Professional Documents
Culture Documents
Crash Course in Vector Calculus
Crash Course in Vector Calculus
as the radial distance from the origin O and as the angle made by the radial
vector with a reference line (usually chosen to coincide with the x-axis of the
cartesian system). The radial unit vector and the tangential (or angular) unit
vector are taken respectively along the direction of increasing distance and
that of increasing angle respectively, as shown in the figure.
y
^
x2 + y 2
y
= tan1
x
By definition, the distance > 0. we will take the range of angles to be 0 <
2 (It is possible to define the range to be < +). One has to be careful
in using the inverse tangent as the arc-tan function is defined in 0 < . If y
is negative, one has to add to the principal value of calculated by the arc - tan
function so that the point is in proper quadrant.
2
Example 1 :
~ has cartesian components Ax and Ay . Write the vector in terms of its
A vector A
radial and tangential components.
Solution :
Let us write
~ = A + A
A
Since and are orthonormal basis vectors = = 1 and = 0. Thus
~ , A = A
~
Ar = A
Note that (see figure) makes an angle with the x-axis () and /2 with the
y-axis (
). Similarly, the unit vector makes /2 + with the x-axis and with
the y-axis. Thus
~ =
A = A
=
=
~ =
A = A
Ax + Ay
Ax cos + Ay cos(/2 )
Ax cos + Ay sin
Ax + Ay
= Ax cos(/2 + ) + Ay cos
= Ax sin + Ay cos
The Jacobian :
When we transform from one coordinate system to another, the differential element also transform.
For instance, in 2 dimension the element of an area is dxdy but in polar
coordinates the element is not dd
but (d)d. This extra factor is
important when we wish to integrate
a function using a different coordinate system.
d
d
d
x
u
y
u
x
v
y
v
The differentiations are partial, i.e., while differentiating x/u = f (u, v)/u,
the variable v is treated as constant. An useful fact is that the Jacobian of the
inverse transformation is 1/J because the detrminant of the inverse of a matrix is
equal to the inverse of the determinant of the original matrix.
Example 2 : Show that the Jacobian of the transformation from cartesian to polar
coordinates is .
Solution :
We have
x
J =
cos sin
J=
sin cos
Exercise :
Show thatthe Jacobian of the inverse transformation from polar to cartesian is
1/ = 1/ x2 + y 2 .
Example 3 :
find the area of a circle of radius R.
Solution :
Take the origin to be at the centre of the circle and the plane of the circle to be the
plane. Since the area element in the polar coordinates is ddr, the area of
the circle is
" #R
Z 2
Z R
2
= R2
d
d = 2
2 0
0
0
- a very well known result !
Example 4 :
Z
2
2
Find the integral e(x +y ) dxdy where the region of integration is a unit circle
about the origin.
2
Using polar coordinates the integrand becomes e . The range of i integration
for is from 0 to 1 and for is from 0 to 2. The integral is given by
I=
e d = 2
4
e d
1
ew dw/2 = [ew ]10 = (1 )
e
Exercise R: R
Evaluate
xydxdy where the region of integration is the part of the area between
circles of radii 1 and 2 that lies in the first quadrant.
(Ans. 15/8)
Exercise :
R
2
Evaluate the Gaussian integral I = 0 ex dx
[ Hint : The integration cannot be done using cartesian coordinates but is relatively
easy using polar coordinates and propertiesR of definite integrals. By changing the
2
dummy variable x to y, one can write I = 0 ey dy, so that we can write
I2 =
e(x
2 +y 2 )
dxdy
= lim
=
dt t0 t
d
= lim
=
dt t0 t
(t + t) (t)
t0
t
(t + t) (t)
lim
t0
t
lim
One can evaluate the derivatives by laborious process of expressing the unit vectors and in terms of constant unit vectors of cartesian system, differentiating
the resulting expressions and finally transform back to the polar form. Alternatively, we can look at the problem geometrically, as shown in the following figure.
^(t+dt)
^ (t+dt)
d
^ (t)
^ (t)
In the figure, the positions of a particle are shown at time t and t + dt. The unit
vectors is shown in red while the unit vector is shown in blue. It can be easily
seen by triangle law of addition of vectors that the magnitude of
and is
1.d = d. However, as the limit dt 0, the direction of
is in the direction
of while that of is in the direction of
. Thus
d
= lim
=
dt t0 t
dt
d
d
= lim
=
dt t0 t
dt
Now, d/dt is the angular velocity of the point, which is usually denoted by ,
Thus we have,
d
=
dt
d
=
dt
Exercise :
Obtain the above relationships directly by using the expression for the unit vectors
and in terms of the cartesian unit vectors, viz.,
= cos + sin
= sin + cos
6
O
x
x2 + y 2
y
= tan1
x
z = z
Exercise :
[Hint : Determine and
Find the cylindrical coordinate of the point 3 + 4
+ k.
