BOOK TWO

part one

B.'SOER & H.DEMiR

'

BOOK TWO
. PART ONE

!)cvyWt

t+cd~t~ 13/1ff:ofl-e
l:;. '2
d/ /-ulle vf 11Ve_-

' ).Jtv<r\\,1\X

C:

ciA/\_

~r

SECOND EDITION

HOsEvt N DHU R

BEDRi SOER-

M.E.T.U.
1980

PREFACE
Being well aware of lhe

:;

of excellent

exjstertc~

textbook~

of similar content, before a~ding another one to'tite mark~t. we

humbly feel that _the nature of the challenge, wh-ich motivated .us
to prepare

~his

work, needs a justifying explanation. Summing up

we may state briefly the following

facts~

a. This work is designed primarily for studen'ts who were

- like the onesat METU - to utmost one year intensive
language trajning in-English,

b. Its content is closely related

t~ th~

syllabus traditio-

nally followed at METU and sim~lar institu~ions,

c. It is a practical answer to the ever_increasing demand,
caused by contemporary currency fluctuations which
effectively cu_rb the

avail~b1lity

of the t,extbooks

etlited_ abroad.
We s1ncerely believe that the topics treated
volumes, each
~nd

cont~ining

in~luding

the

the two

two distinct parts, are self ct>ntained

compact. Each_ part and eaFh

exercises,

i~

a~swefs

cha~ter

is provided with numerous

corresponding to the even number-

ed ones ..
We

~xpress

enco~rage_ments

1980
ANKARA

our gratitude to our colleagues for their constant

and to Miss Zehra Uner for her careful typing.

CONTENTS
CHAPTER 1 - . SEQUENCES AND NUMBERS

f. 1.

SEQUENCES OF NUMBERS,
A, DEFINITIONS,

Determination of sequences by recurrence relations, 3

Algebra of sequences, 5
Subsequences,

B, NATURE OF SEQUENCES,
1~

Monotonicity, 6

2.

Bound~dness,

3. Convergence,,

6
L

.7
9

C, SOME IMPORTANT CONVERGENT SEQUENCES,

EXERCISES (I -10)-,
ANSWERS,

1, 2.

14

16

SERIES OF NUMBERS,
A, DEFINITIONS,
Arithm~tic,

17

17

geometric, harmonic series,

Bo SERIES OF POSITIVE TERMS,

20

1. Integral test {of McLAURIN),

p-series,

3. Intrinsic tests,

26

RAABE-DUHAMEL's test,
ALTERNATING

21

22

2. .Comparison with other series,

c.

12

SERI~S,

28

27

23

~8

D. SERIES OF. ARBITRARY TERMS,

E. EVALUATION OF SERIES,

EXERCISES (11-30)
ANSWERS,
1.~

30

32

34

36

3, PowER SERIES, _38

A, DEFINITIONS,

38

B. REPRESENTATION OF FUNCTIONS BY POWER SERIES, 40

Binomial Series,

44

C, ALGEBRA OF POWER SERIES, - 46

Euler Formula,

~7

D, THE -REMAINDER THEOREM,

. 50

E, SOME OTHER APPliCATIONS,- 55

a; Evaluation of lfmits in indeterminate forms (0/0, =-=), 55
55

b. Evaluation. of inte.grals,

c. Determination of the equation of

a given point with-. y'
d.

Evaluati~n

of series

-EXERCISES (31-60);
ANSWERS,

61

A SUMMARY,.

63

f(x, -y),

curve through

56

constant terms,

58

MISCELLANEOUS EXERCISES,
ANSWERS, 7o .

~f

th~

(61~120),

65

57

CHAPTER 2

2. i

MATRICES AND SOLUTION SYSTEMS OF LINEAR

E~UATION BY. f'lATRICES

MATRICES, .

71

A, DEFINITIO~S,

71

Transpose of a matrix,

72

Some s~ecia1 square matrices,

73

14

B, OPERATIONS WITH REAL MATRICES,

C, PROPERTIES,

79

.~

82

D, INVERSE OF A f!IATRIX,

Square case,

84

Determination of the inverse of a square matrix,

.

1. by the definitions,

.. \

86

86

2. by the'adjoint matrix, .8.6

3. by elementary row operatfons,
4. By Cay1ey-Hafuilton Theorem,
EXERCISES.Ci-30)~

ANSWERS,

2. 2.

91
93

95

100

SoLUTION OF SYSTEMS OF LINEAR EQUATIONS BY MATRices,

1.

~olution

by inverse

matrix~

.2. Scilution by Gauss Method,

EXERCISES C3t-45),.
ANSWERS,
A SUMMARY,

1~3

108

111

112

MISCELLANEOUS EXERCISES

Answers.

102

115

(46-65),

113

101

ANALYTIC GEOMETRY IN R

. CHAPTER 3

.3, 1. . VECTORS.
117

A, DEFINtTIONS,

Rectangu1 ar coordinate - systems,

121

Length of a .Vector,

s.

ALGEBRA-OF VECTORS,

1. Addition;

119

122

122

2. Multipfication .by scalar,

123 '

Direction angles., cos in.es and numbers,

3. Multiplication of vectors,

126

126

a) Scalar product o~ t~o vecto~s~

126

~e~metri~ interpret~tions,. 127

Physical 'interpretation, _127
b) Vector

p~oduct

'

Properties,

of two

vettor~.

13.2

Geometric interpretation,
/

Physical interpretation,
Triple products,
i. Mixed

Group, 139,

133
133

135

product~

135, Geometric. i nterpreta ti on, l37

ii. Tr..iple vector product,

C, VECTOR SPACES,

137

139

field, 141,

Vector space, 141

Linear dependence and independence 1n

Linear dependence in

EXERCISES {l-70)
I

ANSWERS,

16o

130

ls-I

Dn-l[I1 (-+, IR),

Rn, 144, Bases, 147

~49

2. PLANES AND LINES,

A, PLANES,

1-62

162

Equation~

of some special planes,

162

Equations of a plane through a p~int and perpendicularto

a vector,
Normal

163.
164

equation~

Equation of a plane
Intercept form,

throu~h

three points,

166

167

Li-near family of 1i nes ( penci 1 of planes L

169

169 - -

B. LINES,

Fnl!ations of some special lines.

169
Equation of a line thrdugh a point and parallel toa vectDr,
C,- INTERSECTION, ANGLE, DISTANCE,
174
Intersections,
Angles,

174

179

Distance,

183_

D, SKETCHING OF PLANES AND LINES,

187

EXERCISES UI -95), 189

ANSWERS,
SURFACEs,

192
193

A. RELATIPNS AND THEIR GRAPH~, _--193

Surfaces, 19'4,

lin~ar f~m~ly

Surface- Sketchint,

of surfaces,

196,

196

B. CYLINDRICAL AND SPHERICAL COORDINATE SYSTEMS,

198

1. Cylindrical coordinate_system (polar cbprdinates

in 3-space),

198

Jrans~orming

relations,

2. Spherical coordinate

198

~ystems,

Tra~~~o~~ing rel~tions,

c,.

199

199

CYLIN~RS, CONES, SURFACES OF REVOLUTION,

Cylinder!, 200,

~ones,

Particular cases,

S~faces

of revolution, 204

206

2.09

D, QUADRICS;

Ellipsoid

202,

200

{~nd spher~).

Paraboloids,

209, Hyperboloids,

209

209

'

E, SURFACES 0~ SECOND DEGREE, '213,

EXERCISES
. ANSWERS,

. 3.

4.

(96-115),

218

221

SPACE CURVES,

223.
223

-A, VECTOR FUNCTIONS,

227

B, SPACE CURVES,

Definitions,

277,

Arclength, . 229,
FRENET

f~ame,

Gepmetric interpretation of ~{t),

PhysicaJ interpretation of r{t),

231,

SERRET~F~E~ET

Cur~at~re of plane curves;

239

Curvilinear motion {in plane),

Curve Sketching,

243

EXERCISES CII6-140),: 244

ANswgRs, . 2.47
A SUMMARY, 2'/8
. MISCELLANEOUS EXERCISES
ANSWERS, 21

ci4I-22QL 25-"z

241

formulas,

234

228
230

CHAPTER I
SEQUENCES AND SERIES

I. I. SEQUENCES OF NUMBERS
A;

DEFINITIO~S

If
set

~P

subs~t,

f: D

i.s a function whose domain

of consecutive integers
then the

infinite~y

p, p+l, p+2,

, f(n),

as a

('1 )

seque~ce

written in this order, is called an infinite

where

, n, . .

th~

many numbers

f ( p) ; :f( p+ 1 ) ,

sequence;

D admits

f(p), f(p+l), , f(n)

or sjmply a

are called the .first

term, the second term, .. , fhe general term.respe'ttively.

, Fo.r

briev~ty

the subscript

n,

one d.eno.tes
say

an,

f(n) usually by a letter with

and the sequen.ce ( 1)

. or more simply. by

(f(n))P

or

1 , 2. 3',,

... :.

(a ) .
. n P

Examples

(n~2)3

' n' .

4 5
n ,
''1~3 :, ii"="2' .,. ,;
~

~Y

1 1 2,

( n!) 0

((-l)n-1)
.

...

n!
(-l)n-1,

-1, 1 -1
. . -2 :
.

.The rtotation (an)p

.is use'<! tO denote a finite sequence

admitting also last term:

(Iii)~

2,

/5, lb. 17.

2/2. 3

In this Section we discuss briefly infinite sequences onfy.

~niq~elj

determined when the first ~nd general

n-1
define the sequence
~
3 : 4, an :. 2

A sequence is
term are

given~

Thus

{2n-1)3.

:whil~

some numbers

..
~ritten

4, B,

in

...

succes~ion

2 n-l
,
,<

...

followed by three dots,

such as

5, 7, 9,

''

do not define uniquely a sequence, since the general term is not

given, and as the 4th term any number can be assigned arbitrar.ily
other than 11 (that one would expect). Indeed, the sequence
with general

gives

(an)l

term~

5' 7. 9

as the first three terms and 17 as the 4th term.

3 .--.:.
. I

'

Determination of Sequences By Recurrence Relations:

A .sequence
. recurrence

(a

-n P can be defined 'ore generally by a

relatio~

and

consecut'ive t~rms. ap .. , ap+k-l"

The following

a~e

two examples for

Example 1. Given the

.a1

. a) obtain

~he

s~quence

defined by

=3 ~

first four. terms,

b) find the general term;

Solution.

n = 2, 3, ... up

b) Writing the -relation for

addi~d

these member to memb~r~ ihe interme.

2' ., .. an-1 are canceled, and .an

. ~:::: a 1 +.2

. 'a.g_

=~

an ;::: a 1

=3

+2

(n'-1 )2

4- 2 n - 2 _ 2 n+l.

is ob

1; an4

':Prms
..

definitiob of a

Anothe~

sequ~nce

first two tefms and i relation between

es differ by

is obtained by giving the

an

whose indic-

and

2:

Example 2. Given ihe sequence defined .by

a) obtain the first four terms,

b) find the

term.

gener~l

Solution.

b) Since indices dif_fer by 2, one evaluates

a 2 n+l

seperatel~.

Replacing

by

2n

and
in the given

relation, one gets

a2n
~hich,

.when

= 2n2n +-- 11 a2n-2

writt~n

n:l,"2,

for

up to

n, gives

2n-l
a2n :: ~- a2ri-2

which in turn, when multiplied member to member yield

~2n-1~ a2
2n+1

=2n+l3

6
=2ii""+T.

a2

.I
)

~6w,

by

si~ilar

replacing

by

2ntl 1 one .has

process.

Algebra of Sequences:
., '

'

Given two sequences

(an)l, (bn)l

domain-~~ we define new .sequences. on

: ca , ca ,
2
1

l. (can )i

2. (an.fbn)t

.3. ("a - b ~)
n n t

4. (an.bn)1

5.

(F:-)~

Thus, if

. n

.::. . nl

...

having the common

.1 namely

CER

.... ; .;!II

' can'

ar+bl' a 2tb 2 ;

....

' an+bn'

al"-bl, a2-b2,

...

' a2-bn'

albl

a2b2,

al

a2

b.J

'

1>2 '

...

. ...

' anbn '

a
n

~- '

. ..

(bnfO for all n>~)

then

1,1-/2,

.. .

1,1, ... ,1,

2
n

rn.

su-bsequence-s:
If every term-of an (infinite) sequence
term of a sequence

(ari) i-, then, ( bn) 1

-. -

~bn)l

is also a

is said to be a subsequence

Clearly,
other

-eve~y

subseq~ences

sequence is a subsequence of. itse1i. Among

of

(an)l

A notation for arbitr&ry

where,- (nk)

we mention the following:

of

~ubsequence

--

is a sequence .of integers.

5ome subsequences of

((-l)n) 2

1 ' 1 ' 1 ' :.

- 1'

are

' '1 ' .

- '1 ' . - ,_'

l,_;J,,.l,
1

~1.

'

:J,

- 1'

. '

.... ,

1 ' -~~'--~n,........:.....1' . ' - 1'

B. ,BEHAVIOR OF .A SEQUENCE

\'
I. Monotonocity:
A sequence

(an ) 1 .is called monotone if

or else

In the .fq:rmer (latter), case the sequence is said to be

monotone-increasing ~decreasing).
(an 1
1

To show that
either

is monotone increasing .. one proves

.\
a~_,;i
an

~1

or

an,..l

0
n ;:...

for all

n~ 1

In the second case, the in e qua 1 ity, i s to be reversed.

Example. Show that the sequence

(n-_1.)
n

2.

.is monotone inc-

reasing.

Solution.

- a

n-1
n
1
n+ 1 - -n- :: n(ntl).

~;

~ 0

for all

2. Boundedness;
A seg,uence
number

(an)l

is bounded (frfm) abov-e,

i'f

there is a

such that

~u

for all

n?l

and bounded (from) below, if there is .a number R. such tliat

A sequence which fs bounded ~bove and bounded below is said

to be bounded~ otherwise unbounded. Jn. a boundetl sequence all the

..

terms 1 i e i.n a closed i nterva.l

[f, u],

where

is an upper bound,

and

. . ..
-~

is a lower bound. 'Any number

bound, and any number smaller than

larger then
1 is

als~

is an upper

a lowir bound of

the sequenc!!.
If
number

(a 0 ) 1
s~ch

is bounded there exists, clearly,

ptsitive

that

-K

an

Pnl~K

K or

for.all

n~l.

Examples.
1.

1,2,

2.

3.

2 , 2,

, n, ... monotone and_ bounded be)ow,

, 1 /2 '

.- .. ,

...

4. -1 , 1 -

...

, 1 /n,
2,

...

( -1) .

monotone and bounded,

monotone and bounded ,

. ..

non monotone, but bounded.

Example. -Show bound~dness of

b) ( n+8)
~4.

Solution ..

a>

PJ.flt81

Is in

nf

since max(sin n)= 1

b>

and

n+8
1
8
L3i~\ = nrn = 7il'" + il7l'i "

sihce

lFn.,. 8 - ~ 7'ifi1f ~ 71+8' ;. '9'

min n :

-1
(K :::.-g-),

min n:::. 1.
1 .

71r

8 1
472r -=

l ( K : 3!2 >

3 ;. Convergenc-e:

A sequence
term

'(an~

is said to be-convergent if the general

has a 1 imit as

an
If

1 im
n +
and one writes

an = a,

Q)'

otherwise i.t is divergent.

one says that

or

Clearly if (an)_1
as

(an)

Q)

an

a, then

converges tb a.'

(a n +p
) 1 + a (for '

nf-p -+

Q)

n +a>).
Since

a n -a

0,

+
.

it follows that ,in a.- convergent .sequence : (an)l; for sufficien"tly

large N, all the terms
(a~)

1)

fall

insi~e

words, the sequence

aN+l'

a~tZ'

a neighborhood
(a )

_n 1

exists a positive integer

an_ (a-e

.. (or almost every term of

(a-e,

a+el

of

~.

In other

converges to a, if given e>O

there

N such that

a+E)

or

lan -al ~

for a.ll

n>N.

We note that omiss1on of finite number of terms from a

sequence does not alter con\l_!!rgence or divergence of the sequence,
and as far as convergence is cbncerned the
may be dropped.
The following staiements hold true:

jnd~x

in

(a~)p

10

1. If a s~quenceis convergent, then every subsequence pf

ii is conver~ent,

l. If a subsequence is diver~ent, th~ originil sequence is.

divergent,
3. If two sub.sequences converge to distinct 1 imits, the
I

o r i g i n a 1 s e qu en c e fs d i v erg en t . \

Example.
l. 3~

' rn' . . .

/iO,

d i verges t 0

2. ((1 t .-k)nh . converges to

3. 1, -1, 1, -1, ..

00

e,

, (-1) n-1 ., ... dive.rges since it has

the subsequences (1) and (.,.1) having distinct, limits 1.

and -1.

(b )
n

b,

a.n d

c e:IR ," the n

~

a+b

(c .an)

c a

b) (a n + bn)

c) (a~bn)

ab

an
a
d) (F:") ~ 1) . (if
n

~)

d) ( lan\

bn,tO, b,tO)

Ia\

Proof. We prove

c) only. Those of the others are similar.

The. proof runs in the same way as. that for functions with continuous
variable.
Let
such that

an~ a,

bn ~b. Then given

e:>O

there exists

~0

ll

Ib n

- . b \ < e:

for a 11 n >N.
.

anbn - ab

I a n bn -

Since for all

n>N,

abj =

bn

I a n bn

:.

I (an

I an

Ibn

and get

- ab.n - ab + abn
- a)b n

- a

- a(b

I I bn.l

I e: + I a I

K,

- bn)

+ fajJb - bnl

e:

lies in the interval

it follows that bn < K for- some positive

(b-,

and one has.

Theorem 2.
a) A monotone bounded sequence is convergent,
'
b) A convergent
sequence is bounded,

c) (an)
.::;}

a,

( c ri)

(bn)
+

Proof. Omitted.

and

and
a~ c~ b .

an~cn~bl'!

for all . n>N

b-e:)

12

C. So~~ I~portant Convergent Sequences

:a,

_ (for

2. ( ( 1 ... ~)n)
n

e.A

p .
3. a) (!..!!___..!!.)
n

riP

4~.

--

converges to l

a>.O)
1,

b) (--n)
e

.
;a, 3;a, ... , n.;a-,

(n+ .A) n
n

or

nrn, . . converges to 1

e.A

p.

for any" .constant

for any constant

r:;
/2, -3 .,..3,

'P

: a 1 ~ -a 2 , . , an, . is a sequence with

positive terms, conv_er;9Lllg to the 1 imit - a, then
If

(an)1

..

al +

a) An

:.

!>) Gn

an

n
n

...

la 1

an

where---" An' Gn are cal~etl the arithm~tit ~ean and geometric mean
of the pos i tive nuinbers al' .a 2 , . , an.

5.

b)

en

Cn;)

The proofs of ! ~ ~ are obtained by 1 imit proces_s considering the fUJJCtions

al/x,

x1fx; (1

~ ~)x,. (lnPx)/x, xP/ex .of

continuous vari ab1 e _ x.

Proof(4).Since
such that

(an) +a, then given

> 0

there is ."Nt > 0

'I 3
\

l ct"n

- al<

for all

e:

A.
n

a) considering the-difference

a 1 + +an

- a

J' An

(a 1 -a)-

<

- a,

lal

. For sufficiently

(a~-a(
.
t

al

...

-llaN

n
\aN+l-aJ

al
t

n, say for

--

l~rge

-(a n -a)

(atlf-1-a)-

-n_.

+ {an-a)

(a 1 ..:.a)+

-(atl-a)

+.

we have-

--~~~~--~~~

n
la 1 -af

aj~

n>N.

(n

(an-al

N)e:

NI >N _the
. . first term

on the right hand side is -less than E:, and

b) Gn

-::

nl"'a

an

by (a} ~

Gn

R.n Gn

a.

+.. R..n .a n

R.n al t
n

12
.!

Proof(5)
a} Applying 4b to . (an}

(ntl)n
n

e, we ha i'e

.+.

X.n a

14

1
."2"

(n+ 1) n' ::. __!!_:!!_ +

n
".{IT!

.n:T

n . nl
n
- - - - - n-IT+
n /ri! - nfll'n !

t' .-

e. 1 :. e

b) We.use the lemma:

Lemma.

+O~an+o.-

Indee-d, since

-n

an+ 0, given

> 0

for all n>N. Taking

there is

N ::0 such

<1,

"ii,fri!

E X.E RC I S E S (1. I)
1.

W~ich

a)

( 1)

ones of the

followin~

b) (-

2. Write two subsequences of

are subsequences of
c) ( 0)

n
(1-(-1} )1 ?
d) ( 2)

15

a) (cos n1r.) 0
3. For each of the f_oll owing cases find
and

(d

n
I

b) {dn - d n

c) (dndln)

(dn)

such that

a) ( dn +d n

dive~gent.sequences

d) ~dn/dln>

is convergent.
4. Examine the following sequences for monoton.oc'i ty1 bound~dness

and convergence:
n - 1

a)(-.-n-)2

3 n -1-1

b)(

3n

1 >o

5. Same ijuesti6n for:

a) (fn n)

b) {1/.fn n) 2

6. Same question for:

a) (Fn

i> 1

b) (an) 0

when

a.n

0 . if

is odd

if

is even

]1

7. Same question for:

. n
b) ([Jll>o

a) ([ nn-1 ] ) 5

-~

8. same question f"or:

a) ( l/ 2-vn ) 1

b) ((-l)nn)]

c) ( t(-1 )n-1}) -3

d) (O,g2 ... 2)l

n

9. Find the general term of the sequence defined by

a) al ::. 1 '

an .:. 3an-i

c) cl ::. 2 '

c2 ::. 1 ' en

=-

cri-2

b) bl

.::.

2' . bn

.:.

3 +- bn-1

d) dl

::.

1'

d2

.:.

2'

d n ::. -dn-2

16

10. Given the

~equence

a1

prove that

1,

a) -is monotone increasingI

~)

b) is bounded

has a limit, and find this limit.

ANSWERS TO EVEN NUMBERED

, 2. _a) (1);(-1),
4. a)_ M, B, C,

~)

( 0) , ( 1 )

b) M, B,

p. ;

6. a) M, ~.

(.

.b)

8. a) M, B,

c'

,b)~~ }. ;.

10. c:)

..

: (/+tl.f)

'yi,

EXERCIS~S

. c) M, B, C,

d) M, B, C.

(a )
. n

17

I 2. SERIES OF-NUMBERS
I

A,- DEEINIIIONS

A sum

al + a2

. ..

a n

. n=lI
co

......

( -1 )

an

of infi~itely many numbers in the given order is cal1ed an infinite

series or simply~ series, where
series.

an~

is the general term of the

In the sum (1) the numbers are to be added in succession,

that is' a2 is to be- added to . a 1 ' next
a3 is to beadded to
aL+ a2, then a4 is to be added-to al + a2 +a3, and so on ..
This definition of ( 1 ) ...lS equivalent to

s 11

lim
11

-+ co

'
j

where
+ a

is called the general -partial sum of the series.

If this 1-i,mi_t exists which is equal to Za , we call (1)
'

convergent, otherwise divergent.

For instance the

seri~s

n=l
i~

(-l)n+l =1- 1.+ 1 - 1 + ... + (-l)n+l +

divergent, since

_sn

1 +

. -

+ (-l)n+l =

has no. limit.

Writin_g
co

n=l

an,;. ~n "'Rn+l

{~

if

is even

if

is odd

18

where
Rn + 1 = an .... 1 + an+ 2 +
is the remainder after the genera.l term, we ha.ve

co

Rn+l

an) - sn = (lim s 0 ) - .rn =

s-s7t.

showing that

for a convergent series.

Arithmetic, geometric, harmonic series:
These series are related to arithmetic.,' geometric and harmoni~

means of numbers. If

a,. b>O, then

th~

numbers

A, G, H

defined by
a+b
A .. -z
are

res~~ctively

means of

,.

~
vau,

called the arithmetic, geometric and harmonic

a, b. Observe that

metic mean of the reciprocals- of

is the recipr6cal of the aritha

and b.

A series whose,.all terms are positive is called an arithmetic, geometric or a harmonic series according
of the series, except the first

on~

~s

every term

an

is the arithmetic, geometric

or harmonic niean of its adJacent terms

an-l

and

an+t'The first

term is an arbitrary positive number.

Arith~etic

series: a + (a+d)+(a+2d)+
:

Geometric series: a +art ar

.
.
1
Harmon1c
ser1es:

at

a+Q + ~

In an arithmetic series
ference, in a geometric series

+(a+nd)+

+ .... + arn +
t

an+l - an d

+ a+nd
1
+
is t.he common dif-

a n+l/a n=ris the c6mfuo~ ratio.

19

-These three series defined for

zations

o negative

a, d

a, d, r>O. have obvious generali-

and

r.

An arithmetic series is convergent only when

an~

divergent in other

a.:O

and

d.:O

ca~es.

A geometric series (for aiO} is obviously. divergent fo.r r=L

In other cases (for at&O)", having.
n1
rn +1 - 1
.+r)=a
r-1
there. is conve.rgence when lrl<
~

when

or when

.-l<r<l, and divergence

irl~l. Hen~e
CIO

arn

J~

if lrl < 1

n=O.
lCIO

The harmonic series

1

CIO

I n

n=l
corresppnds.
.series. To

to. a:l , and

~~oye

.r r

= 1- + '2'

+ '! + ..

n.

d:l. We show that it is a divergent

divergence we assume its convergence and produce

a co~tradiction by esiablishing a relatiun between

where

hn

is

t~e

general pariial

h2n

and

h.n

sum~

+(

1 + 1)
zn-,1-...
.. rn

'(1 + '!
1) + (1~. ~-~
1}
>.
.

'! + 1 + ! +

. \

+ .!.-_1
n '! \ hn

~ h2n.- hn >.} con.tradic_ting h2n-h\t-+h-h .. 0 .

20

Theorem.
a) In a convergent

~eries,

th~_general

te~m

tends to zero,

b) If, in a series, the genera 1 term 'does not tend to zero,

the seiies is divergent.
We Call-Part

otest f.or divergence by general-term lest

Proof.
a) Let

I:a

s.

.an

k~O,

' n

b) Let

Then

suppose the series is convergent.

b~t

'"'
Then by

a),

a0

contradicting

k/0.

Remark. Jl.s ~-t~e harmonic .series shows, the approach of

an

to zero.do~s no~ imply the c~nvergence Of the s~ri~s~ that i~,

the series in which
~sts

for

0 may

c~nvergence

ctin~erge

dr

and divergence

diverge~
ar~ g~nerally

given.

for series of positie terms.

_1

B, _SERf ES OF POSITIVE TERMS

A se.ries hav,ing almost every term (all terms except perhaps

finitely many ones) positive is called a_series of po-sitive terms.

Thus the series
f

.
-2n-3 __ .
n2

--are series of positive :terms-, while

3
~

(-1 /2)n

0 + .5 +

"'

~2n-3 ~-

n2

is~ not.

.T"Jle tests that are g_iven here are comparison tests by which
a given series of -positive ter'ms is, <:ompared eJ.ther with a positive
.

improper i~tegr~l or wiih a~other seriJs of poiitive terms. Ther~

are also intrinsic tests
series itself;

~hich

are applied directlx to a given

21

:1. Integral test (of McLAURIN):

~

'

'

co

Theo~em.

an: f{n)

Let

an be a series of p9sitive term with

p
Of; [p, oo). If f{x)
is positive, continuous

and

and dec rea s i n g on

l}l , "') 1

N::> p , then

(X)

~ an
conve~gent

are both

or both

~ f(x) .:dx

and

diverge~t.

s.
n

Proof. Consider the partial sum

lp.,

terms a p , . ~ .' , ' a N'-l . of

the -hypothesis r

N-U

on which f(x)

may not satisfy

we set '

sn

= SN-1 +~

' '

lim sn

Separating from it the

exists i f

lim s '
n
otherwise divergent.

( SN-1

is finite).

.s4' ''

exists. Then th.e series is convergent,

Consider-}he graph of. f{x).

f{X') being decreasing,one has the
d6uble inequality
.
.
.n .
- f( n') < J f ( x) d x<:f ( n- i
n-1

.which,. wheri written for a 11 subi nterva 1 s on [N, n] and added:...gi ves
n

s'- s

If the integral'has

N < NI

f(x)dx<s'l --sN- 1
n-

a limit

when

n.-+ oo, this limit is an

upper bound for the-.; ncreas ing sequence . (s~ .::-sN). Then
and hence

~an

i_s ~onvergen~ .

If the_ in.tegral diverges to

cio.

the same is true for

is diverg_ent.

(S

I)
n

22
p-series:
The.series

...

1 +-/ + ]
l + 1 +
"'
~
2p
3p
nP
n ""
\

n=1
..
is called a p-ser.ies. Clearly the harmonic series ' s a p-series
for

p:l .

.I

Theorem. The p-series is convergent when

when

p>l, divergent

p~l.

~oof.

For

p:l

p-serie~

the

is harmonic and

the~efore

divergent.
The function
decreasing on

is positive, contin~ous and-

f{x) = 1/xP

[1,

=); Then
CX>
CX>

which is
when

..

dx
x.p

if

conver~ent

CX>

1\

X-p+l

X1-p

1
x-pdx =
---::ptT =r:p

or when

1 -. p < 0.

X:l

p> 1. and divergent

p<L
Example 1. Test the series.

CX>

1+1+'5"
+
"T'7 = 1+ "2"

n +J

1 "+
--z-:-:n

+1

for convergence.
Solution. The

seri~s

is of

po~itive

terms. The function

= l/{x 2 ~1) being positive, continuous.

--[1, -) the integral test is applicable:
f{x)

r1 CX> ~
X
.

+1

arctan

X ICX>l

Hence the given series is convergent.

=I

and decreasing on

(convergent).

23

Example

Test the seris

2~

Th~

-Solution.

[2.

i~ ~f

for convergence.

positive terms. The function

fulfills the conditions of the inte~ral test

f(x} i 1/(x 1nPx}

on

series

2 n R.nPn,

co):
00

J
X

converging when

co
1 .dx
d 1n X
1
J
=
= 1-p (Rn :r;jP
Rnp X
1nPx
2
/ n

2.

Comparis~n

-2

p<l. Hence the gi.ven series

p> 1 diverging when

is con.vergent only when

I"

P>l.

with othr series

Below we give two tests by comparison of a given series of

positive terms with other such series.
Theorem 1. (Test by inequality}
Let

Ian

ld.n - be

be a given series of positive terms,' and let

two such series whioh are convergent arid divergent

-lan.- is_ convergent

respectively. Then

fo.r all

n>N

.for some

N.
n

~ro_o1f.

a. limit as

an~ ~n
-+ co

.Qr divergent according as

=>

an~=r en. By hypothesis

.

the sequence

n
(I an}
.

len

having

is bounded above by

this 1 i mit. Since-it is increasing, has a limit, and

ran/is

convergent._
Divergence case can be proved
A

generaliz~tion'o~

the above Theorem. 1:

The Theorem holds true

where

and

sirnil~rJy.

~hen

the inequaJities are replaced by

are positiVe numbers:

2~

In tests for

convergenc~
.

a series

is posibility for an ea-sy selection of

co

'

compared often either

or with p-ser~es

with a geometric serie; Irn

2n+3

Example 1. _Thst I

i~

I 1/~P. ~ince

there

r, p for comp_a:ison.

for convergence.

2n+3 . 2 3
an = -z-:. n + 7 ,_t;he series of posi, n
n
tive terms is com~i~able wi~h har~onic series:
Soluti-on. Writ,ing

an >~ ~Ian> 2I

r __n__

Example 2. Tes't

Solutio~.

[an

::::::;;>

divergen-ce~

for convergence.

inn'+ 2n

Having

is convergent since the geometr.ic series

(.J/4)n

is

convergent.
and its generalization may be~r~phrased

This Theorem
as follows:
-corollary.
If for series

~a

/. n'
a

Ibn of positive-terms,

a.<......!!.~
,
n " b
b

h6lds for positive

num~ers

(or

a.b n~a~
n b.bn)

a,.b .for all

n>N~

then

Ian

are both convergent or both divergent.

Theorem 2. CTest by limit .ratio).
1et

Ian :Ibn

be two serie~ ~f positive terms with

. an lim- = >.(>0)
bn
a) If

(>.~0.

}.:0,

or >.

AiO, the series are pf the same na ttir.e J

co)

Ibn'

. 25
b) If

A:O, the convergence of Ibn

implies
that of
.

(or divergence of . Ian

c) If 1

= oo~

implies that of- Ian

I~n

the. convergence of

Ib)
n

implies that of

Ibn

(or divergence-of
P~oof.

a) If. 1:;0, giVen

e:>O

a
1-e:<.:....!!. < He:

and less then 1, there -is

for all

ft>O such that

n > N.

bTl.

Then by above c,o.ro 11 a ry the two series' have the same nature.
b) Let

1:0. Then for

n >"N

an
bn < e:

or

for some

NJ

.an< e:bn

holds. By Theorem 1 the assert~on is.}rue~

c) Let 1 =

oo

Then t.h ere i s

a :>;O

such that

an > ab n.
f o r a 11

l'l ~:>N. Ag a i n-~ Co r o 11 a r y t h e a s s e r t i o 11 i s t r u e .

Example. Test-the convergence by lim1t rafi.

b)

st:

t.' tan n

.l

Sol utfon.
a)" .Comparing

::. ~im

a
n

= 1 im

bt,
b) Comparing

n3

= 0 :=:::;. conv. of

en

1
.an =tan n

with

an

t an.l
- n

b'lt

1 = l.im

we have

.with

.. = lim

bn = 1 /ri

=l

n e

-n

we have

div. of

. tan 1
n

26.

J. Intrinsic tests
By an intrinsic test we mean one
of the

gi~en

related~only

serie of positive terms without

to the terms

refer~nce

to other

series. The follo0ing two theorems express such tests:

Theorem 1. (Roo~ and ratio tests of CAUCHY)
A series
a number

Ian

of positive terms is convergent if there is.

such that

a) n .ran~ k < 1

or

b)

an-tl
k< 1
. n .
.-a-~

and divergent i f
~1
a' ) nra
. n

for all

n '.N

for some

or

I )

an+l
an

N.

Proof.
n

a n"' kn' the series is comparable

~n.,; k < 1
with convergent geometric series Ikn. Hence Ian is convergent.
a) Since

. b)

a'~

bI

The following tests are more useful in practice, than the

above test because of the. deter~ination of k.

27

Theorem 2. {Lim roo.t, .lim ratio testsof-CAUCHY)

of. positive series is convergent if
n
.
an+l
< 1_,
or
a) 1 im' ran< 1
b).lim
an

A series

and divergent
aI

) .

Ian

if
~ran> 1

1 im

or

an+l >
1.
an

b-') lini

Thst fails-if limits are equal .to 1.

Proof.
a) let

If

lim nran = r.
.,
k
1 there 1S

1 i mit . nran ~ k

test,

Ian

o..

holds f. or all

r<k<l. Si_nce

r is the

n >" k for some . N. Then by root

is conv.

a I) If
a t~
n

such that

nra: > 1 ho fds for all

n

r > 1 , . then

n.> N and

{div.)
The proofs of

b~

bI

are similar.

Remark. If one of the lim root, lim ratio test fails, the
other fails too .
In the failure case, one way apply. the fo

tnq:

RAABE-DUHAMtl 1 s .Test:
A series Ian

of positive terms is

conver~-

or

divergent

according as
an
an+ 1

1 i m n{ - - - 1)

is greater or le'ss than -1. Test fails if 1 imit is equal to 1.

Example. Test the following series of
convergence:

positfv~

terms for

28

.a} r(n+l>n
n

. b}

2n.

T n;

c)

Solution.
.a} Since
b}

an 4- e
en+l
(n+l}!.
' .

an+l
an

, o,
n!

n2

C.
. n vn;
.1 .
= - - 4- -< 1
n
e
d)' a

4-

(it

n!

I n
2. n

'd}

. n

ril"'+2"

- 1

the series is diverges.

=

n:tT

4-

0 <1

(it converges}

('it converges}

diverges}

C. ALTERNATING SERIES
A series
~

(-1}

a n =a o -a 1 +a 2-- ...- + (-1)

.

a n ......

~.

(an>O)

in which th~ter~s aiternate in sign is called an alterna~ing

/

series.
The series--

n +

is an alternating

ori~.

known as the alternating harmonic -series.

Sin~e an. alternating s~ries is not a series' of p-ositive

terms-, the p1rev1-ous tests. cannot b_e appl i;d. However there is a . . .
test special to alternating series which is the following:
Theorem (LEIBNIZ}
-The alternating series
a

+ (-l}n an +

0'

-is convergent if
a }. a 0 ~ a 1 ~ a 2
. b} an

4-

0.

.. ~ ail

~ ..

'

Proof.

.29

1-h

will suffice to prove that

s 2n

and

s 2 n+l-

have

the same 1 imH.

{from l}

s 2.n

ao~(al-a2)- : -(a2n-2-a2n-l)~a2n~ ao

5 2n;

>10 .~ s 2n ~
Hence (s;nf

(from 1 )

ao

is bounded, and being monotone increasing i't converges

. to :a limits. Now,

Corollary.

I1\. a convergent alternating series-s = a 0 -a 1 +a 2 - ... + (-1) an+Rn+l-

witn g1ven -hypothesi~, the inequality

\

holds, that is the error made in taking

sn

for

an+ 1
Proof.
R

n+ 1

n\l

( -1} (an+ 1 -- an+ 2 + -)

~IRn+ll=lan+l - an+2 + .. I

Example. Given the alternaiing harmonic s~ries

a) show its convergence

is less than

30

b) to have max error

be

~f

10 -2 in sum s, how many terms should

taken~

Solution.
' 1. . . . 1
1 ) 21 . 'II'
3 ;:. .. ~ n >. \ an d-

a) Since

th. e r.e 1. s

convergence.
b)

<
ii+Tl()l

~ n-t 1

> 100 9

{l09terms)

n > 99

D, SERIES OF ARB1TRARY TERMS

'..

If a series is one uf positive terms, one

appli~s

a test

given for such series,. if .ttj.e series is alternating, one applies

LEIBNIZ'test {tes~ for alternatin~ ser1es). For a~ arbitfary~series
the following theorem holds:
Theorem~

A series
~I an = a 1 + .a 2 r +

+ a ri + ...

. 1

is convergent if the series,

of absolute values is c_onvergenL

Proof. Let

sn' S0 \be the corresponding partial sums. The sum

sn ~ontain~ non negative and n~gative terms and we write

sn
.where

I
II

that

sn "'Pn+ Qn

P ,-Q are sums ~f positive and negative terms respectively .

n

.n

(Pn)' {Qn)
by

= pn...;dn;

a~e monotone increa~ing sequences bounded above

S = lim sn. Then

s n -+ p ...
-Q

P0 _,. P,

Qn-+ Q. and

0~ P +

Q; It follows

.I

31

1-

.>

I ..

A series

Ian . SUC'h thatr, rlanl

is conver:-gent is called an,

absolutely conver-gent series; and the above theorem states that an

absolutely convergent series is convergent.
As

alternating harmonic series show~ a series may be

convergent witho~tPeing"' absoluiely convergent. Such series are
~he

.t

ca 11 ed simp 1y co,nvergent. series:

(conv)
(conv)
or
:::9>
. I:a. (div)
n

Ilanl (conv._)

rjan'~ (div.)

abs ." conv. of >I n

simply conv. of I: an

Ian

{'""

There is an essential .difference between the absolutely

convergent series and

~imply

conVergent ones. The

ab~olutel8

con-

vergent series. have 'the following ;hlo properties among others:

"--..

1 . The terms can be rearranged in any order

does

~ot

(re-arr~ngement

alter the sum)

2. Finitely or infinitely ~any terms may_be replaced by

their sum.
These

pro~erties

may not be shaed by stmplyconvergent

series, th!lt is, a !rearrangement of terms. in a simply convergent

s~ries

'

m_ay give a different sum as illustrated by the following

ex~mple:

Consider the

~imply

conVergent

1.

~lternating ~armonic

~1

S:.l-'2"+!-4+ ... + ( -1)

series

n+l 1

ii+ ...

Let us rearrange the te)!IIIIS to have the series

1
1
1
S l = ( 1 - I1 - 4)+('!
- -o
I

...-

,_

1
1
ir)+ (~ -

1
+ (ffiT

1
nr1 - .nL
1

.1

~ ."lfiHZ - 4'ii"t4)

t ...

' .
In many texbooks conditionl convergent or semi-convergent terminologies are used 1nstead of s1mply convergent.

-containe~in

Observe that every term in one series is

the

other exactly once:

S

"'

I<
0

4n+2"

= "2" {1

'2J"il"+4

.'

1: . 1
1
4 +
"2" + 3

'

1
~)

( 1 .
2 oI 2n+T-

...

t {-l)n+l 1n +

... ) ...

