UPSC Civil Services Main 1992 - Mathematics

Calculus
Sunder Lal
Retired Professor of Mathematics
Panjab University
Chandigarh

January 9, 2010

Question 1(a) If y = eax cos bx, prove that y2 2ay1 + (a2 + b2 )y = 0 and hence expand
eax cos bx in powers of x. Deduce the expansion of eax and cos bx.

Solution.

y = eax cos bx y(0) = 1

y1 = eax (a cos bx b sin bx) y1 (0) = a
y2 = aeax (a cos bx b sin bx) + eax (ab sin bx b2 cos bx)
= eax ((a2 b2 ) cos bx 2ab sin bx)
Thus y2 2ay1 + (a2 + b2 )y
= eax ((a2 b2 ) cos bx 2ab sin bx 2a(a cos bx b sin bx) + (a2 + b2 ) cos bx) = 0

as required.
Thus y2 = 2ay1 (a2 + b2 )y. Differentiating n 2 times, yn = 2ayn1 (a2 + b2 )yn2 .
Thus we have a recurrence relation for yn , which we can use to compute yn (0).

y(0) = 1
y1 (0) = a
y2 (0) = 2ay1 (0) (a2 + b2 )y(0) = a2 b2
y3 (0) = 2ay2 (0) (a2 + b2 )y1 (0) = 2a3 2ab2 a3 ab2 = a3 3ab2
y4 (0) = 2a4 6a2 b2 a4 + b4 = a4 6a2 b2 + b4
...
r n
2  
X n n2r 2r
r
yn (0) = (1) a b
r=0
2r

1
The last formula can be proved by induction the coefficient of an2r b2r in yn (0) from the
RHS of the recurrence relation is
     
r n1 r n2 r1 n 2
2(1) (1) (1)
2r 2r 2r 2
          
r n1 n2 n2 n2 n2
= (1) 2 + +
2r 2r 2r 1 2r 1 2r 2
      
n1 n1 n1
= (1)r 2 +
2r 2r 2r 1
   
n1 n1
= (1)r +
2r 2r 1
 
n
= (1)r
2r
Now by the Taylor-Maclaurin formula

ax
X xn
y = e cos bx = yn (0)
n=0
n!

where the yn are given as above.

Putting b = 0, all terms except the first in each of yn (0) vanish, so

ax
X xn
e = an
n=0
n!

Putting a = 0, all terms in the odd yn (0) vanish, and all but the last in the even ones
vanish:
X x2r
cos bx = (1)r b2r
r=0
(2r)!
Note: It is possible to expand eax cos bx in powers of x directly without using the formula
y2 = 2ay1 (a2 + b2 )y as shown below however in this question we are required to use
the formula and proceed as above.
Put a = r cos , b = r sin , so that

y1 = eax r(cos cos bx sin sin bx) = reax cos(bx + )

y2 = reax r(cos cos(bx + ) sin sin(bx + )) = r2 eax cos(bx + 2)
...
yn = rn eax cos(bx + n)

Thus yn (0) = rn cos n. Using the standard formula for cos n, and then using a = r cos , b =
r sin , we see that this is the same as the above result (or alternatively using these results
we have derived the standard formula for cos n).

2
Question 1(b) If x = r sin cos , y = r sin sin , z = r cos then prove that
dx2 + dy 2 + dz 2 = dr2 + r2 d2 + r2 sin2 d2
Solution.
dx = sin cos dr + r cos cos d r sin sin d
dy = sin sin dr + r cos sin d + r sin cos d
dz = cos dr r sin d
We square and add these three equations.
Coefficient of dr2 = sin2 cos2 + sin2 sin2 + cos2 = 1
Coefficient of d2 = r2 cos2 cos2 + r2 cos2 sin2 + r2 sin2 = r2
Coefficient of d2 = r2 sin2 sin2 + r2 sin2 cos2 = r2 sin2
Coefficient of 2 dr d = r sin cos (cos2 + sin2 ) r sin cos = 0
Coefficient of 2 dr d = r sin2 cos sin + r sin2 cos sin = 0
Coefficient of 2 d d = r2 cos cos sin sin + r2 cos sin sin cos = 0
Thus dx2 + dy 2 + dz 2 = dr2 + r2 d2 + r2 sin2 d2 .