~ = d + d + d
dl
z
and the volume element is
dV = dddz
The Jacobian of transformation from cartesian to cylindrical is as in the polar
coordinates since z coordinate remains the same.
^r
r sin
r cos
The surface of constant r are spheres of radius r about the centre. the surface of
constant is a cone of semi-angle about the z-axis. The reference for measuring is the x-z plane of the cartesian system. A surface of constant is a plane
containing the z-axis which makes an angle with the reference plane.
Example 5 :
8
P
^k
z
^j
+)
= sin k+cos
= cos( +)k+sin(
2
2
^i
^r
x2 + y 2 + z 2
9
= tan
y
x
Range of the variables are as follows :
x2 + y 2
z
= cos1
z
r
= tan1
0 r < 0 0 2
Exercise :
A particle moves along a spherical helix. its position coordinate at time t is given
by
cos t
sin t
t
x=
,
y
=
,
z
=
1 + t2
1 + t2
1 + t2
Express the equation
(Ans.
of the path in spherical coordinates.
r = 1, cos = t/ 1 + t2 (t) = t)
z
d
r sin
r
r sin d
rd
dr
Example 6:
Find the volume of a solid region in the first octant that is bounded from above by
the sphere x2 + y 2 + z 2 = 9 and from below by the cone x2 + y 2 = 3z 2 .
Solution :
Because of obvious spherical symmetry, the problem is best solved in spherical
polar coordinates. The equation to sphere is r = 3 so that the range of r vaiable
for our solid is from 0 to 3.
10
dV =
r 2 sin dddr =
/2
0
/3
sin d
r3
/3
. [ cos ]0 .
=
2
3
9
=
4
"
r 2 dr
#3
0
Exercise :
Using direct integration find the volume of the first octant bounded by a sphere
x2 + y 2 + z 2 = 9.
1.2 Line and Surface Integrals of a Vector Field :
Since a vector field is defined at every position in a region of space, like a scalar
function it can be integrated and differentiated. However, as a vector field has
both magnitude and direction it is necessary to define operations of calculus to
take care of both these aspects.
1.2.1 Line Integral :
If a vector field F~ is known in a certain region of space, one can define a line
integral of the vector function may be defined as
11
~
F~ dl
12
Fy dy =
(1,2,3)
C3
dy = 2
(1,2,0)
z 2 dz =
z3 3
| =9
3 0
C2
(0,0,0)
C1
(1,0,0)
~ =
F~ dl
=
(Fx dx + Fy dy + Fz dz)
1
= 2m3 + m2 + 27
m4 1
| = 9.75
4 0
Example 8 :
Find the line integral of F~ = y + over an anticlockwise circular loop of radius
1 with the origin as the centre of the circle.
Solution :
y
The length element dl has a magnitude 1.d = d. Since the unit vector
~ makes an angle of (/2) +
along dl
with the positive x axis,
dl
R=1
~ = | dl | cos( + )+ | dl | sin( + )
dl
2
2
= sin + cos
In polar coordinates, x = cos and y = sin (since the radius is 1). Thus
Z
~ =
F~ dl
Fx dx + Fy dy
Z
sin d +
= + 0 =
cos d
Exercise :
A force F~ = xy + (x2 z 2 )
xz 2 k acts on a particle. Calculate the work done if
the particle is taken from the point (0, 0, 0) to the point (2, 1, 3) along straight line
segment connecting (0, 0, 0) (0, 1, 0) (2, 1, 0) (2, 1, 3). What would be
the work done if the particle directly moved to the final point along the straightline
connecting to origin.
(Ans. 16, 13.8.)
Exercise :
A vector field is given by
F~ = (2x + 3y) + (3x + 2y)
14
Evaluate the line integral of the field around a circle of unit radius traversed in
clockwise fashion.
(Ans. 6)
Exercise :
Evaluate the line integral of a scalar function xy along a parabolic path y = x2
connecting the origin to the point (1, 1).
q
[]Hint : Remember that the arc length along a curve is given by (dx)2 + (dy)2.
