Professional Documents
Culture Documents
Untitled
Untitled
2
Please read this disclaimer before proceeding:
This document is confidential and intended solely for the educational purpose of
RMK Group of Educational Institutions. If you have received this document
through email in error, please notify the system manager. This document
contains proprietary information and is intended only to the respective group /
learning community as intended. If you are not the addressee you should not
disseminate, distribute or copy through e-mail. Please notify the sender
immediately by e-mail if you have received this document by mistake and delete
this document from your system. If you are not the intended recipient you are
notified that disclosing, copying, distributing or taking any action in reliance on
the contents of this information is strictly prohibited.
3
ME8692 FINITE ELEMENT ANALYSIS
4
Table of Contents
2 Pre Requisites 06
3 Syllabus 07
4 Course outcomes 08
6 Lecture Plan 10
8 Lecture Notes 13
9 Assignments 45
10 Part A Q & A 47
11 Part B Qs 49
15 Assessment Schedule 55
5
I - Course Objectives
01 To introduce the concepts of Mathematical Modeling of
Engineering Problems.
02 To appreciate the use of FEM to a range of Engineering
Problems.
GE8292 Engineering
Mechanics (Semester 2)
CE8395 Strength of
Materials for Mechanical
Engineers (Semester -4)
5
III - Syllabus
UNIT I INTRODUCTION 9
Historical Background – Mathematical Modeling of field problems in Engineering
– Governing Equations – Discrete and continuous models – Boundary, Initial and
Eigen Value problems– Weighted Residual Methods – Variational Formulation of
Boundary Value Problems – Ritz Technique – Basic concepts of the Finite Element
Method.
TOTAL PERIODS : 45
7
IV - Course Outcomes
(COs)Course Outcome (CO) - Statement
CO. No. Knowledge
Level
Course Augmentation
Units I II III IV V
Assignement 09.03.2021 23.03.2021 10.04.2021 20.04.2021 07.05.2021
Quiz
8
VI - COs-POs Matrices
POs CO PO1 PO 2 PO 3 PO 4 PO 5 PO 6 PO 7 PO 8 PO 9 PO 10 PO 11 PO 12
COs
C305
.1 CO1 3 3 3
C305
.2 CO2 3 3 2
C305
.3 CO3 3 3 3
C305
.4 CO4 3 3 3
C305
.5 CO5 3 3 2
C305
.6 CO6 3 3 3
1: Slight (Low) 2: Moderate (Medium) 3: Substantial (High)
9
UNIT V: ISOPARAMETRIC FORMULATION
Natural co-ordinate systems – Isoperimetric elements – Shape functions for iso
parametric elements – One and two dimensions – Serendipity elements – Numerical
integration and application to plane stress problems - Matrix solution techniques –
Solutions Techniques to Dynamic problems – Introduction to Analysis Software.
LECTURE PLAN
Session Mode of
Topics to be covered Reference
No. Delivery*
1 Natural co-ordinate systems 1,2
2 Isoparametric elements 1,2
3 Shape functions for iso parametric elements 1,2
4 One and two dimensions 2,3
5 Serendipity elements Chalk and 2,3
6 Numerical integration and application to plane Talk/Online 1,2
stress problems Webinars
7 Matrix solution techniques 1,2
8 Solutions Techniques to Dynamic problems 1,2
9 Introduction to Analysis Software. 1,2
*Chalk and Talk, PPT / OHP, Quiz, Group Discussion, Student, Seminar,
Case Study, Field Study, Video (NPTEL/EDUSAT/ONLINE), Crossword
Puzzle, Tutorial Session
10
ACTIVITY BASED LEARNING
Across Down
2. If the number of nodes used for defining the geometry is same 1. Derivative of Field Variable
as the number of nodes used for defining the displacements, then 3. Name the element whose extremity is -1 and +1
24-04-2021, 12:40
ACTIVITY BASED LEARNING
Across Down
2. If the number of nodes used for defining the geometry is same 1. Derivative of Field Variable
as the number of nodes used for defining the displacements, then 3. Name the element whose extremity is -1 and +1
it is known as -----------element 4.Which type of numerical integration method is used to
8.Stiffness Matrix is always calculate the definite integrals?