1
'2" s.

E, EVALUATION OF SERIES
Each series can be'evaluated by neslecting the remainder
Rn tl

for certain

n with

some~pproxi~ation.

Exactevaluation is impossible in general ,_except for convergent geometric series and sonie .s~ries whose g_eneral term

ari

is rational function of- n. There are other.possibilities by

t)~e

use o( power series . {1; 3).

Example; Evalua~~ the. geom~tr1c series/

a)I.
o

c-}>"-

b)

e-n"

I
0

Solution. Recalling
a{l

+ .. r +

... )

we have

Example. Given
.a) write it ai a

t = 2,1Ji,
gerimetri~

series,

a
1-r

{\rl<l)

33

b) disscuss the conver.gence and fjnd

t,

as a ratio of two

integers.
So-lution.
a) t_ .. 2 +

J...

~
+ ...........,3,....7
I VVV
100000

IV

.-..?<'

21
37
/ 1
= TO + 1000 <1 + 1 oo
-

37
+

+. :

+ lOOO.lOOn-l
+

.. _.

1
.
n-1 + .. )
1 00

bJ The serieswithin the paranth~~is is a geo~etrtc series

r = 1/100

with.

w.hfch is. absolutely less than 1. Then

it fs convergent:
21
= 10

37
TOmJ

- 1

37 .

21

1 .. 10 + "9mY =

ron-

1 -

2116

9'9'U""

e:

Example~ Fi~~.the sums:

1
_a) 'T:7 +
b)

D +

"1
. :r:4 -,l

1' . +
n(n+l)

...

1
+
(2n) 2-l

...

Solution.
a) an

1
n(n+1L

B
.A
+ ii+T~

. 1

8 -1

1
n+1 .

= (.l - ~)+(~ - !) +

.. 1-

b} a

...
n

1
, (2n)2.,.1

. ====P an ..- ~

n+T

.+

1
<n1 -- ii+T)

-s-1
--or

A +

an

rn:T

tiHT

1 -.

<.'2'ii'=T - '2'i+T)

~ A _.. '1
~

8 -

34 .

.
1

:: "2"

( 1 - '2ii+T)

s = 1 /2

====!;>

EXERCISES (L 2)
11. Are the following

d~finea

series?

~ive

reason
co

L n!

_c)

and

12. In the following series forming

show

convergence or divergence
1
. 1 - 1. + '""1 -. '""
c.
c.

+ .. +

1
n. - -n +
'

13_ . Show the fo 11 owing for divergence by the use of genera 1 term test:
I)) }:si11 n_

c)

L (1

~) n

'+

14. Thst the followin~ forconvergence;by i~tegral test:

a)

L
1

c)

n -

b)

;z-;1

rn

.11.1'1

rn
2

1
n .tn ( R.ll

~)

d)

I0

1
t.n2 n
l

7n2:1
n +1

15 . rest for convergence by comparison.

a)

2" +
r 3n
+ n

b)

d)

c)

l ...

2~

r n2 2n+

-- 1

2n

16. Same question for:

a)

3n
r~
n+
5

b)

1.

.41 sn 2 ,.. 3ln +

n!

35

]7.

~st

an~

the series given in Exercises 5

CAUCHY's root orr.

t19

6 by the use df

test.

18. Same question for the 1ollowing series:

a) I--r~T"

b)

10

c) j

(1

~) n

d)

n .

2 .

n!
2n,-1

. ( n )n

.n+T

o_

19 .. Test. the following a 1 tern a tl n g s e r i e s f o r co n v e r.g e nc ~ :

a)

{:-1 ) n

c)

b)

.tn n

( -1 ) n sin 1
n

( -1 ) n

I. :--n:;,-

> I c.;.1 >"

2'0. Same question for:

a)

I~
n.

b)

I (-1 ) n
l

cos n

21. Test for-convergence:

a)

n
R.n n:T

, n2 R.n ( 1

b) [.

1 F .
n) .
2 .

22. Test for. convergence:

a)

I
"1

23. Test
a)

I
0

n!
nn
fo~

b)

I ( n2-t-n1 ) n .
l

convergence:

nn
( 2n)!

b)

(n+1~{~t2~ ... 1 n+n)

2 n

2 4 Di s c u s s a b s o 1 u t e 1y a n d ,s imp 1y con ve r g en c e :
1 - -.1- )
a..) \ (t.

.n

n+a.n '

(0. <a.n< 1)
..

1 . b') ?T'

n1

25. Same question for:a_)

(-l}n-1
tr(n+l)

b)

( -n )n .
" 10

I.1

2 n- 1 )

...

1 - Tri'iT
1 t.
t 7il=T

36

26. Discuss the .convergence, and

a)

the sum when convergent:

sn - 3n
6n

l:
0

b)

fi~d

(-l)n+l

(use the identity:

n
1. 1
-. 2 + 3

and express the

ri~ht

n+n

hand side as a RIEMANN sum)

27. Find the sums:

a)

l:

b)

L-

2n 2 +6n+3

n(n+l)(n+2)(n+3)

(Hint~

28. Find
a)

and evaluate lim sn:

2- ..
16n
sin 4(4n-tl).
2
2 Sln
(4n 2 -1)2.
( 4n -1 )
sn

2 9 .. Let [A 0 B0]
by

Use decomposition irito partial fractions)

be the i 11 t e r va 1

b)

l:

[0 , 1] . Let

;n;l

2;ri +
2 n+l

[A 0 Bol

be b i sect e d

s 1 ,(A 0 s 1)by

A1 ,: ... "(An-lBn1by An and[Ansn)by.

Bn+l" Sh6w that lt~iting position of ~An (or Bn) trisects

[ 0' 1] .

30 .. A ping-pong ball, when dropped from a. height rebounds a distance of three foOrth of that high~ .. Find the total distance
travelled by the ball if the initial

ANSWERS
12. s2n

..

0,

14. a) Di v. ,

~0

~eight

16. a) Conv., b) div.,

h em.

EVEN NUMBERED EXERCISES

'.

- 1
s 2 n+f:: 2ri+T .... 0, conv.

b) conv.,

is

c) d i v .-

d) div.

37

J 8. a) Conv.,

b) di v .

"20. a) Conv. ,

b) div.

22. a) Conv.,

b) Con\.1.,

d) conv.

c) d i v ..

24. a) S_imp1y conv.,

b) abs. conv.

26. a) Conv.,

S:4,

28. a)

Sn =

- z1

b)

Sn =

ln+1
2n+l

7k

em

30~

b) conv.,

s =

1 cos 4
cos 4 + z
"2n+T

1 + 1 (1 +
"2"
4

Rn

s
1

.. z1 ( 1 -

.. + :n:T) -

..

s "' 0

cos 4)

38

POWER SE~I~~

I. 3.

A. DEFINITIONS
An infinite series
i
.

.
n
+ an(x-x 0 ) +.:..

fn powers of

x-x 0

where

is a variable .and

x0

an

(1)

are cons- ,

tants, is called a power series.

If

xo =- 0' the power series is -simply

..
L an xn .:;. ao f.al X .+ . . + ~lx n+
00

(1I

li

Some power series are ~convergent for a 11 . va 1ues of the va riI.

able x,

~orne

others for no value of

except for

x~x

0 ; sttll

others are convergent for all ialues is an interval of non zero

.

, t

length and divergent outside this interval. Such an interval 1 i?

#

called the interval of

conver~ence

of the seiies (1).

There are series convergent (divergent) at.both ends of.the

interval of convergence

or

corive~gent

at one-end and divergent at

A main problem, in connection with power series, is the

determination Qf the interval of convergence. It is

determined~by

CAUCHY's.l~mit .root or ltmit ratio test applied to ~eries Ilan(x-x 0

.

of absolute values, yielding an inequality involving

set of which is the required interval of

co~vergence

x;t~e

solution

I.

in the. obtai ned interval I ' the power series

will be absolutely convergent, and acc'ordingly the i nterva 1 I may
For every

also be ca 11 ed the interval of absolute convergence.

In the interval of (abiolute) corivergence I the half length
R .is referred to as the radius. of convergen.ce(l).
1) The reason for calling it radius of convergence is that the power series
Ian(z;_z 0 J' for complex numbers admits a circle of convergence.

39

Applying to (1) the CAUCHY's limit root or limit ratio

say limit

test, for absolute

~atio

we set

converg~nce,

~est,

un ::a n ( x-x o )

and have
lim

~u~~l~

=lim

n-1
an+ 1 ( x_-xo)
a (x-x )tl
n . o

,.

for convergence, or

where

j:~,
= R-
an-1.

1 im

is the radius of {absolute) con~er~enc~, ana

~

jx

x0 j < R

I = ()( 0 -R, -x 0 +R)

~~-limit root test hasbeen app~i~~. we wouJd get

-1

. lim, n~

;;:. R.

Example. Find the interval of (absolute) convergent of the

power series:
nn

(X -1-)

b)

Solution.
n .

(-~-1 )n

'9 tun! ;

-h

- - jx-lj

nrn:

=*

I x-11

<

-e <x-'~

< e '*

jx-ljn
-+-

-eJx-.11 < 1

1- e <

x < 1 +e

I ::. ( 1-e, _.l+e)

/

40

b) .In this case we evalu,te

. .
1

n -t 1
_,mw=""

R:

I . : ( - co, co).

B, REPRESENTATION OF FUNCTIONS BY POWER SER.l.E.S.

For the present the only functiori

resentable by power series is
series

\ x n
L
0

t~at

is known to be rep-

f(x) = l/(l..:x). indeed the geometric

with common ratioxls known to be equal to 1/(1-x):

which is valid. in

I= (-1, 1).

The queition now. arises as to which functions are representable by~pow~r series. The answer to this question is ii~en by the
following theotem borrowed form Advanced Caiculus:
Theorem.
1. A function
a point

x0

f{x)

admitting derivatives of all orders at

has a unique power series representation

f(x)

(1)

valid in an interval I of abso)ute convergence.

2. The power series

'

r~presenting

the derivative f'(xJ is

...

obtained from (l) by terrn'-by-term differentiation:

n-1
f'(x) "'a 1 -t2a 1 (x-x 0 )+ . ; 1n an(x-x 0 ). t
valid in the spme interval as that of (1)

(2)

3-. The power series representing the integraJ

is obtained from (1) by term-by-term

f(x).dx

Xo

integ~ation.

f. f(t)dt ::. ao ( x-xo )Lr

x
0

valid in the same

int~~val

+ n=T
an ( x-xo
)n-1
.
+
.

as that of (1).

( 3)

:.,'

'

41

.. -

Th~

..

'

coefficients

a0

and

are immediately obtaniable

from(l) and .{2):

Differentiating (2)

ll-1 times, one gets

f(n)(x) ::. n(n-1) . .": 1 a n +powers of

(x-x o )

and

Thus we arrive at the power series:

f(x)

::.

of f(x)

at

interval

f(x 0 )

XO)
(x-x 0 )
1!

I (

...

fCn>cx )
0
( x-xo) n
n.

. ..

xo' ca 11 ed the TAYLO'R Series of f(x) at X ;: xo

which is valid for x(:l an.d non valid i f X is exterior ~ this
of absofute

convergen~e.

TAYLOR series becomes what one calls the McLAURIN seriesif

x0

is taken at the origin:

f(x.) ::. f(O)-+
Ex~mple.

.at

indi~ated
.

a) e ,

Fi~d

_f~ ~O'). x

t ...

f(~~ (O)

x"

the power series of the following functions

points:
.

x:O,

b) sin X, COSX,

X:O,

C)

en

X,

X: 1

l.

Solution.
a} f(~) ::.ex

has derivatives of all orders at

f ( n ) (x )

ex
-~.

Since

ex

::.

.: ex 9

a n = f ( ~ ~ ( 0 J ::. ll n ~

1
1 x2 +
1 +.,-: X t
IT

~.

1
l'l

-\. -::T

x" + ...

is convergent for any .xe:R, we have

..........

x::.O:

(-<D. (I)) -

42

e
e 1

..

TT

=l -

TY

::

~ .1

n;:

2"":-

...

...

+.(-l)n

n;:

. ..

b) f ( x ) : s i n x has de r i vat i v e s of a 11 orders, and a r ~ do s x ,

-sinx,

sinx - periodically,
and
.
.

:-COSX,
'

f ( 0) :. 0 > f I ( 0) : 1 >
~3

fll ( 0) ":. 0 >

~5

.!".I ( 0) :. 1 ;

n x2n~l

sinx = x -r;-t-s: ... t(-1)

.~.(-co,

--zn+Tt

The expansion of cos x .is obtaihed more

~asily

co)

by differen-

tiating the series of sinx:

COSX

..

. 4

. X

~tel(!"

2n

(-1 )n

22-n-

. (co, co)

Note ttiat the odd function sin x (the everi funct.ion cos x)
involves only odd (even) powers of
c) f(x):. 1n x
. f

X)

:.

i:

x~

has deriv~tivds of all orders at

X-

11

~)

:.

( -

1)X-

, f

111

~)

:.

( -

x = 1.

1 ) ( - 2J X -

~,

f(n)(x) :. (-1)(-2), .. (-n-l)x'" 0

~

/f

( n)~

\ 1 ) ::. ( -

Ut

X ::.

1n

X :.

>:
1

.n n -1 ( n-1) ~

n:. 1, 2, .;.

f,(n)(l)
(x-l)n;:.}: (-l)n-1 (x-l)n
n!
n
. 1

x-1
-1-

(x-1)

2
+

(x-1)3
3

(x-l)n
. + (-l)n-1
n

... ( -1 ,1)

Part 3 of the .theore]l is usefull in obtaining expansions,of

functions

whos~

derivatives are (simple)

1n(l+x), arctan x, Argth x.

ratio~al

functions such as

43
ExamEle. Find MclAURIN -series 'for
b) arctan

a) !n(l+x).

c) Argth

Solution. Knowing the sum of' geometric series.

-2

nn
1
d

a)l-x+x - .. +(~1) x + ..... =,l+x=-rx1n(l+x),

..
we have by term by term integration ,.
x3
.
, n x-n-+1
x2
!n(l+x) = x ,- 2 + T - + ( -1) n+T - j( -1; l) ~
I~

/_

a imilar way

d a-rctan
b) Ox
arctan
C)

d X Ar g t h

- 3x3

X :. -'X

X :;

- :-2 =
l.+x

Argth x = x

-t

x5
+T
i

_X a. =:

x3
T

-t

Ts

- ...
.f

+ ( -1) n ~2n

+ x4

( -1 )"

+-

...

-+

x2ntl
TntT +

. ( -1 1 )

x2nfl

(-1, 1)

+ . -+ 7il+T t

The serj!s.of !n(J+xi being convergent at the end point .1

of

I = (-1, 1 ), .we .have

1

1' .

!n 2 = 1 - .! + J -

(.:.l)n+l

(-l)n

{why convergentj

The same being true for arctan x we get

1T
1
1
4 ~ 1 - J ~ 5- ...

+-

These are not preferable for computing !n 2

.rwhy

2rJ+T.

and .rr/4

as the

S!ries converge very sl.owly-

. If

f(x) is a polynomial of degree

n.

the power series

representing it, is a finite series, since all derivatives of order

heigher than

are ze'ro-.

Example. Expand

P(x) :::. x

- x.+

in powers of

x~l.

conve~genJ

44

\ ...

'
Sol ut.i on. p ( x)

x3

::.

P1"(i')

:.

6 ~ p ( 1) = l,.J> 1 (l)

P(x) :. P(O) +

x 3 -x+~

+ 1 pI (X)

-qp (x-1 f

= 1+2(x-l)

3x 2

- 1,

pII ( 1 ) = 6. . p111( 1 )

2.

-t P" (:0) (x-1 ) 2

3(x-1) 2

P11 ( x):6x,
::

+ p~;(:O) (x-1 ) 3

(X-1)~

Binomial series
The expansion of

(ltx)a

for any

aiR

series which becomes a finite series when

.ltx

=-

ltx

(l+x) 2 ~

(ltx) 3 :.

is called a binomial

a is a p6sitive integer n:

2x t x2
3x

3x 2 t x3

(") xn
. n

coefficients given by

where

( n)k ..._

ck _ C(n k) _ n(n-1) ... (n-kil)

n '
.k!

In the binomial series th:e coeffidents of

den6t~d

by (a)
. n

a::...-_.:.l...t..)___:_;'.,.~(=-a-....:n.:...;-i:....:.l...t..)

.l..:(

= (ltx)a.

Let

f(x)

f(x)

= (1-fX)a

f II (X)
f(n)(x)

:.

wi11 be

which we show to be

= ;.:.~

f I (X)

xn

0:

Then
~

f(O)

a(lU)a-l

=)

a(a-1 )(Hx) a-2

~-,

II ( 0 ) .::.

with

a-n

=a(a-l) .. (a-n+l}(l+x}.

0)

~
.P

1
a

-., f(n)(O) .::.

a( a- 1 )

ci(a-1) ... (a.:.n+l)

45

convergpn~e~

As to the radius of

I ::. lim

an
= nlfl anil

2 +c .+. ~a~(~a-_l~)~~~~(~a--n~'+~l~)
n ..

we have
:
(n+l)! . '
la(a~l) ... (a-n~l)(a-n)l

la(a-l) ... (a-n-1)1

n!

= lim_ln+lj
. a-n ::.1
.

Hence the i nterva 1 of convergence is (:-1, 1) as in geometric series.

For

a = -1

we have _the geometric series

1

14X
and for

a ::. 1/2,
1

7T"=X '

th~t

series when a<O

1

0-< a<

-(-1)

and their corresponding ,series.

s~ries

tor

x>O

is an

a ::. p+r

(p

a)

(1~~)~ is an alter-

Expand the following in powers of

7ITX

Solution.
a) .

ynx.=-

is a positive

we have

nating s-E!ries.
Exa~ple.

alternatir~

1.

has f1nite num~~r of terms and

jl+x)P

. .. '

the familiar algebraic functions

-and non integer, say

integer and . 0 < r < 1)

. .

the binomial

or

Wherr a : :.

-.....

-1/2, l/3, -1/3

.

.-Fx. -r:;;:=
.;r:x

.Observe

where

1-x+x

b) (l-tx) 5/ 3

x:.

:::. 1 -

'1

)
(- "2") ( - "2" - 1 )
-l 12 x. '+ .----...-r--- x 2 +

1!

2!

2.1!

1. ~ x2
-~

1
1
'1.'
/ ( - "11") ( -"2" .. 1 )... (_- -.r - n+1 )
+
c.

c.
xnt

'

n.

t(-l)n L .. (2n.:.l) xn _
2n.n!

~ ( 1 +X ) 5 I 3 ;. ( H

. 5
1
2
b). a .::.j.::..
+3

X )( l i X )

2/ 3

Jn the second factor the exponent being less t~an 1,-it can be expanded as ~bove.

or by direct expansion ..

(l+x.)5/3=l_+-~xt.l

which is an

c.

5(5-1)x2-+

'2:1'1'

,J

al~ernating

5(5

j'!"jj

l)(~-2)x3t
.)

series after the second term.

ALGEBRA OF POWER SERIES

The fell o'wi ng tpeorem _permits us to write the power series

f(x) + g(x), f(x) - g(x), :f(x) g(x)

of

and

f(x)/g(x)

as soon

as 1:_!le power series.of

f(x)

and

g(x)

are known. For brievity the

theorem will be stated for McLAIJRIN series:

Theorem. Let

I; an xn ,.

f (X) ::::

g(x) ;

I
0

be absolute1y'convergent on
ponding powerseri~s of

If,- Ig

respec-tively. Then the corres-

~g, fg, f/g

f~x) g(x) :.

I0

b ) xn
(a. +'
n - n

f(x)

pnxn

00

g(x) .::

ffi+
.

where

::.

-I0

::.

I0

(aobn+albn-1

are given by

anb 0 )x

(CAUCHY Product) .

n
qnx ' (b 0 tO)

. .. )

qn (n:O, '1 2.
.an
I .

::.

is given by the recurrence rehti ons

bo qn -+ b1 __qn-.:r + ... bn q_o,

47
-

each series being absolutely convergent at least in the intersection

lfn Ig.

If
points
at

>,<

f(x), g(x) _are 'expanded into power. series at distinct

2 , then_ f(x) mu~~ be expanded at x 2 or g(x)

;---- but i f not possible expand both at a point x 3 .
x1

and

i~

When

is necessary index shift will_ be helpful in the above

ope-rations. (Ind:_x shift:

-,

'

! a 0 _r

xn =-

'an

n-o.

xn-r)

n:;r

Example 1. By the. use --of

-Ch x:. 1 (ex+ e-x), Sh x =~(ex- e-X")

/' '2"

c.

obtain MclAURIN series for Ch

and Sh x.

Solution. Since
ex_ "" 1 i. x t x 2 _.

TT

e -x:. 1

TT

"2T
x2

..

"2T

2n

...

T2riT!"

2n-l
X .
_,_
(2n-1):. -

x2n
""(7ii1:

2n-l
1

(2n-l):

...

(-co,. co)

(-co, co)

by addHion and subtraction, we get

Ctix .
Shx

1 +

= "'
X

x2

"2T +
x3

3T

x4

2n-

X
v. ... + 1TrllT

.
"5:
x5

(-co, _a>)

2n-l
( 2n-1 ) : t
X

(-co, co)

Observe that the even (odd) function Chx(Shx) inVolves only

the ~ven (odd) powers of
-~

x.

EULER Formula:
----.--------

Assuming
that
.
.
-f

T; i -2-;

is absolutely

n-

2 '

ez :. 1

conver~ent fo~ ~11

z = -ix, the expansion;'

.. -1

h.+

complex numbers

z, we have for

48

eix

.1

ix

TT

( lX
. ) 2 .i

x2
-X
= 1 + i TT
"2!
x4
x2
( 1 -- "2! + T
5
x3
X
-t i(x +
1:
"5T

(ix)n

----n;-:-

,.
:r:

x3,

- ...

+ ( -1 ) n

in

x2n

...

n:

T2"iiJT
X

... )

n-1 .

... )

+ ( -1 ) n (2n-1)! t

cos x ; i sin x

eix ~ cosx + isinx

(EULER)

Example 2. Obtain the McLAURIN series for

So 1uti
- on
-:

Wr i .t i ng

R.n/f%::