^n
^n
If, on the other hand, the surface is tilted relative to the flow, as shown to the right,
the amount of flux through the area decreases. Clearly, only the part of the area
that is perpendicular to the direction of flow will contribute to the flux.
~ as the dot product of the vector field
We define flux through an area element dS
~ When V~ is parallel to dS,
~ i.e. if the surface is oriented
~ with the area vector dS.
V
perpendicular to the direction of flow, the flux is maximum. On the other hand, a
surface oriented parallel to the flow does not contribute to the flux.
15
Example 9 :
A hemispherical bowl of radius R is oriented such that the circular base is perpendicular to direction of flow. Calculate the flux through the curved surface of the
~ to be constant.
bowl, assumuing the flow vector V
Solution :
Since the curved surface makes different angles at different positions, it is somewhat difficult to calculate the flux through it. However, one can circumvent it by
calculating the flux through the circular base.
V
^n
2
^n
1
^n
1
^n
2
Since the flow vector is constant all over the circular face which is oriented perpendicular to the direction of flow, the flux through the base is R2 V . The
minus sign is a result of the fact that the direction of the surface is opposite to
the direction of flow. Thus we may call the flux through the base as inward flux.
Since there is no source or sink of flow field (i.e. there is no accumulation of
water) inside the hemisphere, whatever fluid enters through the base must leave
through the curved face. Thus the outward flux from the curved face is +R2 V .
We may now generalize the above for a surface over which the field is not uniform
by defining the flux through an area as the sum of contribution to the flux from
infinitisimal area elements which comprises the total area by treating the field to
be uniform over such area elements. Since the flux is a scalar, the surface integral,
defined as the limit of the sum, is also a scalar.
16
X
i
~i n
V
i Si =
^n
i
S i
Vi Si cos i
= V~ dS
S
If the surface is closed, it encloses a volume and we define
=
~ dS
V
Example 10 :
Evaluate the flux through each
~ = xy + yz
A vector field is given by B
+ zxk.
face of a unit cube whose edges along the cartesian axes and one of the corners is
at the origin.
Solution :
Consider the base of the cube
(OGCD), which is the x-y plane on
~ = xy.
which z = 0. On this face B
The surface vector n
is along k direction.
Thus on
this surface flux
R
R
~
Bn
dxdy = xy kdxdy
= 0
z
E
B
O
1
D
C
2
0
0
x
In a similar way one can show that flux from left side (AEOD) is zero while the
contribution from the right side (BFGC) is 1/2. The back face (EFGO) contributes
Z
Vz dxdy =
xdx
dy =
17
zero while the front face (ABCD) contributes 1/2. The net flux, therefore, is 3/2.
Exercise :
z
~ =
Find the flux of the vector field V
2
Ax + By through a rectangular
surface in the x-y plane having dimensions a b. The origin of the coordinate system is at one of the corners of the rectangle and the x-axis
along its length. (Ans. Bab3 /3)
b
y
Example 11 :
Calculate directly the flux through the curved surface of the hemispherical bowl
of Example 10.
Solution :
Use a spherical polar coordinates with the
base of the hemisphere being the x-y plane
and the direction of the vector field as the zaxis. We have seen that an area element on x
the surface is givenR by dS = R2 sin dd.
~ n
/2
sin2
= | V | R .2
2
2
"
cos sin d
#2
^n
zaxis
O
= | V | R2
Exercise :
Find the flux through a hemispherical bowl with its base on the x-y plane and the
origin at the centre of the base. The vector field, in spherical polar coordinates is
~ = r sin
V
r + + .
(Ans. (1 + /2))
18
Example 12 :
A cylindrical object occupies a volume defined by x2 + y 2 R2 and 0 z h.
Find the flux through each of the surfaces when the object is in a vector field
~ = x + y + kz.
V
Solution :
Because of cylindrical symmetry, it is convenient to work in a cylindrical (, , z)
coordinates. The vector field is given by
~ =
V
+ z k
z
^n
Bottom face has z = 0 and the normal to the face points in k direction. Thus the flux from this face
R
R
~ n
V
= Vz dd = 0.
h
The top face has z = h and n
= k.
