,

L-2/T -2/ME Date: 12/01/2021

_________ B,A.l\JG~DESI:U.!l\JIVERSJJY.QF EN~JNEERING AND T.E~HNQLO(3Y, --------
BUET
L-2fT-2 B.Sc. Engineering Examinations 2019-2020
Sub: ME 261 (Numerical Analysis)
Full Marks: 180 Time: 2 Hours
The figures in the margin indicate full marks.
Symbols used have their usual meaning and interpretation.
USE SEPARATE SCRIPTS FOR EACH SECTION

SECTION-A

There are FOUR questions in this section. Answer any THREE

1. (a) Find the root of the following equation using the Newton's method, (15)

8 - 4.5(x - sinx) =0
Use initial guess, xo = 2.0 and Es = 0.1%. From your solution show that Newton's method
converges quadratically.
(b) A simplified model of the suspension of a car consists of a mass, m, a spring WIThstiffness, (15)
k, and a dashpot with damping coefficient, c. A bumpy road can be modeled by a sinusoidal
up-and-down motion of the wheel y = Ysin(wt). From the solution of the equation of motion
for this model, the steady-state up-and.down motion of the car (mass) is given by x =
Xsin(wt - <{». The ratio between amplitude X and amplitude Y is given by:

X mew3
- = 1.,..,.,--':";-:--,-.,-"
Y 2
k(k - mw ) + (we)2
Assuming m = 2000 kg, k = 500 kN/m, and c = 38x103 Ns/m, determine the frequency w
for which XIY = 0.2. Rewrite the equation such that it is in the form of a polynomial in wand
solve using false position method using WI = 2.0 and ~u = 10.0. Show three ITerations.
2. (a) In a linear coiled spring, the relation between spring force (F) and displacement (x) is (18)
described by F = kx, where k is the spring constant. Testing on a certain spring has led to
the data recorded in the following Table (All values are in consistent physical units).

Displacement, x 0.2 0.3 0.5 0.6 0.8 0.9

Force, F 43.5 65.7 109.8 133 176.2 198.2

(i) Using least-squares regression, find a straight line that best fits the data.
(ii) Using the linear fit, find the estimated value for the spring constant, and the
displacement corresponding to F = 150.

(b) What modifications are needed to fit the data of problem 2(a) with (12)
(i) weighted linear regression
(ii) quadratic polynomial regression

Page NO.1
•

3. (a) (i) Solve the following system of equations ([Aj{x} = {b}) by using Gauss elimination method (18)
. with partial pJyoting-'-- ., ._ _ .... . ....__
Xl + X2 -- X3 =2
2xl + 3X2 - 5x3 = -3
3xl + 2X2 - 3X3 =6

(ii) using the results obtained in (i), find the [Lj matrix required for solving the system with
the LU decomposition method.
(iii) using the results obtained in (i), find the determinant of the coefficient matrix [Aj.
(b) Consider the following linear system, (12)

1
5
1
~1],b= 1~7)' = 1::), .
6 13.
x
~
xeD)
Initial vector = 1~1
1
(i) Solve the above system using the Gauss-Seidel method. Show at least one iteration.
(Ii) Will this system converge to a solution using the Gauss-Seidel method?
4. (a) Discuss how the total error can be minimized for solving a problem numerically. (6)
(b) The mid-point deflection of a transversely loaded rotating shaft can be estimated (15)
experimentally by the following equation,

WL3
Y = 48EI

where, W is the weight of the shaft, L is the length of shaft, E is the modulus of elasticity of
the shaft material and I is the moment of inertia. During an experiment following data were
recorded, W = 3100:!: 75 N, L = 2.5:!: 0.1 m, and 1= 0.0003:!: 0.00005 m4.
Considering, E = 2 GPa for the shaft material to be exact, estimate the error in deflection y
for the given data.
(c) Use zero- through third-order Taylor series expansions to predict f(3) for f(x) = 25x' - 6x2 (9)
+ 7x - 88 using a base point at x = 1. Compute the true percent relative error Et for each
approximation. Discuss the meaning of the results.

SECTION-S

There are FOUR questions in this section. Answer any THREE

5. (a) The record of velocity (v) and acceleration (a) of a particle is given below: (20)

t (min) 1 2 3.25 4.5 6

v = dydt (mls) 5 6 5.5 7 5.5

a = dv
dt
(m/s2) 0.01 0.02 -0.007 0.02 .. -0.02

(i) Determine the total distance (y) travelled by the particle using Trapezoidal Rule. What
is the order of Truncation Error associated with the result?

Page NO.2
• ,.
(ii) Can the acceleration (a) data be used to obtain a more accurate value of distance (y)?
If yes, determine that more accurate value and the order_o.fJ:r.uJ)c~lion Er[QUI~s.Qci<lt\ld
with it.
(b) For Simpson's 1/3'" Rule, determine the minimum number of divisions required to (10)
guarantee a result within the 0.2% of the exact result of integration of the function-

{(x) = lSe-x from x = 0.2 to 1.5

6. (a) With necessary diagrams, describe Forward Difference, Backward Difference and Central (14)
Difference approach of Numerical Differentiation. Explain their applicability and relative
advantage/disadvantage in context to practical problems.

(b) A Simply Supported Beam (length, L = 6m, Bending Stiffness, EI = 26 x 10' N-m') has (16)
the following deflections under a partial distributed loading-

x (m) 0 1 2 3 4 5 6

Y (m) 0 0.0102 0.0282 0.0413 0.0368 0.0173 0

Where, x = axial location, y = vertical deflection.

Determine bending moments, M = EJ d'Y, at x = 2m and 4m with-
dx

(i) accuracy of O(h')

(ii) accuracy of O(h') using Richardson's Extrapolation Scheme.

7. (a) The general form of Newton's interpolating polynomial is- (14)

(nIx) = bo + b,(x - xo) + b,(x - xo)(x - X,) ..... +bn(x - xo)(x - X,) ..... (x - Xn-l)

Show that, the coefficients bo, b" b2, ...• , bn actually indicate derivatives by Newton's
Divided Difference Method.

(b) The table below gives the values of ((x) from x = 2.0 to x = 4.0. (16)

x 2.0 2.5 3.2 4

((x) 8 14 15 8

(i) Is ~ possible to formulate interpolating polynomials of different orders to estimate (2.8)

from the given data? If yes, how many different pOlynomials of different orders are
possible?
(ii) Estimate (2.8) using the interpolating polynomial of the highest order by a suitable
method.
8. (a) "Heun's method is a Predictor-Corrector approach for solving 1st order ODEs"- justify. (10)

(b) The governing equation of a particle under vibration is found to be- (20)
d'x dx
dt' + 0.6 dt + 0.8t = a
Where, x = displacement of the particle, t = time. x (0) = 4, x' (0) = O.
Determine the veloc~y at t = 2 and 4 by 4th order R-K method.

Page NO.3