-~

R.n

R.n

Jm

.J

....;:-.

~~~:: ~

from

[R-n(ltx) - R.n(l-x)],
2n-l
X
zn:T
+ 2n
x2n

( -1 ' 1 )

2n-l
X
2n=T

x2n
2n

..>

( -1 ' 1 )

we get
R,n

vn-::;
8
:
1:-x

X .. .f.

T.

.;,._ i
o

2n-l
X
; zn:T +

C0 S

b ) [ R, n { H

.. 3
X

11

's i n . X

X )]

Solution.
a)

(-1,

3. Obtain the MtLAURIN series for

E~ampl~

a) X3

X ;

COS

2n-+3

T2"iiJT

b) We have

[in ('1 ~x)} sin x

(-l)n-1

+ (-l)

... )

x2n+l
(2nfl}!- )

.-----

49

an

bn

3 ..
1

- 2

4
I

-4

Pn :_/0

"2"
2

1 '3

[R.n(Hx)]sinx =X. --2 X

I X

where the g.eneral term is omitted s i nee it J s unnecessary, and we

have all pro.perties of the series, because, those of R.n(Hx) and
sin x are known.
Example 4. Obtain power series expansions of the following
rational functions at t~e indicated points:

3+x
1-x '

a)

X :.

b)

3+x

X .::.

Solution.
a) Direct divisi~n gives (since
3+x

1-X = 3" + 4x + 4x

b0

+ + 4x

to)

~..

Observe that it involves a geometric series with common ratio x. Then

it is convergent for

lxl < L

obtain the same series by performing

(3~x)(ltxi ... + xn ~ .. ~).

a n d a 1 s o by d i ; f e re n t i a t i n g

-~(3 -t >.< ~ 1 - x )

'b) Th_e series being in powers of

X -

or
I

successively at x:. 0.
x-,1, use substitution

1 + t. Then

2
4*t
3*X : 3t(l-1-t)
l+t . = Ht :. 4 - 3t + 3t

;x

convergent for

3(.<-1)

lx-11<1.

3(x-l)~

." ..

- ...

+ ( -1} n 3tn

.., (...:1 )n _3(x-l )n

50

When the series of

f(u(x)),

on may set

illustra~e

f(x)

u(x)

is given, to obtain the series of

for

x. The following are examples to-

the process.
.

a)

s1n 2

b)

-x2

= 1 + ( -,x!2)

c) esinx

(x/2)
3!

= (7)

T!
.

1 +

'

1 -+- (x

. 2

+ Sln

x3

x5

3: -+ "5:

...

2:

1!

... +

...

(-x2)n
n!
't

2n
X
.t
t (-l)n ~

2T

(-x2)2
t
2!

x4

-i

(x/ 2 )2n-l
(2n-l)! 1

-t ( -1 ) ~

.. 2
X

'.

sinnx

---:n:

- ... )

1
x3
x3
3
2 1
+2(x.--r.:-r ... ) +o(x-J:-+ ... )
,. 1 .

(x -

2
esin;:::;
x
_1-+x-+z-:

D THE

~EMAINDER

TW) t

1 x4

(-y - T) +

~,

(o -

-y)

1 x5
TO

8'

THEOREM

Recall the MVT and its first extension for a function

defined on

[a, b], namely.-

f(b) ::. f(a)

unde~

f''((aa)
t---,--:-

(b-a)_ -t R2 ,

differentiability and

~ontinuity

--z:

R ::._ f""((.22) (b-a)2,

2
conditions.

The genera 1 i zati ons of. these is expressed bel ow as the

Remainder Theorem ~i Generalized MVT:
Theorem . I f

f (x)

de f i ned on .[a , b)

sat i sf i e s

(Hence continuous in
2. f(x)

is continuous at

and

b,

(a, b))

f(x)

51

e~ists

then there

in

a number

f(b) ::. f(a)

(a, b)

(b-a) + ...

such that

+ f(n}:(a)

(b-a)n -+. Rn-t 1

(1)

where the remainder

Rn-t 1 . is given by
f(n+1)(c . )
~
n+1 (b~a)nt1
R
nt1.,
(n+l)!

Proof. Transposing every term on the right hand side i I) ( 1 ) ~

. except

Rn+ 1 ' to the, left hand side and setting

'
f(n)(x).
(b-x)n,
!f( X) -.. f(b)-f(x) - fll(p (b~x) n.

ionsider the function

-x)n+l
. F ( X) ; Cl( X) - (b
I

(b-a)n+J

'('(a ) .

This function satisfies the three ROLLE conditions:

1) F(x) e: DCa, b)
2):

F.(x)

is continuous at

3)

F( a )

= Cf( a ) - '('( a ) = 0

and

b
~

and

F(b) = Cf(b) - 0:; 0.

Hence by ROLLE s Theorem there is a number . . en+ 1e: (~) such that
1

F~(_cnt1):: 0.

Now,
F I (X)

= lP
t

t (nt 1)

{b- x) n

(b-a)n+l

<P ( )
I

:; {:- f (X) - ( fIll(!~) ( b- X) I

(f(ni~~ (x)

'-

f., (t

J )_ -

(b-a)n - f/nn_,({!) {b-x)n-1 )}

'

(b-x)n
n+
'f( a )
(b:-a) 1
f(n+1)(x)
n
(b-,a) -t (n+1)
n .

-t ( n +1 )

(b-x)n
( b- a ) n +1 .

'f( a )

52

f(n+l)(c 1)
Cb-cnfl)n
----.-....:n~tc...:...._ (b-e +l )n+(n+l)
Cf(a)
n.
n
(b-a)n+l
f(ntl)(c
)
.
n+l
ntl
<f( a) =(n+l)!
(b-a).

0 :. -F ' ( c n t 1 )

-which is' inview of ( 1

the remainder

Rn+ 1 E!i
The Remainder Theorem when written in the form.

f(x).

= f(x 0 ) +

>

XO)

1:

(x-x_o l t

f(n)(xo)
(x-xo)n + Rn+ 1'
n

where
::.

-f(n'il )(c)
( n +l ). !

is called TAYLOR's Formula of

is the McLAURIN's
If

- o

'

f(x) at x0

Rn+l is neglected, then. f(x)

and the one

at
\

:.

Formula~

. by a polynomial of

If

(X X ) n~

= 0,

~pproximated

n:

degr~e

the

is said to be

ini~ial

approximation

s 0 (f)

is the constant

_ polynomi a 1
~:=f(x 0 )

The first and second approximafions

s2 (f)

s (f)~

are the-polynomial

functions
y

f(x 0 ) + f'(x 0 )(x-x 0 )

y::. f(x 0
of-which the first is the

+ f'(x 0 )(x-x
tang~nt

f"(x 0 )(x-x 0 ) 2 '

line, and the

tangent to the graph of f(x) at x0 .

y

s~cond

is a parabola

-53

If one approxi_mates J(x)-=: s(x)

error
of

en+~'

IRn+ll :, ~ince

is

Rn 4 l

by-

'sn(f)~

then the

is uncertain due .to the presence

'.

one must find an upper bound

. Example. Compute

~he

K: .

numbers
. -4

, a) _e

b) R.n

c) cos 3 .

correct .to 'three decimal places, th.at is with an error

less

than 10 -3.).
Solution.
a) We have
.. n

ex = (
X:l

k"

6 f: )

cE(O,X).

Rntl'

gives

( Y.~II..
_Jr.~ ). ;._

e::.

ec
E =- (nt1)!

=:pj

3
< 1
{nt-1)! lQ"J'

(0 < c < 1)

ec
{~:~+1)!

.e
3
E < {nt-1)! < lnTIJ!

:::?! ( nt 1)! > 3000

::. 7T (5040 ..... 5040 +.2580

-4''

-=i!!)

n+ 1 > 7

840 + 210-+ 42 + 7

+ 1)

=.. 13700/50..!1-b ~ 2,718

n
k-1
b)R.n(Hx)= (
(-:1)

xk
~T)

lxln-+l
Rn+l' IRn-+11< . nt1

(since the s_eries is alterllating'.

. 54

1 t

R.n

j ~(

X :::.-

(-l)k-1"

E::. IR ntl I _<

Then

~)-+Rn~l

(n+l )3

1 ....... (ntl")3nl> 1000

n-tl < :-:3' .....,
10

n > 4. Then

::ij>

4,..,

R.n

1/3.

s=

1.

,.-:-3"- 2.32 i

:. 0,333 3

~-

0,055 5

'4:'3'4
0,012 3 -0,0031

= 0,287.
c) We have
n

cos X :.

0:0

x2k )
R
. IR
I
"'['2"J<'J"T + 2n-+) '
2n-+ 1 =-

( -l)k

. 2k .
x
) + 0 + R
IR
I
"'['2"J<'J"T
2n+2,- _ 2n-t2 ~

or
n

COS X ::

o:0

w.hich are being- valid for

x ='1T
60

l S2ni ..nI c) J - I x I 2 n t 1

( --1 ) k

lcos
cJixl2n-t2
2n.;2).

"
in radian~ one has

rad (=- 3 ). Then

cos 3

' :,; cos

n
1T

60

. :. :\'

(-

1)k

L .
T2l<TT

1T )

60

2k

.t

The error in computation must be less than the larger bf

IR 2 ~ 41 1. IR 2 n .. 2 1.

Then

I R2 n .; 1 I =-

ffiTITt (

IR 2 n -12 I

lcos
cj
2 n +2 ) :

6'ITO ) 2 n -+ l < ( 2 n ~ 1 ) : (

i;-) 2n -+ 1 < 10 - 3

('IT)2nt21
(.,;.)2n+2<_ 10.-3
60 < ( 2 n-+ 2 )_ : o

The first inequa1ityjimp1ying the second) is verified for

;j
!~

II

il
'i

!!
~I
,\

:!

n >-0,.

in partiCular for

n :. 1:

55

9.8696

1 (1T)2
'1'
1T
-1
=- 1- 2" 60 .=- ... -m=-

7200

- 1.:. 0,0014.,. 0,9986=. 0,998.

approximately with an error lesi than 10- 3

Obtain cos 3

by the use of error property in alternating series.

E, SOME OTHER APPLICATIONS

1. Evaluation of limits in indeterminate forms!

lim
f(x) or
1i m
[ f (x ) - g( x ) j
c a n be eva 1 ua ted
X+X(9TXJ
X + X0
.
0
by expanding both f(x) and glx) into power series at ~.
'

Example. Evaluate
a)

lim
X +

(x)l/(x-l)

b)

I'

ex - l - x
SlnX - X

1i m
X + 0

'

Solution.
1

a) y :. xx--1 ~

R.n y = x=T.

R.n

y = -1

"'

(x-1) 2 +
2!

{x-1 )

R.n

x-1
- 2T

lim
.x + b"'

ex - 1
SlnX

R.n

x-1 )

( 1. i m y) :.1
X+ 1

=i-)

x2

f<
X

...

+ (powers of

1 i m R.n y = 1
X+ 1
b)

x. Then

R.n

"' Xli+m.0+

lim-y,: e.
X+ 1

... ) -,X'-/
x3
- . . ) ~/
V- 1:'

(..r""t/t z;-

-+

(power~ of x)
:

-co

(no limit)

- 1: + (powers of x)

z;.Evaluation of integrals_.
In case the
f(x)

eval~ation

is impossible, or

offer~

of

jf(x)dx

some

f6r a given function

di~ficulties,

it is performed

56

by po\'ter series.
'3

Ex~mple. Evaluate

. 2

A = ~ e~x dx . appioximately

Solution.
0,3

A =- ~

l- ( -oo,

. (1

-+

:. 0 , 3 -

( 0 3 ) 3

1!--3-

x2no~-1
0.3
n! ( 2n +l ) -t 1
0
.
n ~(0,3)2n-tl
+ (-l)
n!(Zn~l) +

is correct with an error less than 10-S

0,29443

(correct to

2! 5

~ 1)

0,089 - 0,0.00343

- 0,3
..

( 0 3}

oo)

five decimal places). Why?

\

3. Determination of the equation y(x) of the curve

through i giveq point with

y'= f{x. y).

Example. Fin~ the series for

cond i't i_on~_

:+Y.

y(~)

with the follo~ing

. 2'

1
::.

y(l)

= -2

(initial condition)

Solution. In the po\'ters series.

Y.(x)

of

y(x)

implicit

= y(l)

Y 1 (l)
y"{l)
2
~ (x-1) + -~ (x.-1)

'
the coefficients
are obtained by successive
c:lifferentiationrof (x-+y)y 1 - x 2 =- 0 using y(l) :'-2:

at

X=~

(x-+y)y 1

~'J. =

(Hy 1 )y 1
:::,

9 (1-2)y 1 (1 )-1 =-

9 y' (i) :. -1.,

(x-+y)y" - 2x .: 0
..:y"(1).- 2:.0 :::?;1y"(1);-2.

y"y 1 t ( 1 +y 1 )y"~ ( 1-1y 1 )y 11 -t-(x+y-}y" 1 -2 =. 0

y 1 y"-+2(1-ty 1 )y"+(x+y)y"' -2

-~

<- 2 ><- 1 >- y" p r- 2

1

=-

or

o, . y"'< 1 >

57

y" 2 -+Y~f"'+2y 112 i2 {1 tY 1 )y!"+ {1 iy' )y'"-+ {xy )y'"':.O

4+2.4..:y"""{l) :. 0

y"" {1) :: 12,

and so on .. Then
.

2' 1

.4 .

y{x) ; -2-{x-1)-{x-1) + y {x-1) t

Evaluation of

Of constant

s~rie~

term~

In il~ 2 evaluation of geometric s~r{es and some s~rie~

with general 1 term exp~essible as simple partial fraction was discus~ed. Now ~t ts time ~o evaluat~ identifying_ the series. In the

-first .step. identify

such that

f{a)

f{x), and in the second, find v-alue

__

... ... n!2n . . ..

1 + 1 +
.. r.7
2!2 2

a) A =

c)

of

is the given series ..

Exampie. Evaluate

"b) B

= 13 - (

c =1

1>

3 t .{

1> 5

1
2 .. 3d
3 ..

-+ { -.1) n

?"1 ..

:, n

(IJ~2nl

'3'11-T

2n+l

...

t ...

Solution~.

x
a) f{x) :. e

= 1

.A~f<l>=
re.
'

=J

r;

c) C:.

/3

x ::

-~

f n(f)n-lj

d
= ax

r xn

x?

xn

'2T t- . .. iiT + .

b) f(x) :. arctan "x :.

and

x3

x - 3

x5
T - ...

B = arctan

f{x) ~

= axd (l

Ln

and

+ ( ) n

x = 2

x2n+l

TntT t

(t":. j-

xn.;.l.

-+ x -+ x

+ .

) d
= ax

1
r=x

1
= l 1-x > 2
...

.58

EXERCISES CL 3)
31.

Find the interval of convergence of-the

.
2
nxn
b) I
(.x-t-1) n
a)
~3 n -1

followi,n~

..

-T:-

c)

power series:.
2
n +1 xn

6 n:

32. Same question for.:

a)

6 . n-1

--~

x2n

b)

33. In Exercises 31 and

32;

(x-2~n
n-t

xn
n

C)'t

discuss convergence at the end poin:rs

of the intervals of c/onvergence if finite.

I'

34. For what values of

'

. a)

r tn

there is absolute convergence:

x.

b)

where

. ry 1n

tri

where

2x ==

35. Determine the interval of convergence:

a)

b)"

P(n)x"

where
3~.

is

P(n)

r0 ..g!l
x2
n.

polynomial (P(n)IO

c_) ~
l.

n!
PTriT
..

x".
.

n).

for- all

Find the tnterval of absolute convergence:

f """"i}
n-1

a)
37. Same
a)

n-1

qu~stiori

.t

b)

6 n2 ( x-1 )n

c)

xn

for:

(x-3}n R.n n_4

n

b),

2n
(-l)n
X
.
.
"[2iiy.

c)

n
n xn
iiT

. 38. 'Expand .
a) cos x, sin x at X:.TT/2

b) Ch x, Sh x at

39. Expand the following at X:.O:

b) sec x, up to x6
a) tan x, up to x7

x~~n2

c) esinxI up to" x6

40. . Expand
/

3-2x
at 0.
a) f(x) =-. x:rr
x2-3xt2
(Hint for (b): decompose1i.h into partial .fractions)
X

at X:2,

b)

.59

41~

1/(l~ex)

.Expand

in poWers of

x~

up to

l/ (. H t ) .::. 1 - t H 2

(-

using

1 ) n t n

"2. Expand :
3
2
a) x +2x -3xtl

in

po1~ers

4
2
b) x -3x -+4

of x-2,

in powers of xt2

In the following exercises 43-50, us~ differentiation or integ-ration of a known power series:
43. Find tpe sum:

r'f

a)

n x

n-1

b)

n-1

44. Find the sum:

P+(atb)x+(2ab)x t
45.

Sho~1

.. ;

t(naib)x- ...

that-

2.lt3.2H4.3a
is convergent for

Ia J

2
+

< 1., and then find. the sum~

46. Evaluate
1

'2'
47. Find the sums:
( -1) n
a)

n2n1l

0 (2~,+1)!
48; Find the sum:

--

J7riil
x4
-T .

x8

9'"

x4n
+ ( - 1 )" ?(iii!

49. Evaluate
nxn

I0 'fTnT!"
50. Evaluate
a)
-

I(2n'-fl)x2ntl
0

51 . sh-ow the following

b)

4 .

60

-~

a) Cos ix:Ch x, Ch ix:cos

.
.2
1
wh ere 1 =\ - .
52. Let

.f(x) :

2 an(x-c)n.

Prove

b) Sin fx:.i Sh x, Sh ix:.i sin x

1i m

f (c) .:.

f ( x) - f ( c )

x-c

X + C

53. Prove that if

x2n
2
~ 2 n(2n}!

.f(x):.}

F(x) :

2n

b~

2
21 (x) - F(x) : 1.

then
'

54. Evaluate the following by series:

a)

1 i ITi R.n(hx)
SlnX
X+ 0
ex

liin

c)

X ->-

b) 1 im
X+ 0
I

sin X
X

d)

cos X
X

rx-:;--2

lim
X +

ex

55.r Evaluate_the following 1 correct to three decimal places:

a) /404

56. Graph
hood

b) 3{27090

s~(f), s (f),s 2 (f)

1
N[0].

c)

for 'f(x) =ex

/6,375

i11 a neighbor-

57. Show by squaring-that'

2

(1-xx

.l.X

2
+ } : 1 ;2x-t3x .j.

58. Show that the terms

by

a 1 :1,

a 2 :1,

an

-i

s~quence

of the FIBONACCI

an=an_ 1 +an~ 2

(n'-t-1 }x .......

defined
n
are the coefficie~ts of the

McLAURIN Jeries of_

(Verify only .6 terms).

I

59. show that the arc length of the ellipse

with.. eccentriCity

e -::. c/ a ( c =-

x:a cos9, y:.b sinS

.,la'Zb2) . is

61

s : 4a

_L
J0 2 'Vl-e
1T/

= (1 - 'll

e.

sin 9 d9
3' 4

- Oif e

)2

1ra

60. Evaluai~ ~orrect to ~wo decimaJ places:

1
1T
2
Sin x dx
b)
a)
sin(x )dx

6
j J64+x
0
X

c)

dx

d) 2

6 e-x

2 .
/4dx

Answers to e\!en numbered exercises

. 32. a) (-1, l),
34. a)

> 1,

36. a) ( -1 , 1 )

If)

38. cos

sin

= 6

r0

Sh

r0

40. a) ~3 t

(-1 )n+_l
(~l )n-+1

42. a) 11

-t

.
( 1. -

4
c) (.., 1 , 1 )

1 (x

n:

b-a

+ hx

1r)2n~1
.

.lln 2)n

R.n 2)n

.n

(x

1 (
n
3nt I x-2)
.
2
8(x-2)

bl 8-2o(x+2)+.21(i<-r2) 2 ;. . 8
4 4 . ( 1- x ) 2

(X -

1
-rr)2n
TZii"JT (x - 2

l..:.}L) n:

(~l)n-+1

17(x:...2)

( - 1 1)

( 2nd )!

(~J'n>

(1 -t

<X<

b) ( 0. 2)

..

Ch

( 1 , 3)

b)_ -4/3

b)

:t" (x-2)

b)

2n+ltl xn
b 2ntl

(~+2) 3 + (xt_2) 4 .

62

46. 1I ( 2 v'e)
48.

-fx arctan

50. a)

54.

1 tx 2 .
X

:--t '
1-x

""b)

a~)
1, ~b) 1,.

dx
. 2
x(1-x )
d) 1/4.

56.
I)(_

60. a) 0,30

b) 1,85

c) 16,11

d) 1 '84

63

'

A SUMMARY
(CHAPTER I)

l.

.. ..

, ) co:
... an,
f(n), then (a np
f: Z.:tR
p ' an
-_ a p'' ap+l'
is called an sequence which is conve-rgent if 1!rn an exists,
I.f

divergent otherwise. (an)l is monotone i f

or

a ~~ ~

2 ..

for some

-~<:n~

.. , and bounded i f

a 1 ~a ~ ~

--K~an~K

.. ~an~

for all_

n~l

K>O. Any monotone bounded sequence is convergent .

.....,

l. 2

f an

is called a series. It.is called

::. al ,. ... ., an.+ ..

convergent if

sn

= a1

~n

has a limit, divergent

otherwise.

6 ar"-

Geometric series:
. otherwise.
Harmonic. series:

T~sts

1
1-r if

I r I < 1 , divergent

1/n is divergent.
f
is conve-rgent for p> 1 ,

p-series: r 1In~
l
In a convergent series
divergent (a test for

.:

an .... 0. If

divergent for p<l

the series is

an -1+ 0

diverg~nce).

for series of positive terms:

1) Integral tests: If

an= f(n).

and f(x) is positive

(I)

and decreasing on

~p; cc),

then

~an__

and

J f(x)dx
p

are both convergent or both divergent.

2) Comparison tests: Let

r C:n

dn~a n~c n . Then convergence of

convergence of

r dn +Aivergence of

ran, and divergence of

ran.

3) Intrinsic tests of CAUCHY: If

there is convergence for
Test fails when

A =- 1.

Test for alternating-series:

"ran~ A or. antl/an~>-,

A<l, diyergence for l>l.

64

~n +

is'decreasing and

0, then the

I Rn+l.I <a ntl"

series is convergent, and

rests. for arbitrary series:

Lianl.~ .convergence of

_ Convergence of
series

ran

is called .

l. 3

L!a 0 1

ra (x-~ )n
n

x-x

(at x

Such a

ra~

Ia~

is' convergent

is divergent.

is called a power sertes, epande~ in powers

o.

of

absolutely convergent.

'is called simply convergent if

:but

La 0

The 1nterval, at the interior of which

).

the serie~ is ~onvergent .and at exterior divergent. is called the interval of convergence. It is determined by appli-.

cAu cHy r 0 0 t

cat i 0 n

Taylor

series~

orders a~

X=Xb

~i.ven'by ~

rat i 0 t est t 0

A function

havin~

f(x)

n~

derivatives of all

f.(n)(x )(x-x )n(n! ,-..called TAYLOR. series which is

. J~0

f'(x):.

= 0, and

f f(n)(x

)(x..;x )n-l/(n-1)!,
0

f (.x) dx :. L:f(n)(x )(x-x )ri-+l/(ntl)!

0
. 0
0
.

--- 0) a n

xn

+ ) b

.-

(\ bnf . tanb 0 )x
o

X0 )

has a ~ni~ue power ~eries expansian it x~J

McL&URIN series if

ra

r I an( X -

xn
n

(CAUCHY Product)

xn .

.rb xn

o n

Remaindei Theo~em: If
a

and

b ; t.h en t lie r e i s

and
c e: ( a , . b )

such that

cont~ouous

at

65 .

n f(~l(a~
(b-a)k t-Rntl'
k!

I
0

Rn+l

:.

f(n-+l)(c}
(-tr,a)n~l
( n" I ) !

n f(k)(a}
.k
(x-a) + ~n+ 1'
k!

:.

f(n+-l)(c)
(x-a) n-1
(n+ I ) !

f(b)

:.

f(x)

=:.Io,

where

CE(a~

n+l

or

x).
\.

'

61

The.~equence. (tn)l

MISCELLANEOUS
EXERCISES
.;
of triangular numbe.rs and the sequence of

square nu~bers are d~fined_below:

. ... . .

.. .
.

{.

..

'

. ..
~.

, n(n-rl) ,
2

...

'

( sn) 1 : 1 ' 41 91 16 1

2
.n

a) show that

is a subsequence of

( 8 t~+.l) 1:

...

s )
n
:b) discuss the numbers 2701 and 1979 to be triangular
numbers
c) show that (sn-tn)2 is i (tn)l
62. Given

'

..

th~

geometric sequence
~allowing

prove that the

a) {gn +2gn_-l )o
63. Let

(an) 0

. respectiVely

(gn)o

(gn)o where

are geometric

gn

:.

.a . r n
1 gn >0 .

sequences~

b) (",!go ... gn.)o

be an arithmetic and geometric sequences

(gn>~). Compute .the constants .A

a~d

G such:

66

that
an + G , tn g n

=1

for a 11

~ 0.

64. If
,!

(gri)o is a positive geometric sequence, sh ow that

.
1/p
(gn+~/gn)
. is independent of th~ integer p

65. Find the general term of the sequence defined by

a0

, an : Jan.,.. 1. - 4 ,

-66. Same question for:

a 1 , a 2, an"'" 2 ~ 2-a n-1 + an = o.

67. Use the equality
"1+'1
1
:7"*:7+
...
2

1
+--z+

(EULER)

to evaluate
s :. 1
68. Test for
a)

1'
1
-t:-z-t:-z+
.. ; ..
3

. 5

( 2n-l)

co~vergence:

b) ~ )

il!

1"5 ~

69. Write the numbers as

as the ratio of.two
a)a:. 1,7

geometric.se~ies,

and then write them

integer~:

,b) b ::.1,"21"

c) c

= l,2J

70. Test for convergence:

l:2

Iii inn

d)

n ..
a
1-an

b)

f (-1}n

b)

c)

71. Test

.;

tnnn
fo~

convergence:
2 n-1"
nP-tn

d) d :. 56,9

67

72.

for

T~st

convergenc~:

+ 2

(n-tl}:

13. Discuss convergence:

[r ...

a> y

b) . y

c2n-1)Jp
2 ... (&) - .

r_

_____,(. .,. 2.,_n-...-11 4 n+3 I 2

(2n)
Tn"i2"

,.....J_

c.

74. Test for convergence:

1

( 2n'" 1 HTn"TIT

b)

1 /

+ e 1/n

75. Test for convergence by comparison:

a) y

+. _b)

~ n -1

ysin_ n
T n 3

.c)

6 n2-n+5n- 5.-

d)

3.

1.

vrr ~n

'

76; Show convergence of

~

..a)
'

-(1

b)

R.n ( 1 -_ -

b)

t"

(n~2)(ni3)(n+4)(~

e .n

C!

an

-~-:

where

an > 0

and

an

oo,

77. Find the sums:

a)

11-=

t ( Zn+ 1)(- 12p+Zn+ 1)

(n+1)n
iiTI"

79; If

-~un,

a)

~l:Gn.

};v~

-b)
~re

series of

means of ~n' vn
Tes~

a)

""'
2
( (n-1 )~n-2))n
.. n

'

positive.term~show

b)

are convergent, where

80.

_ .n(n+l)

for.convergenc~~

78. Test
a)

(pe:N)

Gn, Hn

that

L Hn
are geometric and harmonic

~espedive1y~

for convergence:

r1

. 1

a "-tn 1/a

81. Test 1or convergence:

b)

a sin

n ,

(af_o)

68

a}
8~.

';/

L-,et

b)

s =

3 .
n:;;.;rn

(b. c > 0)

(an>O).Show 'that

6 an

b)

' .

s<s .

83 .. Test for convergence:

n
1
/ i
r
(.1 ) ' (:rn-:rr + rn~
n-0

(X)

of

1
1iit3')

i f P> l.
I'

85. S~ow th_at, (n-+l)! + (n-~>2)! t < n!n

A1 , 7 , a 0 , a 1 , ... ar.~ ~he'areas of cotresponding
region~, evaluate (P T , P1T 2 , are tangent lines)
0 1

. efo

'86. :If

A0

~-

,.
l37~

A1 , A2 , are the areas under the arcs

prove that
(-l}n-lAn _is convergent
lf

88. Find interval of convergence:

a)

r0 xn

where

Ii.
!

o(

Sh na

b)

xn
n

R.n na

is a positive constant.
find the interval of convergence of

an > 0

8/J. lf
~

i
i

90. Evaluate the sum:

I1.
I

91. Given
.
f{x)

x2

""~

+n +

-t

n-1
n(n-1)., '

a) find the interval o( conv:.ergence,

b) find

f 1 (x), f"(x)

and then ~ed~~e f(x).

69

Identify
93.

, xn = 20 (n+l){n-t3}"

f(x)

f( X):

Evaluate
a)

f _R.n (1

~)
n

co

94. "Find the sum

n-O

((n+l)!

t _k3)
n

.....
g(x)

97.

=6

( -1) n a (a..; 1 ) .- ( a~

n -

co

. 96. Find

"l

f nfn+l}Cn+2}

co.

b)

( n+3)!

n+l

(n~l)!

The se_quences
f~

:. 0 , f 1 :. 1., f n_2

1 o-

2-

1 1 =-

1 n~t

-~

f n+f n...; 1

1 n~ 1 n-l

are respecti~~ly called the FIB6NACCI (1180-1250) and LUCAS

(1842-1891) sequences respectively-'
Obtain the terms.up to

12

, t

12 , and then prove the

following:
a)

fn.+l/fn-+- T and .l'.n-+l/R.n-+-: where T is the positive root

of: x 2 - x - 1 : 0 (considering
.existence of limits)

b)

L
k-0

98~

k-0

.1'. k

= .1'. n t 2

- 1

For the sequences defined in Exercise 97 show:

2 c) f2n = fn .l'.n (LUCAS)
- d) f2ntl :. f2n + f n-1 (LUCAS)

f2 :. ( -1 ) n (SIMSON, 1687-1768)
n-1 f n-+1
n
n
-n
Tn-*(-:-T)n
f) fn- . -1
.l'.n ::. T +(-T)
(BINET} .

T+T
Determine the interval of convergence and-determine behivior

e)

99.

fk :. fn.+2" - 1 '
' .

at th~ end pofnts:

a)

+ 2.! X 2 t 3 !. X 3

-+

+ n! xn

70

-;z ..

b
b) 1
a +

CXl

100.

"Prove

e-1 <

bn-1
b2 3
, x +- i - - x"
a"
a
2"
I .. 3 ... ( 2n- l"J < 2e

..

(a' b > 0)

ANSWI:RS TO EVEN NUMBERED EXERCISES

66 .an .... (n-l)a - (n-?)a
2
1
68. a) Di v. , .
b) d i v.
70. a) Conv.,

b) d i v. '

c) Conv.,

72.. Conv.
74. a) Conv. ,

b} di v ..

a) Conv.,

b) conv.

78~

80. a) Conv. for .a < 1 ,

b) div.

e:..2
e
,
._ a) 2(e~l)
b) 2(e-1)
-2a
88. a) ( -e :..2a ' e
b) ( -1 ' 1 )
)'

~6

90. -3if (e + e- 1 )
/

92. - '21" [.~.n(1-x)

94. e/4
96.-ex/x

-X

R.n(1-x)
X

- x?-a]

d) conv. for lal<1

71

'i

CHAPTER 2
MATRICES
2. I.

MATRICES

'A.

DEFINITIONS:

A.rectangular array of the form

all

of

mn

mxn

...

~lj

a .
1n
i .. l, ...

ail

aij

a in

aml

a nj

amm

entries {elements) is

sfze {shape)
II

...

j:l, ... , n

c~lled

(reit~ngular) ~atri~

of the

(m by n). Some author:-s use the symbols (

lfinstead 0~ []to represent matric-es. If'

the matrix

, m,

is,call~d

a real

mat~ix.

An

mxn

or

aijtR

for all i, j

matrix

con~ists

of m

rowsand

n columns.
The .element _a1..J. lies in the ith row ana
.
the jth column. A matrix consisting of a single row (column) is

called a row matrix (column matrix).

The above matrix o.f size
following

is abbrevia-ted by one of the

~ymbols:

In some cases the subsecipt

If

mxn

= n, the

matri~

mxn may be omitted.

is called. a square

to be an nth ordered matrix.or:- a matrix of ordei"

aii

in~

squa~e

m~trix

and is said

n. The elements

matrix lying on the main diagonal are called the

dia'gonal elements.
In an

mxn

(m .f. n)

matriX the elements

similarly the diagonal elements.

aii

may be called

72

If in a square matriX all- entries below (above) the diag-onal

are zero, the __matrix is called

,.

an~

(lower) tri'angular inatrix.

-The following are"triangular matrices:

3
0

[:

oc

0 0 3
An upper triangular matrix

4 -5

A lower triangular ma tri.x

-;--.

(a .. = 0 for
1J

(a .. = 0 for i>j )'JJ

i<j)

Tae above. definition may. be ex-rended to any matrix, with

:Ls

.1

Q 3

[:
0 4x3

'Transpose of a matri x:
The tr.anSEOSe of a matrix

A= (aij]mxn..

is the matrix

AT =
[ajiJnxm According to this definition the transpose is obtainT T
ed by changing rows into columns and column into fow.s. Why' (A ) =A?
ExamEle. Wiite the transpose of the following matrices:

and give reasons foi unalteration of the diago~al elements.

I

Answer.

AT

r:l~ : :]
'f

BT :l

~3

0 -lJ

. 5 .-2

~. ~ ~. =[: ]
,

. : '7.

73

In a

matri~

and its

~rinspose,

= aji

aij

when

and these

f:j

matrices havethe same diagonal-elements.

Some special square matrices:

ln. a square rna tri x (a .. ]

1J

a .. = a . (for all i, j)

a) If

1J

_J

a ..

is called a symmetric

rna trix.
b) If

a .. =
1J

(for a 11

J1

symmetric matrix, and

c)/ If

i lj ~

when

aij = 0

A.

d) I f a diagonal matrix
A

i~

a scalar

is called a skew

is zerd axial since

is a diagonal ~atrix.

= 0.

aii

A have equa.l diagonal elements

matrix~

e) If in a scalar matrix
(unit matrix)

i, j)

A::.l,

is an. identity matrix

is a zero matrix

In[::.oij]n

f) If'in a scalar matrix

).=0,

On.

The matrices

0 ... 0

z
I

0
0

; - .n

... 0

are diagonal, scalar, identity and zero matrices of order

0
n

respectively
An identity matrix
the (kronecker

is the matr.ix[o ij] where

o) deffned~by

when
when

hj

0 .

1J

is

74

Accordingly [>..i :oij]' [Xoij]

rices

are diago_nal and scalar mat-

respectively.~-

For a square matrix

A= JaijJ' the symbols

I Aj,

det A,

det[a~'-]
are used to -denote the determinant of A:
. 1J

I a i j I ~- de t

I A I,=
We note that i f

-A

de t [a i j]

I A I is

is a non square matrix,

not defined.

B. Operations with real matrices:

1. Equality: .The matrices [a .. ], [b.-.] are equal i f tlrey
.

1J

. 1J

are of the same size and corresponding elements are equal:

2. Addition: The sum of two matrices [a .. ], [b.-.] of the

1J

1J

same size is a matrix o"f the same size whose elements are .the sums
of

thei~-corresponding

elements:

3. Multipllcation by a scalar: The product .of a matr.fx with

a sc~lar is a matr~x of the same size obtained b~~multiplying every
el~ment

or

the matrix by that scalar:

c[a;j] :. rc aij] :. [aij]c

I f c = -1, then

cA

= -A,

.and it follows that

.4; Subtraction: The difference

and

B of the same. size is the matrix

5. Multiplication: The-p~odutt
number bf columns in

A~B

A .-+(-A
):0 _
mxn
mxn _ -mxn

of two matrites

A .. (-:-B):

AB

is defin~d only when the

is equal to ~he number of rows in

B.

75
r-

'
If A has the- size

AB

is

t~e

matrix

::

mxn

and

of siie

mxp

where

[ci}mxp

-""
In words, the element
is obtained by taking the
B (having

thasanH~

ha_s the size

c ..
lJ

nxp

the ptoduct

defined by

clJ
..

=-

k=l

a i.k bkj

of the product matrix

ith ro\'1 from A and

AB

::.

.c

jth column from

number elements) and adding fhe products

of
. i

their corresponding elements:

.The matrices where :product is defined are said to be_con_,..

f o r t a b1e .

Example.
I

T. Coll)pute

Solution.
\

If
then

o.A t

..

A, B,
SB +

are matrices of the same size and' a, B,

is called a

linea~combination

2. Compute, when matrices ~re conformable:

a)

of

A~

B,

76

Solution.
matric~s

a) Sizes show that the

and the product

C is of size

3x4. The

are
cij

ele~ent

C is

of

B and

the products_of their respective elements:

[ 2.1+3.2
-3.1+1.2

AB =

:2.2+3(-1)

2.3t3.0

-3.2~1.(-:-1)

-3. 3... 1 0

0.2+2(-1)

.0.3+2.0

O.lt2.2

_;: l

=E

'

-7

-9

-2

inner

.2.4+3.2]
-3.4tl.2
0.4+2.2 .

. 4

b) Since the sizes of the two matrices are

and

conforma~le

A . and jth column of

obtained by taking the ith row of

addi~g

A3 x2 , B2x4

numbers are. not equal (213),

2x2

mul~iplication

and

3x2

is-undefin-

ed.
3.

ferform the multiplications of row and columns rriatries

'

a)

[2

-1]
4] [-:

-3

4]

-3

Solution.
a) Since sizes are

lx3, 3xl

the product is of

[2(-l) + (-3)(-2) + 4.~

.t-:::.1.2

b) The p_roduct is of size

'

[ ::] p

-3

4]

lxl:

~ize

:[r6l

3x3:
-1 (-3)'
-2(-3)
3(-3)

-1.4]
[2-4
-2.4
::.

3.4

3
6

6 -9

-4]
-8
12

77

Comparinp:the ~bove res~lts

...._. general

AB

AB

and

BA, note tha~ in

BA ..

4. The product of two noo zero matrices maybe zero. Deter-

mine the niumber''alsuch that

[2s 1o4J - [-21 - a]3 == [oo .olo]

Solution.

[:

t
5.

If

:J

[_, H :J
1 -3.

'''l["
Sa-30

:J

:::>

2a-12 = 0
Sa -.30

= 0

=?

a::.6

is a square matrix, the following matrices are

defined:
AO

=.I ,

, Al =-- A, -A 2 =- AA , ~ ..

Then evaluate the following linear combunation

of

A2 ,, A; ' I

wher,e

Solution.

(nEIN) .

78

0
2A :. 2

2
0

_, r,

A 2 -2A-3~I3

~-

2:J

J: _: :1
l

0 -3_

= r-13.-1,
.

3 -1

Remark. Note tha~ multiplication of a matrix by a scalar

and that of a determinqnt by a scala~.a~e defin~d differentiy: a
I

. ,

matrix is multiplied by a scalar by muliiplying ev~ry element by

that scalar, while a determinant is multiplied by a scalar by
multiplying only

~ne

row {column) by that scalar.

Thus
I

t'
~

.c
...._./

6]7 :.

3c

5c

6cl
7c

4c

9c

2c_

6'

9 c. 2

...

.C

I.

I
i. . . .
I' .. .
~

As a result we

,.

!'

.1

(.

hive~f~r

];

8c

6.c

;J .

7 = 3

5c

9c

matri~

A of order .n,

79

C, PROPERTIES

1. A;.B :. B't-A

. (cbm!J.!.Utafivity)

2. (ArB}-C::: A-T-{B-!-C)
(At-B) T : A\sT

3.

4. IA ,;,_ AI ::. A'

is said that

(The transpose of a sum is the sum of

.transposes)
AB t BA

but
A

and

5. (AB)C-=- A(BC)
6. (ABJ T = BTAT

(Associativity)

in genera 1. If

AB

::.

BA, it

commute.

(Associativity)

~.

(The transpose Qf a product is the product

of transposes in opposite order) .

7. I ABI =- I AtiBI = I BAI (The determinant of a product is the

. prod~ct of determinants)

8.

(AB)C :. AC:-BC

(The right. distributfvity)

C(A"-B) - CA"-CB

(The ~eft. distr ibuti'vity.)

1

'

Froofs.
/

1.

"

[a ..J
1J

"'-

[b 1.J.}/= [a .. .;.b ..]

1_J .lJ

=-[b .. .;.a.J
1 J 1J

=[b.]1J

[a ..
1J

2. Proved similarly.

1= .I a . ' I L
"b ..-I =

3 . (A.,. B) T:.; .(a .. + b ...] T = [ a .. +b . .

.1J
1J
J1 J1

4. IA = [oi}[a;j1
s~

!ilarly

AI

=(?

"1"

J1

a;) =fai) ::.A

ATT
.;. B

A.

5; Let the symbol (A) ij

we prove

oik akj] {o;;

-J1

denote th~ ij-entry of

((AB)C)ij =- (A(BC))ij" Let

A. Then

80

p
((AB)C)ij

;.

(AB)ik ckj

k:l

::.

k=l

\ bn,, ckJ.)
Lk
AA
.6. Let

and

B " [b; j]nxp

[a; j] mxn'

= [cij] pxq

Then

CR-:.1
I

=I

(aiR. b.9.k)ckj

a. (BC). = (A(BC)) ..
1~
J
1J

R.

'

=-. (aij]mxn'. B = [b;j]nxp' We prove

(B

Settin9 . aji

::.

A )ij::. ((AB) );j

and

aij

::.

bj i

b i j ; we have

;:.

T
((AB) )ij = (AB)ji

Ik

:.

a j k bki

:.

a I kj b I ik

TT

b I k a I kJ' =- ( B A ) ..
1

1J

1. The proof.maybe found in Linear Algebra.

8. (A+B)C:. [a .. +b .. ](c,.]
1J 1J
1J

= [Iaik ckj"
=The left.

AC

= [Ik

l'b;k ckj]

(aiktbik)ckJ.]

[Iaik ckj] [Ibik ckj]

BC

.dts~ributivity

is proved similarly_.

' 1 e 1.
Ex amp
1. Find all

2x2

matrices which commute with

81

Solution. Let [a b,] be such that

c . d
.

21 [a

-~ [a
c!l = c

-1 -1J c

21
dJ -1 -d
bl [ 1

We have
at-2c

b!-21
-b-d

[ -a-c

a-2c

pro.ve

'

b+2d

= 2a-b

::. c-d,

-b:-d

2c-d

btd

::

-2c,

d-a

.::

2 c'

Example 2. If
2
(MN) ~ 0.

2c-dJ

a-b,

::.

-a-c

2a-b1

=[ac~-db

M, N

a'

b = -2c

:9

:.

-2c,

2 .

commut-e and .M 2 = 0 .or

~
. -2
?
Solutiqn. (NN) - (MN)(MN) _ (MN)(NM) ::. M(N'-)M

::. 0,

d-2c

then

HOrl :. 0.

Is it.possibl'eto define multiplication of two determinants?

Yes!. The propertry IAIIBI ~ABI can be used tr. define the
produtt of two determinants ai a single
are obtained by the-same rule for
Example 3. Verify
.A -

multiplicatio~

(AB.)T:. BTAT
3]

determi~ant

for

B=-[1

- -1

whose elements

of matrices.

82

Solution.
AB

BTAT
'

=. [-2

sl ::;

(AB)T

[:n.

[; lGJ [::J
=

0, INVERSE OF A MATRIX
-If for a given matrix

Amxn' there exists a matrix

Bnxm

such th~t

B is called a right inverse of

A, and if there exists a matrix

CA _ In

C is called a left

inverse~f

A.

Observe that a right and a left inverse of

A-mxn (m

# n.)

are not of the same size.

Example. Find a right inverse
if

a~y,

B and a left inverse

C,

of the rectangular matrix

1

A :.
. I

[3
-l

-2l
J2x3

Solution. If a right'inverse exists it will be of size 3x2.

Then, writing

we have

83

a.- 2e

2e -1 ;

::.

3a+4e

3a

-:)

b - 2f

T,

;.

- 0.)

4e

3a

1Oe+ 3,

o,

;.

. ( 4 equations,
6 unknowns)

4f =.

::.

2f

:.

3af-4f-l

10 f -1

2e-l
and

B :.

10et3
[

....e

. showing the existence nf infinitely many right

one .is obtained by taking

e::.l,

inver~es,

of

whi~h

f -::. -1, namely

3 -11-3].

13

- 1

For

. -1

left inverse, writing

0

we have
a't3b'
c't3d'
e't3f'

=I

b'

:.

0'

=
=0
:.

-b

-d

= 0

-2a -r4b'

::.

:.

-2C t4d

::.

-f'

::.

a' = 1 ,

So given recta ng u1a r matrix

of

-:-2e'+4f'.

=-

(9 equations,
I

6 unknowns)
:.

(no solution)

A has no left inverse.

The example shows that the determination of.a. right inverse

2
Amxn leads, from 1 A~xn Bnxm: Im~ to m linear equ~tions
I.

84

-with

mn

unkn_owns. As to the left inverse of Amxn' one has

~quat~on~ with mn unk~owns.
Square case:
The

rig~t

A of order.

and left inverses-. B- and- C of a square matrix

n, if exist, are of order

AB : I

II

It is proved in

~inear

and

n
CA

ands_atisfy
~I.

Algebra that (considering

syste~

of

,I

'I

eq~~tions

linear

right~(left)

arisii'lg from-the aoove equations) if

inverse it has also a left (r~ght) inverse which are

-
equal to each other, and this is the case when
AB :. BA

IAI f 0:

-The Oniqueness of stich a matrix

.Suppose

A has a

B can be proved as follows:

C is another inverse of A. Tnen

AC-: CA:. I

and
\

B
Thi~

. ed by

-1

':