The flux is
Z
Vz ddr = h
0
2
^n
d
y
R
= h .2 = R2 h
2
^n
x
The normal to the curved face is along direction. An area element on the curved
face is Rddz Thus the flux from this face is
Z
Rddz = R2
(R
+ z k)
2
0
dz = 2R2 h
19
T
T
T
T =
x +
y +
z
x
y
z
T( r + d r )
T
T
T
dx +
dy +
dz
x
y
z
Note that the change dT involves a change in temperature with respect to each of
the three directions. We define a vector called the gradient of T , denoted by T
or grad T as
T T
T
+
+k
T =
x
y
z
using which, we get
dT = T (dx + dy + kdz)
= T d~r
20
T( r )
Note that T , the gradient of a scalar T is itself a vector. If is the angle between
the direction of T and d~r,
dT =| T || d~r | cos = (T )r | d~r |
where (T )r is the component of the gradient in the direction of d~r. If d~r lies
on an isothermal surface then dT = 0. Thus, T is perpendicular to the surfaces
of constant T . When d~r and T are parallel, cos = 1 dT has maximum value.
Thus the magnitude of the gradient is equal to the maximum rate of change of T
and its direction is along the direction of greatest change.
The above discussion is true for any scalar field V . If a vector field can be written
as a gradient of some some scalar function, the latter is called the potential of the
vector field. This fact is of importance in defining a conservative field of force in
mechanics. Suppose we have a force field F~ which is expressible as a gradient
F~ = V
The line integral of F~ can then be written as follows :
Z
~ =
F~ dl
=
V d~l
dV = Vf Vi
where the symbols i and f represent the initial and thec final positions and in the
last step we have used an expression for dV similar to that derived for dT above.
Thus the line integral of the force field is independent of the path connecting the
initial and final points. If the initial and final points are the same, i.e., if the particle
is taken through a closed loop under the force field, we have
I
~ =0
F~ dl
Since the scalar product of force with displacement is equal to the work done by
a force, the above is a statement of conservation of mechanical energy. Because
of this reason, forces for which one can define a potential function are called conservative forces.
Example 13 :
Find the gradient of the scalar function V = x2 y + y 2 z + z 2 x + 2xyz.
21
Solution :
V
V
V
+
+ k
=
x
y
z
= 2(xy + zx + yz)( + + k)
(iii) ln x2 + y 2 + z 2 (Ans. (x + y
+ k)/(x
+ y2 + z2)
Gradient can be expressed in other coordinate systems by finding the length elements in the direction of basis vectors. For example, in cylindrical coordinates the
length elements are d, d and dz along , and k respectively. The expression
for gradient is
1 V
V
V
+
+ k
V =
z
The following facts may be noted regarding the gradient
1. The gradient of a scalar function is a vector
2. (U + V ) = U + V
3. (UV ) = U(V ) + V (U)
4. (V n ) = nV n1 V
Example 14 :
2
2
Find the gradient of V = e(x +y ) in cylindrical (polar) coordinates.
Solution :
2
In polar variables the function becomes V = e . Thus
2
e
=
2
= e .(2) = 2
V
Exercise :
Find the gradient of the function V of Example 15 in cartesian coordinates and
then transform into polar form to verify the answer.
22
Exercise :
kz)/( + z ))
In spherical coordinates the length elements are dr, rd and r sin d. Hence the
gradient of a scalar function U is given by
V = r
V
1 V
1 V
+
+ k
r
r
r sin
Exercise :
Find the gradient of V = r 2 cos cos (Ans. 2r cos cos
r r sin cos
r cos sin .)
Exercise :
V 0
~
F~ dS
V
where V is the volume (enclosed by the closed surface S) in which the point P
at which the divergence is being calculated is located. Since the volume shrinks
to zero, the divergence is a point relationship and is a scalar.
Consider a closed volume V bounded by S. The volume may be mentally broken
into a large number of elemental volumes closely packed together. It is easy to
see that the flux out of the boundary S is equal to the sum of fluxes out of the
surfaces of the constituent volumes. This is because surfaces of boundaries of two
adjacent volumes have their outward normals pointing opposite to each other. The
following figure illustrates it.
