Вы находитесь на странице: 1из 8

UPSC Civil Services Main 1994 - Mathematics

Calculus
Sunder Lal
Retired Professor of Mathematics
Panjab University
Chandigarh

January 9, 2010

Question 1(a) Let f (x) be defined by



1 2 2
2 (b a ),
0<xa
2 a3
f (x) = 12 b2 x6 3x
, a<xb
1 b3 a3

3 x
, x>b

Prove that f (x), f 0 (x) are continuous but f 00 (x) is discontinuous.

Solution. Clearly f (x) is continuous for 0 < x < a, a < x < b and x > b. The problem
points are x = a, b.
1. x = a:
1 1
lim f (x) = lim (b2 a2 ) = (b2 a2 )
xa xa 2 2
1 2 x2 a3 1 a2 a3 1
lim+ f (x) = lim b = b2 = (b2 a2 )
xa xa 2 6 3x 2 6 3a 2
1
Since lim f (x) = lim+ f (x) = f (a) = (b2 a2 ), f (x) is continuous at x = a.
xa xa 2
2. x = b:
1 2 b 2 a3 1
lim f (x) = b = (b3 a3 )
xb 2 6 3b 3b
3 3
1 b a
lim+ f (x) =
xb 3 b
1 3
Again lim f (x) = lim+ f (x) = f (b) = (b a3 ), so f (x) is continuous at x = b.
xb xb 3b

1
Thus f (x) is continuous everywhere.
Clearly f (x) is differentiable for all x > 0 except possibly at x = a, x = b, which we
examine next:
d 1 2 2

1. L.H.D at x = a is (b a ) = 0.
dx 2
d  1 2 x2 a3  a a3
R.H.D at x = a is b = + 2 = 0.
dx 2 6 3x x=a 3 3a
Thus the L.H.D = R.H.D at x = a, therefore f (a) exists and f 0 (a) = 0.
0

d  1 2 x2 a3  b a3 a3 b 3
2. L.H.D at x = b is b = + 2 = .
dx 2 6 3x x=b 3 3b 3b2
1 b 3 a3 a3 b 3
R.H.D at x = b is = .
3 b2 3b2
Thus f (x) is differentiable at x = b also.

Moreover it can easily be seen that



0,
0<xa
0 x a3
f (x) = 3 + 3x2 , a < x b
1 a3 b3

3 x2
, x>b

It is obvious that f 0 (x) is continuous for 0 < x < a, a < x < b and x > b.

a a3
lim f 0 (x) = 0, lim+ f 0 (x) = + 2 = 0 = f 0 (a)
xa xa 3 3a
3 3 3
b a a b
lim f 0 (x) = + 2 = 2
= lim+ f 0 (x) = f 0 (b)
xb 3 3b 3b xb

Thus f 0 (x) is continuous at a and b also, so it is continuous everywhere.



0,
0<x<a
00 1 2a3
f (x) = 3 3x3 , a < x < b
2 a3 b3

3 x3 , x > b
f 00 (x) does not exist at x = a, because LHD of f 0 (x) at x = a is 0, while RHD of f 0 (x)
2a3
at x = a is 31 3a3 = 1.
3
f (x) does not exist at x = b, because LHD of f 0 (x) at x = b is 31 2a
00
3b3
, while RHD of
0 2 a3 b3 2a3 2
f (x) at x = b is 3 b3 = 3b3 + 3 .
Hence f (x) is not continuous at x = a and x = b, since it is not defined at these values.

2
Question 1(b) If , lie between the greatest and the least values of a, b, c, prove that

f (a) f (b) f (c) f (a) f 0 () f 00 ()

(a) (b)
(c) = K (a) 0 () 00 ()
(a) (b) (c) (a) 0 () 00 ()

where K = 21 (b c)(c a)(a b).

Solution. (Note: The question should have been more clearly worded as : Show that there
exist , lying between the greatest and the least values of a, b, c, such that . . . .)
Consider the function

f (a) f (b) f (x)
(x a)(x b) f (a) f (b) f (c)


F (x) = (a) (b) (x) (a) (b) (c)
(a) (b) (x) (c a)(c b)
(a) (b) (c)

Clearly, F (x) is the linear combination of functions f (x), (x), (x) and a polynomial, so
it is twice differentiable note that it is implied from the statement of the question that
f (x), (x), (x) are twice differentiable. Since F (a) = F (b) = F (c) = 0, it follows that F (x)
satisfies the requirements of Rolles theorem in the intervals [a, c], [c, b] (we assume wlog
a < c < b). Thus there exist x1 , x2 such that a < x1 < c, c < x2 < b and F 0 (x1 ) = F (x2 ) = 0.
Applying Rolles theorem to F 0 (x) in the interval [x1 , x2 ] we get a real number (x1 , x2 )
such that F 00 () = 0.
Now
f (a) f (b) f 00 (x) f (a) f (b) f (c)
2
F 00 (x) = (a) (b) 00 (x)

(a) (b) (c)
(a) (b) 00 (x) (c a)(c b) (a) (b) (c)

from which it follows that



f (a) f (b) f 00 () f (a) f (b) f (c)
2
(a) (b) 00 () = (a) (b) (c)
(c a)(c b)
(a) (b) 00 () (a) (b) (c)

f (a) f (b) f (x)

Note that if H(x) = (a) (b) (x) = Af (x) + B(x) + C(x) (say), then H 0 (x) =
(a) (b) (x)
f (a) f (b) f 00 (x) !

