Professional Documents
Culture Documents
Calculus
28.2
Introduction
A vector eld or a scalar eld can be dierentiated with respect to position in three ways to
produce another vector eld or scalar eld. This Section studies the three derivatives, that is:
(i) the gradient of a scalar eld (ii) the divergence of a vector eld and (iii) the curl of a vector
eld.
'
Prerequisites
Before starting this Section you should . . .
&
Learning Outcomes
After completing this Section you should be
able to . . .
Consider the height above sea level at various points on a hill. Some contours for such a hill
are shown in the diagram.
10
20
30
40
50
C
60
B
A D
Figure 1
We are interested in how changes from one point to another. Starting from A and making
a displacement d the change in height ( ) depends on the direction of the displacement. The
magnitude of each d is the same.
Displacement Change in
AB
40 30 = 10
AC
40 30 = 10
AD
30 30 = 0
AE
20 30 = 10
The change in clearly depends on the direction of the displacement. For the paths shown
increases most rapidly along AB, does not increase at all along AD (as A and D are both on
the same contour and so are both at the same height) and decreases along AE.
The direction in which changes fastest is along the line of greatest slope and orthogonal
(i.e. perpendicular) to the contours. Hence, at each point of a scalar eld we can dene a vector
eld giving the magnitude and direction of the greatest rate of change of locally.
A vector eld, called the gradient, written grad , can be associated with a scalar eld
so that at every point the direction of the vector eld is orthogonal to the scalar eld contour
and is the direction of the maximum rate of change of .
For a second example consider a metal plate heated at one corner and cooled by an ice bag
at the opposite corner. All edges and surfaces are insulated. After a while a steady state situation exists in which the temperature at any point remains the same. Some temperature
HELM (VERSION 1: March 18, 2004): Workbook Level 1
28.2: Dierential Vector Calculus
heat source
35
30 25
20 15
10
35
30 25
20 15
10
ice bag
(a)
ice bag
(b)
Figure 2
The direction of the heat ow is along ow lines which are orthogonal to the contours (see the
dashed lines in Figure 2(b)); this heat ow is measured by F = grad .
Denition
The gradient of the scalar eld = f (x, y, z) is
grad = =
i+
j+
k
x
y
z
Often, instead of grad , the notation is used. ( is a vector dierential operator called
i+
j+
k. As a vector dierential operator, it retains the
del or nabla dened by
x
y
z
characteristics of a vector while also carrying out dierentiation.)
The vector grad gives the magnitude and direction of the greatest rate of change of at
any point, and is always orthogonal to the contours of . For example, in Figure 1, grad
points in the direction of AB while the contour line is parallel to AD i.e. perpendicular to AB.
Similarly, in Figure 2, the various intersections of the contours with the lines representing grad
occur at right-angles.
For the hill considered earlier the direction of grad is shown at various points in Figure
3. Note that the magnitude of grad is greatest when the hill is at its steepest.
10
20 30
40
50 60
Figure 3
Key Point
is a scalar eld but grad is a vector eld.
(b) = xy 2 z 3
Solution
2
(b) grad =
(xy 2 z 3 )i +
(xy 2 z 3 )j + (xy 2 z 3 )k = y 2 z 3 i + 2xyz 3 j + 3xy 2 z 2 k
x
y
z
(a) grad =
Example For f = x2 + y 2 nd grad f at the point A(1, 2). Show that the direction of
grad f is orthogonal to the contour at this point.
Solution
grad f =
f
f
f
i+
j+
k = 2xi + 2yj + 0k = 2xi + 2yj
x
y
z
y
A
Figure 4
To nd the change in a function in a given direction (given in terms of a unit vector a) take
the scalar product, (grad ) a.
HELM (VERSION 1: March 18, 2004): Workbook Level 1
28.2: Dierential Vector Calculus
Example Given = x2 y 2 z 2 , nd
1. grad at (1, 1, 1) and a unit vector in this direction.
2. the derivative of at (2, 1, 1) in the direction of
(a) i
(b) d = 35 i + 45 k.
Solution
grad =
i+
j+
k = 2xy 2 z 2 i + 2x2 yz 2 j + 2x2 y 2 zk
x
y
z
1
r
(d) = rn
3. If = x3 y 2 z, nd,
(a)
(b) a unit vector normal to the contour at the point (1, 1, 1).
(c) the rate of change of at (1, 1, 1) in the direction of i.
(d) the rate of change of at (1, 1, 1) in the direction of the unit vector
n = 13 (i + j + k).
4. Find a unit vector which is normal to the sphere x2 +(y1)2 +(z+1)2 = 2
at the point (0, 0, 0).
5. Find unit vectors normal to 1 = y x2 and 2 = x + y 2. Hence
nd the angle between the curves y = x2 and y = 2 x at their point of
intersection in the rst quadrant.
Your solution
1.)
1 (3i
14
Your solution
3.)
r
r
r
(a) , (b) 2 , (c) 3 , (d) nrn2 r
r
r
r
Your solution
2.)
(a)
(y x)i +
(y x)j = i + j,
x
y
(b) 2xi + j,
(c) [
Your solution
4.)
(a) Find the vector eld where = x2 + (y 1)2 + (z + 1)2
(b) Find the value that this vector eld takes at the point (0,0,0) to get a vector normal
to the sphere.
(c) Divide this vector by its magnitude to form a unit vector.
c)
1 (j
2
+ k)
b) 2j + 2k
a) 2xi + 2(y 1)j + 2(z + 1)k
Your solution
5.)
F1 F2 F3
+
+
.
x
y
z
F1 F2 F3
+j
+ k ) (F1 i + F2 j + F3 k) =
+
+
.
x
y
z
x
y
z
the vector eld representing velocity) at a point in a uid (liquid or gas) is a measure of the
rate per unit volume at which the uid is owing away from the point. A negative divergence is
a convergence indicating a ow towards the point. Physically divergence means that either the
uid is expanding or that uid is being supplied by a source external to the eld. Conversely
convergence means a contraction or the presence of a sink through which uid is removed from
the eld. The lines of ow diverge from a source and converge to a sink.
If there is no gain or loss of uid anywhere then div v = 0 which is the equation of continuity for an incompressible uid.
The divergence also enters engineering topics such as magnetic elds. A magnetic eld (denoted by B) has the property B = 0, that is there are no sources or sinks of magnetic eld.
Key Point
F is a vector eld but div F is a scalar eld.
Solution
(a) div F =
(x2 )
x
(b) div r =
(x)
x
(c) div v =
(x)
x
(y 2 )
y
(y)
y
(y)
y
(z 2 )
z
(z)
z
= 2x + 2y + 2z
=1+1+1=3
(2)
z
= 1 + 1 + 0 = 0
Example Find the value of a for which F = (2x2 y + z 2 )i + (xy 2 x2 z)j + (axyz 2x2 y 2 )k
is incompressible.
Solution
F is incompressible if div F = 0.
(2x2 y + z 2 ) +
(xy 2 x2 z) + (axyz 2x2 y 2 ) = 4xy + 2xy + axy
x
y
z
which is zero if a = 6.
div F =
Find the divergence of the following vector elds, in general terms and at the
point (1, 0, 3)
(a) F 1 = x3 i + y 3 j + z 3 k
(b) F 2 = x2 yi 2xy 2 j
(c) F 3 = x2 zi 2y 3 z 3 j + xyz 2 k
Your solution
1.)
i+
j+
k
y
z
z
x
x
y
Physical signicance of Curl
10
The divergence of a vector eld represents the outow rate from a point; however the curl
of a vector eld represents the rotation at a point.
