
110 cross product
√42+ 52+ (–12)2
cross product (vector product) Many problems in
three-dimensional geometry require physicists and
mathematicians to find a
VECTOR
that is perpendicular
to each of two given vectors aand b. The cross prod-
uct, denoted a×b, is designed to be such a vector.
One begins by positioning the two vectors aand b
at the same point in space. These vectors define a plane
in space, and one sees that there are two possible direc-
tions for a third vector to point so as to be perpendicu-
lar to this plane. Mathematicians have settled on the
convention of following the “right-hand rule” to deter-
mine which direction to choose:
Take your right hand and point your fingers in
the direction indicated by the first vector a.
Now orient your hand, with your fingers still
pointing in this direction, in such a way that
your palm faces to the side of the plane con-
taining the vector b. (If you curl your fingers,
they will consequently turn through the small-
est angle that leads from ato b.) The direction
in which your thumb now points is the direc-
tion the vector a×bwill take.
Thus a×bwill be a vector that points in one direction
while b×awill point in the opposite direction. (In fact:
b×a= –a×b.)
Mathematicians have settled on a second conven-
tion to define the magnitude of a×b:
The magnitude of a×bis the area of the
PAR
-
ALLELOGRAM
defined by the vectors aand b.
If θis the smallest angle between aand b, then the par-
allelogram defined by the two vectors has side-lengths
|a| and |b|. Taking |a| as the base, the height of the par-
allelogram is then given by |b|·sinθ, and consequently
the
AREA
of the parallelogram is |a|·|b|·sinθ. Thus:
a×bis defined to be the vector of magnitude
|a|·|b|·sinθwith direction given by the right-
hand rule.
For example, if i= <1,0,0> is the unit vector pointing
in the direction of the x-axis, and j= <0,1,0> the unit
vector in the direction of the y-axis, then i×jis a vec-
tor pointing in the direction of the z-axis, with length
equal to the area of the unit square defined by iand j,
namely 1. Thus:
i×j= <0,0,1> = k
If two vectors aand bare parallel, then the angle
between them is zero and a×b= 0.
There is an alternative method for computing cross
products. If ais given by a= <a1,a2,a3> = a1i+ a2j+ a3k
and bis given by b= <b1,b2,b3> = b1i+ b2j+ b3k, then
one can check that the
DOT PRODUCT
of the vector:
(a2b3– a3b2)i+ (a3b1– a1b3)j+ (a1b2– a2b1)k
with each of aand bis zero. Thus this new vector is
perpendicular to both aand b. Mathematicians have
shown that it also has direction given by the right-hand
rule and magnitude equal to the parallelogram defined
by aand b. Thus this new vector is indeed the cross
product of aand b:
where, for the final equality, we have written the for-
mula in terms of the
DETERMINANT
of a 3 ×3 matrix.
(To prove that this new vector does indeed match the
quantity a×b, rotate the system of vectors a, b and
a×bso that apoints in the direction of the x-axis and
blies in the xy-plane. Then, for the rotated system we
have: a1= |a|, a2= 0, a3= 0, b1= |b|cosθ, b2= |b|sinθ,
b3= 0. One can now readily check that the formula
above yields a vector of the required length |a|·|b|·sinθ
pointing in the correct direction. One then argues that
the formula continues to hold when the system of three
vectors is rotated back to its original position.) Thus,
for example, if a= <1,4,2> and b= <3,0,1>, then a×b
= <4·1 – 2·0, 2.3 – 1·1, 1·0 – 4·3> = <4,5, – 12>.
According to the
DISTANCE FORMULA
, this vector has
length = √
–
185, which must be the
area of the parallelogram formed by aand b.
In two-dimensions, the determinant
gives a vector the same length as
a= <a1,a2> and perpendicular to it. In four-dimensional
space one can always find a fourth vector perpendicu-
lar to each of any given three vectors.
See also
ORTHOGONAL
;
TRIPLE VECTOR PRODUCT
;
VECTOR EQUATION OF A PLANE
.
iaa aa
21 21
−=<−>j,
ij
aa
12
=
ab×= − + − + −
=
()()()ab ab ab ab ab ab
aaa
bbb
23 32 31 13 12 21
123
123
ijk
ijk