FACULTY OF MATHEMATICAL STUDIES

MATHEMATICS FOR PART I ENGINEERING

Lectures

MODULE 15 VECTORS II
1. Triple products
2. Differentiation and integration of vectors
3. Equation of a line
4. Equation of a plane

1. Triple products
There are two triple products which make sense and are used. These will be introduced briefly and then
considered in turn.
Triple scalar product a × b · c
(It should be clear from the definitions of scalar and vector products that the only possible interpretation
of the above is (a × b) · c , when the answer is a scalar. The other possibility a × (b · c) does not make
sense.)
Triple vector product e.g. (a × b) × c
(Here the brackets are essential since a × (b × c) also exists but gives a different answer. Both expressions
lead to results which are vectors.)
(i) Triple scalar product a × b · c
Draw a parallelepiped with vectors a, b and c as sides.
k

c
b
a
Now
(a × b) · c = {(area base) n̂} · c = (area base) (n̂ · c)
= (area base)(± height) = ± volume .
The alternative signs occur above because it is not known whether the unit vector n̂ points in the same
direction as c , or in the opposite direction. It follows from the previous formula that

volume of parallelpiped = |(a × b) · c|,

where the modulus sign in this situation means taking the absolute value, since the triple product inside is
a scalar.
A number of results can now be proved, either geometrically by considering the equivalent parallelogram or
algebraically from the properties of scalar and vector products.
Properties:
(a) If two, or more, of the vectors a, b and c are parallel then a × b · c = 0
(e.g. a × b · a = 0) .
(b) If a, b and c are coplanar (i.e. lie in the same plane) then a × b · c = 0 .

1
(c) a × b · c= a · b × c (can interchange · and × )
(d) a × b · c=b × c · a=c × a · b
(can keep the operator signs fixed but rotate the three vectors in cyclic order a → b, b → c, c → a)
(e) The component answer for the triple scalar product is easily written as a determinant
 
 a1 a2 a3 

a × b · c =  b1 b2 b3 
 c1 c2 c3 

(If you have not yet covered the matrix section then calculate the triple product (a × b) · c directly in two
stages using the component forms of the vector and scalar products.)

(ii) Triple vector product a × (b × c)

It can be shown, from the component forms of the vectors for instance, that

a × (b × c) = (a · c) b − (a · b) c

The above result, which is stated on the Formula Sheet and in the Data Book, can sometimes be useful in
manipulations but it has no simple physical interpretation.
The corresponding result for the triple vector product with brackets in different positions can easily be found
using the above identity and known properties for vector products, as follows:

(a × b) × c = −c × (a × b) = − {(c · b) a − (c · a) b}
= (c · a) b − (c · b) a

2. Differentiation and integration of vectors

In mechanics vectors are usually time-dependent. For instance the position of a particle normally depends
on time so we write r(t).
An obvious extension of the derivative of a scalar function of a single variable gives
 
dr r(t + ∆t) − r(t)
Def. = lim
dt ∆t→0 ∆t

Q
r(t+∆ t )
P
r(t)

−−
→ −−
→
Now r(t + ∆t) − r(t) = P Q, and as ∆t → 0 then P Q becomes closer to the tangential direction at P . The
dr
magnitude approaches the rate of change of distance with time and, therefore, the above definition for
dt
gives the velocity of a moving point with position vector r(t).
In components, if position vector r(t) = (x(t), y(t), z(t)) then
 
dr dx dy dz
velocity = , ,
dt dt dt dt
 
d2 r d2 x d2 y d2 z
acceleration = , ,
dt2 dt2 dt2 dt2

2
Properties:
d dλ da
(a) (λ(t) a(t)) = a+λ
dt dt dt
d da db
(b) (a(t) + b(t)) = +
dt dt dt
d da db
(c) (a(t) · b(t)) = ·b+a·
dt dt dt
d da db
(d) (a(t) × b(t)) = ×b+a×
dt dt dt
Ex 1. If P has position r = sin t i + cos t j find (a) the acceleration of P ,
(b) the speed of P .
(a) Since i and j are constant

dr
velocity = = cos t i − sin t j
dt  
d2 r d dr
acceleration = 2 = = − sin t i − cos t j
dt dt dt

(b) Using the above

 
 dr 
1/2
1/2
speed =   = (cos t)2 + (− sin t)2 = cos2 t + sin2 t = 11/2 = 1 .
dt

Ex 2. The acceleration of a particle is t i + cos t b . Find the position of the particle at time t given that
dr
r = 0 and = 0 at time t = 0 .
dt
It is given that
d2 r
= t i + cos t j,
dt2
and so we must integrate. Now integration is the inverse of differentiation, and since the vectors i and j
are constant it is only necessary to integrate the components of the vector. Hence, integrating with respect
to t implies
dr t2
= i + sin t j + C1 .
dt 2
A constant of integration must be introduced, as usual, but since each term in the vector equation is a vector
dr
the integration constant must also be a vector. Applying the initial condition that = 0 when t = 0
dt
implies
0 = 0 i + 0 j + C1 ,
hence C1 = 0 . Integrating the simplified equation for the velocity then yields

t3
r(t) = i − cos t j + C2 ,
6
with C2 denoting the new constant of integration, again a vector. At time t = 0 it follows that

0 = 0 i − 1 j + C2 → C2 = j .

The final answer for the position, therefore, is

t3
r(t) = i + (1 − cos t) j .
6
3
3. Equation of a line
The points A and B have position vectors a and b respectively. The vector equation of the straight line
through A and B is now derived.
A P B
r
a b

O
Consider a general point P which lies on the line, with the position vector of this point being r. It is clear
from the diagram that
−−→ −→ −→
OP = OA + AP
−→
i.e. r = a + AP
−−→
Now AB = b − a and, since the point P always lies on the straight line passing through A and B , the
−→ −
−→
vectors AP and AB are parallel. Thus
−→ −
−→
AP = t AB = t (b − a),

where t is any scalar, i.e. −∞ < t < +∞. Combining the two results the vector equation of a line
through two given points can be written

r = a + t (b − a), −∞ < t < +∞.

In almost all cases you will be working with the vector equation of a line and not its Cartesian equivalent.
However, the Cartesian version is easily calculated. If r = (x, y, z) , a = (a1 , a2 , a3 ) and b = (b1 , b2 , b3 ) ,
then substituting into the vector equation of the line and equating components gives

x = a1 + t(b1 − a1 ), y = a2 + t(b2 − a2 ), z = a3 + t(b3 − a3 ).

Rearranging each of these expressions to produce a formula for t leads to the equations
x − a1 y − a2 z − a3
= = , (= t),
b 1 − a1 b 2 − a2 b 3 − a3
which is the Cartesian equation of a line.

Ex 3. Show that the line through A (0, 4, 4) in the direction of b = (−1, 2, 1) intersects the line through
the points C (1, 4, −1) and D (1, 2, 3).

b
P L1
A C
a r
D
O L2
For the first line, L1 , we are given a point on L1 and a vector parallel to L1 . If P is any point on the line
−→ −→
then it is clear that AP is always parallel to the line, hence AP = t b = t (−1, 2, 1) . It follows immediately
that for line L1
−−
→ −→ −→
OP = OA + AP
or r = (0, 4, 4) + t (−1, 2, 1).

4
The second line, L2 , passes through the two points C and D, and is identical to the general case considered
at the beginning of this section. Thus
−−→ −−→ −− →
CD = OD − OC = ((1 − 1, 2 − 4, 3 − (−1)) = (0, −2, 4)
so equation of L2 is : r = (1, 4, −1) + s (0, −2, 4).
−−→
(Note in the above that L2 is the equation of the line passing through the point C which is parallel to CD.
−−→
We could have written down the line passing through the point D which is parallel to CD. This would
also lead to a correct equation for the line – the details would be different, but the final part of the question
would lead to the same answer.
It is also important to note that line L2 involves a different scalar constant to that used for L1 . This is
essential because a general point on the first line has no connection with a general point on the second line.)
The two lines intersect if there is a point in common between the lines. This will occur when ALL THREE
coordinates of the general points on the two lines are equal. Thus the lines intersect if

(0, 4, 4) + t (−1, 2, 1) = (1, 4, −1) + s (0, −2, 4)

i.e. (−t, 4 + 2t, 4 + t) = (1, 4 − 2s, −1 + 4s),

which requires
−t = 1 4 + 2t = 4 − 2s 4 + t = −1 + 4s.
The first of the above equations implies t = −1. Substituting this value into the second equation yields
4 + 2(−1) = 4 − 2s, → 4 − 2 = 4 − 2s, hence s = 1. Finally, substituting the calculated values for s and
t into the third equation gives 4 − 1 = −1 + 4, i.e. 3 = 3. This equation is clearly satisfied identically so
the lines do intersect.
(It is essential to look at the third equation because if this was not satisfied then the lines would not intersect.)
The point of intersection is easily calculated by substituting the calculated value for t into the equation for
L1 (or the value for s into the equation for L2 ) leading to

point of intersection = (0, 4, 4) + (−1)(−1, 2, 1) = (0 + 1, 4 − 2, 4 − 1) = (1, 2, 3)

Shortest distance between two skew lines

The calculation of this distance is illustrated by the following example.

Ex 4. Find the shortest between the two skew lines

r = (0, 9, 2) + t (3, −1, 1) and r = (−6, −5, 10) + s (−3, 2, 4).

From the earlier discussions of equations of lines it is known that the first line above passes through the point
(0, 9, 2) and is parallel to the vector (3, −1, 1), whereas the second line passes through the point (−6, −5, 10)
and is parallel to the vector (−3, 2, 4). It should be obvious after a little thought that the shortest line which
joins the two given lines will be perpendicular to both lines. Thus, if the ends of this shortest line are P1 and
P2 with P1 and P2 lying on the first and second lines respectively, then the shortest line will be parallel to
the vector product of the vectors parallel to the two given straight lines. Define n = (3, −1, 1) × (−3, 2, 4)
then from the components written
3 −1 1 3 −1 1
−3 2 4 −3 2 4
it follows that
n = (−1(4) − 1(2), 1(−3) − 3(4), 3(2) − (−1)(−3)) = (−6, 15, 3).
The line with shortest length will be parallel to n and it is helpful to calculate the unit vector n̂. Since
√
|n| = ((−6)2 + (−15)2 + 32 ))1/2 = (36 + 225 + 9)1/2 = (270)1/2 = 3 30,

5
 n 1 1
n̂ = = √ (−6, −15, 3) = √ (−2, −5, 1).
|n| 3 30 30
Since P1 and P2 have been defined to lie on the first and second lines respectively, then
−−−→
P1 P2 = {(−6, −5, 10) + s (−3, 2, 4)} − {(0, 9, 2) + t (3, −1, 1)}.
−−−→
When the joining line has shortest length then P1 P2 = d n̂, where d denotes the required distance. Equating
the two expressions gives

{(−6, −5, 10) + s (−3, 2, 4)} − {(0, 9, 2) + t (3, −1, 1)} = d n̂. (1)

The unknown d can be found by equating the three components of the above equation, and then solving the
three resulting equations. However, it is much quicker to use the fact that, by construction, the vector n̂ is
perpendicular to the vectors parallel to the two original lines. Therefore, distance d is found more quickly
by taking the scalar product of the above equation with n̂, leading to

{(−6, −5, 10) + s (−3, 2, 4)} − {(0, 9, 2) + t (3, −1, 1)} · n̂ = d n̂ · n̂.

Using the calculated value for n̂ and expanding leads to

1 1
(−6, −5, 10) · √ (−2, −5, 1) + s(−3, 2, 4) · √ (−2, −5, 1)
30 30
1 1
− (0, 9, 2) · √ (−2, −5, 1) − t(3, −1, 1) · √ (−2, −5, 1) = d n̂ · n̂ = d.
30 30
Evaluating the scalar products gives

1 1
√ (12 + 25 + 10) + s(0) − √ (0 − 45 + 2) + t(0) = d,
30 30
√
47 − (−43) 90 90 30 √
i.e. d= √ = √ = = 3 30.
30 30 30

[To find the coordinates of the end-points P1 and P2 it is necessary to equate the three components in
equation (1) and solve:
   
−2 √ −2
−6 − 3s − 3t = d √ = (3 30) √ = −6, 3s + 3t = 0,
30 30
   
−5 √ −5
−14 + 2s + t = d √ = (3 30) √ = −15, 2s + t = −1,
30 30
   
1 √ 1
8 + 4s − t = d √ = (3 30) √ = 3, 4s − t = −5.
30 30

The solution of the three equations in s and t is s = −1 and t = 1 leading to the points P2 = (−3, −7, 6)
and P1 = (3, 8, 3).]

6
4. Equation of a plane
The key feature about a plane is that at any point a line perpendicular to the plane is always pointing in
the same direction.
n

P
A

a r

O
Consider a plane which passes through the point A with position a. If the vector n is perpendicular to the
−→
plane and P , with position vector r, is any point on the plane, then the vector AP must always lie in the
−→
plane and hence always be perpendicular to n. Now AP = r − a , so the vectors being perpendicular implies
(r − a) · n = 0
→ r ·n−a · n= 0
i.e. r ·n= a · n
The right-hand side of the above contains two constant vectors and so the general vector equation of a
plane is usually written
r · n = C,
where n is the normal to the plane and C is a constant determined by a point on the plane.

Ex 5. Find the equation of the plane which passes through the point A (2, 3, −1) and is perpendicular to
the vector 2 i − 2 j + k.
The equation of a plane with given normal is

r · n = C, or r · (2, −2, 1) = C.

The constant C is determined by recalling that A (2, 3, −1) lies on the plane so must satisfy the above
equation when the position of the general point is replaced by the position of A. It follows that

C = (2, 3, −1) · (2, −2, 1) = 2(2) + 3(−2) + (−1)(1) = 4 − 6 − 1 = −3,

and so the required equation of the plane is

r · (2, −2, 1) = −3.

[The Cartesian equation for the plane is easily found from the above, if required:
(x, y, z) · (2, −2, 1) = −3 or 2x − 2y + z = −3. Note that if you were given the latter Cartesian equation of
a plane then the coefficients of the unknowns give you the components of the vector which is perpendicular
to the plane, namely (2, −2, 1) in the above example.]

Perpendicular distance from a point to a plane

Consider the plane introduced at the beginning of this section, that is passing through the point A with
normal n. Suppose we need to calculate the distance from point Q to the plane.
n

N
A

7
The easiest way to find the answer is to say that
−→
QN = component of QA in the direction perpendicular to plane
−→
i.e. QN = |QA · n̂| ,

where n̂ denotes the usual unit vector and the modulus sign is necessary because we do not know the relative
−→
directions of n̂ and QA.

Ex 6. Find the distance from the point Q (2, −3, 4) to the plane x + 2y + 2z = 13.
As mentioned at the end of Ex 5 the given equation of the plane shows that a vector perpendicular to
the plane is (1, 2, 2), = n, say. It is necessary to find the coordinates of a point that lies on the plane.
Clearly, any point is sufficient so choose y = z = 0 in which case the equation x + 2y + 2z = 13 is
satisfied provided x = 13. Hence the point (13, 0, 0) is on the given plane. Call this point A, then
−→
QA = (13 − 2, 0 − (−3), 0 − 4) = (11, 3, −4).
Now  
n (1, 2, 2) (1, 2, 2) 1 2 2
n̂ = = 2 = = , , ,
|n| (1 + 22 + 22 )1/2 3 3 3 3
hence
    
−→  1 2 2   11 6 8 
distance from Q to the plane = |QA · n̂| = (11, 3, −4) · , , = + − = 3.
3 3 3  3 3 3