9.Which coordinate system is used to defdine any point inside 5.The family of elements involving curved boundaries is known
the element by a set of dimensionless numbers? as _ Elements
11. Natural Coordinates are 6. Shape functions of the 2D elements can be developed by
12. If the number of nodes used for defining the geometry is less having nodes only on the boundaries for _ _ element
than number of nodes used for defining the displacements, then it 7. Higher order lagrangian elements contains ---------------nodes
is known as ----------element 10. Magnitudes of natural coordinates lies between
24-04-2021, 12:41
Natural Coordinates and Coordinate Transformation
The parent line indicating varying from A’ (−1) to B’’ (+1) is also shown in the
figure. An appropriate linear transformation between x and is given by
𝑥(𝜉) = ( 𝑥 1 +𝑥 2) + ( 𝑥 2+𝑥 1) 𝜉
2 2
1−𝜉 1−𝜉 2
=( ) 𝑥1 + ( )𝑥
⏟ 2 ⏟ 2
𝑁1 𝑁
= 𝑁1 𝑥1 + 𝑁2 𝑥2
𝑥
𝑥 = [𝑁1 𝑁2 ] {𝑥1 }
2
For a point P ’ = 0.5, correspondingly
𝑥 2 +𝑥 1
𝑥 = 𝑥1 +
4
𝑥 = 𝑥 1+3𝑥
4
2
𝑥(𝜉) = ∑𝑁𝑖 𝑥𝑖 = 𝑁1 𝑥1 + 𝑁2 𝑥2 + 𝑁3 𝑥3
Serendipity approach
N1 = (SF)( − 0)( − 1)
where the scale factor (SF) is determined from the requirement that N1 be unity at
point 1, i.e. ( = −1). We readily find that SF = 0.5. Thus the required shape
function is
N1 = 0.5( − 0) ( − 1)
On similar lines, we can directly write the other two shape functions as
N2 = 0.5( − 0)( + 1)
N3 = − ( + 1)( − 1)
In this way, we can also write down very easily the shape functions for any complex
finite element. Considering the example of quadratic coordinate transformation once
again, we get, the required transformation
If point 3 in the physical space x-y were at the middle of the line, i.e. x3 =
(0.5) (x1 + x2), then we observe from the above equation that the quadratic
transformation just obtained reduces to the simple linear transformation obtained
earlier. However, if point 3 is located away from the mid-point, say for example, x3
= x1 + (0.25)(x2 − x1), then we require that a quarter point in the real Cartesian
plane be transformed to the mid-point in the plane.
x() = 0.5( )( − 1)(0) + 0.5()( + 1)(4) − ( + 1)( − 1)(1) = (1 + )2
Lagrange's interpolation
x1, x2, …, xn and the corresponding function values at these data points as ƒ1, ƒ2, …,
ƒn, we can then write
ƒ (x) L1 ƒ1 + L2 ƒ2 + ⋯+ Ln ƒn
We observe that the Li given above are simply the required shape functions Ni. Thus
we can write the shape functions for the quadratic coordinate transformation above
as follows: Given the data points (1 = −1), (2 = 1) and (3 = 0) and the
corresponding function values x1, x2, and x3, we can write the required Lagrange
interpolation polynomials as
it is easy to see that the coordinate transformation would be simply a scale factor
and is given by
𝜉 = 2𝑥ℓ
𝜂 = 2𝑦
ℎ
For a general quadrilateral element, we have to obtain the transformation
(x,yf(, )
This transformation, which yields the coordinates of a point P (x, y) within the
element in terms of the nodal coordinates (just as we interpolate the
displacement/temperature field), can be written as
xP = Nixi, yP = Ni yi
where i = 1, 2, …, NNOEL which stands for the number of nodes per element (= 4
here). Considering that Ni must be unity at node i and zero at all other nodes, the
shape functions Ni can be obtained as
where the scale factor (SF) is to be chosen such that N1 is unity at node 1 and the
product of equation of lines 2 − 3 and 3 − 4 has been taken to ensure that N1 is
zero at nodes 2, 3 and 4.