23
We can generalize the above to closely packed volumes and conclude that the flux
out of the bounding surface S of a volume V is equal to the sum of fluxes out of
the elemental cubes. If V is the volume of an elemental cube with S as the
surface, then,
1
~ V
F~ dS
V
0
V
S
S
The quantity in the bracket of the above expression was defined as the divergence
of F~ , giving
Z
~ =
F~ dS
XZ
F~ dS = lim
~ =
F~ dS
divF~ dV
24
^n
^n
Fy
F( y + dy) = Fy +
dy
y
(x,y,z) dz
dx
Fy
Fy
dxdydz =
dV
y
y
(x+dx,y,z)
dy
~ =
F~ dS
Fx Fy Fz
+
+
x
y
z
dV
Fx Fy Fz
+
+
x
y
z
3. (F~ ) = F~ + F~
4. In cylindrical coordinates
1
1
(F ) +
F + Fz
F~ =
z
5. In spherical polar coordinates
1
1
1
F~ = 2 (r 2 Fr ) +
(F sin ) +
F
r r
r sin
r sin
6. The divergence theorem is
I
~ =
F~ dS
F~ dV
Example 15 : Divergence of ~r = x + y + kz
Divergence of position vector ~r is very useful to remember.
~r =
x y z
+
+
=1+1+1=3
x y z
One can also calculate easily in spherical coordinate since ~r only has radial component
1
1
~r = 2 (r 2 .r) = 2 .3r 2 = 3
r r
r
Exercise :
Calculate the divergence of the vector field ~r/r 3 using all the three coordinate systems.
(Ans. 0)
Example 16 :
Find the surface integral of the
A vector field is given by F~ = 4xz y 2 + yz k.
field from the surfaces of a unit cube bounded by planes x = 0, x = 1, y = 0, y =
1, z = 0 and z = 1. Verify that the result agrees with the divergence theorem.
Solution :
Divergence of F~ is
Fx Fy Fz
+
+
x
y
z
4xz (y 2 ) yz
=
+
+
x
y
z
= 4z 2y + y = 4z y
F~ =
26
F~ dV =
dx
dy
~ =
F~ dS
F~ dxdz
Z
dx
(4z y)dxdydz =
z
E (0,0,1)
^n = ^j
G (0,1,0)
D
(1,0,0)
dz = 1
3
2
Consider
the top face (ABFE) for which the normal is k so that the surface integral
R
is Fz dydx. On this face z = 1 and Fz = y. The contribution to the surface
integral from this face is
Z 1
Z 1
1
ydy =
dx
2
0
0
For the bottom face (DOGC) the normal is along k and z = 0. This gives Fz = 0
so that the integral vanishes.
R
For the face EFGO the normal is along so that the surface integral is Fx dydz.
On this face x = 0 giving Fx = 0. The surface integral is zero. For the front face
ABCD, the normal is along and on this face x = 1 giving Fx = 4z. The surface
integral is
Z
Z
1
dy
4zdz = 2
Adding the six contributions above, the surface integral is 3/2 consistent with the
divergence theorem.
Exercise :
Verify the divergence theorem by calculating the surface integral of the vector
3 for the cubical volume of Example 17. (Ans. Surface
field F~ = x3 + y 3 + kz
27
div F~ dV =
32 dV = 3
3 d
5
0
dz = 120
In order to calculate the surface integral, we first observe that the end faces have
Since the field does not have any z- component, the
their normals along k.
contribution to surface integral from the end faces is zero.
We will calculate the contribution to the surface integral from the curved surface.
Using the coordinate transformation to cylindrical
x = cos y = sin
and
= cos sin
= sin + cos
k = k
Using these
F~ = 3 (cos4 + sin4 )
+ 3 (sin3 cos cos3 sin )
The area element on the curved surface is Rddz , where R is the radius. Thus
the surface integral is
Z
F dS = R4
2
0
(cos4 + sin4 )d
28
dz
= 16.
3
.5 = 120
2
Fx Fy Fz
+
+
x
y
z
3
3
x
y
z 3
=
+
+
x
y
z
2
2
2
= 3(x + y + z ) = 3r 2
F~ =
where r is the distance from origin. The volume integral over the hemisphere is
conveniently calculated in spherical polar using the volume element r 2 sin dd.
Since it is a hemisphere with z = 0 as the base, the range of is 0 to /2.
F~ dV
= 3
= 3
r dV = 3
r dr
6
1
1 2 =
5
5
/2
sin d
~ = R5
F~ dS
+ R5
/2
0
/2
sin d
cos d + R
cos4 sin d
R /2
R 2
0
/2
sin d
sin4 d
XZ
i
X
i
Ci
~
F~ dl
~
F~ dl
Si
Si
Ci
Si 0
~
F~ dl
n
i
Si
Ci
as the curl of the vector F~ at a point P which lies on the surface Si . Since the
area Si is infinitisimal it is a point relationship. The direction of n
i is, as usual,
30
along the outward normal to the area element Si . For instance, the x-component
of the curl is given by
(curl)x =
lim
y,z0
~
F~ dl
n
i
yz
Ci
Thus
I
~ =
F~ dl
curlF~ dS
This is Stokes Theorem which relates the surface integral of a curl of a vector
~ and dS
~ are fixed by the
to the line integral of the vector itself. The direction of dl
right hand rule, i.e. when the fingers of the right hand are curled to point in the
~ the thumb points in the direction of dS.
