CALCULUS II

Chapter 1: FUNCTIONS OF SEVERAL VARIABLES

LE THAI THANH

HCMC UNIVERSITY OF TECHNOLOGY

Unit 2: DIFFERENTIALS AND DIRECTIONAL DERIVATIVES

DIFFERENTIALS
Now consider a function of two variables, z “ f px , y q, and
suppose x changes from a to a ` ∆x and y changes from b
to b ` ∆y . Then the corresponding increment of z is
∆z “ f pa ` ∆x , b ` ∆y q ´ f pa, bq

Thus the increment ∆z represents the change in the value

of f when px , y q changes from pa, bq to pa ` ∆x , b ` ∆y q.
By analogy with the functions of one variable we define the
differentiability of a function of two variables as follows.

Definition 1: If z “ f px , y q, then f is differentiable at pa, bq

if ∆z can be expressed in the form
∆z “ fx pa, bq∆x ` fy pa, bq∆y ` ε1 ∆x ` ε2 ∆y
where ε1 and ε2 approach 0 as p∆x , ∆y q Ñ p0, 0q.
Theorem 1: If the partial derivatives fx and fy exist near
pa, bq and are continuous at pa, bq, then f is differentiable at
pa, bq.

Example 1: Show that f px , y q “ x exy is differentiable at

p1, 0q and find its linearization there. Then use it to approx-
imate f p1.1, ´0.1q.

Solution : The partial derivatives are fx “ exy `xy exy and

fy “ x 2 exy . Therefore, fx p1, 0q “ 1, fy p1, 0q “ 1. Both fx
and fy are continuous functions, so f is differentiable. The
linearization is

f px , y q “ f p1, 0q ` fx p1, 0qpx ´ 1q ` fy p1, 0qpy ´ 0q “ x ` y

The corresponding linear approximation is

ñ x exy « x ` y

so
f p1.1, ´0.1q « 1.1 ´ 0.1 “ 1
Compare this with the actual value of f p1.1, ´0.1q “
1.1 e´0.11 “ 0.98542.
Definition 2: For a differentiable function of two variables,
z “ f px , y q, we define the differentials dx and dy to be
independent variables; that is, they can be given any values.
Then the differential dz, also called the total differential, is
defined by
Bf Bf
dz “ fx px , y qdx ` fy px , y qdy “ dx ` dy
Bx By

If we take dx “ ∆x “ x ´ a and dy “ ∆y “ y ´ b in this

formula, then the differential of z is
dz “ fx pa, bqpx ´ aq ` fy pa, bqpy ´ bq
So, in the notation of differentials, the linear approximation
can be written as
f px , y q “ f pa, bq ` dz
Example 2:
(a) If z “ f px , y q “ x 2 ` 3xy ´ y 2 , find the differential dz.
(b) If x changes from 2 to 2.05 and y changes from 3 to
2.96, compare the values dz with ∆z.

Solution :
(a) We have fx “ 2x ` 3y , fy “ 3x ´ 2y , and so

dz “ p2x ` 3y qdx ` p3x ´ 2y qdy

(b) Putting x “ 2, dx “ ∆x “ 0.05, y “ 3, and

dy “ ∆y “ ´0.04, we get

dz “ r2 ¨ 2 ` 3 ¨ 3s0.05 ` r3 ¨ 2 ´ 2 ¨ 3sp´0.04q “ 0.65

The increment of z is

∆z “ f p2.05, 2.96q ´ f p2, 3q

“ rp2.05q2 ` 3 ¨ 2.05 ¨ 2.96 ´ p2.96q2 s
´ r22 ` 3 ¨ 2 ¨ 3 ´ 32 s “ 0.6449
Example 3: The base radius and height of a right circular
cone are measured as 10 cm and 25 cm, respectively, with
a possible error in measurement of as much as 0.1 cm in
each. Use differentials to estimate the maximum error in the
calculated volume of the cone.

Solution : The volume V of a cone with base radius r and

π
height h is V “ r 2 h. So the differential of V is
3
BV BV 2πrh πr 2
dV “ dr ` dh “ dr ` dh
Br Bh 3 3
Since each error is at most 0.1 cm, we have |∆r | ď 0.1,
|∆h| ď 0.1. To estimate the largest error in the volume
we take the largest error in the measurement of r and of h.
Therefore we take dr “ 0.1 and dh “ 0.1 along with r “ 10,
h “ 25. This gives

2π ¨ 10 ¨ 25 π ¨ 102
dV “ 0.1 ` 0.1 “ 20π
3 3
Thus the maximum error in the calculated volume is about
20π cm3 « 62.83 cm3 .
THE CHAIN RULE
The Chain Rule (Case 1)