1
𝑁1 = ( 4) (1 − 𝜉)(1 − 𝜂)
1
𝑁2 = ( 4 ) (1 − 𝜉)(1 − 𝜂)
1
𝑁3 = ( 4 ) (1 − 𝜉)(1 − 𝜂)
1
𝑁4 = ( 4 ) (1 − 𝜉)(1 − 𝜂)
While we have derived the shape functions by the Serendipity approach, the reader
is advised to obtain the same using Lagrange’s approach. Figure shows a plot of
these shape functions.
x = Σ Nixi = −N1 + N3
Thus we have
1
𝑥= [(1 + 𝜉)(1 + 𝜂) − (1 − 𝜉)(1 − 𝜂)]
4
5
𝑦 = [(1 − 𝜉)(1 + 𝜂) − (1 + 𝜉)(1 − 𝜂)]
4
x = (1/2), y = (5/2)
x = (1/2), y = −(5/2)
Thus the equation of the line in the (x-y) space is y = −5x. These two lines
are indicated in above Figure. We observe that the - axis lines, when transformed
into the physical (x-y) space, need not be parallel to the x and y axes. The
transformed lines need not even be orthogonal in the x-y space. We reiterate the
fact that our nodal d.o.f. (such as displacements) are along real, physical x-y axes,
and not along -.
For the four-noded element, with the above linear shape functions, we have
been able to transform a parent square element in (-) space into a general
quadrilateral child element in the physical ( x-y) space. If we need general curved
edge element, we require a higher order transformation and we will now discuss the
eight-noded element for this purpose.
Consider the general quadrilateral element shown in Figure below with eight
nodes.
The parent eight-noded element in (-) space has also been shown in the figure.
The attention of the reader is drawn to the convention followed in numbering the
nodes. The shape functions Ni can be obtained as
where the scale factor (SF) is to be chosen such that N5 is unity at node 5 and the
product of equation of lines 2−6−3, 3−7−4 and 4−8−1 has been taken to ensure
that N5 is zero at all other nodes.
𝑁5 = ( 1 ) (1 − 𝜉 2 )(1 − 𝜂)
2
𝑁6 = ( 1 ) (1 + 𝜉)(1 − 𝜂 2 )
2
𝑁7 = ( 1 ) (1 − 𝜉 2 )(1 + 𝜂)
2
𝑁8 = ( 1 ) (1 − 𝜉)(1 − 𝜂 2 )
2
We observe, for example, that N5 is quadratic in but linear in . Similarly, N8
is quadratic in but linear in .
Referring to the plot of N1 for the four-noded element, we can infer that we
can readily modify that shape function to obtain the shape function N1 for the
present eight- noded element by m1aking it vanish at nodes 5 and 8 also. Thus,
1 1
𝑁1 |8𝑁𝑜𝑑𝑒 = 𝑁1 |4𝑁𝑜𝑑𝑒 − 𝑁5 − 𝑁8
2 2
1 1
= (1 − 𝜉)(1 − 𝜂) − (1 − 𝜉)(1 − 𝜂)(2 + 𝜉 + 𝜂)
4 4
1
= (1 − 𝜉)(1 − 𝜂)(1 − 𝜉 − 𝜂)
4
On similar lines we can obtain the shape functions for other nodes as
1 1
𝑁2 |8𝑁𝑜𝑑𝑒 = 𝑁2 |4𝑁𝑜𝑑𝑒 − 2 𝑁5 − 2 𝑁6
1
= (1 + 𝜉)(1 − 𝜂)(−1 + 𝜉 − 𝜂)
4
1 1
𝑁3 |8𝑁𝑜𝑑𝑒 = 𝑁3 |4𝑁𝑜𝑑𝑒 − 2 𝑁5 − 2 𝑁7
1
= (1 + 𝜉)(1 + 𝜂)(−1 + 𝜉 + 𝜂)
4
1 1
𝑁4 |8𝑁𝑜𝑑𝑒 = 𝑁4 |4𝑁𝑜𝑑𝑒 − 2 𝑁7 − 2 𝑁8
1
= (1 − 𝜉)(1 + 𝜂)(−1 − 𝜉 + 𝜂)
4
As an illustration, shape function N1 is plotted in the above Figure. Using these
shape functions, it is possible to map a straight line in the (−) space to a curve in
the physical (x−y) space. Consider edge 1−5−2 ( = −1). On this edge, we have N3
Thus,
2
𝑥 = −𝜉 (1−𝜉
2
)
𝑥1 + 𝜉 (1+𝜉
2
)
2 + (1 − 𝜉 )𝑥 5
𝑥
−𝜉(1 − 𝜉) 𝜉(1 + 𝜉)
𝑦= 𝑦1 + 𝑦2 + (1 − 𝜉 2 )𝑦5
2 2
For the general values of nodal coordinates ( x1, y1), (x2, y2) and (x5, y5), this
would represent a nonlinear transformation. For example, if the nodal coordinates
are given by (0, 0), (3, 9) and (2, 4), then we have
𝑥 = 3 (1 + 𝜉) + 1 (1 − 𝜂 2 )
2 2
9 1
𝑦 = (1 + 𝜉) + (1 − 𝜂 2 )
2 2
Therefore, the equation representing the edge 1−5−2 in the Cartesian space
is
𝑥 1 +𝑥 2
𝑥5 =
2
𝑦 1 +𝑦 2
𝑦5 =
2
then the transformation and reduces to the simple case of mapping a straight edge
1−5−2 in the (−) space to another straight edge in the physical ( x−y) space.
