Chapter 2: Solutions to Problems

Problem 2.1.
Work is calculated from the integral of F.dl. In general, the work is path dependent but not always, depending on the
vector F.
F = φ1 , dl = rdr + φ rdφ + zdz , F⋅dl = dφ
r
a. First we transform the coordinates of the points into cylindrical coordinates, (Eq. (1.64)):

r = x 2 + y 2 , φ = tan −1 y , z = z
x
Points P1 (a, 0, 0) to P2 (a, b, c ) are converted into:

P1 r1 = a, φ 1 = 0, 0 , P2 r2 = a2 + b 2 , φ 2 = tan −1 b , z2 =c
a
Performing the integration:
φ2 φ2
−1 b/a
F⋅dl = dφ = φ tan
0 = tan −1 b
φ1 φ1 a

b. Now we integrate from P1 (a, 0, 0) to P2 (a, b, c). First we convert the coordinates of the two points into cylindrical
coordinates:
P1 r1 = a, φ 1 = 0, 0 P2 r2 = a2 + b 2 , φ 2 = tan −1 b , 0
a
The integral between P1 and P2 is:
φ2
F⋅d l = dφ = tan −1 b
φ1 a

Now we integrate between P2 (a, b, 0) to P3 (a, b, c ). Transformation into cylindrical coordinates gives:

P2 r2 = a2 + b 2 , φ 2 = tan −1 b , 0 P3 r3 = a2 + b 2 , φ 3 = tan −1 b , c
a a
and the integral between P2 and P3 is:
P3 φ3
tan−1 b/a
F⋅d l = dφ = φ tan−1 b/a = tan −1 b − tan −1 b = 0
P2 φ2 a a

Note that the work along the two paths is the same.

Problem 2.2.
To see if the integration is path dependent, we perform two integrations. One path is from P1 (0,0,0) through P3 (1,0,0),
P4(1,1,0) and then back to P2 (1,1,1). The second starts also at P1 (0,0,0) then goes through P'3 (0,1,0), P'4 (1,1,0) and then
back to P2 (1,1,1).
(1) P1 0, 0, 0 →P3 1, 0, 0 →P4 1, 1, 0 →P2 1, 1, 1
Between P1 and P3 , dl = x dx, y = 0, between P3 and P4 , dl = y dy, x = 1, and between P4 and P2 , dl = z dz, x = 1, y = 1.
Therefore, performing the scalar product, we get:
P2 P3 P4 P2 1 1 1

A ⋅dl = A ⋅dl + A⋅dl + A ⋅dl = 0dx + 2dy + dz = 3

P1 P1 P3 P4 0 0 0

(2) P1 0, 0, 0 →P3" 0, 1, 0 →P4" 1, 1, 0 →P2 1, 1, 1

Between P1 and P'3 , dl = y dy, x = 0, between P'3 and P'4 , dl = x dx, y = 1, and between P'4 and P2 , dl = z dz. Therefore,
performing the scalar product, we get:
" "
P2 P P P2 1 1 1
3 4
A ⋅d l = A ⋅d l + A⋅d l + A ⋅d l = 0dy + dx + dz = 2
" "
P1 P1 P3 P4 0 0 0

18
Thus, because two different paths yield different results, the integration is path dependent.

Problem 2.3.
The element of length dl is written, followed by the scalar product A.dl. Then this is integrated along the paths specified.

A = x2 − y5 d l = xdx + ydy + zdz → A ⋅d l = 2dx − 5dy

a. From a to b along the parabola:
0 0

2dx − 5dy = 2 + 10 = 12
x=−1 y=2
or, since: y = 2x 2 , dy = 4xdx
0 0
0
2dx − 5dy = 2dx − 20xdx = 2x + 10x 2 −1 = 0 − − 2 − 10 = 12
x = −1 x = −1

b. From a to c, then to b :
1 2 0 0

2dx − 5dy + 2dx − 5dy = 4 − 0 + − 2 + 10 = 12

x=−1 y=2 x=1 y=2

c. From b to a :
−1 2

2dx − 5dy = −12

x=0 y=0

The work in the closed path a →c →b →a is equal to zero.

∴ All three paths give the same result. The work is independent of this particular path.

Problem 2.4. z φ=π /3

The elements of surface on each of the 6 surfaces of the
volume are calculated and then the scalar product with the z=2
φ=π /6
vector A is evaluated, all in cylindrical coordinates. The flux
A
is evaluated as the surface integral over the six surfaces to
yield the total net flux. Although in this case the flux must ds2
ds1
be zero, we will calculate it explicitly to show that it is.The z=1
vector A and the elements of surface are: r=1 r =2 Figure A
A = r4z, d S r = ± rrdφdz , dS φ = ± φdr dz , dS z = ± z rdφdr
Because A is in the r direction, the scalar product A.dl is zero on the top and bottom surfaces (at z = 1 and z = 2) as well
as on the φ = π/3 and φ = π/6 surfaces:
A⋅dS r = 0, A ⋅dS z = 0
However, on the surfaces r = 1 and r = 2, the scalar product is nonzero.
On the left surface (r = 1 ), the element of surface points in the negative r direction while A points in the positive r
direction:
π/3 2
π ≤ φ ≤ π, 1 ≤ z ≤ 2, r = 1, ψ1 = A⋅ −dS1 = r4z⋅ −rrdφdz = − 4r z dz dφ
6 3 φ=π/6 z=1
π/3 π/3 π/3
2
= − 4r z2 dφ = − 4r 3 dφ = − 6r dφ = − 6rπ = − π
φ=π/6 2 1 φ=π/6 2 φ=π/6 6

where we set r=1. On the right surface (r = 2 ), the element of surface points in the positive r direction while A points in
the positive r direction:

19
 π/3 2
π ≤ φ ≤ π, 1 ≤ z ≤ 2, r = 2, ψ2 = A⋅ dS2 = r4z⋅ −rrdφdz = 4r z dz dφ
6 3 φ=π/6 z=1
π/3 π/3 π/3
2
= 4r z2 dφ = 4r 3 dφ = 6r dφ = 6rπ = 2π
φ=π/6 2 1 φ=π/6 2 φ=π/6 6

Thus, the total flux is the sum of the two fluxes (all other surfaces contribute zero flux):
ψ = ψ1 + ψ 2 = − π + 2π = π

Problem 2.5. S2
The problem statement defines two surfaces. One
S1
(defined as 0 ≤ R ≤ 1, θ = π/6, 0 ≤ φ ≤ 2π) is a conical π/6
surface (similar to an ice cream cone, see Figure A). 1
The second is a hemisphere of radius 1, topping the
cone (Figure A). Integrate the spherical surface and
the conical surface separately and add together. Figure A

A.ds = A.ds + A .ds = A .θRsinθdRdφ + A .RR 2 sinθdθdφ

S1 S2 S1 S2
2π 1 2π π/6
R π 1 dR dφ +
2
= RθsinθdRdφ + R sinθdθdφ = sinθdθ dφ
S1 S2 0 0 6 2 0 0
2 1 2
= R π 2π π /6
+ 2π (−cosθ) 0 = π + π (2− 3)
2 6 2 0 12

where R = 1 was substituted. For the given values:

2
A.ds = π − π ( 3 − 2)
12

Problem 2.6. y
y dxdz
Integration should be performed on each side of the 1
cube separately. With the aid of Figure A, the ds2
integrals on the various sides are as follows: − z dxdy
1. On the right side, ds = x dydz and the limits are 0 ≤ − xdydz x dydz
y ≤ 1, 0 ≤ z ≤ 1. Thus, the surface integral is: ds 3 ds 1
x
z dxdy 1
1 1 1 1

A.ds1 = xx 2 + yy 2 + zz 2 .xdydz = x 2dydz = x 2 z 1

S1 y =0 z =0 y=0 z=0 − y dxdz
Since x = 1 on this side: Figure A.

