HIGHERENGINEERING

MATHEMATICS
(40thedition)
e" *"a"W" Uynt t"*t

for 8.8., B. Tech.,M.8., M.C.A.,

and
B. Sc. (ComputerScience),M.Sc.(Physics/It4aths)
EquivalentExams.(7th edition)
Contents: * Approximations and Errors in Computation
Solution of Algebraic, Transcendental and Simultaneous Equations
Matrix Inversion and Eigen-value Problems
Empirical Laws and Curve Fitting
-
Finite Differences and Interpolation
Numericai Differentiation and Integration
Difference Equations
Numerical Solution of Ordinary and Partial Differential Equations
Linear Programming
Use of Computers in Numerical Methods
Numerical Techniques using FORTRAN -77, C and c'- languages.

s' i#ffi wF.ffiffi f.:trir.f.:,4n

for I & II Semdstersof 8.E., B.Tech',B. Arch. of Indian Technical (Jniversities. (I2th edition)
Infinite Series
Analytical Solid GeometrY
Differential Calculus
Partial Differentiation
Integral Calculus
Multiple Integrals & Beta, Gamma Functions
vector calculus & ortirogonal curvilinear coordinates
Differential Equations of First order
Linear Differential Equations
Solution of Differential equations in series & Special functions
Partial Differential Equations
Fourier Series ; Lapalace, Fourier & Z'ftansforms
Complex Numbers
Solution of Equations & Curve fitting
Matrices
Statistics & Correlation
Probability & Distributions
Numerical Methods
 40th Edition

[.-
B.S. GREWAL,Pu.D.
Professor of Applied Mathematics
Principal Scientific Officer (ex.)
DefenceResearch & Deuelopment Organisation,
New Delhi
Formerly of
College of Military Engineering;, Poorta
Delhi College of Engineering, Delhi

J.S. GREWAL, M.I.E., I. Ensr- (tl.K-), M'1. Mar' E' (London)

NinefeenthReprint : 2010

KHAI\NA PUBLISHERS
Operational Office: 45751L5,Onkar House, Room No. 3-4, Ground Floor,
Darya Ganj, New Delhi- 1 10002
Phone: 23243042 I Fax: 23243043
Despatch Office: 11, Community Centre, Ashok Vihar, Phase 2
Delhi-1 10052. Phone : 27 224L7I
Regd. Office: 2-8, Nath Market, Nai Sarak, Delhi-110006.
P h .2 3 9 1 2 3 8 0
 lrrrl
Piblishedby': .:,:
:: :: ' RomeshChanderKharlna
/oTICIANNAPUBLISHERS
, I :,2,-8,Nath Market, Nai Sarak
." ,', '' Dblhi-,110006(India) ,

@All Ri$hts Reserved

This boohor part thereof cannot b,etrani,slated,or reprod,ucedin uny farm (exceptfor reuiew ar
critiiijsm)witlrcitthewrittenpeimis..sio,n::oftheAuthorandthePubtisi;hers.

ISBN No. 81-?409-195-5

4ffih&dition:2OOz

Priae: Rs.36O.OO
/f"
11 "

Typeset;
CouerDesignandIllustrations by:
SOFTSER\IECOMPUTERSYSTEM
Iftishna Nagai, Delhi

Printedat'
Print India, Dilshad Garden, Delhi and
New A.S. Offset Press,Laxmi N agar,Delhi
 To
My Fathcr
who sented the cause of education
with a missionctry zeal for ovcr five decades
 The book has run through the 39th edition alongwith its twelve heavy reprints in just
two years. This has encouraged the author to retain the same features of the book which
'Z-Transform' has been given in a separate new
have made it so popular. The treatment of
'Objective Type of Questions' have been updated
chapter. The text has been revised and
andgiven chapter wise. The number of solved examples has been ineraased considerably
in each chapter. It is hoped that the book in its revised form will enjoy its ever increasing
popularity.
The author thanks the numerous readers in India and abroad for their unsolicited
letters of appreciation and fellow professors for their suggestions and patronage of the
book. tn pailicular, he is grateful to Prof. Bablu Samanty, W.B.T.U., Kolkata; Prof. Anil
Gupta, S.B.S. College of Engg. & Tech., Ferozepur ; Prof. G.C. Shukla, Institute of
Technolog:f & Management, New Delhi ; Prof. J. Thomas, St. Joseph's College of Errgg. &
Tech., Choondacherry (Kerala) ; Prof. B.K. Yadav, Chouksey Engg. College, Bilaspur
I (C.G.) ; Prof. Jeevargi Phakirappa, V.N. Engg. College, Bellary (I(arnataka) ; Prof. J.C.
Prajapati, Charotera Inst. of Tech., Changa (Gujarat) ; Prof. Ganapathy Subrammanian
K.S., bnn{ Engg. College, Potheri (TN) ; Dr. P.K. Sahoo, Tech. School, Bhubaneshwar ;
Prof. Shiny Philip, Govt. Engg. College, Cochin ; Dr. Kuldip Kumar, N.I.T., Kurukshetra
h
Si
; Prof. Rajni Sharma, DAV I.E.T., Jalandhar ; Prof. Sheikh Ahmed Hussain, Haldia
Institute of Technology, Medinipur (WB) ; Prof. Punit Srivastava, fnst. of Engg. & Tech.,
$ Indore (M.P.) ; Prof. Sabita Mahante, NERIM, Guwahati and Dr. Latika Bhandari, RVS
5 College of Engg. & Tech., Bhillai. '
Suggestions for improvement of the text and intimation of misprints will be thankfully
acknowledged.

New Delhi B.S. GREWAL

 The majority of Engineering and Technolory students in India are obliged to study
Mathematics partty from r ,ro*b"r of books quite ofben not easily available and partly
from sketchy iyclortyled hand-outs prepared by their teachers. Also the teachers are faced
with a real problem while selecting .suitable problems with technical bent. The author has
of
all along been conscious of this want while teaching the subject to a large number
students in Electrical, Mechanical, Civil, Chemical and Telecommunication Engineering.
This book has, therefore, been written with the sole aim of ploviding a suitable textbook
covering all topics common to the various Degree and Post-graduate Engineering courses
in India. \
This book has been written with the firm conviction that a good book is one that can
be read by a student with minimum of guidance from his instructor. Clarity no less utility
has been the author's constant goal which is a result of his long and varied experience as
a teacher and examiner. To achieve this, more than the usual number of solved exanoples
given.
together with carefully designed diagrams and properly graded problems have been
Many of the examples rnd problems have been selected from recent papers of various
the
univlrsity and other engineeiing examinations so as to make the reader familiar with
type of questions usually set therein. As far as possible, it has been planned to make each
A
chapter an independeni unit so that the various topics could be studied in any order.
fairiy comprehensiye index is also given at the end for ready refere-nce.
The author acknowledges,with thanks, the permission given by the authorities of Agra
University, Aligarh Musll* Uninersity, Banaras Hindu University, University of Delhi,
Institution of Electronics and Telecomrnunication Engineers etc. to reprint questions from
their preyious Engineering papers.
The author is indebted to many authors, the faculty members and other friends for the
advice and help in preparing this book.
Suggestions for the improvement of the text as well as intimations of misprints will be
most gratefully received.

Delhi College of Engineering, B.S. GREWAL

Delhi
 Contents

L Solution of Equations I
2. Linear Algebra : Determinants, MaLrices 19
3. Vector Algebra and Solid Geometry 81

4. Differential Calculus , 145

5. Partial Differentiation 211
6. Integral Calculus 255
7. Multiple Integrals -\ zs4
8. Vector Calculus 343

g. Infinite Series
10. Fourier Series

11. Differential Equations of First Order 464

12. Applications of Differential Equations of First Order 491
13. Linear Differential Equations 512
14. Applications of Linear Differential Equations 546
15. Differential Equations of Other Types 579
16. Series Solution of Differential Equations and Special Functions 591
17. Partial Differential Equations 632
18. Applications of Partial Differential Equations 657

li;f 19. Complex Numbers and Functions

20. Calculus of Complex Functions

21. Laplace TYansforms

22. Fourier Transforms
28. Z-Ttansforms

24. Empirical Laws & Curve-fitting 886

25. Statistical Methods 906
26. Probability and Distributions 937
27. Sampling & Inference 982
28. Numerical Methods 1007
29. Finite Differences and Interpolation 1038
30. Difference Equations 1084
31' Numerical solution of ordinary Differential Equations 1096
 \ 1" 1t6
s2. Numerical Solution of Partial Differential Equations
1144
33. Linear Programming

1204
34. Calculus of Variations
1226
35. Integral Equations
1246
36. Tensor Analysis
1267
37. Discrete Mathematics
1304
Appendix -l : Answers to Problems
Appendix 2 : Tables
1378
Table-I : Inverse Quantities, Powers, Roots & Logarithms
1.379
Table-II : Trigonometric Functions
1.380
Table-III : Exponential, Hyperbolic & Trigonometric Functions
Table-IV : Gamma Functions 1381
1.382
Table-V : Bessel Functions
Table-VI : Area under the Normal curve 1383
1384
Tabl+-VII : Values of t
1385
TabIe-VIII : Value s of Y2
1386
Table-IX: Values of F
L387
Ind.e-r
 I. BASIC DATA
l. Uceful Conetants
e = 2.7183 U e = 0.3679 log"2 = 0.6931 log" 3 = 1.0SG
l f = 3.1416 L / n = 0.s183 log" L0= 2.3026 logro e = Q'4,343
r/2 = L.4L42 { g = t.732 1 rad.= 57"17'45" 1'= 0.01?4rad.

2. Convereion Factors
1ft. * 30.48 GrIl= 0.3048 m I m = 100cm = 3.2804ft.
I ft2 = 0.0929 m2 1 acre = 4840yd2 = 4046.77m2

I
;
I
l fts = 0.0283 mg
l dsec = 3.2804fVsec.
1m3 = 35.32ft3
l mile/tr - 1.609km/h.
i I, Syetdms of Unite

l,ength foot (ft) centimetre (cm) metre (m)

Mase pound (lb) gram (gm) kiloeram (kg)
fime second (sec) second(sec) second(eec)

Note. The M.K.S. system is also known as the International Systemof Units fSf .Sysfem).
4. Greek Lettere UEed i

ct alpha 0 theta K kappa T tau :

p beta 0 phi tt mu x. chi
Y gamlna v psi vnu (r) ornega
0 delta gxi npi r cap. Eamma
e epsilon tl eta p t
rno A cap. delta
i iota g zeta o sigma E cap. slgma
l' lambda
6. Some Notatione
€ belongs to u unlon
# doeanotbelong to n intersection
!9 implies )t such that
€ ies &
Fsctor'ialn i,e.n l'='n (nr; 1) (n - 2).....3 .2 . L.
D o u b l ef a c t o r i a l s(; 2 n ) ! =
! 2n(2n-2)(2n- 4),,,"6 . .4.2.
( 2 n * 1 )! ! = ( 2 n - l l ( 2 n - 3 ) ( 2 n ' - 5 ) . . . . . 5 . 3 . 1 .
|
n
Stirling's approximation. When n. is large n t,^, ,tffi . nn e- .
(r)
(iD HIGIJER
ENGINEERING
MATHEMATICS

IT. ALGEBRA
l. Quadratic equation o*2 + bx +c = 0 has roots

a --_ --i a t r-
za

c t + B= - L , crp=g
a
Roots are equal if b2 - 4ac= 0
Rootsa.re real and d.istinct if bz - 4ac> 0
Roofs are imaginary if bz - Aac< 0
2. Cubic equ a ti o n , * 2 * l x 2 + m x + n = 0
cc'lreten
",s'lnetho'ctffi,:";,::T"T::ffi;; l:? -(rr3
3uvy =0
+v3) r ) =,,*,,
(iii) Find u3& u'1.Then llnd u & r.,.
(lv) Get .\,=r, + y and .r=v * l/3.
3. Biquadratie equation , 13 * kx| + lx2+ mx + n. = O
I' Ferruri"s rttctlpcl (i) Conrhinex4 and *'1termsinto a pert'ectsquareby adclingterms
ol l..
{1i)Make R.H.S.a perfectsquareto finrl },.
(iii) Solveresultingquadraticequations. .
II. Dt,stttrtr;'.t
ttrctltrtrl(i) Rcmou.*-t termhy putting r.-
J.= f f- /l
\
(tt) Equatetransformed expression to (r.3+ p), + q) (t: * P1,
+ ci\
(iii) Equatecoetk. or'rikepowersfiom trorhsides.
(iy) Findp,q & q' ancJsolveresultingquaclratics
4.' Cross-multiplication i a1x * bry + c1z= 0
a2X+b'Y+C2z=O
x,-.-= Y - z--
Thcn,-
op2 - b2 c 1 c 1 a 2- c 2 a 1 a 1 b 2* a 2 b 1

6. Method of Least squares : curue of best y - a + bx * rr2

fit
[t{orntal equations : Ey = na + bf:l-,+ cLxz,
Irr = a l :x + b rr2 + c Z x S ,E * Zy -aLxz + bl r3 + cf-x|
Tn find a. h, c, solve these equations.
6. Progressions
(r ) NumbPr$i a. o, + d, a + 2d,...
are said to be in Arithmetic progressionA.p.)
I t s n t h t e i r mT , r = e * r . - t d a n d s u m S , , - | e o + n - L d>
( ir ) Num ber s a. o ,,u ,2 ,......
a re s a i d to b e i n Geom,etri cprogressi on(G.p.)
- I
It s nt h t er nt 7, ,= a r' t = U ,* :-t .,S * =
a n d s rz rn ,S,, (r< 1)
T' ;
 INFORMATION
USEFUL (dD
(iir) Numbers L/a, L/(a + d,),L/(a + fuJ),......are said to be in Harmonic progressionH.P.)
(r,e.a sequenceis said to be in H.P. if its reciprocalsare in A.P.)
Its nth term Tn- L/(a +FL dr.
(fu) If c and b be two numbers then their

Arithmetic mean = + b), Geometric',lnean= ffi, Harmonic mean = la.b/(a + b)

t ,"
(u)Natural numbersare 1, 2,3, ......,n.

={"';"1'
,En3
7. Permutations and Combinations

n' p - n! nC,r_'r=nCr; rCo=L=nCn.

= : r ev-r=- , , r ! . , =- 3
; ., :
r- (n_r) ! r iI(n_r) ! i
8. Binomial theorem
(i) When n is a positive integer
( 1 + * ) n = 1 + t C t " + n C z x z+ n C g r 3+ . . . . . . .+ o c n x n .
(ii) When n is a negatiueintegeror a fraction
(n 2)"8
(1+r)n = 1 + *ff x z+ n * . . . . . . .o. .o .
"* I.L)(\-

.' nC-=
(iii) Binomial coefficients , '!'

g. rndice E i a* xon =a**n"'";;::=r::"; o0 = 1.

10. Logartthms
(i\ Notural logarithm log.r has basee and is inverse of e".
Commonlogarithrn loglg * = M log * where M =logrc e = 0.4343.
(ii) log" 1 = 0 ; logoQ=-oo (a ) 1) ; logoa =L.
(ar) log (mnl =log rn+ logn; log (m/n) -log m- logn;log(m')=nlogm.
11. Partial Fractions
ao{ +-ar{.-|-..+-"'+a,
A fractionof the,or
--- -'
bu{t+ hy'r- | + ....,,+ bn i
in which nr and n are positive integers, is called a rational algebraicftaciian. When the
numerator is of a lower degree than the denominator, it is called a proper fraetion.
To resoluea given froction into partial fractions, we first factorise the denominator into real
factore. These will be either linear or quadratic, and some factors repeated. Then the proper
fraction is resolved into a sum of partial fractions such that
(r) to a non-repeatedlinear factor x - a in the denominator corr€spondsa partial fraction of
the form A/(r - a) i
(rr) to a repeated linear fantor (x - a)' in the denominator correspondsthe sum of r partial
At 42 4t o'
fractionsof the form * =* , +... + ,
x-a (x-a)" (x-a)o (r-a)''
(iv) HIGHERENGINEERINGMATHEMATICS

Giil to a non'repeated. quad.ratic factor (*2 + ax + b) in the denorninator, corresponds a

partial fraction of the form * U-
--- # ,
x ;z+ a x + b '
(iu) to a repeated' quad.ratic factor (*2 + ax + b)r in the denominator, corresponds the sum of
* *
r partialfracrions
of the form t*-* "t, * , !* * lo
"r,.r,+......
.*'+ax+b (*'+ax+byz ( f ' * a x +",b ) r
Then we have to.determine the unknown constants A,AL,B1 etc.
To obtain the partial fraction corresponding to the non-repeated linear factor x - c, in the
denominator, put x = d. eu€rlwhere inphe giuen fraction except-in the factor tc- a itself.
In all other cases, bquate the $iven fraction to a sum of suitable partial fractions in
accsrdance with (j) to (iu) above, having found the partial fractions .d"responding to the
non-repeated linear factors by the above rule. Then multiply both sides by the d*tto*inator of
the given fraction and equate the coefficients of hke powers of or rrrbstitute convenient
numerical values ofr on both sides. Finally solve the simplest of the" resulting equations to find
the unknown constants.
12. Matrices
(D A system af mn numbers arranged in a rectangular atcty of m rotils and n columns is
called a matrix of order rrl x n.
In particular if rn = n, it is icalled, a square matrix of ord,er n.
('il Two matrices of the same order can be added,or subtractedby
adding or,subtracting the
corresponding elements.
(ii) Product of a matrix A by a scalar k is a matrix whose each element is & times the
corresponding elements of A.
Gu) Two matrices can be muttiptied only when the number of columns in the first is equal
to the number of rows in the second. If A is of order m x n and B is of order nxp, thenihe
product A-B is a matiix of order ffi x p, obtained by multiplying and adding the ,o* Ll.-ents of
4 with the corresponding column elements of B.
-(u) Transpose of a matrix A is the matrix obtained by interchangrng its rows and columns
and is denoted by A'.
A square matrixA is said to be symmetric if A=A'and skew syrnmetric if A=-A'.
fui) If A and B are two square matrices such thatAB = I (i.e. a unit matrix), then B is called
,the inuelse of A and is denoted by A- 1. Th*r, M- L = A- 1A = I
13. Determinant
(i) A determinant is defined for a square matrixA and
is denoted bV '--
lA ''| . Unlike
- a matrix
it has a single value e.g.
d ,1 b L c 1
b 2 c2
lAl= a2 bZ c2 - c I 1 - b r a 2 c 2 * c l a2 b2
bg cg ag ca as bg
a,g bg cg
-
= eL @zce bscil - bt (azcs - ascz)+ cr (azbs - aebz)
In this w&Y, determinant can be expanded in terms of any row or column.
Gi) Properties. I. A determinant remains unaltered if its rows and columns are
interchanged.
II. A determinant vanishes if two of its rows (or columns) are identical or proportional.
III. If each elemerrts of a rqw (or column) consists of m terms, the determinant can be
expressed as the sum of rn determinants.
ry. If to each elements of a row (or column) be added equi-multiples of the corresponding
elements of two or more rows (or columns), the determinant remains unaltered,
 INFORMATION
USEFUL

III. GEOMETRY
l. Coord.inatesof a point : Cartesian (x, y) and polar (r, 0)'
tc=rcos0, !=rsin0
or r ={ (*2*yz), 0 =tan-t,gt*).(Fig. 0'1).
-
Distancebetween two points (rr, y1) and (xz, yz)= { [(rz *t)z + (f z- yr)2]
Point of d,ivision of the line joining (rr, y1) and (xz, yz) in the ratio rnl i tn2 rs
( mpz+ mzq mgz+ mzltt\
r n 1 *m 2 ' *r* mz
)
t
In a triangle havrng vertices (r1, !t), (tcz,y2)and (la, ye) Y
)c1 tt L
(i) Area =| xc2 lZ 1
s3ls1
(ii) Centroid'(pointof intersectionof medians)is
(h*!r2*)cs tr+tz+v3 )
l.s't)
(iii) Incenfre (point of intersection of the internal bisectors
of the angles) is
(*r+bx2*crcs aYL+bYz+t t )
q,+b+c ' a+b+c
{. )
where d,b,c are the lengths of the sides of the triangle.
(iu) Circurncentreis the point of intersection of the right bisectors of the sides of the triangle.
(u) Orthocentre is the point of intersection of the perpendiculars drawn from the vertices to
the oppositesides of the triangle.
2. Straight Line
tz- tt
(i'l Slopeof the line joining the points (rr, yr) and (r2,y il =ffi

Slope of the line ar + by +c = 0 i, - *b i.e. -g""gi{"

coeff.ofY
i

,/'
(ii) Equation of a line '
(a) having slope rn and cutting an intercept c on y-axis is y = rrbtc
+ c.

(b) cutting intercepts a & b from the axes is { + ! = t.

(c) passirrgthrough (rr, yr) & having slopem is y -Jt = m (x - x1)

(d) passing through (rr, yr) & making an ,j|with I e r-axi, i, 4 =ffi =
"
(e) through the point of intersection of the lines a1x,*bfl +cl =0 and d,2x*bzy + cz=0 is
a1x * bg + c1+ k (a2x + bzY +ce) = 0
* 1 mL* tnZ
(iii)AngtebetweentwoIineshavingslopesm1andnz2istan.ffi

T\po lines are'parallel if frl= rflz

T\vo lines are perpendicular if mt?z='I

Any line parallel to the line ax+by+c=0 issr+by+k=0
Any line perpendicular to ax+by*c=0isbr-aY+k=0
 (vD
HIGHERENGINEERINGMATHEMATICS

(iu) Length alFthe perpendicular from

(xt,yt) to the line ar + by +c = o, ffil t
' i, lDYr=+
'lPz + bz)
$. Circle
G) Equation of the circle having centre (h,
k) and radius r
{s (r - h)? n (v * k)z = (Fig. 0.2)
"2
Qi) Equation x2 + y2 + 4, + Zfy+c = 0 represents
a circle
tlaving cerrtre(-g, -fl and radius-= rl@z+fzl
r).
(iii) Equation of the tangent
at the point (nL,;ito the circle
i'*y' =o2 i* ffit+!!t=o2.
Gu') Condition for the line y = rnx+
- c- ' -to
r touch the circle
*2*y2=o2i*c=a{1r+rnz)t.
(u) Length of the tangent from
the point (rr, yr) to the circre
*2 * y2 + 2sx + zfy+ c = o is {1rf *
r?* Wr; ;;r + i).
4. Parabola
(r) Stand,ardequationof the parabola
, i, y2 = 4ax.(Fig. 0.3)
Its parametrie equationsq,rex,= at2, =
! Zat.
' Laqus-rectum
LL, = 4a, Focusis ,S(a,
\--' 0)
-'
Directrix ZhI is * + a = 0.
(rr,Yr)
Gi) Focal distance of any pbint p (xt,yr)
e on the parabola
!- = 4ax is ,SP= x1* e.
Gii) Equution of the tangent at (x1,y1)
, to the parabola
=
l- 4ot is yyr = ?.a(r + rr)
,ir, Condition for the line y = mx,* c to touch
,, the parabola
y =rlofflBc=a/m.