Thus, by making the nodes 5, 6, 7 or 8 move away from the middle of their edges in
the physical (x−y) space, we can achieve curved edge elements. It is observed that,
irrespective of their locations in the physical ( x−y) space, nodes, 5, 6, 7, 8 are at the
middle of their edges in the − frame.
𝐿2 = 𝐴𝐴2 = 𝐴𝑟 𝑒 𝑎 𝑜 𝑓 𝛥1𝑃
𝐴𝑟𝑒𝑎𝑜𝑓𝛥123
3
𝐿3 = 𝐴𝐴3 = 𝐴𝑟 𝑒 𝑎 𝑜 𝑓 𝛥𝑃12
𝐴𝑟𝑒𝑎𝑜𝑓𝛥123
L1 + L2 + L3 = 1
𝑁1 = 𝐿1 , 𝑁2 = 𝐿2 , 𝑁3 = 𝐿3
Comparing with the shape functions obtained for the three-noded triangular
element, it is observed that we can very easily write down the required shape
functions in terms of the natural coordinates just described. If we require a higher
order transformation, for example, to model curved domains, we can then use a six-
noded triangular element as shown in Figure. The natural coordinates of the nodes
are indicated in the figure. Following our earlier procedure, the shape functions can
be obtained as
where the scale factor (SF) is to be chosen such that N1 is unity at node 1 and the
product of the equations of lines 2−5−3 and 4−6 has been taken to ensure that N1
is zero at all these nodes. Equation of line 2−5−3 is L1 = 0 and that of line 4−6 is L1
= 1/2, i.e., 2L1 − 1 = 0. At node 1, L1 = 1 and, therefore, SF = 1. Thus the desired
shape function is readily obtained as
𝑁1 = (𝐿1 ) (2𝐿1 − 1)
𝑥1
𝑦1
𝑥2
𝑦2
𝑥3
𝑥 𝑁 0 𝑁2 0 𝑁3 0 𝑁4 0 𝑁5 0 𝑁6 0 ]
𝑁6 𝑥 𝑦3
{ 𝑦} = [ 01 4
𝑁1 0 𝑁2 0 𝑁3 0 𝑁4 0 𝑁5 0 𝑦4
𝑥5
𝑦5
𝑥6
{ 𝑦6 }
Comparing with our discussion in section 5.2.3, we readily observe that the natural
coordinates as defined here for the triangular coordinates significantly simplify the
derivation of shape functions. We will now discuss the formulation of the element
level equations for these elements, viz., three- and six-noded triangles, four- and
eight-noded quadrilaterals, etc. for both structural mechanics problems and fluid
flow problems.