BI

B(AC) = (BA)C

:::.

IC

:.

c.

urli qu_e matrix /B . is called t'he inverse of A denotand if A-l exists then A is said to be an inver-

tible matrix,
Si

nee

it follows that
'

To show that a given matrix

A, it will suffice to show
both.

AB :.

B is the
o/r

BA: I

in~erse

of a .matrix

but not necessarily

. 85

Example. Show that

=[

~oof.

_ -] ;s

We only need to_show that

BA ;

l-_:

-2

-1

Evaluate
Som~

not. From

the invme of

_J [: . ~-

;j

~]

BA =I:

~I

= [:

1 ._ lj

J-

AB!

square matrices are invertible, but some others are

II :

one/has

invertible,but since

OB

- 1
I~

BO

= I
::

and_identity matrices ar_e

0 f I

zero matrices are not

'

invertible.
--.

A . i's invertible, then(

Corollary. If

I AI
1\--oof. AA-l

--

.;.

and

1
IA - 1

I AA -ll; I I I

= IAI- 1
IAIIA- 1 1::

'

.~

Note.
is invertible

Th~

lA I i 0. and

I A-ll

1 I lA I.

converse 1s-'a-"l so true, i . e. i f . IAI 'I 0'. then

(See 1.3, Corollary

Corollary. If

:>f CRAMER'sRule_, BOOK ONE, Part two)''

A, B ~re invertible square matrices of the

same order, then AB is invert-ible and

(-AB)-l :: B-l A-}
Froof. (B-lA- 1 )(AB):.: B- 1 (A-lA)B:. B- 1IB:

s"' 1 s:.

I.

''

'.

BQ

Determinati6n of

th~

inverse of a square matrix:

Given a ~~uare matrix A, its-inver~e A- 1 can be deter~
mined, namely
1 by the d.e f i n i t i on ,

'2. bY adjoint matrix

3. by

element~ry

row operations (sweepout

proce~s)

4. _by CAYLEY- HAMIL TON Theorem.

1. By the definition:
.This is the method used in the determination of a right or
left

i~verse

of a matrix, and

~pplicable

also to square ones.,

2. By the adjoint matrix:

By the classical adjoint(l} of a square matrix

A = [aij]

is meant the tr~nspose of the matrix [Ai) of the cofactor.s of

aij's. In this text we simply use

"adjo~nt"

to mean the classical

adjoint:

Example l. Find the adjoint of

.

"

a)

=[

-2

-J

b)

Solution.
i!) Since

All :.' -3,

A2l :. -: l ,
A3l

:.

l ,

Al2

.:.

-3,

Al3

A22

::.

3,

A23 = -1

A32 = 3 ,

A33

::.

-5

[Aij}

['
-1

l.

-3
3

~]

3 -5

1 ) In Linear Algebra the term"adjoint" is used in two meanings that are entirely
different. The'one we used here is -n.o;.:med "classical adjoint'~ to ,distin,suish it
from the other.

Adj A

b) Ad j

B :

J:
-:
:1
.[3 -1 -5

~:

Th eo rem. A-l :.
A .: l'
ra.lJ,
l

[_Aji]

.. - :]
1 . A.
TJ\f
dj

(AJ.i]
:::-TJ\1

l'f

IAlt

0, i.e. , if

is invertible.

Proof. We need to show that

Adj A
A ~:.

or

A Adj

A= IAII

Indeed,

~ [oijlA]tlAI~[aij]=by Theorem 6 on determinant.&( Book I)

Exa~p1e 2. find the inverses of the matrices

and

in

Example l, if any.
Solution.
a) The classical adjoint of

was obtained as

the matrix

-~j
-5

and the inverse is obtained byd_ividing this matr1x by

IAJ= -6

sa

lA\ =

1 -2 -"1
.1 .1

-1

~1

l -1

:.

--

2(-2-:1) -- 1 . 0

-=

-6

Hence

-l./~6

"[1/2 . l/6
.
=.__112 -1/2 _..;l/2_

. - -1 /2 '

B =. 0,

b) Since det
3~

Examp)e

1/6

5./6

B is not invertible.

Find the inverse of

A-.

[ac~.

_bdJ .

1.

1~1

..

ad...:bc 'I 0

Solution. Since

A- 1 .~-[

we have

d .

-c

b1 liAI.
a

3. Elementary row operations

1

Let
matr-icesbe

b~

R1 ,

rectang~lar

.~._.

matrix of shape

mxn. Let the row

Rm.The followi-ng operations on rows are

called -the. ~lementary row operations:

Ri~

Rj

Interchangini of ith and jth rows,

Rj

Addi ng the j t~ row to the ith row,

R;

c R;

Multiplying a row by~ non zero ~caiar.

89

An elementary operat-io~.follo_wed_by another is an operat.ion_

calle_~

_a row operati_on. :The .row operation

-.
~as an o~vious m~aning\-

A finite

num~er

A produte a matrix

df

A'

el~mentary operati~~saponied

which is saidtobe

to,a matrix

rbw ~quivalent to

A,

written

A - A'
.In transforming a matrix 'A
useful ncitations

f~r

'1
-a 1.

ajl

ai n

- . . . .
- . - ajn

. . .. .

akl

A'_

R-++
R.
1
. J

'

A mairi

to a row equivalent ma\trlx

~r6ble

. . . ,. akn
i~ t~

here

y
),

transform a given matrix. A

to

what we call a echelon matri. By an 'echelon matHv -(echelon form) .

I

i~. meant' a matrix'in which t."e _nu~ber t?-f iero'elerr.o!nts, in eachrow, preceeding .the first non- zero 'e_lement (the distinguished e'lement)
--

increases from row to row until the-last row,: The iast

\.

'

ro~ and some rows prece~di~g-it may consists of-zero '1ements un1y

.I

This mean that:

( j .. 1 ; - k )

. ' k+ 1 . at . lea_s t)

.l

90

The following is an echelon matrix.

0

-2- 3.

The~di~tinquish'd el~ments

rows are

-2' -1 , 8 ~ -5,

ro~, whi~h;

8 .. 0

5.

-5

:o

{the first non zero.elements}'in

and the numbers. of zero elements i,n the

-rows p.receeding these .are

the last

-2

;..,_

1, 2, 4,7.;

and they" increase untii_

this-exa~ple~con~ists

in

of zero elements

only,
Threemore

I
l

~
~

~xamples

are:

-3 .

'

o.

2
0

'

I)

~].

'4

,.

o .1

.0

ii\.l

Example. Reduce. the given-matrix

-0

to an echelon form

A=-

.3

.4

4 '

I'

Solution. A is not . of .e-che 1on for!ll. Then

.I

-2

Zero matrices are considered to be in echelon form.

,j

91

,.,
/

.0

8'

6'

4 . 1

6 .

'3

. 2

2.

.o

o.

-4

.1

.0

8,

.,0

-4'

1.

0 .. 1

1.

-1

0.

1'

'3

1.

0'

3/2

~~

.f"l

)
/

~-'

Method for f.inding A-;1...

Let
i.n~

be a squa r'e matrix of order

n. llhe m.etho'd for find-

A-1 . consi.sts of considering the matrix

1Air.'lnx2n

and b.y app1ication of row operations; reducing it

the matrix

lr!afnx2n

.'

where . B

is. the'de~ired
1nverse;
if not reducibie to this 'form
.
.'
.

in~ertib1e~

is not

'

.The p-~oo.f of above statement will be given_i~l. 3.

92

Fin( the inverses of the matrices

Example~

1] _

r~-1
l~

a)

B=

. b)

-2 -:

rz
L6

~1

91

Solution.

IA lhx6 =

a)
.

-2

-1

...

.,
I

:.

f.

-2

I
I,
. I

I .

3
,

-2 . -1
5

. 0 . '2/5
0

1
0

0
1

lI n
I
I
I
I I

1-

-1

1 -2
1 ' 1
~ T5'
~
2

"5

112 1!6

I 1/2 -1/2

-I
1-1/2 1/6
I

:lr
;J

~J_:

.oj

. ~-

-2
5

'.1

I
I

-2

roJ

r 0

-J

.G

. 0

. 1

.:l

'

0- .

I
t

.-2

I
I

( 1/3)

~1

iJ~:;:~ .-- G
~

.I

3/5 :

. 1

.rJ-0

(-'l/5) .

-1/6]
-1/2
. 5/6

A-

-. ~

I.

c...

1/2

"~"

- 6

. 3/5: 1/5 -2/5

=.
[

1/6

0 .

-t

l/2'

=t

1'

I--~

-0

i:_ i

.o

r .,I ~2
1 3'51
I
~
'!. ~-

o I

.-2

_: -:

'1 -4-1/2
. ~

1/2

1/6
-1/2

-1/2.

li'6

1/6

-1/6]
-1/2
. 5/6

5/

j.
-.3/5

~,

93

b)

[s1 ~]2x4

[:

I
I

. o-

I
I

3/2 II 1/2.

.,.,_
. 0

I:

0
whe~e th~

1-

,,..,

{lf2)

,.

I\ l/2

3/2

I
I
I

0,

"]t6)
1 ..

~]

-3-_1 '.

matrix on the l~ft cannot be r~duced to th~ identi~y

matrix. Then

B ;is not invartible ..

A. By CAYLEY-HAMILTON Theorem
.

The following theorem is proved 'in Linear Algebra:

.

Theore~

own

(CAYLEY-HAMILTON). A sguare matrix satisfies its

characte~istic

equation.

The characteristic poly'nomiai of a square matrix

A'-~ [~ijln

is defin~d
~o- I . be-the
p~1yn6mial
.
,
.
.

P{A) ;!A-AI
. I:
' .

a ..
ln

a-11-).

al2

a:n

~22.-).

I
k:.O

..

k
ck). .

anl
-and

P(X).= 0

is the

ch~racteristic

nn

-).

equatiDn of

A.

C1 e.a r 1y
P_(O} :
~he

IAI:

~o.

theorem states that

n

p (A) :.

ck At ::

k-0

0n.

which is
c A0 + c
n

n-1

An-l
.

+ c' A
1

t I~;

o0

'I

94

-M~ltiplying both side of this ~qu~tio~ by

A- 1

(if exists),

we have

since

AkA-l

= (Ak~fA}A~l

~ Ak-l{AA-l} ~ Ak-~.

This latter equality is solvable for.-A-l

.which is the same condition-in
If IAI

= ~--0 = 0,

then

M~thod

0hen

~-~

A = .

-2

G 1

~J

Solution.
~) . The c~aract~ristic equ~tibn

2-).

is

. 1
-2-).

-1

.-:: 0

2-).

Expansion gives

a~d

by

CAYLEY-HA~ILTON

- -A3 + 2A~
~

-A2

=)

-6A-l

T 2~
:.

A2

Theorem we .have

5A +
+ 51

61~"' 0

6A-l

= IAI* ij

A is not invertible and such matrix_

~l

p (). } :::.

2.

Example. Find the inverse of

a}

"" 0

- 2A - 51
'

'

95

-l

..

3]1

1 z2

-1. 4

-l

l--4
-2

.:2,4 --?]2 ..+

4 .

-2

-2' -4

1/6

-i

. -1/2

0' OJ0 = [-3-3 -1-1 l1iJ

0 -5

'

1/2 --1/2

A . .:.

[5

-1 -5

t61

..

5/6

12 ....
31
'

0 -5

-1/2 .

l/6

.
b).co=-IBI.:.

=- 0. Then

is non invertible.

EXERCISES (2. 1.)

1.

Let A :. [a ..]
l.J

be a square matrix of size

when

i+j

nxn

such that

is even .

What is the num~er of non zero elements in

2.

A._ [

If.

1'

2 -3.]

-2 -4. 6

B
l

= [~4

i ] ,.

' 2

A?

find

b) BA .

3~.

Given

A =- [ai

~i~ ~xrri ~nd .. B =[ai 1 cSi~ mxn

, show that

:AB.:. BA:. In.

..,

4.

Let

A [:
a)

-:

and

B .

G -:J
5

Then find

3A-2B

- :l

.'

s.

If

..
A

:.

:J.

[:

B :

a) AB

(: j

b): BA

find
c) 2~ 2 -r3B 2 -5A+4B

'
-~

6 . Verify the associative law for the product

-1

-2]

0 .2

"

t 2,

2.

'

"1

7. _Show that the matrix

['2l
1

2 - 3

1. -2 -3

-,
is a zero matrix.
8.

lf

_P,
.

:.[~
1

-:

-:j-

-2

-3

~[-: _:<l
-1

~J

..

the'n -prove

c) PQ=O,

9.'

If -A

-~find

is 'a -matrix and

the shape of A. -

~1{:1

QP=Q-

97

10. If

An. I:l
find the matrix

and

A.

a)

A ~atisfies

A2 - A

b)

8 satisfies

83 - 28 - I ~ 0

2I ~ 0

where

c)
14. If

and apply CAYLEY-HAMilTON Theorem to gel

ATA :. I .then

a) Find

A is ca 11 ed an orthogor'

in
A

for

~in

cos g

sin.

A to be orthogona 1.

b) Prove that the inverse of ari orthogonal matrix is an or-thogonal matrix.

98

15. If

find a set of numbers

and an'orthogonal ma~rix

x, y, z

such that

i6. Show that

=A

a)

A i.s symmetric' if

b)

A is skew symmetric i f . AT = -A.

17. If

A and

th'at

AB

B are both

skew symmetric matrices~ prove

2x2

BA.,_ Is the property true for

18. Prove that i f

a) , A"t AT

AT

3x3

matrices,

A is square matrix, then

is symmetric,

b)

AT

is skew symmetric

19. Use Exercise 18 to show that every square matrix can be written
as the sum of a

s~mmetri~

and skew symmetric

matri~.

and apply

this to

20. If

S and

T are symmetfic matrices which commute, show that

..

ST

i~

symmetric.

21. Prove that the pr_oduct of t;,wo

'

99

a)

Scalar matrices is a scalar matrix,

b)

diagonal matrices is a .diagonal matrtx.

22. If

A and

B are both diagonal, prove

23. Find the 1eft inverses of

a)

~he

AB

= BA

following matrjces; if any:.

UJ

24. Find the right inverse of t~e matrices in Exercise' 23, if any.
25.Ftnd the inverses of the following square matrices by the use
of all given methods:
a) [ :

:J

26. Find the inverses of

only by CAYLEY-HAMILTON Theorem.

27. Find the inverses of the matrices by inspection.
0

a)

... , 0
,

. I

,
, ,

( d'l .. dn;'O).

b)

"

0
0

0
0

0
0

0
0

dl 0

c)

'

'

, dn

d)

II

100

II

I[
:1

;i

28.

if

JL3; -:

:J

a n d B -= [ :

-1

-2

.
. A-lB, B-lA
evaluate

-~-

and show that each is the inverse of

the-'other ..
29.

P~ove

a)

(A-1) T:. (AT)-1

(tnanspose each side of.

b)

the inverse of an

orth~gonal

30. "under wnat condition for

. -1

A~

matrix is its transpose.

x, the matrix

is non invertible .

. ANS\IIERS TO EVEN NUMBERED EXERCISES

2.

4.

10.

12 ..

a)
a)

k :J
[6' -11
5 ...;10

[i

b)

Lt-~ -!J

61

c: )

b)

-~8~; fJ"

-2
-1 ]
0

r- 1 .()]. ' r- 3 OJ r- 3
La -1

Lo-3

OJ ,

Lo-1

1- 1 . OJ.

lo-3

_,:4]

'

~I).

or

21

10

II

II,
dI
!I

101

iql

14. A = -cos g

i[
I,

24.

'

,,,,

a) No right inverse,

b)

26.

il'I

"2"

28.

1-2t

-2s

9t-4
---;--

---;--

tl~7

A-1 :::."2"

c) No

rig~t

inverse

9s-1

-q

1
4
-3

!l,,

If

,,
il

,,H,,
I'

-1

r
~

-2
1
3

-J
1

-1

::. 44

[5

-2
13

-10
4
18

;n

(C.rttff'nll.p

30. xc{O, -2, 2}.

2. 2.

SoLUTION

OF

SysTEM

OF

LINEAR EQUATIONS By MATRICES

Let

( 1)

be a system of linear equations involving

m equations wif~11.

unknowns.
Setting

.[~1]

B=

.
bn
I

(1) becomes

.I

102

AX

=.

(1I

To solve this equation we have two methods:

1) By inverse matrix,
2) By GAUss Method (by elementary.row operations)
1. Solution by

lnv~rse m~tfix:

Square case ( ni :.n ) : .

If
by

A-l

A-l

exists multiplying both sides of

(l'}

from left

we have

a .unique solution .matr.ix (column

matrix)~

A-l can be determined by any method given inl. 2 D, and.:

th~

the method of solution has

hod to find
If

same name as that

~f ~he

used met-

A-1 .

A is .n-on inv-ertible, i.e. if IAI

=.

.0, the syste.m has

no solutio_n or has a solution matrix involving one parameter at

least depending on the degree of freedom.
Example. Solve the square system
2y -

+ y

2x -

+Z

y t

:::.

= 6

z :.

Solution. Writing the system in matrjcal

.. and

findi~g

the inverse of

A by any method

w~

f~rm:

have

.1 03

9
Rectan~ular

x:l,

y;:::.2,.z:.3.

case: (m, n

are arbitrary)

The system is .
.AX = B

where

A, X, B are matrices of size

Since the

s~uare

r'espe~tively.

mxn, nxl, mxl

case has been discussed, there remain the

cases:
1) m>n: One takes

equations, obtaining a square system

, and solves.
2)

.. .

m<n: One chooses suitable

n-~ unkno~ns

as parameters

, t n-m and solves.

It is advisable in all cases to solve any system of linear

equa~ions

by GAUSS Method (by elementary operations):

2. Solution by GAUSS Method:

This methdd is applicable successfully

~o

any rectangular

system
AX

=B

( 1)

which we represent, conveniently, by the matrix

: [Ais]mx(n-tl)
called the augmentel matrix of

A.

. The method consists of reducing by row operations the.matrix

[A/B] to.an echelon form [A'!B'] which represents a system (1') of
linear equations equivalent to the system (1). Since
echelonform,the number of nqn zero

t~rms

[A'i B'} is an

decrease as we go down

104

until the last equation, and solving (1') is much more simpler.
Examele 1. Solve the system

Solution.
2 -1

-1 . 1

+ 2y,- z

:::

+z

:.

~X

y t z

( 1)

=3

1:J)(-l ,~"'# r: -~ -: 1:J-r'l

:

1 3"'

2 -1

: -:

~ ~ Jts)~ l:

2 -1 ! 2.,]
-1

0-71-21

This echelon matrix represents the system

2y - z :. 2

X t

(1

-y + 2z =- .4

I )

-7z =- -21
which,is equivalent to ( 1 ) since the row operations (interchanging
two equations, addin_g two equations, miltiplying an equation by a
n6n

..

~e~o scala~)

do. not. alter the solution of the system.

Starti-ng from bottom and going upward we have

\

-7z = -21 ~ z = 3
-y + 2.3
X

2.2

-4

~ y

=2

X.

::.

=1

s =[ 1'

2, 3]

x 1 - 2 ( -2 t t) - 2t - 4 :. 1

s ; [1 ' -:2

-+ t'

o:::

t' 4 ]

Remark. If the la.st row in an echelon matrix consists of

zero elements only, this" row is to be discarded. If the lastro~

is-

o .
.the system is fnconsistent.

ol
_, a

(a. f 0)

'.

. 'l.

'it

106

Aft~r

discarding "zero rows, if the remaining is consisteflt,

the given system is consistent.. In the consistency

case starting
.,
from the bottom and gof~g upward considering the equation corresponding to each row one can find the unknowns successively
.(xn' xn-l, ... ) , some of which are taken as parameter when possible.
o~

When the echelon form

the system is inconsistent

the system is, for instance

0 -3

0 o

I
I
I
I
I
I
I
I

:J

(no solution)
,. for instance,

If' the echelon fl"

4 .I

-3

u.

:J

we have consistency. Then

2x 1 + x 2 - 3x 3
~

2x 1
X1

x 2 - 3x 3 ~ -11

:.

S, '

:. t ~

X3

[s, -

S "

4.3 :

11 . .: 2

When the echelon f?~m is

-3
0

-...
/

I
I

.I

I
I

0 1-1

I
I

I
I
I
I

0
0

X2

t.

: . - 1 1 - 25. ~ 3 t

3 t, t, 3]

107

the system is inconsistent

If .it is

,.
1 . - 3
2

I
.I

I
I

0
0

I
I
I

'

we have consistency. Then . Then

2x 2 = 3 =) x2
3
xl -. 3 - 2

If

3/2

4 ~ xl = 17/2

:.

[17 /2.

S =

"

::.

3/2]

the echelon form is

-3
0

.4
I

I
I
I

0
0
0

there is ,c.onsisten.cy. Then

I

x1 - 3x 2 ::: 4

~.

x 2 :. t

x1

=4

+ 3t

= [4 + 3t;t]

Proof' of the va 1 i dity of the sweepout prcess (by row

operations) f~r find~ng the inverse of a matrix:
Let
consider a

A be a matrix of order
squa~esystem

n. To find its in~efse

A- 1

bf
AX -::: Y

( 1)

- 108

of linear equations. If

A.

is invertible the unique solution is

giv.en by
(2)

(
Introducing the identity matrix
(2) can be written

,I_ .. of order

a~

. where the solution (2

1
.)

AX= IY

( 1 1.

IX= ~A-lY

(2I

can be obtained from (1' 1 ) by row oper-at'ions.

This means that the matrii

A I: I

r~pre~enting

row operations to the matrix Ii'A-l

A..: I
I

-v

verse of
If
form

I!

(1

is reduced by

represe"'ting (2

):

1: A-l
-'

So starting with the matrix A:' I

reduce it.to

n, (.1) . and

elementary row operations

IiA~t whe~e the second side ~f the_l~tter is the in-

A.
A

is not i nverti bl e

A! I

cannot be reduced to the

A . where the. reft hand side .i.s the identity matr1x .

..:.

.... '

109

EXERCISES <2.2)
31. Given that

rna tri X

- -solve i t for the

32. Solve
2
x 2 +5/-z

a)

2 2
2x 2 -+3y -z
x 2 -/+2z

\-J.

b)

:.

-i

;.

-t z

+ 2y

;
'

y + z

2 :.. 6

-y

:.

3
z

-1
;.

33. Solve
'-

4x 4

:.

2x 1 +' x 2. t 4x 3 +

x4

:.

xl + 2x' 2 t 3x 3

3x 1

4x 2 + Xf~+''5x 4

2x 1

3x 2 + 5x 3 + 2x 4 :. 0

:. 6

34. Discuss the solution of

2y + z . .:: -9

3x
X

2x-

+ y t z___,::.
+ y - 5z .:::

-7

35. Solve

cos 2>..

2
z sin

c .::

sin 2 >..

z tan

tan

co.s 2A + y cos 2 B

2
z cos c

2
sin 2 A + y sin B

tan A

.;.

a) in the general case,

b) when

:.

X1

y tan B

.::.

>..

.,-

A -t- .8-tC :. Tr/2

'

''

110

36. Discuss the solution-of the system:

y+ z :. 1 '

2x-2y~z

:. 0,

4x-+3y-+5z

=i

37. Discuss the solution by the use of augrrented matrix:

-3
0

=
[:]
[:]
:J

38. Find right in~erses, if any, of the fo,ll owing matric-es:

a)

[ ~]

b)

[3, 7]
matrices given in Exercise 38.
(adf F 0)

and evaluate
2
2 .
2
. 2
.
(a -a)x+(b -b)yt(c -c)z-(~ -p)
42. Find the inverse of:

a)

43. .Find

~
~

I-

!'
l

--il
2
2
x,

_j

Y~.

3
-2

b)

i f any:

1J4 .

Fl
'-X

;~ =

14

-16

'

~
1:]

-1

:J

111

44. If

A2

x2 ['] ~ rx I]~

point (x,

y1

~h~

to

we say

point

that~the
yj.

(x,

n1atr.ix. A maps the

Fi11d

i\'w~ich

to ( 1 , 0) , and ( -1 , 1 ) to .( 2, - 5)

ANSWERS TO EVEN
32. a)

(2,

.o.

:tl]

NUM~ERED

EXERCISES

(four solutions), b) (2, 1, -3]

3 4 [- 2 ' 2 ' 1]
3~.

[1-k, l-2k, 2k]

38. a)

40.

42. a)

44.

-z1

1 s + 1

if

-a<r

0'

Ga10
-5

t - if

~1

1
...;1

[: _:]

']

3 t

;-if

"2'

3
1
- "2' s ~. if

.b)

.
2d

be-cd.

---aar
e

-"aT
. 1

-q
-4

Lll2

b)i-; ~vi

'

8
-1

J]

[1 '_5

,.J

maps (2, 3)

112 ..
\'

A SUMMARY

2. 1.

Operations with .matrices:

A+B - [aij}+ [bij]=(aij+bij}:. BiA
cA

= c [a i ~
-

= [c a i jl

= Cmxp

Amxn 8 nxp

::: .[cij]

Definitions:
(trarispose of the

mat~ix

A)

(adjoint of A)
( A- 1 i s the i n.v e r s e of A)

!I

,,j!

= AdjA/IAI

A formula for A-l is A-l

where

IAI

det A

Echelon matrix: Is a matrix [a ij ] mxn

a .i j = 0.

,,

aitl
An example

=h

(j

...

such that
k)

= 0 (j:::l, . . .

k+l at least) . for any i

echelon matrix:

~f

5 . 0 -7

6 -3

0 -2

12

0: 0

0.

-4

o o o _o o o o o o
2. 2.

Solution of a system of linear equans: -AX:B

.a) Square case:
By in~erse matrix:
b) General case:
By GAUSS Method:

X:A~ 1 s' when

A is invertible

Obtaining an echelon form of ihe

augmeted matdx lA. Bl and solving
step by step from ottom to top.

11 3

MISCELLANEOUS EXERCISE
46. Let A

-- J

~ :: _:
1

B :.

Find

[:]
2'
b) A3 + A ,. . 6A - 17I3

a) AB
47. Prove

a)
/

T ~ ~]
=

b)

B su'ch that. B-lAB

49. List all

echelon matrices.

50. For an

nxn

matrix

-=

0 1

48. Find a matrix

2x2

rr
1

0
0

= lAin

Sl. Obtain an echelon form of:

0

-2

i .4

-2

-2

-4

52 .. If

ad

be

ad

cd
-ac -c 2

:.

then

-ab -be -1
a 2 ac

~2

-b2

-bd

-be -cd

'ab

ad

bd

.1

ad
:.

~ n~n+ 1 )]
.1 .

is diagonal, where

A, prove IAI ladj AI

bd

-ac

-be

-ab -b2
cd d2

a2
-c 2

ab
-cd

.:be -bd

ac

ad

11'4

53. If

....

Al,

, An . are invertible matrices of the same order,

b) prov_e

-1
Al

A ) -1 :. A-1
n
n
(An)-1 = (A-1 )n

a) prove . (Al

...

54. Find the inverses of:

a)

:1

55, Show that the prod~ct of two u~per (lower) square triangular
matrices is an upper (lower) triangular matrix.
56. Prove that ~he inverse of a non singular d~a9onal matri~ is a
non s i n9 u1a r d i a 9on a 1 . r,; . r'i x .

l,.

57. Evaluate

..

a~
)10

a/;bJ

c
7IJ
where
58. If

D :. a + d + 2/ad

(A-l)n

be

=~

-J

-2

>O

[fl(9)]n

.:

-1

v=
5
-1

59. Prove

u-2, V-2 where

A-n , then evaluate

is denoted by
I

-5

fl(n9), where
2

fl(9) =

cos e
cos9 sin9
sin 29

-sin 29

'j

sin 9
cos 29 . -sin9 cos9
cos 29 .
sin 29

11 5

60. Find all

a). with

matrices which commute

2x2

11 1]
~

b) withevery 2x2 square

,
o~

61. For which values

m~trix.

t, will the system

4x

-t

y - 4z = 0

2x

ty

+ z = o.

(t-l)x -y f 2z =- 0

be consistent?
. 62. So 1ve the sys teni.S:
a)

2x' - 3y

b)

X t

= 0,

y - Z

=.

0,

xt3z:.o~

2z

=0

,;0,

3y

X -

y+ Z

2z =- 0

63. Show that th~ solution of

a x t a 2y
1

-t

b2y

b x +
1

a 3z

+ b z :. 0
3

is
64. SolVe:
1: 2

-3

-3

.:1

-7 15

65. Show that the following system is inconsistent:

2x 2 - 3x 3 t 4x 5 = 2

xl '+ 2x 2 ,..x3 + X4 1. 7x 5 ::.'3

4x 3 - 5x 4 +.8x 5 ::. 8
2x 1 t 4x 2
3x 1 + 6x 2 - .5x 3 -i 6x'4 t 15x 5 :: 11
x1

116

ANSWERS TO EVEN NUMBERED EXERCISES

4 6. a)

48. B

(17

1] T ;

[ ~]

where

.: l

b)

w
w2

54. a)

t:

58~

u-1=-i

1 -'

-3

-1

a>

62. a)

= -13

-31

b)

[:

,,

60.

[-3t. -2t' t]

64. [-5k+7.

0,

-:]

1
0

2'

'

k]

-1~
-9

-1

5j

.,.1

.:

w;"l.

-1

in

-:3

-1

3 .
=1

,I

['

\'1

'

v-2_

[ 15
9

-15

-1

...,

.,.3

-3

-87

-229

87

J
b)

25

[o o o]

25 . 14

117

CHAPTER 3
ANALYTIC GEOMETRIC IN iR3
3, l, VECTORS
A I. DEE IN IT IONS

In Physical science are quantities such as velocity and

.acceleration that 9re determined not.only by

ma~nitude

(length)

but also by difection and sense. Such quantities have the common
name vector. Vectors are very

po~1erful

tools in treating analytic

.I

geometry.
The

~et

of all parallel lines in space

A~

6, and each line

thi~

defi~e

~irect1un,

of the set has

indicated direction and

defines two senses, one from

A to B, the other from B to A

. A.

i i ne segment [AB] with .

end pojnts

and

has a direction (that of its

su~port

line),

and a length lAB!.

A line segment [AB] oriented from one end to the other,

sayfrom A to B,

~s

called a line

vect~r.

written

initial point or the point of applicatiqn and

AB.

A is the

the extremity

of

AB.

line AB
(has a direction)

1 i nE! segment (AB] '

(has' a direction .ancl
a 1e~gth)

vector AB
(has adirection
l.ength and sense)

118

A vector having

coinci~ent

end points or having a

+

length is the zero vector

~enoted

by

PP

zer~

or

0 -or

sim~ly

with

indefinite direction and sense.

I

Vectors may as well be denoted by single letters with an

+

+
a, +u;. .e.

arrow on top:

Equality:
Two vectors

AB, CU

having to same direction,- the same

sense and the same length are considered equal, writteh

If

A'S:.

CD,

then .ABCD

An : CD.

is a parallelogram (which may be

degenerate one) 'a'nd one of these

ve~tors ~~n be cibtained from the

oth;!r by a translation (under

- which direction, sense andlength are preserved).

In the figure,
fo 11 o~s that

Ks

ABC.D

and

CDFE

='CD ~ EF. Note that

are parallelograms. It
ABFE

is a degenerate

parallelogram.
From this definition,

vector~

can always be drawn to. have.

the same initial, point.:

Two vectors are unequal if they differ from each other
eithe-r in. direction or in sense or in length.
A vector with fixed _initial point is a bound vector (or a
position vector), one ~estricted to lie on a fi~ed line is a
sliding vector and on~ with no such restrictions are free .vectors.

A free vector remains invariant under any tr~nslation,

while a sliding vector remains invari~~t und~r a translatio~
the direction

of~the

vector.

in

119

'I

:I

'

Line vectors can be expressed analytically bi the use of

coordinate systems. Thi~ representaiion ~ill be called a c~ordinate
vector. Coordinate vectors 9re more convenient in all _analytical
treatments than'the line vectors. We introduce l>elow rectangular
coordinate system to define coordinate

~ectors.

Rectangu'lar coordinate systems:

Consi_der in 3-space three .mutually perpendicular axes Ox,
Oy, Oz

with a common origin

coo~di~ate system,,written

O, called a r-ectangular (or cart,esian)

.Oxyz. A 3-space provided with such a

system will- be called an analytic 3-space.

.\ '

In the followirig two figures two.distinct

coo~dinate

systems

are shown:

A-negative

A positive system
(A right hand

~ystem)

(A left hand

On a poshive system an observer, standing on

at

x~axis

~lon~

positive

z-axi~

syste~
sy~tem)

Oxy

plane

observes that the angle from positive

to positive y-axis is positive

In a

(counterclb~kwise).

negative system the same observer observes that the same

angl~

negative (clockwise)
The positiveness .and
translati~n

negative~ess
I

is invariant

~nder

'

and rotation of the coordinate axes.

In this book we will always use positive

syste~s.

anl

is

120

In an

Oxyz

the axes taken two at a time determine

syste~

three planes called

coordi~ate

planes: xy-plane, yz-plane,

i
.

xz~plane.

These planes separates fr6m

R3

eight regions,

1I

called

octants, .numbered I, II, ... ,VIII.

The one containing the positive
parts of the axes is taken as
the I. octant.
..

analytic

Referring to the Figure, let. in

3-sp~ce,

be any point (usually taken in the I. octant)

and let the projections of

P on

coordinates axes and planes be

X, Y, Z and
points arid

X', Y', Z'. These

0, P are vertices
I
I

of a rectangular parallelepiped.
The abscissas.
of

X, Y,

~x,

dfrected

"

---------- _v
z/

y, z

on respective axes

are called the rectangular_ (or cartesian) coo rd ina tes of

qne writes

/"

and

P(x, y, z). It is seen that these coorc;tinates are

di~tanctes

A point

of

P(x, y, z)

from the respective coordinate planes.

~

in usually

skeiched as shown in ~he Fig~re.

The set
{(x, y, z):

x~o. y~o.

z~O}

represents tlearly the I. octant.

Though

~re

ordinate, and z

not in use, we call

the cote of

P.

x the abcissa, y the

1 2]

A bound (position) vector

r = OP

point at the origin and extremity at

the point

~nd

p.,

r, (x, y, z)
+

P(x, y, z), defines uniquely

conversely, and we make

+

OP, P,

no difference between the symbols

+

having its initial

and write

= OP = P = P

(x, y, z)
o .....

which is a coordinate vector.

Any free vector can

"

b~

. . . . . . '~,

:i: ,~, r ~"

I

-.

. ',..._r:L~

-----:i-~-- _.::;vz/

expressed ilS'\a coordinate vector.

Since

(x, y, z)

is an ordered triple,

as an element of the ca:rtesian product

We

r~present

the position

vec~or

[ : ] = [x,

y,

~t ca~

be considered

tRX{RxiR = IR
(x, y,. z)

as

th~

column

matrix

z] T

Length of a vector:

loP!

Referrin~ to above Figure we have as the length

the vector

OP:

I-OPI 2 == IOPI 2 = I.OZ'I 2 t IZ'PI 2

= (IXI 2 + IYI 2 ) + IZI2 I p I = I OP I =- Vx 2 .. i . z 2 (~0)

X2-ty2.fZ2

. Example. Find the lengths of.

A:. (1, -2, '2),

8:.(2,3,6),

Solution.

I AI = vlr-=2----'-t'-(---2)-=2-+.:-2~2
I BI ::

./'4 + 9 i

.36

=7

:.

Vl

-t 4 t 4 .: 3

:.

(~

9 '

~)

of

122

A vector haying length equal to 1 is called a uriit vector.

In the above

e~ample,

C is seen to be a unit vector.

B, ALGEBRA.OF VECTORS
l.. Addition:

The sum

;+b

of two vec.tors . ; . and

b, in this order, i",s

the vector whose initial point is that of the first vector and

th~

extremity is that of the second when the latter is translated tb

I

'i

'I
I:

have its. initial point at the

I~

e~tremity

of the first vector:

1/''
./

1!

!. iI

;:

The following

II

l'l

!,;

,;.:

-;

~nd

commutati~elaw

a +- b

jl

):
i
I,

the

figur~

+a

parallelogram law:

tl

\';

1.':
!,,.

lhe ~ssociative law

+

(a + b)+c

+ +
+
~ a~(b't c)

is the

Because of this law the

by

(a+b)ic

or by

ar(btc).

su~

a4b1c has a

mean~~g

as defined

123

2 . Multiplication by scalars
+

+++

~+

.......

Denoting the sums a+a, a+a-la, .. by 2a, 3a,

+ -+
' -+
. -+
defining la,, Oa as a and 0, the vector na (ndN)
denote a vector having the same direction and sense as
'Z

For 'any
(1}

will.
-+

and

-~

n times that of
1

and

a:

.. .,
cf.P

1&R, we define

-+

as a vector of length

1a

. -+

.parallel to

according as

and agreeing.or
1>0

or

;:_,

di~agreeing

....
a

in sense with

~.d'!:

---')>

<

-+

-+

In particular for 1: -1; we have the vector. (-1)~

....
....
which is opposite to. a, called the additive inverse of a,
....

since

-a

-+

a+{-a) : 0.
-+

1a
calle~

and

-+

'

have parallel supports, and are

collinear vectors.
The difference

a- b

is by definition the sum

a+(-b):

'

(1) Two coplanar.fines {lying on the same plane) having no common

point are called parallel.

Bu~

iQ this 'book parallellism of

ltnes is defined to 1nclude also the coincidence of lines.

124

f~ojectioni
~

Let
+

OP

be a vector and

along it. If

~ntersections

of

on

be an axis with a unit vector

is the projection of

on

with ' the plane through

Ot

OP' is the vector projection

coordinate of
OP

P'

Ot

-~

Ot), then
+

of vectors:

(~ector

Ot

{which is the.

perpendicular to
compoment), and the

P' is the scalar projections (_:;_calar _compoment) of

Ot. (see Fig.)

Now consider

.!!

second

+'

vector

and the vector

....
+
uR = OP. + OQ where OQ ::. PR.
+

OQ

I
I

I
I

....

kr

Then the projection of OR

....
+
on ot is OR' ::. OP' t P'.f
-)-

,.
f.,
I'

which is the sum of

-c

R'

proj~cti6o

(compoments).
Obviously the scalar compo,nent of the sum
of

seal~

components of

OP~

~nit

is the -sum

OQ.

Now, taking coordinate axes

respective

OR

OX, Oy, Oz

instead of Ot with

vectors
i . ::: ( 1 , 0, 0) , j :. (0, 1 , 0) , k :. .( 0, 0, 1 )

along positive parts, and denoting the projections of.

l,i

X, X,

on the axes and xy-plane by

....
+
+
OP = QP I T P'P
+

: :. ox
Since

ox

...

OY

t- 0-tY

OP' and

;,

'll

Xi,

::e

1re have

Nx;#,=r~

oz = P'P

~...-...:---t-__..;..Y__,. 4
I

+ yj + zk

where

F'

+ oz

Then

I'

z and

P(x, y, z)

yj,

Zk

are vector

...

I I

'

---------~I

(/"

125

components, while

x, y, z

(coordinates of

P)

are scalar com-

ponents of

OP.

Let
0p1 :

X1i

+ y 1j

OP 2 :. x 2 i

-1-

Z1k

[x l . y 1

Z 1] T

y 2j + z 2 k = [x 2 y 2 z 2]

Then we have

by properties of projections.
Also
+

OP 1 -'-0P 2 .:: (x 1 -x 2 )i + (y 1 -y 2 )j

Accordingly any free vector

A(a 1 a2 a 3 )
vector
+

since

to

OA t AB
Wh~n

B(b 1 , b2 , bj)

= DB

or

AS

AB

(z 1 -z 2 )k

extending from the point

can be written as the,position

OB - OA.

a vector is mu1tiplied by a s~ala~its ~11 components

being multiplied by the same scalar, we have

+

).P

= >.(x,

y, z) :. (>.x, >.y, >.z)

by the prnperties of projections.

Observe

~nalogy

between matrices and vectors in the

. operation of addition and multiplicationby scahrs:

[al a2 a3] (bl b2 biJ .: (albl
>. (~1 a2 a3]

(>.al

a2b2
>.a2

a3b3} ~

>.a3]

126

.Generalizing, we have
+

c 1 P1
whi~h

-t

c 2 P2 ,. ..

-t

cnPn

is called a linear combination of

Vectors.

Direction-angles,-cosine ~nd-numbers~
+

a,

~.

r = OP

Let

be a position vector with length i. The angles

+

that are made by

with

positiv~

direction angles, their cosines cosa,

cosin~.and

cos~,

sides of axes are

are direction

cosy

proportional. to

nu~bers

direction cosines are direction numbers

+

of

OP.

Since scalar projections of

P(x, y., z)
x~

:;

on the .axes n re

y, z. one has
'

cosa =

'X

~nd consequently

COSf3

x, y, z

cos'(

and for k

to;

z
r
i<x, ky, kz

are direc-.

tionnumbers.
3. Multiplication of Vectors:
.There are two kinds of multiplication lor vecto~s. scalar
and v_ector multi-plications, the results of which being scalar and
vector:
. a) Scalar product of two vectors:
A =- (a 1 , a 2 , a 3 ),

Let

B =- (b 1 , b2 , b3) be two vectors:.

Then

A.B =-!AliBI
is cailed the scalar product of
b.etween them
/

(0

9 ~ rr)

cos8

A and

B, where

is the angle

127

_The.product is also called dot product or inner

and also denoted by

AlB,

p~oduct,:

(A, B), <A, B> or A, B.

The scalar product