\v) &quation af the normal to the parab

ola yz = 4axin terms of its slope zl
is
!omx-Zam-o^3.

6. Elllpse
(i) Standard,equation ofthe
ellips" i, 4 * 4 = f .
do b"

Itspurametric equationsare x = acos

0, rt:;t;rl?.
Eccentricitye={tf - b2/a2y,
Fig. 0.4
Latuq-rcctumLSL, = Zb2/ a.
Foci ^9(-ae,0) and.S,(oe,0)
Directrices ZM (r - - a/e) and Z,M, (x =
a/e).
6il sum of the focal d'istancesof any point
on the ellipse is equal to the major axis
i.e,
SP+S? =Za.
(iii) Equation of the'tangent
at the point (rr, yr) to the elripse
 (v,,)
INFORMA1ION
USEFUL

*Z n, xtcl . tlt _
3.h=1isi."F=L' g.
*2 n,2 t. 2
-- c fo toachtheeuips" t is c = 4@'*', + b2)
for theriney rm,c*
(iu)cond,ition 5.fi=
6. Hyperbola
(i) Stand'ard,equation ofthe hyperbolais
*2 n,2
( F i so' ' 5 ) s
o\-b=t'
Its parametric equationsare
x--ct,sec0, !=btan0'
EecentricitYe ={qf + b2/a21,
Fig. 0.5
Latus'rectum LSLi'- 2b2/ o'
(* = - a/e)'
DirectricesZIaI (x = a/e) and'ZluI , ,
Tri'r-it"irA "irn, at thepoint(rr'vr) to thehvperbola
tu'ngent
*2 -t-=1'-ffiL U]-''
7_7=rrs:l_-62=,.
(iiilcond'ition
*"n" " i u'
;: &==,,"' =:,'"::: :li',*o"'uorf *n.;
o2 b2 '.:.
2.2^'r '";:in1
1 are:.t= 0 andI-t= o'
of the hvpei:rbota
(iu)Asymptates
3 b=
.,s axesis *y = o2'
(u) Equation of the rectangula,rhypetbola with asyrmptotes
=c/t'
Its parametricequationsare x=ct'!
7. Nature of a conic
+ c = 0 represents
The equation'axL+Zhxy +byz + zgx +Zfy
a.hg
(il a pair of lines, if h b f (=A)=Q
sfc
( i i ) a c i r c l ei,f a = b , h = 0 , A * 0
(iiil a parabola,if ab - h2 =0, A * 0'
if ab - h2 > 0, A * 0'
(iu\ an elliPse,
(u) a hyperbola,if ab - hz <0, A * 0'
o +b = 0'
and c recta,ngularhyperbolaifin addition'

-i1 l
(vitD HIGHER
ENGINEERING
MATHEMATIGS
8. Volumes & Surfaee areas
Solid Volume Curaed suface area Total surface area
Cube (sideo) o3 4az 6az
Cuboid (length r, tbh 2(l+b)h 2 (lb +bh + hll
breadth b, height h)
Sphere (radius r)
i *"t 4nr8
Cylinder nr2h Zr,rh 2rv1rl,+h)
(baseradius r, height fr)
Cone 2h nrl trr (r + l)
(base radius r, height h) ln
where slant height / is given by / = {(rz + lrz).

fV. TRIGONOMETRY

l. 0o- 0 30 45 60 90 r80 270 860

sin 0 0 u2 U,lf ,lTtz I 0 -1 0
cos 0 I ,l{tz Ll'.tT u2 0 *1 o I
tan 0 0 Ll..[3 1 {5 oo 0 .a oo 0

2. Signs & variations of t-ratios

Quadrant srn 0 cos 0 tan 0

I + + +
(0 to I) (1 to (r) (0 to
"a)
II +
(I to o) (Oto-1) (-*to0)
III +
(0 to -1) (-l to 0) @ to "")
N +
FI to O) @tol) -a to O)
8. Any t-ratio of (n . 90" + 0) = t scLmerAtio of 0, when n is euen.

rhesign+ or- is t;obe;;;:;';n"{^2';#;;::,"#;r*o n . e0.* E ties.

e.g. sin5?0o=sin (6x g0"+B0o) =-sin SO"=-
*;
tan 315o= tan 13x 90"+ 45")= - cot45"= - I

4. sin (A t B) = sinA cosB t cosA sin B

cos(A + B) = cosA cosB T sinA sin B
sin 24 = 2 sinA cosA = 2 tan A/(l + tanzA1

cos?ll = cos2A- sinz4 = l - 2 sin2A=z.os2A - l = t -tan?A

1+tanzA
B=
s. tan (A ! b=' .!TA ltun = ZtanA
1 + t a n A t a n B ; tan 2A
 INFORMATION
USEFUL

sinA cosB = t [ritt d + B) + sin (A -Bt

cosAsin B = t trin (A + B) - isin(A *B)l

cosAcosB = | l.o, (A + B) + cos(A -8)l

sinA sin B = t t"o, (A - B) - cos(A + Bll

sinC+ sinD =2*i" ff;.o, $!

s i n C * s iD
n =2**gPri"gf,
D =2*"ry
cosC+ con ,o*9iD
gf,
cosC- cosD =*Z sint#ti"
a sin x +bcost = r sin (r + 0)
a cosr + b sint * r cds(r - 0)
|
Wherea= fcos 0, b =rein 0 so that 7 =,l1al +b21,0= tan- (bto)
9. In anyL ABC: (l) a/sin A = b/sin B = c/sin C (Sineformulal
(ir) coso=W' (Cosineformula)
(iii) d's b cogC + c cosB (Proiectionformula')
(iv) Areaof LABC=llz-sinA = @ where = (a + I>+ c)'
" ;
1O. $eries
('il Erponential series:er = t * . + """ oe
ft *+ *
(ii) ,Sin, cos,sinlr, cosh serdes
.*2*4
sin r * iE*fi + - ......s' cosr=1-fi+;l-
5T

sinh r =tr . . * .......*, eoshr=1+

*.*........&
* *
(iii) LaS seriee

r o s( l + r ) * . - * . - ..
*
GulGrcgoryseries
t",
*8'*6 . t^1a- -n- rh- *t*1r ^=-i.l r -t -o1* i *! x - -=- .. r + t ' * {i t
t a n * l t = n - S + T - - . . . . .@
It. (i) Complernuftber z3n+iy tr (cos0 +l sin 0)=rudo[*uuFig.0,tl '
{lf Euleds th€or€fr : cos0 +'i sin 0 = cd
(ird)Demoivretg theorern : (cos6 + i sin 0)t = cosne + i sin n0
(x) MATHEMATICS
H|GHM ENGINEERING

'
12. (i) Hyperbolic functions. (r) sinh* =+; cosh*={Ju '
2',
titlt:
tanhr= c o t h r = S * 3 : s e c h* = - 1 - - : c o s e c h t = . f -
cosh.r' srnhr' cosnr. slnn f,
Gil Relations between hyperbolic and circular functions :
sin lr = i sinhr; coshr = coshr ; tan ix =itanhr.
|jii\ Inverse hyperbolicfunctions:
t
sinh- 1r = log h * {f* rl; cosh-1* = log Ix +'11x2-'1)l; tanh- * =*t"*
z i#
V. DIFFEREI{TIAL CALCULUS
1. Standard limits :
. ,[-an n -*r L, s
(t) Lt = ha,n ft any rational number(ii) tr; Eln = 1
x - + a +x ' a .r-+o x'
(iiil Lt (L + x:L/' - Gd Lt *r/x = 1
x-+0 " x-+@

- ax-l r
(u) Lt =log"a
r-+0 T x

2. Differentiation
d (u\ udu/dx-udv/dx
6*(uu)="#*r# aFJ=T
du = du dv (chainRule)
ffi tt e *r*+b)n=n(ax+b)n-t.d

GD*(e')=e' r., - ax lose@

*.
.I d,r \ 1
x) = l/x lto9a *) =
;;(loge * * log o
,1
=
(trt) (sinr) = cosr
# fi<"o"r) sinr
tl
=
fr tt"" r) sec2
r ft<"otx)=-cor""2,
d (sec = = - cosec
x cot,
x
r) secx tan x
; ft {"or""r)
dr1
(iv);;1sin-'r)=
t-A St.o"-1")=-fr|F
d . . - r- 1 -1-
(tan x)= {drrc' -o- r l t ) =
4r 1+F l+xZ
d - r- x ) = ; 1 . " z1_ 1 1 d - r- . -1
(cosecx) =
4l(sec ;;
"./A_r1
(q *) = coshr f.orr,r) = sinhr
*(sirrh $
d
= L* rcotnr)= - cosech2 r
ft tta"ttr) sech2
r .t
-
(ui) D" (at + bl^ = m(m - L)(m * 2)......(m- n + lXar + b)^ 4' a'n
- 1 -
D" log (ax + 6) = (- l)n (n 1)! an /1ax + b\"
 INFORMATION
USEFUL

Dn(e*) =nnew D"(a*, = tnn(log a\n ' a^*

D" fsin (u +ail = o" fsin (w + b + nn/Z))

Itot. " J lcos J
* ,u* * b')n'2'* @*:,c+ ntan-1
:'")
H: ,:,)]=i"' h:
( u i i l L e i b n i t z t h e o r e m :( u v \ n = t t r n * n C t r n - l u t + n ' C 2 u n - Z v Z + . . . . .*.n C r u n - r u r * """ +nC,rrtn

=/(0)+
S.(i) Maclaurin'sseries.f(r) fr/'(0) * f
+ ...
*hf "rr,** f "'(o)
frf"'(
2.3
(ii) Taytor's : f(x+ a) =f(a)*
serdes 't"l *ftf "{of* f "'(a')+
ft.f h
4' curlrature
,2t)B/2 (r2 * 17)3'2 d;"
(i) Rad.ius of curuature P =\-, P=F;zrTrrr; P='Tf

(ii)Centre :i =*
ofcuruature i =y*1 (r *v?1.
#,
(iii\ Evolute is the locusof the centreof curvatureof a eurue.The curve is called the involute
of the evolute.
(ru) Envel ope of a curvef(*, y, O)= 0 is the 'o, eliminant frorn
= 0 o"a =
f (x,Y,cr) ffi(x,!,0) 0'
Theenuetopeof the normals to a curt)efs its euolute.
6. Asymptotes
(i) Arymptotesparallel to x-axisare obtained.by equating to zero the coefficientof the highest
power of r in the equation, providedthis is not merely a constant.
Asymptotesparallel to y-axisare obtained by eguating to zero the coefficientaf highest power
of y in the equation, providedthis is not merely a constant.
Gi'tOblique asymptotesare obtained as follaws:
Put r = 1,J = nz in the highest degreesterms getting 0"(m)
Put Q,r(nt)= 0 & find the values of nz.
Find c from c = - Qn- {m)/Qn'(ml
If two values of m are equal, then find c from
n2
* c['n - t (m) + 0,,- ztn) = o '':,'
i 0"n{*)
-
The asymptotes is y ftur + c.
(iii) Asympttotes of polar curue l/r =ffe) ie r sin (0 - a) = l/f(u) where a is a root off (0) = 0,
6. (i) A curve is symme*ricatabout x-a^tis;if only even powers of y occur in its equation.
(iil A curue is syrnmatrsecfsbouty"axii, if ontryeven powers of .t occur in its equetion
(iiil A curue is symmetriealahout the line ! = x, if on interchanging r and y, its oquation
remains unchanged.
(iu) A curueposJesthraugh the origin, if there is no constant term in its equation.
(u') Tangentsto curueat the origin are found by equating to eerothe lowest degqoeterms.
7. Pertial Differentiation
(i) Ealer's theorem.If u is a homogeneousfunction in r and y of degr@ R, then
(xfl) HIGHERENGINEERINGMATHEMATICS

du
* a * * .Y --du
a y = n---
u ': - a x d t*y2+
*r* +zw;fu =n(n*r)u
dxz ,-t ayr-"
(iitCha,in =*# .# = (x,y,\,r=O(t)r=V(r).
rulez # #,if u f
$ i i D * = - 3 * / # , i r o( r , y=
)c
(iu)Jacob u' ' ) d (u' u) | autar autb
ianJ I
.,y r aunvI
)-'aGt= larnx I
If J = d (u, u)/0 i.r,y) &,J' = d (x,y)/d (u, v), then JJ' = t

#=HH
(u)raylor's : f (a+h,b +k\ f (a,b).
series !** u'# .t t..
Io *
)t. [t $ # )
(vi) IVIarimaMini,ma(a')# = O, =O
$
(b',# #,#.; #,0 rorminimum.
#, "",'"1*,*,,*
VI. INTEGRAL CALCI,]LUS
l. fntegretion
I *ni.t
( t ) J " "d x = ; * (n*-1) r
I I o.=los"
I ,*d*="* '
=
J a*d,x a*/log, a
(tt I sinn fls =- cosr =sin r
I cos.rd'x
=log sin r
J t"r, x d.x=- logcosn I *, r d,x

J t*. x d.x=log (secr + tan r) r - cotr)

J .o*", x d.x=loB(cosec

J ,"r' r d*= tanr J .or"r' x d.x=- cotr

I d.x -ltr
$iil[j#=It""LL lffi-;r='i'-tf
I dx 1 t e*x I dx .nh-l{
I;\t=r'tos d-n, J qoz*"1 =tt a
I dx 1 , r.-d
I Tli=fr tosn+o I trfq,='o'h-'i
- x\ d.x
Jl@2
r;uy =ry +$ "in-t L
xzt -tL
I u,o,+ xz)4* =r'li'.z-+ + { *i'r,
 INFQRMATION
USEFUL

Juo,-or)*=W-{"o",lr-
<ulI ,*sin bra* =ffi@ sinbx- bcosbr)

bra, = br)
cosbx+ bein
I uorcos ffi(a
=sinhr
tra J sinh.rdr = coshr J .*r, x dx
=logsinhr
J turrt x d.x
=log coshr J cothr d.x
=- cothr
2
J .n".t x (x
=tanh r J .or*.h2x d'*

,,,Io,,sinnr*=II,,cos'rd*=ffi-F,onlyifni,eu*,,)
x d'r=
ff'' sinmr cos'
x ft' o"tYif bothnzandn are"u"tl
(z' )

(uiii)[i ro,* =[: f @- x\d'x

\f (x)isaneuenfunction
I - t o, dx= 2J, r o, d'x, f
"
=$ , iff (r) is an odd function'
pb ?o
dx,itf (2a-r) =f(r)
Jo f $\ dx= 2 Jo f @)
=0 ,iff(?n-x)=-f(x).
2. Multiple integrals
=f:, f;,2d.xdv* fl, f;, f:,dxdvd'z
,, or"r"=1"' fl,o.d,y; voturne
II'pdrdv =IW IIrPd'xdY
la,mina,i-_
ei)e.c. ofa ptane ,!
1jTpd* d,
JIJ xpdxdyd'ziIIJ vPd&dY&- JIJ zildx,ty,tu
. C.G. of a salidi' =
pd*dy dz TIFo,u;'
IIo-d*dv Ilordxdv
A
,zl: .. .,

A
-FT-l-
ftii't Centre of pressurex = ,f =
I J Pd x d Y
A
(xiV) HIGHERENGINEERING
MAIHEMATICS

Gu)M.L about x-axisi.e.L = IJ J- v t 02 * d*dyd"

"21
M.I. abouty-a,tcis
i.e. Ir= J J I, v , ,"' + x\1d^rdydz

M,I. aboutz-axi,si.e.I"= J J I , ,*' +y\ dx d,yd.z

v
-
-
8. GammafunctionI'n = ft o e-' xn L d*= (n - 1) ! , I'(n + 1)= n I.(n)= n!, F(1,,/2)
= .Jn
r1 Fm Fn
t (t - *r)n-L
\ r " , = .Jr 0xm-
Betafunction |F (m,rr) " \' , *d*=
-f1m+n1

VII. VECTORTI
1 . ( i ) I f R = r I + y J + z K t h e n r = l R l = . / ( " 2* y 2 * r 2 1
(rt) & = Position vector of gPosition vector of P.
2 . I f A = a l I + a 2 d + o g K B = b 1 | + b 2 . I + b 3K , t h e n
G\ Scalarprod,uctt A.B =ab cos0=ar br +a2b2+ agbs
Qil Vectorptrod,uct'A x B = ob sinO ft = Area of the parallelogram having A + B as sides
I I K
a1 a2 ag
b1 bs bg
(ttr)B A .l-if A.B=0and AisparalleltoBif AxB=0
a1 a2 ag
8. (i) Scalar triple product tA B CJ = b1 b2 bs = Volume of parallelopiped
C1 c2 ca
(tt) If [A g CJ = 0, then A, B, C are coplanar.
Gii\ Vectortriple productA x (B x C) : (A . C) B - (A. B) C
(Ax B) x C = (C . A) B - (C . B) A

4.(i)srad
f =vf=# r. $ t .#*
divF=vF=+ *y**
dctl&
IatK
curlF=VxF=
aaa whereF=fi l+fz.I+/3K
hfz fs
(tt) If div F = 0, then F is called a solenoidal uector
(ttt) If curl F = 0 then F is called an inotational vector
6,Velocity=d'R/d't;Acceleration:dzntdtT;Tangentuector=d&/d.t;Normaluector=VQ

G.Green's
theoren,,C gd' + vdyt=
I "--'- [ f +dr - P
'v'r I .l[ .r"
Ay
)rr"
I )-
 INFORMAIION
usEFuL (xv)
Stoke'stheorem:
'Ic F.dn= "[s c u r l F . N d s
I
Gaussdivergence theorem: 'Is F . N d s = J d.iuFdu

7. Coordinate systems
Polar coordinates Cylindrical coordinates Spherical polar
{ r,0 ) (p, 0, z) coord,inates(r'. 0.6)
Coord,inate r=rcoS0 x=pcosQ N=rsin0cosg
transformations !=rsin0 y=psinQ !=rsin0sing
Z=Z z=rcos0
Jaeobian d(r, y), -'
_- d(x, Jr,z) _ ^
-
E1r,
e; E(p,0, z) Y #t#=r2sino
(Arc-Iength)2 (ds)z=1d,r)2+rz1dIyz d.s)z=(d,r12+rz1doyz
dxdy=rdQdr + (r sin o)z (dotz
VoIume-element dV=pdpdQde dV =rz sin g d,rd,0
VIII. DIFFERENTIAL EQUATIONS
1. Equations of firet order
G) variablesseparabte: f Qr)dy = g(x)e, Ifo ay = I g(x)d,x
+ c.
Gi)Homogeneous equation#=f@,y'1/S@,y)wher ef(x,y)& 0(r,y) are of the samedegree.
Puty = us so that dy/dx = u + xd,u/dx.
$ii) Equations reducible to homogeneous dy ax+br,+c
form ' .+-

dx- a'x+b'y +c'

Whena/d' * b/b',put * =X+ h, y =y + k
Whena/a' = b/b'o put ax + by = tr.
$ul 7s;6txitz'slineq,requatio" , py =e where p,
#+ eare functions of r.

I.F. = r!'or,then solutionisy (tr.F)= e (I.r.) d,x+ c.

J
fu) Bernoulli's equation.d.y/dx + py =ey', reducible to Leibnitz,s equation by writing as
- it
-n -'
" * + Pyt = Qand putting yr
fui) B**t equation : M (x,y) di + N (x,y) dy = O
Solutionit 'gt
J. a* * I (terms of N not containing x) dy= c, provi dediMtcly = lN/tb.
cons.)
biil Ctair&ut's equation : y = px +rfu) where p =dy/dx
Solution is obtained on replacingp by c,
2. Linear equatione with constant coefficiente

hr#+ ..... r#* k,r=x

+ kn-
#.
- I
S y t nbolicf orm : (D n + k 1 D n * .... * k n _ 1 D + hn)! = X .
(xvr) HIGHERENGINEERING
|v|AT}'IEMAIICS
I, To find C.F.

Roofs of A.E.
(il my fit2, tttg,....
clemtx + cgem* * caemar .u ...
(ii) my tn.1,ntg,.,..,,
(cr + c6x)emrx+ ca4s + ....
(iii) cr + ip, cr - rB, zrs, ....
e* (cr soefu + cg sin pry + cadn# a ....
(iu) cr t rp, cr!ip, ms, ....
eu [(cr + cosFr + (es+ ear) sin Frl + egem#+ ..,,
"ar)
II. To find P.f.
(i) x = eoY,P.I.
= ro", put D =c,
*
Art) [/1o)* 0
=* putD =a, ffic)=o,f'(a)*a
fu,effi,
*"t*effi.put
'r'- D a e ,
f"(Dl- v'@)=a,f"(a)*0
(ri) X = sin (ar + b) or cos(cr + b)
P.I. = l*- sin (ar + 6) [or cos(ar + A)1,put D2 = - o2,
0(D*)
[g1-ozy* 0l
=t I
sin (o* + b) [or (cosar + b)1,put D2 = - q2, [o(- a2)= o, o'(- oz)* ol
;65
=*' (ar + b) [orcos(a*+ b)1,putD2= - n2, [0'(- a2l=0, 0" ? azt* o
Vfusin
$iil x=s"', P.I.= = tf@)l-r l".Expendtf(D)l*llo powers
ofD as for as
h** "rcending
D" & operate on #'n.

tiu)
x=e'''v, =
P.I. =e*
#""'v ffir.
III. Complete solutlon : C.S. is y = C.F. + p.I.