Af 0 (x)+B0 (x)+C 0 (x), H 00 (x) = Af 00 (x)+B00 (x)+C 00 (x) H 00 (x) = (a) (b) 00 (x) .
(a) (b) 00 (x)
Now we consider

f (a) f (x) f 00 () f (a) f (b) f (c)
2(x a)
g(x) = (a) (x) 00 () (a) (b) (c)
(a) (x) 00 () (c b)(b a)(c a) (a) (b) (c)

3
Clearly g(a) = 0 and g(b) = 0, as proved above. Thus using Rolles theorem for g(x) in the
interval [a, b] we get (a, b) such that g 0 () = 0. Thus

f (a) f () f 00 () f (a) f (b) f (c)
2
g 0 () = (a) () 00 ()

(a) (b) (c) = 0
(a) () 00 () (c b)(b a)(c a) (a) (b) (c)

or
f (a) f (b) f (c) f (a) f 0 () f 00 ()
1
(a) (b) (c) = (b c)(c a)(a b) (a) 0 () 00 ()
(a) (b) (c) 2

(a) 0 () 00 ()
as required.

Question 1(c) Prove that of all rectangular parallelopipeds of the same volume, the cube
has the least surface.
V
Solution. Let V = xyz, then S = 2(xy + yz + zx) = 2(xy + x
+ Vy ).

S  V  S  V
=2 y 2 , =2 x 2
x x y y

For extreme values, Sx


= Sy
= 0. Thus x2 y = xy 2 = V = xyz x = y = z (as x 6= 0, y 6= 0
being the sides of a parallelopiped).
 2 2
2S 4V 2 S 4V 2 S 2S 2S S 16V 2
Now = , = , = 2 = 4 = 16 4 >
x2 x3 y 2 y 3 x y x2 y 2 x y x3 y 3
0, because V 2 = x2 y xy 2 = x3 y 3 .
Thus S is minimum when x = y = z, V = x3 i.e. when the parallelopiped is a cube.

Question 2(a) Show by means of the beta function that


Z z
dx
I= 1
= , 0<<1
t (z x) (x t) sin

Solution. Put x t = u in I.
Z zt
du
I = . Put u = (z t)v. Then
0 (z t u)1 u
Z 1 Z 1 Z 1
(z t) dv dv
I = 1 (z t)1 (z t) v
= 1 v
= v (1 v)1+ dv
0 (1 v) 0 (1 v) 0
= B( + 1, ) = B(, 1 )
()(1 )
= = ()(1 ) =
( + 1 ) sin
The last step follows from a standard result, which we prove here.

4
Z 1
B(m, 1 m) = (m)(1 m) = xm1 (1 x)(1m1) dx
0
2
Put x = sin so that dx = 2 sin cos d.
Z Z
2 2
2m2 2m
B(m, 1 m) = sin cos 2 sin cos d = 2 tan2m1 d
0 0
dz
Put tan = z sec2 d = dz d = 1+z 2
.
Z
dz
B(m, 1 m) = 2 z 2m1
0 1 + z2
2
Put z = t, so that 2z dz = dt. Thus

tm1
Z
B(m, 1 m) = dt 0<m<1
0 1+t
t
xm1
Z
Consider the integral dx. Put x = y1 so that dx = dy
y2
, and
1 1+x
Z t m1 Z 1 1 m1 Z 1 m
x t
y dy t y
dx = = dy
1 1+x 1 1 + y1 y 2 1 1+y
Z m1 Z 1 m
x y
Letting t , we get dx = dy. Thus
1 1+x 0 1+y
Z m1 Z 1 m1 Z 1 m Z 1 m
x x y y + y m1
I= dx = dx + dy = dy
0 1+x 0 1+x 0 1+y 0 1+y
k+1 k+1
Taylors expansion gives us (1 + y)1 = 1 y + y 2 . . . + (1)k y k + (1)1+yy , 0 < < 1.
Since
Z 1 k+1m Z 1
y + y k+1+m1 1 1
dy < (y k+m + y km+1 ) dy = +
0 1 + y 0 k+m+1 k+2m
1 1
Clearly k+m+1 + k+2m 0 as k , so we get
Z m1 Z 1
x
dx = lim [y m1 y m + y m+1 . . . + (1)k y k+m1 ] dy
0 1 + x k 0
Z 1 
m 1m 2m k km
+ [y y +y . . . + (1) y ] dy
0
(1)k (1)k
   