A.~

vanishes when

~ertainly

A~

orB::. 0. For non zero vectors, the product is positive, zero or

hegative acco~ding as

is an~cute, right or obtuse.~ngle.

Geometr'i c i nterpreta ti ons:

+

1. If _OB' ' is the projedion of

+

OB

on

OA.OB : OA.OB'
a2 . =- b 2tc 2-2bc co sa

2. The cosine 1aw

for a triangle ABC can be. ex pres s'ed

in the form
+

IBCI2 =- IABI2

2
IAC I . - 2AB.AC

'
3. Two 'non zero vectors are perpendicular
(orthogonal) if

and.only if their dot product is zero:

A,.LB <=;> 9 = 1f/2

A.B = 0

4. The dot product of two vectors

~ith

known lengths (of

variable directions) is maximum. or minimum when they

are parallel
physical

in the same or opposi-te senses.

i.nt~rpreta ti

on:

If a particle is moving on a line in 'the di_rection of a

vec.tor

R, under a force
+

the effective force 'Fe

vector of
of motion:

F on

F, then

~s

the projection

R in the direction

128

Properties:
1. A.B:. B.A

1aw)

(com.

2. (>.A).B =- A.(AB)= >.(A.B)

3 . A. ( B+C) =- A. B + A. C

(>.e:R)

(dis t . 1 a w) _

Prouf:
The first two properties are direct consequences of the
definition.
I(

To prove the distributive law

,A. ( Bt C) = A. B
+

8 _________----;

A; C

I
I

.+

consider 'BR ::.'OC

projections

(See Fig.), and

B', R'

of

B, R on

OA. Then
A.(B+C) = A.R

=A.R'

(Geom. interp. 1 )

= A. (B'+C:')
= A.'B + A.C' _ (collineari.ty of vectors)
I

:.

A.B + A.C

(Geom. interp. 1 )

Now we de.rive the analytic expression

A.B :. aibJ + a 2 b~
for

( a 1 , a 2 '. a 3 ) ,
Expanding
A.B

by

dis~ributive la~

a 3 b3

B = (bp b2 , b3 ).

= (a 1 ita 2B+a 3 k).(b 1 i+b 2jtb 3 k)

we get nihe terms; six of which are zero by.

properties
i.j:.O,
for orthogonal vectors

l;.~ .

t:

rr~-

1 bp a 2b2 , a 3 b3
for unit vectors

&

j.k:O,
i, j, k,

k.i:O
~nd

by the pr-operties.
i, j 1 k

the remaining ter~s ar~

i.i:. 1,

j.j

= 1,

k.k.:. 1

129

Then

If.t~e v~ctors

-- and column vectors

are written as row

then

Example 1. Given the vectors

A ::. { 2, -3 , 4) ,

B ::. ( 5, 6, -1)

find
a) scalar project1on of

B on

b) vector projection of

a on A

c) angle between

A and

8'

Solution.
a) A.B = IAI.IBI cos 8
scalar proj.
~1

2. 5

{-3)6 +. 4'( -1 )
+

:::!)

::./29 lOBI cos

-12

= /449+16

lOBi

S g

lOBI cos
~b)

OB'

-. -

12

= -12//29

TAT~

(A/I AI

is the unit vector in the

direction and . sense of

::.

1 2.

72"'9'

( 2' -3, 4)

:.

A)

36
48)
(- 24 ; zg
-zg
'2"9'

130

-12

A.B.
IAII BI

-12

9 = arccos

1179167

Example 2.' Given vectors

A
determine

= (2a,

-4, a+ 1 ) ,

( 3 , a -1 , 2)

;::.

such that

aiR

a) Aj_B

b) A// B

Solution.

'b) A//B

+ 2 ( a.; 1 ) :. 0

6a-4

a+l

(no solution)

z-

tors

Example 3. uiv
A ::. ( 1 , 6,

wri

if possible,

a ::. - 3/2

C:.(9,1,7)
C

combination of

A and

B.

So 1u t i.on.
::. tA
~

of wh1

= \1 ,

-1- S B ~ ( 9, 1 , 7)

t < S' = 9 ,

6t t 9s ; 1

the first two give

t ::. 80/3,

A and

9, 7)

4 t t 7 s ::. 7

not satis'ty the thi-rd one. Hence

linear contination of

6 ; 4) + s ( 1 ,

s = -53j3. But these do

C .cannot be expressed as a

B.

b) Vector product of twrr vectors:

The vector product of two vectors

A and

B, in this order,

is the vector
+

Ax B.:. n IAIIBisin 9
where

is the angle between them

vector such that

A, B, ;

(0~9~n),

and

is the unit

form a positive system ( ; .l.A, B).

131

AxB

In case of collineari1y of
B {9:0
-+

or

B:n)

denote~

prod~ct

0,

and

als~ call~d

is

cross product or outer

produc~

also by the symbols

Since
and

AxB

is uncertain.
Vector

and

one has

A and

9~0,

IAIIBisi_n

then

AxB

{for non collinear vectors)

have the. same sense.

va~ishes when

AxB

A:O

or

8:0

or when

A and

B are

collinear.
It follows from the definition
1)

AxB

~hat

is perpendicular to both

2) AxB

is oriented so

th~t

A and

A, B, AxB

{direction)

is a positive

system, not necessarily rectangular (sense)

.3)

IAxBI~IAIIBisin

{length).

For non zero vectors, the vector product

if and only i f

the projection of
+

-+

2.

-r
OB~

vanishes

A//8.

Corollary. If a. is a

1 OA

AxB

,X.

-+

OB

on a. is

DB =- OA
-7

plan~

DA x OB

perpendicular to

-+

DA

and if

-+

OBI' then

OBI
js

obt~ined

from

DB'

r.t:,

-+

.H!

OBI '.

about 0 through 90 a~d magnifying it

Proof.
1) The two products have obviously

the same direction and sense {See Fig.).

To show the equality of lengths, we
IAxBI

= IAIIBisin'B =

~ave

IAIIB'.I

=- IAIIB'I'sin ~ = IAxB'

8'

132

B";; AXBJ..A, B

OB".LOB',

= IA I

a.

Properti.es.
(~nticommutative la1~)

1. BxA ;. -AxB

2. (AA)xB : Ax(AB)

= AxB

3. Ax(BtC)

A(AxB)'

+ Axe

(AR)

(distributive law)

Proof.
The first two poperties are direct consequences of the
definition.
To prove the th'

use the above Corollary.

rl.
~

. OC and
pro

'-+

OR{~

~ction

OB+OC)
+

of

OA

magni_

ing by

....

OA (the

::.

o(

p~rallel~

I AI

pnd

c'

we get

the pa.ral t e 1 ogram. OB"R"C".

,,

are projections of. OB,

is the ,

about

by TI/2

o~

a perpendicular to

Rotating the
OB'k

and

Inthe figure,

R'

Now by the Corollary we have

....
....
....
....
....
....
OAx{OB + OC) ;. OA x OR ~ OR"

= O+B"

'

+ O+C"

....

.... . ....
....
: OA x OB .I+ OA x OC. s
.T~e

analytic

expres~ion

for

AxB = (a 1 i' i, a 2 j + a 3 k)x{b 1 i t b2 j + b3 k)

is obtained by expanding it by the use of distributive law .
Expansion gives nine

term~

three of which are zero by the relations

I

ixi ::. 0,

jxj :. 0,

kxk..=. 0 1

133

and ,the .sum of the remaining six terms will be.

AxB

::.

(a2b3

a 3 b2 )i + (a 3 bl

alb3)j + (alb2

a 2 b1 ) k

or

...

AxB

(a2b3

a3b2, a3bl-~b3, alb2-a2bl)

since
jxk

:.

-kxj = i

Observe that

kxi

:;:

-i xk ::. j.

i xj ::. -jxi

= k.

AxB ' is equal to the symbolic determinant

k

al a2

bl

b3

b2

(::. AxB)

Geometric interpretations: .
+

A = (a 1) a 2 a 3 ),
al
a2
a3
-bl -- -b2 -:::- ru3.
-

1. For non zero vectors

+

A//B

* =0

Axu

B :. ( bp b 21 b3 ) :.

2. IAxBI =-(Area of the para-llelogram DARB with adjacent sides

IDAI. IDBI):.IOARBI2
. where. the index "2"

rep~esent's

the dimension. of the measure.

Proof.
The first one is a direct
. For the seco'nd one, we

consequence of the definition.

8

h.a~e

IAxBI ::. lniiAIIBISin 9 =-IAIJB[jin 9

h

:observe that

th~

= IDARBI 2

4~;'Jb

distance between the end point of one of

the vectors to the second-vector can be obtained by the use of

cross-product.
,Physical interpretation.
Gi~en

a sliding force vectoi

F and.a fixed point

134

{Se~ Fi~.), then-the vector ~roduct

+

OP x F

is the moment vectnr of

respect to the point

with

in~

0, which is

d:pendent of the -point of applicatio!1

P

on

.JI. .,

+"

0.

and

1. OPxF

is perpendicular to the plane determined by

Jl.

(fixed direction),
+

2. Sense of
+

is u:nalter.ed,

3 1oPx F 1 =- 1~ 1J oP~ s...i n.Jl

::.

1F lh .

A_:. (t, 4, t 2 ),

Example l. Given the vectors

a) find

tR

such that

B:. {-2, 4t, 1),

AxBHyz-plane,
for ',t

b) find the area IOABiz

determined in (a)

Solution. '

a)AxB:.

-t

-2

4t

(Ax~):i =-

b)For

t::.l,

AxB:Li
3
4-4t = 0

Exam p1 e 2 .

A:.(l,4,1),
1

IOABI2

-::. {4-4t 3 )i - {2t 2 H)j + {4t 2 +8)k,

=>

AxB//yz-pl ane
==>

= 7 IAx&l :.

z l:-

3j

Gi v en the poi n t

t ... l
B:.(~2,4,l).Then

+ 12kl

= 7 ;ill=- 7

A( 1 , 9 , 7 ) ,

. .

/i?.

B( g, 4 , 2 ) ,

C ( 0, 1 , 3) I
a) find the area IABCI 2
~

b) find a vector

V//ABC

and perpendicular to z-axis, if any.

135

Solution.
a)

I ABC I 2

~ I AB

:;.

AB :. B-A

x AC I
(8, -5, -5)
\

AC ..,. . C-A :. (-1' -8, -4)

+

ABxAC
1Asc 12

=-

(-20, 37, -69)

2
~ /2o + 37 2 + 69 2
~

-)

b) V :::. A ( 8 , - 5 , ..: 5.) t ll ( 1 , 8 , 4-)

:

(8A.j.ll, -5At8ll, -5A+4Ji )

V.k:: 0

:9 -5Ai4ll:: 0

ll:: 5A/4

V ::: A(- , 5, 0) .

Triple products.
cross product of two vectors i_s multiplied scalar.ly

~fa

or yectorally by a third vector, one

obt~ins

what one caJls a

triple product.
If

A, B, C are three vectors, then

Cx(BxA), B.(CxA)

(AxB).C,

(AxB)xC,

are examples_pf triple products.

AxB_.C, being a scalar and involving two operat-ions_, is

call_ed a triple
ior

(AxB)xC
Now

~e

sc~lar

product or a mixed product, while the vec-

is called a triple vector

~roduct.

obtain analytic expressions for a mixed product and

triple vector product:

i. Mixed product:
Let
A:. (a 1 , a 2 , a 3 ),B ::.(b 1 , b' 2 , b3 ), C::. (c 1 , c 2 , c 3 )
be any three vectors. The11
al
'AxB.C

:.

a2

a3

bl

b2. b3

cl

c2

c3

136
Indeed,
j

AxB.C

al a2 a3
bl b2 b3

::.

( c 1 c2 , c3)

= (Cll c12 cl3).(c,. c2' c3) .

where

i j . k. Then

are cofactors of

c,, cl2' cl3

Ax.B .C :.Cll cl -t c12 c2 + c 13 c 3

:::.

a2

a .
3

_a2

a1
a3 :. bl

6'1.

b3

b2

b3

c,

c"l.

cl

c2

al
bl

c3

I'

Corollary. In a mixe~ produci the interchange of the

and "

"xJ

does .not alter the mixed product:

Ax.B.C ::. A.BxC
!'"roof.
I

A.BxC

:..

(commutativity of \dot product)

BxC.A

:.

We denote

bl

b2

b3

c,

c2

CB

al

a2

a3

AxB. C or

A.BxC

al

a2

a3

b,

b'l.

6.3 = AxB.C

c,

c'l..

c.3

by the symbol

(A 's C)

so

that we ha v.e

(A B C)

Since

::.

(A B C)

AxB.C

::.

A.BxC =

al

a2

a3

bl

b2

b3

c,

c2

c3

is a determinant it obeys the properties:

l37

1. (aA, 13~. '(C), = ti13V(A B C)

2. (A 1 -tA B C)
2

(Al B C)+(A 2 B C)

:.

3. (A A B). ::. 0

c B) = -:(A B C)

4. (A

5.:- (A B C) ::. (BC A).::. (CAB)

The last p~operty st~tes that

is un~ltered under

(A B C)

,cA~
e,.-...:, c.

a circular permutatiom of the letters:

Geometric interpretations

If A(a 1 , a 2 , a 3 ), B(b 1 , b2 , b ), C(c:l' c ,c }, then

3
2
3

/(A B

C)/ ::.(volume

of the parallelepiped

built on.IOAI, lOBI, lOCI)

::. 6IOABCj 3

AxB

Proof.
.
(ABC)-: AxB.C
= IAxBI~
tl

= jOAC'BI2

h ::. IOAC'RB'CA'BI3

: 6jOABCI 3
It'-follows ti1at

(A B C)

.aa

0 ~ OA.

are coplanar.
.

ii. Triple Vector product:

The triple vector product
perpendicular to

i'

(A~B)xC, being a veLtor

AxB, is parallel to the plane OAB,and hence can

be expressed as a lin~ar combination of A and

(AxB)xC = (A.C)B - (B.C)A
Indeed, 1et

B:

138

We have

I"

S iiarly,
W

~.

.,

(A. ,.

',

~ (A. C) p-3

w -:
:: (A.- )B - (BC)A. s
Simi l.
r-(~c~)

(A.C)B - (A.B)C

Indeed,
Ax(BxC) : -(BxC)xA
F- -[(A.B)C - (A.C)B) :

(A.C)B-(A.B)C

The two multiplic-ationrhave the same rule of expansion,

namely

remote
~
(AxG)xC
~----

= (C.A)B

- (C.B)A

remote:
---~
h:;o.xc):-'
(A.C)B - -(A;B)C.

139

J
Generaliz~tion-of

triple scalar and tr.

vector products

are the following with their expansions:

1. (Ax8).(CxD)

= (A.C)(8.0}-

(B.C)(A~D)

2. (Ax8)x(CxD) : (ABD)C-(A8C)D : (ACD)8 - (8CD)A

Proof.
1. (Ax8).(CxD)- (AxB)xt.o

(interchange. of signs)

:. [ ( A C) 8 - ( 8 C) A] 0 .

:. (A.C)(8.0) - (8C)(A.D)
2. (Ax8)x(CxD) : Ux(CxD)
~

:. (U.D)C - (U.C)D

= (Ax8.D)C (Ax8)x&Qj

(Ax8.C)D

.:.

(A8D)C - (A8C)D

:.

(Ax8)xV

= (A.V)8

- (8.V)A

= (A,CxD)B.- (8.CxD)A
.:

(ACD)8 - (8CD)A".

VECTOR SPACES

The concept of vector &pace is based on tht

group

and~

fi 1el d.

A group isa non empty set

operation, denoted by
aobtG)

. cepts of

"o", is

G of

def~ned

element~

(that is

1
d,

b~;;G,

single
then

satisfjing the three axioms below:

G1 . F6r any three elements
distinct:

a, b, c

(aob)oc :. ac(boc)
G2 . There is an element
element; such that

"e"

in

of

G, not necessarily

(associ~tive

law)

G, called the identity

140

e a a

=a

a o e

for any

aG~

Given any element "a" in G, correspondirig to it there

. is uniqu~ element, denoted by a_-l ,'called the inverse
of a, such that
a-l o a

=a

o a-l

=e

If furthermore
A4

aob

holds, then th:

boa

for

very

0up

a, bG

(commutative law)

a commutative group or .an

ed

~b_elian

group ..
-

'

If the op_

is theaddition "-+",the group isan

~ation ".'~the

ad .. itive.group,

.,

.'

ca t: qroup.
The i de~ _i t;,
by

group is a multiplf-

a 'n d i n

m_u lt i p 1 i ;..

.y " 1 "

n additive r.1oup th1

one a

or "I "

element a~ 1 of "a" is written

-a,.

additive group is denoted

the addi. ,~ lnVEi'::.c

u1

"a"

~-

in

a multiplicative

s called the multiplicative inverse of "a"

~rizing:

For

additi

ommutative group the axioms are.

a+( b+c)

(associative law)

2. 0-+a : a-10 : a
3. (-a}+a ;::. aq-a)
4. at b :::: b-+a
i:
.,

''

(existence of zero element)

.=.

(existence of inverse)
(commutative law)

while for a multiplicative commutative grpup;

1. (ab)c

= a(bc)

2. la ::. al = a
-1
3. a a ::. aa
- 1
4. ab :::

'

141

It." ts easy to verify that the set

R of all real numbers

R*dR-{0}

is an additive commutativegroup; while

is a multipli-

cative commutative group.

The set

Minxn

0f all real matrices of the same size is a

commutativa additive group, and the set. Mnxn of all non singular
square real matrices is a non commutative

~ulti~li.cative

group.

nu~erous.

Examples are

Now we defjne a fietd: A set

= {a,. f:l,

is called a

field, if
--

i. F

is an additive abelian group,

ii. F*

is a multiplicative ab~lian group,

iii. Distributive law holds: a(SH')

The elements of a field are called

s~alars.

The familiar examples of fields are the set

real numbers, and the set

= aS+a;r
IR

of all

of all complex numbers.

Vector Space:
Let

V =- {u, v, .;.}

F ={a, S, : }

be a non.empty set, and let

be a fi.eld. Supp~se in

opera~ion of addition. Then

the field

V th~re is defined an

V is call~d a Vector S~ace over

F, if
I.

v is an additive abelian group:.

1. (u
2. 0

v)

,u - u

3. (- u)

4. u -\

Wjo:. U

-t

-t

u - u

."::.

(v

-t (- u)

w)

-4

':

142

II. For an.g

and any

u' Yt.V

a, S F:

1. lu ::. u
2. (aS) u, ::. a (S u)

3. (as)u

- au ... su

4. a(u-tv)

.:

This field is

au t av

d~~oted

scalars while the elements

Among many

vector~

V(, R). The elements of

by

are

V are called vecfors.

spaces we mention.the following as

examp.l es:
Example 1. The
i

R 7 { (X ' y ' z) :

I! I'

,,

li
I

X'

y ' z t.R}

of all ordered friples ~r vectors in 3-space is the v~ctor sp~ce

3
R3 (+, ~)which become an inner product space R {+, ., R) when

i:
I

s~t

((

'!
i;:
;i

inner product (d6t product) is defined.

l:

Exampl~

l,i

The se.t
I

C[a, bJ = {f:

~~

.;

is a veCtor space

i,.l

space

I!

I[

b) ( +, IR)

C [a,

C[a, b](+, . , R)

.is continuous.on [a, blJ

which becomes an inner product

if an innerproduct

<f, g>

is defined

il

tlj

'I

~~

<f,_ g>

i
)

~~ '

~l

The. functions

i:

'I

:';1

.if

<f, g> ::. 0,

lrJ i

is called the !!!.In of

~I!

!ji

~xample

; i ~I

;Jir
:!:

lii!,,

,.1 1
::1,11

,::~;li

hlli

3.

.f

on

[a'

b]. '

'\

2
1 2
f (t)dt) '

The set
Mmxn =-{[aijJmxn=

H!

f f(t) g(t}dt
~
are said to be orthogonal on [a, bj

.::.~>~(

nfJI

q.

and

[i;

.f, g

ai:ie:R}

143 '

is a Vector space

Mmxn (t, ffi)

where the inner product (dot

product) is defined as
<A, B>

= A.B

- allbll"'
t

a2lb21 +

+ anlbnl

.j.

alnln

a2nb2n

ann bnn

A', _by II A"ll :. /A.A

The vector space R 3 (~. R) has the.natural .generalization

and the norm of

where
Rn .::. {(x 1 , .. , x0

xie:R}

):

is the set of all ordered.n-tuples or vectors in n-space,in which

the operation of addition, multiplicatidn by scalars and i~ner
product are defined as:

'-.'

X )

n.

, yn)

=(A. X l , . . .
~

, A.
.

x 1y 1 +

X )

xny n

The vectors
, 0 ) , e 2:.( 0 , 1 ~ 0 , . 0 ) , . . . , en::. ( 0 ,
in this spate are unit vectors and
~pplication

pair~ise

' 0 '1 )

orthogonal as seen by

inner product:

e 1 .ej

1
0

They are said, to lie on

when

i:..j

when

itj

n (> 1) mutually orthogonal axes

sketch of whiGh cannot be realized

wh~n

n>3.

Ox , .. , Oxn.
1

144

Any vector
combi nation of

P:. (x , ... , xn)

1
e , . . . , en. Indeed
1

xn) =- xl el

(xl
Lin~ar dependence and

The k

+- + xnen

independence in

non - z e r o v e c to r s

dependent if there exist

can be written as a lfnear

u 1 , .... . , u k
.

Rn:
a r a c a 11 e d 1 i n e a.r 1y

scalars c , ... , ck'. not all zero,

1

such that

=-

clul + + cknk
otherwise

u1 ,

, uk.

ar~

ca.lled linearly independent. In other

words. u 1 , .. , uk are 1 inearly

according as a 'relation

implies or does not imply

t

independent or 1 i nearly de-pendent

c :.o,
1

Theorem. In thevector sp~ce

ck~o~

Rn

1. n . vectors
. ul
u2

::.

( u 11

= _(u21'

..

un

...
.. .

-- ( un 1

uln)

u2n)

unn)

are linearly dependent or independent according as

det[11ij]
is zero or non zero,
2. There is a set of
3. Any 'ntl
Proof.
1. Setting

lin~arl~

independent vectors,

vectors are-linearly dependent.

145

-t

. n n :.0,

. -tCU

one has

or
-tcnunl' ' . ' Cl Ufn +- -tCnUnn) :..

'

i~plying the homo~eneous ~quare system

'\tel .-t

unl en.-:. 0

of linear equations in the unknowns

determinant is
If
cl'::.O,_

of which the

D =.detjuij~

Dto, the system admits only'the trivial splution

, cn=O ;neaning that. the vectors are -linearly :independent .

.If D~o. the system admits solution other tfian the trivial

one, meanin~ that not all

linearly dependent.

c~s

are zero, and the veit6ts are

'

2. ~n

contains the qnit vectors

,.
e 1 :. (1, 0, . , 0), , en:. (0, .... , 0, 1),

which.are linearly independent since detluijl is 1Inl=- 1 t 0.

3. let
be~non zero vectors in
Rn.
The theorem is proved if . n

of them, say

linearly dependent, in which casd

linearly dependent.
let then

u ,
, un
1
theo suffi~e-to prove that u
tJl' 'un:'

u , . ; , un are
1
ul, ~ , un, un-+1 are

be linearly independent. It will

= uhtl

is a linear combination of

146

Ii .
I'

This equality is equivalent to

...

1 u 1 ,-~

cnunl

ul

::.

c;:nuln +
with

cnunn :. un

detlu1 JI. 10. Hence there is a unique solution in

. ing that

ntl

vectors

u1, .. , un' u

(-2, 3, 0)

and

are linearly dependent.

A= {l, 2, 3), B =- {4, 5, 6}

in 1R 3 show that A, B, C

Example 1; Given the vectors

c =-

cs, mean-

0 :. {0, 11, 6}

.are ljnearly independent, and then express

as a linear combi-

nation of them._
Solution. Since
2

-2

.::: 24

A, B, C are linearly indep!ndent.

Writin_g
{0, 11, 6} =- a{l, 2, 3}

+ b{4, 5, 6} + c(-2, 3, 0}

we have
a

4b - 2c :. 0

2a ~ 5b t
3a

solution of which being

'

+ 6b

:. 6

a:.O, b:.l, c:.2,

we have

0 =- OA t B

-+- 2C

Exam(!le 2. Gi ven the vectors

A.-

=-

{9. -1, 0, 0),

::.

{1 , 2 .. 3. 4)

c
in

a') show that

..,,

11

3c :.

:.. {0. 0 ' 7. 9).

IR4,

A,

s,

C, 0

{ 1 9, 0, 0}

0 =- { 0. 0, --9, 7}

are mutually 9rthogona1.

and

147

t, D for linear independence, an-d

b) Test. A, B,

the case, express E in terms of

i f it is

A, B, C, D.

. Solution.

=-

a} A,.B =- 0' A.C ':... 0' A.D

9

b) '

o.

.0

-1

-9

:. .-10660'7" 0

=9 a-t9b - 1,

bB

.f

0, C.D

,'

E =- aA

~linear

independence.

cC t dO

= 2,

9a-b

a" 19/82,

=;)

0 , B.C ~ 0, .B;D

~,

7c-9d .:.

= 7/82,

9c+7d = 4

c =- 57/130,'

d =- 1/130.

Bases:

V( -+ ,

'A set :B ::. {b 1 , , bk' } of vectors in a vector space

F) i s c a ll'e d a ~ f o t
V( +- , F) i f
1 ) bl ' b2'

are linearly independent

2) bl ' b2'

generate

v
If

(~pant .. V' 'that is any vector in

b IS

is a 1 i.near combination of

bi 1 s a.re mutually ortbogona 1, then :B

~rthogonal system and if further~o~e

b IS
1

is ca l1 ed an

are unit vectors J

is said to tie an orthonormal system~ ..

Clearly
{e 1',

... ,

en}

{i, j, k}

is a basis for IR 3 (-t, R),

Rn ( + , IR) , and each set i s or tho norm a 1.

i s a bas i s for

F6r the Vect~r space

1
[_o

0]

ro

and

M2x2 ' a basis is

1]

o ' lo o.

ro
'L

0]

ro

o_ ' Lo .

0]
1

Indeed,
1) theke four matrices are

linear~y independent~

since

148

0] + [0 ~r
d

r: :H:

~,

[ac

:H: :H: :J ~

bdJ -- 0 -,

[ac djbl

r: :] ~ a[:

2) Any matrix

B ....

I I.

. Reca 11 that in
b1 b2
is
. b.J btt

M2x2

<A, B>

.' : i~

:.

d:.O

i~

in. M2x 2
of the mentioned four matrices:

0 .

c .... o.

a:o, ; b=O,

-la..

li~ear

b[: }c[o 0]
0

t.

1:

the inner "product of

albl + a2b2

-t

a 3b3

-t

combination

~I
A

~
a't~

a4 b.f

Example 1. Show the following:

a) Any non zero vector fn R1 . is a basis for. ~ 1 (, R).
b) Any two non zero non parallel vectors in
for IR 2 (t-, IR)

~ 2 is a basis

c) Any three non zero non coplanar ~ectdrs in 4R 3

is a

~ 3 (1-', R).

basis for
, Solution.
a) Let

A= (a)fo be'any. ve,ctor in

is

with

aiO. Since

'

detlal t' 0,
B .:. (b)

JR 1

A islinearly indepen'dent, and any vector

.

equa~

to

A.

A =- (a 1 , a 2 ) _'I 0, B ::. (b 1 , b2 ) '1",0 be two non

-~
~
I
parallel [b
~ b )
vecto~s. They are linearly

b) Let

'

2 ..

149

indep~ndent,

.. :.

J:~ :~~

a 1 b 2 -a b f 0, and any vector

2 1
can be written as a linear coobination of A,_B:

since

C:. (c 1 , c 2 )

C = a.A SB --t.=)

c2 - aa2 _.. sb2

1:

the latter admits a unique solitionsince

c) Let

n~n

/to.

B ::L(b 1 , b , bi)to,
2

vect~rs

be non coplanar
are

1 . :1
2
2
in

~3

l)

in

Since A, B, C

coplanar we have

(A B C) =-

showi'~Jthei_r

al

a2

a3

bl

b2

b3

cl

c2

(;3

t 0

linear indeprendence. Any vector

~ritten as a linear combination of

D ,., aA

SB

-t

can be

A, B, C:

:+ C
.

_Yi,elding a systemof linear equations i.n <X, s.x

solution

of

which exists s.ince d~terminant of the system is not zero.

Linear dependence in

f 1 (x), ~.~, fn(J!:) e: Dn-l[Ij, that is,let

Let
fn(x)

~e

Dn-l[Ij (~. IR).

f (x), ... ,
1
f~nqttons differentiable up to the order n~l on

an in te rv a 1 I.
Theie functions are said to be
exists scalars

cl' .. ,

en

c 1 f 1 (x)

line~rly.dependent if there

not all zero such that

-t

-t cnf (x) ::. 0;

11

otherwise they are linearly independent.

Differentiating it succes~ively up to the order
get the

~omogeneous

system

n-1

we

150

ci f 1. ...

., cnfn -

t-

cl f

1 ....
~-

1-

c nf'n - 0

of linear equations, implying

fl

w : w[f 1 ,

f n1-=-.

(non triv.sol.)

f'n

_(only triv.sol.).
f (n-1)

f(n-1) ...

where

W fs-caJJ ed the WRONSKian of' the given set of functions.

Hence a set of

ind~pendent

functions

f 1 , . , fn

or linearJy dependent according. a~

is linearly

W[f 1 ,

is i denti ca lly zero 'or no.n zero

~xamp]e

1~

Show that

a) two constant functions,

b) three lin~ar polyhomials~

ci

four quadratit polynomials

are linearlty dependent in al'iy interval.

Solution.
a)
The~

Le~

P(x)

= a0 ,

Q(x)

b0

be two constant functions.

forming the WRONSK1an, we have

'

w-(P,. Rj
b) Let
three

1 i'near

P(x)

--

ao

bo

':.

a 0 1a x,
1

polyl_lQmials. Then

Q(x

~>

linear dependence
be

. 151

a 0 -+a 1x
W[P, Q, R]

b0 'fb 1x

0 ~c 1 x

al

bl

. cl

. ~

1i near dependence.
c) Proof is similarly done.
dependen~e

Example 2. Test_thefollowing for linear

. e -x ,

c) sin x,

C.h

b) 5,

X,

d) sin 2x,

COS X

cos 2x

Solution.
a) w[ex,

(Co~pare

-x

'

Ch xJ

:.

X
X

- '0

dep~ndent.

x,

Ch
Sh
Ch

the first and third rows).

They are linearly

b) w[5,

ex e-x
ex-e-x
ex. e-x

x2]

.::.

x2
2x
2

o.
0

: 10 t 0

sin 2x

(linear,i'nd.}

COS X

cos2'l<] .,.
2sinxcosx
sin x

-2cosx sinx

COS X

= 2 sinx cosx

: -sin2xt.O (linear ind.}

C0 S

- S, i

X .

tXERCISES (3, I>

1. Plot the followi~g points on a positive rectansular co9rdinate
system:

152

a) A(2, 0, 1 )

b) B(O, 3. 2)

c) C(2. 2, 0)

d)D(l, 1 1 )

e) E(l, 2, 1 )

f) F(O, 0, 3)

2. Examine positiveness (negativeness) of a cartesian system

which is symmetric of a positive cartesian coordinate. systeo
with respect to:
a) a point
3. Find the

b) a, line

symm~trics

of the.point

c) .a plane
A(l, 3, 2). with respect to the

a) origin

b) x-axis

c) y-axis

e) yz-plane

f) zx-plane

g) xy-plane.

4. ABCDEF

])Oc
, ,"

is a regular hexagon with unit side.

+

. write

d) z-axis

AB + AC

AD + AE

of a vector and find

5. ABEF and BCDE

it~

+
~

AF

as a multiple

length.

compute

lengths~

A
+

b) FC-FE
5. By

d) FC-DB+DC

c) 2FEtDC

a line inclined at an angle

....

8
-7

the sides of an equilateral triangle

proj~cting

1.

are unit squares.

thei~

Construct the following line vectors

~nd

1.

e.

ABC

onto

Q td one

of them, show that

cusQ

= cos(Q

sinQ.

sin(Q

: 0

7. Consiruct two vectdrs whose sum and difference are the given
vectors

u and

v.

8. Theree forces of 'lengths

5'oo

.. 120 . 21S 0

3. 5, 4

units are in the .di.recti ons

from East. Find components of these forces. in

the direction of Nortn and East. Then find the .direction of the
I

153

resultant force.
9. Compute

a) (3i ... 2J

10~

Find

I.e:~

4k). (3 i

2j

..

for the vectors

7k) ' b) (7, 8, 4}.(5, -.9, 4)

i

..

2j

3k and

4i.

..

5i + Ak

, to oe orthogonal.

n.

Find

(A-8).(A-8)

.12. Given a vector from

A(2, 3, -4)

to

8( -4, 5, 6) ,
magnitude and direction cosines of the vector.

find the

13. Expand and simplify:

-+

-+

-+
-+
u .::. (2a
3b).(a + 2b)
+
-+
I a I :. 3, . I b I = '2,. .:}: -+a, --+b

where

:=.

rr/3

14. E:valuate co.s i nes of the angles a,

a. y of the triangle
.having verti'ces at A( 1 , 0, 0) , 8(0, 2, 0) ,
C(O, 0, 3)
-+
-+ 2
+
15. Show: (a + b)2 = a
+ b 2 - 2 *a.b+
16

Given A
a)

--

( 1 , 2, 2),

A.8

b) .fA

:.

( 3 , -1 , 1 ) ,

find

c) vector component of B in the

direction of
17. a) Given
find
b), Find

lA

::.

[3,

.,

A.

,..

tl -5' 20' -25]

A.A

is a minimum,

-1'2, 15J

te: R such that

'.

s, t

not both zero such that

s [~, -12, 15]
t[-5, 20, -25] -::, 0

"'

18. Give.n A=- (1, 9, 7),

a minimum.
19

a ) Gi v en

s, te:iR

8 =- (9, 4, 2),

.A :. ( 1 , 2 , 0 ) .- s ( 4 , 5 ,
such that

A.A

z)

find t

for lA

t81

to be

t(l, -1, 2)~ find

is a minimum,

II:1_1

i~

II

''~"fil-----:-------:----------=1~5~4~~---~==------------

b) find

a, b, ce:R

a ( 1 , 2, 3)
2 0 . Gi ven

-~

not all zero such that

b ( 4, 5, 2)

A =. (- 1 , 2 , - 2 ) ,

c {1 , -1 , 2) ,., 0

B =- {1 , - 4 , 8 ) ,

f i nd u n i t vectors

in the direction of
a) A

c) A-B

b) B

21. Place three- vectors of lengths

7, 24, 25

units to have a

zero resultant.
+

22. Three mutually orthogdnal vectors. a,

t, {, /l"
a) find

b~

have lengths

respectively;
+

Ia

cl

b) find angles betweeri

23. 'A certain force

and

a, b, c

3i- 4j- 2k kg pushes an object in t'he

direction of the vector

to

A = 2i - 3j t 5k

by a distance equal

IJ!.l

a) find the work done

b) find magnit~de of th~ force and displate~ent
c) what is the angle between two vectbrsi
24. A vector makes an angle of 60

with the x-axis, 60 with

y-axis. Then find the anglesthat

25. Let

A~

G = B-Ar,

A.G

~ 0~

(2, 4, 4),
H
A.H

(3,-'3, 0),

=.

C-As-Bt i where

~
~

0, s;H

~t

makes with z-axis.

C

=.

r,' s, te:iR

0, Find

{1, 5, -1)

and

are such that

r, s, t.

26. a) show that in a parall_elogram the sum of the squares of the

sid~s

is equal to the sum of the squares of

diagonal~.

b) show that in a triangle, the sum of the squares of the

sides is t!qual to 3{4 of the sum of the squares of the medians

155

27. Expa~d an.d simpli.fy:

(A~B)x(A+B)

28. Find

{2A- 3B)x(A

2B)

and interpret its absolute value geometri-

cally.

=0

29. Show: AtB+C

30. Find

AxB

BxC -:. CxA :.. AxB

=)

where

=-

i+2j-3k,

= 4i-5j-6~

and find

sinEl

for these vectors.

31. Find unit vectors perpendicular to both . i tj - and

A(2, 1 4)

32. For what value of A wi 11 the vector from

8(3, -1 2)

jtk.

to

C( 4' -2, 1 )

be parallel to the vector from

..
to

0{2, 2' ;.)?

33. Let

A(O, -2, 1),

B(l, 0, 3)

and c(3, 1, 1).

be the vertices

of a triangle. Find
+

a) AB

AC

3 4 Gi v en

A :. (2 , - 1 . 3 )

and

B :. ( -3 , 0 , 2 ) ,

f i nd a u n i t

vector which is orthogona 1 to each of the vectors

and

B. '

Is there more than one solution?

A =-

35. Let

D :-

(2~

-14, 8)

c = ( 0. 1 -2) '

( 1 -1 ' 1 )

a), compute

AxB

b) find

a' b

c) find

.a, b, c

not both zero such that

A: (3, 2, -1),

a) find

aA

not a 11 zero, such that

d) find a, b. c .. d
36. Let

B .: ( -3. 21 -12)

not

a 11

-\

bB

aA

zero such that

-1

=0
bB -+ cC =-

aMbB-IcC4dD :::. 0

B:. {1, -3, 5),. C:.. (4, -1,_1)

AxB

c) find the volume of the

are . [OAJ., (OBJ , [OCJ

/ b) AxB.C
p~rallelepiped,

three of whose edges

]56

37. Given

A :::. ( 2, -3. J),

:.

find a vector

(1, 2. -1)

X, i f

any, as the solution of the equation:

a) A.x
38~

Prove

=5

c) (AAX).B = -2

bJ AAX = B

ih~t

the tfiangle

.and R(-1, 4, 5)

wi~h

vertices

P(-1, 4, 3),

Q(2, 0, 3)

is a right triangle, and the ~r~a IPQRI 2 =5

39. Let 1rnr denote the oriented area of .a triangle ABC in xy~plane
that is, AlfC is positive or negative according as A, B, C are
or tlockwise order. Then if A(a 1 , a 2 ),
B( b1 , b2) ., ' c ( c 1 ' c2)' prove
al a2 .1
. 1
1\lrr="2"
bl b2
cl c2
40. Prove:
in

co~nterclockwise

a) jABCj 2

=2

jOB x oc + oc x OB

b) . I ABc I 2

=~

(j AB

BB I

X'

AC I + jllC

OA x OBI

I CB

CBI ) '

41. Prove the tdentity

IA~Bj2 ~ jAXBj2 ~ IAI21BI2

'42~

Show that the indicated six

(LAGRANGE)

point~

a~ .~idpoints

of the edges in a

right parall_elepiped are the vertices

of

a plane

(AxB)~C

43. Evaluate

and

Ax(BxC)

where A : i~j-k,

B :. 2i-itk,

C ~ i+2j-k

and compare the results. Does there exist the

~ssociative

law for cross multiplication?

44. Prove:
45~

hexagon.

Ax(BxC) ! Bx(CxA)

Determine scalar

>.eR

Cx(AxB) 0

such that
."1

157

A C) = >.(ABC)

. (Ax(AxB')

46. Pro've
a) AxB~(AxC)xD - (A.D)(ABC~

= (A,Dj(ABC)

b) AxB Ax(CxD)

(A.C)(ABD)

47. Expand

48; Prove

a) (ABC) UxV:

b) (~Bt)(UVW)

A.U

A.V

B.U

B.V

c.u

c.v

A.U

A.V

A.W

B.U

B.V.

B.W

c.v

c.w

jc.u
49. If

A, B,

p :. aA

50. For (ABC)

tf
'0

then
I

a . .,_
1A

..are non collinear,v~ctors, prove

+ as + ac

-t

0,

prove

CxA

- '(ABC).

s:-

- .if.ill_
' a - (ABC)

ABC)

(a b c) '{:- 0,

BxC
"("A'i3CJ',

bxc
""" (a b,c}

' =-WaH-

:=)

cxa

(a'bCT

'('-:

{ASP}
(ABC)

fhd/c

:.

:.

'Axs

(~BC) '
axb
(abc)

51 . Prove that the fol,lowing s.ets are vector spaces for the given
operatio'n, over the given field:

b) R* ~. ,

c) It;

+,

52. Same question f6r:

a) M2 x 2.; i; IR

c) M;x2

=ll: :1 : 1:

b)'C[a,b];

:1t

0}; . , R

.. , R

d) C* ; , R

I
I

1 58

53. Given

A ~-

<A,

a) find

~]

[1

B =-

[~

:]

in

M2x2{+, ' R)

II.A II II B II .
for the -rna trices A,

b) find

B>

<A
c) evaluate cos9 :: IIAII

B>

II ~II

B.

54. Same question for

C"' [1_8 _42] "

55. Test .[:

:] ,

M2x2(+, IR),

[:

:],

.D -: . [-4
[:

2]

10

:] , [:

and if it is a

b~sis

:]

for a basis for

as a linear

express [ 43

combination of basis vectors.

56. show that

a)

b) {1-i,

{1, i}

are bases for the vector space

z - 5-3i

1-ti}

It{+, IR),

and express

in terms .pf the elements in each basis

57 .. Sho~ that the three vectors

dependent if

A~B,

A,

AB+C

A, B, C are so.

58. Find a linear combination of

[o

l] T

and

i] T.
which is orthogonal to [o
. 1 . - 13 .
/3.
1 .
59. Given vl ::.2 1 -TJ,
v2 :: -21 - 2J
that v1 v2 is an orthonormal system.
60. Show that there exist
such that

are linearly in-

scala~s

a, b

and

[1

o] T

in ~ 2

show

not all zero

159
61 . Gi ven
l

A - ,!J" i

l
,13 k ,

= ,12l ..

A :. (a, 1_, 2) ,

b,

B - (1 ,

2)

and

Hhat must be relation between. a, b, c

linearly

a)

3J T

expres~

+ __;,..

vo

J' - ~ ..,
vo "

C:.{l,2,c).

A, 8,

for

to be

dependent,

63. Prove with out any computation that

[o

l .
l ,.
/2" ,-~ , C = 7b

1 -

A, 8, C is an orthonor~al syste~i

show that
6 2. Let

-f . v'3 j

constitute a basis for

I 3. 6

9IT .

as

[1 2 2) T,
3
lR . Then

[o

2] T'

"

linear combination of the vectors

of this basis,
b) which of the vectors of .this_basis may be replaced by
[3
64. G:iven

5]T. and the resulting set still being a basis~

A(O, 1, 1),

8(1, -1, 1)

.. linear combination of
65. Let . A

= (3/5,

4/5,

0~

A and
a) ,

= ( 0, 0, l t/2, l//2),

and
8

C(ll, 2, 0),

find a
~o

that is orthogonal

8'::. (-4/5,

C.

3/5, 0, 0).

D .::. (0, 0, -l//2,

l/.12).

Show that .A' 8' C,. D is an orthonormal system.

66. Test for linear dependence, and determine the dimension of
the space generated ispaned) by these vectors:
a) (1, 3, 5), (1, 2, -1), (4, 6, 3)