B. Homogeneous linear equation , .t + kpz hr.. + hgr - x

* #. *
rcducesto linear equation with constant coefiicientsby putting

.r=(,r." =DJ,,
r' D (D- l)y,.t (D- 1)(D*z)y
# #= *=D
4. Lagrange'e Linear partlal difrerential equafion
Pp + Qo = R, P, e, J?being functions oifx,y, z.
To solucif (it firnn the subsidiaryequationrf =t =#
(ii ) solvethese'equationsgiving r, = ar u = b.
(rii) Cornpletesolutionis tr, ,l = 0 or u =
0 f(u).
5' Honro* atronl w,:: constantcoerficients
":.lt."Tuu*ri
U *rt + "" +n"#=F (x,Y)
a.." rr',5
S l r n b o l i cf o r n t i ( D ' , + k 1 D , t - l D , + . . . . . .+ h , r D ' , t y Z = F ( x , y l
USEFUL
INFORMATION (xviD
I. To find C.F.
Rootsof A.E.
0) h 0 + mfc) + fz 0 + mzx)+ fs U + mgx)+ ....
(ii) ft 0 + m$) + xfz g + mp) + fs (y + msx)+ ....
Qii)
h (y + mfi) + xfz g + mp) + xzfy (Xr+ msx)+ ....
II. To find P.f.
G ) F ( x , ! ) = - e o x + b tP, . I . = , , + ^ ^ = a , ,D , = b .
r lu, u') " o * * b ! , p u t D
Uil F (x,/) = sin (mx +ny) or cos (mx + ny)
-
P'I' = or cos(mx + ny),put D2 = - *2, DD' - - ffift, D2 = - n2
1d*,.D4sin
(iii) F (x, y) *-i*yn,PJ. - v (D, D')l-' **yn.Expand v (D, D')J-I and operate on x*yn.
(iu) F (x, y) is. any function of r & y, p.I. = F (x, y).
# r,
' ResoWe
L/f(D, D') into partial fractions consideri ng f (D, D') asa functio n of D alone and
operate each pertial fraction on F (x, y).
III. Complete solution : C.S. is y - C.F. + p.I.
t*. FoURIER SERIES
.,
l. f ( x ) = *, oo* d 1 c o src* d ,2 c o s2 x + .....+ b1 si n x + b2 si n2x+ .... i n (0 2.n),

, 1f2n tr?n tPn

oo=i Jo f@)d.x,
where flr) cosnxd,x,u,=*Jo /(r) sinnxd,x
",=i,J,

2.f(x)= * : @

) ancosnx+ I 6,, cos nx in any interval (0, 2c),

f "o n=l n=l
- r P" tf\c
whereoo=* I;" fo, d,x,a,= J;"/(r) cos* *,b, =: I, Oo sinw dx
;
3. For enen function f(x), Fouri,er expansion contains only cosine terms.
.
i.e.as = zlt ^-. 2 lc
= @ d.x,bn- e.
t Jo f(x)d.x,on t Joft*y.o,
For odd function f(x), Fourier expansion contains only sine terms.
t lc
i.e.as= o,en= o,bn= Jo/(r)sin a..
i ff
X. TRANSFORMS

1, Laplace Transforms. L (f (t)= e-'t 11t1

dt
l;
I
(i)L(l)== (ii) L (tnl=-4: 1
s G i i ) L ( e o t'\ =
sn+I s-d
(id L (sin af) = =Q"'i (u) L (cos af) = --j- (ui)L(sinh at)* =o,=
,2*o2 so+ao ,2 -' o2
(xviii) , HIGHER MATHEMATICS
ENGINEERING
:
tuii) L (cosh at) = (uiii) L eotf(t) -F (s - o) (id L f '(t) - sL f(t) - f(0)
*
Sn [r I f-
(x)L Lt"f(t)l = (- I)" fr F(s)l (xilLLin,, J" &s)g__
,
l
ds" J= - o s
t < a- ' *as
( x i i ) u ( t - a ) - t | ,0: w" :h-e- ;n- - s -
@iiil L^i-G -a)l -=e (xiu) LE (t - a) - s
I l w h e n t >a

2. Inverse Laplase Transforms

'J=
(ilL-'[* t
(iu)L-'[ri" sinar
J=*
(ui)L-tl ,1=+sinhor
"t
lt"-o" )
a
(uiii) t sinlor
*
"-' [ @;g]=
l*
3. Fourier Transforms : F(s) = J__ f(t) e

Fourier sine transformi 4(s) = J; fQ) sinst dt

Fourier cosine'transform;F"(s)= J; flr; cosst dt

(^2 \
Fl + l=- t?F(u\.
la*" ) - .€
-n
{, Z-Transforms zZ (unr= I un z
. n=0

- - . . . .- . r \ *
(ilZ (l\ =;= (ii)Z(n)=--z ( i i i ) Z ( n z ) = "" 2 "
(z - L)o -
(z 1)3
_? o"
Gu\Z (a;'\= fu\ Z (nan1=' u
z-a (z - a)'
z sin o
fuil z(sin no) -
,2'- ?i coso + 1
tuii)z(cosno)=ftffi
(z - coqho\
(uiiil Z (sinh no)= - z sinh o (id Z(cosh no)- -z
."2-z"cosh0+1 zo -22 cosh 0 + 1

XI. STATISTICS & PROBABILITY

E f;x;
l. A.M.i ==?
Lli
USEFUL
INFORMATION (x/x)
Eft(xi-Ot
3. Moments about
h n t r l ]erae
A o ,tue.L*t.,
the wnnn
mean.:: ) , , --= 1
pt --= /0,
lro rr1, pl
.- lLO r .. =
[t2 = gg.2,
ur F2
- -2 .- =
[r" = 4-
, lar
4,-
4. coeff. of skewness= (mean-mode)/owhich lies between- 1 and 1.
5. Kurtosis : |9z= i
' ' 4 It+/tt|
r-*' r-a
,'
, ,
n n ^ f a - -
6.uoeff. ofcoirelationr= 1<r<
\Lt " u; - (H*)"I l"ui - @dl"| )': - 1

of y on x -:y -, =, ?(r
7. L,ineof regression - D

of x on y : x -l=
Line of regression 0 -t)
?
8.ProbabtitJ.p$]t ,p*e=r.
U -_J

\.. (i) p (A.orB) - p (A) + p(B\,,

Gi\p(A & B) = p(A).p(B\
,\
-9. Binomial distribution : p (r) =nC, prqn-,
Mean = hi, iariai:ee (o2)= npq
'
10. Poisson distribution z p(r) = ,- ^
#
Mean - m, Var,iance(o2)= *.
- @-tt)z
11. Normal distributioir f\x) = 1 e-f , Stand,ard,uariatg _ x - p
o.b" o
(i) Probableerror l, = 0.6?45o. \
(ii) 6go7o
of values lie betweenr = p - o & r = p + o.
95Voof valuesliebetween x=p_ 1.96o 6i*= p+ 1.96o
99voof valuesliebetyreen x=p-2.59 o &x= p + 2.5go

:ilI. NUMERICAL METHODS

l. Solution of equations
(il Bisectionmethod.: xr= _ t* xn _z)
l{*,
Ui)Method,
of Falseposition
i x2=,r,s
nffif@d.
(iii) Newton-Raphson method, : t t n + r - x n f@")
T6
$u) Iteratiueformula to find I/N is &n+ r= x, (Z _ Nrn)

fu) Iteratiue formula to find ,lN is .rn* , = (xn + N/xn)

;
2. Finite differencas Ay, =lr +t-lr;YIr =lrr
lr -ti for,_ L/z=ln - ln _ |
-
( i )F a c t o r i allx l r x ( x * 1 )( r - 2 ) . . . . ( r - r + 1 ) ' | .
G i \L - E - 1 , V = ! - E L , 6 = E L / 2- E - L / . 2E, = e h D
(xx) MATHEMATICS
HIGHMENGINEERING

3. Newton's forward interpolation formula :

!p=!o+p a y,*P'@i! L2 .1-.e*f.? +....

A3yo
Newton's b ackw ard, interpolation formula

lp =!n+ p v ,^ * &# v2yn+"ft*qtt? v3y,,* ....

t(
4.f'(a)=; ^f@)-*o' f(a)+* otf,o,-i */(') * ...I
[
't'( )
f " (a)= - tsf(o).+ * f@)f
n'. f@)
tnI ;
6 . Tr apez oid. alr u l e, I:." o ,r* , d * = * [Oo + y ,) + 2 (yt+ l z+ ' ;, + y, - 1;]

Simpson's rule
gz s + nh h
J;, f @ )d * = * [ u o + y ) + 4 ( y r + l s + . . .* ! n - r ) r -2 ( y z u a + . . . . . i y n - i l ]
 Unit I - Algebra & Geometry

Solution of Equations
1. Introduction. 2. General Properties. 3. Relations between roots and co-efficients. 4.
Tbansformations of equations. 5. Reciprocal equations. 6. Solution of cubic equations-Cardan's
method. 7. Solution of biquadratic equations-Ferrari's nrethod ; Descarte's method. 8. Graphical
solution of equations. 9. Objective Type of Questions.

I.T. INTRODUCTION
-t
The expression f (x) = a&n + a!fin * .,. * &n- 1 x * an
where o's are constants (ao * 0) and n is a positive integer, is called a polynarnial in r of degree
I =
n. The poly4omial f (r) 0 is called an algebraic equation of degree n, If f (r) contains some other
functions such as trigonometric, logarithmic, exponential etc. ; then/(r) = 0 is called atranscendenta.l
equatinn.
The value of r which satisfies f (x) = 0, ...(1)
is called its root. Geometrically, a root of (1) is that value of r where the graph of y -f 1r) crosses
the t-axis. The process of finding the roots of an equation is known as solution of that equation.
This is a problem of basic importance in applied mathematics. We often come across problems
in deflection of beams, electrical circuits and mechanical vibrationd which depend upon the
solution of equations. As such, a brief account of solution of equations is given in this chapter.

I.2. GENERAT
PROPERIIES
l.Ifc^isarootoftheetquationf(x)-0,thenthepolynomialf(x)isexactiyd,iuisiblebyx-
and conuersely.
For instance,3 is a root of the equationf - O* - 8* - 3 = 0, because I = 3 satisfiesthis equation.
.'. r-3 divides*a -6*2 - 8r- 3 completely,i.e.*-3 is its factor.
IL Euery equation of the nth d,egreehas n roots(real or imaginary).
Converselyif 41, %,......, c6 be the roots of the nth degreeequation f (x) = 0, then \
f (x) = A(r - af(r - oe) ... (r - cr) where A is a constant.
Obs. If a polynomialof degreen vanishesfor morethan n valuesof r, it rnustbe identicallyzero.
Example 1.1. Soluethe equation 2x3* f - 13x+ 6 = 0.
So.t.Bf inspection, we find r = 2 satisfies the glven equation
t . ' . 2 i s i t s r o o t ,i . e . x - 2 i s a f a c t o r o f Z r s + * 2 - 1 3 r + 6 . D i v i d i n g t h i s p o l y n o m i a l b yx - 2 , w e
get the quotient ?.rz+5* - 3 and remainder 0.
tl
I Equating the quotient to zero,we get 2tz +5r - 3 = 0.

[, - 5 t{tsz-,1=(z)= (= 3I 5 tJ - --o e
'E1
'
solving this quadratic,we get" = --
"
ZxZ 4
 HIGHER
ENGINEERING
MATHEMATICS

Hence the roots of the given equation are 2, -3, L12.

Note' The labour of dividing the polynomial by x - 2 can be saued. consid,erably by the
following simple deuice called synthetic division :
-13 6
10 -6

lExplanation; (i) Write down the co-efficientsof the powers ofr in order (supplying the rrrissingpowiis
of r by zero co-efficients)and write 2 on extreme right.
(ii) Put 2 as the first term of 3rd row and multiply it by 2, write 4 under L and add,.giving b.
(iii) Multiply 5 by 2, write 10 under - 13 and add, giving - B. r
(iu) Muitiply. S by 2,urite - 6 under 6 and add giving zersl. : I
Thus the quotient is ?.tc2+5* - 3 and remainder'is zero.
obs. To divide a polynomial by x + h, we write - h on the extreme right.
III. Intermediate value property.If f (a) and f (b) haue dffirent signs, then the equation
f 1x1= 0 has atleast One root between :t = d and, x -=b.
The polynomial f (r) is a continuous function of r.
So while * changes from a to b,f (r) must pass through all the values from f (a) to /(b). But
since one of these quantities f (a) or f (b) is poiilive and th-eother negative, it foliows that atleast
foi one value'of'* (say a) lying betw eel a anrl b, f 1ry mus t,be zero. th"r, s is the ;il:
NI. In an equation with real coefficients, imaginary roots occur in conjugate ""q"iruJ
pairs, i.e. if
cr+ipisarootoftheequationf(x)-0,thend-'BmusLalsobeitsroot.
Similarly if a + {b is an irrational root of an equatio4, then - ^tit must also be its root.
obs. Euery equation of the odd degreehasatleast one'real root. "
This follows from the fact that imaginary roots occur in conjugatepairs.
Example L'2. Sol,uethe equation 3x3* 4f +* + 88 = 0, one root being Z +lfi
sol. since one root is 2 +',1 7i, the other root mustbe 2 - ^lzt.
.'. The factors corresponding to these roots are
(x-2 .,12il and(r -Z+lZi1
or (x - 2 - r h D 1 u- 2 +,,lZt1=
@ - 2 ) z+ 7 = * 2- 4 * + L L
is a divisor of 3.r3- 4x2+r + 88 ...$)
.'. Division of (i) by *2 - 4*+ 11 gives 3s +8 as the quotient.
Thus the depressed_ equation is 3r * 8 = 0. Its root is - 8/3. Hence the roots of the given
equation are 2 Xi,'li, - 8/3.
V. Descartets rule of signs.* The equation f (x) = 0 cannot huue more positiuerootsthan
the changesof signs in f (x) ; and more negatiue ro,otsthan th,echangesof sigis in (- r).
f
For instance,considerthe equationf @) =b)7 -x5 + 4x3- b = 0 ..:(1)
Signs of f (x) are + +

Clearlyf (r) hasr YX X*, (from+ ro- or- to +).

ftT"';H;::?rl'TTsi
i,Tr:.,""'fs
= - z x T + * 5 - 4 x g- E !
\/V
This showsthatf (r) has 2 changesof signs.Thus (i) cannothave more than Z negativeroots.
* After the French rhathematician and philosophBrReru Descartes(1b96-1680),
who invented Analytic
geometryin 1637.
SOLUTION
OF EQUATIONS

Obs. Existence of imaginary roots, If an equatinn of the nth dngreehas at the mostp positiue roots
and at thi rnost q negatiueroots, then it follows thot the equation has at least n - (p + q) imaginary roots.
Evidently (1) aboveis'an equation of the ?th degSeeand has at the most 3 positive roots and 2 negative
roots. Thus (1) has atleast 2 imaginary rcots.

I.3. RELATIONS
BETWEEN
ROOTS
AND COEFFICIENTS
If c*1,d2, a,B,...... , d.n be the roots of the equation
-
qgn * a1frn | + a2xn -2 + ...... * &n - fi * an = Q ...1
()
1
a1 ag ag
).(I1= - Zf,Xt0rZ=;O, =-
2,01C[2C[3
,rO, OO
\'""
\
...... Cf,;'= F D"
C[1C[2C[g
X.
Proof. Since o,1.o"2to3,...... , cln are the roots of (1), we have
av{+a'l{-'
::# -:;;:Xi"-::,i i: ."- a,)
- og k" - (crr + crz+ .,.... * a^) *- I + (414,2+ flzcra+ ...) xo - 2
- (ararae + c,2ougo,a+ ......) xn-3 +...... + (- 1)" drdzas .... aoJ
Dividing both sides by oo and nniting,
A1a. d2f €t3* ...... + 0,r=2A"1
crrcb + creaa+ ...... = Ecrrcz
ut%crg + c[zc[gc[4] ...... = Ecrugds and so on, we get
-r),
nn*|.A]"" - t * [9'lr. -z + ......*["" * ""
l."oj (oJ (. oo oo
)
- * -(Ear) gn-t + €arnd g-z
Equating coelficients of the like powers of r from both sides,we.get the desired results.
Example 1.8. Solue tlrc'equation *3 - 7f + 36 = 0, giuen that one root is d,ouble of another.
SoL Let the rootsbe u, B,y such;thatF = 2u.
Also a+B*T=7,aF+Fy+T€r=0,gFT=-36 :
3a+Y=7 ...(r)
2* +3cr1=g ...(t,)
2&T= - 36 ...(ttt)
Sblving(z)and (ii), we get c = 3, T- - l.
[The values cr= 0,'l = 7 are inadmissible, as they do not satisfy (jjj)I.
Hence the roots are 3, 6, and - 2.
Exampl e 1.4. Soluethe equation x4 - ztcs+ ai + 6x - 21 = 0, giuen that the sum of two of its
roots is zero. (Cochin, 2005; Madras, 2003)
So1. Let the roots be u, F, T, 6 such that cr + F = 0.
Also cr+ Ft y+ 6.;,2 i,,,.', T+ 6 =2, . , :
Thus the quadratic factor correspondingto cr,B is of the form *2 -O*+p, anfl that
correspondingto 1, 6 is of the form af x2 -Lr + q.
,. *4 -.hg + 4x2+6r- ZL=(xz+il@z -Lr+q) ...(r)
 HIGHERENGINEER]NGMATHAMATICS

.-Equating coefficients of *2 and* from both sides of (i), we get

' 4=p*e, 6=-2p.
- -3,
P Q=7.
Hence the given equation is equivalent to (*2 - 3)(xz- 2.x+ ?) = 0
.'. The roots arer =+ {8, 1 t t{0.
Exampie L'5. Find. the cond,itionthalthe cubic *3 - I* + rlx - n - 0 should.haue its rootsin
(a) arithmeti,cal progres sion. (Madras,2000S)
@ geometrical progres sion.
Sol. (a) Let the roots be o - d, d, a + dso that the sum of the roots =3e =t i.e. a = I/3.
Since a is the root of the given equation
at-1,o2+rnq,-n=0 ...(r)
S u b s t i t u t i n g c= l / 3 , w e g e t ( I / q s - I ( I / q z + m Q / B ) - n = 0 .
or 2t3- 9lm + 27n =O which is the required condition.
(b) Let the roots be a/r, e,,ar, then the product of the roots = o3 = n.
Puttingo= (n)l'!, in (i), we get n-ln2/3 +*nl/g -n=O or rn-,lntrg
Cubing both sides,we get *3 =13n, which is the required condition
Example 1'6. Soluethe equation14 - 2x3- LIf + 22x + 40 = 0 whose rootsare in A.P.
Sol. Let the rootsbea-3d,a- d, a+d,a+3d, so that the sum of the roots- h-2.
a=L/2
Also product of the roots = (a2- gd,2)(o2- d2)= 40
or B - nat) (* - o'\= no - Aodz-6Be= 0
or L44d,a
t4
i' -gt+ or - 7/86
Thus d=!3/2,t}rre other value is not admissible.
Hencethe required roots are - 4, - L,Z, E.
Example L'7. Saluethe equation2x4 - 15x3+ 35f - 30x+ 8 = 0, whose rootsare in G.p.
Sol. Let the roots be a/r3, a/r, ar,cr"3,so that productof the roots = a4 = 4.
Also the productsof a/r3, er! and,a/r,@r are each=d,2=2.
.'. The factors iorresponding to a,/r3, ar3 and a/r, ar ^r" *2 + px + Z, x2 + qx + 2.
Thus b4 -1br3 + BExz- B0r * g = 2 (*2 + ptc+ z) (*2 + qx + 2)
Equating coefficients of rB and 12, we get
- 1 5= 2 p . + 2 q a n d B b= g + Z p e
P--9/2,8=-3'
Thus the given equationi" z(xz -* el1r2 - Br + 2) = o
\o)
"*
Hencethe roots areI/2,4 and !,2 i:.e.,|, t, Z, +.

Example L.8.If a, 9, y be'the rootsof the equation*3 + p* I q = 0,

fi,nd the ualue of (a) zozp, (b) >n4 (r) >a2F.
Sol. We have 0+B*T=0 ...(r
)
sF+Fy*!d.=p ...(ii)
aFr= -g ...Gii)
 EQUATIONS
SOLUTION,OF

(o) MultiPlYing (i) and (ii), we get

g
crzp+ cr2T+ FzY* F2o* fo * fF * 3up1=
hv (t,tl
or EcrzP= - SaPY=39
* =0
(b) Multiplying the given equation by x,we get *4 + p*2 Qr
get laa + pSn? + qla = 0
Putting N = a,B, y successivelyand adding' we ...(iu)
or >n! =-Pl'o,z -q(0)
=0
Now squaring (i), we geta2+ 02+ f + 2@9+ 9T+ Tu)
lbv (it)I
fnz =-2p
(iu)' we obtain
Hence, substituting bhevalue of Ecrzin
l,u4=-p (-2p) -2p2.
- =- - (- q) (0)= - 2p2'
Ip trl >oap=tu'Z I,oF crFyIa 2p (p)

(Madras,2000 S\
roots are 3 + i and rl7'
1. Form the equation of the fourth degree whose
2. Solvethe equation (i) ra + 6r+20 = 0, one root being 1+ 3i'
(Kattavam'1996)
(;;;"t -*t -iu'+G2,-- tb=0, giventhat2 +{3 is aroot'
(Cochin,2005)
g. Show that r? - gxf +Zxs - I = 0 has atleast four imaginary roots:
and two imaginary roots'
;: ;;;; i*,n" *a + !&xz + 7x -11 = 0 h1s on; nositivl,^onenegative
"qoatioo + 48:3- 4r - 13 = 0'
E. Find the numler and position of real roots of xa
6 .S olv et heequati o n 3 r3 _ L L x 2 + 8 r+ 4 = 0 ,g i v e nthattw oofi tsrootsareequal .
pairs of e.qualroots. Find the -valugsof o and b'
?. The equation *a - 4*B+ a* + 4* +[ = Qhas two
SolveQheequations 8-14 :
\
g. rB - g,,.z-16r + 48 = 0, the sum of two of its roots being equal to zero'
(Madras' 20A3)
g. ra -6rs + lgxz * t2x+4 = 0, given that it has two pairs of equal roots'
in the ratio 3 : 2'
10. rB -gxz +L4x +24=0, given that two of its roots are
connectedbytherelation".lJ;.}.
11. rs -4ez -zox+4g=0grventhatthe rootsaandpare u.,2007)
roots is equal to the sum of the other two'
!2. *4- grg + zl*2 - zhx+ b = 0 given that the sum of two of the
(Madras'20015)
18. rB -t2*+3gr -28=0,rootsbeinginaxithmeticalprogression'
progression' (osmania" 1999\
i;: ;t - t6xz + 7x -1 = 0, roots being in geometrical
that the distancesof A, B, c from o arethe roots
!5. o,A,';', c are the four points on a straight line such azd.-t:!t +2b3 =0.
of equatio^*{*iu*i il +d,=0. If t iB the middlepoint of Ac,show that
(s.v.r.u.,2006)

theequati"*J;l
16'sorve
5::r -?fr:'H:;: : : ;l::: ;: ; #l
,. 1
- - n= 0, find the value of
1?. If a, p, 1be the roots of the equation *s t*' + rttn
(i) Ea2P2, (tt) (P+?(T+ crXa+ F) ":::
f - af + lk - 6 = 0'
1g. Find the sum of the cubes of the roots of the equation .*{:
t!:lg:

- g= 0, find the value of (i) o1.l+ F-1+ fl

it. it o, F, T are the roots of rB + 4x, \.
{j:'-'.