Question 1(c) Find the dimensions of the rectangular parallelopiped inscribed in the ellip-
x2 y 2 z 2
soid 2 + 2 + 2 = 1 that has the greatest volume.
a b c
2a
Solution. See 1997 question 1(b). The dimensions are
3
, 2b3 , 2c3 .

Question 2(a) Prove that the volume enclosed by the cylinders x2 + y 2 = 2ax and z 2 = 2ax
is 128
15
a3 .

Solution.
Clearly
on the given volume z varies from 2ax to 2ax, y varies from
2ax x2 to 2ax x2 and x varies from 0 to 2a (note that x2 + y 2 = 2ax is a circle in
the XY-plane in which x varies from 0 to 2a).
Z 2a Z 2axx2 Z 2ax Z 2a Z 2axx2
V =
dz dy dx = 2
2ax dy dx
0 2axx2 2ax 0 2axx2
Z 2a Z 2a
= 4 2
2ax 2ax x dx = 4 2a x 2a x dx
0 0
2
Put x = 2a sin
Z 2
= 4 2a 2a sin2 2a cos 4a sin cos d
0
Z
2 21 128 3
= 64a3 sin3 cos2 d = 64a3 = a
0 531 15

3
Question 2(b) Find the center of gravity of the volume formed by revolving the area bounded
by the parabolas y 2 = 4ax and x2 = 4by about the x-axis.

Solution.

To find the intersection points of the

parabolas: x4 = 16b2 (4ax) x = 0, x3 = x2 = 4by
1
64ab2 x = 0, 4(ab2 ) 3 Thus the centroid is
given by (x, 0), where
R 4(ab2 ) 13
0
x(y12 y22 ) dx
x = R 1
4(ab2 ) 3
0
(y12 y22 ) dx
R 4(ab2 ) 13 x4
O
0
x 4ax dx
= 16b2
R 4(ab2 ) 31 x4

0
4ax dx
16b2
4(ab2 ) 13
4ax3 x6

y 2 = 4ax
3 96b2 0
=  4(ab2 ) 13
5
x
2ax2
80b2 0
4a 64ab2 (64ab2 )2 256 2 2 256a2 b4
ab 2 2
= 3  96b2  = 3 6b2  = 128a b 5 = 20 (ab2 ) 31
64ab2 2

2 2
2 ) 3 2a
4a 48(ab2 ) 3 6a 9
16(ab2 ) 3 2a 2
16(ab
80b 5
 20 1 2 
Hence the center of gravity is a3 b3 , 0 .
9
Question 2(c) Evaluate the following integral in terms of the Gamma function
Z 1
(1 + x)p (1 x)q dx (p > 1, q > 1)
1

and prove that ( 13 )( 32 ) = 2

3
.
Z b
Solution. Consider the integral (x a)m (b x)n dx. Put
a

x = a sin + b cos2
2

dx = (2a sin cos 2b cos sin )d = 2(a b) sin cos d

xa = b cos2 a(1 sin ) = (b a) cos2
bx = b(1 cos2 ) a sin2 = (b a) sin2

4
and x = a = 2 , x = b = 0. Thus
Z 0
I = 2 (b a)m+n+1 cos2m sin2n ( sin cos )d

2
Z
2
m+n+1
= 2(b a) cos2m+1 sin2n+1 d
0
2m+1
m+n+1 ( 2 + 21 )( 2n+1
2
+ 12 )
= (b a)
( 2m+1
2
+ 2n+12
+ 1)
(m + 1)(n + 1)
= (b a)m+n+1 = (b a)m+n+1 B(m + 1, n + 1)
(m + n + 2)
Substituting a = 1, b = 1, m = p, n = q we have
Z 1
(p + 1)(q + 1)
(1 + x)p (1 x)q dx = 2p+q+1
1 (p + q + 2)

Now put p = 23 , q = 13 to get

1
( 31 )( 23 ) 1
Z Z
23 13
2 sin 3
= (1 + x) (1 x) dx = 2 1 d
(1) 1 0 cos 3
Putting tan = z, Z
1 2 1 dz
( )( ) = 2 z3
3 3 0 1 + z2
2
Put z = t to get
1 2
t 3 1
Z Z
1 2 t 6 dt 2
( )( ) = 2 = dt =
3 3 0 1 + t 2t 21 0 1+t 3
by proceeding as 1994, question 2(a).