Consider the ow of water down a river. The surface velocity v of the water is revealed by
watching a light oating object such as a leaf. You will notice two types of motion. First the
leaf oats down the river following the streamlines of v, but it may also rotate. This rotation
may be quite fast near the bank but slow or zero in midstream. Rotation occurs when the
velocity, and hence the drag, is greater on one side of the leaf than the other.
bank
bank
Figure 6
Note that for a two-dimensional vector eld, such as v described here, curl v is perpendicular
to the motion, and this is the direction of the axis about which the leaf rotates. The magnitude
of curl v is related to the speed of rotation.
For motion in three dimensions a particle will tend to rotate about the axis that points in the
direction of curl v, with its magnitude measuring the speed of rotation.
If, at any point P, curl v = 0 then there is no rotation at P and v is said to be irrotational at P.
If curl v = 0 at all points of the domain of v then the vector eld is irrotational.
Key Point
Note that F is a vector eld and that curl F is also a vector eld.
11
Solution
(a)
i j k
v = x y z
x 2 0
=
(0) (2) i +
(x)
(0) j +
(2)
(x) k
y
z
z
x
x
y
= 0i + 0j + 0k = 0
A oating leaf will travel along the streamlines (moving away from the y axis and
upwards - see Figure 14 of Section 29.1) without rotating.
(b)
i
j k
v = x y z
y x 0
=
(0) (x) i +
(y)
(0) j +
(x)
(y) k
y
z
z
x
x
y
= 0i + 0j + 2k = 2k
A oating leaf will travel along the streamlines (anti-clockwise around the origin )
and will rotate anticlockwise (as seen from above).
12
Solution
(a)
i j k
curl u = F = x y z
x2 y 2 0
2
2
2
2
=
(0) (y ) i +
(x )
(0) j +
(y )
(x ) k
y
z
z
x
x
y
= 0i + 0j + 0k = 0
curl u = 0 so u is irrotational everywhere.
(b)
curl F
i
j
k
=
F =
x
y
z
xy xz 3x2 yz
2
=
(yz) (3x ) i +
(xy xz)
(yz) j
y
z
z
x
2
(3x )
(xy xz) k
+
x
y
=
zi yj + 5xk
At the point (0, 0, 0), curl F = 0. At the point (1, 2, 3), curl F = 3i j + 5k.
13
1. Find the curl of each of the following two-dimensional vector elds. Give
each in general terms and also at the point (1, 2).
(a) F 1 = 2xi + 2yj
(b) F 2 = y 2 i + xyj
(c) F 3 = x2 y 3 i x3 y 2 j
2. Find the curl of each of the following three-dimensional vector elds.
Give each in general terms and also at the point (2, 1, 3).
(a) F 1 = y 2 z 3 i + 2xyz 3 j + 3xy 2 z 2 k
(b) F 2 = (xy + z 2 )i + x2 j + (xz 2)k
3. The surface water velocity on a straight uniform river 20 metres wide is
1
modelled by the vector v = 50
x(20 x)j where x is the distance from
the west bank (see Figure 7).
20 m
Figure 7
(a) Find the velocity v at each bank and at midstream.
(b) Find v at each bank and at midstream.
4. The velocity eld on the surface of an emptying bathroom sink can be
modelled by two functions, the rst describing the swirling vortex of
radius a near the plughole and the second describing the more gently
rotating uid outside the vortex region.
are
These functions
x2 + y 2 a
u(x, y) = w(yi + xj),
wa2 (yi+xj)
v(x, y) = x2 +y2
x2 + y 2 a
Find curl u and curl v.
14
Your solution
1.)