( i i ) a - z+ 9 - 2 * { 2 . \
g0. If s, F, Tbe the roots of *3 +pr + q = 0, shourthat
(t) ai + F5+ t' = EcrFy$v+ tot + ap), (tt) 3Ect2xcr6= StnB Icr4'
 HIGHERENGINEERINGMATHEMATICS

I.4. TRANSFORMATION
OF EOUATIONS
(1) To find an equation whose roots are m times the roots of the given equation,
multiply the second term by m, third. termby *2 and, so on (allmissing terms supplied with zero
coefficients)"
For instance,let the given equation be
3x4+6x3 + 4xz -8r + 11 = Q ...(t)
To multiply its roots by m, put y = trl* (or r = y/m) in (l).
Then 3(y/m)a +6(y/m)s + A(y/m)z+8(1r/m)+ 11"= 0
or multiplying by *4,we get Bya* m(6y3)+ m2(4y\ - -3(gy) + nza(ll) = 0 ,
This is same as multiplying the secondterm by m, third term by *2 and so on in (l).
Cor. To find an eq..atiorr whose roots are with opposite signs to those of the given equation,
changethe signsof thc eueryalternatiue terrn of the giuen eguotion biginning with the secind,.
Changing the signs of the roots of (d)is sarneas multiplying its roots by - 1.
Therequired
equati""-lt"ii
(* 1); .,- ,rJer, -l u3 8" + ( - 1)411= 0
or 8 1 4 = 6 1 3 +4 x 3 + g r + 1 1 = 0
which is (i) with signs of every alternate telm changed beginning with the second.
(2) To
find an equation whose roots are reciprocal of the roots of the given
equation, changex to 1/x.
Example ir'L. Solue6x3- IIf - 3x + 2 = 0, given that its rootsare in harmonic progression.
(Triputi, 1998)
$ol. Since the roots of the given equation are in H.P., the roots of the equation having
reciprocalroots will be in A.P.
The equation with reciprocalroots is 6(L/x)s - tt(ttx)z - BlLt*) + z = 0
?l3-BxZ-11*+6=o ...(r)
Since the roots of the given equation are in H.P., therefore, the roots of (i) are in A.P. Let
the roots be s - d, e, a + d. Then
Ba = B/2 and,a(az _ a\ = _ g.
Solving these equations,we geta - l,/2, d.= E/2.
Thus the roots of (1) are - 2, llL, g.
Hencethe roots of the given equation are - LlZ, Z, !lg.
Example 1'L0.If a, $,1be the roofsof thecubic*3 - p*2 + qx - r = o,form the equation whose
rootsare By+ 1/a, ya+ 1/$, u$+ I/y.
Henceevaluate> (aF +t/y) (pT+ L/a).
Sol. If q11sa root of the given equation andy a root of the required equation, then
'
y =Fr+L/u=ry=$
(Ia t.., upr=
rl

r
"
+
Thus substitutirrg r = (r + t)/y in the given equation, we get

i.gf _oft*tf
't v --('*_I)_
r
-_
[-r ) )*ot J-r=o
Iq

f j , . t r - " ' ' , . . J

 OF EQUATIONS
SOLUTION

y - (r+ 1)3= 0, which is the required,equation.

or rys _q v +L)y2 +p (r +L)2
HenceE (crpi trrl$T+ L/a) -:pi (r + !12/r: o ^
equation whose rqotsare cubesof of x3 - 3f + 7 - 0' "'(t)
t;he root"s
Exampre r.1r. Form an
= x3 i "'(tt)
sol. If y be a root of the required equation,then !
l Now we have to eliminate r from (i) and (ii)
.': Rewriting (i) t3 + ! = 3x2
", =
Cubing both sides,19 + 3rG+ 3r3 + 1 2tx6
-24y2 + 3y + 1= 0
Substitutingrs =/, w€ getys
which is the required equation
(x) bv x - h successiuely.
(B) To diminish the roots of an equation f(r\ -0 b{ h, d'iuid''ef
equation'
Then t},e successiue remaind,ersd.eterminithe toiffici'i"ntsof the required
t Let the given equation be
'
a f f i n * ( r 1 x n - 1 + . . . . . '* a n - 1 f i * Q ' n = a
- x -h (or x =! + h) in (i) so that
To diminish its roots by h,puty
I =
ao U + h)" + at (Y + h)"- + """ * o,, Q
On simplification, it takes the form
'r ' | ir : 'Aoyo*A.1'n-l*...,.,*Arr=Q :
.,
the help of syntheticdiuision (p' 2)'
n carTeasilybe found with
For this, we put y - i'- i itn(iii) so that ' '

of (iu) byr - h, the remainder isA, and

clearly (i) and (iu) are identicar. If we divide L.H.s.
rheduotienre=;;@-hio-t*iit* -D"-:+......+An-r limjlarliifweailae^Qbv l,:!:
bvx-h'An.zwill be
irr" ttt" q";ii"tttis 4r ('"v1'Againdivfding
Qr
""i"irra"jtt;;:1and
obtainetl;l*::1"
*- rootsbv h,we takeh negative "
,t'l

- g = 0 inro anot'herin which the second

Example L.12.Tlansform the equatiorylt 6f + 5x +
tertnisrnissing.Hencefind,theequationofits,qio,ud'.d'ffirences..
Sol. Sum of the roots of the grven equation = 6'
is missing, the sum of the roots
In.order that the secondterm in the transformed eguation
is to be zero.
by 2, the sum of the roots of the new
since the equation has 3 roots, if we decreaseeach root
equation will becomezero.
Dividing *3 -6*2 + 5* + 8 by r - 2 successively,we have
-6 58',Q
-8 '6
2
i8 HIGHER
ENGINEERING
MATHEMATICS

Thus the transformedequationis rB - 7x +2 = 0. ...(t)

If cr,p, 1be the roots of the gtven equation,then the roots of (i) are o - 2, P - 2, y - 2.
Let thesebe denotedby e, b, c.
Then b - c= B - 1. Also a + b + c =0, abc = - l.
Now (b - (b + q2 - 2bc= (a + b+c - q)2-ry =a2 + 4/a
")2= a
.'. The equation of squared differences of (i) is given by the transformation ! = 12 + 4/x
or x3-*y+4=0 ...(rt)
Subtracting (ri) from (l), we get
-7x+xy -2=O or x -2/(y -7)
Substituting for r in (r), the equation becomes
t2t1y- Dl3 - 7 P/b'- 7)l + 2= o
or ys - zSyz+245y - 682 = o ...(iii)
Rootsof this equationarc (b - c\2,(c - o)2,(n - b)zi.e.(F -,()2,(T- c)2,(c - g)2
Hence (iii) is the required equation.

I .5.RECIPROCAI
EQUATIONS
If an equation remains unaltered on changing x to L/x, it is called,a reciprocal equation.
Such equations are of the following stand,ard, types :
I A reciprocal equation of an odd degree having coefficients of terms equid.istant
from the
beginning and end equal.It has a root - -1.
[. A reciprocal equation of an odd degree hauing coefficient of terms equid,istant
from the
beginning and end equal but opposite in sugn.It has ioot = 1.
nI. A reciproca,l equation of an euen degree hauing coefficients of ternts equid.istant
- ftom the
beginning and end equal but opposite in sign. Such an equation has hro roots = 1 and -1.
The substitution x + L/r -y reduces the degree of the equation to half its former degree. ,
Example 1.13. Solue6x5 41xa+ g4xa-gTxz + 4lx - 6 = 0. (Coimbatore,2001 S)
Sol. This is a reciprocal equation of odd degreewith oppositesigns. ...r = L i s a r o o t .
binidi.tg L.H.S. by r - 1, the given equation reducesto
6xa-Bb*3 + 6Lx2- Bbr + 6 = 0
Dividing by ,2, we have
6 (*2 + t/x21- 35 (r + L/x) + 62 = 0
Putting x + L/x =y and x2 + !/x2 =y2 -2, we get
6 g z - 2 ; t - B b y + 6 2 = O o r f u z - B l y+ b 0 = 0
or (3y-1)(2y-5)=0
/.
x + L/r =! = L/3 or E/z
i.e., 3rz-10r+B=0 or fuz -5x+2=0
i.e., (3{-1)(r-B)=0 or (%c-1)(r-Z)=0
x = L/3,3 or L/2,2
Hence the rootsare 1, L13,3,L/2,2.
Example 1.14.SolveOrc6 -25x5 +51# -Slf +25x-6=0. (Mad,ras,Z00g)
Sol. This is a reciprocal equation of even degreewith oppositesrgns. ;, ,t = 1, -1 are its rogts.
 OF EOUATIONS
SOLUTION

Dividing L.H.S. by r - 1 and r + 1, the given equation reducesto

614- 25x3+ 37xz- 25x* 6 = o
Dividing by *2, we get
6 (*2 + L/x2) - 25 (x + l/r) + 3? = 0.
Putting x + l/x=y and x2 + L/x2 =y2 - 2, it becomes
6 (yz- 2) * 25y+37 =0 or 6yz- 25y+25 = 0
or (2y-5)(3y-5)=0
x+L/x=!=5/2 or 5/3.
i.e., b2-5x+2=0 o r } x z- 5 x + 3 = 0
x=z,!/z or "=E#1
Hence the roots of the given equation are 1, z,L/'z,*fE

1 . Find the equation whose roots are 3 times the roots of 13 + 2x2 - 4r + 1 = 0.
2. Find the equation whose roots are the reciprocals of the roots of x4 - 7xs + 8* + 9x - 6,= 0.
8. Find the equation whose roots are the negative reciprocals of the roots of
:,,.,r x 4 + 7 x 3+ S * - 9 * + 1 " 0= 0 .
4. Solve the equation 613 - Ll-rz - 3r + 2 = 0, given that its roots are in H.P.
o . Find the equation whose roots are the roots of
(j) rB - 6xz + Lk - 6 = 0 each increased by 1.
(ii) xa + - 3x2- x + 2 =0 each diminished by 3.
"3
(iii\ x5 -1xa + 10*3 -LOxz + 5*+ 6 = 0 each diminished by 1.
,6. Find the equation whose roots are the squares of the roots of *3 - x2 + 8x- 6 = 0.
7. Find the equation whose roots are the cubes of the roots of xg + pxz * q = 0..
8. If d; p,y are the roots of the equation ?.x3+58-x-L=0, form the equation whose roots are
(t':- cr)- 1, (1 - F)- 1 and (1 - y;- 1.
9. Ifo, b;caretherootsoftheequation *3+p*2+qx+r=0,findtheequationwhose roots are ab,bc
(Madras' 2003)
,r. f:,"f 1be the roots of r3 + ntx + n= 0, form the equation whose roots are
(o)a + F -f, F +y- cr,T* cr- p,
11
(b)pT/q,Ia/F, orF/T t, .c1) 6 + t ' t *1;1' a+ 1^p* '
11. Find the equation of squared differences of the roots of the cubic xs +6* +7x+2=O.
12. Solve the equations :
(i) 6ra+ 5*3- 381 + 5r + 6 = 0. (Madras,2000 S)

$, (ii) 6x5+ x4 -43n3- 43xz+r + 6 = 0.

(iii) 8x6-zz# - b5r3+ 55xz+ZZx- 8 = 0.
(s.v.T.u.,2006)
(Madras,2000l
(iu) 4xa- 20xa+ 33xz- 20x+ 4 = 0. (Madras,2003)
6 * 5- T ? x a + 2 7 1-2r - B = 0 -
( u ) 3 1+
,18. Show that the equation xa - l0x3 + 23* - 6x - 15 = 0 can be transformed into reciprocal equation by
diminishing the roots by 2. Hence solve the equation.
14. By suitable transformation, reduce the equation 14+ 1613+ 83x2+ L52x+ 84 = 0 to an equation in
'wh*ch terrnin *g is absent and hence solve it. (Madras, 20021
,,,
 l0 HIGHER
ENGINEERING
MATHEMATICS

I'6. SOTUTION - CARDAN'S

OF CUBICESUAT]ONS METHOD-
Consider the equation o * 3+ b x I + c x l d , = 0 ...(1)
Dividing by o, we get an equation of the form xB + lx2 + rnx * z = 0.
Toremouethe*2tur*,puty=cc-(-l/3')orx=!-l/Ssothattheresultingequationisofthe
form

y3+w*q=o ...Q)
To solve (2), put t=u+u
sothat y 3 = u 3 +i 3 + } u u ( u + u ) = u } + u g + B u u y
or ys - \uuy - (u3+ u3;= g, ...(3)
Comparing(2) and (B),we get
- -ps/27
u3u3
u3,,3 ii:,:::"1;;":;; ;:;'::,._:r),;:and
""e
which gives ut =*,-n *'^6ffi1=13 (say)
and ut=t.q -"{F+ >
The three values of u arel,, l.crl,lnrz, where o is one of the
imaginaly cube roots of unity.
. From uu = _ p/8, we have u = _ p/Bu
.'. when u =?r,lnr and lnr2,

t)=-_h,-$""u-# [
'.' tos= 1]

Hencethethreerootsof (Z)arel. - - ^r, -

fr*, Ior #, H ( b e i n g- u * u )
Having known y, the corresponding values of r can be found
from the relation r - y - t/8.
obs' 1' If one value of u is found to be a rational number, find
the corresponding value of u giving one
rootv =tt+u' Then find the.ott".po"dtd;;;=ti+il't*ur,
-
n d''grving the remaining quadratic equat-ionfrom whichile iirria"'*rr" teft nana side of (1) by
ottr"r"two roots can be found readily.
obs' 2' If u3 and u3 torn out to be conjugate complex numbers,
obtained in neat forms by employing De MoiwE's theorem. ($ the roots of the *"o cubic can be
lg.b)
Example 1.15.solue by cardan's method,*3 - s f +
LZ x + 16 =0. (U.P.T.U, 2001)
S o l . G i v e n e q u a t i o n i s r-sB x 2 + L 2 x + 1 6 = 0
...(r)
To rerrlouethe secondterm frorn (i) diminish each root of (i)
by B/B = L, i.e put / = r - L or
x - y + | ['.' Sum of roots = 3]. Then (i) becomes
0 + 1 ) 3 - g 0 + 1 ) + t Z ( y + 1 ) + 1 6 = 0 o r y s+ 9 y z + 2 6 = 0 ...(ii)
T o s o l v e ( i l )p,u t! = u . + u s o t h a t y s - B u u y - ( u s + r t ) = O -
...Gii)
Comparing (tt) and (jjz),we get uv - B and u3 + r3 = - 26
=
:. u3, r3rru the roots of the equation F + 26 t - 27 = 0
or ( t _ + 2 7()f - 1 ) = 0 w h e n c e t = - 2 7 , t = L .
or u3 =- 27 i.e,,u,= -B and u3= 1 i.e.,v = l
t = u + u = - 3 + 1 = - 2 a n dx = ! + l , = - 1
t Named after an Italian
mathematician Girolamo Cordan (rb01-1826) who was the first to use
complex numbers as roots of an equation.
 ll
OF EQUATIONS
SOLUTION

- =0
Dividing L.H.S of (r) by r + 1, we,obtatn'x2 4 i+ 16
or =4.P =z!t2{3
- 2 {3
Hence the roots of the given equation are L' 2 tt
-9 Card'an'srnethod"
Example 1.16.Solvethecubic2Sxs f + 7=0by
putr =1/y in the gtven equationrsothat the
Sol. since the term in r is missing, let us :: (t)
transform"d - ev +28 = o
"q;;;t;;;"t -3 uu y -(u3 +'3) =0 "'Gi)
, u t ! = t t ' + u s ot h a t y g
T o s o l u e ( l )p
= - 28.
comparing (ii) and (iii), we get ut) = 3 and u3 + r3
i . u 3 , r 3 u " " t h e r o o t so f t 2 + 2 8 t + 2 7 = 0
or (t+1)(t+27)=0or t--_!,.2.7 o'1_-t,u=_3
...y-u+u--4.OirriaingL.H.S.of(i)byy+4,weobtainyz-4y+'I=}whence!=2tt{3'
.'. roots of (i) are - 4, 2 t it/3.
- -1, Q-i.,B)tl,(2+il3Dn '
Hencethe rootsof the givencubicare I,#ftor.
- 76x+ 20 = 0
Example l'17. Soluethe equationxS+ f
by - 113, we frrst multiply its
Sol. Instead of diminishlng tt roots of the given equatio:r
"
roots by 3, so that the equation becomes
...(i)
x3 + 3x2- L44x+ 540 = 0
- (i)
T o r e m o u e t h e# t u ' * ' p u t y = x ' - ( - 3 / 3 ) o r f r = ! 1 i n
sothat A*1)3+30 D2*L44(v-1)+540=0
...(tt)
or Yg-L4lY+686=0 ., .,,:, r,,
T o s o l v e ( i i )l ,e t : u = r ' t + u ' s o t h a t r'' : r' '
...(iii!l
Y 3 * 3 * ' Y - ( u 3 + u 3 1= g
Comparing (ii) and (iii), we get
, uv = 49, u3 + oB= -686.'so that z9u3* (343)2'
J.'u3,u3 the roots of the quadratic
"re
t2 + 686t +(343)2= 0 or (t+ 343)2=Q
t='343 i.e,rug=u3=-343,ot rL=n=''7:'
:- 15.
Thus ! = u + u = - 1 4 a n dx = ! - t
Dividing L.H.S. of (i) by r + 15, we get
(x '6)2=0 or r=6'6'
.'. The roots of (i) are - 15, 6, 6'
Hencethe,ro,otsof the glvenequationar€ - 5'?.'2'
(s.v.r.u,, 2006)
Example 1'18. Solue*3 - gf + 3 = 0' :
...(t)
Sol. Given equation is *8 - 3r2 + 3 = 0 t
=y * 11
To remouethe *2 te'*' put y =N'- 3/3 o:5
d O
=
so that (i) becomes (v + 1)" - 3(y + L)" + 3 0
...(ii)
or Y3-3Y+1=0
Tosolueit,Put!=u+u
= ...(ttt)
so that Ys - SuuY- (us u9; 0
 t2 HIGHER
ENGINEERING
MATHEMATICS

Comparing (ii) and (iii), we get ut) -- L, u3 + r3 = - |

i. u3, 13 .r" the roots of the equation t2 + t + 1 = 0
-1+i{3
_ 3=--Z , r -1-t{B
Hence U- and U"=
z
(- t +;^/sf/3 . 1 -;=r
,
put cos0 and lSlZ = r sin 0
"=[ ,,|
= [r (cose + isin 0)]1/3 sothatr=1,0=2n/3
- [cos(0 + 2nn) + i sin (e + 2nn'1ll/3,where
n is any integer or zero. Using De Moivre's
theorem(p ...).
0 +Znn 0 +Znn
u=cos +rsrn
3 3
Giving
nthe"":::l ': ':"*'l;"""?:T,"'u'ooi
cos :,:T':::T"1";J":':
+, srn cos-F + z sin cos;- + i sin
* nn
F 3-, ]-,, ?
2n 2n 8n ._.,_ 8n L4n
i.e. cosT+r srnT, cost+l sin c^^_L4tE
os . . s_,_-
intil
rf, n-+i
The correspondingvalues of u are
_ 2 n 2 n-,8 n
cos sln cos - r s l n l r c o s T - r s m g.
T- 5, 5
. ' .T h e t h r e e v a l u eos fy = u + u a r e

Hence
theroots
3,ff#ffi:"il:r:'=t f r;ff
L + 2 cos2n/9, t + 2 cos8n,/9,t + Z eosLhn/g.
"1
Solve the following equations by Cardan's method :
1 . . x 3 - 2 7x + b 4 = 0 (U.p.T.(l.,2008) 2, f-18*+Bb=0 (Osm.ania,2003)
3. .r3- LEx = L2G (S.y.". U., 2007) 4 . 2 x 3+ S f + x - 2 = 0 (U.P.T.U,2003)
5.913+Gx2-L=0 6. r3 -Gx2+6r-b=o
7.x3-3r+1=0 8.27*3+54I+198r-?3=0

I.7. SOLUTIONOF BISUADRATICEOUATIONS

(1) Ferrari's method
Tlnis method of solving a biquadratic equation is illustrated by the following examples :
Example L'19. Solvethe equationx4 - 12f + 4If - LBx- zz - 0 by Feryari's method,.
6.V.T.(1., 2007)
Sol. Combining 14 and t""ms into a perfect square, the given equation can be written as
"3
(*2 -6x+7")z + (5 - 2?r)xz + (IZL- 1g)x- (?rz+72)=g
or ( * 2 - G r * I ) 2 = t ( z f - 5 ) * 2 + ( 1 g - L Z l v ) r + 0 2+ 7 2 ) l ...(r)
This equation can be factorisedif R.H.s. is a perfect square
i.e. if (18 - LZL)z= 4(21,- 5XA,2
+72) [b2 - +ac
i.e. if 27'3- 4L?e+ 25ZL- 441 = 0 which gives ),.= B.
 t3
OF EQUATIONS
SOLUTION

-
.'. (t) reducesto (*2 -6r + 3)2- 1x 9)z
- =
i.e. @2- 5x - 6)(*2 7x + 12) 0'
- 6'
Hence the roots of the given equation are 1, 3, 4, and
-2xs -5xz + 10*-3- 0by Ferrari'smethod'
Example 1.20.Soluetheequation#
the given equation can be written as
Sol. Combining 14 and t*r*, into a perfect square'
"3 is a
@r:;;;;
--(;io>*r- (21.+L0)r +(L2 +B). This equationcan be factorised,if R'H'S
=
tb? 4ac!
perfectsquare i.e.1f(21"+ t0)2 =;(t^ *ul (?v2 + 3)
-
or 2t"3+ 5?r2- 47"' 7 = 0 which gives ?"= 1'
-
.'. (r) reducesto (xz - x - L)2= 4 x2 8 x + 4
or qy-B)(r2-3x+1)=0
@2-,c-L)a-erc-Z)z=Oor
- ltl'fE 3 trr'9:Z
*=- ,.,"
LrI
2

Hence rr" ff,

theroots ryz
(2) Descarte's method
removlls the tu$ in'r3 and then
This method of solving a biquadratic equations consistsin
best illustrated by the following
expressing the new equation as proaort of t*o quadratics. rt is
examples:
- - 24 r + 7 = 0 by Desca'rte'smethod'
Example 1.21. Soluethe equationx4 I *
W.P.T.U.,2001)
so we assume that
Sol. In the given equation, the term in 13 is already absent
...(t)
*a-g *2-z4x+7=(x2+px+q)(xL-px+q)
Equating coefficients of the like powersof r in (i), we get

8 = 4 * 8 ' - P 2 , ' 2 4 = P ( q '- q ) ; ' l = q Q '

q+q'--P2-8' q-q'=24/P
-
.'. @2- 8)2 Qa/fiz = 4x7
-576= 0 wheret =P2
p6 - 16pa +36 pz -ffiA= 0 or fB- 16 t2 + 36 t
Now t = 16 satisfies this cubic so that P = 4'
q + q l '= 8 , 9 - q ' - - 6 ' q=7,q'=t
.'. (t) takes the form (tz + 4x + 7) (*2 - 4r + 1) = $

whence

Hence tc=- 2 t{3i, 2 t{3. \.

equation x4 - 6 x3 - 3x2+ 22r - $ = A by Desarte'smetht\
Exampl e L.ZZ. Salue the \
\
Sol. Here sum of roots = 6 and number of roots = 4'
(= }tz) whic.hwill be a
... To remove the secondterm, we have to diminish the roots by 914
- ys yZ + L76y- 96 = 0 wherey =
problem. Therefore,we first multiply the roots by 2. y: L2 + LZ
- + 297 = 0 where z = ! ' 3.
2r. Now diminishing the roots by i, we obtain "4 42 z2 + 322
- pz +'q') "'(t)
Assuming that ra - 42 22 + 32a+ 297= ("2 + pz + q) @2
l4 MATHEMATICS
ENGINEERING
HIGHER

and comparingcoefficients,we get

- 42 = q + q' - p2 ; 32 = p @'- g) ;297 = q q'
q + q' =pz - 42 ; q - q' =- 32/p,e e' =297
(q' - n4! - 1- 32/p)z- 4x2e7
or t s * 8 4 P + 5 7 6 t - L 0 2 4 = 0 w h e r et = p 2
Now t = 4 satisfies this cubic so thatp - 2.
: . q + q ' = - 3 8 , g - q ' = - 1 6 , : . q - - 2 7 , Q '= - L L .
, Thus (i) takes the form ("2 + 2a - 27) @2- 2,2- ll) = 0
Whence

or

Hence*=1+^h,2t{3.