Consider the parent four-node square element in – frame as well as the
general quadrilateral element in the physical x-y space as shown in Figure. The
coordinates of any point P (x, y) are interpolated from nodal coordinates as follows:
4
𝑥 = ∑ 𝑁𝑖 𝑥𝑖
𝑖 =1
4
𝑦 = ∑ 𝑁𝑖 𝑥𝑖
𝑖 =1
The displacements of the interior point P are also interpolated from nodal
deflections using the same shape functions for this element, which is therefore called
an “isoparametric element”. The displacements can be written as
𝑢1
𝑣1
𝑢2
𝑢 𝑁 0 𝑁2 0 𝑁3 0 𝑁4 𝑁04 ] 𝑢𝑣32
{ 𝑣 } [ 01
𝑁1 0 𝑁2 0 𝑁3 0 𝑣3
𝑢4
{ 𝑣4 }
Let us now show that the displacements in this element are compatible across
inter-element boundaries. Along edge 2−3 ( = 1), for example, the shape functions
N1 = 0 = N4. Thus, irrespective of the shape of the element in the physical x-y space
(i.e. even for a general quadrilateral), the displacements of any point on edge 2−3
are given as
u = N2u2 + N3u3
v = N2v2 + N3v3
Recall our observation that the nodal d.o.f. are in X-Y directions and not along
−. Also the −. axes, as transformed into real physical x− y space, may not be
orthogonal. Thus we write the strain displacement relations in an orthogonal,
Cartesian frame of reference, as
𝜕𝑢
𝜕𝑥
𝜀𝑥
𝜕𝑣
{𝜀} = { 𝜀𝑦 } =
𝛾𝑥 𝑦 𝜕𝑦
𝜕𝑢 𝜕𝑣
+
{ 𝜕𝑦 𝜕𝑥}
We observe that u and v are given in terms of shape functions Ni which are
expressed in the −. coordinates rather than x−y. For a general function ƒ, using
the chain rule of differentiation, we obtain the equations
𝜕𝑓 𝜕𝑓 𝜕𝜉 𝜕𝑓 𝜕𝜂
= +
𝜕𝑥 𝜕𝜉 𝜕𝑥 𝜕𝜂 𝜕𝑥
𝜕𝑓 𝜕𝑓 𝜕𝑥 𝜕𝑓 𝜕𝑦
= +
𝜕𝑦 𝜕𝑥 𝜕𝜂 𝜕𝑦 𝜕𝜂
𝜕𝑓 𝜕𝑓 𝜕𝑥 𝜕𝑓 𝜕𝑦
= +
𝜕𝜉 𝜕𝑥 𝜕𝜉 𝜕𝑦 𝜕𝜉
𝜕𝑓 𝜕𝑓 𝜕𝑥 𝜕𝑓 𝜕𝑦
= +
𝜕𝜂 𝜕𝑥 𝜕𝜂 𝜕𝑦 𝜕𝜂
𝜕 𝜕𝑥 𝜕𝑦 𝜕
𝜕𝜉
𝜕 = 𝜕𝜉 𝜕𝜉 𝜕𝑥
𝜕𝑥 𝜕𝑥 𝜕
{ 𝜕𝜂}
[𝜕𝜂 𝜕𝜂] { 𝜕𝑦}
𝜕 𝜕
𝜕𝜉
i.e. { } = [𝐽] { 𝜕 𝑥 }
𝜕 𝜕
𝜕𝜂 𝜕𝑦
where
𝜕𝑥 𝜕𝑦
𝜕𝜉 𝜕𝜉
[𝐽] =
𝜕𝑥 𝜕𝑦
[𝜕𝜂 𝜕𝜂]
𝜕𝑁 𝜕𝑁
∑𝑖 𝑥𝑖 ∑ 𝑖 𝑦𝑖
𝜕𝜉 𝜕𝜉
[𝐽] =
𝜕𝑁 𝜕𝑁
∑ 𝑖 𝑥𝑖 ∑ 𝑖 𝑥𝑖
[ 𝜕𝜂 𝜕𝜂 ]
Here, [J] is known as the Jacobian matrix relating the derivatives in two
coordinate frames. This generic expression for [ J] is true for all 2-d elements, but
the actual coefficients in the matrix will depend on the shape functions being
employed in a given element and the nodal coordinates.
For this particular element, for example, the Jacobian is
(1 − 𝜂) (1 + 𝜂) (1 − 𝜂) (1 + 𝜂)
(𝑥2 − 𝑥1 ) + (𝑥3 − 𝑥4 ) (𝑦2 − 𝑦1 ) + (𝑦3 − 𝑦4 )
[𝐽] = [ 4 4 4 4
]
(1 − 𝜉) (1 + 𝜉) (1 − 𝜉) (1 + 𝜉)
(𝑥4 − 𝑥1 ) + (𝑥3 − 𝑥2 ) (𝑦4 − 𝑦1 ) + (𝑦3 − 𝑦2 )
4 4 4 4
𝜕 𝜕 𝜕
𝜕𝜉 𝜕𝜉
𝜕𝑥 1 𝐽 −𝐽12
= [𝐽] −1 𝜕 = |𝐽| [ 22 𝐽11 ]
𝜕 −𝐽21 𝜕
{ 𝜕𝑦} { 𝜕𝜂} { 𝜕𝜂}
𝜕𝑢
𝜕𝑢 𝜕𝑥
𝜕𝑥 𝜕𝑢
𝜕𝑣 1 0 0 0 𝜕𝑦
{𝜀} = = [0 0 0 1] 𝜕𝑣
𝜕𝑦 ⏟0 1 1 0
𝜕𝑢 𝜕𝑣 [𝐵1 ] 𝜕𝑥
+
{ 𝜕𝑦 𝜕𝑥} 𝜕𝑣
{ 𝜕𝑦}
𝜕𝑢 𝐽22 𝐽12 𝜕𝑢
− |𝐽| 0 0
𝜕𝑥 𝜕𝜉
𝜕𝑢 |𝐽| 𝐽21 𝐽11 𝜕𝑢
− |𝐽| |𝐽| 0 0
𝜕𝑦 𝜕𝜂
= 𝐽22
𝜕𝑣 𝐽12 𝜕𝑣
0 0 −
𝜕𝑥 |𝐽| |𝐽| 𝜕𝜉
𝜕𝑣 𝐽21 𝐽11 𝜕𝑣
0 0 −
{ 𝜕𝑦} [⏟ |𝐽| |𝐽| ] { 𝜕𝜂}
[𝐵2 ]
For this element, for example, the shape function derivatives in the - frame
are as follows:
𝜀𝑥
𝜀
{𝜀} = { 𝑦 } = [𝐵]{𝛿}𝑒
𝜀𝑥 𝑦
The stress-strain relation matrix [D] remains as given earlier (see, e.g. Eq. (5.91)).