Suppose that z “ f px , y q is a differentiable function of x

and y , where x “ gptq and y “ hptq are both differentiable
functions of t. Then z is a differentiable function of t and
dz Bz dx Bz dy
“ `
dt Bx dt By dt

Example 4: The pressure P (in kilopascals), volume V (in

liters), and temperature T (in kelvins) of a mole of an ideal
gas are related by the equation PV “ 8.31T . Find the rate
at which the pressure is changing when the temperature is
300 K and increasing at a rate of 0.1 K {s and the volume is
100 L and increasing at a rate of 0.2 L{s.
If t represents the time elapsed in seconds, then at the given
dT dV
instant we have T “ 300, “ 0.1, V “ 100, and “
dt dt
T
0.2. Since P “ 8.31 , the Chain Rule gives
V
dP BP dT BP dV 8.31 dT 8.31T dV
“ ` “ ´
dt BT dt BV dt V dt V 2 dt
8.31 8.31 ¨ 300
“ p0.1q ´ p0.2q « ´0.04155
100 1002
The pressure is decreasing at a rate of about 0.042 kPa{s.
The Chain Rule (Case 2)

Suppose that z “ f px , y q is a differentiable function of x and

y , where x “ gps, tq and y “ hps, tq are both differentiable
functions of s and t. Then z is a differentiable function of s
and t and
Bz Bz Bx Bz By Bz Bz Bx Bz By
“ ` “ `
Bs Bx Bs By Bs Bt Bx Bt By Bt
The Chain Rule can be used to give a more complete descrip-
tion of the process of implicit differentiation. We suppose
that an equation of the form F px , y q “ 0 defines y implicitly
as a differentiable function of x , that is, y “ f px q, where
F px , f px qq “ 0 for all x in the domain of f . If F is differen-
tiable, we can apply Case 1 of the Chain Rule to differentiate
both sides of the equation F px , y q “ 0 with respect to x .
Since both x and y are functions of x , we obtain
BF dx BF dy
` “0
Bx dx By dx
But dx {dx “ 1, so if BF {By ‰ 0 we solve for dy {dx and
obtain
BF
dy Fx
“ ´ Bx “ ´
dx BF Fy
By
Now we suppose that z is given implicitly as a function z “
f px , y q by an equation of the form F px , y , zq “ 0. This
means that F px , y , f px , y qq “ 0 for all px , y q in the domain
of f . If F and f are differentiable, then we can use the Chain
Rule to differentiate the equation F px , y , zq as follows:
BF Bx BF By BF Bz BF Bx BF By BF Bz
` ` “ 0, ` ` “0
Bx Bx By Bx Bz Bx By By By By Bz By
 Bx By By Bx
But “ 1, “ 0, “ 1, and “ 0, we obtain
Bx Bx By By
BF BF Bz BF BF Bz
` “ 0, ` “0
Bx Bz Bx By Bz By
We com to the following formulas

BF BF
Bz Fx Bz By Fy
“ ´ Bx “ ´ ; “´ “´
Bx BF Fz By BF Fz
Bz Bz
DIRECTIONAL DERIVATIVES
 Suppose that we now wish
to find the rate of change
of z “ f px , y q at px0 ; y0 q in
the direction of an arbitrary
unit vector Ñ Ý
u “ pa; bq.

To do this we consider the surface S with the equation z “

f px , y q (the graph of f ) and we let z0 “ f px0 , y0 q. Then the
point Ppx0 ; y0 ; z0 q lies on S. The vertical plane that passes
through P in the direction of Ñ Ýu intersects S in a curve C .
The slope of the tangent line T to C at the point P is the
rate of change of z in the direction of Ñ Ý
u.
If Qpx ; y ; zq is another point on C and P 1 , Q 1 are the pro-
ÝÝÑ
jections of P, Q onto the xy -plane, then the vector P 1 Q 1 is
parallel to Ñ Ý
u and so
ÝÝ Ñ
P 1 Q 1 “ hÑ
Ý
u

for some scalar h. Therefore x ´ x0 “ ha, y ´ y0 “ hb, so

x “ x0 ` ha, y “ y0 ` hb, and
∆z z ´ z0 f px0 ` ha, y0 ` hbq ´ f px0 , y0 q
“ “
h h h
If we take the limit as h Ñ 0, we obtain the rate of change
of z (with respect to distance) in the direction of Ñ Ýu , which
is called the directional derivative of f in the direction of Ñ
Ýu.
Definition 3: The directional derivative of f at px0 ; y0 q in
the direction of a unit vector Ñ
Ý
u “ pa; bq is

Bf f px0 ` ha, y0 ` hbq ´ f px0 , y0 q

px0 , y0 q “ DÝ
u f px0 , y0 q “ hÑ0
Ñ lim
BÑ
Ý
u h
if this limit exists.

