Communication & Computer Engineering Department

Subject: Advanced Engineering Mathematics

Dr. Bessam Z. Hassan

Chapter 3: Vector Calculus and Integration Theorems

3.1 Differentiation of Vectors

Previously position vectors were discussed in detail and we saw that a position vector gives the displacement from the origin as well as the x-, y-, and z- coordinates of a particle locating its position in the 3D space as shown in Fig. 3.1. Now if this particle is moving with respect to time then its x-, y- and z- coordinates are functions of time. Furthermore the particle will have velocity and might have acceleration as well which can both be presented by vectors.

Fig. 3.1 The Position vector of a particle moving through space is a function of time

In this section we will restrict the definition of the vector function to the domain of a single real variable t. This will result only in 3-D curves as shown in Fig. 3.2. Later on, we will define vector functions with a domain in the plane or space which will give us a vector field.

Fig. 3.2 Space curves defined by position vectors

To define the derivative of a vector, let O be the origin and P be the position of a moving particle at time t as shown in Fig. 3.3 The position vector of the point is given by . Now if this particle is moving with time then let Q be its position at time t+t From Fig. 3.1 it can be seen that
O

Fig. 3.3 Position and Velocity vectors

) ( The definition of the change in position is given by Since this change occurred during a small instant of time t then the motion can be expressed by the velocity

which is a vector. As t0, Q

tends to P and the chord PQ becomes the tangent at P. The general definition of the velocity vector is

Note that

is a vector in the direction of the tangent at P. It gives the velocity of

of the particle at P which is along the tangent to its path. This derivative can be found by finding the derivative of the components of the vector .

is also called the differential coefficient of with respect to t.

In a similar way we can define

as the second derivative of . Also

gives the acceleration of the particle at P. | |. ) | | ( ) | |

In this sense speed is defined as So ( )(

Theorems of Differentiation: Using the theories of vectors it can be shown that (i) (ii) ( ( ) )

(iii) (iv)

( ) ( )

where the order of the functions is not to be changed Ex: A particle moves along the curve in the direction Sol. The position vector The velocity

where

is the time. Find the component of its velocity and acceleration at time . ( ) ( )

and when t = 1, we have

Now to find the component of in the direction of orthogonal projection from chapter 1 ( | | | | )( )

we use

Now acceleration

and when t = 1, we have Component of acceleration in the direction of ( | | | | Hw1: The position vector of a particle at time is ( ) ( ) Find the condition imposed on by requiring that at time t=1, the acceleration is normal to the position vector. )( is given by )

3.2 Application: Projectile Movement

When we shoot a projectile into the air, we usually want to know beforehand how far it will go (will it reach the target?), how high it will rise, and when it will land. We get this information from the direction and magnitude of the projectiles initial velocity vector, using Newtons

second law of motion. This application is highly important to communication engineers to understand the physics and geometry of placing satellites in orbit. To derive equations for projectile motion, we assume that the projectile behaves like a particle moving in vertical coordinate plane and that the only force acting on the projectile during its flight is the constant force of gravity which always points straight down producing a downward acceleration of g. We will neglect the effects of Earths motion, air friction and gravitational force change. We will assume that the projectile was launched from the origin at time t=0 into the first quadrant with an initial velocity vo at an angle with the horizontal (| | ) (| | ) The projectiles initial position is ro as shown in Fig. 3.4 Newtons second law of motion says that the vertical downward acceleration due to gravity is g so if r is the projectiles position vector and t is the time then

This is a vector-differential equation subject to the following initial conditions (| | ) (| | )

The solution can be found by integration giving the following motion vector (| The | ) of ((| r | give ) the ) parametric
Fig. 3.4 Projectile Motion

components

equations ( ) (| | ) and ( ) (| | )

The position vector at any time t can be found and hence the position of the projectile can be simply drawn Projectile Height, Flight Time and Range To find the maximum projectile height, maximize y(t) | | (| | ) Substitute this value of t get (| | )

To find the maximum projectile range, set y=0 and solve for t to get the time of flight T ( ) (| | ) | |