Paper II

Question 3(a) Examine f (x, y, z) = 2xyz 4xz 2yz + x2 + y 2 + z 2 2x 4y 4z for

extreme values.

Solution. For extreme values,

fx = 2yz 4z + 2x 2 = 0 yz 2z + x 1 = 0 (i)
fy = 2zx 2z + 2y 4 = 0 xz z + y 2 = 0 (ii)
fz = 2xy 4x 2y + 2z 4 = 0 xy 2x y + z 2 = 0 (iii)

Subtracting (ii) from (i) we get

z(y x 1) + x y + 1 = 0 (x y + 1)(1 z) = 0 z = 1 or x y + 1 = 0

5
Putting z = 1 in (i) and (iii) we get 2y + 2x 6 = 0 x = 3 y and xy 2x y 1 =
0 y(3 y) 2(3 y) y 1 = 0 y 2 4y + 7 = 0. This gives imaginary values, showing
that z = 1 is not possible.
Putting x = y 1 in equation (iii) we get y(y 1) 2(y 1) y + z 2 = 0
y 2 4y + z = 0 z = 4y y 2 . Substituting x = y 1, z = 4y y 2 in (i) we get
y(4y y 2 )2(4y y 2 )+y 11 = 0 4y 2 y 3 8y +2y 2
+y 2 = 0 y 3 6y 2 +7y +2 = 0.
2
Factorizing this, (y 2)(y 4y 1) = 0 y = 2, 2 5.
Case 1: y = 2, x = 1, z = 4. fxx = 2, fyy = 2, fzz = 2, fxy = 2z, fxz = 2y 4, fyz = 2x 2.
Thus the matrix associated with the quadratic form

d2 f = fxx (dx)2 + fyy (dy)2 + fzz (dz)2 + 2fxy dx dy + 2fxz dx dz + 2fyz dy dz

when x = 1, y = 2, z = 4 is given by

2 8 0
8 2 0
0 0 2
   
2 0  2 8 
of which the principle minors are    ,   or 4, 60, showing that the quadratic form
0 2  8 2 
d2 f is indefinite. Thus
the point (1, 2, 4) is neither maximum
nor minimum.
Case 2: x = 1 + 5, y = 2 + 5, z = y(4 y) = (2 + 5)(2 5) = 1. The matrix B
associated with d2 f in this case is

2 2 25
2 2 2 5

2 5 2 5 2

 
 2 1
which again is indefinite as one principle minor is   = 0. Thus (1 + 5, 2 + 5, 1)
2 2 
is also not a maximum or minimum.

Case 3: x = 1 5, y = 2 5, z = y(4 y) = (2 5)(2 + 5) = 1. The matrix B
associated with d2 f in this case is

2 2 25
2 2 2 5

2 5 2 5 2

which again is indefinite for the same reason as Case 2. Thus (1 5, 2 5, 1) is also
not a maximum or minimum.

6
 Note: Consider the similar question: Examine f (x, y, z) = 2xyz 4xz 2yz + x2 + y 2 +
2
z 2x 4y + 4z for extreme values.
For extreme values,
fx = 2yz 4z + 2x 2 = 0 yz 2z + x 1 = 0 (i)
fy = 2zx 2z + 2y 4 = 0 xz z + y 2 = 0 (ii)
fz = 2xy 4x 2y + 2z + 4 = 0 xy 2x y + z + 2 = 0 (iii)
Adding (ii) and (iii) we get x(z + y 2) = 0 x = 0 or y + z = 2.
Case (1): x = 0. Putting x = 0 in (i), (ii) we get yz 2z 1 = 0, and z + y 2 = 0.
Solving these, we get (z + 2)z 2z 1 = 0 z 2 1 = 0 z = 1 y = 3, 1. Thus
(0, 3, 1), (0, 1, 1) are extreme values.
Case (2): y + z = 2. From (i) we get x = 1 + 2z z(2 z). Using (ii) we get
z(1 + z 2 ) z + 2 z 2 = 0 z 3 z = 0 z = 0, 1. Thus (1, 2, 0), (2, 1, 1), (2, 3, 1) are
also extreme values.
Now fxx = 2, fyy = 2, fzz = 2, fxy = 2z, fxz = 2y 4, fyz = 2x 2.
Case (i): x = 0, y = 3, z = 1. The matrix associated with d2 f is