A.ds1 = 1
S1

2. On the left surface, ds = −x dydz. However, because x = 0 on this surface, the integral is also zero.
3. On the upper surface, ds = y dxdz, and the limits are 0 ≤ x ≤ 1, 0 ≤ z ≤ 1. The integral on this surface is:
1 1 1 1

A.ds2 = xx 2 + yy 2 + zz 2 .ydxdz = y 2 dxdz = y 2

S2 x=0 z=0 x=0 z=0

Since y = 1 on this surface,

A.ds2 = 1
S2

4. On the front surface, ds = z dxdy and the limits are 0 ≤ x ≤ 1, 0 ≤ y ≤ 1. The integral is:

20
 1 1 1 1

A.ds3 = xx 2 + yy 2 + zz 2 .zdxdy = z 2 dxdy = z 2

S3 x=0 y=0 x=0 y=0

Since z = 1 on this surface:

A.ds3 = 1
S3

On the bottom and back surfaces, the integral contributes nothing because y and z are zero, respectively

A.ds = 3
s

Problem 2.7.
To calculate the flux of a vector field we must perform a surface integral over the product A.ds, where ds is the element
of area in the direction normal to the surface. First we evaluate the element of area ds. This is ds = n ds , where the unit
normal vector is:
n = k = k = z1
k
Thus, ds is:
ds = z ds = zrdrdφ
The scalar product A.ds:
A.ds = r5 + z3 .zrdrdφ = 3rdrdφ
Thus, the flux is:
a 2π a
Φ= A .ds = 3rdrdφ = 6π rdr = 3π a2
S r=0 φ=0 r=0

Problem 2.8.
In (a) we must integrate over the volume of the cylinder to find the total mass. In (b), we first transform the mass density
equation into spherical coordinates to simplify integration over the sphere.
a. The mass density equation is:
ρ v = r2 + ra + z 2 + zd
The total mass over a cylinder of radius a and length d centered at the origin is:
2π a +d
2
Q= r2 + ra + z 2 + zd rdrdφdz
φ=0 r=0 z=− d
2
The integral over φ equals 2π since the density is independent of φ. Thus:
a +d +d +d
Q = 2π
2
r3 + r 2 a + z 2 r + zdr drdz = 2π
2
r4 + r3 a + z 2 r2 + zd r 2 a dz = 2π 2
a4 + a4 + z 2 a2 + zd a 2 dz
r=0 z= − d z= − d 4 3 2 2 0 z= − d 4 3 2 2
2 2 2
4 4 3 2 2 2 +d 4 4 3 2 3 2 4 4 3 2 3 2
= 2π a z + a z + z a + z d a 2 = 2π a d + a d + d a + d a − 2π − a d − a d − d a + d a
4 3 6 4 z= − d 8 6 48 16 8 6 48 16
2
4 4 2 4 23 3
= 2π a d + a d + d a = 7π a d + π d a
4 3 24 6 12
b. First we transform from cylindrical to spherical coordinates using Eq. (1.93):
r = R sin θ , φ = φ, z = R cos θ
substituting these into the expression for mass density we get:
ρ v = R 2 sin 2 θ + Ra sin θ +R 2 cos 2 θ + Rd cosθ
2
The element of volume in spherical coordinates is dv = R sinθdRdθdφ .
The mass density, integrated over the volume of the sphere of radius a, centered at the origin gives the total mass:

21
 R=a π 2π
Q= R 2 sin 2 θ + Ra sin θ +R 2cos 2 θ + Rd cos θ R 2sin θdRdθd φ
R =0 θ =0 φ =0
R=a π
2 2 2 2
= 2π R sin θ + Ra sin θ +R cos 2 θ + Rd cos θ R sin θdRdθ
R =0 θ =0
π
a5 sin 3 θ + a5 sin 2 θ +a5 cos 2 θsin θ + a4d sin θcos θ dθ
= 2π
θ =0 5 4 5 4
5 3 5 5 3 4 2 π 5 5 5
= 2π a −cos θ + cos θ + a θ − sin 2 θ a
+ − cos θ + a d sin θ = 2π a 2 1 − 1 + a π + a 2
5 3 4 2 4 5 3 4 2 0 5 3 4 2 5 3
5 5 5 5 5 5 5 5 2 5 5 5 2
= 2π 4 a + a π + 2a = 2π 4a + a π + 2a = 8a π + 2a π + 4a π = 4a π + a π
35 8 15 15 8 15 15 8 15 5 4
5 5 2
The mass of the sphere is: Q = 4a π + a π = 4.98a 5
5 4

Problem 2.9. z
In order to find the total number of particles, the
particle volumetric distribution N v must be integrated
a
r (z) = a h− b z + b
over the entire volume. h0
0
N= N vdv where: Nv = 105 r3 +103 r(h−h0 )2
v
First, we must know what the volume itself looks like
(see Figure A). This is a 2-dimensional plot of the r
dz
cone and the cone is symmetrical. The equation of
the line defining the side of the cone gives the radial r
distance from the axis of the cone: b
Figure A.
r = a − b z +b
h0
Because of the circular nature of the cone we use cylindrical coordinates. Thus, we will first integrate with respect to r ,
then with respect to z and finally with respect to φ . The ranges of integration are:

r from 0 to a − b z +b , z from 0 to h0 , φ from 0 to 2π .

h0
Thus, we can determine N by performing the triple integral of N v over the volume. Also, recall that for cylindrical
coordinates, the element of volume is dv = rdrdzdφ :

2π h0 a − b z +b 2π h0
5
N=
h0
105 r3 + 103 r(z − h0 )2 rdrdzdφ = 20,000 a − b z +b + 1000 z − h0 2 a − b z +b 3 dz dφ
φ=0 z=0 r=0 φ=0 z=0 h0 3 h0
2π

= 50 600a 5 + 600ba 4 + a 3 h02 + 600a 3 b 2 + 3a 2 bh02 + 600a 2 b 3 + 6ah02 b 2 + 600ab 4 + 10h02 b 3 + 600b 5 h0 dφ
0 9

= 100π h0 600a 5 + 600ba 4 + a 3 h02 + 600a 3 b 2 + 3a 2 bh02 + 600a 2 b 3 + 6ah02 b 2 + 600ab 4 + 10h02 b 3 + 600b 5
9

Problem 2.10.
By direct volume integration on each of the scalar components of the vector.
a. Total force is:
1 1 1 1 1 1 1 1 1

F= Fdv = x 2xy dxdydz + y z dxdydz + z y 2 dxdydz = z 8

v x=−1 y=−1 z=−1 x=−1 y=−1 z=−1 x=−1 y=−1 z=−1 3
b. Total force is calculated as in (a) but the limits on the three coordinates is between zero and 2:

22
 2 2 2 2 2 2
F= Fdv = x 2xy dxdydz + y z dxdydz +
v x=0 y=0 z=0 x=0 y=0 z=0
2 2 2

z y 2 dxdydz = x 16 + y 8 + z 32 [N ]
x=0 y=0 z=0 3

Problem 2.11.
We use the definition of acceleration as time derivative of velocity to set up the integral. Integration over time gives the
velocity. Thus,
a = dv → dv = adt
dt
5 5 5 5
3 2
3tdt = x t − t + y 3t = x162 + y371
2 2 2
v= adt = x (t − 2t) + y3t dt = x (t − 2t)dt + y
t t 0 0 3 0 2 0 3 2
m
s
There is also a constant of integration V 0 (velocity at time t = 0) but we can take this to be zero. Thus:

v = x 162 + y371 m
3 2 s

Problem 2.12.
This is a regular vector integral along a line and therefore we integrate each of its scalar components separately as
follows:
F= r2 dl = x 2 + y 2 xdx + ydy = x x 2 + y 2 dx +y x 2 + y 2 dy
C C C C

where C is the path of integration (the straight line between the origin and P(1,3). The equation of this line is found as
follows. Since it is a straight line, its general form is y=ax+b. Thus:
at (0,0) → y = ax + b → 0 = a0 + b → b=0
at (1,3) → y = ax + b → 3=a → a=3
or:
y = 3x and x=y
3
Substituting for y in the first integral and for x in the second integral we get:
x=1 y=3

F=x x 2 + y 2 dx + y x 2 + y 2 dy = x x 2 + 9x 2 dx + y y 2 + y 2 dy
C C x=0 y=0 9
x=1 y=3 2 33
10y dy = x10x 31
=x 10x 2dx + y + y10y = x10 + y1 0 [N]
x=0 y=0 9 3 0 27 0 3
Thus:
r2 d l = x 10 + y10 N
C 3

Problem 2.13.
First find the gradient of V = xy 2 + yz. Substitute the coordinates (1,1,2) to evaluate the gradient at the given point. To
find the derivative in the direction of A, we first find the unit vector in the direction of A and then take the scalar product
between the gradient and the unit vector:
The scalar may be written as V = xy 2 + yz. Its gradient is:

∇V = xy 2 + y(2xy + z) + zy
∇V(1,1,2) = x + y4 + z
The unit vector in the direction of A:

23
 A = x 2 − y + z2 → A = x 2 − y + z2 = x 2 − y + z2
x2 − y + z2 3
The projection of the gradient of V on A is:
∇V. A = x + y4 + z . x2 − y + z2 = 1 (2 − 4 + 2) = 0
3 3

Problem 2.14.
The gradient is calculated directly from the formula for gradient in Cartesian coordinates. The direction of the gradient
equals the unit vector in the direction of the gradient:
a. The gradient of P(x,y,z) is:
∇P(x,y,z) = x ∂P(x,y,z) + y ∂P(x,y,z) + x ∂P(x,y,z) =
∂x ∂y ∂z
2 2 2
d (x − 2) d (y − 2) d (z + 1)
x +y +x = x 2(x − 2) + y 2(y − 2) + z 2(z + 1)
dx dy dz
The magnitude of the gradient is:
∇P = 2 (x − 2)2 + (y − 2)2 + (z + 1)2 = 2 P

The direction is (unit vector in the direction of ∇P:

∇P x (x − 2) + y (y − 2) + z (z + 1)
∇P = =
∇P (x − 2)2 + (y − 2)2 + (z + 1)2
b. The derivative of the pressure in the normal direction is:

∇P.z = 2(z+1)
c. The derivative at 45° degrees is the scalar product between the unit vector at 45° and the gradient. The unit vector at
45° is:
a=x+y
2
The derivative in this direction is:
∇P.a = x 2(x − 2) + y 2(y − 2) + z 2(z + 1) . x + y = 2 (x + y − 4)
2

Problem 2.15.
The gradient of f(r,φ,z) is calculated directly from the definition of the gradient in cylindrical coordinates. Then the
vector obtained is transformed into Cartesian coordinates and into spherical coordinates. Alternatively, the scalar
function f(r,φ,z) may be transformed first into Cartesian and spherical coordinates and then the gradient calculated in the
corresponding system of coordinates. We show the first method.
a. The gradient of f(r,φ,z) in cylindrical coordinates is:

∇f = r ∂f + φ 1 ∂f + z ∂f = r cos 2 φ + φ −2r cos φ sin φ + z cos φ + zsin φ

∂r r ∂φ ∂z r

b. The transformation from cylindrical to Cartesian coordinates is (see Eq. (1.70)):

Ax cosφ −sinφ 0 Ar cosφ −sinφ 0 cos 2 φ
Ay = sinφ cosφ 0 Aφ = sinφ cosφ 0 (−2r cos φ sin φ + z cos φ )/r
Az 0 0 1 Az 0 0 1 sin φ
The scalar components in Cartesian coordinates are:

24
 A x = cos 3 φ + 2 cosφ sin 2 φ − z cos φ sinφ
r
A y = cos 2 φ sinφ − 2 cos 2 φ sinφ + z cos 2 φ
r
A z = sinφ
Before writing the gradient, we need to transform the coordinates r, φ, and z, into Cartesian coordinates. For this purpose
we write Eq. (1.63):
x y
x = rcosφ , y = rsinφ , → r = x 2 + y2 , cosφ = , sinφ =
x + y2
2
x + y2
2

Writing the various values, the vector A in Cartesian coordinates becomes:

x3 2xy 2 xyz x 2y 2x 2 y x 2z y
∇f = x + − + y − + +z
3/2 2 3/2 2 3/2 2 3/2 2 3/2 3/2
x +y 2
2 2
x +y 2
x +y 2
x +y 2
x +y x +y 2
2
x + y2
2
3 2 2 2
= x x + 2xy −3/2xyz + y x z − x3/2 y +z y
2 2 2 2
x +y x +y x + y2
2

c. To convert the gradient into spherical coordinates we use the transformation in Eq. (1.92)):
AR sinθ 0 cosθ Ar sinθ 0 cosθ cos 2 φ
−2r cosφ sinφ + z cosφ
Aθ = cosθ 0 − sinθ Aφ = cosθ 0 − sinθ
r
Aφ 0 1 0 Az 0 1 0 sinφ
The scalar components become:

A R = cos 2 φ sinθ + cosθ sinφ , Aθ = cos 2 φ cosθ − sinθ sinφ

A φ = −2 cos φ sinφ + z cosφ
r
Transformation of coordinates from cylindrical to spherical coordinates (Eq. (1.93)):
r = Rsinθ , φ = φ, z = Rcosθ

Substituting into the scalar components and writing the vector A in spherical coordinates gives:

∇f = R cos 2 φ sinθ + cosθ sinφ + θ cos 2 φ cosθ − sinθ sinφ − φ 2 cos φ sinφ − cosθ cosφ
sinθ

Problem 2.16.
Write the equation of the plane as a scalar function of the variables. Then the gradient of the scalar functions produces a
vector normal to the plane. Normalization of the vector by its magnitude produces the required unit vector.
a. The equation of the plane may be written as a constant value function:
f(x,y,z) = 5x + 3y + z=0
The gradient is normal to the constant value function:

∇f(x,y,z) = x ∂(5x) + y ∂(3y) + z ∂(z) = x5 + y3 + z1

∂x ∂y ∂z
and the normal unit vector is:
n = x5 + y3 + z1 = x 5 + y3 + z1 = x 5 + y 3 + z 1
x5 + y3 + z1 52 + 3 2 + 1 35 35 35

b. The constant value function is:

f(x,y,z) = 4x − 3y + z = 5
The gradient is:

25
 ∇f(x,y,z) = x ∂(4x) + y ∂(−3y) + z ∂(z) = x4 − y3 + z1
∂x ∂y ∂z
and the normal unit vector is:
n = x 4 − y3 + z1 = x 4 − y 3 + z 1
42 + 3 2 + 1 26 26 26

c. f(x,y,z) = ax + by − z = 0

∇f(x,y,z) = x ∂(ax) + y ∂(by) + z ∂(−z) = xa + yb − z1

∂x ∂y ∂z

n = x a + yb − z1
a2 + b 2 + 1
To check that this is correct, find a vector A in the plane and then calculate the scalar product:

A ⋅ n̂ = 0
To calculate the vector A, we need two points in the plane.
For (a): P1=(1, -1, -2) and P2 = (-1, 1, 2). The vector A is:
A = −x̂2 + ŷ2 + ẑ4
# 5 3 1 & 10 6 4
A ⋅ n̂ = (−x̂2 + ŷ2 + ẑ4 ) ⋅ % x̂ + ŷ + ẑ (=− + + =0
$ 35 35 35 ' 35 35 35
The same can be shown in case (b).

Problem 2.17.
Write the equation of the surface as a scalar function of the variables. Then the gradient of the scalar functions produces
a vector normal to the surface. Normalization of the vector by its magnitude produces the required unit vector.
a. The equation of the surface may be written as a constant value function:
f(x,y,z) = 3xy + yz + z = 0
The gradient is normal to the constant value function:

∇f(x,y,z) = x ∂(3xy) + y ∂(3xy + yz) + z ∂(yz +z ) = x3y + y( 3x+z) + z(y+1)

∂x ∂y ∂z
and the normal unit vector is:
n = x 3y + y( 3x+z) + z(y+1)
9y 2 + ( 3x+z)2 + (y+1)2
b. Using identical steps as in (a):
f(x,y,z) = x − z 2 − y 2 = 0
2 2
∇f(x,y,z) = x ∂(x) + y ∂(−y ) + z ∂(−z ) = x1 − y2y − z2z
∂x ∂y ∂z
x 1 − y2y − z2z
n=
1 + 4y + 4z 2
2

c. Using identical steps as in (a):

f(x,y,z) = x 2 + y 2 + z 2 = 8
2 2 2
∇f(x,y,z) = x ∂(x ) + y ∂(y ) + z ∂(z ) = x2x + y2y + z2z
∂x ∂y ∂z
n = x 2x + y2y + z2z = x x + yy + zz
4x 2 + 4y 2 + 4z 2 x 2 + y2 + z2

26
Problem 2.18.
2 2
a. ∇. xx 2 +y1−zy 2 = ∂(x ) + ∂(1) − ∂(y ) = 2x
∂x ∂y ∂z
2 2
b. ∇. r2z +φ 5r−z3r = 1
2 2 ∂(2z r) +1 ∂(5r) − ∂(3r ) = 2 z 2
r ∂r r ∂φ ∂z r
2 2 2 2
∂ x +z ∂ x + y x y
c. ∇. x x 2 + z 2 +y x 2 + y 2 = + = +
∂x ∂y x + z2
2
x + y2
2

Problem 2.19.
By direct application of the formula for divergence in Cartesian coordinates.
∂ x2 ∂ y2 ∂ z2
∇⋅ xx 2 + yy 2 + zz 2 = + + = 2x + 2y + 2z
∂x ∂y ∂z
∇⋅A = 2×1 + 2× (−1) + 2× 2 = 4

Problem 2.20.
By direct application of the divergence in cylindrical coordinates:
∂ rAr ∂A φ ∂A z
∇.A = ∇. r2rcosφ − φ rsinφ + z4z = 1 +1 + =
r ∂r r ∂φ ∂z
2
1 ∂ 2r cosφ − 1 ∂ rsinφ + ∂ 4z = 4cosφ − cosφ + 4 = 3cosφ + 4
r ∂r r ∂φ ∂z
At (2,90°,1):
∇.A = 3cosφ + 4 = 3cos(π /2) + 4 = 4

Problem 2.21.
By direct application of the divergence in spherical coordinates:
∂ R 2A R ∂ A θ sin θ ∂A φ
∇.A = 12 + 1 + 1 =
R ∂R R sin θ ∂θ R sin θ ∂φ
5 2 3 3 3 2
1 ∂ 0.2R φ sin θ + 1 ∂ 0.2R φ sin θ + 1 ∂ 0.2R φ sin θ
R2 ∂R R sin θ ∂θ R sin θ ∂φ
= R 2 φ sin 2 θ + 0.6R 2 φ sin θ cos θ + 0.2R 2 sin θ
At (2,π/6,π/2):
2
∇.A = 4 π 1 + 0.6(4) π 1 3 + 0.2(4) 1 = 3.6
2 2 2 2 2 2

Problem 2.22.
To verify the divergence theorem we integrate both sides of the theorem and show they are the same.
The divergence theorem is given as:
∇.Adv = A .ds
v s

First we evaluate the divergence of A. However, because we will need to integrate the divergence as well as the vector
itself over a cylindrical surface of radius 3 and length 4, it is easier to do so in cylindrical coordinates. Thus, we
transform the vector into cylindrical coordinates using Eq. (1.71):
A r = Axcos φ + Aysin φ = 4x cos φ − 2y 2 sin φ
A φ = −A xsin φ + Aycos φ = − 4x sin φ − 2y 2 cos φ
A z = − 2z 2
The coordinates transformation is (Eq. (1.63)):
x = r cos φ , y = r sin φ , z= z

27
Substituting these, the components of the vector in cylindrical coordinates are:
A r = 4r cos 2 φ − 2r 2 sin 3 φ , Aφ = − 4r cos φ sin φ − 2r 2 sin 2 φ cos φ , Az = − 2z 2
The divergence of the vector A in cylindrical coordinates is:
∂ rAr ∂A φ ∂A z
∇⋅A = 1 +1 + =
r ∂r r ∂φ ∂z
3 2 3
8cos 2 φ − 6rsin φ + 4sin φ − 4cos 2 φ − 4r sin φ cos 2 φ + 2r sin φ − 4z =
= 4 − 4r sin 3 φ − 4r sin φ cos 2 φ − 4z
Now we integrate this over the volume of the cylinder of radius 3 and length extending from z = −2 to z = 2, using
dv = rdrdφ dz as the element of volume in cylindrical coordinates:
3 2π 2

∇.Adv = 4 − 4r sin 3 φ − 4r sin φ cos 2 φ − 4z rdrdφdz

v r=0 φ=0 z=−2
2π 2 2π

= 18 − 36sin 3 φ − 36sin φ cos 2 φ − 18z dφ dz = 72 − 144sin 3 φ − 144sin φ cos 2 φ = 144π

φ=0 z=−2 0

Now we turn to the right hand side. First we write the elements of area for the cylinder as:

dSr = rrdrdz, dSφ = φdrdz, dS z = z rdφdr

Thus, on the top of the cylinder we use dBz on the bottom −dBz and on the side, dBr . Integrating over the three surfaces
separately we get:

On the side (circular surface), the conditions are: r = 3, −2 ≤ z ≤ 2. Substituting these into the components of the vector
we get the contribution of this surface:
2π 2 2π 2
A.dsr = 4r cos 2 φ − 2r 2 sin 3 φ rdφdz = 36 cos 2 φ − 18sin 3 φ rdφdz = 144π
sr φ=0 z=−2 φ=0 z=−2

On the top surface: z = 2:

2π 3 2π 3
A.dsz = − 2z 2 rdφdr = − 8rdφ dr = − 72π
stop φ=0 r=0 φ=0 r=0

On the bottom surface: x = −2:

2π 3 2π 3
A.dsz = − 2z 2 − rdφ dr = 8rdφ dr = 72π
sbottom φ=0 r=0 φ=0 r=0

Thus:
A.ds = A .ds + A .ds + A .ds = − 72π + 72π + 144π = 144π
s top bottom side

Since both sides yield the same result, the divergence theorem is verified.

Problem 2.23.
To show that the divergence theorem applies we write the divergence theorem and evaluate each of its sides separately.
The divergence theorem states:
∇.Adv = A .ds
v s

First we need to evaluate the divergence of A. Since A only has an R component, its divergence is:

∇.A = 12 ∂ (R 2 ×R) = 3
R ∂R
2
The element of volume in spherical coordinates is dv = R sinθdRdθdφ. Thus:
28
 π 2π a π a

∇.Adv = 3R 2 sinθdRdθdφ = 6π R 2 sinθdRdθ =

v θ=0 φ=0 R=0 θ=0 R=0
π π
3 a
2π R sinθdθ R=0 = 2πa 3sinθdθ = − 2πa 3cosθ π0 = 4πa 3
θ=0 θ=0

On the right hand side we can evaluate the surface integral provided the element of surface is first written. The element
of surface on a sphere of radius R is:
ds = RR 2 sinθdθdφ
The right hand side becomes (setting R = a):
π 2π π 2π π
A.ds = Ra . Ra 2 sinθdθdφ = a3 sinθdθdφ = 2π a3 sinθdθ = − 2π a3 cosθ π0 = 4π a3
s θ=0 φ=0 θ=0 φ=0 θ=0

Since the two sides yield identical results, the theorem is verified for the given vector.

Problem 2.24. y
y dxdz
a. Integration should be performed on each side of 1
the cube separately. With the aid of Figure A, the ds2
integrals on the various sides are as follows: − z dxdy
1. On the right side, ds = x dydz and the limits are − xdydz x dydz
0 ≤ y ≤ 1, 0 ≤ z ≤ 1. Thus, the surface integral is: ds 3 ds 1
x
1 1 z dxdy 1
2 2 2
A.ds1 = xx + yy + zz .xdydz
S1 y=0 z=0
z 1
1 1 − y dxdz Figure A.
2 2
= x dydz = x
y=0 z=0

A.ds1 = 1
Since x = 1 on this side: S1

2. On the left surface, ds = −x dydz. However, because x = 0 on this surface, the integral is also zero.
3. On the upper surface, ds = y dxdz, and the limits are 0 ≤ x ≤ 1, 0 ≤ z ≤ 1. The integral on this surface is:
1 1 1 1

A.ds2 = xx 2 + yy 2 + zz 2 .ydxdz = y 2 dxdz = y 2

S2 x=0 z=0 x=0 z=0

A.ds2 = 1
Since y = 1 on this surface, S2

4. On the front surface, ds = z dxdy and the limits are 0 ≤ x ≤ 1, 0 ≤ y ≤ 1. The integral is:
1 1 1 1

A.ds3 = xx 2 + yy 2 + zz 2 .zdxdy = z 2 dxdy = z 2

S3 x=0 y=0 x=0 y=0

A.ds3 = 1
Since z = 1 on this surface: S3

On the bottom and back surfaces, the integral contributes nothing because y and z are zero, respectively. The closed
surface integral is the sum of the integrals on the 6 sides of the cube:

A.ds = 3
s
b. First we find the divergence of A in Cartesian coordinates:
∂A y ∂A z ∂(x 2 ) ∂(y 2 ) ∂(z 2 )
∇.A = ∂A x + + = + + = 2x + 2y + 2z
∂x ∂y ∂z ∂x ∂y ∂z

29
The volume integral is:
1 1 1 1 1 1

∇.Adv = 2x + 2y + 2z dxdydz = 1 + 2y + 2z dydz = 1 + 1 + 2z dz = 3

v x=0 y=0 z=0 y=0 z=0 z=0
Thus:
∇.Adv = A .ds
v s
as required.

Problem 2.25.
By direct application of the formula for curl in Cartesian coordinates for (a) and (c) and in cylindrical coordinates for
(b).
∂A z ∂A y ∂A x ∂A z ∂A y ∂A x
∇×A = x − +y − +z − =
a. ∂y ∂z ∂z ∂x ∂x ∂y
∂(− y 2 ) ∂(1) ∂(x 2 ) ∂(− y 2 ) ∂(1) ∂(x 2 )
x − +y − +z − = − x2y
∂y ∂z ∂z ∂x ∂x ∂y
∂B ∂ rBφ
∇×B = r 1 ∂Bz − φ + φ ∂Br − ∂Bz + z 1 − ∂Br
r ∂φ ∂z ∂z ∂r r ∂r ∂φ
b.
= r 1 ∂ −3r 2 − ∂ 5r + φ ∂ 2
2z − ∂ 2
−3r + z 1 ∂ 2
5r − ∂ 2
2z = φ(4z + 6r ) + z10
r ∂φ ∂z ∂z ∂r r ∂r ∂φ
∂Cy ∂Cy ∂Cx
∇×C = x ∂Cz − + y ∂Cx − ∂Cz + z − =
c. ∂y ∂z ∂z ∂x ∂x ∂y
2 2 2 2 2 2 2 2
x ∂( 0) − ∂ x +y + y ∂ x +z − ∂( 0) + z ∂ x +y − ∂ x +z = y z +z x
∂y ∂z ∂z ∂x ∂x ∂y x 2 +z 2 x 2 +y 2

Problem 2.26.
a. See Figure A.
b. The curl of v in cylindrical coordinates is:

& 1 ∂vz ∂vφ )

∇ × v = r̂ ( −
& ∂v ∂v ) 1 & ∂ rvφ
+ φ̂ ( r − z + + ẑ (
( )
∂v )
− r+
' r ∂φ +
∂z * ' ∂z ∂r * r ' ∂r ∂φ *

& ∂ (1 / r ) ) 1 & ∂ (1) )

= r̂ ( − + ẑ ( =0
' ∂z +* r ' ∂r +*

The curl of v is zero because the vector v does not Figure A

vary in either the r or z direction.

