OK, but what about the boundary where there are supports? What happens to the reactions there?
For example, the cantilever plate example above:
R2 1 2 3 4
They don't disappear and should
be included in the structural
1 R1 2 3 4
stiffness matrix. R10 5 11 2 6
5
5 6 7
9
9 10 p
5 R9 2 6 6 7 10 8
8 p
We know that there will be
R18 4 33 8
7
7 12
11
11
unknown reactions at global yy 4 8 12
nodes 1, 5 and 9. We could call 9 10 11 12
9 R17 10 11 12
xx
these reactions R1, R2 , R9 , R10 , R17 and R18 (consistent with global displacements).
So we have the free body of the structure:
K 1 2 3 4 5 6 7 8 9 10 11 12 {q}
1 X X X
q1,2
2 X X X X X
q3,4
3 X X X X X
q5,6
4 X X X X
q7,8
5 X X X X X X
q9,10
6 X X X X X X X
7 X X X X X X X
8 X X X X
9 X X X
10 X X X X X
11 X X X X X
12 X X X X
q23,24
OK, now one last step. We have to apply displacement boundary conditions. The structure is fixed at
1. Zero out the row and column on the left side matrix (the [K] matrix) corresponding to each B.C.,
and zero out the row of the right side (the {Q} matrix) corresponding to each B.C.
2. Place a 1 on the diagonal of the left side matrix (the [K] matrix) corresponding to each B.C.
You will notice that every dof that has a B.C. also corresponds to a dof where a support reaction (R)
occurs. Applying B.C. as described above will thus eliminate the reactions from the equilibrium
equations.
A theoretical reason why we don’t have to worry about reactions in structural equations of
equilibrium? Because these support reactions R do no work (displacement is zero at support) and
hence do not affect equilibrium of the structure!!!
K 1 2 3 4 5 6 7 8 9 10 11 12 {q}
1 1 0 0 0 0
1
0 0
0 0
q1,2
2 0 0 X X 0 0 X
0 0
0 0
q3,4
3 X X X X X
q5,6
4 X X X X
q7,8
5 0 0 0 0 1 0 0 0 0 0 0
0 0
0 0
1
0 0
0 0
0 0
q9,10
6 X X 0 0 X X X X
0 0
7 X X X X X X X
8 X X X X
9 0 0 1 0 0
0 0 1 0 0
10 0 0 X 0 0 X X
0 0 0 0
11 X X X X X
12 X X X X
q23,24
(1) q1 0 , etc.
Now we are ready to solve for the element strains and stresses. For each element, we can substitute
the 6 global displacements corresponding to that element into (0.13):
The stresses for each element can then be obtained by substituting the strains for that element into
Error! Reference source not found.:
Evaluation of stress results based on stress components in the Cartesian coordinates directions
may not necessarily represent the largest stresses and we need these in order to consider yielding or
failure. You already know that you can calculate principal stresses and maximum shear stress using
stress transformation equations or Mohr's Circle. Hence, stress results (stress components) are often
represented in two additional ways:
Principal stresses can, as noted above, be obtained by either stress transformation equations or
through the use of Mohr's Circle. An alternate approach to define principal stresses is to write:
xx p xy xy
yx yy p yz 0
zx zy zz p
Expansion of the determinant provides a cubic equation that can be solved for the three principal
stresses p . Comparing principal stresses to a tensile yield stress provides some measure of
evaluation; however, one has to keep in mind that comparing the principal stress (obtained from a
three-dimensional stress state) to a yield stress obtained from a uniaxial tension test is risky at best.
The von Mises stress provides a means to extrapolate uniaxial tensile test data (for yield stress) to a
three-dimensional stress state. In effect, the von Mises stress provides an "equivalent" uniaxial stress
approximation to the three-dimensional stress state in a body through the following equation:
1
( xx yy ) 2 ( yy zz ) 2 ( zz xx ) 22
VM 12
6 xy 6 yz 6 zx
(0.45)
or
VM 1 ( p1 p 2 )2 ( p 2 p 3 ) 2 ( p 3 p1 ) 2
2
(0.46)
This representation has been used quite successfully to model the onset of yielding in ductile metals
and collaborates well with experiment. It is widely used in industry. For a material to remain elastic,
VM y (for no yielding)
Equation (0.46) forms an ellipsoid in 3-D (ellipse in 2-D) when the stresses are plotted in principal
stress space. As long as the stress state represented by the principal stresses is inside the ellipse (the
yield surface), the material is elastic.