Solve by Ferrari's method, the equations :

1. ra - 10rs + 35 x2 - 50r + 24 = 0 (U.P.T.U., 2003\
'
2.'xa + 2 x3 - i x2- 8r+ 12 = 0 (1.P.7.11., 2002)
s . 1 4 = 1 0 1 2 -2 0 x - 1 6 = 0 4. x4-8rs -L2' .2+ 60r+ 63= 0 (U .P .T.\J.,2005)
Solvethe following equations by Descartesmethod :
6.ra-613+ Bx2+22x-6=o 6.ra+ 12fr-5=0
7 . x a - 8 r 3 - 2 4 x + 7= o
'
(U.P.T.tl., 2001) 8. xa -10.r3 + 44 rz - L04n+96 = 0
Obs. We have obtalned algebraic solutions of cubic and biquadratic iquations. But the need ofibenarises
to solve higher degree or transcendental equations for whichno algebraichethods are available in general.
Such equations can be best solvedby graphical rnethods (explained belour) or by numericol meth,ods (928.2).

I.8. GRAPHICALSOLUTIONOF EQUATIONS

Let the equation be f 1r1= g.
(i) Find the interval (a, b) in which a root of f(x) = 0 lies.
lAt least oneroot of f (x) = 0 lies in (a, b) itf (a) and /(6) are of oppositesigns-$1.2On) p. 21.
(ii) Write the equation f (r)=0 as 0 (r) =V(r) where ty(x)contains only terms in x and the
constants
Gii) Draw the graphs of y = Q(r) and y = V(r) on the sa;rlescale and with respectto the sa,me
q.xes.
(iu\ Read the abscissaeof the points of intersection of the cu.rvesy = 0 @)andy = V(r). These
are required,real roots of f (x) - 0.
Sometimes it maynot beconvenientto write the givenequationf(*) = 0 in the form g (x)= y(r). In such
cases,we proceedas follows :
(i) Form a table for the value ofr andy =f (x) directly.
(ii) Plot these points and pass a smooth curve through them.
(iid) Read the abscissaeof the points where this cunre cuts the r-axis. These are the required roots cf
/(r) = o. ,
Obs. Tl.reroots, thus located graphically are approximate and to improve their accuracy,the curves are
replotted on the larger scale in the immediate vicinity of each point of intersection. This gtves a better
approximation to the value of desired root. The above graphical operation may be repeated until the root is
obtained correct upto required number of decimal places.But this method of repeatedly drawing graphs is
very tedious. It is, therefore, advisable to improve upon the accuracy of an approximate root by nunrerical
methodsof 928.2.
 l5
OF EOUATIONS
SOLUTION

roatof tlte equatian

Exanfple 1.23"Find,,graphicallyan approximateualue of the
3-x=e*-7.
f(i)=dE-L+x'-3=0
...(r)
Sol. Let
firi=1+1-3= -ve
- =
and f (2)= e +2- 3 2'7LB- ]' + ve
'
.'. A root of (i), lies between# = 1 and Jc=2'
Let us write (i) as = 3 - x'
"'-!
The abscissa of the point of intersection of the curves
t={-L
...0r)
...(iii)
and !=3-x
will give the required root.
To plot (ii), we form the following table of values :
t.4 1 .5 1,6 L.7 1.8 1.9 2.0
s= 1.1 1.2 1.3
1 .3 5 1,.49 1.65 1.82 2.01 2.23 2.46 2.72
n t=, e x - l 1.11 t.22
= 0'02, we plot these points and
Taking the origin at (1, 1) and,1 small unit along either axis
pass a smooth cunre through them as shown in Fig. 1'I.

a 1.4 1.6

Fig.1.1.
t6 HIGHER MATHEMATICS
ENGINEERING

Ta d,raw the line (lll), we join the points (L,Z)and (2, 1) on the same scale and with the same
axes.
From the figure, we get the required. root to be r = 1"'44nearly.
Example L.24. Obtain grapltically an approximate ualue of the root of x = sin x + n/2.
Sol. Let us write the given equation as sin x -- tc- Ir/2
The abscissa of the point of intersection of the curveJ=sinr and the line !:x-n/2will
give a rough estimate of the root.
To draw d, cltrue ! = sin f, we form the following table :

JC 0 n/4 n/2 3n/4 n

v 0 0.71 1 0.71 0
Taking 1 unit along either axis = */4 = 0.8 nearly, w€ plot the curve as shown in Fig. 1"'2.
Also we draw the line ! = x - n/2 to the same scale and with the same axis.
From the graph, we get x - 2.3 radians approximately.

Fig.1'2.
Example I.25. Obtain graphically the lowest root of cosr cosh x = - 7.
S ol. Let f (x )= c o s .r c o s h x * 1 = O ...(r)
f (0)= +ve, f (n/2) = *veandf (n)= -ve.
.'. The lowest root of (i) lies between x = n/2 and r = n.
Let us write (i) as cos r = - sech r.
The abscissa of the point of intersection of the cun/es
J=cosr ...$i)
and y- -sechr ...(iii)
will give the required root.
To draw (li), we form the following table :
x- n/2 =I.57 3n/4 = 2.36 n:3.L4
Y=cos.f 0 - 0.71 -1
Taking the origin at (1.57,0) and 1 unit along either a)os =n/8 = 0.4 nearly, we plot the
cosinecurve as shown in Fig. 1.3.
SOLUTION:OF
EOUATIONS 17

x=7,.5

Fig. 1.3.

To draut (iir), we form the following table :

L.57 2.36 3.L4
coshr = 2.58 5.56 11.L2
y=-secrrltr - 0.39 - 0.18 - 0.09
Then we plot the curve (ssi) to the same scale with the same $(es.
Frorn the figure we get the lowest root to be approximately x = L.57 + 0.29 = 1.86.

Solvethe following equationsgraphically :

1. .r3- r - 1= 0 (Mad.ras,2000S) 2.x3-Sr-5=o
B.*3-Ex?+9r-3=0. 4.tanx=L.2x
5. r= 3 cos(r -n/4) 6. e' = Sr which is near x = 0.2.

I.9. OBJECIIVE
TYPE
OF AUESIIONS

Choose the correct answer or fiII wp the blanks in the follawing problems :
1. If for the equation 13 - fu;z + kx + 3 = 0, one root is the negative of another, then the value of ft is
(a) 3 ( b )- 3 (c) 1 (d)-1.
2. If the roots of the equation { - L = 0 are 1, G1,&2, ...,.., ar, _ 1, thgn
(1- cr) (1- az) ......(1 - an_ r) is equal to
(a) 0 (6) 1 G)n ( d )n + L .
8. If a, p, 1 are the roots of ?*3- 3x2+ 6r + 1 = 0, then cr2+ 92+f is
(a) L5l4 (6) - 3 k) - L5/4 (il s3t4.
4) x+2 is'afactorof
(a) xa +2 ( i l x a- x z + ! 2
(c) xa -'h)g - x+2 @ ) xa+ 2xB- x-2.
5. If a+B+1=5;crB+0y+fq=7;q0y=S,thentheequationwhoserootsarecr,pandyis
'(o)13-7=o
$)x3-7r2*3=o
( c ) 1 3- 5 x 2 + 7 r - 3 = o @)x3+7x2-3=0.
r8 HIGHER MATHEMATICS
ENGINEERING

*..m of the roots of,the equation *3 - 6i + 1Lr - 6 = 0 is 2, then the other two roots are
, '{6) 1and 3 (b) 0 and 4
l(s)-1andb @\-2and6.
?..15*,*Un*tion whose roots are the reciprocalsof the roots of x3 + prz * r = 0 is
@) L/r.13 + L/p .x* 1= 0
(d.)rx3+px,+1=0.
8. If 1 an'd2 are two roots of the equation *4 - *3 - Lli + 49x- 30 = 0, then the remaining two roots are
(a)-3and5' (b) Sand-5 (c) -6and5 (d) 6and-5.
9. If the r,oots ofr3 - 3x2 + px + L = 0, are in arithmetic progression then the sum of squares of the largest
and the smallest roots is
(a) 3 (b) 5 (c) 6 (d) 10.
10. Aroot ofr3 -&;2 +px+g=0 wherep and q arerealnumbersis g +i{g. The realrootis
@)2 (b) 6 (c) 9 (d) 12.
1 1 . O n e o f t h e r o o t s o f t h e e q u a t i o nf ( x ) = x n + a , n - L f r n - 1 + . . . . . .* a 1 x * 0 9 = 0 w h e r e d o , a L , . . . . , . a n - 1 a r e
real, is given to be 2 - 3i. Of the remaining, the next n - 2 roots are given to be 1, 2,3, ..,...,n -2.Tlte
nth root is
(a)''n. (bln-L k)2+3i ( d ' )- 2 + 3 i .
lS. If areal root of f (r) = 0lies in [o,6], thenthe signof f (a).f (b\ is ......
S" ryafies rule of signs states that ....
l*., [f,ci'F.,T arethe roots of the equationx3 - px* g = 0, then I L/ a = .......
16. If a,,B,y arethe roots ofrs = ?, then E aBis ......
16. A roo{ of rB .- \xz + 2.5 = 0 lies between1.L and 1.2. (True or False)
lf. hrsn equation with real coefficients,imaginary roots must occur in ......
1 8 .I f f ( a ) a n d f ( p ) a r e o f o p p o s i t e s i g n s , t h e n f ( r ) = 0 h a s a t l e a s t o n e r o o t b e t w e e n a a n d B p r o v i d e d

in*'Y*:l yaretherootsoftheequationrs
+Zr +3 = 0,thena+ 3, p+ 3, y+ 3 aretherootsoftheequation

E0. [f one root is doubleof another in r3 - Txz+ 86 = 0, then its roots are ..........
ttr. the equationwhoseroots are 10 times those of 13- zfi - 7 = 0, is ........
42. If o, p, T are the rootsof xs + pxz + qx,+r = 0, then X (L/crF)= ........
8& {3 and - 1+ i are the roots of the biquadratic equation........
m-* ffd.,,F; T are the roots of *s - Bx+ 2= 0, then the value of a2 + F, * f is ........
ftS. If there is a root off (r) = 0 in the interv al [a,6], then signof f (a)/f (bt is ..".....
8eIf+qx+r=o,thentheconditionforcr+F=0is..'''..'
87, The three roots of rB = 1 are ........
2t. one real root of the equation *3 + - 5 = 0 lies in the interval
(r) (2, 3), (ii) (3,4), " (iii) (L,2), (iu) (- 3, - 2)
-
&, If two roots of 13 Bx2+2 = 0 are equal, then its roots are .......
30. Ihe cubicequationwhosetwo rootsare b and 1- i is ..........
El. rhesum and productof the rootsof the equationx5=z are .........*d .....,..
'&. The equation *6 - *5 - 10r
+ T = 0 has four imaginary roots. (Ttrre or False)
$8. ooe real root of the equationx3+2x2 + b = 0 lies between............
.
'&L,trftherootsoftheequationx4+2x3-a*2-Z?.x*40=0are-b,-2,
land4,then a=...........

/
 Linea,r Algebro, : Determinants, Matrices
l. Introduction. 2. Determinaots, Cofactors, Laplace's expansion.S. Properties of determinants.
4. Multiplication of determinants. 5. Matrices, Special matrices. 6. Matrix operations. 7. Related
matrices. 8. Rank of a matrix, Elementary transformations, Elementary rnatrices, Inverse from
elementary matrices, Normal form of a matrix. 9. Partition method. 10. Solution of linear system
of equations. 11. Consistency of linear system of equations. 12. Linear and orthogonal
transformations. 18. Vectors ; Linear dependence,14, Eigen values and eigen vectors. 16.
Properties of eigen values. 16. Cayley-Hamilton theorem. 17. Reductionto diagonal form.
18. Reduction of qua&atic form to canonical form. 19. Nature of quadratic form 20. Complex
matrices. 21. Objective $pe of Questions.

2.I. INTRODUCTION
+et
: Linear algebracomprisesof the theory and applications of linear system of equations,lineai
transformations and eigen value problems. In linear algebra, we make a systematic use of
matrices and to a lesser extent determinants and their properties.
Determinants were frrst introduced for solving linear systems and have important
engineering applications in systems of differential equations, electrical networks, eigen-value
problems and so on. Many complicated expressions occurring in electrical and mechanical
systemscan be elegently simplified by expressingthem in the form of determinants.
Cayley* discoveredmatrices in the year 1860.But it was not until the twentieth century was
well-advanced that engineersheard of them. These days, however, matrices have been found to
be of great utility in many branches of applied mathematics such as algebraic and differential
equations,mechanics,theory ofelectrical circuits, nuclear physics,aerodSmamics and'astronomy.
With the advent of computers, the usage of matrix methods has been greatly facilitated.

2.2. DETERMINANTS
lrl

(1) Definition. The expr"r*ior, lll l1l * called a d,eterminantof thc seeond,
ord,er and,
laz ozl
standsfor'a1bg- a2b1'.Itcontains4 numberr'o1, bt, az,b2 (called elements)whichare arranged
along two horizontal lines (called rows) and two vertical lines (called columns).
l"t bL r1l
Similarly b,z cz is calleda d,eterminant of the third order,It consistsof g elements
lc2 Ds I
loe csl
which are arrangedin 3 rowsand 3 columns.
lot b 1 c1 dy.....lt
lo, b 2 c2 d2......12
In general,a d.eterminantof thenth ord,erisdenotedby l.--"
I bn cn dn""",I,
lo"
-* Arthur Caytey(1821-1895) was a professor at Cambri-dgeand is known for his important contributions
.\
to algebra, matrices arrd differential equations.
:\.
19 i\
20 HIGHER
ENGINEERING
MATHEMATICS

which is a block of nz elements arranged in the form of a square along n rou)s and n columns.
The diagonal through the left hand top corner which contains the elements o1, bz, cs,....../o is
called the leading or principal diagonal.
(2) Cofactors
\
The cofactor of any element in a determinant lp obtained by deteting the row and, column
which intersect in'that element with the proper sign.\The sign of an element in the ith row and,
*i. The cofactor
ith column is (- L)i of an element is u{ually denoted by the corresponding capital
letter.

1 " , b t r rl
For instance,in A= 1", b2 c2l, the cofactorof bsi.e.,Bs=(-1)3.rl:t :tl and
l"t bs ttl
,\ ro, brr
"r=- lo, asl.
(3) Laplacets expansion.* A d,eterminrantcan be expand,ed,
in terms of any row (or column)
as follows :
Multiply each element of the row (or colurnn) in terms of which we intend erpand,,ingthe
determinant, by its cofactor and then add up aII theseterms.
.'. Expanding by Er (r.e. lst row),

A,=atAr
+ b1814r r / l=
e,c, - -r.,1?' - orl"' "rl *
c:'l ?4
Te tlrr
gl "'lo, "rlo,
" s l "log ogl
= at(b*s * bs"z)- b1(a2cs
- ascz)+ c1(a2bs
- asbz)
Similarly expandineby Cz G.e.Znd column)

-orl7,
A-btBt+b282+b#s= '^rl - brl:t :tl
- rgl
*urll,
- :tl
log log .gl " czl
lo,
= - bt(azcs- asc2)+ b2(aps- agct\- bs(ap2- azct)
and expandingby Rs (i.e.Brd row), A,=a#s+ bsB3+ caCs.
Thus A'is the sutn of the products of the eletnents of an.u rntrt (nr onlrtt44l hrt l!12 a^ri^^$^$ ):--
4--:--
.od,uccts;r^-;;;rr,
If, howeuer,the sum of the products of the ele; *"*,o* ,oo,*ru*n,; r";r -";:r:::,:,
another row (or column) be taken, the result is zero.

e . g .i n ' l o t c r l-
, L , o # z l b # z + c s c 2 = - , . l b t " r l * ot Io,
"t lu, "rl lo, ';l ? tl
b;l
=- " lo,
as(bfs - bscr)+ b3(o1ca
- oscr)- ca(o1bs- o3br;= g
fn general, a; Ai + b; Bi + c; Ci = L when i =,1
=Q wheni*7

Example rA
2.1.
Expand,
o=|fr !
?l ls,r
sor. bynr, o=olur ^l! [l. rl! ,A
Expanding
-
N'_
= a(bc f2,{- h@c- gfl + g(hf - gb)= abc+ 2fgh - af?..- bgz;*chz,
- * Nameddftera greatFrenefrmath/maticiarPierce
SimonMarquisDeLaplace(174g:l&-D. Uetade
important cgrtributionsto-prqbabilpy'theory,specialfunctionr, - ' rfo6"ti"l
- theiry
------J uri"o"ppqy-Sm-;,
? 't'
professor in Paris, he taught Napolean Bonail-ari for a year. ""a
 LINEAR
ALGEBRA: DEIERMINANTS,
MATRICES 2l

lo 1 z 3l
theuatue o=lt g t, ?l
e z.z.Find.
Exampl
'"f
l3o12l
Sol. Since there are two zeros in the second row, therefore, expandingby Rz, we get

Ir z Bl lo 1 sl
a= - 13 0 1l +0-312 B 1l+0
lo r 2l 13 o 2l
(Expand by C1) (Expand by fir)
--i1":";lili - 3t0- (2x2- 3x1)+3(2x0- 3x3)l
: -'J?;:;.';?;:'
2.3.PROPERTIES
OF DETERMINANTS
The following properties, are proved for determinants of the third order, but these hold good
for determinants of any order. These properties enable us to simplify a given determinant and
evaluate it without expanding the given determinant.
i't
I.A d,eterminant remains unaltered,by changing its rows into columns and, colurnns into rows.

l"t b1
Let A = laz b2 "l [Expandby Rr]
"zl
.' 1", bs rrl
- a,t(bzh - b3c2)- b{a2cs- ascz)+ cr(a,zbs- agbz)

lot a,2 oeI

Then l'= bz Url [Expand
byRll
lar c 2 c g l
l"r
- bscil- a2 (bps - bgti + as (bf2- bzci'
= a,1(b2cs
= a.1(bscs-becil- b{a2cs- oecz)+ c1(a2bs-asb2)= 6.
Obs. 1. Any theorem concerning the rows of a d.eterminant, therefore, applies equally to its columns
anduice-uersa.
2. When a row or a colurnn is referred to in a general manner, it is called a line.
ll. If two parallel lines of a determinant are interchanged,, the d,eterminant retains its
nurnerieal ualue but ehanges in sign.

l"t bt rtI
Let L - laz bz ,rl [Expandby Rr]
l"t bs rr I
= a1(b2cs- bscil - bt@zce - a*il + c1(a2bs- asbz)
Interchanging C2 and C3, we have

l"t c1 arl
tt =loz c2 Url lExpandbyEr] ]
l"r ca arl ; .
= a1(csbs- csbz)
- cl(azbs - asbz)+ U1i,@4gg,i'u,scz)
= - {ar(b2ca- becil- br(orps- cplffieg ba- csbt)}= + A .
22 ENGINEERING
HIGHER MATHEMATICS

Cor. If a line of Abe passed.

over two parallel lines,i.e., if the resulting d.eterminontis like
la, cl orl
o,=lu) then a,=(- r)za.
"; ";l
lbr ca "rl
In general, if any line of a determinant be po,ssed ouer rn parallel lines, the resulting d.eterrninant
A, = (_ I)* L.
rrr. A determinant uanishesif two parallel lines are id,entical
Considera determinant A in which two parallel lines are identical.
Interchange of the identical lines leaves the determinant unaltered yet by the previous
property, the interchanges of two parallel lines changesthe sign of the determinant.
Hence A=A'=-[ Or 24=0,orA=0.
TV. If each element of a line be multiplied, by the samefaetor, the whole d.eterminantis
multiplied by that faetor.
1", pbr "rl 1", b1 "rl
i.e. lo, pbz c2l- p la2 bz czl
l"u pbe l"r bs 'gl
For on expanding by Cz, "l
L.H.S. = - pb1(azcs- ascz\+ pbX@ps- ascr)- pbs(apz- azct)
= p {- brBr + b2B2- bBBB}
= R.H.S.