Problem 2.27.
By direct evaluation of the curl in Cartesian coordinates. If the curl is zero, the field is conservative.
a. The curl of A in Cartesian coordinates is:
∂A y ∂A y ∂A x
∇× A = x ∂A z − + y ∂A x − ∂A z + z −
∂y ∂z ∂z ∂x ∂x ∂y
Since A has only a y component, and this component only depends on x and z, we can write:
∂A y ∂A y ∂ 3x cos (ω t + 50z ) ∂ 3x cos (ω t + 50z )
∇×A = x − +z =x − +z =
∂z ∂x ∂z ∂x
x 150x sin (ω t + 50z ) + z 3 cos (ω t + 50z )
b. A is not conservative since ∇×A ≠ 0.
30
Problem 2.28.
By direct evaluation of the curl. In (a) this is done in Cartesian coordinates. In (b) it is done in cylindrical coordinates.
∂A y ∂A y ∂A x
∇×A = x ∂A z − + y ∂A x − ∂A z + z − =
a. ∂y ∂z ∂z ∂x ∂x ∂y
= x 0 − 2(x+1)y + y z 2 + z 2yz − 2y = −x2(x+1)y + yz 2 + z 2yz − 2y
∂A φ ∂(rAφ ) ∂A r
∇×A = r 1 ∂A z − + φ ∂A r − ∂A z + z1 − =
r ∂φ ∂z ∂z ∂r r ∂r ∂φ
2
b. r 1 ∂(3) − ∂(4rsinφ ) + φ ∂(2rcosφ ) − ∂(3) + z1 ∂(− 4r sinφ ) − ∂(2rcosφ ) =
r ∂φ ∂z ∂z ∂r r ∂r ∂φ

z 1 − 8rsinφ + 2rsinφ =− z6sinφ

r

Problem 2.29.
We start with Stokes' theorem and evaluate each side of the theorem separately. If the two sides are equal, the theorem is
verified for the given vector.
Stokes' theorem states:
∇×A.d s = A .d l
s C

To evaluate the left hand side we need to first find the curl of A in Cartesian coordinates. This is:
∂A y ∂A y ∂A x
∇× A = x ∂A z − + y ∂A x − ∂A z + z − = x − 4zy + 4zy + y 0 − 0 + z 0 + 1 = z
∂y ∂z ∂z ∂x ∂x ∂y
Now, the surface over which we need to integrate is the upper half of the sphere x 2 + y 2 + z 2 = 4. The normal unit
vector to the sphere is R but, to be able to integrate over the sphere we must first transform the vector ∇× A into
spherical coordinates. This is accomplished as follows (see Eq. (1.89)):
AR sinθcosφ sinθsinφ cosθ 0
Aθ = cosθcosφ cosθsinφ − sinθ 0
Aφ − sinφ cosφ 0 1
This gives:
A R = cosθ, A θ = − sinθ, Aφ = 0
and the vector in spherical coordinates is:
∇×A = Rcosθ − θsinθ
The element of area in the R direction on the sphere is:
2
dsR = RR sin(θ)dθdφ
Thus, the surface integral becomes an integral over the open surface defined by the half sphere:

∇×A.ds = R cosθ − θsinθ . RR 2 sin(θ)dθdφ = R 2 sinθ cosθ dθ dφ =

S S S
π/2 2π π/2 π /2
2
R2 sinθ cosθ dθ dφ = 2π R 2 sinθ cosθ dθ = 2π R 2 sin θ =π R 2
θ=0 φ=0 θ=0 2 θ =0
For R = 2 we get:
∇×A.d s = 4π
s

31
Now we turn to the closed contour integral over the rim of the half sphere. Again, since the sphere is given in Cartesian
coordinates, the rim is:
x 2 + y2 = r2 = 4
This suggests that cylindrical coordinates should be used to integrate along the contour. Thus we transform the vector A
into cylindrical coordinates using Eq. (1.71)):
Ar cosφ sinφ 0 2x− y
Aφ = −sinφ cosφ 0 −2yz 2
Az 0 0 1 −2zy 2
The scalar components of A in cylindrical coordinates are:
A r = 2x− y cosφ − 2yz 2 sinφ , A φ = − 2x− y sinφ − 2yz 2 cosφ , A z = −2zy 2
Now we transform the coordinates x, y, z using Eq. (1.63):
x = rcosφ , y = rsinφ , z=z
This gives:
A = r 2rcosφ −rsinφ cosφ − 2rsinφ z 2 sinφ − φ 2rcosφ −rsinφ sinφ + 2rsinφ z 2 cosφ −z2r 2 zsin 2 φ
This is simplified somewhat from the fact that we integrate it at z = 0 (on the x-y plane). Thus we can substitute z = 0:

A = r 2rcosφ −rsinφ cosφ − φ 2rcosφ −rsinφ sinφ at z=0

The integration is along the rim of the half sphere which may be defined as dl = φr dφ . Now we can write the integral on
the right hand side of Stokes' theorem as:

A.dl = r 2rcosφ −rsinφ cosφ − φ 2rcosφ −rsinφ sinφ . φr dφ =

C C
2π

− r 2rcosφ −rsinφ sinφ dφ =

2
r2 sin φ −2r 2 sinφ cosφ dφ = r 2
φ − sin2φ − sin 2 φ 2 π = r 2 π
C 0 2 4 0
Since r = 2:
A.dl = 4π
C

and therefore the theorem is verified for the given vector.

Problem 2.30.
Parts (a) and (b) are evaluated directly. From Stokes theorem there are two ways of solving the third part of the problem.
One is to calculate ∇×F, find the normal components of ∇×F by taking the scalar product between ∇×F and ds, and
integrate over the surface of the hemisphere. The second is to integrate the vector F on the closed contour integral
formed by the hemisphere on the x-y plane. The latter is considerably simpler but we will show the former since they are
equivalent.
∂F 2
a. ∇⋅F = ∂Fx + y + ∂Fz = ∂(3y) + ∂(5−2x) + ∂(z −2) = 2z
∂x ∂y ∂z ∂x ∂y ∂z
2 2
b. ∇×F = x ∂(z −2) − ∂(5−2x) + y ∂(3y) − ∂(z −2) + z ∂(5−2x) − ∂(3y) = 0 + 0 − z5 = − z5
∂y ∂z ∂z ∂x ∂x ∂y
c. Converting the curl of F from Cartesian coordinates to spherical coordinates we get:

∇×F = R (−5cos θ) + θ(5sin θ)

To calculate the surface integral we also need the element of surface perpendicular to the R direction. The latter is:
2
ds R = R R sin θdθdφ
Thus, setting the radius of the sphere R = 2:

32
 ∇×F ⋅dsR = R (−5cos θ) + θ(5sin θ) ⋅ R4sin θdθdφ =
s s
φ=2π θ=π θ=π π /2
2 2 2
−20sin θcos θdθdφ = −40π sin θcos θdθ = −40π sin θ = −20π
φ=0 θ=0 θ=0 2 0

Problem 2.31.
There are a number of methods that can be used to solve this problem. We will show here two methods.
Method A. Calculate the components of the vector ∇×F perpendicular to the x-y, x-z and y-z planes. Integrating these
components over the projections of the triangular surface on these planes gives the total flux through the triangular area.
First we need to calculate the vector ∇×F :

∂Fy ∂F ∂F
∇× F = x ∂Fz − + y ∂Fx − ∂Fz + z y − ∂Fx = − x y − y ∂Fz − z ∂Fx = − x − y − z
∂y ∂z ∂z ∂x ∂x ∂y ∂z ∂x ∂y
Now, the elements of area perpendicular to the three planes, assuming the surface vector points out of the “volume”
defined between the planes and the triangular area (see Figure A):

dsx = − x dydz, dsy = − y dxdz, dsz = − z dxdy

Now we can write the flux as follows:

Φ= ∇×A .dsx + ∇×A .dsy + ∇×A .dsz =

sx sy sz

− x − y − z . − x dydz + − x − y − z . − y dxdz + − x − y − z . − z dxdy

sx sy sz

= 1dydz + 1dxdz + 1dxdy

sx sy sz

To integrate we use Figures B through D and write:

For sx, using Figure B
z= b c− y
c
For sz, using Figure C
y = c a− x
a
For sy, using Figure D
x= ab− z
b
With these we can now write:
c bc−y b ab−z a c a−x
c b a
Φ= 1dydz + 1dxdz + 1dxdy = 1dydz + 1dxdz + 1dxdy =
sx sy sz y =0 z =0 z =0 x =0 x =0 y =0
c b a c b a
bc−y ab−z c a−x
zdy cz =0 + xdz b + ydx ay =0 = b c − y dy +
a b − z dz + c a − x dx =
x =0
y =0 z =0 x =0 y =0
c z =0 b x =0 a
2 c b 2 a
b cy − y + a bz − z
2
+ c ax − x = bc + ab + ac = ab + bc + ca
c 2 0 b 2 0 a 2 0 2 2 2 2
The answer is:
ab + bc + ca
Φ =
2
The flux can be negative or positive because the direction of the normal can be taken in either perpendicular direction.