2 2 2
2 2 2
2 2 2
 
2
2 2 
d f is indefinite as one principle minor is   = 0, so there is neither a minimum nor
2 2
maximum at (0, 3, 1).
Case (ii): x = 0, y = 1, z = 1. The matrix associated with d2 f is

2 2 2
2 2 2
2 2 2
 
2
 2 2
d f is indefinite as one principle minor is   = 0, so there is neither a minimum nor
2 2 
maximum at (0, 1, 1).
Case (iii): x = 1, y = 2, z = 0. The matrix associated with d2 f is

2 0 0
0 2 0
0 0 2

d2 f is positive definite so there is a minimum at (1, 2, 0).

Case (iv): x = 2, y = 1, z = 1. The matrix associated with d2 f is

2 2 2
2 2 2
2 2 2

7
  
2
2 2 
d f is indefinite as one principle minor is    = 0, so there is neither a minimum nor
2 2
maximum at (2, 1, 1).
Case (v): x = 2, y = 3, z = 1. The matrix associated with d2 f is

2 2 2
2 2 2
2 2 2
 
2
 2 2
d f is indefinite as one principle minor is   = 0, so there is neither a minimum nor
2 2 
maximum at (2, 3, 1).

Question 4(a) Find the upper and lower Riemann integral for the function defined in the
interval [0, 1] as follows (
1 x2 , x Q
f (x) =
1 x, x 6 Q
and show that f (x) is not Riemann integrable in [0, 1].

Solution. For 0 x 1, (1 x)2 = 1 2x + x2 1 2x2 + x2 = 1 x2 , so 1 x 1 x2 .
For any interval [a, b] [0, 1], 1 b 1 x 1 x2 , a x b. Thus for any partition
P = {0 = t0 < t1 < . . . < tn = 1} ofP [0, 1], we have
Lower Riemann Sum L(f, P ) = ni=1 (1 ti )(ti ti1 ) = L(g, P ) where g(x) = 1 x, 0
x 1. Thus Z 1 Z 1
f (x) dx = sup L(f, P ) = sup L(g, P ) = g(x) dx
0 P P 0

as g is Riemann integrable. Therefore

1 1 1
x2
Z Z
1
f (x) dx = (1 x) dx = x =
0 0 2 0 2
.
Similarly, for any [a, b] [0, 1], 1 a2 1 x2 and since 1 x2 is integrable over
[0, 1], we get
1
Z Z 1 Z
2
2
f (x) dx = inf U (f, P ) = inf U ( 1 x , P ) = 2
1 x dx = cos2 d =
0 P P 0 0 4
R1 R1
Thus f (x) dx 6= 0
f (x) dx, showing that f (x) is not Riemann integrable on [0, 1].
0

8
Question 4(b) Evaluate
ZZ s
a2 b 2 b 2 x 2 a2 y 2
dx dy
a2 b 2 + b 2 x 2 + a2 y 2
x2 y 2
over the positive quadrant of the ellipse + 2 = 1.
a2 b
Solution. Put x = aX, y = bY so that the given integral
r
1 X2 Y 2
ZZ
I = ab dX dY
X 2 +Y 2 1 1 + X2 + Y 2
X0,Y 0

Transforming to polar coordinates by setting X = r cos , Y = r sin so that

Z Z 1r Z 1r
2 1 r2 1 r2
I = ab r dr d = ab r dr
0 0 1 + r2 2 0 1 + r2

Put r2 = cos t so that 2r dr = sin t and

Z 0r Z s
1 cos t ( sin t) 2 1 cos2 t
I = ab dt = ab 2
sin t dt
2
2
1 + cos t 2 4 0 (1 + cos t)
Z Z
ab 2 sin2 t ab 2
= dt = (1 cos t) dt
4 0 1 + cos t 4 0
  2  
ab ab ab
= t sin t = 1 = ( 2)
4 0 4 2 8

which is the required value.