I "t b1 c 1| 1", b1 crl

Similarly, lko, kbz kc2l - k la, b2 ,rl
I "r bs ', I 1", b3 'rl
- _Co". If two parallel lines be such that the elements of one are equi-multiptes of the elem.ents of the other,
the deterrninant uanishe s.

ie l:; i: i: lll=,,0,=o
ii:l=,lzi
l"r h pbtl bs arl 1",
Y. If eaehelernentof a line consistsof m ierms, the d.eterminantcan be expressed.
as the sum
of m determinants

consider
thedeterminanr
^=l:; i; :;:3;-:;l
bs
cs + d,s- e3
lor |
eachof whosethird column elementsconsistsof three terms.
Expanding A by Cs, w€ have
.
A = (cr + dt- ei@2bs- debil - @z+ dz - e)(a$s - asbl)+(cs+ ds- es)(a$2- azbt),
= lc1(a2bs- asbil - c2(a1bs- asb;) + cs(a1b2- azbt)l
+ Id{azbs - asbil - d2@1bs- aebi + d,s(a1b2 - azbil
- [e1@2bs- ag,bz)- e2(a1bs- aebt)+ es(a5bz - azbt)l
1", bt rtl lot bL dl lon br ti
=
lo, bz c2l+ la2 bz drl _ 1", bz ,zl
lot bs "rI l"r fu otl l"r bs ', I
- Further, if the elernentsof three parallel lines consistof m, n ond,p terrms respectiuely,the
dcterminants can be expressedas the sum of m x n x p d.eterminants.
LINEARALGEBRA: DETERMINANTS,
MATRICES ffi

l" a2 "t-tl
Example 2'3. tf lU b2 bs - 1l = O tn which a, b, c are different, show that abc*. I, ,
ir ,2 ,3-tl
Sol. As each term of Cgin the given determinant consists of two terms, we expFBBsit as a
sum of two determinants.

l" o2 o3-rl l" o2 otl l" o2 -11 lr a o'l l" oz 1l

= +
lu b2 b3-rl lo bz a3l la b2 - r l = a b clt b o'l-lu b2 rl
l' "2 "3-11 l" ,2 "ul l' ,2 -11 lr , ,21 l' "2 1l
-[Taking common a., b, c from Rt, Rz,Rs respectively of the first determinant and - 1 from
c30fthesecond
determt"i:''t
,?r a ,21
- o b " l L "b a r l - lrrr b b r l
lr c lr c
"l "l
[P as s ingC3ov e rC 2 a n d C 1 i n th e s e c o n d d e te rmi nant]

lr a o2l lr q, o2l
.. I r b uzl {oo"- 1)=0 Henceabc=1, since| 1 b b?l *0 asa,b,carealldiffsent.
lr c"l lr c"l
UI.If to each elernentsof a line be added equi-multiples of the correspond.ingefunve56'Sbne
',.
or rnoreparallel lines, the determinantsrem,ainsuna,Itered
lot bL "tl
Let n= loz b2 trl
lor bs rrl
la1+pb1-qcL b1 rrl l"t b1 rtl lnbt f u t r l l - e o r ' & 1 : 6 r I
Then L' = laz+ pbz- qcz b2 c2l= la2 b2 c2l+ lnbz b2 trl *l-atz b2 rul
I lpas bs rrl l-ara b9' rsl
llrllr
rs cs
I log os
bs
las+pbs-ecs
=A+0+0=A. 0$y.t'I*[hnl
Obs. ?hrs property is uery useful for simplifuing dctenninants. To add equi-multiples of p l lings,
we shall employ the following notation : , .,:
Supposeto the elements of the secondrow, we add p times the elements of the first row a1td,ff ,
the element of the third row ; then we say : ',
, ,':l
OperateR2+pR1+qRs. .':"r'
Similarly Operate 'Cst- mC1- nC; .'';"" r' :
means that to the elements of the thiid column add nr times the elements of the first column ant'..is:*ir&B
the elementsof the second."r"T;;
n 7 rcl
Example 2.4.Eualuate ,r, 9 tgl
ltt,
-
16 7 I 2l
Sol. Operating ftr R2- Ra, Rz- 3Rr, Rg - ?.R4,th'egtven determinant
l-e -Lz o -zl
-z
+ - 6 o - 111l
_ ll - 6
o^ =
o
15 7 1 2l
24 HIGHER
ENGINEERING
MATHEMATICS

=_ -:i -?l=o
l-3
l-4
-6 -11
'
Rt = 2Rz)

I x+2 2x+3 3x+4 |

Example2.6. Soluetheequation
lry * t_ Bx+! 4x + S | =0.
l3x+5 5 x + 8 1 0x+171
Sol. OperatingRa- (Er +.8/, we get
x+2 ?.x+3 3*+4
?*+3 3x+4 4x+5 -0 (Operate Rz-R1 and ^R1+.R3)
0 1 3r+8
x+2
x+L
?.x+4 6x+L2
x+L x+L =Q or (r+1)(r+2)11
Ir z 6 |
1 1 l=0
0 t 3r+8
To bring one more zero in C1, operete ^81- R2.
lo 1 Br+Bl

lo I =o 1 s
(r+_lxtr.t'll
i ,,t*rl
N o w e x p a n d b y C r . . . - ( " ' * 1 ) ( r +Z ) ( B x + 8 - b ) = 0 o r - B ( r + 1 ) ( r+ 2 ) ( x + 1 ) = 0 .
Thus t c , =- L , - 1 , - 2 .
It+a I I rl *
1, I +b L t I r
Example Z.G.prouethat
- | 1 I I /)
| I 1 l+c
I+dl ' '/
Sol. Let A be the giveJ ut"*r*irrt".*. tjr.rrru d, b, c, d common from Rb Rz, Re,Rt
respectively, we get

lo-t* L a-r a-l a-l

I
L - o ; b c d l 'b _ r b _ 1 + 1 b-r
-1+
b-r
I c-r c-L , I c-l
I a-t d-r
[operate Er + (Rz+.83 +Ra) and take out the commonfactor from R1l

- abc d( 1 + a - 1 + b - 1 + I + d - r)
"-

c11

=abcd,[t.*****.*)

Obs' If all elements on onersideof the leading diagonal are zero, then
the determinant is equal to the
product of leading diagonal elements and such
a &terminant is called atriangular determinant.
lo2 +X ab ac ad. I
Example 2.7,Show thatl ab bz +), bc bd- | -
I ac bc ,i*x ;; l=1't(ot+b2+c2+d2+7"1
I ad bd cd. d2 + 7,1
 ALGEBRA
LINEAR MAIS!q!!
: DETERMINANTS,

Sol. Let A be the grven determinant'

by abcd' Then
Multiply C1, C2, Cs,C+by a, b, c, d' respectively and divide
Io3+ o)" abz o"2 d)2 t
. 1 | o'u b3+blu bcz bdz I
o = -ob"d cd'z t
I o'" bzc '3 +
"1" + d?"1
bzd ,2d ds
| "'o
we get
Taking a, b, c, d common from R1.,R2,83 and '84 respectively'

lo2+x bz ,2 d'2 t
. abcd.I o' b2+L ,2 d'2 |
o=ob"dl
o, bz c 2 + 7 , d,2 t
bz ,2 d,2+Ll
I o2
[operatec1+ (c2+ c3 + c1) and take out the commonfactorfrom c1J

Iines become id.entical when )c= dt then x'a is a factor of L"

Let t=f(x)
Since A=0when x=Q, :. f (a)-Q
i.e. (x - a) is a factor of f (x).
Hence x - a' is a factor of A.
Obs,IfhparallellincsofadetenninantAbemmeidentical'wlrcnfi=a,tlun(x-Ok-LisafactorofA'

Similarly, a - c and a - d are also factors of A'

-c is afactor of A.
Again putting b =c,Rz=.83 and henceA=0. :. b
Similarly b - dand c - d arealso factorsofA'
any other algebraic
Also A is of the sixth degreein a, b, c, d and therefore, there cannot be
factor of A.
constant'
...supposeA= h(a-b)(a-c)(a-d,xb -c)(b -d)(c -d), where kis anumerical
=
The leading term in A = o\bL".The correspondingterm on R.H.S. ka\bZc'
k-r.
Hence A = (o- bXo- cXo- d)(b- cxb- dXc- d).
 HIGHER
ENGINEERING
MATHEMATICS

Exampte2.9.Factorize
^=l'* :' A
l"t bs
Sol. Putting a = b, Cr = Cz and hence A "=l0.
.'. a.- b is a factor of A. Similarly b -c andc -a are also factors of A.
:. (a - b)(b - c)(c - o) is a third degree factor ofA which itself is of the fifth degree as is judged
from the leading term bzc!.
.'. The remaining factor must be of the second degree. As A is symmetrical in a, b, c, the
remaining factor must, therefore, be of the form k (o2 + bz + c2)+ t(bc + ca + ab).
.'.A- (u.-b)(b -cXc -d.){k@2 +b2 +cz)+t(bc +ca+ab)\.
If k * 0, we shall get terms like aab, b4c etc.which do not occur in A. Hence ft must be zero.
L,= l(a - b)(b - cxc - a)(bc + ca +ob). The leading term in A = b2c3.
The corresponding term on R.H.S . =l bzc!. :. I = L
Hence L = (a - b)(b - c)(c - a)(bc + ca + ab).

lp*r12 o2 o2 |
E x a m p l e 2 ' 1 0P.r o u e t h aI t b 2 @+a,)2 b2 l=2abc(a+b+c)g
I t'"2 put a = 0,
(a + b)21 rJ.N.T.u.,rssl)
Sol. Let the given determinant be A. If we

l{u*")' o ol
A=l o brl_O
",22 b'l
I t'
:. a is a factor of A. Similarly b and c are its factors.
Againifweputo+b+c=0.

ll."f o2 o2 |
A=l b2 eb\ b2 l=o
| "' (- r)21
"2
In this, three columns being identical , (a * b + c12is a factor of A.
As A is of the sixth degree and is symmetrical in a,b, c the remaining factor must therefore,
be of the first degree and of the form k (a + b + c).
T h u sL , - h a b c ( a + b + c ) g
To determinek, put a = b =c = 1, then
l+ 1 1l
It 4 Ll=ZZn or54=27k i . e .k = 2
11 1 4l
H e n c eL , = 2 a b c( a + b + c ) 3 .
Otherwise : Operating Ct - Cs and C2 - Cs, we have

Ita*"\2-oL o a2 I
A=l 0 1c+a12-b2 b2 | ttute(o+b+c)commonfromClandC2l
- (a + b)z c2- (a + b)2 p +bfl
lt'
lb+"-o 0 o2 |
=(a*b+c)2 O c+a-b b 2 | t O p " . a t e R s - R 1- R z l
|
lc-a-b c-a-b 1a+b121
LINEARALGEBRA: DETERMINANTS,
MATRICES 27
b+c-a 0
=(a*b+c)z 0 c+a-b
"'l
;, | e, * cs,cz* *
[operate *
-Zb -2a, zabl "rt
b+c oz/b
=(a* b+c)z "'l [Expand by.Esl
bz/a c+a b'l
0 0 zabl
= Zab(a + b + q2 t(b+ c) (c+ a) - abl= 2abc(o+ b'+c)3.

2.4.MULTIPLICATION
OF DETERMINANTS
Theprod,uctof two d.eterminants
of the s(nneord,eris itself a d.eterminant
of that ord,er.
l"t b1 cll lt, ftt1
Let A1=lo, bz c2l and Lz-llz rrl2 "tl
cg "zl
l"t 4 I ms
llt "tl
then their product is defined as
lot tt + b 1 m 1 1 c 1 n 1 , a 1 l 2 + b tmZ * c{L2t al l s + btms+ cl ng I
Af AZ =la2l1+bgnl* c2tt1, a2I2+ bZnZI c2tt2, a2lg+bZrny+ c2ngl
+ bsm1 l cgn.1, agl2 + bsmZ * cgrt2, aglg+ bsms + can3
lott, |
Similafly the product of two determinants of the zth order is a determinant of the nth order.
lo2+t2 ab+c7',"ca-b\rl I r c -bl
Example 2.11.Evaluatelab - c7, b2+Lz bc+ alrl x l- 1,
" -a "l
l c a + b L b c - a 7 , c 2 + l 2l I b Il
Sol. By the rule of multiplicationof determinants,the resultingdeterminant
ld' dn drs|
t= ldzt dzz dzsl
lo' dsz asel
where dtt= (o2+ [2ll + (ab +c],)c + (ca -b]'X- b) =t.(az + bz + cz + L2l1
dtz= (oz +lzx- r) + (ab + clv)}r.
+ (ca- bl)a = 0
d13= o'
d z t = O , d z z = M a+
2 b 2+ " 2 + * 1 , d z g = 0 .
d g t = o , d g z = 0 , d g g= 7 4 a 2+ b z + c z + 7 1 2 1 ,

lr4' + b2 + c2 + 7,2'1 o ol
Hence o=l o \.(oz+bz+cz+?rz.y
o o L ( a z +b'o* ^11
I
\'s(oz
+b2+"2+?&1s.
"',*
lo, Br c, | 1", bL "rl'
Example%.l?. Showthat lA2 Bz C2l = b2 trl whereA, B etc.are the co-factors
la2
lo, B; "rl l"; 6; ';l
of a; b, etc. respectiuelyin the determinant (albgcg\.

l"t b1 rtl
la, 81 crl
soLLetA=l:?""u t?
^'=l,r:
i? ::l S:l
28 HIGHERENGINEERINGMATHEMATICS

Then

loret + b1I.1*c1C1, atAz + b182* c1C2, a{s + b1Bs+ c1Csl tO 0 0l

A A ' - - l " A r + b 2 8 1 * c 2 C 1 , a $ z + b 2 8 2 * c 2 C 2 , a f i s + b 2 B s + c 2 C s=l l O A 0l =lt
*caC1, a#z+bsB2*caC2, a#s+bsBs+csC| 0 Al
l"*tt+b381 lO
HenceN = L2.
Obs. A'is called the reciprocal or adjugate determinant of A.

lzb,- o' ,2 bz I
Example 2.18. Express| Zca-bz o2 |
",
I o' o2 2ab- czl
as the squareof a determinant, and heneefind its value.
Sol. Given determinant
l".t- a)+b.c+c.b, a.(-b)+b.a+c.c, a.(-c)+b.b+c.al l" b cl l-" c bl
= l b . ( - a ) + c . c + a . b , b . ( - b ) + c . a + d . c , b . ( - c ) + c . b + a . a=l l a c d , l* l - a a rl
l r . ( - " ) + a . c + b . b , c . ( - b ) + a . a , + b . c ,c . ( - e ) + a . b + b . a l I r q , b l l - c b al
lTake out (-l):common from C1 and interchangeC2,Cs.l
labtl b
=lb c o[x1-ry2 c "ol=tz
l
lc a bl d, bl

where o=l; 2;l =-(os+bs+cB-Babc)

lc & bl
Hence the given determinant = A2 = (aB + b3 + - 3obt12.
"3

1. Prove, without exBanding, that

ll ; ;: -i:lvinishes
11 c cz-abl
l" xz 1+r'l
z. tf
ll y2 r *ytl = o, th"o prove,without expansion,that 3ry2= - 1 where x, !, zare unequal.
l" '2 | + z3l
* nd.hra,Iggg ; Assam,
Iggg)
l xr l m lt m 1l
s.shoedtn*ll : ll =o-aXr-Fxr-e.
;
laPrll
l.a b c I
4 . S h o w t h a lt b + c c+& a+bl- -@-b)(b-cXc -a)(a+b+c).
l"' b2
"2 |
rra,b,carealldiffererr."u ;: ;i-il =o,rn"rrshow
thatabc(bc+ca+ab)=a+b+c.
ll
lr ,3 "o-rl
 29
ATGEBRA
LINEAR M
: DETERMIryANIS,

l12 22 gz 42
2 3 +l gz
334l.
244l|
o,tl'r1 42
42
s2
52
62
2 3 5l ln' s2 o2 72
Prove the following results : (7 to 12)
b cl 2b 2cl
la+b b+c c+al la la-b-c i,
rL+n n+ll*ll rrl nl=2 8.1 2a b-c-a l=(o+b+c)3
Z.lt+m c-a-bl
q+r '*ol lr q r'[ l2o
2b
ln*t
2b -2b l'
lr*o'-b' 2a * a perfectcube'
9. | 2ab L-az +bz -"1
% -2a2 t-az -tzl
|
6rl lr s o 2 a3 + bcdl
la5
'"l: ? ; :l
i' aPerfectsquare' ,r.
' ^ ' [t
11
b 6z bB+ cdalvanishes.
co+ dnbl
" 2P I + abcl
" d
lxY z 0l lL d
ft cosA ti"4l B-C. C-A-,-A-B
u. l1 cosB tf{l =4sin-isinn-t* 2'
It cosC sinC[
Factorize each of the following determinants : (13 to 15)

1 L tl 1b2"2 + o2 bc + a
a b cl (Andhra, 1998) L4. lc2a2+ bz ca + b
ab + c
o2 b2
"'l l o' b' + ' 2
o2 bz c2 d'l
b t dl
15. a
1 1 1 1l
bcd cda dab abcl
lo-* c
(Andhra, 1999)
1 6 .I f o + b + c = 0 , s o l v e l c b-x
lI b a

';t
theequat,""
12.sorve I T:tt #l ll =t
f a r +r 6x+4 8r+41

lbz*"z ab ac I to c bfi
18. Show that I ba + o2 bc | = l" 0 = 4a2b2c2
"2 a, "l
0l
| "" cb oz+uzl lb
?-o1+}bl
,lTr;:::
19.Provethatthedeterminan i::i::Sbslis a multiple of the determinant
?lo,z+
+ ca ca+ 3ag 2o3+ 3b3l
lzb3
lot bt .tl
lo, bz
"rl.
l"r bs trl
2.s.MATRICES
I
(t) DCfinition.A systemof mn numbersarranged,in a rectangularformatio.nalong m rows
;;'lr,"
as
onati'rot\t;lr;'dhi|;6;;;'i by ihe brackeist I is tilttd qn rn bv n matrix ; whieh is written
ri i is also denotedby a single capital letter'
Trpgba{.*.lAlmatrix
"
30 HIGHERENGINEERINGMATI-IEMATICS

att atZ ...a1j ..,atn

aZL az7 .,.agj ...azn

Thus A =
;,;; i')
Aml dm2 Amj ...Qmn

is a matrix of order mn. It has nz rows and n columns. Each of the rnn numbers is called an
element of the matrix.
To locate any particular element of a matrix, the elements are denoted by a letter followed by
two suffixes which respectively specify the rows and the qolumn$. Thus a;; is the element in the
i-th row andT-th column ofA. In this notation, the matrixA is denoted AV btl.
A matrix should be treated as o single entity with a number of componcnfs,rather than a collection of
numbers. For example, the co-ordinatesof a point in solid geometry, are given by a set of three numbers
which can be represented by the matrix lr,y,zl. Unlike a determinant, a matrix cannot reduce to a single
number and the question of finding the value of a rnatrix never arises.The differencebetweena determinant
and a matrix is brought out by the fact that an interchange of rows and columns does not alter the
determinant but gives an entirely different matrix.
(2) Special matrices
Row and column matrices. A matrix hauing a single row is called a row matrix e.g.,
u 3 5 n.
A matri"tchauing a single column is called a column rnatrix, e.9.,13l
Izl
L5l
Row and column matrices are sometimes called row vectorsand column uectors.
Square matrix. A matris having n rows and n colurnns is called & squsre matrix of order
\
n.
The determinant having the same elements as the BquarematrixA is called t}rredeterminant
of the matri* and is denoted by the symbol I A | . For example,if
[r 2 sl lr z sl
l,=lz B 4l,then;a;=lz g 4l
L3 4 5J 13 4 5l
The diagonal of this matrix containing the elements 1, 3, 5 is called t}neleading or principal
diagonal. The sum of the diagonal elements of a square matrix A is called the trace of A.
A square matrix is said to be singular if its determinant is zero otherwise non-singulbr.
Diagonal matri* A square matrix all of whose elernents except those in the leading diagonal,
are zero is called a diagonal ma,trin.
A diagonal matrix whose all the leading diagonal elements are equal is called a scalar matrix.
For example,
[s ool [sool
lO -2 0landl0 3 0l
lo o,6l [o o Bl
are the diagonal and sc6lar matrices respectively.
Unit matrix. A diagonal matrix of order n which has unity for all its diagonq.lelemenfs,is
called a unit matrix or an identity matrix of order n and is denoted by .In.For example, unit
matrix of order 3 is

[i:tl
LINEARALGEBRA: DETERMINANTS,
MATRICES 3l
Null matrix. If all the elernentsof a rnatrix q,rezero, it is called a null or zero rnatrix and is
denoted!y'0' ; e.g.,
l-q 0o 0o OJ
ol,^^
*tnullmatrix'
L0
Symrnetric and skew-symmetric matrices . A square matri* 4 - la;il is said.to be
symmetric when aij= ajifor atl i and,j.
If aij: - aji for all i and j so that all the leading diagonal elements ere zero, then the matrix
is called'a skew-symmetric matrix. Examples of symmetric and skew-symmetric matrices are
respectively.

la h sl I o h -sl
lh b /landl-a 0 f
ls f cJ -f oj I
ls
Triangular matrix. A square matrix all of whoseelementi below the leading diagonal are
zero' is called an upper triangular rnatrix. A square matrix all of whose elements above the
leading diagonal are zero, is called alower triangular matri,x.Thus
[,e h sl [t
llN 7l andlz
llo \cJ 1 L
are upper and lower triangul-ar matrices-respectiveiy.

2.6.MATRIXOPERATIONS
(1) Equality of Matrices
Two matrices A and B are said to equal if and. only if
(i) they o,re of the same order
and (ii) each elernent of A is equal to the correspond.ingelernent of B.
(2) Addition and subtraction of matri ces.If A, B be two matrices
of the san1eord,er, then
their surn A +B is defined as the matrir each element of which is'the sum of the correspond,ing
elements of A and B.

2. Addition of matrices is commutative,

L.e. A+B=B+A.
3. Addition and subtraction of matrices is associatiue.
(A+B) - C =A+ (B- A = B+ (A- C).
(3) Multiplication of matrix by a scalar. The prod,uct of a rnatrix A by a scalar k is a
matrix whoseeachelementis k times the eorrespond,ing "
elementi,ofA.
|
b1 ,r] =l!",
rhus h l"t bz
'"
c2l-
!!, urrl
kbz
Lo, lka2 nr")
The distributiue latp holds for such products, i.e. k(A + B) = kA + kB.
i ,

Obs. All the laws of ordinary algebra hold for the addition or subtraction of rnatrices
and their
multiplication by scalars.
32 ENGINEERING
HIGHER MATHEMATICS

Example 2.L4, Find x, !, z and w giuen -that

l
o[* yl-[' 5 ,[- 6 x+yl
- l z * )--5l z; x+cw+ v l s -]
-'1" *l_[
s o t . w e h' -a1v3e" I l 3tI 't*l)=- L[ '- 1 +rz++6w 2w+5 J
Equating the correspondingelements,we get
3 x- x + 6 , 3 y= 5 + x * ! , 3 2 = = - - +
1z +w,3w =2w +5.
b-6,2y=5+X,%=w-L,w=5
H e n c ex = 3 , y = ' 4 r z = 2 , w - 5
I s s -71
Example 2.L5. Express| - A 11 a las the sum of a lower triangilar matrix with zero

Ieadingd.iasonal
anaan *olulrtrr";;k", iJr*.

Sol.Let, =f l 3 3lo" thelowertriangular

mutrixwithzeroleading
diagonal.
Lb c 0 J

and , =l'o f I I ," theuppertraingularmatrix.

L0 0 rl
[s b -71 [o o ol lt rrl n1
Thenl-8 11 al=la o ol+l o p ql
-14 6l Lb c oJ lo o rl
L 13
Equatingcorrespondingelementsfrombothsides,weobtainS=1,5-m,-7=n,-8=d,
1t =p, 4= q,13 = b,- L4=c,6 =r.
o o.l [s b -71
Hencer=[-B 0 0landU=10 11 4l
-r4 oJ
113 [o o 6l
(4) Multiplication of matrices. Two rndtrices can be multiplied only when the number of
columns in the first is equal to the number of rows in the second. Such matrices are said to be
conformable. \

Forinstance,the0""u".,[i
li if "f;l t]
ton;:, :l L"; ";l
lort, + b1m1*c1n1 a1l2* btmz+ cp2]
isd,efined.
asthematrixl::i::i:::::::: ::i::i:::::#:)
b+mt* c4n1 a4I2+b+mz+can2;)
La4l1+
at2 ...dLn-l bn urof
["tt [u1
o" a22 "'azn bzz 'rol
ifA =l
In general, l"ra n -lbn
t...tt......1
dm2 ...o*n) bnz U"ol
lo*t lb",
betwo tnxn andnxpconformablematrices,tl,""theirproductisdefinedasthe rnxp matrix
 LINEARALGEBRA: DETERMIN,ANTS,
MATRICES 33
.r

|l: ::?
o"=1, ".:: :.: ,r:
,r.1cmz
"ol
I c,npl
Lt-t
where cij = air6y + aizbzj+ a;sbsi+ ... + dinbnj,i.e.,theelJmunt in the lth row and the.Tth
column
of the matrixAB is obtained by weaving the lth row ofA withTth column of B. The
expression
fot ciiis known as the inner product of the lth row with theTth coiumn.
Post-multiplication and, Pre-multi,plication In the product AB, the matrix A is said to be
post-multiplied by the matrix B. whereas in BA, the matrix A is said to be pre-multipliecl
by
B. In one case the product may exist and in the other case it may not. Also the products
in both
casesmay exist yet may or may not be equal.
9b". t. Multiplication of matricesis associatiue. i.e. (AB)c = A(BC|
providedA, B areconformablefor the product^48and B, C arcconformablefor the product
BC. 1Bx.Z.L6).
Obs. 2. Multiplication of nzatri.cesis distributiue. i.e.A(B + C) =AB +AC.
providedA, B ate conformablefor the product AB and A, C areconformablefor
the product AC.
Obs. 3. Power of a matrix. If A be a squarematrix, then the productAA is definedas A2. Similarlv
we define higher powersof A. i.e.AA2 = A3,A, . A2 = Aa etc.
If 42 = A, then tlzemcttrixA is called,iclempotent.
ltJ I a, | | I -l:r]*,-
_zl

Example rrA=lr,
2;16. 3r1"",, =l-:, theprod,uct
ofAB rsBAdefined,
?
i "
Sol. since the tt*n.l"of ;";1, ofA = ttlu ru*ri, urrows of B (eachbeing = B).
.'. The productAB is definedand
-
[ o . r + 1 . 1 + z . z 0 . - 2 + 1 . 0 + 2 .- 1 l [ s -z]
= l 1 . 1+ 2 . - I + J . 2 t._Z+2.0+8._fl=lS _ 51
-81
Again sincethenumb"L:;:hf::, .?;;il';:JJ ;*l
.'. The product BA is not possible. "ll
Exampr ', 3 4 1
e2.r7. rrA=l-t,t ?1,,=l 2 3l, computeAB and BA and show
7 AB*BA. Lo o 2) [-r 1zil
Sol. Considering rows ofA and columns of B, we have
-
t L . 2+ 8 . 1 ' +o . 1 , 1 . 8+ 8 . 2+ 0 . 1 , r.4+S.B+0.21 s 1B.l
AB =l-1.2+2.t + 1. 1, - 1 . 8 + 2 . 1+ 1 . 1 , - 1 . 4 + 2 . 3 + 1 . 2[ s1 = l -21 4 l
-
L 0 . 2+ 0 . 1+ Z . t , 0 . 8+ 0 , 2+ 2 . L , 0 . 4 + o . B + z . zl -Jz 2 4 l
Again consideringthe rows of B and columnsofA, we have
-
I z . t + B . 1 + 4 . 0 , 2 . 8+ 8 . 2 + 4 . 0 , 2.0+ 8.1+ a.zl [-r 12 11.l
B A = l 1 . 1+ 2 . - 1 + 8 . 0 , 1 . 8 + 2 . 2 + 8 . 0 , 1 . 0 + 2 . L + 31.=
2l - . 1 7 8l
l - 1 . 1 + 1 .- 1 + 2 . 0 , _ I . S+ 1 , 2 + 2 . 0 , -1.0+ 1.1+ 2.2J [-z -l bj
EvidentlyAB*BA.
-,
s.xample
2.18.O=lt! i r*o thematrix
Bsuch
thatAB= ? 11
3 4l [
" L 5 [s 6 4 )
1
"l, (Bontbay, 2005J
 MATHEMAiICS
ENGINEERING
HTGHER
34
fs z z l I t
q "r1I
m
Sol.Letaa=l 1 B lllo
L5 3 4)1" u w)
I et+zp+ 2 u 3 m+ 2 q + 2 v 3 n+ 2 r + 2 w l
=l l+.Bp+u m+3q+u n+3r+w I
+4u 5m+3q+4u 5n+3r+4w )
[5]+3r
[s 4 21
= l 1 6 1 | (given)
L5 6 4l
Equating corresponding elements, we get
3t + 2 p + 2 u = 3 , I + 3 p + r' L= L , 5l + 3p + 4u = 5
i,,
3 m ,+ 2 q + 2 v = 4 , m + 3 q + u = 6 , 5 m + 3 q + 4 u = 6 ...(tt)
3n +2r +2w =2, n +3r* Ir= 1, 5n +3r + 4w = 4 ...(iiil
Solving the equations (i), we get I = 1, P = 0, u = 0
Similarly equations (ii) give nz = 0, I = 2, u = 0
--
and equations (iii) grve n = O, r = 0, w L
f t o ol
T h u s r = l0 z 0 1 .
lL o
J[r
o 1l
321
--
Example 2.19. Prouethat A3 - 4A2- 3A + 111 0, whercA -12 0 -11.
LI 2 3 1
[tg 21 l-r 3 21 | ue+z 3+o+42-3+61[o / o l
S o l . A 2 = A xA = 1 2 o -rlxlz 0 - t l = l z + o - r 6 + 0 - 24 + 0 - 3 1 = l4r 1 l
Irz sJ i t 2 al [r+++s 3+o+6 2-2+e) L8 e e l
l-g 7 5l [r 3 z1 [g+t+*s 2?+0+1018-?+151lza 37 261
A3=Az xA =l t 4 t lxl z 0 -1 l=l t+g+t 3+0+2 2-4+3 l=l 10 5 1 l
2 s] ls+ra+o 24+0+t8 16-e+27 42 3 4 )
[a e eJ [t I 135
A3-4a,2-3A+11r
lzeBT zal [g7 5l fr3 z] ftool
=lto b rl-+lr4 rl-slz0-rl+rrlolol
Ltu42s4l [aeo] lt2 sj [oot]
[28-86-3+t1, 37-28-9+0, 26-20-6+0I [o 0 0l
=l 10-4-0-0, 5-16+0+11,1-4+3+0 l=l 0 0 0l=0.
lBb-82-3+0, 42-36-6+0,3a-36-9+11j L0 0 0J
Exampl e Z.ZO.By mathematical ind,uction, proue that if

"=1l',:|yhenA"=['
- 10"
;:i;"]
ot = lt
\ n = !, Angives
Sol.Whe :'rl
...(i)
L
Let us assumethat the result ii true for any positive integer ft, so that
toa -25k1
At,=[ r +4k
|
=lr.nI*,:ikl
**t _N.AL
-
L-lOk,J
-:3]
Ii'
 MAIRICES
LINEARATGEBRA: DETERMINANTS,

- 100ft - 25 (L + 10ft)+ 225k1

= I[ rr (1 + 10fr)
4 4 h + 4 ( 1- 1 0 f r ) - 1 0 0 f t - e ( 1- l o h ) J
- iLt * 1 0 ( & + 1 ) -25(k+L)l
4(fr+1) 1.-10(ft+1)l
This is true for ru- lz+ L ...Gi)
r We have seenin (i) that the result is true for n t.=
.'. It is true for n = L + L =2 lby (lt)
Similarly it is true for n = 2 + 1 = 3 and so on.
Hence by mathematical induction, the result is true for all positive integers n.
Exampl e 2.2!, Proue that (AB)C =A(BC), whereA, B, C are matrices ton!?:Yo_blefor the
products. (J.N.T.U., 2002 S)
S ol. Let A - larl b e o f o rd e r mx n ,B-[b i ;! be of order nxp and C -Ici i J be of order of
p xq.
,n
' Then AB = [c,,rl [buf,=\ a,r"bu
h ,= L
p

(AB)C = prr=[ T l; o,uuu,),,)=f

=fi o,ubu,f ; 1 lL= 1o,ruu,')
-1 --
Lu=t I L. I [fr=l ) I Lk= I

Similarly, BC = lbnil. [c1il- Z bu, tj

' = t l "ri
I p
A(BC)-tarn)
| I
l - n7.
untrul=l T,
il ,,rlLao,,,;l'l=f
i **b;u,',j\
) ) L u = [,i
fu*1.
=',
L'=1 J L [,=t t
Hence (AB)C -A(BC).

-. 'u
7 1. For what valuesof r, the matrix[3
l-z 4
-* ?li, ,irrgohr r
-1-*j

2. Find the valuesof x, y, z andawhlchsatisfythe'matrixequation I -? f ;]= :;]

F: [3
B. MatrixA hasf, rowsandr+5 columns.MatrixB hasy rowsand 11-y columns.BothAB andBA
exist.Findr andy.
-l]
4. rrA* = o- u= theproduct
l] .ur.o,"te AB.
" [l ""u [l
u.oo=fl Z I ll ""u"=[3 t !1 *. ABor.BA,whicheverexists.
lslosl Llo1l
I r '),"=l?
6.IfA =l_; ; =A(BC)andA(B
+c)=AB+AC.
l],"ur=r; l],""*that(AB)c
: .[s ? '"ll4..ft 1-1lf l 1l ru sl
7.Evaruate(i)
rr,r,ov _i
(,,)L_$ ,
? tj |1j; 3J.L:3iJ"L_;ij
 MATHEMATICS
ENGINEERING
HIGHER
36
lrl lzl

8. Prove that the product of two matrices

cos0 sin ol cosOsin Ql
[ .ort e u.ra | .or'O
lcos0sin0 sinz0 l*'-[cosqsinq sin"0 J
is a null matrix when 0 and 0 differ by aq odd multiple of n/2.
l- n
u,. - tan *u/2]
''l'sho v'tthatl
-r^, cos u, -sin cr-l
9. IfA=1. + A = ( I - t l,' f
1ra^urz LSrno cosctl
io el
-GA+8.I,wherelisaumtffiatrix'of secondorder' @.P.T.U.,2006)
10. IfA=l| | l,n"ather,'alue of A2
L' "-J
:,,,,', [ | '2 3ll , . . A+9/.
1 1 .I f A = l t , t - r l ; a n d / i s t h e u n i t m a t r i x o f o r d e re3v,a l u a t e A- 3
z
2)
L-3 1
-1.l, -l
,r.,ro =[]', =[3 (A+B)z=Az+ BA+
]1.u"-* theresuu AB+ Bz.
r 2) " -3 2J
1_o l_+
[o1ol ioool
tg.IfE=i0 0 llandF'=i1 0 01,
loool loloJ
calculate'the products EF and FE and show that EzF + FEz = E.
r- -"l
o' t;;;l,.no*ttute" =[- :t: l': n is apositive integer.
14. rfA=l "o" a coscrr'
[-sin cosnsr'when
ro :l::l,
[r _z ,+-.u
1b. FactorizethematrixA=17 1 -5 lintothe formL(J,where.Lislowertriangular andU isupper

1374)
triangular matrix.

2 7. RELATED
MATRICES
(1) Transpose of a matrix, The matrix obtained from any giuen matrix A, by interchanging
rows and cblurnns is called the transpose of A and is denoted by A'.

matrix of A=14
Thus the transposed
l-t ?1. a, [r 4 7]
S is A'=
; ; ;]
17 8l
[Z
Clearly the transpose of an nxxrimatrix is an nxm matrix. Also the transpose of the
transpose of a matrix coincides with itself, i,e. (A'1' - 4.
For a symmetric matrix, A' = A and for a skew-symmetric matrix, A' = - A.
Obs, 1. The transpose of the prod,uct of the two matrices is the product ot''tlteir transposes taken in the
reuerseorder l.e. (AB)'= B'A'.
For, the element in the lth row andTth col. of (AB)'
= element in theTth row and ith col',ofA.B = inner product ofTth row ofA with ith col. of B
- inner product ofTth col. of A'with ith row of B'= element in the ith row and.Tth col. of B'A'
Hence (AB)' = B'A' .
Obs. 2. Eucry squdre matrix can be uniquely expressed,u,sa sutn of'a syrnmetric and. a skew-sytnmetricl
nz'atrix' (J'N'T'U" 2oor)

LetA bethegivensquarematrix,then A=+(A +A')*+6 -A')'

Let n =l<d * e')andc =|te - e'i

 LINEARALGEBRA: DETERMINANTS,
MATRICES
37
r - f

a'=lla+a'l-j =iro'+ (A')'l=!te'+A)=B i.e.a = j(A+A,)isa symmetric

matrix.
- 4

Again c' =l - o', = - (A')'l= -

L**
-J.22 | *ro' f,a' e1=- f, i.e.c = *ro- A,) is a skew-s5rmmetric
matrix.
Hence A can be expressed as the sum of a symmetric
and a skew-symmetric matrix.
uniqueness' assume that P is a symmetric marrix and-e
,n"r l"lil"A:n" i, u ,f."--ry**etric. matrix such
Then A'=(P+8)'=P'+Q,=p-e
Thus P = + ( A + A ' t a n d 0 = * ( A- A , )
rvhich shows that there is one and only one way of
expressingA as the sum of a symmetric and skew-
symmetric matrix.
Exampl e 2'22' Express the m'atrix A as the sum of a symmetric
and, a skew-symmetric matri^tc
where

I a z _rl
A=l I 3 _61
L_5 0 _71
1 -5.l
> Sol. We have
l+
A ' = l 2 3 ol
L*3 -6 - 7 )
3 -8.] t z]
rhen A-o,-L 6 -Gl and = Il _
o1
o _61
i -6 - 14_J
L-2 6 0l
tl .r

:31.[-rgo3 -;] .
A=*ro+A,)*|a-A,)=ltf
z'
(2)Adjoint or a square*'t"i*. rhe detef;r-""; 0J
3rrn" :{"L"*trix
f o r b ' rJ l ttl bL rrl
A. =ll-or , c z i s o =l o , b z , r l
lv z3 I I o , og cg
L*r "3 | lo3 I
7 rhematrix
ft
formed
o!?. i;*;x{ ofrheih"ft,"T1,,
r) rt1

;: !ll. .n*"thetranspose
l;: D3 orthis
matrix
,r.,.lt: A: ftl
lrr n ^-l
L 3 rrJ c2 c'|
|c, J
iscalledtheadjointofthematrixAandiswrittenasAd"j.A...
Thus the adjoint of A is the transposed. matrix
of eofactors of A.
(3) Inverse of a matri x. If
A be any nratrin, then a *otrt* B if it exiqts,
such that
AB = $4= r, is aalled' the inverse olA which
is denoted by4- r so that AA- r - [.
Also 4-r' =4di' A
lAl

) .-
* " -Ipr,4
lAdjA)=lo, bz ,rl
|or br "r l
[", bL rrl

Fi ,i,
uil]{r ^,fi
,i,l ;il
 MATHEMATIQ
ENGINEERING
HIGHER

or l.ffi=t t'.' lAl*01 or

ffiistheinverseofA.
Obs. L. Irouerseof a nratrix, is unique.
For if possible, let the two inverses of the matrixA be B and C,
then AB=BA=I and AC = CA= I
CAB=(CA)B=IB=B and CAB = C(,48) = CI = C
Thus B = C.
Obs. 2. The reciprocat of the prod.uct of two matrices is tlw product of their reciprocals taken in' the reuerse
ord,eri.e. (AB)-1= 3-1 6-1 14sscrrn,1999)

If A, B be two matrices, then the reciprocal of their product is (AB)-1.

Clearly (AB) (r-Lq-r) = A(BB-L\A-L [by Associative lawl
'
= AIA -L = A/;Y -1.'
Similarly, ( B-LtrL). (AB) = /
Hence B-tA"l is the reciprocal of AB.
- Obs. 8. Multiplication by an inverse matrix plays the same role in matrix algebra that division plays
in ordinary algebra.
i.e.if tAltBl= tcltDl then tAl-lHllBl = [A-l]tqta
or B=^drtcllDl i.e.ryi =A-1[cltD].
lt I 3l
Exampl e 2.23.Find.the inuerseofl - 1 3 3I
-4 -4)
l-2
Sol. The determinant of the given matrixA is
I r 1 Bl l"t bL crl
A=l 1 3 - s l = l o z b z c z l( s a y )
-4 -41
i--2 1", b3 'al
If '41, Az,......be the co-factors in A, then At= -24, Az=- 8, Ag = - !2 i Bt = 10,
of a1,a'2.....
B Z = 2 ,B e = 6 ; C f = 2 , C Z = 2 ,C g = 2 .
Thus A = a,LAL+aAz+ a#s= - 8.
lat A2 Asl l-z+ -8 -1zl
Bz Bgl=l ro z
and,ad.ju^-lBt Gl.
L"' cz "rJ L 2 2 2)
Hence the inverse of the given matrix A

1r?i z -'il=fj jJ
-adjA-
=-1-=-8
, 2) 1',i La
-r _4)
Note. For other methodsseeExamples2.25 ;2.28 and 2.46.

1. IfA =f t9t 1 :i" 1.l,verify thatAA'= I =A'A,where/is the unit matrix.

L-smct coscl'
m,at*bri
2. Expresseachof the followingmatricesas the sumof a symmetricand a skew-s5rmmetric

(')[s -2'r 6lll

L3
,,,,['
I l]
LINEARALGEBRA: DFTERMINANTS.
MATRTCES 39
[r z sl
S. lfA = | 3 -Z 1 l, compute adj A andA-l. AIso veriff 1yr^191q-L= L
L4 2 lJ
4.If A is a non-singular matrix of order n, prove thatA adj A= | A I I. (Bombay,2006)

A) = (ad.j
verifyrhatA(ad.j A)A = | A I r,whereo= [l 3 ;.l
Ll 5 Lzl
Ir s B l I s -z 4l
5. Find the inverse of the matrix (l) 1
| 4 3l uill-z 1 1l @.P.T.U.,2005)
Ll 3 4 J L 4 1 0J
l-g -s 41
a . r r a = 1 2- B 4 l
l0 -1 1l
(i) findA-1 ; (ii) show thatAB =A-1.
.,[ L z 21
7.lfA=;l 2 1 -2l,provethatA-L=A'.
" l - 2.
rtl
2 -11
-sin 0l I r -tanetz]l
8. Showthatlcos0 = tan.a/Zlr
lsin 0 cosel ltan eD 1 Jl- ,tur-etz r _1
-1.l
g . l f A = 1[3r z 0 2landn=lZ
frool
| 0l,verifythat(AB),=B,A,,whereA,isthetransposeofA.
l+ b o.J [o 1 sj
[r'L21 [r z ol
l 9 ?l,B=12 B -ll,verifyrhat(AB)-1-B-rA-1.
r O . A =1
-1 3l
Ll 42) Ll
11. If A is a square matrix, show that (t) A + A' is symmetric, and (ii) A - A' is skew-symmetric.
P'T'U" 1999)
12.If D :diag fd1,d,,2,
d.s7,d.y d2,d,a* o,provethat D-l = diag ldr-r,d,r-r, dr-rl
13. If AandB are square matrices of the same order andA is symmetrical, show thatB'AB is also
symmetrical.
lHint. Show that (B'AB)' = B'ABI
14. If a non-singular matrixA is symmetric,show thatA-l is also symmetric.