1 1 1 1
= lim + ... + + + ... +
k m m+1 k+m 1m 2m k+1m
 
1 X 1 1 1 X 2m
= + (1)k = + (1)k 2 2
=
m k=1 m+k km m k=1 m k sin m

by using the identity csc z = z1 + k 2z


P
k=1 (1) z 2 k2 2

5
ZZZ
dx dy dz
Question 2(b) Prove that the value of taken over the volume bounded
(x + y + z + 1)3
1  5 1 256
by the coordinate planes and the plane x + y + z = 1 is log 2 = log 5
2 8 16 e
Solution.
Z 1 Z 1x Z 1xy
dz
I = dx dy
0 0 0 (1 + x + y + z)3
1xy
1 1
Z Z 1x
2
= dx (1 + x + y + z) dy
2 0 0 0
1 1
Z Z 1x  
2 1
= dx (1 + x + y) dy
2 0 0 4
1x
1 1 (1 + x + y)1 y
Z 
= dx
2 0 1 4 0
1 1
Z  
1 1 1x
= (1 + x) dx
2 0 2 4
1
3x x2

1
= log(1 + x) +
2 4 8 0
 
1 5
= log 2
2 8

as required.

Question 2(c) The sphere x2 +y 2 +z 2 = a2 is pierced by the cylinder (x2 +y 2 )2 = a2


(x2 y 2 ).
8a3  5 4 2 
Prove that the volume of the sphere that lies inside the cylinder is +
3 4 3 3
RRR p
pSolution. Clearly V = dx dy dz, where the limits of z are from a2 x2 y 2 to
a2 x2 y 2 and x, y vary over the cylinder D given above. Thus
ZZ p
V =2 a2 x2 y 2 dx dy
D

Because of symmetry, it is enough to compute the integral in the first octant. Let x =
r cos , y = r sin , dx dy = r dr d. (x2 + y 2 )2 = a2 (x2 y 2 ) r4 = a2 r2 cos 2 r2 =
a2 cos 2, so the limits of integration are 0 4 , 0 r a cos 2.

6


Z
4
Z a cos 2
1
V = 8 (a2 r2 ) 2 r dr d
0 0


Z
42 2

3
 1 a cos 2
= 8 (a r2 ) 2 d
0 3 2 0
Z
8 4 3 3
= a a3 (1 cos 2) 2 d
3 0
Z
8a3 h 4 3
3
i
= 2 sin d
2
3 4 0
Z
8a3 h 3 4
i
= 22 sin (1 cos2 ) d
3 4 0
8a3 h 3
h cos3 i 4 i
= 2 2 cos +
3 4 3 0
3h
8a 3
h 1 1 1 ii
= 2 1 +
2
3 4 2 6 2 3
3

8a 5 4 2
= +
3 4 3 3
as required.
Paper II
Question 3(a) Let the function f be defined on [0, 1] by the conditionZ 1 f (x) = 2rx when
2
1
r+1
< x < 1r , r > 0, show that f is Riemann integrable in [0, 1] and f (x) dx = .
0 6
Solution. Clearly f (x) = 2(r 1)x when 1r < x < r1 1
, therefore
limx 1 +0 f (x) = limh0 2(r 1)( r + h) = 2(r 1) r = 2 2r .
1 1
r
limx 1 0 f (x) = limh0 2r( 1r + h) = 2r 1r = 2.
r
Thus f ( 1r 0) 6= f ( 1r + 0), so f is discontinuous at x = 1r , r = 2, 3, . . .. Since the points
of discontinuity have only one limit point, f is Riemann integrable. Now
Z 1 Z 1  r1
X r X
2
f (x) dx = 2rx dx = rx
1 1
0 r=1 r+1 r=1 r+1
 
X 1 1
= 2
r
r=1
r (r + 1)2
     
1 1 2 2 3 3
= + + + ...
12 22 22 32 32 42
1 1 1 2
= 2 + 2 + 2 + ... =
1 2 3 6

7
Question
Z ZZ 3(b) By means of the substitution x + y + z = u, y + z = uv, z = uvw evaluate
(x+y+z)n xyz dx dy dz taken over the volume bounded by x = 0, y = 0, z = 0, x+y+z =
1.

Solution. Clearly the region of integration is the interior of the tetrahedron bounded by
x = 0, y = 0, z = 0, x + y + z = 1. The substitution x + y + z = u, y + z = uv, z = uvw
implies 0 u 1, 0 v 1, 0 w 1, and conversely, when (u, v, w) is in the unit cube,
then (x, y, z) falls in the interior of the tetrahedron. Moreover

vw wu uv vw wu uv
(z, y, x)
= v vw u uw uv = v u 0 = u2 v
(u, v, w)
1v u 0 1 v u 0
Hence
ZZZ
I = (x + y + z)n xyz dx dy dz
x+y+z1
x0,y0,z0
Z 1Z 1Z 1
= n
u (u uv)(uvw)(uv uvw)u2 v du dv dw
Z0 1 Z0 1 Z0 1
= un+5 (1 v)v 3 (1 w)w du dv dw
0 0
Z 10 Z 1
1 41 21
= v (1 v) dv w21 (1 w)21 dw
n+6 0 0
1 (4)(2) (2)(2) 1 1
= =
n + 6 (6) (4) n + 6 5!

Вам также может понравиться