~
diurection of dl,
1.5.1 Curl in Cartesian Coordinates :
We will obtain an expression for the curl in the cartesian coordinates. Let us
consider a rectangular contour ABCD in the y-z plane having dimensions y
z. The rectangle is oriented with its edges parallel to the axes and one of the
corners is located at (y, z). We will calculate the line integral of a vector field
F~ along this contour. We assume the field to vary slowly over the length (or
the bredth) so that we may retain only the first term in a Taylor expansion in
computing the field variation.
z
D(y,z+ z)
C(y+ y,z+ z)
^n
A(y,z)
B(y+ y,z)
y
31
Contribution to the line integral from the two sides AB and CD are computed as
follows.
R
~ = R F~ dy = R Fy dy
On AB : F~ dl
R
R
~ = R F~ (
ON CD : F~ dl
)dy = Fy dy
Using Taylor expansion (retaining only the first order term), we can write
Fy
z
z
Thus the line integral from the pair of sides AB and CD is
Fy |CD = Fy |AB +
Fy
Fy
zdy
zy
z
z
In a similar way one can calculate the contributions from the sides BC and DA
and show it to be
Z
Fz
Fz
ydz
yz
y
y
Adding up we get
(curl F )x =
lim
y,z0
Fz
y
Fy
z
yz
yz
Fz
Fy
y
z
In a very similar way, one can obtain expressions for the y and z components
Fx Fz
z
x
Fy Fx
=
x
y
(curl F )y =
(curl F )z
One can write the expression for the curl of F~ by using the del operator as
curl F~ = F~ =
Fx Fy Fz
1 Fz
F
z
!
!
1 F
1
F Fz
+
(F )
k
z
32
F
F~ =
r
(F sin )
r sin
!
1 Fr
1
1
(rF ) +
+
r sin
r
r
Fr
(rF )
r
Example 20 :
y
B(1,1)
A (1,0)
~ =
F~ dl
=
=
Fx |y=0 dx +
x2 dx + 0 + +
Fy |x=0 dy +
x2 dx +
1
1
+0 +2=2
3
3
Fx |y=1 dx +
Fy |x=1 dy
2dy
Fy Fx
=20=2
x
y
Thus
Z
=
( F~ ) kdS
33
2dxdy = 2
(Hint : To calculate the line integral along a straightline, you need the equation
to the line. ForR instance,
the equation to the line BO is y = 2x. Check that
R
~ = ydx + R xdy = 1. )
~ dl
F
AB
Example 21 :
Calculate the line integral of
A vector field is given by F~ = y + z
+ x2 k.
the field along a circular path of radius R in the x-y plane with its centre at the
origin. Verify Stokes theorem by considering the circle to define (i) the plane of
the circle and (ii) a cylinder of height z = h.
Solution :
The curl of F~ may be calculated as
F~ = + 2x
+ k
Because of symmetry, we use cylindrical (polar) coordinates. The transformations
are x = cos , y = sin , z = z. The unit vectors are
= cos sin
= sin cos
k = k
Substituting the above, the field F~ and its curl are given by
F~ =
+
F~ =
+ (sin
+ 2 cos2 ) + k
34
^
k
~ =
F~ dl
~ = R2
F~ dl
^
k
Let us calculate the surface integral of the curl of the field over two surfaces bound
by the circular curve.
(i) On the circular surface bound by the curve in the x-y plane, the outward normal
is along k (right hand rule). Thus
Z
( F~ ) (kdS)
=
dS = R2
(ii) For the cylindrical cup, we have two surfaces : the curved face of the cylinder
The contribution from
on which n
= and the top circular face on which n
= k.
the top circular cap is R2 , as before because the two caps only differ in their
z values (the z-component of the curl is independent of z). The surface integral
from the curved surface is (the area element is Rddz )
Z
Rd
For both the terms of the above integral, the angle integration gives zero. Thus the
net surface integral is R2 , as expected.