33
 y
(0,c,0) P3 ds z
z y
(a,0,0) b c
ds x
P1 x
z y
P2
z (0,0,b) c y a x
ds y y c-y x a-x
Figure A Figure B Figure C

y
(0,c,0) P3
x
n
a (a,0,0)
x P1 x
z P2
b
x b-x z (0,0,b)
Figure D Figure E
Method B.
First we find the vector normal to the triangle by defining two vectors along its sides, calculating the vector product
between them and then finding the unit vector in the direction of the normal. This gives the direction of ds. On the
triangular area. Then we calculate (∇×F).ds and integrate over the projection of the surface on one of the planes (x-y, x-z
or y-z).
Taking a vector between P1 and P2 and one between P1 and P3 we get (Figure E):
Figure E

P 1 P 2 = x −a + z b , P 1 P 3 = x −a + y c , P 1 P 3 × P 1 P 2 = x bc + y ab + z ac
This is the vector normal to the triangle so that its projections on the axes are positive. The normal unit vector to the
triangle is (see Figure E) therefore:
x bc + y ab + z ac
n=
b 2c2 + a 2b 2 + a 2c2
Now we argue as follows: the flux through ds is F.ds and may be calculated by integrating over the projection of the
surface on the x-y, x-z or y-z plane. That is:

(∇×F).ds = (∇×F).ndxdz = (∇×F).ndxdy = (∇×F).ndydz (1)

s sxz n.y sx y n.z syz n.x
where the normal was calculated above and sxz is the surface of the triangle formed by P1, P2 and the origin, sxy is the
surface of the triangle formed by the P1, P3 and the origin while syz is formed by P2, P3 and the origin. We choose
(arbitrarily) the first of these. First we calculate the curl of F:
∂Fy ∂F ∂F
∇× F = x ∂Fz − + y ∂Fx − ∂Fz + z y − ∂Fx = − x y − y ∂Fz − z ∂Fx = − x − y − z
∂y ∂z ∂z ∂x ∂x ∂y ∂z ∂x ∂y
The two scalar products in (1) are:

x bc + y ab + z ac − bc − ab − ac , ab
(∇×F).n = − x − y − z . = n.y =
b 2c2 + a 2b 2 + a 2c2 b 2c2 + a 2b 2 + a 2c2 b 2c2 + a 2b 2 + a 2c2
Thus:
F.n dxdz = − bc − ab − ac dxdz = − c − 1 − c dxdz (2)
sxz n.y x z ab x z
a b

34
The limits of integration (see Figure A) are as follows. We can take the limits for x from zero to a but the limits for z
need to be calculated from the equation of the line connecting points P1 and P2. The later is:

z + (b/a)x − b = 0
Thus, the limits of integration for z are zero to z = (b − (b/a)x). With these, we can write:
a b − (b/a)x a a

F.n dxdz = − c − 1 − c dxdz = −c− 1− c

b − (b/a)x
dx = − c − 1 − c b − b x dx
sxz n.y x=0 z=0 a b x=0 a b 0 x=0 a b a
a
2 x=a
= − bc − b − c + bc2 + b + c x dx = − bc − b − c x + bc2 + b + c x = − bc + ba + ca
x=0
a a a a a a a a 2 x=0 2
The negative sign simply indicates the direction of the flux with respect to the normal to the surface (that is, the normal
points upwards, but ∇×F is negative). If we were to calculate the normal as P 1 P 2 × P 1 P 3 instead P 1 P 3 × P 1 P 2 , the
normal would have been negative and the flux positive. Thus, we may write:

(∇× F).ds = ± ba + ca + bc
s 2
The same result may be obtained by using Stokes' theorem and integrating along the path bordering the triangular area.
In this particular case, this is actually somewhat simpler. Stokes' theorem is:

∇×F.ds = F.dl
s c

where c is the contour around the triangle in Figure A. We integrate on the three segments separately. First, the scalar
product F.dl is:
F.dl = x y + yz + zx . x dx + ydy + zdz = ydx + zdy + xdz
To integrate, we note the following:
1. On the segment between P2 and P1, the following holds:

z = b − (b/a)x, y=0 → dz = − (b/a)dx, dy = 0

On this segment we have:
F.dl = ydx + zdy + xdz = xdz = − (b/a)xdx
2. On the segment P1 and P3:

y = c − (c/a)x, z=0 → dy = − (c/a)dx, dz = 0

On this segment we have:
F.dl = ydx + zdy + xdz = ydx = − c − (c/a)x dx
3. On the segment P3 and P2:

z = b − (b/c)y, x=0 → dz = − (b/a)dy, dx = 0

On this segment we have:
F.dl = ydx + zdy + xdz = zdy = − (b − (b/c)y)dy
With these we can write:
P1 P3 P2 a a c

F.dl = F .dl + F .dl + F .dl = − (b/a)xdx − c − (c/a)x dx − (b − (b/c)y)dy

c P2 P1 P3 x=0 x=0 y=0
2x=a 2x=a 2 y=c
− by − by
2 2
= − bx − cx − cx = − ba − ca − ca − bc − bc = − ba + ca + bc
2 x=0 2a x = 0 2c y = 0 2 2a 2c 2
This is the same result (including the sign) and was obtained by integrating in the positive direction (so that the normal is to
the left of the path of integration). A negative flux is obtained if the integration is taken in the opposite direction. Thus:

ab + bc + ca
!∫C
F ⋅ dl = ±
2

35
Nevertheless, in Method A we obtained a positive value. Here we got a negative value. This simply indicates that the flux
coming through the triangular area must exit through the sides or vice versa. Therefore the two methods give opposite signs
as expected.

Problem 2.32.
a. Operations (1), (4), (5), (8), (9), (10) are valid. All others are not. Operation (2), (3), (6) and (7) are not defined.
Operation (11) cannot be defined because the gradient does not operate on a vector while operation (12) is not defined
because the curl cannot operate on a scalar.
2 2 2
b. (1) ∇⋅ ∇Φ = ∇⋅ x ∂Φ + y ∂Φ + z ∂ Φ = ∂ Φ2 + ∂ Φ2 + ∂ Φ2
∂x ∂y ∂z ∂x ∂y ∂z
2 2 2 2 2 2
(4) ∇ × ∇Φ = ∇ × x ∂ Φ + y ∂ Φ + z ∂ Φ = x ∂ Φ − ∂ Φ + y ∂ Φ − ∂ Φ + z ∂ Φ − ∂ Φ ≡ 0
∂x ∂y ∂z ∂y ∂z ∂z ∂y ∂z ∂x ∂x ∂z ∂x ∂y ∂y ∂x
∂ 2
A z ∂2 A y ∂ 2
A x ∂ 2
A z ∂2 A y ∂ 2
A x
(5) ∇⋅ ∇ × A = − + − + − ≡0
∂x ∂y ∂x ∂z ∂y ∂z ∂y ∂x ∂z ∂x ∂z ∂y
(8) ∇× ∇× A = ∇ ∇⋅A − ∇2 A = ∇ ∇⋅A − x∇ 2 A x − y∇ 2 A y − z∇ 2 A z
2
∂2 A 2
where: ∇ ∇⋅A = x ∂ A2x + y 2y + z ∂ A2z
∂x ∂y ∂z
2 2 2 2 2 2
∂ A ∂ A ∂ A ∂2 A z + ∂2 A z + ∂2 A z
∇2 A x = ∂ A2x + ∂ A2x + ∂ A2x , ∇ 2A y =
y y y 2
+ + , ∇ A z=
∂x ∂y ∂z ∂x 2 ∂y 2 ∂z 2 ∂x 2 ∂y 2 ∂z 2
Thus:
∇× ∇×A =
2 2 2 2 2 2 2 2 2
∂ A ∂ A y ∂ A
x 2x + y 2 + z 2z − x ∂ A x
+ ∂ A x
+ ∂ A x − y ∂ A y + ∂ A y + ∂ A y − z ∂2 A z + ∂2 A z + ∂2 A z
∂x ∂y ∂z ∂x 2 ∂y 2 ∂z 2 ∂x 2 ∂y 2 ∂z 2 ∂x 2 ∂y 2 ∂z 2
We use the identity in Eq. (2.140), by first substituting B = ∇×A . With this, we have:
∇ ⋅ (Φ∇ × A ) = ∇ ⋅ (Φ B) = B ⋅ ( ∇Φ ) + Φ ( ∇ ⋅ B) = ( ∇ × A ) ⋅ ( ∇Φ ) + Φ ( ∇ ⋅ ( ∇ × A )) = ( ∇ × A ) ⋅ ( ∇Φ ) =
) # ∂Az ∂Ay & # ∂Ax ∂Az & # ∂Ay ∂Ax &, ) ∂Φ ∂Φ ∂Φ ,
(9) +*x̂ %$ ∂y − ∂z (' + ŷ %$ ∂z − ∂x (' + ẑ %$ ∂x − ∂y ('.- ⋅ +*x̂ ∂x + ŷ ∂y + ẑ ∂z .- =
∂Φ $ ∂Az ∂Ay ' ∂Φ $ ∂Ax ∂Az ' ∂Φ $ ∂Ay ∂Ax '
& − )+ & − )+ & − )
∂x % ∂y ∂z ( ∂y % ∂z ∂x ( ∂z % ∂x ∂y (

where we made use of the fact that ∇⋅(∇×A) = 0 (see Eq. (2.113)).
∂A y ∂A y ∂A x
(10) Φ∇× A = x Φ ∂A z − + y Φ ∂A x − ∂A z + z Φ −
∂y ∂z ∂z ∂x ∂x ∂y