Substitute this value in x(t) to get the maximum horizontal range R | | | | ( ) (| | ) Ex: A bullet projectile is fired from the origin over horizontal ground at an initial speed of 500m/s and a launch angle of 60 (a) (b) (c) Sol. (a) (( (b) , (c) (( (|
(| | | |

Where will the projectile be 10 seconds later? Find the maximum height, flight time and range of the projectile Find the speed of the projectile when it hits the ground back

(|

| )

((|

) )

( (

) )

| )

((| )

) | |

In conclusion a bullet fired in the air will reach a maximum height of less than 12km, travel a distance of less than 25km and take less than 2

minutes to fall down Hw2: Find the launch angles that will give the maximum 1. Height 2. Range 3. Flight time Hw3: A baseball is hit when it is 1 m about the ground. It leaves the bat with an initial speed of 50 m/s making an angle of 20 with the horizontal. At the instant the ball is hit, the wind was blowing horizontally opposite to the ball direction adding a component of -2.8i (m/s) to the balls initial velocity (a) (b) (c) Find the position vector of the path of the baseball How high does the baseball go when it reaches its maximum height? Assuming the ball is not caught , find its range and flight time

3.3 Arc Length, and Unit Tangent Vector

Airplanes making turns and twists are subject to very high forces due to the high accelerations. Turns that are too tight or climbs that are too steep can even breakup the airplane or spin it out of control to crash to Earth.
Airplane.wmv

Depending on the shape of the curve we can determine the sharpness of its turning. One of the features of smooth space curves is that they have a measurable length. This means that we can locate points along the curve by knowing their direct distance s along the curve from some base point or reference as shown in Fig. 3.5. In this figure the time variable describes the motion of the particle while s describes the distance traveled along this curve
Fig. 3.5 Smooth curves can be scaled like number lines

Definition: Length of a Smooth Curve The length of a smooth curve traced only once as t increases from ( ) ( ) ( ) ( ) to ( ) ( ) , is that is

Please note that the square root in this definition is the magnitude of the velocity vector = || | | so we can say || Ex: An airplane is gliding upward along the helix ? Sol. || ( ( ) ) ( ) ( ) ( ) ( to ) .

How far does the airplane travel along it path from

Fig. 3.6 Example

If we choose a base or reference point ( ), each value of t will gives us a point ( ) point. If if ( ( ) ( ) ( )) on the curve with respect to the reference , then ( ) is the positive distance from ( ) to ( ) while

, then ( ) is negative. For any value of t, the distance s(t) will be ( ( ))

called the are length parameter and will be defined as ( ) ( ( )) ( ( )) |( )|

HW4 Show that if

is a unit vector, then the arc

length parameter ( ) along the line ( ) ) from the point ( ) where

) itself

is equal to

Note: since the curve is smooth then ( ) is differentiable with respect to and the derivative is a scalar function that gives the speed of the |( )| will be an increasing function of time. That is never zero for a smooth curve Unit Tangent Vector T We already know that the velocity vector and the vector || is therefore a unit vector tangent to the curve as shown in Fig. 3.7 Definition The unit tangent vector to a smooth curve ( ) is given by ||

object

since |( )| is

is tangent to the curve

Fig. 3.7 Unit tangent vector

Ex: Find the unit tangent vector of the curve Sol. Thus
||

||

HW5: Given a counterclockwise moving phasor in 2D plane as shown in Fig. 3.8 which can be described by the position vector ( ) find the unit tangent vector to the circle ( ) ( ),

Fig. 3.8 Homework

3.4 Curvature, and Unit Normal Vector

As a particle moves along a smooth curve in the plane,

turns as the

curve bends. Since is a unit vector, its length remains constant and only its direction changes as the particle moves along the curve as shown in Fig. 3.9

Fig. 3.9 As P moves along the curve in the direction of increasing length, the unit tangent vector turns.

The rate at which T turns per unit of length along the curve is called the curvature Definition If T is the unit vector of a smooth curve, the curvature function of the curve is || The curvature is a scalar function and if | | is large, turns sharply as the particle passes through , and the curvature at is smaller Ex: On a straight line, the unit tangent vector always points in the same direction, so its components are constants. Therefore | | | | as shown in Fig. 3.10 | | | | | |

Fig. 3.10 Along a straight line, T always points in the same direction. Therefore the curvature is zero.