Exercise :
Check the validity of the Stokes theorm
A vector field is given by F~ = k2 z .
35
h=4
y
R=2
Exercise :
Let C be a closed curve in the x-y plane in the shape of a quadrant of a circle of
radius R.
y
(0,1,0)
x
(1,0,0)
Example 22 :
A vector field is given by F~ (r, , ) = f (r) where is the azimuthal angle
variable of a spherical coordinate system. Calculate the line integral over a circle
of radius R in the x-y plane centered at the origin. Consider an open surface in the
form of a hemispherical bowl in the northern hemisphere bounded by the circle.
36
Solution :
~ = Rd.
Hence,
On the equatorial circle dl
I
~ =
F~ dl
^r
1
1
(f (r) sin )
r
(rf (r))
r sin
r r !
f (r) cos
f (r) f (r)
=
r
+
r sin
r
r
F~ =
/2
f (r) cos 2
R sin d = 2Rf (R)
R sin
= 2Rf (R)
/2
cos d
Exercise :
Verify Stokes theorem for a vector field 2z + 3x
+ 5y k where the contour is an
equatorial circle of radius R and is anticlockwise when viewed from above and
the surface is the hemisphere shown in the preceding example.. (Ans. 3R2 )
1.6 Laplacian :
Since gradient of a scalar field gives a vector field, we may compute the divergence of the resulting vector field to obtain yet another scalar field. The operator
div(grad) = is called the Laplacian and is written as 2 .
If V is a scalar, then,
2 V
= (V )
!
!
V
V
=
+ + k
+
+ k
x
y
z
x
y
z
2V
V
2V
=
+
+
x2
y 2
z 2
37
Example 23 :
2
=
2
2
2
2
x x + y + z
x (x + y62 + z 2 )3/2
2x2 y 2 z 2
=
(x2 + y 2 + z 2 )5/2
3x2 r 2
=
r5
Adding similar contributions from 2 /y 2 and 2 /z 2 , we get
2
3(x2 + y 2 + z 2 ) 3r 2
3r 2 3r 2
1
=
=
=0
r
r5
r5
1 2
2
=
+ 2 2+ 2
z
!
In spherical :
1
r2
2 = 2
r r
r
+ 2 2
sin
r sin
1
2
r 2 sin2 2
Frequently the Laplacian of a vector field is used. It is simply a short hand notation
for the componentwise Laplacian
2 Fz
2 F~ = 2 Fx + 2 Fy + k
Exercise :
Show that
( F~ ) = ( F~ ) 2 F~
38
is zero everywhere, other than at one point, the integral of the function over any
region which includes this singular point is finite. We define
(x a) = 0 if x 6= a
with
where f (x) is any function that is continuous at x = a, provided that the range of
integration includes the point x = a. Strictly speaking, (x a) is not a function
in the usual sense as Riemann integral of any function which is zero everywhere,
excepting at discrete set of points should be zero. However, one can look at the
function as a limit of a sequence of functions. For instance, if we define a function
(x) such that
(x)
1/
1
for < x < +
2
2
= 0 for | x |>
2
(x) =
/2
0 + /2
Example :
Show that 2 (1/r) is a delta function.
Solution:
As r = x2 + y 2 + z 2 , we have
x
r
y
r
= ,
= ,
x
r
y
r
using this it is easy to show that
r
z
=
z
r
2 1
3x2 r 2
( )=
x2 r
r5
Thus
1
=
r
2
1
2
2
3(x2 + y 2 + z 2 ) 3r 2
( )=
+
+
=0
x2 y 2 z 2 r
r5
!
However, the above is not true at the origin as 1/r diverges at r = 0 and is not
differentiable at that point. Interestingly, however, the integral of 2 (1/r) over
any volume which includes the point r = 0 is not zero. As the value of the
integrand is zero everywhere excepting at the origin, the point r = 0 has to be
treated with care.
Consider an infinitisimally small sphere of radius r0 with the centre at the origin.
Using divergence theorem, we have,
Z
Z
Z
1 3
1
3
2 1
dr=
d r = i dS
r
r
r
V
V
S
where the last integral is over the surface S of the sphere defined above. As the
gradient is taken at points on the surface for which r 6= 0, we may replace (1/r)
with 1/r02 at all points on the surface. Thus the value of the integral is
Z
1
1
2 dS = 2 4r02 = 4
r0 S
r0
Hence
2 1
= 4(r)
r
40