Thus we have the element stiffness matrix
𝑇
[𝐾]𝑒8×8 = ∫ [𝐵] [𝐷][𝐵]𝑑𝑣 = ∬ [𝐵]𝑇 [𝐷][𝐵]𝑡𝑑𝑥 𝑑𝑦
𝑣
1 1 𝑇
= ∫ ∫ [𝐵] [𝐷][𝐵]𝑡|𝐽|𝑑𝜉𝑑𝜂
−1 −1
where the coordinate transformation has been taken into account and the integrals
need now be evaluated over the parent square element in natural coordinates. Since
[B] in general varies from point to point within the element, we look for ways of
evaluating these integrals numerically within a computer program rather than
attempting to derive them explicitly. We will discuss numerical integration schemes
As per our generic formula, we attempt to express the integral in the following way:
I W1 ƒ1 + W2 ƒ2 + ⋯ + Wn ƒn
𝑝
𝐼 ≈ 𝑊1 (𝑐0 + 𝑐1 𝑥1 + 𝑐2 𝑥12+. . . +𝑐𝑝 𝑥1 )
𝑝
+𝑊2 (𝑐0 + 𝑐1 𝑥2 + 𝑐2 𝑥22+. . . +𝑐𝑝 𝑥2 )
+. . .
+. . .
𝑝
+𝑊𝑛 (𝑐0 + 𝑐1 𝑥2 + 𝑐2 𝑥𝑛2 +. . . +𝑐𝑝 𝑥𝑛 )
Comparing the coefficients of c0, c1, c2, … in Eqs. (5.207) and (5.209), we can
write
𝑊1 + 𝑊2 +. . . +𝑊𝑛 = 2
𝑊1 𝑥1 + 𝑊2 𝑥2 +. . . +𝑊𝑛 𝑥𝑛 = 0
2
𝑊1 𝑥12 + 𝑊2 𝑥 22 +. . . +𝑊𝑛 𝑥 2 =
𝑛 3
⋮
𝑝 𝑝 𝑝 1
𝑊1 𝑥1 + 𝑊2 𝑥2 +. . . +𝑊𝑛 𝑥𝑛 = [1 − (−1)𝑝 +1]
𝑝+1
These equations can be solved for the locations of the sampling points xi (i =
1, 2, …, n) and the corresponding weight factors Wi (i = 1, 2, …, n). It is easy to see
that our number of unknowns (viz., 2 n) and equations available ( p + 1) match
exactly when the polynomial degree p is such that
p + 1 = 2n
or 𝑝 = 2𝑛 − 1
We now list the sampling points and the weight factors for Gauss quadrature
for one-dimensional case in Table .
It is easy to observe that the exact numerical values of the sampling points
and the weight factors, obtained as a solution of Eq. depend on the integration
interval. Thus the formula is standardised for the interval −1 to 1, keeping in mind
the finite element computations using natural coordinates. If the interval is different,
we need to perform an appropriate change of coordinates such that the limits tally
with −1 and 1. We will now illustrate the use of the formula through the following
examples.