b) (1, -1, 2), {3, 4, -2), (4. 3, 0), (1, 6, -6)

c) (1, 0, 4, 3), (2, 1, 8, -6), (3, 1, 4, 8)

d) (2, 0, 3, -l), (4, l, 1, 2), (0, 1' -5, 4)

67~ The distance'&etween the points

(3a, 2a, 2a;1, 4)

values for

a?

in

IR 5

(a, 2a, -a, 3, 1)

and

,--

is

,126. What are the possible

160

68. The vector

[a

41 T a.nd

2a 2a

[4

2a

2a

aj T

are orthogonal. What are the possible values of

in

R4

a?

69. Test the functions f.or linear independence:

. a) sin
c)

l. x.

'

sin 2x
. n

....

b) 1

cos x, cos 2x
2x ~
d) 1 ex, e . ~nx

. X

70. Given sin x, sin 2x

.in

c [o;

21TJ( +; IR)

a) test them for orthogonality on. [0, 2r]

r.

b) find their norms

ANSWERS TO EVEN NUMBE.RED EXERCISES

2.

b) posHive

a) negative

c) negative

-+

4.

3AO,

8.: .3 cos 40, 5 cos 30. -4 cos 55._, 3 cos 50, 5 cos 120~ 4 cos 215;

tan 9 =

3' CO,S 40,.. 5 co.s 3.0 -/ 4 cos 55

3 cos 50

..

5 CO.S'l20 ... 4 cos 215

10. -14/3
14 ~ cos 'a =

5~~'2"

16~

b): 3,.

a) 3,

cos

'*.

5t ..65

1-.2. '2;a; . '-3//35,: lt/35.

=-

9
cos ":" /1 30 .

c) (l/3, 2/3, 2/3)

18 .. -59/101
20. a) _(1/, -2, 2)/3,

l2.

a)

c) (1, -3, 5)/135

hR. 2:~ R. ,

b)

24. i/4.

a.

e.'

=- e,::..

arccos

1:29. 2 -!1/.'

.31T/4

28~. 2 AxB,

ICDEF1. 2

30. {-,27, -6, -13),

32. 5

b) (1; -4, 8)/9,

'-=

2jOARBI 2

sine::. Ml/(7/il).

.l

161

....

34. n

= ( 2.

13, 3 }/li82.

36. a) {7. -16, -:11},

5~'

a) 13.

58~

a[o, 1, 1JT

b)

yes,
b) 33.

185, 11,

....

--n

c} 33

c) 13/ ( 11185')

2a [1, 1 o] T

for

aER

62. 4a _. 2b .f c - abc ._ 6
64. 9a(O, 1 1 ) - 2a ( 1 , ';" 1 , 1 )
66. a) i nd. , 3~

b) dep~, 2.,

c) ind. 1 3,

d) dep .,2,

6&. 0, -1
70. -orthogona 1,

ITT,

ITT

il

'"
11

I\

'.t i

(-

162

'~
I'

3 2
I

,1-

PLANES AND

LI NESI

A, PLANES

Planes will be denoted by small Greek letters

Equations of

som~

u, a,

s,

special planes:

a) Plane parallel to a coordinate plane

(perpend~cular

to

-a coordinate axis):

n//Xy-plare
(Horizo~~~l

Tr //

yz- p 1an e

x:a

plane)

Z:.C

b) Plane perpendicular to a coordinate plane

a

coordinat~

to

axis):

i'r

1:

xy-plane

.;

uJ..xz-plane

c) Plane passing through a coordinate axis (passing also

origin in case (b)):

= mx

163

Equation of the plane through a. point

a~J

to a. vector:
Let
-;,

and

P0 (x 0 , y 0 ,

==(A, B, C)

be the given point

0)

be the vector. Let

be, any point of the plane n through

P(x, y, z)

and

N. Then

perpendicular to
+

vector

N j_ IT'

Then the

=- 0 ( eguation

re~uired

equ~tion

is
( 1)

or
n:

where

Ax+ By

Cz

D:.

~(general

equation)

(1')

D :. - Ax 0 - . By 0 - Cz 0 .
+'

N is called a normal vector or a direction

The vector
vector,
plane

~rrd

its components

A, B, C are direction

number~

of the

n.

Observe that direction numbers

A, B, C

app~ar

as

coefficients in (1) or in (1'), and hence the planes

Ax+BytCz+D

and

A'x~B'y+C'z~o

are parallel i f and only if

~

-;

(N//N'.)

Similarly
AA'

BB' + CC' :. 0

<;:=:>

n.l_n 1

Remark. The plane represented by the general equation

Ax + By
when

A~O,

Cz

B:O

is parallel (perpendicular) to

~y-plane

(C#O). Similar results hold when parallel (per-

pendicular) to other coordinate planes.

164

equation.

~ormal

'

If the direction vector

N of

'

is a unit

(cos .a. cos e. cos:o)' we have

+.

n. {P- P0 )

-++

n.OP

where

OH

;9

n.Or 0

:..

+;t-

n. P... n .P 0

++

+
OP

is the prbjeition of

Then, i f

=:

IOHI

=)

+
=+
n.OH

n.OP

.+

or
+

OH

onI

n.

= p{cos

a. cos 13 cosr) we get

n.OP ::: p

or
~:

Cos a + y cos 13 +

cos 1(

which is c~lled tha normal equation since

p

=0

{2)

cos 2a.+cos 2 etcos 2 ~1.

is the distancc ;;f the origin fr.om the plane.;

called the EULER's equtionor HESSian form of

{2) is also
~

The equation {2) can be written in the form

~= ax +-by+ cz -i d
which is also a

=- 0, a 2 .; b2 t c 2 .::.

{2I

equation.

no~mal

The general equation {1_') can be normalized by dividi~g

/A 2 + s2

its; every term by

1

-~=
~a

Ax

+ BX

h2

t ..

"<

s2

that (2);'{2'), {2 11 )

+ c2

Cz
1

c?

{f 0):

il' ~ 0
-

{ 211)

are all normal equations.

The interesting fact about normal equation is the simplicity

6f t~e

expression of the distarice of a point

P0 {x 0 , y 0 , z 0 )

from the plane:

1x 0 cos

a-t

y 0 cos e

1 Ax0 t

By 0 +.

Cz 0

/A2 + s2 -+- c2

-+

+ z0
D1

cosY- PI

{3)

165

Indeed, 1et

be the plane //'IT ,

'~~'o

passing through . P 0 (x 0 ~ y 0 , z 0 ):
'11'

X COSa

y cosa

.COS

(pd}=O.

xocos a ... YoCOS a + zocos 'll - (pd}:O

d - XoC?S 'a + Y(,COS. a + zocos y

=).

....:)

e.

I X0 c 0 s

- p

a .. y 0 c 0 s a .. z 0 c 0 s r - p I.

Example l; Gfven the point

A(3, ~2~ -1.)

.,a) Write the equation of. the plane

perpendicular to

7T

-~

and

N= (1, -8, 4),.

through A and

N,

b) Obta i.n the normal equation of

'IT

c) Find the distance of 8(2,2, -5 j .. from

'IT.

Solution.
a)

(x-3)

8(yf:2) ,._ 4(ztl) - 0

X - By :+..4~ - 15/:. 0

-~

b)

Since

f12 + <~8)2 + 4 2' = 9' . we have

X -. 8~ ... '4z
15'
.:-0
9

c)

d(B, 1r)

Exam~le

'IT: 2x

= 12

2.
2y

~ 8.2 + 4(-5) - lSI~

Give~

a) show .that

the planes
3 .--. 0

and

b) find the distance

d('IT, 'If I

Solution
. 2

-2

'If': 4x

'11'7/ 'IT'

a) 4,= =-4=2

4~ 19

5> n- II 'IT'

4y + 2z + 7 = 0

166

distanc~

b) The

between these parallel planes is the

distande of a point on orie ffom_the ether:

A( 0 , 0 , 3 ) e: 11 ,

then-

d ( 11 ,

11

14.0 - 4:0 + 2.3 ~ 71

..[36"

1:

Observe that

:.6i.~

= d (A ,

:.

11. 1

13

-3

31 (difference of distances of the

origin from the planes)

' Example 3. Given the planes

1f:

2x

2y + z t 3

a) find

ae:R

0 and

.::

such that

11

111_71

1.

4x - 4y t az ; 7

::.

'

b) write a unit vector parallel to

11

Solution.
a)
b)

+'(-2)(-4)

2~4
1f 11f

NI II

=)

-')

n ""

I'

Any uriit vector

I,

!
'

:.

Nl
I N'l II

11

-=)

n =

--=)

11

:::..
-"

( 4, -4, -16}
-=
4l/T8'

II

1 a

Pz

.-1 ! -4)
3,12"

to

._,

N .. if

=.(cos a, cos

s,

cos)

is perpenducular

-~

N.U = 0

implying

2cos a - 2c.os 13 + cosY =- 0

with

infinitely many-solutions.
.

~guations

of a plane through three points;

let

Pl(xl' yl' zl),

determine a plane

'il.

P3(x3,y3, ~3)

P2(x2' y2, z2)1,

The unknown eq'uation

I~

II

-1,6

Ax + By

4 C~

+ D_ 0

and the three conditions

Ax 1

Byl

:*

Cz 1 . + D

Ax 2_ + By2 + cz + D
2
Ax 3 t By3 t Cz 3 _t D
determine
C, D if

HLS

.::

.::

-::

which admits a non trivial solution in A' B,

167

,x

xl

yl

z .

x2

y2

-Z2

x3

.Y3

z3

::.

(4)

Which is the determinantal equation of the plane

w.

When expa_nded gives an equation. in the form (1').

~ntercept

form:

This is the equation of the plane determined by three

points

A(a, 0, 0),

B(O, b_, 0),

C(O, 0, c)

where a, b, c

are

intercepts of the plane with coordinate axes.

The equation is simply

since it is satisfied by the coordinates of

(5)

A, B,

c.

can .be obtained from (4) directly.

Compare (5) with -intercept form of a line is l-space.

Equation of a the
to two
L~t
~

~on

collin~ar

the given

U =- (.u , u , u ) ,
3
2
1

pl~ne

poin~

through a given point and parallel

vectors:

and

ve~tors

be

-1>

V .... ( v: , vi, v ) .
3
1

If. P(x, y, z)

is any point

will be a linear combination of

on the plane

U and

w, then

0~

V:

or
P:.P 0 +sU-4tV
which is a (parametric) vectoral equation of

(6)
w.

.168

The matrix

fo~~

of (6) is

(6I

wtrtch is equivalent to
X

=- X0 + su 1

z : z0

tv1
~ su
2 t tv 2
t su 3 f tv 3
t

~alled the.Rarametric eguatio~s of

p1 an e s

li ilea r fa mi 1y of

(6")

( pen~ il of" p 1 a n e s ) :

The set.of pTa9es through i;i. given l.ine R. is call'ed a 1i~

fa mil y of pm e. s or a pen c i 1 of p 1 an e s of wh i c h R. i s the ax i s .
If

are equations of tWo dfstinct planes of the pencil, then the

family is represented 6y the equation
.

'

v(;l.l' }. :):

where.

}. 1 ,

;l.

:A (A]x-tB ytC

1 z.to 1 ) 1- ;l. 2 (A 2 x-IB2y-IC 2 z-ID 2 ):.0

(7)

2 e:~.

Observe ~hat i f , v 1 //v 2 , the

family consists of parallel planes
(with no axis).
Example 4. Given the points

c( -1

2, 0)
a) ABC

and

D(O, l,

in. the

2)

gene~al

A( 1 , 1 , 1 )

fin:d the equation of the plane

form Bttd then get its intercept and

.parametric form,
b) through

B(2, 0, -1 )

D and para 11 el to . ABC,

c) through BC and p'erpendicu.lar to

AB'C.

169

'
So 1 uti
on.
I

a)

.1
1

-1

-1

. 2.

.7/3

0. .=9

::.

t..Lt...L_7/5

3x + 5y

:. S ,

y ::.. t,

3S

:.

)A( -t f'~t
f
.J

x, y

5t - 7

b) 3(x-0) t 5(y-l)- (z-2) = 0

rr' :

1=.

(general form)

(intercep..t form)

-7 -

Taking two of the coordinates, say

one gets
X

=)

as parameters,

(parametric form)

3x t.5y- z::. 3

1:- +f) -=--1.-.

-#we..~
3A-r ..r11Tt j_ A-IJ(
13

C'

(;=-.

rr
7r

C'~~f .IJ~;)
rr:

R.,

"l

A -

c =o
c. t- ;J :::-v

_ A-t .,_lg t

1J -'--,

~--t r~.
v /f

Lines Lstraight lines} w111 oe_aenotea I:?Y sma11 letters

a,_ b, d,
Equations of some special

line~:

a) Line perpendicul~r to a coordinate plane (parallel to

a coordinate axis):

170

R
6
0'

)-

.L xy-pl ane

xz-pl ane

t .l..

i .L yz-pl ane

(vertical line)
~::.t, Z=C
'X'=d, Z::C
'
x=a, :t= 6,
b) line parallel to a. coordinate plane (perpendicular to
a coordinate

axi~):

--

R.
i

X
o'a

II xy- p1an e
y

R.

= 1 .

Z::O

t: -a + L=
1
b

Z:C

c:

;'. I

_p_

II xz-plane

R.

X .. z
a.
c-= 1 Y=.b

Observe that

ea~h

II y z- p 1 a ne

+- . c~ .:.
u

line is represented by a set of two

equations.
Equation(s) of the line passing through a
par~llel

Let
and

'->

to a

~e

a vecto_r.

b, c)

is any point of the line

+

o,

parallel to

a given
If.

poin~

P(x, y, z)

through

P0

and

one has

'+

pop //aD
p

=)

t:

and

v~ctor:.

P0 (x 0 , y0 , z0 )

= (a,

~oint

=.

=}

Po

pop

:..

t ..1)

tD

Po + tD

(1)

= Cl..-

lll

called the (parametric) vectoral equation of

R., the matrix form

of which is
(1I

R.:

The ~quality (10 is equivalent to

R.:

.z -= z

0
0

-1-

tb

-+

tc

( 1" )

which are called the ~arametric cartesian equations of

Elimma tj ng the parameter

1.

from ( 1") one gets. the

equations
x-x 0 _ y-yI 0.
a
- --o
ca1led the symmetric

eq~ations

-:

( 2)

of

1.

-4

The vector
a, b, c

D : (a, b, c)

the direction numbers of

is the direction vector and

1 .

Observe that the symmetric equations (2) of'a line

are

equivalent to two simultaneous linear equations, namely

x-xo
y-yo
a - --o

z-z oc

which are the equations of two planes containing the line

1.

Th~n

any two linear equationi

Ax+fly+Cz.fD "0,

.A'x .. B'y+C'z-tD' ... 0

(3)

of tw.o i.ntersect.ing planes tepresent their line of intersection.

Consequently ariy line in 3-Space can be represented by two simultaneous linear equations (2) in x, Yi z (non parametric case).

172

Example 1. Given the points

A(l, 2, -3)

and

8(4, -1, 2),.

obtain t'he
I

a) symmetric ~quations,
b) vectoral equation
c) parametric cartesian equations
d) simultaneous equations
of the line

AB.

Solution. Since the line

A( 1 , 2 , -3 )

AB

passes through the point

and admit t i n.g

AB - B - A .,_ ( 3 - 3 , 5 )

as a d i r e c-

tion vector, we have

a)

v--~

x-1

,-- ::. "":: r

Taking

z+3
= ---;- <: t >

B. instead of

A- as a point .on .{,we also have

x-4

y+l
-r- ::. -=r
::.

z-2

-5-

as symmetric equations.
b) P ::. A

-t

t AB

or

B "' s AB.

P ::.

c) Solving from (a) ,

y, z

in terms of

t,

we have
I

::..

lr3t,

::.

2-3t,

':..

-3'i5t

I'

d) From (a) we have

-3(x-l) : 3(y-2).

5 ( y- 2) =. -3 ( z '\' 3)

or
3x~3y-9

.::. 0.,

5y -\ 3 z - 1 ::. 0

Remark. Observe that in the symmetric equations (2) ?f

line, the coefficients of
If

x, y, z

o~

the

numer~tors

are

~11

1.

this is not th.e case, divide every term in a each fraction, by

the corresponding coefficient. For instance,

173

2x-l
y+2
-3z-l
-.- = -=r:. -z-

x-l/2

y+-2

z-l/3

~:. - : j :.~

==)

Example 2~ Write the equation of the linea~ family of planes

passing through the line
x-1
y+2
-:-r
"' ---r

R.:

z-3

= -,-

The n find the rnember of the family.

a) passing through the potnt

A(O, 2, 1)

, b) perpendicular to the plane

c) parall~l to the line

n: 2x-y ::. 3

1:

~ ~ ~ ~

Solution.
x;.-1

-r

-:;

y+2

-y-

x-1
z-3
-r " -,-

>.(2x - 3y

8)

- 8)

a) A{O - 6

- 3z

ll(X

-+

ll

=.

8)

3y

:::-)

+ ll(O

Taking the values A -:. 5,

2x

'=)

-t

14

3z

8 - 0
+ 8

"' o

8) :. 0

-t

8)

::.

=)

-14;..

-l

5ll

we get

5(2x
9

24x

=l

Bx

3y
15y

5y

+ l4(x - 3z

- 42z
142

72

.24

::.

-1

8)

.:::.

=0

b) Dot product of normal vectors

(2A~ll.

-3A' -3ll)

( 2' -1 ' 0) and

of the planes must vanish:
2,{2A+ll) . 3A t 0 ::. 0 ..=;) 7A 1 2ll :. 0

Taking the values A .':- 2' ll

- -7, we get
2(2x - 3y - 8)
7(x - 3z
=)

=)

-3x - 6y - 21z - 72 = 0
.

2y - 7z

't

24 :: 0

8)

,_ 0

:::

174

c) Dot product of direction vectors

(2A~~.

-3A,

4,

and

of the line and the member must vanish:

-3~)

2 A. t

Taking

(1, 0, 3)

fl - 9 ~ =

=)

A - 4ll : 0

we have

4(2x - 3y - 8)t(x - 3z
9x- 12y

8)

-3z- 24.=- 0

C, INTERSECTION, ANGLE, DISTANCE

Intersections:
1 . Intersection of a line and a plane: Let th.e lineR. and

plane w be given by the equations:

x-x 0
y-y 0
z-z 0
R,;
-a- ~ -o- :. -c- = t

Ax

By + Cz +-D

(D:. (a, b, c))

+

=0

(N::. (A, B, C))

cases:

~~
]).
N:-o

If direction vectors
has
P0 tn

R. II n

D and

i n whi c h case either

R. e; n

N are perpendicular, one

or

R. n n

.:: 0 . .In the f i r s t case

andthe second P0 iTr

_;,-;- ........ -+

If

D.N

0, the line

inte~sects

.the plane at a

poi~t

,which is determined by setting the coordinates of any point

P(x 0 + at,
o~taining

y0

bt,

z0

ct)

of

a linear equation in t. If

It

I,.
I'

r:

1:'.

P(t 1 )

R.

is the required

int~rsection

in the equation of
t 1 is the solution
point.

w, and
th~n

175

Example 1. Find the point 6f intersection, if any, of

x-1

z+2

-r=-:,-=-,-=.t
x t-

TI:

{D.N :.

perpendicular

:..10

3z

--4-

2-2~3

Pe:R.

;
~

N = {1, 2, -3)

D ~- {2, -1, 1),

Solution. Since
-1 4

2y-

are not

t 0}

intersecting point exists.

=>

P{l + 2t,

{1~2t)

=.>

-3t

-2

t)

2{-t) - 3{-2it)

-t,

= 10

10

tl ::. -1 .

P{-1, 1, -3).

Example 2. Find a relation between

. x-1

"' -r =

l
z+2
a =- -,-- '

TI:

bx

a
+

and
2y -

such that

3z ~

10

a1 have no common point,

b) te:TI.

Solution.
a)

R-n7T ::: ~

_,

.....

,.\""'

p 0 {1, 0, -2}i1T

=> b l 4 ~

b)' 2a

2b

::.

2a
3,

-t

2b - 3

b - 4

'=)

::.

2b L2a

=)

=0

6 .;,. 10

a . .:. -5/2, b

-::.

4.

If the line is given in non parametric form as intersection

of two planes and .the plane by its general equation, the intersect i on

i s o bt a i ned by so 1vi n g a s y s t em of t_l}__r e e 1 i near e qua t i on s .

If the equation are given by vector'aJ one:;:
P

-A

tD

P ::: B t c:tU t SV,

176

the solution is obtainedby equating

equatio_ns in

P's

.obtaining three linear

t, a, 13. solution of. which gives

t = t

and the

.... A t- t o.
1
1
Solve Example1after transforming the'equations into i/e:toral

required point is

forms.
2. Intersection of two lines. Let the lines be given by
x-xl
y-yl
R.l: ~ -=-~ -

x-x
y-y
z-z
2
2
2
R.2: --a-- =- ~ =- c
2
2

-=. t

if

the direction ve~to~s o , 0 2 ~re parallel

1
cl
.
(~ -::. D::' -: - ) then either R. ::R. . or
R. n R. 2 ::. 0. The former
1 2
0
1
1 2
c2
holds when
P (x , y , z )e:R.
arid th~ la,tter when .P i.R. .'
1
2
2
1 1
1
1
If Ol, 0
are non 'paral.lel; either R.l, R. are intersect2
2
ing, ata. point or else
R.1~R.2 =--0. In the latter case
R. ~ R.
2
1

a1

b4

are called skew lines. To determine the point of intersection or

1 tb 1 t, ~ 1 tc 1 t)
in the equations of R. obtaining two linear equations in i. In
2
case of consistency (inconsis-tency) there is intersection, (skewskewness, one sets the coordinates of/

P~x ta t, Y

ness).
Example 2.
x-a

1: -,- -:.

~iven.

y+l

determjne "a" for

:.

z-1

:. t,

1':

R., 1'

a) to intersect at a pqint

A, and f i n d~

b) to be skew.

Srilution. The direction vectors being,non paralleli they

either intersect or are skew:
a) Setting
we have

x:. att,

j:. -1,

z-: lt2t

from~ into

1',

177

aH

-;a

.'
-1-3 . l-2t

:.

2t

.:

4-2a,

2a-t2t

~ =-~

l/2 "=9

:.

-2a-4t

1 t3'
t

:.

3/2

':;)

:::.

-l-2t

A(2, -1 ' 4).

b) a t l/2.
The same problem can be solved by the following technique:
I
/

Using

points

par~metric

P(a+t,
of R.,

R_

2s+3,

p(-~.

2t:+ 1)'

-1 '

-2s)

we have.

I '

"=)
::;:")

-s ,

a+t

:.

-2'

:.

-1

= 2st3,

t ,._ 3/2'

2t-tl

:::.

-2s

a ': 1/2

A{2, -l, 4). 8

Discuss the cases where lin~s <!.teg_iven by vectoral. equations,
and also by pairs of general equations

(fou~

linear equations
I

with

3.

unknowns).

thr~e

Inter~ection

of two planes. Let the planes be given by

their general.equations
( 1)

We have the following three ex~austi~e cases:

P,

r--------

..___ _
\. _ _ _-:q-2..

~ /I.....N,_

Al

Bl

.A2 1r2

Cl

s
(i ) '

01

.02

Al

Bl

A2 :. .Vi (i i )

The two planes coincide when (i) hold~ and ha~e no common

:j

I!

178

point when (ii) holds.

N/N 2 , the.two planes intersect

In the remaining case of

along a line

i.

The-direction

=(a, b, c) {Since
point

P0

on

LLN 1 ,

v~ctor

of i being

N1 xN 2

2 ), it will suffice to find a

as a particular solution of (1): Select two uni.LN

knowns having non proportional coefficients and assign any numerical value, say
P0

on

0, to the third unknown , solve then (1) to get

i. Ttien

which could also be obtained

rectangula~

(i~

parametric form) by solving the

systl"h (1).

Example 4. Fi1d the line of intersection of the planes

w2 : x i 2y - 2z : 5
w1 : 3x- y * z = 1,
in symmetric and the ~ector form.
Solution. Since the coefficients are not proporiional, the
plane~

intersect along a line

Coefficients of

x, y

being not proporti ona 1, setting

z :. 0, we have
3x - y ::. 1,

X -\ 2y
j

-1

,,t:

i'

1.

=)

I,~

= 1'

y ,. .2

==)

p0 (1' 2, 0)

= {0, 7' 7) //{ 0' 1 ' 1 )

x-1

y-2

--u- =- -,-- ::.

i:

ii

Jl

.:.

'z
T

If the planes are given by their vectoral equations

-w 1 : P -:.A 1

w2 : P

= A2

s1 u1

+ t

1 1

t s 2u 2 t t 2 v 2 J

then the vectoral equation

179

P :. A t

.11,:

tD

of the intersection is obtained by equating P 1 s

and

obtaini~g

s 1 , t 1 , s 2 , t 2 Solving three. of them

in terms of the fourth, and setting in one of the ~quations, one

three linear equations in

the equation of

obt~ins

1.

Example 5. Find the vectoral equation of the line of intersection of the

1T

planes~

1T:

( 1 1 1 ) 1 s(O, 0, 1 )

I :

:.

( 1 1 -1 )

-4

.S

..

0, 1 0)

t ( 1 0. -1 )
-4

I (

0 0 1 1)

SolUtion. Equating P 1 s
( 1 1,1)1(0, 0, s)-+(t, 0. :-t) :. ( 1 1 -1
=5>
~

( t. 0, s-t)

(t, -s 1 -t 1

=::.)

t:O,

:::::)

t :. 0
t

= (0, 0, :..2)-+(0,

( l , 1 , -1 )

O)f-(0, t

-+t I

I )

s-t-t 1 ) - (0. 0. -2)

-s 1 -t-1 :. o, s... t-t 1 .: -2 .
:.

0,

s-t 1

=.

-2

s 1 = -s-2

s+2,

P = (1, 1, -1 )-(s-+2)(0, 1, O)+(s-12)(0, L

P

..S I o

s 1 +t 1
....

)+(0~

+ s(

t)

0, 0, l )

Angfes:
1. Angle between two lines:
The angle

(t , t )

between two 1 i nes

obtuse angle between the lines

d1 , d 2

patallel to the given Jines.

Referring to the ffgure,
have

~e

1 , t 2 is the non
t~rough a point A drawn
t

I 0

I)

180

-+

vD~

-+

02) if non obtuse

-+

-+

-+

if obtuse

( -01 ' 02):(01-02)

The measure of the angle
will be denoted by .1(1 1 , 1 2 )1
equal to
-+

arccos
where
value

-+

-+

(1 1 , 1 2 )
which is

-+

I 01 21
-+ .
-+
'I 011 I 021

o1 o2 ar~: direction vectors of 1 1 , 1 2 . The

-+
-+
of o1 :o 2 is take~ since cosine of non obtuse

positive or

angle is

zefo~

Exampl~.

Find the cosine

X -8

11 :

absolute

----=tr- =l_::9 -

z
5

~f

the angle between the lines:

xty~z

12 :

-. 0,

2x-y13z = 1

Solution.
Of=- (4, 9, 5),

02

j
l

-1

:.

:::.

(4, -1 ' -3)

1 51
.
--...1164
cos9
,n
J~
, . -:. _,.,...,...:.__,......,..
71

<..<..
Y
/6T /13'
L.U

2. Angle between a line and a plane

The an g 1e ( 1 ,

between a line

1r )

angle between the line

i~

non

If

obtu~e~

dl. 1r ( dII N), a n d

=1(1 , 1!) I ,

9' = 1(1, d)l, wehave

g

=%-

and a plane ris the

and its wojection 1 1 on

Observe that the'angle de~in~d

.above

9 ~ cos9':::.sin9

7T: (1,7T);:;(1,1 1

).

181

Q =arcsin
-+

D is a directitin vect~r of

where

1:

3. ~ngle between two planes (dihedral angle):

\

'

The angle (n 1 , :rr 2 ) of two intersecting planes n 1 , n2 is

the angle. between the 1 i nes .!!. 1 , 1 2 of intersection of n 1 , n 2 .
with a plane perpendicular to
line of intersection

their

1.
I

where
Example 1. Given two lines
x-y+2z

-::.

y-2
2Z.-l
1 'X = -3= -=-r = t
2. T

and

R.l :

x;3y-z = -5
a) s.how that

t li:l,

1101 2

b) find the eq~ation.of the pl~ne n determined by R. 1 ~ 1 2 ,

c) find planes n', n" through 1 1 making an argle of
n/6

with
1

'IT.

Solution.
X-::.t,

a) Setdng
X -. y

..

weget

2z - 2'

whic~ s~tisfy

::')

1 't
I -' "2'

':

::.

-1 '

from

into'

12

Y. = -1 ' 'z

z + 5)

= -5.

b) Planes through

n(l l p):

2-t3t,

t - -1

3y - z

A(x -

'::.

R. 1

is

t 2z - 2)

(~+p)X t

p(x

-t

3y

(-A+3p)y t (21-p)Z +(-2At5p)

=0

\'

182
.-4

N{ n} = ( A_.~,

-A+ 3 ~, _2 A-~}

A direction vector of R- 2 being N(R- 2 ) =- (1, 3, -l/2),

and sinc.e R- 2 _intersects R- 1 , thenfor TI(~~)

....

_.,

tocon~ain

R- 2 ! one must haye

'

N{1 2 ). N(n} = 0

. iJ11p1ying
l(A'~)
=)

3(-1;~3~)

(2A-~)

"2'

=- I 2.. -~ :.

::: 0

7.~-=2

A=-

4 ::.

n:' 9 x - y -' 12z -

say.

~))I

1T (A'
/

=)

i3

-=

IN ( n) N( nh ll) 11
N(n) I IN (n (A,~)

I9(A+}l~+A- 3!! + 24A- 1211)1

/226 6A 2 - 8A~ + 1l~2

-=

(22f

,'6'1A~4 -~-8~c}
---- - Z'
A~
11 ~

134. - G k I

-f

_=

36~/k~

Y'Zlf

16

- l896k - 278 = 0

8k

-t

"1 '

~1

Example 2. Given a tetrahedron

B(1, -1, 0),

=- ~/!..}

which nas two real roots say

"2
k2 "= i l . These pairs of values
2
required planes.

A(l', 1, 1),

(k

llk2'

C(O, 1, 2),

and

ABCD

"z

IJ-2

Al
IJl
give the
kl

with vertices at

0(2, 0, -1), find cosine.

or sine: of the angle between

I

!I

:I
I'

a) the edges

[DA1

and

[DB]

b) the faces DBC and ABC

c) the edge

[oA}

and-the fa,ce

ABC

C.

183

Solution ..
a)

-+-

-+-

DA ... A-D ., (-1 , 1 , 2)

-+-

= B-D

DB

-+-

cos g = IDA DB I = h -12i =

/'6'.13
0"1\' I I D"B
-+-

b)

j k
-1 \2 2
1 -1

-+-

Nl :. BC x BD :.

-+-

N2

::.

BC

.i

j . k

-1

2
2

-+-.

x BA

:.

INl
cos g - N .
1

:.

(-4, 1.. -3)

=(-:-2'

2
1

( -1 -1 ' 1 ).

1 ' -2)

N21
5
>8+1+61
, "'v~
N2 =

-+-

-+-

IDA

N21
C) .sin g -
IDA I IN2 I

.:.

12+1 -'4\ .
:;
16

.a

.r

TJf:

Distan.ces:
1. Distance between

t~o

points
-+-

d(P , P ) =- IP P2 1
(1')
1
2
1
2
Analytically: d(P , P ) = /(x -x 2 ) 2 -(y -y ) -{zl'"'z )z (1")
1
2
2
1
1 2
whe.re. P Cx 1 ,_y 1 , z 1 ), P2 (x 2 , y 2 , z 2 ).
1
Vectorally:

2,. Distance of a point from a line:

Let _P 0 be a given point and 1 be a given
The distance
to be min d(P 0
obtainedwhen

.where

P)

d(P 0

for a11

P0 PJ..1,

1)

line~

~efined

is
P1;

which is

and

H _is the projection of

on

1.

'

"
1

184

a) Vectorally:

A.,.....,..;r;

'<:

b) analytically:

:d
I

H .

:b

.1-71"--~:- .f

-""'Po

. :.-.:._......;._Po-l-f\:'1--l.\

D
-!:

1Dx.AP 0 I... IDI~

+

d - 1P 0 HI

tzJ

is evaluated by (1")

where H is obtained by finding

d "' I DxAP o I

. (2)

th~

Iff I

intersection of

plane th.rough,

with the

P0 and ..L to 1.(2")

3. Distance of a point from a plane .

. The distance of a point
in

2. , and mi n i s a 1 t a i ned

P0

~hen

from a

~lane

v is defined as

P~ P1. v .
\

b) analytically:

+. +

IN.AP 0 I"'"lNIJAP 0 !~os -~

+

( 3 .)

l'o I

IP 0 HI

is evaluated by.()")

where H is "obtai ned by finding

IN ~ APol
. d -

t~e

intersection of w with the

line through P0 and L tow. (3")

4. Distance of a line and
distanc~

The

defined as min

d(1, v)

d(L, P)

zero when 1
If

inters~cts

R. II w

attained when

i {D.N

( D N, t 0 ) ,

'
LPl_w,
and

d(1, w):. d(~1 .w.)

~1here

Ft

E1_

between a line 1 and a plane.w is

for-all
+ +

a_plan~.

iLE1,

PEw

t 0).
mi n d ( L , P) 4

which is certainly_

185

a)_ Vectorally ( R./1 rr):

+

d(~, 1T)

b) Ana lyti ca lly (_RJ/rr ) :

-.

IN~ LlPll

(4I

+'

d(R., rr) :. d(Ll' rr)

IN I
1Ax 1

By 1 '". Czl ... 0 1 1 ( ")

4
~~ 2 .. s 2 i c~

where

Li (x 1 , y 1 , z )e:R..
1

-+

::.

5. Distance between two planes

The distance, d(rr, rr') ::min .d(P, P'.)
rr ,. rr'

is zero when
If

rr, rr'

intersect

rr// rr' (NxN' ::. 0)

is attained when

between two planes

(NxN' f 0).

the min d(P; P')

L!
.

PP1_rr. Then

, . ~?r'

:d

\ ..

~7t

P,

where
a) Vectorally (n// rr')
+

d(rr, rr') .::.

b ) An a 1y t i c a 11 y ( rr // rr ' )

IN.PlP11

(5I

d(rr, rr') :. d(P , rr')

P'~l,

IA'x 1

B'y

v'A

'f

C'z

B I 2 ..

+ D

c;=z=

--

( 5" )

6. Dista11ce between two lines:

+

Let the 1 i nes

If

t;l //

Rz

R. 1 ,

!1.

( D/12),

be determined by

dete.rmined by

= d(A2'

!1.1).

'

!1._1 f/!1. 2 (intersecting or skew).

Consider the plane rr through

.t

.t,

then

d(Al' !1.2)
Let then

R. 2 and

'7

Al' Dl' A2' Dz

2 and parallel to
d 1 // !1. 1
!1.

oAt

Jl: /----- - 7 - :J- /

/J.I
I

. I . /

I/

cl'
.:
I

,.

A,
.
I

Ji.

.186

Si nee

R. 1 I In,.

then

b) Analytically:

a) Vectorally:
(6 I
+

where

o1

d(R. , R. 2 ) ::. d(A ,.n)

1
1
where n is the plane

o2

~l.

Exainpl e 5. Given the tetrahedron ABCO with

0(2, 0, --1 ) ,

q0,-1, 2) ,

.I

a) .fin<'

D)
~)

c) find
by vector method,
e) Solve ParL

A( 1, 1, lJ,

LYA
1J

:I

,,

par a 11 e 1

through R.. 2
to

-B ( 1 , -1 , 0) ,

b) find

d{D, ABC) .

d) find

d(AO, BC)

by anantic method.

Solution.
a). d(A, D) ::. IA+DI -=- /{1-2}
+

2 + (l-0) 2

(1-(-1 )) 2 -=-

b) d(D, ABC) -::.IN}Ql

where

IN I

.Ae:ABC,

N = AB x AC:. (0, -2, -l)x(-1, 0, l):.(-2, 1, -2)

+

AD ::. (11 , -1 , -2)

1.::.2-1-\41
. d :. - 14 +I .. 4 I :..L
+

c) d (D, BC)
+

BC

=
-

( -1 , 2. 2) . BD

11 3

.+

BC

=-

IBC X BD I
I Btl'.

BO

:.

:.

( 1 ' 1 ' -l)

(-4, 1 ' -3)

/16. + 1

-1

g'

(6")

12613

/6

187
+

I N! BI

d ) d (AD ' Bc ) =

IN I
+

N _ ADxBC::. (1, -1, -2)x(-l, 2, 2):(2, 0, 1)

+

AB::. (0, -2, -1)

I_ 1 j

d-:;
e)

,..-

~'""

1/15

15

x.:l
AD . -=-r

y;...l
z-1
: . -,-= -r
-,
k
2
-2_

-:1

-1

::.

n: -2(x-l) -! O(y-1)

..:2x

d '- d(A, n)

-1 3
-::.

(-2, 0' -1 )

- 1 ( z>

:.

;. . . 1 -+ 31
l-2.1,___...---.14 t 1

.:

=- 0

1//5

D, SKETCHING OF PLANES AND LINES

A plane is sketched js general by

deter~ining

its intercepts

with co6rdinate axes or bydeterminirig its traces (lines of intersection) with coordinate planes.
Example. Sketch the
a)

plane~

'!tYi7-::..l

b) x - y t 2z - 2

solution.
a) The equation being given by
cepts are

a :.3,

b-: 1 ' '

interc~pt

C-::2.

b) We obtain intercepts as follows:

x-intercept:

y:.O,

Z::.O

y-intercept:

X:.O,

z::..O

z-intercept:

X.::.O,

y:.O

=)

0)

._.)

a:.2
b.:.-2
c :.1

form, the

inte~-

188

2,

A lin~ is. sketched in general by finding its ~races (P?ints

of intersections

plane~,

with two coordinate

or by sketching any

two of its points (or by the method given in Curve Sketching)

Example. Sketch the lines
. .x-1
y
zt2

a ) -,- ,"'"

. b)

X -i

2 =- -=T
:::.

2,

2z

-4

::.

Solution.
a) xy-trace:

z :.0

x-1
-,-

:::;:>"

. '

=lx

::.

=- ~-.:. -2
-1

y =- -4

A(-1, -4, o).

xz-trace:

y=.O

x-1

z-2

1""" :::. T. =- -=T

=)

=) X . ::.

1,

.Z

.-:: -

13(1, 0, -2) '

b) xy'"trace:

z:O .

=)

y =-. 2,

=)

:.

-1.,

x - y = -4
y :. 3

.A(-1, 3, 0)

xz-trace:

y.::-0 .:..:) x "' 2,

'==)

B(2, 0, -3)

'

X :::

2,

x t 2z
Z :: -

-4

189

EXERGISES (3, 2)
71 :

a) A{2; 1.. 6),

8(5, -2, 0),

C{4, -5, -2)

b) A{O, b, c) ,

B(a, (i), c),

C{a, b, 0)

72. Find the plane

plane
73.

3x

A{l, -1, 3)

~hrough

Which one'of the points

Given the 1i ne

and parallel to the

nearest to the plane

74.

ABC, where.

Find the equation of the plane

A{8, -2, 1);

+ 4y

- 8z

B{-5, 2, -1)
9 = 0 ?

= A + AD and plane -+N.P -+0 P = 0

a) find condition for intersection at a single

is

'

poin~.

b) find the point o_f intersection.

75.

Determine pairwise rel(\tive positions of the given lines

as

pa~allel

(concident.or non -intersecting), intersecting

or skew

a)

i6.

X-1
y . :
-z
= -=4 = U
.

'.

x ,_ y ,_ z=2
. -z-=,---,-

C)

Fi nd t 11 e e q 1J a t i on- of the 1 i ne pass i ng through

and intersecting

77.

X
y*3
z-1
-=,= -z o-

, , b)

4 = ~-1

orthogonally.
coeline~r?

Which three of the following points are

A(l, 2, -3),
. ,.

B{2, 1,. -1),

C{3, -2,_2),

D(3, -2, -3) .

:/78.

Find the equptions of the line t.such that

a) A<l.

3,

-2ht

. b) B( 2, 0 ,, 1 )E:t ,
. ~) ~(-2

79.

Given

1, l)El ,

A(l, 2,

the Hnes

AB

3)~

and

A( 1 , 1 , T)

t//{2, ,o, 1)
~I I

(1 ,,;3; -2)

B(l, 2,. -1 )Et

B(-2,7, 0),. find the
t: x4y-6

0,

z-2

?'

dista~ce

between

190

80.

IfA=<ar, a 2 ,.a 3-) ~nd6=(b 1 , b2 , b3)

vectors, show that" the lines

are twonon collinear

y-a
. 2

~
intersect at a single point
81.

P = A+AU>

Given lines

C.

= B+llV

p
thei~

a). find condition for

C and find

intersection

b) determine their intersection under(a}

82 .

Given

.A {1 , 2 , 3 ) ,

B( -1 , 2 , 1 ) ,

C( 4 ,

3 , 2) ,

D( 2 , 2 , 9) ,

find
- , b) d(D, ABC)

a) d(A, BC),

83.

Find tile distance between the lil)es

B(4, 5, 6),

84.

C(-3, -2; 1 ) ,

'-

D{1 , 1 , 4)

7TJ..'R. : A(2, 3, -4)

Show that the fourplanes

x+y~5z-8

= 0,

x+y+z-3

3x+3y+7z~4

= 0 are

= 0,

B{5, ' 1 , -1 )
2x+2y~z+l

= 0,

concurre~t.

Find the equation of the line

a) through
x-2y~z =

b) through
2x+6y+9z
8~.

A{1., -2, 1 ) ,

1T/ /7T I: 3x+5,Y-7z = 11

b) B{l, 5, 9):7T~

86.

if

Ft.nd the equ!ltion of the plane 7T if

a) A(2, -3, 5):7T

85.

AB, CD

A(3, -l, 6)
2,

2x+y-z

B(-~~ 4, 3)

and parallel to the planes

=5
and perpendicular to the plane

=0

Find the parametric equations of

a)

Z+5
~ = -!,- = 0
",

b)

X _ y_;2 _ Z
3- ,--r- =2

. 191

c) x+y-5z = 7
88.

Find the distance between

~:

89.

X+y-6 = 0,

Z=0,

t~e

skew lines

~:

P1 (1, 2, 3) P2 (-2, 7, 0)

Find the distance of p 0 (2, 3, 1)

y-2
if""

x-1

--,--

90.

d). 3x-y+z =- 5

from the line

= "2"

Examine f the point

where

face~

x-y+z

= 5, x-y-z

pl~nes:

are the

'
is inside the tetrahedron

P(l, 2,. 1)

X+y-z = 1,

X+Y+Z =3,

Find the equation of the perpendicular biSector of the

following
a) (ABJ

where

A(l~-2,

rcoj

where

t (-1 '

b)
92.

93.

2, -3) '

A(l, 1, l)JB"(l, -3, -1)

a) { P:

I~~~

2 .}

D( -3,

0, 1 )

find the locus:

b)

{P:

~ = l!

Find the di s_tance 'between two lines below:

and

b) X = Y+. 1\ '

95.

8(3, 0, -1 ) .

3 ). '

Given

a) x-y :; z

94.

of 1 i ne:

segment~

z =

Same

~uestion

~1:

-aX = cz cos a

~2=

X z
-a = -c cos

= 2y-l,

X+

XY+2Z+ -4 = 0

and

y = X+ 1 ,

1-z

for:

t = cz
t = cz

- sin a,

a +.

sin.

a'

sin a+ cos a
sin

- cos

a) Show that the equation of the common pP.rpendicular to

z-axis and the line
a X+ by = 0 ,

~=

x =

az. p'

y = bz+q,

Z ---~
_c. c.
a +b

b) th~ distanc~ between z-axis and~ 1s

=IU ~!1
a2

is

--------------------c-

192

ANSWERS TO EVEN'NUMBERED EXERCISES

72.

....,

=!= 0,

a) N.D

76.

-=-r

78.

a)

x-1
y-3 _ z+2
--r
= -a - -,-

c )

--r

b)P=A+

"N.o

x+2

y-1

= -,-

C .= AtB

82

a)J8li,

y+1 _, z-6 .

--s-.

b. )

x+9 _ y-4 ..;... z-3

-r-r -,-9-

88.

3/vff

90.

Out's ide

92.

2
2
2
a) 3x +3y +3z -6x+26y+Hat41

-,-- "::. :-y-

2y+Z+2

=0

z-1

b) 3x-:2y+3z=20

a)

~)

-:;: : 3 = --=-2

z-1
= --="2""

. a) 3xt5y-7z-:=-44,

x-3

x-2
b) . -,-

86.

94.

x-1 = y-1 .::- z-1

1 -,-

80.

84.

-N. AP

74.

193

3. 3. SuRFACEs
A,

RELATION~

A statement
(x, y)eR 2

and

be

<P

or

p(XI y, z)

zeR

a subset.

points is

AND THEIR GRAPHS-

involving variables

y,

X,

with

1s called an open statement which is true for

~3

of

f~lse in

and

R3 -s

where

may

3 .

IR ~

lxamples.

1. p(x,. y, z) : lxi+IYI+Izl
is true for all points

P(x,y, z)

in

RJ.

2 2 2
X +Y +Z < 0