15
4. The Laplacian
The Laplacian of a function is written as 2 and is dened as: Laplacian = div grad ,
that is
2 =
i+
j+
k
=
x
y
z
2 2 2
=
+
+ 2
x2 y 2
z
2
Solution
2u 2u 2u
+
+
= 2y 2 z + 2x2 z + 0 = 2(x2 + y 2 )z
2 u =
x2 y 2 z 2
16
Solution
(a)
2
2
j + k)
(2yzi x yj + xz k) ( i +
x
y
z
2yz
x2 y
+ xz 2
2x2 yz 3
x
y
z
(A ) =
=
=
=
=
(b)
(2x2 yz 3 )i +
(2x2 yz 3 )j + (2x2 yz 3 )k
x
y
z
3
2 3
2
2
= 4xyz i + 2x z j + 6x yz k
So
A =
17
2
2
2
2
3
2
3
(2x
yz
)
+
(2x
yz
)
+
(2x2 yz 3 ) = 4yz 3 + 0 + 12x2 yz
x2
y 2
z 2
Example Consider
(a) grad(div A)
(b) grad(grad )
(c) curl(div F )
(d) div [ curl (A grad ) ]
In each case determine whether the quantity can be formed and, if so, whether
it is a scalar or a vector.
Solution
(a) A is a vector and divA can be calculated and is a scalar. Hence, grad(div A) can be
formed and is a vector.
(b) is a scalar so grad can formed and is a vector. As grad is a vector, it is not
possible to take grad(grad )
(c) F is a vector and hence div F is a scalar. It is not possible to take the curl of a
scalar so curl(div F ) does not exist.
(d) is a scalar so grad exists and is a vector. Agrad exists and is also a vector
as is curl Agrad . The divergence can be taken of this last vector to give
div [ curl (Agrad ) ] which is a scalar.
(A) = () A + ( A)
(A) = () A + ( A)
(A B) = B ( A) A ( B)
(A B) = (B )A (A )B
+A B B A
or (A B) = (B )A + (A )B
+A ( B) + B ( A)
or () = 0
or ( A) = 0
or
or
or
or
div curl A = 0.
18
Solution
div curl A =
=
=
=
=
N.B. This assumes
i
j
j
div
x y z
A1 A2 A3
A3 A2
A1 A3
A2 A1
div
i+
j+
k
y
z
z
x
x
y
A3 A2
A1 A3
A2 A1
x y
z
y z
x
z x
y
2
2
2
2
2
2
A3
A2
A1
A3
A2
A1
xy zx yz yx zx zy
0
2 A3
2 A3
=
etc.
xy
yx
Example Verify identity 1 for the vector A = 2xyi 3zk and the function = xy 2 .
Solution
A = 2x2 y 3 i 3xy 2 zk so
(xy 2 )i +
(xy 2 )j + (xy 2 )k = y 2 i + 2xyj so
x
y
z
() A = (y 2 i + 2xyj) (2xyi 3zk) = 2xy 3
A = (2xyi 3zk) = 2y 3 so A = 2xy 3 3xy 2 giving
() A + ( A) = 2xy 3 + (2xy 3 3xy 2 ) = 4xy 3 3xy 2
So RHS = 4xy 3 3xy 2 = LHS
=
19
1. If F = x2 yi 2xzj + 2yzk, nd
(a) F
(b) F
(c) ( F )
(d) ( F )
(e) ( F )
2. If = 2xz y 2 z, nd
(a)
(b) 2 = ()
(c) ()
3. Which of the following combinations of grad, div and curl can be formed?
If a quantity can be formed, state whether it is a scalar or a vector.
(a) div (grad )
(b) div (div A)
(c) curl (curl F )
(d) div (curl F )
(e) curl (grad )
(f) curl (div A)
(g) div (A B)
(h) grad (1 2 )
(i) curl (div (A grad ))
Your solution
1.)
(b) (2x + 2z)i (x2 + 2z)k, (c) 2yi + (2 + 2x)j, (d) 0, (e) (2 + 2x)j
HELM (VERSION 1: March 18, 2004): Workbook Level 1
28.2: Dierential Vector Calculus
20
21
a), d), g) are scalars: c), e), h) are vectors and b) and f) are not dened.
Your solution
3.)
(a) 2zi 2yzj + (2x y 2 )k, (b) 2z, (c) 0
Your solution
2.)