ME046 : FINITE ELEMENT TECHNIQUES

Time: 3Hrs Max. marks: 100

PART - A (10 × 2 = 20)

1. Distinguish between variational approach and weighted residual approach in finite

element technique.

2. What is a functional? Write down the functional to be minimized for the following
linear differential equation
d2 φ
D + Q(x)φ + R(x) = 0
dx2
with boundary conditions

φ(a) = 0 and φ(b) = 0.

3. Define the band-width of global stiffness matrix and explain how it is related to
node numbering.

4. Differentiate between global, local, and natural coordinate systems.

5. What is an eigenvalue problem? Write down the differential equation and associated
boundary conditions for a one-dimensional eigenvalue problem.

6. Write down the expression for total potential energy of a structural element in
standard form and identify various terms in the equation.

7. Discuss the method of obtaining stress components in an elasticity problem, if

displacement method is used for obtaining nodal displacements.

8. Explain how do you take care of surface tractions, in two-dimensional elasticity

problems, while formulating the element equations.

9. What is meant by geometric isotropy? How do you construct an element which

possesses the geometric isotropy?

10. What is an isoparametric element? How does the gradient of shape functions for
an isoparametric element is related in the global and natural coordinate systems?

1
 PART - B (5×16 = 80)

1. Derive the condition for a function y(x), which makes the functional,
Z x2
I(y) = F (x, y, y 0 , y 00 )dx,
x1

stationary with prescribed boundary conditions,

y(x1 ) = y1 , y(x2 ) = y2 , y 0 (x1 ) = y10 , and y 0 (x2 ) = y20 .

2. Using Rayleigh-Ritz method, obtain an approximate displacement equation for the

simply supported beam shown below using the trial solution y(x) = sin(πx/H).
The governing differential equation is

d2 y
EI − M (x) = 0.
dx2

Use appropriate boundary conditions.

OR

3. (a) With the help of a pseudo-code, describe the algorithm for Cholesky method
for solving system of linear algebraic equations.
(b) Discuss how you will incorporate the Dirchlet and Neumann type boundary
conditions in the finite element formulations.

4. Consider the one-dimensional heat transfer through the composite wall (see figure
below). Using FEM, determine the interface temperatures and calculate the heat
flux though each layer. The left side of the wall is maintained at 20 o C and the
right side is maintained at −15 o C. The thermal conductivities of the material are
K1 = 0.02, K2 = 0.005, and K3 = 0.0035 W/cm o C.

2
 OR

5. Evaluate the Jacobian [J] at ξ = η = 1/2 for the linear quadrilateral element shown
below. Also, find the first partial derivatives of each shape function with respect
to x and y at the same location.

6. Develop the global [K] matrix and [F ] vector for the two-dimensional steady state
temperature distribution problem shown below. The solution domain may be dis-
cretized using eight bilinear rectangular elements. Assume that the material is
isotropic and homogeneous with K = 2 W/mK. The convection heat transfer
coefficient, h = 10 W/m2 K and the air temperature is 300 K.

3
 OR

7. Derive the expression for the shape function N1 for a two-dimensional, 9-node,
quadratic, Lagrangian element and a 8-node, quadratic, serendipity element. Node
’1’ may be located at the bottom left corner of the element.

8. The buckling of a simply supported column is governed by the differential equation

d2 y P
+ y=0
dx 2 EI
where the buckling load, P = 80 kN and the section modulus EI = 2 × 105 N m2 .
The boundary conditions are y(0) = y(9) = 0. Determine the first three critical
loads and compare it with the theoretical value given by Pcr = n2 π 2 EI/H 2 .

OR

9. Calculate the element stiffness and the thermal force vector for the plane stress
element shown below. The element experiences a 10 o C change in temperature.
The following data may be taken: t = 0.5 cm, E = 15 × 106 N/cm2 , ν = 0.25, and
α = 6 × 10−6 /o C.

Also determine the stress vector [σ] for the following displacement components:

U2i−1 = 0.0 U2j−1 = −0.0006 U2k−1 = 0.0

U2i = −0.006 U2j = 0.007 U2k = −0.0096