2. q(x, y. z) :

is true for no point-in

tR

3. r(x, Y; z ): X+y-z = i

is true for all. points on the plane

'Ji; X+Y:-Z

and

false for all other points.

I

4. 's(x, y, z): X+y-1

~-z.

is true f-or all points on the -plane

for points,above
'

_7T,

-5. t(x. y. z):'x = 2,

1r:

X+y-1 = z

and false for points

y = 3,

z >

~elow

or
'IT.

is true fer all po.ints on the vertical line

x = ?, Y= 3,

,above xy-pl ane.

If

p(x, j, z)

is an open statement, then the set

{ (X' y' z) - ( X' y )ER 2 ' ZEIR'

consisting of all points

(1)

(X' y, z) . for which the statement

'

p(x, y, z)

p. (X'
y, z)}
I

is true is called a relation from

IR3), written

p(x, y, z)

R2

to

IR . (or . in

1.94
I'

where

p(x, y, z)

is the

rul~

D~={(x,

y):

for the relation.

The set
p(x, y,

z)}~

IR

and

Rt; = z: p (X , y , z)

J~~

are respectively called the domain and range of the relation

When a relaiion is plotted in 3-space, the resulting set

of point.s is -called the 9..r.2.Q.b of

) Depending upon the case either

-the set is empty (no graph) or the graph is a surface (for an

relation~

equality relation) or a curve (for two equality

or a

solid (for an inequclity relat{~n).

Examples.
Domain
i. { (x , y, z):

X
"2' + ~+Z=l}

-IR

surf~ce

"

~~-

z~O}

= y = z}

Surfa.ces:
Since a

2
3. {(x, y, z): X2+Y +Z 2 -~ 1

5. { (X' y; z ): x

m.t~

R2

2. {(x, y, z): Z.= 2}

4 .. {(x, y, z): . X >,:. 0, y>,O'

Graph

----..,(

'!,==!' ~:
ID"-'~
@r..

.i

)?~
~

'

is the graph of an equality relation, its

equation can be written as F(x, y, z)

0. For instance
2 2 2
. 2
.
2
and (X+Y) -(2y-z) = 3 are equations'of two surfaces.
x +Y +Z
Any relation
conversely any

surfac~

F ( x , y , z ) = 0 represents a surface , and

can be

including degenerate cases.

rep~esented

by such a relation

195

A .surface may as well be' represented by a sys'tem of three

equalities involving

t~o

parameter~

f ( u , V) ,

X =

and .v:

y = g( u , V) ,

Z =

h ( u , V)

which are called the parametric equations of the surface.

The parametric equations can be transfor.med to pr_evious
one by solving

u, v

from any two of them and setting in the

third, theoretically.
E~ample 1. Gfve~

2x- y

3z

= 6,

a) show that it is the equation of a surface,

b) find two parametric equations.
Solution.
a) Since it, is an equality relation involving
s~rface

represents .a
b) X=U,
z,

= -}

Y=V,

~ ~

x, y, z

(plan~)

(6-2u.v)

or

x = s.t,

s-t,

(6-s-3t).

Example 2. Given the system

2x-y.3z

= 6,

a) show that the system represents a curve r,

b) write the parametric equations 1of

r.

i f possible:

So'lution.
a) Since the system consists of two equality
inters~ction

represents a curve (line of

v

b) Setting
5

x;=u

and solving'

and

ref~~ion~,

of two planes)

z, we have

x = u.,

y = - 7 u.6,, z = - 7 u.4.
It is clear that any curve (system of two equality relation)

196

can be represented by a system of three equalities with one pa,r-ameter.

Linear family of surfaces:
Let
F(x, y, z) = 0,
be the equations of.two
F(x, y, z)

and

s~rfaces.

G(x, y, z)

;~.F(x,

G(x, y, z)

y, z)

+ ~G(x,

;~.,' ~eiR

tion, if they

intersecting.

Surface

t~e

linear combinations of

equated to zero, that is

are surfaces for

ar~

Then

= 0

y, z)

passing t'hrough the curve of intersec-

~~f:.tch i ng:

If the equation of the surface is given in the tmplicit

form

F(x, y, z)

the latter may be thought as ihree relations

(x, y)

t~

2) from pairs... (x, z)

to

3) from pairs

to

1) from pairs

having domidps on

(y, z)

xy-, x.z- and yz-planes respectively. o.f which .

the first is considered generally.

Visualizing the shape ofa surface is done in three steps:
1~

tietermination of the domain, say on

xy-pla~e.

2. Determination of symmetries with respect to cobrdinate

planes:
There is symmetry in xy-plane if
There is symmetry in xz-plane i f

F(x, y, -z):: F(x, y, z)

F(x, -y, z) - F( x, y, z)

There is symmetry in yz-plane if.

F(-x, y, z) - F(x, y, z)

Presense of symmetry with respect to another coordinate

197

plane reduce the work in sketching.

3. Determination of. cross sections of thesurface with
planes p~rallel to coord~nate planes.
The cross sect.ions parallel to xy-plane are called the
1 eve 1 curves of the surface .
th~

xy-, xz- and

By

the

mean~

bo~ndedness
)

cmd ones lying on coordinate p 1 anes

yz~traces.

of cross

s~ction~

one gets the. information about

of the surface.

Example. Sketch the surface

S : 4x 2+/-z =

Solution.
l. The, domain in

far a 11

. '

xy~plane is

(x, y.)ER . Having

z.~ 0,

IR

since the relation holds

the surface lies above xy-plane

an~ passes t~rough the .origin.

2. Since the relatfon is uin.altared \</hen

y

~y,

suffice

'-x

anq

there are symmetries in the yz- and iz-planes. So jt will

' ~

t~

sketch the surface on the .I. quadrant over

xy~plane

and get the whole surface by.taking symmetries.

3. Cross sections II xy-p 1 ane {level curves):
2
z = k ~ 4x 2+y = k (ellipses for k> 0' origin for
cross section-s II xz-pl ane: y = k -;:> z = 4x 2;k 2
'4k 2
.
r { para ba 1 as ) ,
cross sere t 1ons
,,''! yz-p 1 ane: x = k ~ z = y 2+
(parabolas).
' There are other coordinate systems in 3-space which are.
\

more convenient than rectangular one.in some cases. The following

are two examples of such systems.

198

B, CYLiNDRICAL AND. SPHERICAL COORDTNATE SYSrEMS

1. Cylindrical coordfnate systems (Pol~r coordin~te systems
in 3-space).

A rectan[ular coordinate system

Oxyz

in which xy-plane

is taken as polar plane (with 'x-axis as polar axis) is called a

cylindrical coordinate

~ystem.

Cylindrical coordinates of a
p o i n t . P a r e.

r, z

Q,

and one

P(Q, r, z).

writes

when the coordinates of a

point
P

P are given, plotting of

is done in the order

z, whi)e having a point

Q, r
P

and

its

coordinates are ribtainedin the

order

z, r, Q.
Transforming relatibns.
The~e

X,

y,

and

are the relation between cartesian coordinates

C

y 1 i1 d r i C a 1

= r cosQ

r sinQ

ne S

Q, r ,

f an arbitrary p0 i n t

Q = arctan lX
or

z =z

r = /x2+Y2

z = z

Q-constant surfaces are planes through

surfaces are circular right cylinders with
z-constan~

surfaces. are

horizo~tal

planes.

Oz

z-axi~ r-consta~t

as axis, and

P:

e-~ons

r-constant surfaces

tant .surfaces

z-constant surfaces

2. Spherical coordinate system.

In a rectangular coordinate system

Oxyz,

taking

xz-plane as initial polar

plane . n 0 (wJth Oz as polar axis)

and~

plane through z-axis as a yariable polar

pl~ne

t~e

(with Oz as polar .axis), then

spherical coordinates of a point are

angle

from

and one writes

9E

(0, 2n] ,

P(9,

~efined

by the directed

to

~.

~-

} where we impose the restrictions

'f'E: [0, n]

and . ~ ~

n),

<:{',

on

0~

P(9, 'f,

Plotting of a point
(determining the plane

and the polar coordinates

'f and

'f)

is done in the order

~(on n )'.

Transforming relations:'
These are the relations between the cartesian coordinates
.x, y, z

and spherical ones of an arbitrary point

P.

the projection of

P. on xy-phne, -fYom P' (9: r)

and

one gets
x =1'

sin~.cos9

y::::) sintt>.siri9
~-=- ~- cosc.r

9
or

= arctan

'f = arccos

f
z

.! c.
f..
z
VX +y +Z

r = vx/r-:.2,-----.,2,.----:2
+y +z

If

P'

is

r:~Sin'f

200

9-constant surfaces are

plane~.through z-axis~

surfaces are circular rihgtcones wit.h_ Oz

surfaces are spheres with center at

~-constant

as axis, and :f-constant

0.

--. -; ,__,
i

'f-constant surface

9-constant surface
(~

s-constant surface

(a cone)

plane)

..

(a sphere)

We have one more set of 'tran'sforming relations among three

coordinate system which is between cylindrical and spherical. one,
namely
9

r ,;, f s i.n <e

=f

=9
i'= arctan
9

9
or

cos~

lrV
7-

r
z

C, CYLINDERSr CONES, SURfACES OF REVOLUTION

1. Cylinders:
A surface
direction
~ecting

generated by a variable line

of given

A, and subject to another condition such as inter-

a curve

r (or remaining tangent to a given surface 4 )

is called a cylinder
. The line
A

is the generatrix,

r the directrix

the direction, and

of the cylinder
in defined by

S, and we say that

A and

201

of a cylinder

Equ~tiori

by~~

The equation of the cylinder.defined

a)

r:

x = f(i'),

b)

r:

~(~. y, ~)

z~h(t)J

y = g(t),
0,

(a, b, c) and

G(x, y, z) = 0.

Solution.
a) Since genera+n.,' fhas directio.n numbers
passes through

y-g(t)
b

x-f(t)
a

and

one has

(f(t), g(t), h(t))

- s...

a, b, c

z-h(f)
c

as the symmetric equations of the cylinder.

When

is eleminated one obtains

r(

s: H(x, y, z) = 0
b) If

P(x, y, z) . is any point on

other point Q on

Taking

Q on

r.

R.

R.

(a point of . s) any

is

one has
0 -1

and. the

e1i'mi nation

of

S: H(x,

gives

y, zr = 0.

Note that cases (a) and (~) are reducible into each other.
Example. Dbtain the cartesian equations of the cylinaers
defined by
a) 6

(-1, 2, 3),

r:

X=

b) 6 =

(-1~

r:

2, 3),

t'

Y= t2'

2. 2z,
. y-z = 1
=

z=0

202

Solution.
y-t 2

x-t

a}

-::-r = -:-:r = 3

,,l

=9

6x-6t

-3y+3t 2

3t 2

:::!)

6x-6t

-3y+3y-2z

--=>

=)

3y~3{x-z/3}

= -y+t

= x-z/3

= 0.

b} Taking : Q(x-t,
y+2t,
I
.
(x-t) 2

= 3y-2z

= 2z

9x 2+Z 2-6xz-9y+6z

=J

i>

3~-3t 2 = 2z

2x'-2t

!'
I

==>

on

r one has
/

2(%+3t),

::,> (x-t) 2

2(z+3t),
t = y-z-1
~ (X-Y+Z+l }2 = 2,{Z_+3y;..3z-3)

2. Cones.
,

'

surface

generated by a variable line 1

through a fixed point

a~

intersecting

passing

and subject to another condition such

. 0

r (or remaining tangent to a given

curve

surface Z:} is called a cone.

The line 1 is the generatrix,

the directrix and

of the cone
is
:,.,'
i~

l,i
',

H
L,

defi~ed

P0

the verte'x

S, and we say that

by

P0

and

r.

Equation of a cone
. The eguat.ion of the cones defined by P0 (x 0
a}

b)

= f(t}, y = g(t),

r:

r:

F(x, y, z}

= h{t}

~------~~~~--~~~--~--~~~-

0,

G(x, y, z}

y ~

and

203'

Solution.
a) Since 2 passes through
are

x~-f(t) ~

y 0 -g(t),

[t~ dir~ctiqn

and

numbers

z 0 -h(f) ,.. one has

z-z 0

as the

~ymmetric ~quations

When

of the cone.

is eliminated one obtains

=0

S: H(x, y, z)
b) If
qf

P(x, y, z)

is any point on the

S), then any other point Q

on

Z+ tz
0
l+t

Setting these

on

F:O,

and. eliminating

G=O

1 (a point

has coordinates

y+ty.o

generatri~

one gets'
;iI

,,,,,,
I,

S: H(x, y, z) ;. 0

I;

Example! Obtain the cartesian equations of the cones

'

!:

defined by
a)

p.o ( 0. 0, 1 ) '

b)

p 0 ( 0, 0, 1 ) '

r:

X=

2t,

..r

x2

y =t

2z,

y-z

So-tution.
a)

,,,,

y
--z
t

--=-!

=)

t 2x

2ty'

=)

tx

:::)

-X

z-1
-x

2y,
-x
"
4y ( z -1 )
X

2 +4yz-4y

2t(z-1)

2t(z-1 ),
-x 2.
=)

4y(z-1)

= 0

1.

204

b) Having

0
Q(X+
T;t

Y+
hl

Z+t)
l+t

F:O,

G:O:

setting its coordinates on

-::).

=>
~
-')

=)
""")

o_n the generatrix, and

x2

2(1+t)_(Z+t),

y-(Z+t) = l+t

x2

2(l+t)(z+t),

2t = y-z-1

x2

(zty-z-l)(z +

2x 2

(Y-Z+l)(Y+Z-1)

2x 2 - y 2-(z-1) 2
2x 2-y 2+z 2-2z'+ 1

l:fl)

0.

3. Surfaces of revolution:
A surface

generated when a curve r is revolVed about

a line A is called a surface of revolution.

is the generatrix and A the axis of the

surface, and we say that

is defined by A and

r.

Spheres, rihgt circular cylinders and

cones are examples of surface of revolution.
Every point of r describes a circle with the center on the
axis

A4 called a "parallel", and

the surface

~(ong

~ny

plane.through A intersects

a curve-called a "meridiah" of

The meridians are congruent curves of

generated by

revol~ing

any meridian about the

S.

S, and
axi~

S may be
A.

An equivalent definition of a surfaie of revolution is the

following: A surface of rev6lution is the locus of a variable
circle with given axis A and subject to

anoth~r c~nditioh

such as

205
intersecting a given curve

r (or remainirig tangent to a given

surface 2.).
Equation of a surface of revolution:
The equation of the surface defi ried by

r: x = f(t),

y = g(t),

6.

"'

z-axis,

h(t)

Solution: The equation of the. "parallel"

-circle- through
X

where

r2

P(f(t), g(t), h(t)

2 2
2
+Y = r.,

being

z = h(t)

f 2 (t)+g 2 (t), we have the parametric equation

Eleminating

between the two relationSy one gets

S:

Example. Find the

F(x,y, z)

ca~tesian

equation of the surface generated

by revcnying.
r : x ~ t ,

= t- 2 ,

z=

about z-axis.
Solution.
.X

=)

=)

=)

=)

+Y

2
2
2
+Y = t +(t-2) ,

z =t

2
2
+Y :2t -4t+4,

z =' t2

2
+Y = 2z-4t+4,

z =t

(x 2+Y 2 -2z-4) 2 =16z.

206

Particular case:
Parti~ular

case~

of cylinders, cones, surfaces of revolutions

are obtained
by taking
the generatrix r nn a coordinate plane.
.
I

f1.

'X

T: f(x, y)=O, z=O

g(x. z)=O, ,y=O

II z-axi s

6.:

h(y, Z)=O, X=O

II x-axi-s

lly-axis

Let the directrix r be on the xy-plane with equation

f ( x , y) = 0 ,

z=0

and d i recti on 6. II z- ax i s . Then the e quat i o rt of

this cylinder is
f(x, y)

Similarly. g(x, z) = 0, and

(for any

z).

h(y, z) = 0

are the equations.

of cylinders having directrices. on xz- and yz-planes and genera trices

'

parallel to y- and x-axis.

These particular cylinders

ar~

right

c~linders

since

genera tri~es are perpendicular to the plane of the directrix.

Observe that .in cylindrical coordinates,
represents a right cylinder with

gene~atrix

F(B, r) = 0

n z-axis.

represents a ~ihgt cylinder having

G(r coia, z) = 0

= 0, and H(r sinG, z) = 0 represents

a right cylinder having directrix o~ the plane 9 = vl2 .
directrix on the plane

207

F ( g, r) =0,

z =0

G(r cos9, z) = 0

, H(r sin9, z)= 0

2. Cones

The

e~uation

of the cone with directrix

r:' (f(t), g(t), O)

z-i 0
which yields
S: F(x, y, z)

when

is eleminated ..
Similar

equati~ns

are obtained when r lies on xz-plane or

yz-plane.
Exampl:. Write the equation of the cone
with Vertex at
~: (y-1)

2+Z 2

(h,O, 0)

and di.rectrix

l,.

X=

0.

Solution. Writing the equations of r in parametric form:

r: x = 0,

y = 1 +cost,

z = s i nt

we have
x-h .
s: n-o
or

z-0

StT = ().:'5 1 n t

""0--...,..(,. ,. ;. +=!.. c-o

208

x-h

l+COSt = -

=)

---n-

- (1 + c o s t )
hy

sint

= - s 1 n t

..::::)

hy
( - - - 1 ) 2 + (- _!13_ /
x-h
x-h

-)

. 2
2
2 2
(x-h+hy) + h z = (x-h)
(x+hy-h) 2

:::::::)

hz

-~

(x-h) 2 + h2z2

'

3. Surfaces of revolution.

If the
equation

generatri~

f(y, z) = 0,

axis of revolution
P

X=O

(or

t., then as

~z-plane

f(x, z),

:r

(or on
Y=O)

xz~pl~ne)

an~

Oz

with

is the

revoLves about z-.axis, any poinf

on r' describes. a circle

X

r lies on

2
2
+Y := r ,

Z = Z

with center on z-axis and

I'
:

radius

r, where r

satifies

f(r,Z)=O ..
Then ~li~inating r
2
between x2+Y 2
r
and
f(r, z) =

o, one gets the.required.equation

' /22 z.) = 0
s: f(l'x>y~,
I

of the-surface of revolution

S,

The other similar cases_, wher.e t. is on .Ox

treated

or

Oy 7 are

~imilarly.

Example. Given the curve

r: Z=Y

2 .

on yz-plane, obtain

the equation of the. surfaces of r.evolution when r is revolved about

------~---~----- ------~-

-~--~

-~-~~--

209

a) z-axis

b) y-axis

Solution.
b)

a).

2 2
2
X +Y = r ,
-:::;>

+Y

X +Z

r
2

=Z

D. QUADRICS

==;l , X +Z

. 't

=y

(In standard forms)

The foliowing standard equations of second degree r~present

surface~called

a, b,

qtiadrics

~ith

certain

oriehtations~

where

c>O.,

a) Ellipsoids

(and spheres)

.X2

z2

--z+:-z+--z=1
b) Hyperboloids
x2
-:-7

a.

'

/ . ,z2 .
+ -:-7'--2'-' = 1,

(~yperboloid

of one sheet)

-~-;. +~
a
b
c

(Hyperboloid of two sheets)

c) Paraboloids:
x2

--z
a

2
x
-:-7
a

= AZ

(Elliptic paraboloid)

l
:Lb.

AZ

(Hyperbolic paraboloid)

1.

210

Sketching:

. 2

a) Since in

the terms are non neg~tive,

:-z+-z
b
c
~

i t follows that lx/al

ly/bl~l.

1,

-b~y,b,

. -t,zEc
Setting

and the surface is bounded.

z = k

( lkl

~c),~

one has
2

y2

-z+:-z
a
b
with
showin~

a=~c 1~ 2 -k 2 ,

a= cb

that cross ~ections parallel to ~y~planes are ellipses.

Similarly cross

sectio~~

parallel to xz- and yz-planes are

ellipses~

Then the surface is called an ellipsoid.

a, b, c

are the semi exes and

the center of the ellipsoid.

If

a=b=C, then the crOSS sections

are.all circles,and the surface is a.

Ell-ipsoid in the

sphere.

r\.

octant

If two ofa))1.c, are equal the ellipsoid is of revolution

arid called a sph@roid, which is oblate (prolate) if the third
axis is smaller (larger) then

th~

se~i

other.

T,he equation
(x-h) 2 . (y-k)2
2
2 +
b

(Z-.Q.)2 - 1

2. -

represents also an ellipsoid centered at

~t

(h, k,

orientation as previous one.

2

b)

x
/
-z
+ :-z
a

z2

--z =
c

(H 1 ),

2 .

---z-~1-~
X

with the same

211

The cross sections of

x=k

o~

y~k

for

HJ

H2 are. hephe~bolas fdr

z=k. Since iross sections are

and

and ellipses for

hyperbolas in two ways, the~ are ~alled hyperboloid~ T~~ cross

sections in

are ellipses for

H
1

Z=k

for any . k and the. surface

consists o'f a single piece (sheet), while i:i

H2

the sections

are ellips~s only for lkl)c. Hence the surface tonsis~s of two
disjoint pieces (sheets). Accordingly. H1
of one sheet,

hy~erboloid

is called a hyperbolo-id

of two sheets.

)-

l)t

A Hyperboloid of

hype~boloid

two sheets

one sheet
a, b, c

of

are semi-axes and

They ar-e of revolution when

0 the center of the surface.

a=b.

Similar surfaces are obtained for other combinations of

The_equatioris
(x-h)
a2

2 .

(y-k)

1'

b2

_(x-b) (y-k) t(z-R.)

2 2
.
2
a
b
c

represent clearly hyperboloids with the same oriencation of axes

but center at
c)

(h, k, R.).

Co~sider

x2

-z
a

the equations

Y2

:7 =
b

).Z

.(EP),

(HP)

_ 1. /
-

212

The cross sections of

parabolas and

EP

and

HP

ellipses (hyperbolas). for

for
z=k

x=k
in

or

Y=k

are

EP (HP). Since
I

sections are parabolas in two ways are called paraboloids. The

third cross sections in EP.{HP) being ellipses{hyperbolas) they
are respectively called elliptic parabaloid (EP), hyperbolic
paraboloid (HP).

,\)o

An elliptic pa~ab~loid

A hyperbolic paraboloid

. The origin in, H 1 and H2 is called the vertex. The

hypirbolic parabol~id is of saddJe shape in the neighborhood of
~h~

origin and the origin is called the saddle.point of

face, and the surface

H2

is sometimes called a saddle

the.sur~
sha~e

face.
Similar r&sults are obtai~ed when
instead of

or

are linear

z.

The equations
(x-h)2 + (y-k)2 = :\(z-R.)
' 2
- . 2
a
b
represent clearJ.y paraboloids having vertex

at

Exam'ple 1'. Sketc'h

a) the sphere'

(x+l')

2 '
+

(z-2)

'
4

(h, k, R.).

sur-

213

4x 2+y.2 +~ 2 .

b) the ellipsoid

16 '

Solution.
a) The center is at

(-1,. 0, 2)

to 2.

and radius

b) Writing it in the form

one has
a=2,

b=4,

c=4

and that part ,in the I.' octant

is shown in the figure.
/

')c,

Eiample 2.. Sketch the quadrtos:

2 . 2
z2
x2
a)

T - ~ - 4' =

b) T

- z

2y

Solution."
a) The surface is a hyperboloid of two sheets. with semi
axes
Ox

a = 2,

b = 3.

= 2 admitting

as axis. Cross secti-ons Uyz-plane

.-cease to.ext.st when -2<x<2.

b) The surface is a hyperbolic

~:r~-
~~

parabol6i~

(or a saddle
~

shaped surface).
x=O

=0

::::9
~

y
y

z2 .
(a parabola)
2
x 2/8

(a parabola)

~.-SECOND DEGREE SURFACES,

The quadrics being

ar~

seco~d

degree surfaces, their equations

included in the general equation

214

A1x2 .+A 2y 2 +-A 3 z 2 + B yz.+B 2zx.+B 3 xy

1
+ c1 X +C 2y + c3 z + D
of second .degree' where the terms with coefficients
and
' :

Ld
' '

( 1)

A;, Bi, C;.

D are called quadratic terms, cross terms, lineaf

ter~s

and constant term respectively.

. If (1) does not
squares, it can be

If

the~e

.involve~

transforme~

is only

o~e

cross term, bj completing the

into

th~

standard form:

cross term, say

xy-plane about z-axis, (by a proper angle)

B3 xy,

a rotation of

elimin~tes.

the cross

terni:
'I
I
- DI
AI lx I 2+.A I 2Y I 2 +A3 2 2 +clx
+c I 2Y I +c3L+

=0

If more tha~ one cross term are ~resent, the elimination. . of

them can oly be performed by a rotation of the. system Oxyz about
a' line through the origtn.
The full discussion of reduction of (1) into a standard

f. '

form (including degenerate cases) is given in LINEAR

l,i

writihg (1) .in matrix form as follows:

ALGEB~A

by

(1) is rendered bomogeneous by introducing a variable of

homogeneity

t:
A1x 2+A 2y 2+A 3 z 2+B 1yz+B 2 zx+_B 3 xy

..

+C xt+C yt+C zt~Dt

!l \
i

!
~

i .
[.;

..

Next one denotes X, y z

and sets

and

by

(1I

= 0

xl, x 2 , x 3

and

x4

215 .

with'

Al

all

A2

a22

A3 = a33 .

Bl

2a23

B2

2a31

B.3

2al3

cl

2al4

c2 = 2a24

c3

2a34

a .. =a:.-. Then ( 1 I
lJ

Jl

D = a44

. I

becomes

2
all xl

.2.
2
a22x2 + a33x3

2a23 x2x3

2a31 x3xl

2al2 x,l x2

2al4 x.lx4

2a24 x2x4

2a34 x3x4

2
a44 x4

or
2

al) xl

al2 xlx2

+ a2lx2xl

2
~ a22 x2.

.+

a31 x3xl

+ a4J

x4xl

al3 xlx3

al4 xlx4

a23 x2x3
2
a32 x3x2 + a33.x3

a24 ~2x4

a34 x3x4

,.

a42. X4>\2 + a43 x4x3

(1

II )

2
a44 x4 = 0

which is seen to be the same as

'

rkl x2 x3 x4]

all

al2

al3

al4

xl

a2l

a22

a23

a24

x2

a 31

a32

a33

a34

x3

a41

a42

a43

a44

x4

=0,_(x 4=1) ( 2')

where the matrix [a .. ) is symmetric.

lJ

When (2) is writ~en in the original notat~on, we have(for

x4

1)

'.

216

A] ~

1
2 82

1
2 C 1

.X

A2

1
'2' 81

1
2 c2

y.

1
'2' 83
[x

'1]

2 82

2 cl

-r

1
2 83

1
2, 81

.- 1
2 c3

A3
1

(2I

2 c3

2 C2

-=0

It is proved that the surface represented by (1) or by (2)

is a quadric

if

det(aij') ~ 0, and degenerate .if

det(aij) = 0,

or (1) is a qua~ric or degen~rate according as

2A
. 1

83

8 .
2

cl

83

2A 2

81

c2

82

~1

2A 3

c3

cl

c2

c3

20

is not zero or'zero.

Quadricsare non degenerate
and

s~ts

~urfac~s,

of two planes are degerierate

while cylinders, cones

~nes

Example 1. Transform the following into standard forms and

test them for, degeneracy
2.
4z
a) 9x - 4y 2 + / + 18x
y 2 + z2 - 2xz. + 2x
b) x2

13 = 0

- 4y -

2z

Solution.
~) Stnce ihere are no cposs terms, standard equations are
\

obtained by completing squares:

t'

,I

217
2 .
2
2
{9x +l8x)-4y +{Z -4z)+13

:=l> 9(x +2X)-4i+{Z-2) -4+13

'2

=)

9(X+l) -9-4y +(Z-2) +9

-~

2
2
2
9 { X + 1 ) -. 4y + { Z - 2 )
= 0

0
0

-,

(corte, vertex at (-1, 0, 2))

Since

--

18

i8

-8

-4

t8

-tt

2G

the cone is degenerate,

- b) Since there is only one cross term, namely --2xz,
standard equation is obtained by rotating Ox.z about
per angle

t~e

by a pro-

9:
2
2 . 2
(X -2XZ+Z )-j +2x-4y-2z-3 = 0
-2
tan 29 - 1-T

co

=)

9 = Tr/4

( l)

=)

cos 9

12

Z=z<x~.z~).-

sin 9 = ,ifi2
Y=YI

Setting these values in (1), we have

12 ( X

) -

4y 1 -/2 (X 1 + Z 1 ) -3

(Y 1 +2) 2 = 0

{two intersecting planes)

218

as in (a)

there is

degen~racy.

Example .2.
_a) Find the locus. of points equidistant from the point
F(O,- 0, 2)

and the plane

z = -2.

1T:

b) Sketch the obtained locus

Solution.
a) Let P(x, y, z) be equidistant from ,F(O, 0, 2) and
1T'

z = -2. Then

=>

"2 2
2. 2
i
2
+Y +(Z-2) = ( Z+2) :) X +Y = Bz.

(EP)

,,
b) Domain: J.flz

"I:races:

xy-tr.ac;:e:

.Z=0

=)

xz-trace:

Y=O .

=)

yz-trace:

X=O

~)

2
+Y
2 .
X +8Z
y 2=8Z
X

(origin)
(parabola)
(parabola)

cross sections:

II

xy~p 1 ane

Z=>k

II xz-,pl"ane

Y=k

II yz-plane

X=k

-J

96.

P}ot the

followi~g

2 2
X +Y
2" 2
X +k
y 2 +k2

(circles fo'r k 0)

8z

(parabolas)

8z

(parabolas)

;...

EXERCISES (3. 3)
points in a cylindrical coordinate system:

A( 1T I 4 , 3 , 4 ) , . B( 1T/2 , 4 , 1) ,
97.

C( 0 , - 2 , 1 ) ,

D( 0 , 0 , 2 ) .

Plot the following points in a spherical _coordinate system:

1

A(1TI4, 1TI2, 2),

B(1TI2~ 1T/4, 3),

~(0, 1TI2, -2)~ . D(1T~4, 1TI4, 1)

219

Sketch each of the.surfaces given by equations in spherical

98.

coordinates:
. a) S'~ 5'
99.

S.~etch

bf'f= 21f/3,

the

100.

~here

d)

s = 2 <;osf.

-surfaces:

2 2
a) X +Y = 36
Is

1T/2,

c) 9

b) y +4z 2
dege~erate

any

2
c) x -z

0'

= ]6

one?

S~me ~uestion.for:

' ..

a) x +16y -.4x = 0

X 2 +Z 2 -4y-::

b)

0,

_c) x.2-y2 .- z2

101. Find the pro-jections of the f,ollowing curves on the coordinate planes:
a) Z

.
?

2. 2

+Y-

4y

c) X2 +y 2 +Z 2

4,

b)

+Z 2

9'

+Z 2 = 4

X+y+Z = 2.

102. Find the equation of the cylinder with directrix r and directiori 6 gfven below:
a)
b)

z2 +2x

r
r

x=2+t;

8, y=O,
2
y

6 //

=t

= t

y-axi s

+1 ,

!!//

X- a Xi S

~.

103. Find the equatiori of the cone with given vertex V

and

directrix r :'
2

2.

a) V(O, 0, 0),

r:

b) V(3, 1, 2),

r : x = 2+ t ,

+Y

16,
y = t

=2

=t 3 + t.

104. Find the equations of the surfaces of rev o 1u t io n ~ i t h given

. generatrix r and axis

a)
b)

6:

r:

y2

4x-16,

6:

x-axis

r:

y2

2pz,

6:

z-axis

220

r:

c)

tJ.: y-axis

--

105. Find the locus of points equidfstan~'from

a) (0, 0, 0)

and

b) (3 , 1 , - 5 )

n: z = -4

and

n:

x +2y- 2 z

106. Construct arid distuss the surfaces:

2

a) z =16
-

d) 4x +Z

.2

f) 16x 2 -4y 2-z 2= 0

e).y-9z=O

16

-2
2
c) 3y + 7 z = 0

4/-25 = 0

b)

107. same question for:

2

a 1 4 x2-+ 9/ + 1 6 z
-4. 29 0
2
C)

108,

.X

Con~truct

a)

t~,

2 +y 2

_a

2
b) X +y 2

2 . 2-

P) 9x -y +9z

144

d)

= j6

.x 2 +Y 2-z 2 = 25

in the I. octant, bounded by

Jli~,

144

z = 2mx(a>O,- m>O)
2
-2ax, z=o
X2 +Y
(a>O}

'

az,

1 09. Determine the reladve. positions of the line and quadric

X+6

.1'. : ~

y-6

z-3

= --=-z- = -=-r '

= 3-i-4t,

b)

.Q.:

-'c)

2:

d)

:x=6t,

b_

=..t
2

'

Q: x 2 +y 2 +4x 2

16

z = -2t,

Q: x'2 -z 2

2y.

~ = z- 5,
y=9+3t,

z=l-2't,

Q: y 2 +4z 2

8x.

110. Write the following in 'standard form:

2

a) x +Y +Z -4x+2y+6z-9 = 0

. 2 .

b) 4x -4y +16x+8z

_111. Same qlies ti on for:

2 2 2 .
a) x +y +Z +2xy = 0

2 2
b) y -z -4xz

Q:

221

112. Write

2
2
2
.
x +3y +2z +2Xy+3X+2y+1

=0

in matrix form.

113. Show that the fo.11 owing are degenerate second degree surfaces:
2
z2
x

a)_,.-~-

at:

b<'

---z
=0
c

c)

114. Find the surface through the curve of intersectian bf

z = x 2 +2/
3x+4y = 0 and passing through the point (0, 1, 4).
115. Write the equation of second degree surface passing through
the nine points:
(0,-0, 1),

(1 ~ 0, 0),

(0,. 0, 0);

(0, 1, 1),

(1, 0, 1),

(0, 1, 0),

(1, 1, 0),

.(1,

1,

(1, 2, 3)

ANSWERS TO EVEN NUMBERED EXERCiSES

96.

98.

a);~

bJ

<)

k J)_

~45.~/

100 .

~~

~~-~~

102. o)2x+z 2

8~ ,b) ~=(li-~J

2 2
104 . . Y+Z

4x-16,

2pz .

'Y

1),

222

106.

108.

,.,.
2
2
2
11 0 ~ a} ( x- 2 ) . +. ( y + 1 ) + ( z + 3 )

--b)

2
(x+2) - /

- ...:2z-4

3/2

1]
0

3/2
2
2
.
114. 2x +4y +3x+4y-2z

--

112.

[x

23

0.

_0,

non

deg.

..

3. 4.

SPACE CURVES

;(of one variable).

A. VECTOR FUNCTIONS

If

is an fnterva 1 on a t-axis, then the function

-+
.3
r: I -+ R

is called a vector function of one-variable from' reals to vectors

i, n 3 - s p a c e

-+

The vaJue.of
-+
r(t)

has scalar

at

com~onents

-+
r(t)

-+r(t.). If the position vector

is

td

x(t), y(t), z(t),

then

= x(t)i + y(t)jc+ z(t)k

or
-+
r ( t) =~ ( x ( t) , y ( t) , z ( t) )
For ihe_furiction

~(t), limit, derivative, differential,

continuity and integrals are-defined as fo.ll ows:

'i

-+

1 i m . r { t)

t-+ t 0
-+

'

d
<IT r(t)

= L 1 im x(t) ,.
t-.t

r< t

1i m
.6. t -+ 0

. <atd
-+
dr(t)

x <t),

r< t l

t,.t) t,.t
d

cl\-

<IT y(t),

dt,

<x.

d y(t),
Y dt,_

y,

d z(t))

i dt)

i)dt

The notations .for the firs.t derivative are

d
'err
.4>

r(t)

r(t),

-+
dr

z(t))
- \

(d x(t),

<x

1 i m y.(t)' lim z{t))

t-.. t
t-+ t 0
0

(ff ,

Dt

-+
r (t) ,

is said to be conti.nuous at

-+
r ( t)
t

if .

224

-+.

-+

= r(to> =

1 im - r(t)

' t .+ t
or i f

x(t),

z(q

y(t),

-+

If

(x(to)' y(to) .. z(t:o

r(t}'

z(t)

are continuous at

is continuous, then the integrals of

x(t); v(t},

exist, and
-+

f
f

(fx~t.

r(t}dt

s. .

(Jxdt,

r(t}dt

fydt,

Jzdt)

ydt,

zdt).

From now ori,the arrows of vector function wil-l be omitted.

Theorem. If the vector functions

are differenti ab] e on . (a._,

the scalar function . u ( t)

. .
/

1. (FG} . =

3. (F G). ~

5. (FGI:l}

F(t), G(t), H(t)

s).

2. (uF}. = uF+uf

F.G+F.G

4. ( FxG) =

FXG+FXG

(F G H')+ (F G H-.}+(F G H).

= (F XG) XH + ( FXG) XH + ( FXG) XH.

where !'dot" -means derivative with resRect to the parameter

F = ( F l , F2', F3 ) ,

G = ( Gl , G2 , G3 ) ,

(H 1 , H;, H3 ).
1. ( F G)
1

then

FG

6 . ( ( FXG) XH)

Proof Let'

and

(Fl~Gl'

F2 G2 '.

F3 G3)

((FlGl) ?_(F2G2)
' (F3 G3)
(FlGl'

F2 G2,

. .

F3 G3)

( F1 ' F2' F3):(Gl' G2' G3)

FG

t.

225

2. (_uF)

.
.

((uF 1 ) , (uF ) , (uF ) )

2
3

(uF 1 + uF 1 ,, uF +uF ,
2
2

uF +uF )
3
3

(uF 1 , uF 2 , uf 3 ): (uF , uF , uF )
1
2
3

3. (F.G)

.
.

(FlGJ" +F2G2 + F3G3) + (FlGl +

F2~2

..
+ F3G3)

F.G + F.G

4. ( FxG)

( (F2G3-F3G2)

( F2G3-.F~G2

, .IF3 G1 -r 1 G3 -

F2G3-F3G2

F3Gl-Fl'G2-

(F3Gl-FlG3)

(FlG2-F2Gl))
FlG2-F2Gl)

(Fl

G2-~2Gl)

F X G + F X G
'

..

5. Since the mixed _product

(or

to

(FG H)

is equal.to

FxG~H

F,GxH)

the deriVative is obtained by

the application
of the properties {3) and (4) .
6. The proof'is obtained by applying twice the

~roperty

.The property {5) gives a rule for differentiating .

.

minant of order 3 where

elem~n~s are~funct~6hs

of

{4) .

-:er-

th~ ~am

variable:~

( FGH)

!
i
'!

226

d
dt

F1 . F2

F3

. Fl

F2

F3

Fl

F2

F3

Gl

G2

G3_

Gl

G2

G3 + Gl

G2

G3 + Gl

Hl

H2.

H3

Hl

H2

H3

H2

H3

Hl

This rule of differentiation

F1-

F2

F3

G2

G3
..
H.
3

Hl . H2

for determinants of any

hold~

order.
Example. Given the vector function
;(t) = Ch t i - Sh 2t j + e-t k
evaluate the following . .
a ) D ; { .t n 3 )

= o___; (t )

r(t)dt.

b)

t=.tn 3

Solutfon.

. a) D r(t)

(Sh t.
D

-D Sh 2t,

D e-t)

. -2 Ch 2t,

e-t)

( n 'ch t.

-:2 Ch 2.tn 3. . -e -.tn

- 3)

r(.tn 3). = (Sh .tn 3.

(e

.tn3 , -.tn3
-e
.

_,

3 - J1
~-

"

2.tn3 +e -2tn3
1
-2 e
- 3)

2~ +

41

R:n 2_)r(t)dt

.tn 2
= (

Ch t dt,

. .v, n 2

=~(Shtl

- J

3" --g
b)

'9"),

.tn 2

OJ

.tn 2

Sh 2t dt,

e-

. . n 2
_t . n 2
1
- 2 Ch2t
,-e
)
10
10

2 - 2

- 7 [-c h 2 .t n 2-1] , - ( ,- 1))

2
1
4 + tr
3
1
tr.- 2- (--z-:-- l),. 2)
2

tr -

8.. 1 .
16 2)

dt)

-- 227
'

.[

B, SPACE CURVES
Definitions:
A vector function
r(t) =x{t)i +y(t)j+ z{'t)k,
considered--as a variabH!_ position vec.tor
curve

r (or an arc

when

r(t)

curve)' as the locus of

te:(n.

s]

P{t)

defines a

P(x(t). y(t). z(q)

[nOs). of which the pa'rametr'ic

varies in the interval

r;

ij

-1:--

equations are
{X=X(t)
Y=Y(t)

. ---- __; ---.-' _ (J-

z = z(t)

r is

The i nit i a l p o'i n t of

i!

the end point is

B(t =

be oriented from

to

A( t

s). r
B

a) and

l")t

is said to

and the sense is--indicated by an errow

put on the curve.

is a plane curve or a skew curve according as it

does not lie. on a .plane. A plane curve or a skew curve is

'lie~~

or

calle~

a spac.e curve ..
El~mination

of

between two-coordinates gives a relati0n betweer

these coordinates showing that r lies on a cylin'der. Such a c linder

is called a_ projectin~ cylinderi. r

define5 three proj~cting

cylinders.
The space curve

r is closed. if

r(n) = r(S) (or i f

and called simple i f r does not intersect itself.

.....

A-'Jf B).

l' is said to be
/

"'
smooth i f admits tangent line of every point of it.

Lines are simple examples of .space curves . Below we give an

interesting curve which is skew.

228

Example.

(Circular helix). Consider the

r(9) = a cos9 i

sin~

c~rye

b 9 k

or
a cos9

(a>O,

a sin9
b 9

Elimination of
-2

+Y

9 .between

gives the

X, y

relati~n

= a 2 which is a rihgt circular cylinder (a projecting

-

~---

,-'

cylinder). Hence r lies on this cylinder. As

P'a cc 9, a

varie~ .

this cylinder,

'e~cribe

Sin9~

b9)

:K_B~~=-~J

lying on

a curve called a

~rc

(0, 2n)

--~---

'

corresponding to the interval

is shown on the figure, starting at

endin'g ilt

'~- .. ~~:~_-:.::-~rJ
'
.. ,,

circular helix.
That

p.,_ ..

A(a, 0, 0)

and

B(a, 0, 2nb).

The whole curve is obtained bytranslating this arc in the

o~

direction

2~L.

z-axi' by multiplies _of