Problem 2.33.
By direct application of the Laplacian of a scalar field to the two fields.
2
∂2 Φ (x,y,z) + ∂2 Φ (x,y,z) =
∇2 Φ (x,y,z) = ∂ Φ (x,y,z) +
∂x 2 ∂y 2 ∂z 2
2 2 2 2 2 2 2 2
∂ (x−2) (y−2) (z+1) ∂ (x−2) (y−2) (z+1) ∂ (x−2)2 (y−2)2 (z+1)2
2
a. + + =
∂x 2 ∂y 2 ∂z 2
2(y−2)2 (z+1)2 + 2(x− 2)2 (z+1)2 + 2(x− 2)2 (y−2)2

36
 ∇2 5r cos φ + 3zr 2 =
2 2
∂ 5r cos φ + 3zr 2 ∂ 5r cos φ + 3zr 2 ∂ 5r cos φ + 3zr 2
b. + 1 + 12
∂r 2 r ∂r r ∂φ 2
2
∂ 5r cos φ + 3zr 2
+ = 6z + 1 5cos φ + 6zr + 1 −5r cos φ = 12z
∂z 2 r r2

Problem 2.34.
In Cartesian coordinates it is best to apply the equation for the vector Laplacian directly. In Cylindrical coordinates it is
2
best to use the vector identity ∇×(∇×A) = − ∇ A + ∇(∇. A) .

a. ∇2 A = x∇ 2 A x + y∇ 2 A y + z ∇2 A z
where:
2 2 2
∂2 A y ∂2 A y ∂2 A y 2 2 2
∇2 A x = ∂ A2x + ∂ A2x + ∂ A2x , ∇ 2A y = + + , ∇ 2 A z = ∂ A2z + ∂ A2z + ∂ A2z
∂x ∂y ∂z ∂x 2 ∂y 2 ∂z 2 ∂x ∂y ∂z
Thus:
2 2 2 2
∇ A x = 0, ∇ A y =0, ∇ Az = 2 → ∇ A = z2

b. The vector Laplacian in cylindrical coordinates may be best calculated as:

∇2 A = ∇(∇. A) − ∇×(∇×A)
We need to evaluate the divergence and the curl of A in cylindrical coordinates. By direct application of the divergence
of A in cylindrical coordinates:
∂ rAr ∂A φ ∂A z
∇.A = ∇. r2rcosφ − φ 4rsinφ + z3 = 1 +1 + =
r ∂r r ∂φ ∂z
2
1 d 2r cosφ − 1 d 4rsinφ + d 3 = 4cosφ − 4cosφ = 0
r dr r dφ dz

By application of the curl of a vector field in cylindrical coordinates (twice), we get:

∂A φ ∂A φ
∇ × A = r 1 ∂A z − + φ ∂A r − ∂A z + z 1 A φ + − 1 ∂A r =
r ∂φ ∂z ∂z ∂r r ∂r r ∂φ
∂(3) ∂(−4rsinφ ) ∂(2rcos φ ) ∂(3) ∂(−4rsinφ ) 1 ∂(2rcosφ )
r1 − +φ − + z 1 (−4rsinφ ) + − =
r ∂φ ∂z ∂z ∂r r ∂r r ∂φ
z −4sinφ − 4sinφ + 2sinφ = − z6sinφ
Now, repeating the process to obtain ∇×(∇× A):

∇ × ∇ × A = − r 6 cos φ
r
Thus, the vector Laplacian of A is:
∇2 A = r 6 cos φ
r

Problem 2.35.
This relation may be deduced in general from the following:
∇2 F = ∇ ∇⋅F − ∇ × ∇ × F
Since ∇ × F = 0, the second term on the right hand side is zero and the relation as stated holds. This may also be shown
to be correct in general by writing the various terms explicitly.
For the left hand:
∂Fφ
∇2 F = r ∇2 Fr − 12 Fr − 22 + φ ∇2 Fφ − 12 Fφ − 22 ∂Fr + z∇ 2 Fz (1)
r r ∂φ r r ∂φ

37
where
2 2 2
∇2 Fr = ∂ F2r + 1 ∂Fr + 12 ∂ Fr2 + ∂ F2r
∂r r ∂r r ∂φ ∂z
2 2 2
∂ Fφ 1 ∂Fφ + 1 ∂ Fφ + ∂ Fφ
∇ 2 Fφ = + (2)
∂r 2 r ∂r r 2 ∂φ 2 ∂z 2
2 2 2
∇2 Fz = ∂ F2z + 1 ∂Fz + 12 ∂ F2z + ∂ F2z
∂r r ∂r r ∂φ ∂z
For the right hand:
∂Fφ ∂Fφ
∇.F = 1 ∂ rFr + 1 + ∂Fz = ∂Fr + 1 Fr + 1 + ∂Fz (3)
r ∂r r ∂φ ∂z ∂r r r ∂φ ∂z
2 2 2
F ∂F ∂ F
+ Fz
∂ ∂F ∂
∇ ∇⋅F = r 1 − 12 Fr + 1 − 12 +1
r r φ φ
r ∂r 2 r r ∂r r ∂φ r ∂r ∂φ ∂r ∂z
2
F ∂2 Fφ 1 ∂2 Fz 2
F ∂2 Fφ 2
+ ∂ F2z
1 ∂ r 1 ∂F r 1 ∂ r 1 ∂F r 1
+φ + 2 + 2 + + z + + (4)
r ∂φ ∂r r ∂φ r ∂φ 2 r ∂φ ∂z ∂z ∂r r ∂z r ∂z ∂φ ∂z
Now we write:
∇ ∇⋅F − ∇2 F = 0
or:
2
∂2 Fr 1 Fr + 1 ∂Fr − 1 ∂Fφ + 1 ∂ Fφ + ∂ Fz
2
∇ ∇⋅F − ∇ F = r 1
2
−
r ∂r 2 r2 r ∂r r2 ∂φ r ∂r ∂φ ∂r ∂z
2 2 2 2 2
+ φ 1 ∂ Fr + 12 ∂Fr + 12
∂ Fφ
+ 1 ∂ Fz + z ∂ Fr + 1 ∂Fr + 1 ∂ Fφ
+ ∂2 Fz
r ∂φ ∂r r ∂φ r ∂φ 2 r ∂φ ∂z ∂z ∂r r ∂z r ∂z ∂φ ∂z 2
∂Fφ
− r ∇2 Fr − 1 Fr − 2 − φ ∇2 Fφ − 1 Fφ − 2 ∂Fr − z ∇2 Fz = 0
2 2
r r ∂φ r2 r2 ∂φ
Collecting terms, and substituting from (2):
2
∂F ∂2 Fφ 2
∇ ∇⋅F − ∇2 F = r 1 ∂ F2r − 12 Fr + 1 ∂Fr − 12 φ + 1 + ∂ Fz
r ∂r r r ∂r r ∂φ r ∂r ∂φ ∂r ∂z
2 2 2 2 2 2
∂ Fφ 1 ∂ Fz ∂ Fφ
+ φ 1 ∂ Fr + 12 ∂Fr + 12 + + z ∂ Fr + 1 ∂Fr + 1 + ∂ F2z
r ∂φ ∂r r ∂φ r ∂φ 2 r ∂φ ∂z ∂z ∂r r ∂z r ∂z ∂φ ∂z
2 2 2 2 2 2
∂Fφ ∂ Fφ 1 ∂Fφ ∂ Fφ ∂ Fφ
− r ∂ F2r + 1 ∂Fr + 12 ∂ Fr2 + ∂ F2r − 12 Fr − 22 −φ + + 12 + − 12 Fφ − 22 ∂Fr
∂r r ∂r r ∂φ ∂z r r ∂φ ∂r 2 r ∂r r ∂φ 2
∂z 2
r r ∂φ
2 2 2
∂ F ∂F ∂ F ∂ F
−z z
+ 1 z
+ 12 z
+ z
=0
∂r 2 r ∂r r ∂φ 2 ∂z 2
Now, returning to the term ∇ × ∇ × F :
∂Fφ ∂F
∇ × F = r 1 ∂Fz − + φ ∂Fr − ∂Fz + z 1 Fφ + φ − 1 ∂Fr
r ∂φ ∂z ∂z ∂r r ∂r r ∂φ
Thus:
2 2 2 2 2
∂ Fφ 1 ∂ Fφ
∇ × ∇ × F = r 12 + − 12 ∂ F2r − ∂ F2r + ∂ Fz
r ∂φ r ∂φ ∂r r ∂φ ∂z ∂z ∂r
2 2 2 2
∂ F ∂ F ∂F ∂F ∂ F ∂ Fφ
+φ1 z
− φ
− 1 Fφ − 1 φ
− 1 +1 r
−
r ∂φ ∂z ∂z 2 r2 r ∂r r2 ∂φ r ∂r ∂φ ∂r 2
2 2 2 2
∂ Fφ
+ z 1 ∂Fr + ∂ Fr − 1 ∂Fz − ∂ F2 z − 12 ∂ F2z + 1 (6)
r ∂z ∂r ∂z r ∂r ∂r r ∂φ r ∂φ ∂z
And clearly:
∴ (5) − (6) = (1) → ∇ 2 F = ∇ ∇⋅F − ∇ × ∇ × F
and, since ∇ × F = 0, the result is correct.