1
𝐼 = ∫ (2 + 𝑥 + 𝑥 2 )𝑑𝑥
−1
It is easy to verify that we need at least a two-point integration rule since the
integrand contains a quadratic term. We will obtain our approximation to the integral
with one and two-point Gauss quadrature rules and show that the two-point rule
result matches with the exact solution.
x1 0, W1 = 2
Hence,
I 2(2 + 0 + 0) = 4
1
𝑥1 = + √3 , 𝑊1 = 1
1
𝑥2 = + √3 , 𝑊2 = 1
Therefore,
1 1 1 1
𝐼 ≈ (1) (2 + + ) + (1) (2 − + ) = 4.6667
√3 3 √3 3
1
1
𝐼 = ∫ (2 + 𝑥 + 𝑥 2 )𝑑𝑥 = (2𝑥 + 𝑥 2 + 𝑥 3 /3)−1 = 4.6667
−1
1
𝜋𝑥
𝐼 = ∫ 𝑐𝑜𝑠 𝑑𝑥
−1
2
𝜋𝑥 𝜋 1
𝐼 = (𝑠𝑖𝑛 / ) = 1.273239544
2 2 −1
(𝜋)(0)
𝐼 ≈ 2 𝑐𝑜𝑠 ( ) =2
2
𝜋 1 𝜋 −1
𝐼 ≈ 1 𝑐𝑜𝑠 ( ) + 1 𝑐𝑜𝑠 ( )
2 √3 2 √3
𝜋
= 2 𝑐𝑜𝑠 ( ) = 1.232381
2√3
5 𝜋 8 𝜋 5 𝜋
𝐼 ≈ 𝑐𝑜𝑠 ( + √0.6) + 𝑐𝑜𝑠 ( 0) + 𝑐𝑜𝑠 ( − √0.6)
9 2 9 2 9 2
= 1.274123745
Since the integrand is not an exact polynomial, our numerical scheme can
never give the exact solution but approximates fairly well when sufficient number of
points are taken.
Problem:
For the four noded rectangular element shown if Fig. determine the following:
i) Jacobian matrix
ii) Strain-Displacement matrix
iii)Element stresses
x1 = 0 ; y1 = 0
x2 = 2 ; y2 = 0
x3 = 2 ; y3 =1
x 4 = 0; y 4 =1
38
0
0
0.003
0.004
Displacement , u =
0 . 006
0.004
0
0
Natural Co − ordinates, = 0, = 0
To Find: 1. Jacobian matrix, J.
2. Strain Displacement, [B]
3. Element stress,
Solution:
J 11 J 12
J =
J 21 J 22
1
J 11 = [−(1 − ) x1 + (1 − ) x 2 + (1 + ) x3 − (1 + ) x 4 ]
4
1
J 12 = [−(1 − ) y1 + (1 − ) y 2 + (1 + ) y 3 − (1 + ) y 4 ]
4
1
J 21 = [−(1 − ) x1 − (1 + ) x 2 + (1 + ) x3 + (1 − ) x 4 ]
4
1
J 22 = [−(1 − ) y1 − (1 + ) y 2 + (1 + ) y 3 + (1 − ) y 4 ]
4
1 1
J 11 (0,0) = [0 + 2 + 2 − 0] ; J 12 (0,0) = [0 + 0 + 1 − 1]
4 4
=1 =0
1 1
J 21 (0,0) = [0 − 2 + 2 + 0] ; J 22 (0,0) = [−0 − 0 + 1 + ]
4 4
=0 = 0. 5
J 11 J 12
J =
J 21 J 22
1 0
Jacobian matrix, [ J ] =
0 0.5
J =1 0.5 − 0 = 0.5
Strain- Displacement matrix for quadrilateral element is,
J 22 − J 12 0 0
1
J 11
1
[ B] = 0 0 − J 21
J 4
− J 21 J 11 J 22 − J 12
− (1 − ) 0 (1 − ) 0 (1 + ) 0 − (1 + ) 0
− (1 − ) 0 − (1 + ) 0 (1 + ) 0 (1 − ) 0
0 − (1 − ) 0 (1 − ) 0 (1 + ) 0 − (1 + )
0 − (1 − ) 0 − (1 + ) 0 (1 + ) 0 (1 − )
− 1 0 1 0 1 0 −1 0
0.5 0 0 0
1 1 − 1 0 − 1 0 1 0 1 0
[B ] = 0 0 0 1
0.5 4 0 −1 0 1 0 1 0 − 1
0 1 0.5 0
0 −1 0 −1 0 1 0 1
−1 0 1 0 1 0 −1 0
0.5
= 0 − 2 0 − 2 0 2 0 2
0.5 4
− 2 − 1 − 2 1 2 1 2 − 1
−1 0 1 0 1 0 −1 0
[B] = 0.25 0 − 2 0 − 2 0 2 0 2
− 2 − 1 − 2 1 2 1 2 − 1
1 0.25 0
= 213.3310 3 0.25 1 0
0 0 0.375
4 1 0
= 213.3310 3 0.25 1 4 0
0 0 1.5
4 1 0
[ D] = 53.3310 3 1 4 0
0 0 1.5
= D B d
4 1 0
= 53.3310 3 1 4 0
0 0 1.5
−1 0 1 0 1 0 −1 0
0.25 0 −2 0 −2 0 2 0 2
− 2 −1 −2 1 2 1 2 − 1
0
0
0.003
0 .004
0 .006
0.004
0
0
− 4 − 2 4 − 2 4 2 − 4 2
= 53.3310 3 0.25 − 1 − 8 1 − 8 1 8 − 1 8
− 3 − 1.5 − 3 1.5 3 1.5 3 − 1.5
0
0
0.003
0.004
0 .006
0.004
0
0
0.036
3
=13.33310 0.009
0.021
480
= 120 N / m 2
280
44
ASSIGNMENT
S.No Question K Level CO
.