Ex: Find the curvature of a circle of radius a Sol. The position vector for a circle of radius a is ( ) ( ) ( ) ( ) ( ) || So Thus || ( ) ( ) ( ) ( )

( (

) )

( (

) )

Hence for any value of t || | |

There are many vectors normal to the vector but one of them is of special interest since it points in the direction in which the curve is turning. Note that will be a function of the parameter

so it can be considered

a curve of . Since has constant length (unit vector), the derivative is orthogonal to . Therefore if we divide

by its length , we obtain a

unit vector orthogonal to as shown in Fig. 3.11

Fig. 3.11 The vector dT/ds, normal to the curve, always points in the direction in which T is turning. The unit normal vector N is the direction of dT/ds

Definition At a point where 0 the principal unit normal vector for a smooth curve in the plane is Where The vector

| || |

| |

points in the direction in which turns as the curve bends.

Therefore, if we face in the direction of increasing arc length, the direction of the vector

is towards the right if turns clockwise and

toward the left if turns counterclockwise. In other words, the principal normal vector will point toward the concave side of the curve as can be seen in Fig. 3.11 Note that we can find without having to find or s Ex: Find and for the circular motion ( ) Sol. We first find , ( ) || From which we find || Notice that | | ( ) ( ) || ( ) ( ) ( ) ( )

, verifying that is orthogonal to . Notice too, that for the circular motion here, points from ( ) towards the circles center at the origin. Note: The binormal vector of a curve in space is orthogonal to both and as shown in Fig. 3.12 , a unit vector

Fig. 3.12 The mutually orthogonal vectors travelling along a curve in space

Ex: Find the curvature for the helix ) ( ) ( Then find the principal normal vector and the binormal vector Sol. The helix is shown in Fig. 3.13. First we find the velocity ( ) ( ) ( || || | | | ) [ ( ] [ ( ( ) ) ( ( )] ) ) ( )

||

Fig. 3.13 Example

| |

[ (

)]

Note that always points towards the z-axis To find the binormal vector we apply

[( ) ( ) ] HW6: For the following position vector find the unit normal, tangent and binormal vectors ( ) ( )

3.5 Gradient Vectors and Directional Derivatives

In the previous section we considered a vector function that is parameterized to become a function of a single variable . In this section we return to the concept of differentiation of functions of two or more variables but we will start with scalar functions of two variables ( Recall that the partial derivative function ( Similarly, the partial derivative ). was the rate of change of the is the rate of change if you walk in ) if you walk in some other

) if you walk in the positive x-direction with unit speed.

the positive y-direction with unit speed. We are left with the obvious question, What is the rate of change of ( direction? Let represent a vector which points in the direction of travel. The length of will be one, to reflect the fact that we are walking with unit speed. We already know two things: Rate of change of Rate of change of ( ( ) in direction of x or in the direction ) in direction of y or in the direction

It is reasonable, then, to expect that Rate of change of ( Ex: Suppose ( ) ) in direction of x or in the direction . Then what is the rate of change of (

), at

the point (2, 1), in the direction

Sol. The answer will be

The rate of change of (

) (scalar function of two or more variables),

in the direction of , is called a directional derivative, and is denoted as ( ). Hence, we have the formula

But notice that the right side of this equation can be rewritten as a dot product:

We can also write this equation shorter by coming up with a new notation for for ( . The best way is to name this last vector as ) then becomes

) and we

will call this vector the gradient of f, and denote it as

. The equation

( ) is a vector as long as

From this definition we can now see that ( ) is a scalar function and that ( Sol. ( ) ) )

Ex: Find the gradient of the function (

at the point (x, y)

What Do the Directional Derivative and the Gradient Mean? The gradient at the point (1, 1) in the previous example, then, would be ( ) Remember that the rate of change of a function is also the slope of a tangent line to its graph, as long as you are traveling with speed one. Heres a nice way to think about the situation. Suppose you are climbing a mountain, and you have a good trail map in your hands. Let ( ( ) be your coordinates when you locate yourself on the map. The function ) is your elevation at that point. Now turn your body to face the direction (on your map). If you sight up or down so that your gaze just grazes the mountainside then you are looking along the tangent line

whose slope is given by

) as shown in Fig. .