The circular helix has the property that when the cylinder
is rut along the line AB
on a

p1an~

the curve

with slope

(or' .the line

becom~s

X=a,

Y=O)

an~

developed

l'~B

a line

-~I
:.2174
tOB
1

2nb/2na = b/a.

IX.

II ..

Geometric interpretation of

Let

;(t).

r: r(t) = (x(t), y(t), z(t)),

be

b&

desi~ed.

space curve.

P(t)

on

js the
P0 P

as

L~t

'

the equatior

li~iting

te:(ei,, 13)

of the tangent line at

Consider a nearby point

r. The tangent line at

..

P0

----~
-~r(/J

position of the line

P0 . The direction numb&rs

//
0

rrlo)

P0 1t 0 }
!;cloj

229

their limits, as

to are

x(to)' y(to}, i(to)

and are

com~onents of a tangent vector ~t _ P Hence

0

ts a _tangent vectof tor at

; ,t )

is a tangent vector at

P{t

= ( x,t

and

},

) , ;

ct ). .- i, t )) _

P{t). The differential

dr = (x,

y,

i}dt

is also a tangent vector.

The vector of equation of the tangent line at. P

is then

Arclength.
Consider two nearby points
P(t)- and
on

r. _THe length

P(t+Lit)
Lis

of the chord

joing th~se points is an approximatfon

to the are from
2

(Lis)

P(t)

P(t~Lit}.

to
. .

2. . 2

'~r)2

sd.:J

from the figure

=. (xCt+Lit)-x(t)) + (y(t+Lit} - y(t)) 2 + (z(t +Lit} - z(t))2

[\X +

:=)

Now,

Lix)2
( Lit

fly

+ LIZ

(Liy)2
[\ t

(Liz)2-.
[\ t

"230

When

X (t),

E_

D(a, S) ,

one has

(dx)2 + (dy)2
dt
+ <-R)2
dt

<{f) 2
=:)

Z( t)

y ( t) '

ds 2

dx

s =
a

+ d/ + dz

e~--2

x,

-2...+ z2 dt.

+ y

Physical interpretation of

r(t):

If one co~siders the parameter. t as time, then the curve

r: r{t)

(x(t), y(t), z(t))

becomes the path (trajector:y) of a moving pa~ticle.

The tangent vector

;(t)

is the Velocity vector

the moving particle, si.nce- x(t), y(t), z(tj

comp~nents

~irections

in the

v(t)

v(tl of

are velocity

of coordinate axes:
=

(x(t)' y(t)' i (t})

. or
vy'

v{t) = (vx'

The ma3nitude ,of the velocity vector is .the speed.of the

particle:

Example 1. Find the arc

r(&)

=a

of the

circ~lar

cosB i + a sinS j + b 9 k

A(~=O),

between the points

le~gth

P(9).

Solution.
y

a sinS,

z = bB

~a sinS,
y
',1
. 2 .2 -2
2
X +Y +Z
= a +bs = JB/a2+b2 dB

a cosB,

z =b

a cosB,.

X
.:::.')
=~

:X

~'

helix
a>O,

b>O

231

Example 2.
3

Let
.'

p = (t ' 2t, t ),
be the paths o{ two p_articles
a) Do the particles collide!
b) If they

collide~

find velocity vectors at the time. rif

collision.
Solution.
~)

For collision, the coordinates

be the sameat the same

time.~Equating

~f

two particles are to

the first coordinates we

have:
3

t = t .f4t-4

. -:.i

t -t -4.t+4 = 0

=:)

t= 1' t:2' t,;,-2

of which only t=l satisfies .the other equations 2t = t 2+1

2t 3 '-l. Hence the p.articles collide when t = 1.
t2

(3~

vl ( 1)

=)

and

2, 2),

FRE:NET Frame.
Given ~ spijce curve r I to . any point Pof r there is as~ociated
a system of three mutually orthogonal

un~t,

vectors

T, N, B as

'

described below. Such a system is called the FRENET Frame.

Wh~n

parameter)~

the aro length

is usedas

be6o~es a function of

param~ter

(instrtnsic

s, and by chairi rule~ we

have
I

r(t)

dr
crs

= datt)

dt
crs

which.is then a vector tangent to r at any poirit

P(t)

232

where
I

dr
Of

dr
ds

r(t)

thi~

These notations will be used throughout

Now we.show that the tangent vector

r'(t).r'(t)
(x

+Y +Z

This unft tangent vector

is denoted by

that

sho~ing

T.T
T'

I (

h~ving

= ds 2 /ds 2 = 1.
the same sense as

r(t) Ps- .

t ) =

1, by diffe~entiation, we get

T .T

is perpendicular to. T.

Definfng a uriit ~ettor

N-

in the same di-"'.. rect(on

and sense as
.
we write

is a unit vector:

=> 2 (~~) 2

)dt /ds

r'

r'(t)

T:
T

Since

Section.

rc..tn ,-.ts .r:. ..t"'

T'
( jS

t/

T' = KN

where

N is called the principl unit normal vector and the

scalar

K the curvature, and its reciprocal

of curvature(l) of r at

= 1/k

P(t).

Havfng defi.~ed two unit Orthogonal vector.s 'T

ctefi ne now.. a third

o~.e

the radius'

and

N, we

B by
. B = T

which is orthogo.nal to both

T and

N, called the binormal unit

vector of rat. P(t).

These three mutually orthogonal unit iectors

T, N; B form,

in this order, a positive system called t.he FRENET system of r at P(r).

(1)

The reason for calling the "radius curvature" is that when evaluated for
a c;:ircle it gives t~e radius of that circle.

233

The linesthrough. P

along

T, N, s

are the tangent line, iprincipal normal

r. These lines detE~rmi ne three

binormal of

planes: The one determined by

normal

~lane

of r at

(NP) of r at

e.

to r at

are tangent

~lane

osculating

and

(OP)

N, B is the

p. The oth-er two

The one formed by

T, N is called the

and the other the'rectifying plane (RP)

P.
The osculating plane can be shown to be that tarigent

th~t fits the cutve

as follows: When
certain

developed onto

best. The name "rectifyi~g" may be explained

r, the RP becomes

P . describes
S containing

su~face

plan~

th~

r.

curve

plane. When developed its

Thi~

cur~e

tan~ent

to a

surface can be

r is transformed-

into a "straight line". Hence the name "rectitying".

The equations of

thes~

planes at

Normal plane:

P0

are:

(P-P 0 ):T=0

Rect.ifying plane: (P-P 0 ).N = 0

Osculating plane: (P-P 0 ).B=O
Exam p1 e . Gi ven the ci r c u1 a r ih e 1 i x
r( t) =
a)

a c os t

i + a s i nt j + bt k

fi n'd T, N, B, "

b0 find the equation of

t~e

osculating plane at

P(t)

Solution.
a)

dr
T = Os
dT
ds

dr dt
dt Os

dT
kN =?> Os

= (-a sint, a cost,

dT dt
dt; Ts

b)

ta<b 2

(-a cost, -a sint, 0)

a 1 +b'2

234

-sint,

a
;z:;z
,.

i
-a sint
- cost

TxN

0) ~
a +b

(-cost,

kN

(-cost, -sint, 0)

a cost
- sint

b
0

.. Princ'ipa1 normal vector is parallel to xy-plane and

curvature is constant.
b) ( x- a cost) b sin t - ( y- a i i n t) b cost + ( z- b t) a

(b sint)x; (b cost)y az = abt

~ ~int -

cost +

t.

SERRET-FRENET formulas;
These are expressions forT', N', B'
T~

tions of

N, B.

T'

was already obtained as

B'~

is derivable fr.om
B.B=l,

s.T=O

B'.B=O,

BI T

=?

B'.B=O,

s.T = 0

Hence there is a scalar

KN.

B.B = 1,

=)

B .J(N

T (ta~)

B.T = 0

B'.B=O,

by'differ:ent.iation:

B'.T+B.T'= 0

=0
B~B,

B'.1.T =;> B'//N:

such that

B' = -TN
where T is called the torsion ofT at
Now,

.as linear combina-

P(t).

235

NI

(BxT)

BI xT

SxT I
BxkN = '-KT

-TNxT

TI

k"

-1(

TB

Thus we have obtained

T'

kN

NI

-KT + 't'B

or

N'

[B [ 0
J

BI

Tj..,[~

-T

T-

_B

which are called the-SERRET-FRENET .formulas, dfscovered by SERRET

in 1851, and by FRENET in 1852, wher.e the square matrix is skew
symmetric.
The reciprocals

~,

cr

of

(Kappa),

'r

(t.au}

are

called the radius of cDrvature and radius of torsion of the space

curve r at

P(t}.

For a straight line,

implying that

I(=

0.

is a constant

v~ctor,,and

T ~-0
1

is a measure of deviation (departure)

I(

of the curve from the tangent line.

For a plane

~urve.,

T = 0. T

implying that

B is a-constant vector and

is a measure of deviation

the curve from the osculating plane.

Conv~rses

are also true, that is,

k - 0

=)

i s a s t r a i g h t 1 i ne ,

T - 0

is a plane curve.

From SERRET-FRENET formulas_, we have

1<2

Evaluation of
Setting,

= T, _. T

= r

11

T", N", B11 :

r 11

BI . B:

B'

(depar~ure)

of

236

the SER~~T-FRENEt formulas are written as

u = Fu
I

which, when differentiated, gives

F1 U+ FU 1

U"

-1
2
(F. +F ) U-

- F1 U+F(FU)
where
!{'

F'=

-K2

-j('

T'

(\

-T'

F2 =

Kr.

2 2 0
-1( -T

kT

-T 2

Then

T"]
r-K2
N"
-1<'
[B" kT
=

or
2
T" = -k T + K 1 N + kTB
N"
-k'T;_(K 2+'r2 )N 't''B

B"

k~T

-T'

N~

2B

Now .we evaluate the mixed products

and

(T T1 T"), (N, N', N")

(B, B 1.,_ B"):

(

T T' T")

(T I<N -k 2r+

(T I<N

K'N+kTB)
_kTB) = IC 2T(T N B) = 1{2T

( N N' N" ) =' (N -I<T+'IB

- k'T~(k 2 +T 2 )N+T~B)

(N -kT TJB}+(N

TB

-I<'T)

-kT 1 (NT B) - K'T (N B T)

kT

K1.T

237

I
(B B 1 B")

(B

-TN

kTT-T 1 N-. B}

(B

-N

kTT) = kT

Thus we have
(r

(T T 1 T")
(NN

r" r" 1}

kT 1

N"}

= k2 T

k1T

(B B 1 B")

show:ing that the sign of

(B 1 B

is that of

I(

B"}, and the sign

(T T 1 T"}.

ofT is. that of

Physical interpretation of r(t)

If a pa~ticL~

moves on the trajectory

r : r(t} = x(t}i

y(t}j

is t~le, then

z(t)k,

is the

where the parameter

acceleration Vector

a(t}

of the particle at time

We show that

a(t}

lies on the osculating plane of. rat

r(t) = v(t)

t.

p ( t}:

d
d ds
. a(t} =at v(t} =at (at T}
2

_d s
-dt7

s T

s T

s
showing also that

th~

ds
at

ds (ds dT}
at

d
at T

n crs

v2

I< N

T + !._ N
~

tangential and normal components are

v2

aN= -

If the motion is uniform (ds/dt =constant} dlong

s

0,

r, then

and the acceleration vect-or lies along the principal

238

normal of r
Example.

A particle is moving on .the path

r: r(t) =inti+~ t 2 j+/z t k.

a) 'find velocity and acceleration vectors,
b) find the components

aT,

aaJ

aN. of

at

1.

Solution.
a)

v(t)

r-( t) = t1 i + tj

.a (t)

r(t)

-:-zl +

./2

b)

aT

where

"' = /x2 +Y 2 +Z 2 dt
ds
ds
t 2'+1
s = dt =. t-

0.

=>
where

K =:lr"l = ldr'l

crs

r' =, r(t) ~ =
.
ds

(.!.t

i'+ tj + /2k)

t
7:1

t
1
1 t2
1
t .
r" ~ [(- -:7 i+j) -:-r:- + (- i+tj+!'2k)'
2 1~
t
t+1
t
(t+1)'t+1

. =)

ri' ( i) = , ( ( - i + j )~:]{- ~ -} '( - i + j )

k(1)=1r"(1)!=12/4.

Then

2
aN(1) = v (1) K(1)

I
!.

= /2.

239

Curvature of plane curves.

Let
r :

f(x)

r(x)

be a plane curve.
(1. + Y 1 J.)

r 1 (X)

(i+YI j)

~s
f.w., th
u:; \!:

1 2
d s = VI
,+
-y-

dx) .

;f1 12
+Y

l2

. yl
c- l+YI2

Y
j]
1 - ~-----,-)

l+YI2

(-yl'i; j)
y"2

For a plane curve the curvature h~s the equivalent definition.

namely
K

da
ds

Indeed,
tan a

o'
=

yl => a

da
K = crs

l+yl2

arctan yl

/l+yl2

==>

da
ds

_y_"_ dx
l+ y I? ds

y"
(l+y12,3/2

?!;,.

240

K when the plane curve r is given in

We have obtained

param~tric

vector or cartesian form. Now, for a

y' =

in

y" '-

j;i.

y = y(t),

x(t),

.. .

y X - y X

- .dx ;I

';'(1
,2)3/2
y
+Y.
.
'

curve

we have
Xy

iX

( x2..+ y-2 ) 3/2

When r is given in polar form:

r: r = f (g)
computing

..

.,

2 .

which can also be 9btained from

= arctan

y = r sin9,

r +2r'- rr"
( r 2+r ,2)372 .

ljJ

~o~9,

and setting in above formulas one gets

j , X, y

X,

x =r

or

= dx/ds

where

a=

1jJ

+g

and

fr

Circle of curvature:
The circle of curvature-of a curve
.

at a point

P of it is the limiting circle

~

of the circle passing t.hrough

ne a r by po i n t s

Q, R whe n

P and two

Q + P

R + P

It can be .shown that the circle center

at

C = P + ~ N a nd r'a d i u s

of curvature
y

~at

(in

~-

i s the c i r c 1e
or 2-space).

lies in the osculating plane at

(orb~. tangent vector

I

T)

si~ce

it is tangent to

and cent~r 'is on principal normaL

r.

241

The coordinates of the center

i;= X . -

y'

Y"
--;rr-'

and the equation of

(X

are, for a pl ane curve,

. 2

l+y'2

1+~.2

n =Y +

i;

x(t) - x

n.

y(t) - y

I:.

.2
+y

~I

.2

.2
+Y

I~-

~.,

is
t.;)

+ (y - n)

2.

= )

Example. Given the cardioid. r = l+cos9, find extrema of

the curvature.
\ Solution .. r'
K

-sin9,

.II

-cos9

2
, 2
r + 2r' - rr"
( r 2_ + r ,2,3/2

.3
K=--Cos3 E!

2 .

3/ltr

is Jllinimum when 9 = 0,

Curvilinear m~tion (~plane).

C'fA

...rE

242

-Jt

(r cos8}=r cos8-(r sin8)9=; cos8-<~~r sin8

-dt (r sin8}=:r sin8+(r cos8}8=r sin8t-wr cos8

d

Of

(r

cos8 - wr. sin8)

tos8-{i sin8:~-~r sin8-(cm" sin8}-(<vr cos8)0')

cos8-~cur

sin8-c.ir si

n8~w 2 r

cos8

or
ax

;:

({-'- u;- r J cos 8- (2.:v r' +U:' r ) s i n8

. ) cos 8
r- .<c2r )s.1 n 8 + (~t'~'r +~r
ay - (''

ar
=)

a8

ar~ a 8
acc.eleration.

where

-~

cuH:

+~?r

(by a rotation)

..A. kL
-~-'-=---.~a.
,\

. . . . :"'T

_,.."""'

~
'
.!)'

'

a.'ll

are the radial

and transverse component~ of the

T~e motion under a central force

(as in .Pla~etary motion)

the lrdnsverse component of the accelation is zero. We have

a8

1
2
1
d (wr 2 )
'.ur+<4-lr'=- ( 2c..Jrr +tt~ r ) = r
Of
r
d R
d
d
r 2 d8
8
(far-central force)
(~):0
r df ---at'
r CIT .. t .

d R

8r

at

con st.

(~EPLER's

..
second law: The area

swept bj ~he r~rlius vector .is pr~po~tional to time).

243

Curve Sketching:
Sketching of a space curve r is usually done by the use of
two of its projecting cylinders.
Let directrices of its two projecting cylinders be
l

say, on yz- and xz-:-planes_. Since a . plane

parallel to xy-plane intersect the cylinders
1 , t 2 the interP_ ofEP -f1.. belong to the cu-rve r.

alongtwo generatrices
section

Then r is plotted by taking some number of

planes

//xy~plane.

A projecting
met~r

cylinde~

is obtained by eliminating

th~

para-

between two coordinates if r is given in parametric form,

a nd b~ e 1 i mi na t i ng a. v a r i a b1 e (; f r i s g i ve n a s

F( x , y , z ) = . 0 ,

G(x, y, z)_== 0, since the equation of a cylinder (generatrices

U are

axi~)

involves only two variables.

Example 1. Find all three projecting cylinders of the

following curves:
a)
b)

r
r

t,
2 .2

X=

=t 2

+4Z = 16,

z = 1+It.X

-y

+Z 2 = 0

Solution.
a) .xy-proj.: cyl.

~:

xz-proj; cyl._;
yz-proj~

b). xz-proj. cyL;

yz-proj. cyl.

( z -1 ) 2
.

+4Z 2

( z -1 )

cyl. :

xy-proj .. cyl.

4/-3x

'

;-

.Y

16 ( g_i ven)
2

= 16

y 2+3Z 2 = 16

244

Ex~mple

2. Sketch the curve_-

r:

' 2
X

2 '2'-.
+y +Z

1 ,

2 2 4
+y --X

(A VIVIANI curve)

in the I. octant.
Solution.
,xy-proj. cyl.

( x-2 )
2

xz-'proj: cyl.

+-/

4
.

z t4x = 16

EXERCISES (3, 4)
116.
'

At what points does-the curve

r(t) := 2t 2 i + (1-t)j + (3+t 2 )k
intersect the plane

117.

3x:-14y+z

10?

--

Find. the unit tangent vector

for:

llB.

a) r (t)

(t3,

b) r(t)

e-2ti + e-2tj + (l+t2)k

Find

2t+ 1 )

angl~

of intersection of the curves

r(t) = (l+t 4 , 2 cos1rt, t-3 ), r(t) = (t+t 2 , t-3t 2 , tel-t')

at"
11 9.

~osine

1-'t,

of the

(2, -2, 1).

Find the are l'ength in the given interval

a) r( t)" = (t Sint, t cost, t)
2
b) r(t) = ti +
t j + t 3 k.
c)

1 20 '

s h0 w

(0,

1T/2]

[o'

21

= t, Y = R.n ( sect + tan t ) ,-- z = R.ri sect,

that the -curve

[o'

1T /4}

r -

r(tf = _etsint i ~ etcost j + etk

. .. 2
2
2
lies on, a cone- ax +by +CZ
angle with z-axis.

0,

and

l(t)

make~

a constant

245

121.

Obtain the equation of the surface generated by the tangent

lines to the curves:
a) r(t) = a cos9 i . + a sin9 j + b9 k
y = t2' z = t3
b) X =-t'
c)

122.

2+y 2 +Z
. 2

l 6'

2
+Y -4x = 0

Find the condition on a' b, c so that the curve

r(t) = at 4 + bt 3 j + ct 6 k
3
4
lies on the surface z2
X +2y

123.

Find the principal normal vector N of the curve

.
t .. t
t.
r(t)
e' 1 + etcost J + e s.1 n I<
at

124.

If

-+

a) r

dr = 0

-+

b) .r

prove

r =.lr(t)

-.9

is constant

is constant.

-+

is a vec'tor function with

r(t)
-+

l ' 0)

(1'

dr = 0

For the curve

X

.,... 6 sin 2t,

Y= 6

COS.

2t,

z..= 5t

a) find. T

b) find

c) find

d) find the osculating

N' K' B

plarie at

for

(0' n)

t = rr/6.

126:

Find the equation of.the osculating plane of

. 3t
-3t . j1)
r: r ( t) = ( e . , e
, 3...- 2 t)
at t = 1 .

127_>

Find the equation of the nor:mal plane:

3 2
a) r(t) "' sin 3t i + cos 3t ~ +. 2t 1_ k at
et 1. + t'J + t 2. k
at
b) r(t)
t = l.

128.

!O,,l, 0)

For the .given trajectories find the velocity vector

and .the accelerdtions vector

a(t)

v(t)

with its c~mponents ar, aN

246-

129. Find the velocity vector ~f-motion

r{t)"' sin 2t i + R.n(l t)j + tk
- 130. At whatpoint cf the parabola

x 2/4

is the radius of

iurvature a minimum?
131 . Find the curvature and equation of the circle of curvature
'of the <;urve
X

132. Find

tn cosy. at

K, T,

( 0' 0)

N . for the circular helix:

r(t)

cost i + sint j + 2t k

133. For the curve

a(3t-t 3 )i+ 3at 2j+ a{3t+t3)k

r: r{t)
prove
K

=T =

3a{l+t 2 ) 2

134. Find the path of a particle which starts from the origin at
t = 0

and moves with the velocity

v{t)"' i cost+ j sint+k

135. Find'the curvature and torsion of

a) x

.,b)

a(t-sin t),

a(l-cost),

a cost,

a sin t ,_

z .= a cos2t

bt

136. Show that the xy-proj~ction of the curve

~
2 2
2
2 2 2
2
f: X +Y +Z -2ax-a = 0
X +y _rz -2a;:+a = 0
is a circle . .
137. Write the equati~~s of the projecting cylinders of t~e
''

following curves:

247

2
4
6
a) r(t) = (t , t ~ t ), .

b) r(t) = (Sh t, Ch t, et)

138. Find the projecting cylinders:

a) x
y

= et

sint

b) x

cost + sint

= Ch

cost

z = 2 sin 2t

-t
t

139. Sketch 'ttie curves:

a) z

xy , - x +Z = y

2
2
b) x -4x+ y

0,

+y

140. Sketch the lines

a)

x,-1 _ .9

z-1

b) x+ 2y+5z = 10,

-::y-T=~

-x+ 2y+ z. = 2

by ;he use of projecting cylinders.

ANSWE~S ~0

(18, 4, 12)

1 1 (. ( 2 ' 0 ' 4 ) '

1 '1

EVEN NUMBERED EXERCtSES.

12/544

122. aJ+ io 4 -c 2 = 0
6
- 3 .
3
126. x-e y- ,Ze z+ 6e
128. a). v

i+6j+l8k,

3j+l8k,

b) v

i+2j .... 2k

a= 2i+4k,
130. at
132. 1/5,
.....

aT= 18,

aN = 3

a =4
N

= -2,

(0, 0)
1

7"'S"'

(-sint i+cost j+2k),

134 r(t) = si nt i

{1-cost)j + tk

-cost

sint j

248

b) xy -::. 1 ,

xty-2z

0,

x+1-=-2xz

r i
140.

A SUMMARY
3. 1.

Dot product: A.B

. A. (B+C)

fAIIB'I

a 1 ~ +~ 2 b 2:+a 3 b

cos9

A1.B .:::~ A.B

A.B+A.C,

;, B.A

Cross product: AxB

n I A I I B I sin 9 _ where

I nj = 1 ;

A, B, n i s

a positive system.

AxB =

.1 AxBI =
l~i

al

a2

_a3

b1

b3

IOARBI2

where

(ABC)

xed product::

OR

BxA,

--

Ax(B+t:)

AxB +Axe

OAi.OB --

AxB.C

A.BxC

al

a2

a3

b;

b2

b3

c1

c2

i:3

T~ip1e vector pro~uct:

Ax(BxC)

(A;C)B -

Inner product in- Rn

<A, B>
II A II

= a 1b 1

++a n b n

2
2 1./2
:::: (a1 + . + an)

(A.B}C,

(AxB)xC
in

c (a,

<f ' g >

= (A.C)B

bt
f

f (,t ) g ( l) d t
.

a b

llfll

- .(B.C)A.

( Jf~(t)dt)~ 12
a

249

3. 2. PLANES:
Plane through

P (v
y
z )
o "'o' o.' o.

and .L

C(z-z)
0

-General equation:

Ax+By+CZ+D

(A, B, C.):

_o

,Normal equaf~ons:
~

. xcos a+ ycos B + ;t:cos y - p = 0,

.
'/.2 2 2
(Ax+By+CZ+D)/VA ~B-+C. = 0

2 2 2
ax+bY+cz+d,;O (a +b +C = 1)

Plane through

+ -

xl

yl

zl

-x2

Y2

z2

x3

y3' z3

. X,

II

and

P = P + tO

+ta,

y-y 0
c

x-x 0

P0 +sU+ tV

D = ( a , b-, c ) :

( vectora 1 ,egua tion)

Vettoral form: p

0.1 ~

= y 0 +tb, -z = t 0 +tc (param.cartes.equ)

z-z 0
-c~

Ax+BY+CZ+D = 0,

A'x+B'y+C'z+D'= 0 (simultanec

_g_~)

3. 3. SURFACES.

A surface is the graph of an equality relation

x

= f(u,

v)~

>.F(x, y, z)

y
+

= g{u, v),

l.IG(x, y, z)

Cylindrical coordinates:

-t"'(B. r, i!.-)

+
f".P/1

z = h(u, v)
0

f(x, y, z)::O.

(parametric egu)

(linear family of planes)

Spherical coordinates:

250

S: x-f(t)
a

(yl i nders:

y-g(t)
b

with direction A= (a, b, c) and directrix.

y

x = f(t),

= g(t), . z = h(t).

--

z-Z0

x-xo

y-yo
- ----,...-;::,x 0 -f(t) - y 0 -g(t)

S:

Cones:~

with vertex
g = g(t),

(x 0

y0

z0 )

z 0-h(t)
x = f(t),

and directrix

z = h(t).

Surface of revolutfon with axis on z-axis:

S:

S:

. whefe

2
2

+y 2
+Y

r 2_, z = h(t),
2
2
f (t)+g (t);

f(t),

g(t),

= f

(t) +9 (t)

= h(t)

= h(t)

is the generatrix .

QUA.DRICS:
.

.3

(y-~)

J!

!i

(x-h)
a2

(~-k12
b2
+

-=

(Ellipsoid}

2
(x-h)
a2
.2
(x-h)
a2

(z-~)

(y-k)
b2
(y-k)
b2

(z-.Q.)

-(Hyperboloid of one sheet)

(Hyperboloid of_ two s \,e e t s

c2

r.

2
( E11 i p ti-c paraboloid)

>,{_Z-,}

2
(Hyperbolic paraboloid)-

=>,(Z-,)

SOME DEGENERATE QUADRICS:

f(X, y)
(X"-h) 2
a2

=0

or

- (y-k)2
2
b

g (x' z)

(z-!1,)2
c2

:or

h (y' z)

(cone)

(c~linders)

251

3. 4.

SPACE CURVES:
Vector function:

r(t)

x(t)i

tangent vector:

r(t)

X(

unit ta~gent vector

t )i

unit principal normal vector

FRENET formulas:

y(t)j

z(t)k

t )j

z ( t ) k'

+ y(

r'(tj ~~dr/ds

unit binDrmal vector B

TxN

N ~ K-l T'

(x(t), y(t), z(t))

( r = dr I dt )

. . 252 .
(

MISCELLANEOUS EXERCISES
"' Show that the midpoi~ts of th~ sides of a quadrangle are the
hf a parallelogram, (VARIGNON,
;i

2. Pre;,

->

---!!>

~n1(et+8)0P

165~-1722).

sina OA + sin 8 OB

\
\
\

Show t.
If

a.A

then

144. 0

L:-~----i---+! t1

yC

+ (.

oO = 0,

ar~

A, H, C, 0

Prove:

-+-

oc+ CR

145. Use dot

0),

coplanar.
H the orthoce~ter df

is the circumcenter and

......

o ,J

(a.; 8 Y

L1

ABC.

-+

+ OC = OH
to provr '

an angle jnscribed in a semi-

drcle is a
1 ~0

Show that

A(O, .., , u,

are the vertices of a

. 3, 4) '
ri~

in

C(4, -1 ' 2' -3)

R4.

14 7. Show that

a) In an isoceles trapezoid the

:.a l s are congruent,

d1

b.) In a right triangle-the midpoint of the hypothesis is

equindistant from the vertices.

148. By projecting the sides a regular pentagon oro a l~ne which

~akes

Cos g

149.' If

angl~

an
+

with one of its sides, p.

(g
Cos ( g + T2n) +Cos

+
lbl

a
+.

= a ab

4n)+ Cos ( g +-;:6n) . -+ C~s ( g

T

show.that
+

+
+

bisects the angle between

b a

b
+

.that

and

b.

8n) = 0
T

253

150. Show that

~)

in a parallelogram diagonals

b) i~ q~adrilateral

bis~cts

the bimedians (segmen~s Joining the

bisect each other .

midpoints of opposite sLdes)

.

151. Show that ln a patallelogram

to
152~

t~e

midpoint

othe~

each

of

ABCD

poi~ing

the line

(BC) trisects the <di-agona 1 [Ar] .

Let

A(a 1 , a 2., a 3 ), B{b 1 , b2, b 3 ) and

vertices of the triangle ABC~ ~rove
:1) the law of cosine

C(c 1 , c 2 ,

3)

be

b) the law of sine

153. Show tha~,i~ in a tetrahedron two pai~s of opposite edges

are othogonal, then the third pain isorthogonal.

'

154. Find the

cond~tions

crx,)xl = ix(gxl)

a, b, c

for nonzero vectors

for,_which

holds.

155. Show that for three ve~tors,

a ) ( b xc ,

c x a , , a xb )

= ( abc )

b)

ax(bxc) - 0

156. Prove that

a)

((AxB)xC) X

b)

{((A_ xA )x~ )xA 4 fx A

,i.

is equal to

Al

157. Given

.Ax((BxC)xA)

A4

A2

Al .A3

A2 ..A?,

.Al.A5

A2 .As

A4 .AS
4 .

3 . 4 .
V.l, = "S"l+sJ,_

show that v1 , v2

3 .

;-l+S"J

in an orthonormal system in

158. Which vectors of the bases { (1

-1

3) T,

IR 2

(2-, 1_

l) T

l,Ii
I'

II

il

254

:I

IIII
il.

(3

II

(1

for the vectors of

can be repl~ced by the vector

!j
!I

4)T

-2

159. Let

-3

3) 1 ?

be a sliding force along a_ line

Let

il

(2

G be the moment of

and

4 be a line.

F with' respect to a point

+

6. Show that th~projection vector of G

of 0

on

on 6 is independent

taken on

?l.

160. Prove that the centroid of equal masses of the verti.ces of a

quadrangle is the center of the VARIGNON parallelogram. (See
Exercises 141).
161. Prove that the perpendicular bisectbrs of the edges io a
tetrahedron are concurrent_
162. Find the

= TT/4,

a.

a)

equatio~

cosa.

of the planes for which

8=TT/3,

cos8

cosy

--:r .= ---::r = --=r

b)

163. Show that the plane

I

point of
and

i
16~.

=2

x + 2y- 2z ..- 9

intersectio~of

2x+3y

the planes

passes through the

x+y+z = 1,

8.

Fihd th~ point of intersect~on of ~he planes

Xt2Y+Z

x-y-z = 1

o,

x-2y-8

= 0 and X+y+z-3

165. Show that the four planes

x-:y+8 = o,, 3x-y-2Z+l8 = 0

(if any):

x+y+2i = 2,

x+y-2z = -2,

passes through

some po'int.

166'. Find the angles between the following pa.irs of planes:

a) x+y-4z

b) 2x-y+z

3y-3z:t 2

X+y+2Z

11

255

167. For what values of

a.

s.

xcos a + yeas S + z cos

and

y -

p=0
b) passes through

II to x z- p1 an e

a) i s

p. the plane

y~axis

J 68. Find the intercepts on the axes and the traces of:
a) 2x+3y+4Z = 24
169. Find the traces of the following lines:
a),2X+y-Z=.2.

b)

x-y+2z = 4

X+ 2y
2x-4y

170. Find the angle between the following lines:

a) X+y-z

x-y = 1

0
and

Y+Z = 0
b) x-2y+z

2y-z

x-3y+Z

x-2y+Z

and
x-2y+2z = 4

171. Show that the following pairs of ti,nes intersect and perpendicular:

/a)lx+~Y = 1 and~x-y = 1

2y-z =

b) 13X+y-z = 1
2x-z = 2

x-:2z = 3

and

12X-y+2Z = 4
x-y+2Z = 3

1.72. Find the equation of the projecting planes of the following

lines:
a),2x+y..:.z
x-y+2Z

o.
3

b)' X+Z
x-z

3
3

173. Find the distance measured along the 1 i ne X = 2 - T3 t.

4 t {rom ( 2. 4. -3} to the in12 t.
z = -3 +TI
y = 4 +TI
tersecti on of the line with plane 4x-y-2z = 6.
17~.

Reduce the equation of the

foll~wing

lines to the symmetric form::

I
I:

256

b) x =mZ+a,

a) 4x-5y+3Z'= 3,
4x--5y+Z+9 = 0

y = nz+b

1"75. Find the equation of the plane determined by the line

.x:-z
Y+3 z-1
. z.- = --:-z- = --,- and the point {0, 3, -4).
176. Find the eq1,1ation of the plane determirted by the parallel
lines
x-3
:-r

and

Y+4

z-1

--y-:-. = -1-

177. Find t.he. distance between the para 11 e 1 planes

and

_2x-y+i

4x-2y+2z = 10

178. Find the dis_tance betw-een the given plane and the point:

{T, 2, -3)

a) 2x-y+3Z = 4,

b) X+2y-z = 5,

'

{2. 3. -1 )

179. Find the distance between tWQ lines:

x-1

z-

.y

"!'

Z+l

and.

= --,-

y-1

.1"=--z=::-r

.180. Find the distance between the- line

2 and the

x. -Y+'l. Z"'
"3"=--z=:-r

point -p(o, 1, 3).

181. Fin~ the dista~ce between the parall~l lines
X+ l'

y+2

---r=T=T

and .. ~- = Y+24
.J

z+l
--,-

182. Identify the surfaces

a)

a u cos v

b) X = a u Ch v

b u sin v
u2

y = b u Sh v
z = c u2

Z=

c) X = a Sh il cos v
y

a Sh u sin v

z = c Ch u

d) X = a Sh u .Sh v
y

a Sh u'Ch-v

z = d'Ch .u.

257
e) x = a

.!:!.:.!
U+V'

y =

uv-1
Z=c--
U+V

b UV+1
U+V

183. Identify the surfaces!

. 2
y2
z2
z2
x2
. a.) x2+~ + T = 1
b) T+ T+ T = 9
2
d') z 2 = X2 + ~+ 4
g) x = a(u-v),

e) X

2
= z2 - y

y = b(u-v),

1 ) r ( u1 v)

= ( u- v+1 ) i

.f) X

2
.= z + l

2 2
= 4 -x

z = uv

i) ~ = TT/4,

h)-):= Z,

c) x2 - 9

k)

z-(rl) (= v)

+ ( 2u +v) j + ( u- 2,v) k

]84. Discuss and construct the loci

a) /+z 2 = 4x

b)

i-z 2

= 4x

185. Construct the fQll6wingsu~faces and shade that part of the

first cut off by the second: (in the I. octant)
/2
2
2
2 2 2
16
a) X +ljy +9Z = 36, X +y +Z
2 2 2 64,
x 2+y 2-Bx = 0
b) X +y +Z
2
x2+4y 2 -z 2 = 0
c) 4x +i-4z = 0,
2 .
186. Given x +2yz-4x = 0, after a suitable transform'ation obtain
.its standard equation and identity.
187.

Fi~d

a)-

the
/!.:

~artesian

y,

equation of the cylinder

r: X

0;

sint,

y = cost,

z = ; i nt cost

0, X+y = 2ax+2ay
X 0;
r: y
2
X+2y+3Ha = 0 and passing through
c) axis: x 2 ..:2y +z+a = 0,

b)

/!.!

z,

(0, 0, 0).
'

l88. Find the equation of the cone with directrix

r: x

a+a

co~t.

a nd ve r t ex a t - ( 0 , 0 , h )

= a sint,

z =0

258

189. Find

equati6n of the cone:

t~e

0, directrix

a) yertex at

x = x-a'

.c)

2t -t,

t2I

is

z. = z-c

y-b'

x = 3t 2 -2at,

b) vertex at (1, 1, l),directrix

=

x = y = z,

semiangle at vertex is

n./6.
whe~

190. Find the equation of the surface of revolution generated

x 2 +Y 2-2ax-2ay = 0

.r: y = z,

is revalved about

z~axis.

191. Find the cartesian equation of the surface of revolution

. 2
z = t3
a) axis: az, generated by r: X=t, y = t
2
0
b) axis: Oz, generated by r: Y=Z, (x-y) -2a(x+y)

r: r=a COS9+b.

c) axis: o:x' g"nerated by

192. Find the h

the points equidistant from they-axis and

xz.-plane.
193. Find the locus of

th~

line~

a) two skew

point lnat are equidistant from

b) a linr

3~d

two pciints

194. Find tli.: locus. of the point equid1stant fr-om the z-axi.s and
the plane
195. Find the

z
loc~s

tangent to two
196. Find

~1.

of center of spheres of radius

inte~secting

"r"

are

t~at

lins having angle 2a between them.

sketch the locus of the points equidistant from the

~and

x-axis. and the point

A(O, j, 0) ..

197. Find th~ locus of the mid-points of the chords of a space

curve

r: x = f (1;) ,

y = g ( t) ,

198. Find the locus of the point

the plane

y = z.

z = h ( t)

e~uidistant

from the

x~axis

and

259

199. Find the locus of the point ratio of its distanc,e from the
point

A(2, 0, 0)

and the plane

is

n: x

2.

200. Find the locus of the points, ratio of its distances from
ihe point
201.

Fin~

~nd

A{O, 0, 2)

B(d, 0, -2)

the equation of the planes whose

2 .

ellipsoid

2'

9x +25_x+T69z

=1

x-x 1

z- z

z-.

3.

in~ersection

with the

are circles.

202. Find the condition that . (x 1 , y 1 , z 1)

point of the chord of the hyperholoid
by

is

should be the middle

x'2-/+4z 2 16 formed

----=-z-

x- 2
y-l . 2 +1
cos a
cos S = cos y
should b~ tangent to the paraboloid. x2 -y. 2 +3z
0.

20.3. Find the condition that the. 1 i ne.

204. Let

....

= {Cos

Show that
dP
a) p xdf
205 .. If
a)

c)

206. Let
let

p ( t)

Jtj

t)A + (Sin

t)B . where :>. ' A; B are constant.

d 2P
b) d"t'+

-Con st.

A p

are vec.tor fun c ti on s , show that

'd 2 P
dP
dP
d
b) p
= I PI dt I PI
df) = p X ::7
dt
dt ..

Q{t)

'

d
dt (P

d {P x dQ . dP
d) dt
dt -. dt

dt IPI

Q)

p X

P, Q be two fixed points of a solid in motion, and

....

V(P),

V{Q)

be

tue

velocity vectors at

and

Q. Show

that the projections of these velocitions on the line

PQ

are equal to each other.

207. Find the arc length of the VIVIANI curve
in the first octant.

r:X+Y:t-Z

4a 2 ,

260

T, N, B, .,:... at the given point of the

208. Find

gi~_en

a) r(t)

.
etcost i + etsint J.+
e tk '

A( 1 ; 0, l)

b} r(t)

(l+t)i + (3-t)j + (2t+4)k,

t .
t
2 Ch -z i + 2 Sh 'Z J + 2tk,

8(4, 0, 10)

c) r(t)

curve:

C(2, 0, 0)

209. Find the equation of the FRENET planes of the-curve

;=sin 3t i +cost j + 2 t 312 k 'at
(0, 1, 0)
210~ Consider the spac~ cufve

_ (t-a) 3 (t-b)
-t

r: x -

.-

(t-a)(t-b) 3

show that if the osculating plane at

through

z = t

a point'

r. the osculating plane at-

Q on

passes

Q passes

t~rough

211. Prove that.

- K I'C) = .(< 5
't'3 ( K 1 't- K~ 1 ) =.,..5
K3 ( (('(I

a) ( T I Th, T"' )
b) ( BI B", BIll )

d (!)
ds K
d
ds (~)

212. Prove that

a= rxr"/Ks 3 ,/

= .( r

r: y;

N =

(sri_sr)/Ks 2 ,

4(

( ..r 2 -s2)/'~
s .

J1" 2 6 /

213. P.rove:
a) r 1 .r" = .0,
b)

r"1

1
"

r 1 .rm = -K 2 ,

= li: 1 K" + U

fC 1

r 1_.r 1" 1 = .:.3(1C 1

2

tK 't'l' 1 + KK 1 'f

c ) T I BI = "' Kf.

. 214. Squari ng r"l = -t2T + l 1 N + K't'B

a) t3 _ 1 r;" 2 - ((2 - ( KI) 2

- KZ .

b) r"l
215. Given

obtain

= -3KK 1 T + ( l:"-1< 3 -Kr )N + (2K 1 T + KT 1 )B

r(n) = anT + b N + ens
n

show that

ana

P.

261

n+ 1

= a1

:n

216. Show that

~ole

.of the

Kb

n'

d~
_r dp

from

.,

n+ 1

where

t~e

1
n+ = T b n;+C -n

= K a + b - Tc ,

c . 1

is the distance

tangerit lirie at

P(9, r).

217. Evaluate radius of curvature:

x3
,
b) 3a/ = X3 '
c) r m = am Cos m9.
a)
= -a-xat the pole.
e) r = sin9 - .cos9
tn Cosx,
d) y
d)
2+2 cos 3 9
Cost sint_._
Cos 2t+ tn sint, y
f) X

9)

h) X

.. 2

(l+Cos t)sint,
3t, y = 3t 2 ,

y_= s1n t cost

2t 3 ,
i) x = a cost, y = b sint, z=ct.

218. show th_at

r: x =. f 1 ( 9) C,o s 9- f 11 ( 9) sin 91.

. on

y = f 1 (9)Cos9+f"(9)Cos9

a(f 1

is anhelix traced

(9)+f"(~))

a cylinder

direction

with

U to zpaxis.

219. Find t'he condition. for- a point on the curve of

'

y = f(x)

to haye a maximum:f

220. Evaluate the radius of curvature of

li

a) y = S'tfl'~ .at
. c)

' 2.
t-+t

)(

X=

b) y =

l'.

(l+t)el/t

tn

at

at the origin .

ANSWERS TO EVEN NUMBERED EXERCISES

154. b // axe
158. (3

-2

or
1

4) T.

162. a) /2X+y+z-12
164.- (2,

~_3,

a//c

4)

b)

2x+y+2z+6

= I

262

166. a) arc cos(5/6)

168. a) x-int: 12

xy-trace:

2x+3y

24

y-int:

xz-trace:

2x+4z

24

z-int:

yz-trace:

,3y+4Z

24

b) x-int:.:.l5

xv-trace:

X+l5

y-int:non

xz-trace.:

z-int: ..,9

yz-trace:

170. a)

rr/3

. 3X+Z

b) X = 2.
174. a) !X

3x+5Z+45

Z+9 = 0

b) arccos (1 /5)

17 2. a) 5X+Y

3.

3y-5Z+6 = 0

z_ = -1

y-3

bx-a
)m-

z = 6'

1 76. 8x+y-26z+6

y-b
n

z
T

-J

178. a) 13/M,
180.

rr/3 .

b)

b) 4//6

V6i/ .fi2

182. a) (EP),
184. a)

b)

(HP)~

cl H~one.

b)

~- ~-~-

.,.

.2.

186. By rotationab<iu,t x-axis by an angle rr/4~ (x-2) 2+y 2-z 2

hyperbol~id

188 .. h(x +Y ) +

of one skeet.
\

2a~(x~h)

= 0

. 2
. 2
2
190. (x-a) +._(y-a) = 2a .
1 92. X2+Z 2-y 2 = 0 cone.
. 2
194. X +Y 2 = (Z+l),2 cone vertex at
1 96. 6y = X2+9,

parabolic cylinder;

( 0' 0, -1 )

4,

263
.

198. z +4y = 4 .parabolic cylinder

202. 2x +v -8z 1 = 0
1 1
1
208 .. a) T =(3 ('t+j+k),

V2

b) T = -:-.=: (i-j+2k}, N =.0,

,! 6
c) T =
\

220. a,

~.

(i+2k),

'1/5.
b)

2 J2- ~

N = i '

c) oa

( i.:. j ) . B
B = 0,

B =

b) l /2/2

-i-j+2k,
0

(2i-k),

K. =

./2-/3

II. = l/10
c) 0

.I

I -,
/

'-~->'

ILK-SAN MATBAASI
LTD. STI
30 16 62