Element Libraries
(or, choose the right element for a structural component and loading, in order to maximize potential
for correct results with the least amount of computation)
Many, many finite elements have been developed for use in modern FEM software. Choosing the
correct element for a particular structural is paramount. For example,
if a structural member behaves like a beam in bending, we should choose a beam element to
model it,
if a structural member behaves like a thin plate in plane stress, we should choose an
appropriate element to model it,
if a structural member looks like a shell of revolution, we should use a thin shell of revolution
element,
if a structural member will experience a three-dimensional stress state, we have to choose an
element that models that behavior,
etc.
Here are some examples of the types of elements available:
Plane stress requires that the only non-zero stresses occur in the plane of the element
(however, strain does occur normal to plane). Generally applicable to thin geometries.
Two displacement dof per node (NO rotational dof).
Plane strain requires that the only non-zero strains occur in the plane of the element (strain
is zero normal to plane, but stress is not zero). Long constrained geometries (for example, a
long pipe, a dam). Elements with curved boundaries will always have 3 or more nodes per
edge.
Plate and shell bending elements (bending and in-plane stresses; flat and curved elements)
Triangular, Quad (both straight and curved sides)
8. Subject Contents
8.1. Synopsis page for each period (62 pages)
8.2. Detailed Lecture notes containing:
a. Ppts
b. Ohp slides
c. Subjective type questions (approximately 5 t0 8 in no)
d. Objective type questions (approximately 20 to 30 in no)
e. Any simulations
8.3. Course Review (By the concerned Faculty):
(i)Aims (ii) Sample check (iii) End of the course report by the concerned faculty
GUIDELINES:
Distribution of periods:
No. of classes required to cover JNTU syllabus : 60
No. of classes required to cover Additional topics : Nil
No. of classes required to cover Assignment tests (for every 2 units 1 test) : 4
No. of classes required to cover tutorials : 2
No. of classes required to cover Mid tests : 2
No of classes required to solve University Question papers : 2
-------
Total periods 70
UNIVERSITY QUESTIONS
1. Derive strain-displacement relations for a 3-D elastic body.
2. (a)What are the merits and the demerits of Finite Element Methods?
9) Using finite element method to calculate displacements and stresses of the bar shown in fig.
10) For the stepped bar shown in figure, determine the nodal displacements, element stresses and
Support reactions. Take P=300kN, Q=500 kN, E=2x1011N/m2. A1=250mm2, A2=500mm2, A=1000
mm2
11) Determine the displacements and the support reactions for the uniform bar shown in
Fig.1. GivenP=300KN
12) Determine the nodal displacements, element stresses and support reactions for the bar as shown
in fig.
13) . Determine the stiffness matrix, stresses and support reactions for the truss structure as
shown in fig.
14) . Calculate the nodal displacements, stresses and support reactions for the truss shown in fig.
QUIZ QUESTIONS
ANSWERS
1. b 2. A 3. C 4. D 5. C 6. C 7. B 8. B 9. B 10. c
II Fill in the blanks
11. Finite element.
12. Nodes
13. Ritz method or Ray-Leigh Ritz method
14. Aspect ratio
15. Shape functions
16. Primary boundary condition
17. [ ] = 1/L[−1 1]
lB 18. Pin 20) Transverse or rotaion
ANSWERS
1. B 2. B 3. D 4. A 5. A 6. C 7. B 8. A 9. D 10. C
II Fill in the blanks
11. [ D ]
12. Heat transfer
13. Sub-parametric
2
14. w/m K
15. 4x4
16. N=[ N ,N ]
1 2
17. 2x2
18. Isotropic
19. N =A /A, N2=A /A, N3=A /A
1 1 2 3
QUESTION BANK
UNIT-1
1. Using finite element method to calculate displacements and stresses of the bar shown in fig.
2. For the stepped bar shown in figure, determine the nodal displacements, element stresses and
Support reactions. Take P=300kN, Q=500 kN, E=2x1011N/m2. A1=250mm2, A2=500mm2,
A=1000 mm2
3. Determine the displacements and the support reactions for the uniform bar shown in
Fig.1. GivenP=300KN
4. Determine the nodal displacements, element stresses and support reactions for the bar as shown
in fig.
UNIT-2
1. Determine the local and global stiffness matrices of a truss element.
2. Determine the stiffness matrix, stresses and support reactions for the truss structure as shown in fig.
3. Taking advantage of symmetry , determine the joint displacements and axial forces in the truss
shown in fig. All members have the same cross sectional area of the same material, A=0.0001m2 and
E=200Gpa, the load P=20KN. The dimensions in meters shown in fig.
4. Calculate the nodal displacements, stresses and support reactions for the truss shown in fig.
UNIT-3
1. Explain with neat mathematical steps to derive beam stiffness
matrix.
2. For a beam and loading shown in fig below determine the slopes at nodes 2 and 3 and
vertical deflection at the midpoint of the distributed load.
3. Why the Hermite shape functions are considered for the beam element? Explain the Hermite
shape functions for a two nodded beam element. And also derive the strain displacement
relation matrix.
4. A beam of 4m length is subjected to point loads at the distances of 2 m and 4 m from the
fixed end of 10KN and 20KN respectively. Calculate the deflection at the center of the beam,
if E= 2.1×104N/m2 and A=450mm2as shown in fig.
Calculate the maximum deflection and slope by using finite element method for the simply supported
beam of length L, Young’s modulus E and the moment of Inertia I, subjected to a point load of P at the
centre.
Unit-4
1. The nodal coordinates of a triangular element are 1(1,3), 2(5,3) and 3(4,6). At a point p
inside the element, the x-coordinates is 3.3 and the shape function N = 0.3. Determine
1
the shape functions and y-coordinates of the point P.
2. Obtain the load vector for following CST element.
3. Determine the Jacobian matrix for the triangular element with the coordinates 1(1.5,2),
2(7,3.5) and 3(4.5, 9.2). And also calculate the area of a triangle.
4. Determine the strain displacement relation matrix for CST.
5. Calculate the strain displacement matrix for the element with the coordinates 1(4,5),
2(9,2) and 3(6,8). And also calculate the strains of the triangle whose nodal displacement
values are u =0.3 mm, v =0.3 mm, u = 0.2 mm, v = -0.4 mm, u = 0.3 mm, v = 0.5mm.
1 1 2 2 3 3
Unit-5
x 2
1. Evaluate ∫ [3e + x + 1 / (x + 2)] dx over the limits -1 and +1 using one point, two point Gauss
quadrature formula. Compare with exact solution.
2 3 4
2. Evaluate the integral ∫ (a + a x + a x + a x + a x ) dx with the limits between - 1 to
0 1 2 3 4
+1 using one point and two point Gaussian quadrature.
3. Derive the shape functions for a four nodded iso- parametric quadrilateral element.
1. Derive one dimensional steady state heat conduction equation and apply to one dimensional fin
problem.
2. Derive one dimensional steady state heat conduction equation and derive the conductivity
Matrix.
3. A uniform aluminum circular fin of diameter 3cm is extruded from the surface whose
temperature is 1000C. The convection takes place from the lateral surface and tip of the fin.
Assuming K=30W/m K, h=1200W/m2K and T∞=30 0C Determine the temperature distribution in
the fin.
4. Composite wall consisting of three materials is shown in Figure below. The outer temperature is
T0 = 200C. Convective heat transfer takes place on the inner surface of the wall with T∞=800 0C
and h=25 W/m2.0C. Determine the temperature distribution in the wall.