Fig. 3.14 Interpretation of the directional derivative

Recall that the dot product is given by the product of the magnitudes of the two vectors, times the cosine of the angle between them. If we fix the point we are at then way to change

is a fixed vector. If ||

, then the only . is

is to change the angle between and

The largest this quantity can be is when the value of cos is largest. This happens when = 0. We conclude that the largest value of given by

|||

and that this value is attained when (the direction we are facing) coincides with the direction of . In other words, if you are on the mountainside and you want to face directly uphill you should point yourself in the direction of the gradient vector. When you do this and sight along the mountainside the slope you see is the magnitude of the gradient vector. (Q) What happens if you are in a valley and you want to face the bottom of the valley? EX: Let ( ) find the largest directional derivative of this function at the point (2,3) and the largest slope for any tangent line at

this point. Sol. At the point (x, y) the gradient vector is ( ) ( ) ) ( ) . The largest slope So, at the point (2, 3) we have | ( )| (

of any tangent line to the graph at the point (2, 3) is then given by

(Q) What if you were standing on the mountainside and wanted to face the direction you would have to travel to keep your elevation constant? In other words, how would you find the direction of your level curve? (A) If you were facing such a direction you would be looking along a horizontal line, i.e., a line whose slope is zero. The only way for to be zero is for and ) to be perpendicular. . In the previous example we saw ( )

EX: Suppose again ( would be

. A vector which points in a direction perpendicular to this (Check this!). Hence, this vector is tangent to a level curve at the point (2, 3). HW7: For each of the functions questions: Find the gradient vector . Find a unit vector that points in the direction of the maximum rate of change at the point (1, 1). Find the largest slope of any tangent line at (1, 1). Find a unit vector that lies in a line tangent to a level curve through (1, 1). 1. 2. 3. ( ). ) at the point (1, 1) in the direction Calculate the rate of change of ( ( ) below answer the following

3.6 Maxima, Minima, and Saddles

At a local maximum or a local minimum of a graph the tangent plane is horizontal. Another way to say this is that the slope of every tangent line to a local maximum or minimum is zero. But recall that the slope of a tangent line at ( ) in direction is given by . The only way for this to be zero for every possible is if EX: Show that ( Sol. The gradient is If x and y are nonzero then this is not the zero vector. This tells us that there is a direction where the slopes of tangent lines are nonzero, and hence we cannot be at a local maximum or minimum. Unfortunately, just because the gradient is the zero vector it does not necessarily mean that there is a local maximum or minimum. EX: The gradient of Sol. The function ( ( ) and ( ) is at ) is the zero vector

cannot have a maximum or minimum at any

point where x and y are nonzero.

the origin. What is the difference between the two? ) has a minimum there, while ( ) has a saddle. In the first-semester we learned to detect local maxima and minima by a second-derivative test. We would like to do the same thing here. The problem, of course, is that there are four second partial derivatives! To keep track of all this information we often write them in a matrix, as follows:

[ ] Now we examine this matrix for several functions whose graph is familiar. Each of these has a gradient vector equal to zero at the origin. 1. . This function has a local minimum at the origin. The matrix of second partials is

* 2. partials is * Let us take other examples 3.

. This function has a saddle at the origin. The matrix of second +

. This function has a maximum at the origin. The matrix of * +

second partials is 4. . This function also has a saddle at the origin. The matrix of second * +

partials is The first and third examples give us a clue as to the quantity we would like to look at. Consider the product of the upper-left and lower-right entries of the matrix. For the maximum and minimum above this quantity is positive and for the first of the above saddles it is negative. However, this alone would not be enough to distinguish maxima and minima from saddles, as the second of the saddles shows. To compensate we must subtract the product of the upper-right and lower-left entries, yielding the formula

However, since the mixed partials are equal we can shorten this to ( )

This is indeed the right quantity to look at, in the sense that if it is greater than zero you have a maximum or minimum, and if it is less than zero you have a saddle. Unfortunately, if it is zero you have no information; you may be at a maximum, minimum, saddle, or something much more bizarre. Nonetheless, we will single this out as our first test.

