is a vector tangent to the curve of intersection and perpendicular to both of the
normal vectors ∇F and ∇G. A unit tangent vector to the curve of intersection is
constructed having the form form
eˆt=
∇F×∇ G
|∇ F×∇ G|
.
Example 7-21. In two-dimensions a curve y=f(x)or r =xeˆ 1 +f(x)ˆe 2 can be
represented in the implicit form φ=φ(x, y ) = y−f(x) = 0 so that
grad φ=∂φ
∂x
ˆe 1 +∂φ
∂y
ˆe 2 =−f′(x)ˆe 1 +ˆe 2 =N
is a vector normal^6 to the curve at the point (x, f (x)).A unit normal to this curve is
given by
ˆen=−f
′(x)ˆe 1 +ˆe 2
√
1 + [f′(x)]^2
The vector T = dr
dx
=ˆe 1 +f′(x)ˆe 2 and unit vector ˆet=
ˆe 1 +f′(x)ˆe 2
√
1 + [f′(x)]^2
are tangent to
the curve and one can verify that ˆen·ˆet= 0 showing these vectors are orthogonal.
Applications for the Gradient
In two-dimensions, let ˆeα= cos αˆe 1 + sin αˆe 2 denote a unit vector in an arbitrary,
but constant, direction αand let φ=φ(x, y )denote any scalar function of position.
At a point (x 0 , y 0 ),the directional derivative of φin the direction αbecomes
dφ
ds = grad φ·ˆeα=
∂φ
∂x cos α+
∂φ
∂y sin α
and the magnitude of this directional derivative changes as the angle αchanges. As
the angle αvaries, the maximum and minimum directional derivatives, at the point
(x 0 , y 0 ), occur in those directions αwhich satisfy
d
dα
[
dφ
ds
]
=−∂φ
∂x
sin α+∂φ
∂y
cos α= 0. (7 .72)
Note there exists two angles αlying in the region between 0 and 2 π radians which
satisfy the above equation. These directions must be tested to see which corresponds
to a maximum and which corresponds to a minimum directional derivative. These
(^6) Always remember that there are two normals to a curve, namely ˆenand −ˆen,