By comparing with the definitions of the partial derivatives,

Ñ
Ý
we see that if Ñ Ý
u “ i “ p1; 0q, then DÝ Ñ f “ fx and if
Ñ
Ý i
Ñ
Ý
u “ j “ p0; 1q, then DÝ Ñ f “ fy . In other words, the partial
j
derivatives of f with respect to x and y are just special cases
of the directional derivative.
Theorem 2: If f is a differentiable function of x and y , then f
has a directional derivative in the direction of any unit vector
Ñ
Ý
u “ pa; bq and
DÝ
u f px , y q “ fx px , y qa ` fy px , y qb
Ñ

Example 5: Find the directional derivative DÝ

u f of
Ñ

f px , y q “ x 3 ´ 3xy ` y 3
ˆ? ˙
Ñ
Ý 3 1
in the direction of unit vector u “ ; . What is
2 2
DÝu f p1, 2q?
Ñ
Solution : We have

fx “ 3x 2 ´ 3y , fy “ 3y 2 ´ 3x

and
?
2 3 2 1
DÝ
u f px , y q “ p3x ´ 3y q 2 ` p3y ´ 3x q 2
Ñ

Therefore
? ?
2 3 2 1 9´3 3
u f p1, 2q “ p3 ¨ 1 ´ 3 ¨ 2q 2 ` p3 ¨ 2 ´ 3 ¨ 1q 2 “
DÝ
Ñ
2
THE GRADIENT VECTOR
Definition 4: If f is a function of two variables x and y , then
the gradient of f is the vector function ∇f defined by
Bf Ñ
Ý Bf Ñ
Ý
∇f px , y q “ pfx px , y q; fy px , y qq “ i ` j
Bx By

With this notation for the gradient vector, we can rewrite

the formula for the directional derivative of a differentiable
function as
Ñ
Ý
DÝ
Ñu f px , y q “ ∇f px , y q ¨ u
This expresses the directional derivative in the direction of a
unit vector Ñ
Ý
u as the scalar projection of the gradient vector
Ñ
Ý
onto u .
Example 6: Find the directional derivative of the function
f px , y q “ x 2 y 3 ´ 4y at the point p2; ´1q in the direction of
Ñ
Ý Ñ
Ý
the vector Ñ Ý
v “2 i `5 j .

Solution : We first compute the gradient vector at p2; ´1q:

Ñ
Ý Ñ
Ý Ñ
Ý Ñ Ý
∇f px , y q “ 2xy 3 i `p3x 2 y 2 ´4q j ñ ∇f p2, ´1q “ ´4 i `8 j
?
Note that ÑÝv is not a unit vector, but since |Ñ
Ý
v | “ 29, the
unit vector Ñ
Ýu in the direction of Ñ
Ýv is
Ñ
Ý
v 2 ÑÝ 5 ÑÝ
Ñ
Ý
u “ Ñ “? i `? j
Ý
|v| 29 29
Therefore, we have

Ñ
Ý
u f p2, ´1q “ ∇f p2, ´1q ¨ u ˆ
DÝ
Ñ
˙
Ñ
Ý Ñ
Ý 2 ÑÝ 5 ÑÝ
“ p´4 i ` 8 j q ¨ ? i ` ? j
29 29
´4 ¨ 2 ` 8 ¨ 5 32
“ ? “?
29 29
Theorem 3: Suppose f is a differentiable function of two
variables. The maximum value of the directional derivative
DÝ Ñ
Ý
Ñu f px , y q is |∇f px , y q| and it occurs when u has the same
direction as the gradient vector ∇f px , y q.

Example 7:
(a) If f px , y q “ x ey , find the rate of change of f at the
point Pp2; 0q in the direction from P to Qp1{2; 2q.
(b) In what direction does f have the maximum rate of
change? What is this maximum rate of change?
(a) We first compute the gradient vector:

∇f px , y q “ă ey , x ey ą ñ ∇f p2, 0q “ p1; 2q
ÝÑ
The unit vector in the direction of PQ “ p´1.5; 2q is
Ñ
Ý
u “ p´3{5; 4{5q, so the rate of change of f in the
direction from P to Q is

Ñ
Ý
u f p2, 0q “ ∇f p2, 0q ¨ u “ p1; 2q ¨ p´3{5; 4{5q “ 1
DÝ
Ñ

(b) According to the Theorem, f increases fastest in the

direction of the gradient vector ∇f p2, 0q “ p1; 2q. The
maximum rate of change is
a ?
|∇f p2, 0q| “ |p1; 2q| “ 12 ` 22 “ 5
That’s all. Thanks a lot

...