EX: Find the location of all saddles of the function

Sol. First, we will need to narrow down the possibilities by finding the critical points. To do this we find the gradient. ( ) ( ) Setting this equal to the vector

gives us the system of equations

The first equation tells us that equation then gives Either (and hence

. Plugging this into the second

) or we can divide both sides of this

equation by y to get Solving then gives us Hence, we have critical points at (0, 0) and (-2/9, 2/9). To determine which of these are saddles we compute the matrix of second partials: [ [ And so ( ) ( ) ) We now check each critical point: ( ) ( ) ( ] ]

When

it would be nice to have a second test to determine ( ) and ( )

whether ( .

) is a local maximum or a local minimum. Such a test can be

easily guessed from our typical examples, Notice that in both cases and

have the same sign. When this sign

is positive we have a local minimum and when it is negative we have a local maximum. This is precisely our second test.

EX: Recall the function

from the previous

example. We found critical points at (0, 0) and (-2/9,2/9), and determined that at (-2/9,2/9) there was a local maximum or a local minimum. To determine which we need only look at as shown in Fig. 3.15. . Since this was 2, and 2 0, we conclude that at this critical point there is a local minimum

-2

-4 1

0.5

-0.5

-1

-1

-0.5

0.5

Fig. 3.15 Example

It is important to keep in mind that if information about the nature of ( example. EX: Consider the following functions: 1. ( 2. ( 3. ( ) ) ) .

then we have no

). We illustrate this in the next

In each case the only critical point is at (0, 0) and maximum, and in the third there is a saddle.

. But at (0,

0) in the first case there is a local minimum, in the second there is a local

HW8: Find the local maxima, minima, and saddles of the following functions. 1. 2.

HW9: For the function sin(x + y) show that D(x, y) = 0 at every point (x, y). Does this function have maxima, minima, or saddles? HW10 If, for some point ( then show that instead of ( ( ) ) you know ( ) and ( ( ) )

. (This tells us that you may use

) when distinguishing maxima from minima.)

3.7 Vector Fields and Gradients

Definition A vector field is simply a choice of vector for each point. So, for example, a vector field on R2 would have some vector at the point (1, 2), some other vector at the point (1, 1), etc. We often draw vector fields by picking a few points and drawing the vector based at those points as shown in Fig. 3.16.

Fig. 3.16 Vector Field

More formally, a vector field is a function from R2 to R2. What goes in to this function are the coordinates of the point where you are. What

comes out are the components of the vector at that point. EX: Consider the vector field field contains the vector the vector . At the point (1, 1) this vector . At the point (2, 4) it contains . If we use a computer to draw it we would

see something like what is shown in Fig. 3.17.

Fig. 3.17 Example

EX: The vector field

is shown in Fig. 3.18.

Fig. 3.18 Example

HW11 Sketch the following vector fields. 1. 2. 3. We have already seen many vector fields, although we did not use this language. Whenever we take a function f and compute its gradient f at a point we get a vector. The set of all such vectors is then a vector field, which we now call grad f. EX: Suppose ( is ( ) ) . Then the gradient of ( ) at the point ( ) . If we draw this vector at various values of x and

y, we get the picture depicted below in Fig. 3.19.

Fig. 3.19 Example

EX: Let (

. Then

HW12 For each of the following functions, f, compute f. 1. ( 2. ( 3. ( ) ) )

3.8 Divergence
In the previous section, we saw that the gradient operator gives us a way to take a function ( ) and get a vector field. In this section, we explore a way to take a vector field and get a function. Eventually, we will see that the value of this function at a point is a measure of how much the vector field is spreading out there. Definition Suppose ( ) ( ( ) ( ) ) ( ( ) is a vector field V on R3 (that is ) ). Then we define the divergence of

V, Div V, to be the function

Note that the first term is associated with the first component of V, the second term with the second component, and the third term with the third component. This, and the fact that the terms are being added, should remind you of the dot product. This gives us a purely notational way to remember how to calculate the divergence of a vector field. We let denote the vector

This, of course, is only a vector in a notational sense. But if we suspend our disbelief for a moment and allow such absurdities, we can write the formula for the divergence of a vector field in a very compact way: