Let
(v,
w)
be an ordered pair of two nonparallel vectors. Representv
and w by directed line segments with a common initial point. Point the fingers of your right hand along the first vectorv
and then curl them toward the second vector w, as in Figure 9. The direction that the thumb points during this process is the direction of the standard unit normal vector for the vectorsv,
w. Look again at Figure8.
The direction of vector n is the standard unit normal for the ordered pair(v,
w)
.Standard unit normal
_. Figure 9
� EXAMPLE 5 Find the standard unit normal for the pairs
(i,j)
and(j,k)
and
(k, i).
Find also the standard unit normal for the pairs(j, i)
and(k,j)
and(i,k).
Solution If you curl the fingers of your right hand from
i
towardj,
then your thumb will point in the direction of the positive z-axis. Thus the standard unit normal for the pair(i,j)
isk.
Similar reasoning shows that the standard unit normal for(j,i)
is-k.
We leave it as an exercise to check that the standard unit normal for the pair(j, k)
isi,
and the standard unit normal for the pair(k, i)
isj.
In addition, the standard unit normal for(k,j)
is-i,
and the standard unit normal for(i, k)
is-j.
<IllINSIGHT
According to the right-hand rule, the standard unit normal vector for the ordered pair (v, w) points in the opposite direction to the standard unit normal vector for the ordered pair (w, v). Example 4 provides particular cases of this general fact.
If v x w i- 0, then we may form the unit vector dir(v x w) =
ll
v x wr
1 (v x w).Because v
X
w is perpendicular to both v and w, the vector dir(vX
w) must be the standard unit normal for either the ordered pair (v, w) or the ordered pair (w, v).Our next theorem tells us which. It also specifies the length of v
X
w and tells us precisely when the equation vX
w = 0 holds.THEOREM 3 Let v and w be vectors. Then:
a.
ll
vX
wll
2 =ll
vll
2ll
wll
2 - (v · w)2.b. If v and w are nonzero, then
ll
vX
wll
=ll
vll ll
wll
sin(O) where 0 E[O, rr]
denotes the angle between v and w.
c. v and w are parallel if and only if v
X
w = 0.d. If v and w are not parallel, then v x w points in the direction of the standard unit normal for the pair (v, w). In particular, dir(v
X
w) =ll
vX
wr
1 (vX
w) is the standard unit normal vector for the pair (v, w).Proot Part a is an identity among the entries of v = (v1, v2, v3) and w = (w1, w2, w3).
We have
ll
vX
wll
2 = (v2w3 - w2v3)2+
(v1w3 - w1v3)2+
(v1w2 - w1v2)2=
(Vt+ V� + V�)(wr + W� + W�) - (V1W1 + VzWz + V3W3)2
=
ll
vll
2ll
wll
2 - (v · w)2.The second equality in this chain is not obvious, but it may be verified by multiplying everything out.
If neither v nor w is the zero vector, then the scalars
ll
vll
andll
wll
are nonzero.In this case, the identity of part a becomes
ll
vX
wll
2 =ll
vll
2ll
wf
- (v · w)2 =ll
vll
2ll
wll
2(i
- (vll
vll ll
wll
=ll
vll
2ll
wll
2(
1- cos2(0))
,where the last equality is obtained by using formula (9.3.6). It follows that
ll
vX
wll
2 =ll
vll
2ll
wll
2sin2(0) and, on taking the square root,ll
vX
wll
=ll
vll ll
wll
lsin(O)I. Because 0 :5 0 :5 rr, it follows that 0 :5 sin(O). Therefore lsin(O)I = sin(O) and
ll
vX
wll
=ll
vll ll
wll
sin(O), as asserted in part b.Theorem 3 of Section 3 tells us that vectors v and w are parallel if and only if Iv· wl =
ll
vll ll
wll
· Using the identity of part a, we conclude that v and w are parallel if and only ifll
vX
wll
= 0. Because 0 is the only zero length vector, assertion c follows.For part d, we will limit our verification to a few special cases. Notice that
i X j
=k, j X k
=i,
andk X i
=j.
Based on our observations in Example 4, we conclude that,for the basic vectors
i,j,
andk,
the operation of cross product produces the standard unit normal. Geometric reasoning can be used to show that the cross product vX
w always points in the direction of the standard unit normal for the vectors v, w. •� EXAMPLE 6 Let v=(l,-3,2) and w=(l,-1,4). What is the standard unit normal vector for (v, w)?
Cross Products and the Calculation of Area
9.4 The Cross Product and Triple Product 759
Solution We calculate the vector
v x w = ( ( -3)
. 4- ( -1) . 2) i - ( 1 .
4- 1 . 2 )j + ( 1 . ( -1) - 1 . ( -3)) k
=
-10i - 2j + 2k
and its length
llv x wll = + (-2)2 + 22 = v'T08 = = 6v'3.
Part d of Theorem
2
tells us thatis the vector we seek. <11111
Now we learn a connection between cross products and areas of triangles and parallelograms. To be specific, if
v
andw
are two nonparallel vectors that are-
-represented by directed line segments
OP
andOQ
with common initial point 0, then we speak of triangleOPQ
as "the triangle determined by the vectorsv
andw."
See Figure
10
for this triangle. The parallelogram determined byv
andw
is shown in Figure11.
Notice that, when we declareOPQ
to be the triangle determined by
v
andw,
the third sideQP
representsv-w.
Had we positionedv
so thatits initial point coincided with the terminal point
Q
ofw,
then we would have obtained triangleOQR
with third sidev + w.
Notice that the areas of trianglesOPQ
andOQR
are the same because each is half the area of the parallelogram.Thus had we declared
OQR
as the triangle determined by the vectorsv
andw
instead of 0
PQ,
we would have been speaking of a triangle with the same area._. Figure 11
THEOREM 4 Suppose that
v
andw
are nonparallel vectors. The area of the triangle determined byv
andw
is� II v
Xw II
· The area of the parallelogram determined byv
andw
isII v
Xw II·
Proof. As we see from Figure