1. Local coordinates
2. Natural coordinates
47
PART - A QUESTIONS AND ANSWERS
UNIT – I INTRODUCTION
S.N O QUESTION AND ANSWER K CO
LEVEL
06 What do you mean by sub parametric element? K1 CO5
The element in which less number of nodes are used to
define geometry compared to the number of nodes used
for defining the displacements. Such elements can be
used advantageously in case of geometry being simple
but stress gradient high.
48
PART - A QUESTIONS AND ANSWERS
UNIT – I INTRODUCTION
S.N O QUESTION AND ANSWER K CO
LEVEL
11 List any three FEA software K1 CO5
The following list represents FEA software as,
(i) ANSYS
(ii) NASTRAN
(iii) COSMOS
1 1
∫−1 ∫−1(3 𝑥 2 + 4𝑦 2 ) 𝑑𝑥 𝑑𝑦
40
S.No. Question K Level CO
05 For the isoparametric four noded quadrilateral K2 CO5
element shown in fig. Determine the Cartesian co-
ordinates of point P which has local co-ordinatesε=
0.5 , η =0.5
52
REAL TIME APPLICATION IN DAY TO DAY LIFE AND TO
INDUSTRY
53
CONTENT BEYOND SYLLABUS
54
ASSESSMENT SCHEDULE
55
PRESCRIBED TEXT BOOKS & REFERENCE BOOKS
TEXT BOOKS:
T1) Reddy. J.N., “An Introduction to the Finite Element Method”, 3rd Edition, Tata
McGraw Hill, 2005
T2) Seshu, P, “Text Book of Finite Element Analysis”, Prentice-Hall of India Pvt. Ltd.,
New Delhi, 2007.
REFERENCES:
R1) Rao, S.S., “The Finite Element Method in Engineering”, 3rd Edition, Butterworth
Heinemann, 2004
R2) Logan, D.L., “A first course in Finite Element Method”, Thomson Asia Pvt. Ltd.,
2002
R3) Robert D. Cook, David S. Malkus, Michael E. Plesha, Robert J. Witt, “Concepts
and Applications of Finite Element Analysis”, 4th Edition, Wiley Student Edition,
2002.
56
MINI PROJECT SUGGESTIONS
The purpose of this tutorial is to explain how to apply distributed loads and use
element tables to extract data. Please note that this material was also covered in the
'Bicycle Space Frame' tutorial under 'Basic Tutorials'.
A distributed load of 1000 N/m (1 N/mm) will be applied to a solid steel beam with a
rectangular cross section as shown in the figure below. The cross-section of the
beam is 10mm x 10mm while the modulus of elasticity of the steel is 200GPa.
57
Thank you
Disclaimer:
This document is confidential and intended solely for the educational purpose of RMK Group of
Educational Institutions. If you have received this document through email in error, please notify the
system manager. This document contains proprietary information and is intended only to the
respective group / learning community as intended. If you are not the addressee you should not
disseminate, distribute or copy through e-mail. Please notify the sender immediately by e-mail if you
have received this document by mistake and delete this document from your system. If you are not
the intended recipient you are notified that disclosing, copying, distributing or taking any action in
reliance on the contents of this information is strictly prohibited.