6. Using finite element method to calculate displacements and stresses of the bar shown in fig.
7. For the stepped bar shown in figure, determine the nodal displacements, element stresses and
Support reactions. Take P=300kN, Q=500 kN, E=2x1011N/m2. A1=250mm2, A2=500mm2,
A=1000 mm2
8. Determine the displacements and the support reactions for the uniform bar shown in
Fig.1. GivenP=300KN
9. Determine the nodal displacements, element stresses and support reactions for the bar as shown
in fig.
UNIT-2
5. Determine the local and global stiffness matrices of a truss element.
6. Determine the stiffness matrix, stresses and support reactions for the truss structure as shown in fig.
7. Taking advantage of symmetry , determine the joint displacements and axial forces in the truss
shown in fig. All members have the same cross sectional area of the same material, A=0.0001m2 and
E=200Gpa, the load P=20KN. The dimensions in meters shown in fig.
8. Calculate the nodal displacements, stresses and support reactions for the truss shown in fig.
Calculate the maximum deflection and slope by using finite element method for the simply supported
beam of length L, Young’s modulus E and the moment of Inertia I, subjected to a point load of P at the
centre.
Unit-3
6. The nodal coordinates of a triangular element are 1(1,3), 2(5,3) and 3(4,6). At a point p
inside the element, the x-coordinates is 3.3 and the shape function N = 0.3. Determine
1
the shape functions and y-coordinates of the point P.
7. Obtain the load vector for following CST element.
8. Determine the Jacobian matrix for the triangular element with the coordinates 1(1.5,2),
2(7,3.5) and 3(4.5, 9.2). And also calculate the area of a triangle.
9. Determine the strain displacement relation matrix for CST.
10. Calculate the strain displacement matrix for the element with the coordinates 1(4,5),
2(9,2) and 3(6,8). And also calculate the strains of the triangle whose nodal displacement
values are u =0.3 mm, v =0.3 mm, u = 0.2 mm, v = -0.4 mm, u = 0.3 mm, v = 0.5mm.
1 1 2 2 3 3
x 2
4. Evaluate ∫ [3e + x + 1 / (x + 2)] dx over the limits -1 and +1 using one point, two point Gauss
quadrature formula. Compare with exact solution.
2 3 4
5. Evaluate the integral ∫ (a + a x + a x + a x + a x ) dx with the limits between - 1 to
0 1 2 3 4
+1 using one point and two point Gaussian quadrature.
6. Derive the shape functions for a four nodded iso- parametric quadrilateral element.
Unit-5
1) Derive one dimensional steady state heat conduction equation and apply to one dimensional fin
problem.
2) Derive one dimensional steady state heat conduction equation and derive the conductivity
Matrix.
3) A uniform aluminium circular fin of diameter 3cm is extruded from the surface whose
temperature is 1000C. The convection takes place from the lateral surface and tip of the fin.
Assuming K=30W/m K, h=1200W/m2K and T∞=300C Determine the temperature distribution in the
fin.
4) composite wall consisting of three materials is shown in Figure below. The outer
temperature is T0 = 200C. Convective heat transfer takes place on the inner surface of the wall
with T∞=8000C and h=25 W/m2.0C. Determine the temperature distribution in the wall.
Sources of Information
2. nptel.ac.in/video.php?subjectId=112106135
3) www.youtube.com/watch?v=NYiZQszx9cQ
The course aims to provide deeper knowledge, a wider scope and improved understanding of the study
of motion and the basic principles of mechanics and strength of materials. It is a concept based subject
and it needs the application capabilities of the concepts on the part of the students.
EVALUATION SCHEME:
PARTICULAR WEIGHTAGE MARKS
End Examinations 75% 75
Three Sessionals 20% 20
Assignment 5% 5
TEACHER'S ASSESSMENT(TA)* WEIGHTAGE MARKS
*TA will be based on the Assignments given, Unit test Performances and Attendance in the class for a
particular student.
STUDENT LIST
4 13R11A0364 ANNAPAREDDY
SATYANARAYANA REDDY
5 13R11A0365 BAIRU RAVIKIRAN REDDY
17 13R11A0377 G BIKSHAPATHI
18 13R11A0378 G ELLENDER