Problem 2.36.
38
By direct application of the formulas in Cartesian coordinates:

a. ∇f = x ∂f + y ∂f + z ∂f = x 4x + y 1
∂x ∂y ∂z
∇.(fR) = ∇. 2x 2 + y xx+yy+zz = ∇. x(2x 3 + xy)+y(2x 2 y + y 2 )+z(2x 2 z + yz )
∂(fR)y ∂(fR)z d(2x 3 + xy) d(2x 2 y + y 2 ) d(2x 2 z + yz )
b. = ∂(fR)x + + = + +
∂x ∂y ∂z dx dy dz
2 2 2 2
= 6x + y + 2x + 2y + 2x + y = 10x + 4y
2 2 2
c. ∇2 f = ∂ f2 + ∂ f2 + ∂ f2 = 4
∂x ∂y ∂z
d. ∇2 R = x ∇2 R x + y ∇2 R y + z ∇2 R z = 0

x y z
∂ ∂ ∂
e. ∇×(fR) = = x z + y 0 + z 4xy − x0 + z(4xy − x) = x z − y4xz + z(4xy − x)
∂x ∂y ∂z

2x 3 + xy 2x 2 y + y 2 2x 2 z + yz

or: ∇× (fR) = x z − y4xz + z(4xy − x)

Problem 2.37.
Calculate the curl and the divergence of A:

∇ × A = x̂ (0 − 0) + ŷ (0 − 0) + ẑ (0 − 0) = 0

∇ ⋅ A = 5+ 2 = 7
The vector A is a solenoidal, irrotational vector field.

Problem 2.38.
Calculate the curl and the divergence of A:
1 % R2 ( 1 R
∇ × A = R̂ (0 − 0) + θ̂ ' − 0 *+ φ̂ ( 2Rsin θ ) = θ̂ + φ̂2sin θ
R & sin θ ) R sin θ
∇ ⋅ A = 4φ R + 2cosθ

The vector A is a solenoidal, rotational vector field.

Problem 2.39.
To see what the fields are, we must calculate the curl and the divergence of all five fields. Fields (1), (3) are evaluated in
Cartesian coordinates, field (2) in cylindrical coordinates and fields (4) and (5) in spherical coordinates. In all cases
direct application of the curl and divergence formulas are used.
1. ∇×A = 0 ∇ .A = 2
2. 1
∇× B = z (χοσφ + σινφ) ∇ .B = 1 (cosφ − sinφ )
r r
3. ∇× C = x − z ∇ .C = 0
4. ∇×D = R2cosθ + θ2sinθ + φ 1 (10R − cosθ ) ∇ .D = 2 sinθ + 5cosθ
R R sinθ
5. ∇×E = 0 ∇ .E = 2k
R
a. Field (3) is solenoidal (its divergence is zero).
b. Fields (1) and (5) are irrotational (their curl is zero).
c. Field (1) is a nonsolenoidal, irrotational vector field.
39
 Field (2) is a nonsolenoidal, rotational vector field.
Field (3) is a solenoidal, rotational vector field.
Field (4) is a nonsolenoidal, rotational vector field.
Field (5) is a nonsolenoidal, irrotational vector field.

Problem 2.40.
2
∇×(∇×A) = − ∇ A + ∇(∇.A)
∂ ∂ ∂
∇=x +y +z
∂x ∂y ∂z
# ∂ ∂ ∂& # ∂A ∂A ∂A &
∇.A = " x +y + z % . (xA x + yA y + zA z ) = " x + y + z %
$ ∂x ∂y ∂z ' $ ∂x ∂y ∂z '
# ∂ ∂ ∂& # ∂ ∂ ∂&
∇(∇.A) = " x +y + z % "x +y + z % . (xA x + yA y + zA z ) =
$ ∂x ∂y ∂z ' $ ∂x ∂y ∂z '
" ∂ ∂ ∂ % " ∂A ∂A ∂A %
!x +y + z $! x + y + z$ =
# ∂x ∂y ∂z & # ∂x ∂y ∂z &
∂ " ∂A ∂A ∂A % ∂ " ∂A ∂A ∂A % ∂ " ∂A ∂A ∂A %
x ! x + y + z$ + y ! x + y + z$ + z ! x + y + z$
∂x # ∂x ∂y ∂z & ∂y # ∂x ∂y ∂z & ∂z # ∂x ∂y ∂z &
" ∂2 A x ∂ ∂A y ∂ ∂A z % " ∂ ∂A x ∂2 A y ∂ ∂A z % " ∂ ∂A x 2
∂ ∂A y ∂ A z %
= x! + + $ + y! + + $ + z! + + $
# ∂x 2 ∂x ∂y ∂x ∂z & # ∂y ∂x ∂y 2 ∂x ∂z & # ∂z ∂x ∂z ∂y ∂z2 &
2 2 2
2 # ∂ ∂ ∂& # ∂ ∂ ∂& ∂ ∂ ∂
∇ = ∇.∇ = " x +y + z % ." x +y +z % = + +
$ ∂x ∂y ∂z ' $ ∂x ∂y ∂z ' ∂x 2 ∂y 2 ∂z2
2 # ∂2 ∂
2 2
∂ &
∇ A=" + + % (xA x + yA y + zA z ) =
$ ∂x 2 ∂y 2 ∂z2 '
# ∂2 A x ∂2 A x ∂2 A x & # ∂2 A y ∂2 A y ∂2 A y & # ∂2 A z ∂2 A z ∂2 A z &
x" + + % + y" + + % + z" + + %
$ ∂x 2 ∂y 2 ∂z2 ' $ ∂x 2 ∂y 2 ∂z2 ' $ ∂x 2 ∂y 2 ∂z2 '

Now:

2
∇(∇.A) − ∇ A =
$ ∂2 A x ∂ ∂A y ∂ ∂A z ' $ ∂ ∂A x ∂2 A y ∂ ∂A z ' $ ∂ ∂A x ∂ ∂A y ∂2 A z '
x# + + & + y# + + & + z# + + & −
% ∂x 2 ∂x ∂y ∂x ∂z ( % ∂y ∂x ∂y 2 ∂x ∂z ( % ∂z ∂x ∂z ∂y ∂z2 (
$ ∂2 A x ∂2 A x ∂2 A x ' $ ∂2 A y ∂2 A y ∂2 A y ' $ ∂2 A z ∂2 A z ∂2 A z '
x# + + & + y# + + & + z# + + & =
% ∂x 2 ∂y 2 ∂z2 ( % ∂x 2 ∂y 2 ∂z2 ( % ∂x 2 ∂y 2 ∂z2 (

40
 " ∂ 2 2
∂A y ∂ ∂A z ∂ Ax ∂ Ax %
x! + − − $ +
# ∂x ∂y ∂x ∂z ∂y 2 ∂z2 &
" ∂ 2 2
∂A x ∂ ∂A z ∂ A y ∂ A y %
y! + − − $ +
# ∂y ∂x ∂x ∂z ∂x 2 ∂z2 &
"∂ 2 2
∂A x ∂ ∂A y ∂ A z ∂ A z %
z! + − − $ =
# ∂z ∂x ∂z ∂y ∂x 2 ∂y 2 &
Thus:

2
∇(∇.A) − ∇ A =
$ ∂ ∂A y ∂A z 2 2
∂ Ax ∂ Ax '
x# + − + & +
% ∂x ∂y ∂z ∂y 2 ∂z2 (
$ ∂ ∂A x ∂A z 2 2
∂ Ay ∂ Ay '
y# + − + & +
% ∂y ∂x ∂z ∂x 2 ∂z2 (
$ ∂ ∂A x ∂A y 2 2
∂ Az ∂ Az '
z# + − + & =
% ∂z ∂x ∂y ∂x 2 ∂y 2 (
Since the two expressions are identical, the proof is complete.

41