2.r. (t) RANKOF A MATRTX

If we select any r rows and r columns from any matrix A, deleting all the other rows and
columns,then the determinant formed by these r x r elernentsis called the minor of A of ord,er
r. Clearly there will be a number of different minors of the sameorder, got by deleting different
rows and columns from the samematrix.
Def. A matrix is said to be of rank r when
(il it has at least one non-zerominor of order r,
a,nd,(ii) eueryminor of order higher than r uanishes.
Briefly, the ranh of a matria is the largestord.erof any non-uanishingminor of the matrix.
If a matrix has a non-zerominor of order r, its rank is > r.
Sall minors of a matrix of order r + 1 are zero,its rank is S r.
The ffi'cf,a matrixA shall be denoted by p("/.\.
 MATHEMATICS
HIGHERENGINEERING

(2) Elementary transformation of a matrix. The following operations, three of which

refer to rows and three to columns are known as elementary transformations :
I. The interchan,ge of any two rows (columns).
II. The multiplicatiort of any row (column) by a, non-zero number
III. The ,tdd,ition of a constant multiple of the elements of any row (column) to the
corcespond.ingelements of any other row (column).
Notation. The elementaryrow transformationswill be denotedby the following symbols:
\i) Riifor the interchangeof the ith andTth rows.
\iil kR; for multiplication of the ith row by k.
(iii) R; + pRj for addition to the lth row, p times theTth row'
The correspondingcolumn transformation will be denotedby writing C in placeof R.
Ele:mentary tranlsforrnatiuzsd,onot changeeither the ord.eror ranh of a matrir. While the value of the
minors *uy g"i changld by the transformationI and II, their zeroor non-zerocharacterremainsunaffected'
(3) Equivalent matrix. Two matrices A and.B are said.to be equiualent if one can be obtained
the
ft'am the 6ther by a sequenceof elernentary transformations. T\po equivalent matrices have
-
same order and the sa.merank. The symbol is used for equivalence.
Example 2.24. Determine the rank of the following matrices :

Ir 23.l lo 1 -s -rl
1 1l
u ) l14 2l t. .' L
. .1 r 0
)ls 1 o zl
tz 6 5l
Ll 1-2 0)
Sol. (l) Operate R2-.81 and Rs - 2R1 so that the given matrix
['r z Bl
- lo 2 -1 i=A (say;
l o z - 1|
Obviously, tt u 3rd order minor of A vanishes. Also its 2nd order minors formed by its 2nd
tt or
and Srcl rows are all zero.But another 2nd order: m i n o r i s l i _ i l = - L * 0 .

.'" p(A) = 2. Hence the rank of the given matrix is 2.

(il) Given matrix

[o 1 -s -1.l [o 1 -B -1j
- l1s1 0 0 0 l - l1r 1 0 0 0 l
1 -3 -11 o o ol
1 1 1 - 3 - 1-1 110 0 0l
ldperating Cs Ct, Cn- Cr] [Oierating Eg - Rr,ha- Br]
[0 1-s-1-l [o1ool
-16
11 0 0 0l 11 0 0 0l
-i6
; ; ol o o ;l=o(saY)
lo o o ol lo o o ol
lOperhting Rs - g&z,Ri - Ril E Ce+ dcr, Cn+ Cz)
[o'peratin
Obviorrsly, the 4th order minor ofA is zero. Also every 3rd order minor ofA is zero.-But, of
all t,he2ndorderminors, = - L + 0. - 2.
:. p(A1
""1, l?
I
ll
Hence the rank of the given matrix is 2.
(4) Etementary matrices. An elem.entarymatrix is that, which is obtuirt1ttfrom a unit
--'4
tnat rix, by subjectingit to any of the elementarytransformations.
 LINEAR.AL9
EBRA: DETERMINANTS,
MATRICES
4l
Exarnples of elementary matrices obtained from

frool lroot [r9gl [r p1 o]

13=10 olarenm=lg 0l
1
o rJ 9 1l=crr;ftn2=lo
.j ""^'-"Lo f 9l;nr+pR2=lo
,,.__ Lo
(5) Trreorem' lo 1l Lo o 1_l
Eleientary (columd't,ansfotrmatioits
,oi of a mitrix A can be obtained. by
pre-multilclying (post-multiplyingl A by the correspond,ing
elementa,,; ;";;;;r;;."

Consider
themarrixo =i:: I: :;j
1", br ',I
Then Rzsx"=|i
i il.|liriif |; ti iil
so a pre-multiplication by Rzs has interchanged the 2nd and'Brd rows of
A. similarlv
lre-multiplication by kR2 will multiply the 2nd row of Aby k and. pre-multipiication
r\ + pRz vrill result by
in thJaddition orj ti*us the 2nd row of A toirs Lst row.
Thus the pre-multiplication 4 by elementary matrices results in the corresponding
elernenta-rv row transformation ^of
ofA. It can *^' veDrr)
easily be
ue seen
see. f,nal
that post-multiplication will perfbrm
the elem-titury corumn
t"urrror*lti;;;.
( 6) G aus s ' oro rd ' a n
me th o d o f fi n d i ng the i nverse,r.. Those el ementary roLU
transforrrzq'tions
which red,u'cea giuen square m,atiix A to the unit
mqirix I giue the inuerse matrix, when applied, to ,rri't
of A.
' Let tkre successive row transformations
by the elernentary which reduce A to^I result from pre-multiplication
matrices RI, Rz, ... Riro ttrut
R i R i _ I ......R 2 R 1 /^_ I
Rr Ri- 1 ...... R2R1AA-L- IA-t
Or
R rR i - 1 ......R 2 R { = A -l '.' Aa-l -
[ /
Hence the result.
' worl-ing rule to evaluate A-1. write the two matrices
A and, I sid,eby side. Then perform
row transformations on both. As soonos
;:r:"*" A is red,ueedto I, the otier matrix represetzts

Exalmple 2'26' (Ising the Gauss-Jordan method,,

find the inuerse of the mah.ix

Il - zi ; 3l (Kuruhshetra,zoo6)
-4 -41
sol' \l\rriting the same matrii side by siae wilh the unit matrix
. l of order B, we have

I I i ii i i ill i (operate
Rz-R1
and
Rs+ ffir)
l_z _4 _4: 0 0

Narned after another German rnathematicianand geodesist

Wilhelm Jord.ang1g4z-rggg).
 1 B: 1 o ol
- 1f t0
2 -6: -1 1 0l (operate n31
-z 2: 2 0 1J L^runa ]
L0
[t 1 s: 1 o ol
-l' 1 -B:-++ "l (OPerateRt - Rzand'R3+ Rz)
-1 r: 1 o +l
Lr
[' o 6: Z -+ "l
-l' 1 -B:-+ + "l R1+3R3,R2-Z*r*u *).4

L'o -z:+ + +l
[onur*r" [-

o o: B t ;-l
ft
-l' 1 o: -i -i -il
L' o 1: -i -i -1.]
tql
13 1 ;l
- - .l-ql - + -1
Hence the inverse of the given matrix is -4 --qli l [cf. Example 2.20

l_r _i _i!
(?) Normat form of a matrix. Euery :r:-".: bv a
*";:) A of ranh r, ca,nbe red,uced'
sequenceof elementary transformations, to the forrn
"'(i)
it |f'"'*othe 4svmalformofA' '
Cor. 1. The ranh of a matrix A is r if and orzly if it gsn (i).
be rcduced,to th,enonnal form
Cot, 2. Since each elementary transforioiion or post-
can be A, pre-mul.tipl.ication we
multiplication with a suitable elerneniary matrix, iri "lSrirra i*is non-rirrgular, therefbre'
elementaty it
Iwue the following result : "o"n
PAQ
Corri,rrporrding to euery matrix A of ranh r, theree*isf
non-sittgular matrices P and. Q such thot
equals (i).
IfA be a m xn matrix, then P and Q are s![\sre rrratrices of orders m and n respectively. --^L
Example 2.26. Reduce the following rnqlrix its rut'*'
irtto its normal form. and hence ft.nd
l2 3 -7 -/l
tr -1 -Z '
o=l; I s -il (U.p,T.t1.,2005\

L6 s o _?l
-1 -z -41
[t
Sol.
"-i3 i-3
3 o
_;1
-7
lby ftrz
L6 I
 LlllEAR ALGEBRA: DEIERMTNANTS,
MATRTCES

ir -1 -z -4
- l lo0 5 3 7
hy Rz - 2R1,Rs - 3.R1,.Ra- 6^Er.
4 9 10
loeLzLT
[r o o ol
- l oo 5 3 7 l
l
LbyC2 * Cr, Cs + 2C1,Ca+ 4C1
4 e 10l
L0 e L2 L7l
[r o o ol
- iloo 5 3 7l
4 e 1ol [byRa-Rz-Re
Lo o o oj
[r o o ol
-6 -3 1
- l oo I
l
Iby R2-.Rg
4 e lol
L0 0 0 0J
[r o -6
o ol
-3 1
- lloo 1 lby ^R3- 4Rz
o Bs zzl
lo o o ol
l-r o o ol
- l o0 1
l 0 0l
lby Cs + 6C2,Ca+ 3C2
o Bs zzl
lo o o oj
[r o o ol
- llo0 1 o 0l
ruv$ c,
o 1 zzl
lo o o oJ
[r o o ol
_lo 1 o ol lby Ca- 22 CB
l0 0 1 0l
Lo o 0 oJ
-V, ol
0l
L0
Hence p(A) = 3.
tl,
Example2.27,Forthematrixo =lt, I
Lo -1 -rJ

)
pe shall affect
-everyelementary row (column) transformation of the product by subjectrng
the prefactor (post factor) ofA to the same
44 HIGHERENGINEERINGMATHEMATICS

0 ol[rool[t -1 -21
operateC2- Ct,Cs-ZCt f l 1 1l=10 1 0lAl0 1 ol
-1 -1.] Lo o 1l Lo 0 1l
L0
Ir 0 o l I r 1 o l [ t -1 -21
Rz- Rt,
Operate t 1l=l-1 o olAlo 1 01.
-1 - 1 1 rl
LS Lo o lJ Lo 0
t-1 o o l I r o o l [ r -1' -1.l
OperateCg- C2, 0 1 0l=l-1 L 0lAl0 1 -1
I0 |
L 1 ol Io o 1J io 0 1j

fr o1 ol Ir o ol fr .'-1 -1-l
OperateB3 * Rz, l0 ol=l-r 1 olalo I -1l.
Lo o ol L-1 1 lJ Lo 0 1j
't
0
which is of the normut ---^^^
ro.* [ ? I
L0 0j
Ir o ol fr -1 -rl
H e n c ep = l _ 1 1 o l , e = l 0 1 -1 and p(A; = 2.
|
L-l 1 lJ L0 0 1J

Determine the rank of the following matrices (1-4) :

[r 4 sl Is -1 z]
1 ,1 2 6 8l (P.T.U.,2005) 2 . 1- 6 2 4l
L3 7 22) L-3 L 2)
[r2sol
[r B
s.ls e rz sl
4 rJ
4 sl
(w,8.7.U.,2005) ^13 ; i 3l (Kottayam,2005)

IrB 16875)

o.

6. Use Gauss-Jordan method to find the inverse of the following matrices:

lz 1 -1-l [e 4 Bl
( r ) 1 02 1 l oitlz 1 1I
-3J
L5 2 Ll 2 lJ
[-r 2 z] [o1zl
liii)l 2 -1 2l @.P.T.U.,2006) riurll 2 sl (Kuruhshetra, 2006 .\

L2 2 *1J, 1311J
[s -B +1
Z. tt A=12 { +lfinaA-1. Also lind two non-singularmatricesPand Q such that PAQ =1, rvhere
lo -1
L J
rl
.I is the unit matrix and verify ihat A-l = QP.
8. Reduce eaclr of the following matrices to normal form and hence find their ranks :
I s 1 B 6l lz 1 -t -6-l
(t)l 0 3 (li)A=lS -g
2 2l (Kurukshetra,2005) I 2l (u.P.T',.U.,
2003\
-1 -s +l 1 1 z)
f-a [t
 LINEAR
,qLGEBRA
: DETERMINANiS"
MATRICES

Ir-7 -2 *3-J
102l|
, . . . l.4
vzz)
l03 I 4l (Bombay,2006)
L0 1 0 2 1
9. Find non -singular matrices p and
I such that PAQ is in the normal form for the matrices :
-1
fr -1.l
(i)A=11 fr 2 B -2)
1 1l (J .N .7 .U .,2 A 0 2 ) ( i i ) A = 1 2- z 1 s l
Ls 1lJ
L3 0 4 1l
2'9.PARTITION
METHOD
OF FINDING
THETNVERSE
^*::*t"?,to.this method of finding the inverse, if the inverse of a matrix A,,
, oforder n is
knordn, then the inverse'of the matrix An
+ r can easily be obtained by adding (n + 1)th row and
(n + l)th column to A,r.

i . e .[o' ; Arl[& , *ri fr- ot

T h e n A A1 = I t t +jt ^
1 , . . . , . . .1 1 , - . : . . .1 = l T ;rJ
l
LAr' r 6xil"r' : LU
girres "J
ArXr*A2[s'-1,,
ArXz+Ag =Q
Ag'Xr * uX3'= Q
Ag'XZ*Gx=l-
From (ii), Xz- -41 'Ar* and using this,
fiu) givesx = (d- As, A1L Ar1-l
Hencer and then X2 aregiven.
Also from (t),Xt =A1.r (In - AzXs,)
and using this, (lli) giyss XB' - - Ae,Ait (o - Ar,
Air Az)-t
Then Xl is determinedand henceA- I is computed.
obs' This is also known as the 'Escalator method'.
For evaluation ofA- 1 we only need to determine tlv.
inverse matrices .A1 I and (a - As, Art
Ar,t- t.

Exampl e 2'28.Llsing the partition method,, 1/l

find the inuers" ,flto 3 - 11.
l,l 531
1
Sol. Let e-l|
ft
4 3 _1l[o, Arl
-l
t...
, l=1o,,crJI
r so that , r
.t'L , -=r
L3 5
[
Ln
r 1
BJ
l * 1 I
J-

-l-a -1.l
s
1l
LV'
 HIGHERENGINEERINGM ATFIEMATICS
46

Let - r=[]i. , "l sothar AA-r

=r.
LXu' i ,cj __
A1L
u- As, B+rB rr i
Az= - to
[- ilt lJ=
x=(cr-AB'AltAz)-t=-*

A,so xz=-AlLo*=l-'u
t]t ljf ,|j= #t ll
rhen 1r=13rr
x3,=-As'A1 21
[-i l]t *)=-+r-11
-l]-#[ zl
Finaily xt=Al,o-Azxs,)=-[_i ? l][ l] r"
L s4 - 1rl1 - 110l-Ld+
=t-- Bl- |- L'4 o'2.l
5 5 - 1 0 . j - L - 1 ' 5o J
I t.+ o.z -o+l
Hence A- L-l - r'r o o'5
t . t - 0 . 2 - o 1 Il
L
Example 2.2e.tle aria ,\u" showthat
I o*0", it*Trr'rnatrices,
I a c J = L - c - 1 B A - tc _ t )
[l o o ol
', (BombaY,2005)
find,inuert,
Hence *lO, 2 SI
Lo 1 0 3J
the
sol. Let the givenmatrixbe M=[f 3]."u
its inversebelurt=L; 3 ] o",nin
partioned form where A, B, C, P, Q, n, S are all matrices'
Mrr,=f4qli{ g'l=r
ciLl? sJ-
L^o
or '"]
i;|rss]=lf
w,:"&
.'. Equating correspondingelements,we have
'\P+OR=1, AQ+OS=0, BP+CR=0, BQ+CS=/'
Secondrelation givesAQ =0 i"e' Q = o asA is non-singular'
1.
First relation givesAP -- I i.e. P =A-
- t
From third equation,BP + CR = 0 i.e- Ci?- - BP = BA-
'o,
'. C-LCR=-e-LB.{l IR=-C-tBA-l or R =-C_LB/'r
F r o m f o u r t h e q u a t i o n , B Q + C S = 1 o, r C S = I o r S = C - 1
IL u'-l
,r':=
H e n cl e l L-
c-LBA-t -tJ
c
 LINEAR
ALGEBRA: DETERMINANTS,
MATRICES
47

ui,Le,,=fi L sl
i il]=t*
Lo 1 o sJ
whence"=[l
g], i l], f; S]
"=f "=f
: A-1-
+lzl],"_l=#[s
l]
c-'@A-')=-#fs
lli'+[s
?]f3ll]
=-#[3
t][st]=#fi' 4o Jl
o
Hencer=l[ t
114- ,(,
Jrn ,3r^ I I
o _t/6
L o vtl

Find the inverse of each of the following

matrices u sing the partition rnethod.
fr Bgl
t.l1 4 sl (Nagpur, lgg7) 2 .[1s 2
z+1
r rl
ir B+l IrBsJ
fr2srl
s B zl fs -t to zl
." q' l1.z1 4
Ll
s sl
11lJ "13 +33
-2
tl
z tj
Ll
5'IrA andc;"
";''1*Tir"are matrices
orordern,
show*n*lf 3 =f_
I' J;l_, f, ]
u"u ',
Hence
l! 4 2 o
SI @ombay,
zooz
s)
lo zl
2.I0. SOLUTION OF LINEARSYSTEM
OF EAUATIONS
(1) llfiethod
of determinants - Cramer's+ rule
Consider the equationsdfr + bg + ctz *drl
a*c+bfl+czz=dzl
...(r)
a4xc* bsy+ tg =dr)

ia, bt
If the determinantof coefficients
beA=loi l-
bz ",rl
tl
la,
l"
bs rgl

* Go,b,iel crcmar oTa+L?62), a swirs mathematieisr.

48 EEI?INGTVlATHEMATICS
HIGHERENGIITI

xat bt 11l
then x L = Xdq b, cel loPerql. C1+yC2+ zCsT
*o; b;
";l
a1x,*bgr+cp b1 b1 C1
ayx+b,nt+c2z b2 t l = l d t bz
"crl c2 lby (t)]
la,
agx*bg+egz bs bs ca
"rl lar
l{' b} .' I l"' bl crI
Thus * = b2 - bz c2| providedA + 0. ...(tt)
lo., "rl lo,
o3 csl q cal
lor los
l"t dr "tl + l"t 6r,rl
Similarly, y = dz c2l b2 trl ...i.ii)
loz la2
l"t dr ruI l"r 6; ,;l
l"t bL atl l"t b 1 c ri
and z= bz d.rl* b; ';l ...(lu)
ldz 1",
lor b3 asl Pt 63 r,I
Equa.tions (ii'), (ili) and (iu) giuing t?u uolues of x, !, z constitu,tethe Cramer's rule, which
reduces the solution of the linear equations (i) to a problem in evaluation of determinants.
(2) Matrix inversion metho_d

rf "=l:: lil,*=Flqnd
D=f3l
L"r bB L'l laoI
"; "J
then the equations (Ij are equival6nt to ths
where A is the coefficient mutrix.
matrix b#",to' AX = D . . . ( )u

Multiplying both sides of (u) by the matrix A-1, we get

".eipro.al
A-r At( =,{7D 01 I x = a-rD y -.. A-rA=11

ie, kl t [t' 42 o,l lol

z:"tl.l::l
or X -,q,*LD ...(ui)
L:j=iL,"l
lor b1 "tl
whereAt, Bt euc.are the cofactorsof 41, 6l etc.in the determinantn:loz be
- - l @ ,+ 0)
crl
las bs "rl
Hence equating the values of.r, y.,z to the corresponding elements in the product on the right
side of (ul), we get the desired solutions.
Obs. When A is a singular tnatrix, i.e. L': q the aboue methods number of'
fail. These alsofait when the
equations and the number of u.nknotuns dt'e uneoual. Matrices ccLn,howeuer, be usefully applied. to deal witlz
t
such equations as will be seer, in # 2 11(2).
E x a m p l e 2 ' 3 0S. o l u e t h e e q u a t i o3nqs. + v + z z = 8 , 2 x - B y - z - - B , x + 2 y+ z : 4
hy determinants (ii) matrices.
(i)
(l) Sol. by deterrninants
ls L zl
Hereo= l? - I - 3 +2) _- 20-4) + (- 1 + 6) - g lExpandingby
1l= 3(- C1l
1
tl
1 2 1l
LINEARALGEBRA: DETTRMTNANTS,
MATRICES
.i

.,13 L2l
,=i l-B - 3 - 1l [Expandbv Cr]
^l 4 2 1l
=* tB(-B + 2) +B(1- 4) + 4(* 1 + 6)l = 1

3zl , ls 1 sl
simirarly
, !=i- ll - 3 - 7 l .- z and , - - *^l z - g -sl
11 4 1[ lt z 4i
HenCe X,=try=2,2=-L.

Note. The use of Cramer's rule involves a lot of labour when the number of equations exceeds fbur. In
such and other cases, the numerical methods given in $ 28'4 to 28.6 are preferable.
(ii) Sol. by Matrices :

ls 1 2l l"t b1 tl
Herel= l2 _B -11= la2 b2 ",rl (say).
11 2 11 1", bs ', I
T h e n A r = - L , A z - 3 , A g = 5 ; B r - - 3 , B z = 1 , B a= 7 ; C t - ' 1 , C 2 - - 5 , C g = - 1 1 .
Also A- atAr+ a#z+a3A, = 3.

H e n C ex = L r y = 2 r Z = - L .
E x a m p l e 2 . 3 1 . S o l u et h e e q u a t i o n st c l - x z * t r 3 * x 4 = 2 i x t * x z - ) c a * x 4 = - 4 ; x1*x2*
rg-x4=4;x1*x21tcgI1E4=0,byfi,p:dingtheinuersebyelemetttaryrowoperq,tians.
Sol. Given system can be written asAX=.B, where

"{l il"=Lll] rl 21
-1
11
1 ["'l -41
tL 4l
1"1 0_l
To find A- 1, we write

l1-1 1 1 I o o ol
IR2 - R1
L { ;""-_l l I1 1-1
1
1 0 1 0 0l
0 1 0l
1-1 0 lR3+.R1
lLr 1 1 1 0 o o rJ lBa + R1

[r -1 1 1 I o o ol
_ l
-l|20 2 -2 0 -1 L o o l E;nz
0 2 o 1 o 1 o l r*o,
0 2 2 1 o o 1.1
L2 Iion
 ENGINEERING
HiG}'IER

Ir -1 1 1 100 ol [R1+R2
1 *L 0 - L/2 L/2 0 0l
=l[0
r o 1 0 L/2 0 L/2 0l
Ll 0 1 1 1/2 0 01 /2 1 lBa - Rs
Ir o 0 I : l/2 L/2 0 ol lBl- R4
1
=ll or o -1 0 : - L/2 L/2 0 ol lR2+ Rs
1 0 :t/20 t/2 ol
'10 0 0 1 :00 - t/2 /2)

Ir o 00 l/2 Llz + t/2 trzl

o | [82-R1
=llr o1 00 o L/2 r/2
1 10 L/2 0 L/2 o | [R3-R1
ttz )
L0 0 01 o0 r/2
Ir o 00 L/2 t/2 t/2 - trzl
1
=ll oo o 00 -L/2 0 0 r/2 |
10 0 -L/2 0 L/2 |
- L/2 r/2 )
L0 0 01 00

I trz l/2 1/2 - trzl

=l| -L/2 00 ttz I
Thus, A-1
o -Ltz 0 vzl
0 -t/2 L/2
L0 )
- trzlt zl
X=A-'" = - t / 20
L/2 l/2 L/2
0 -i f _il
Hence, -t/2
0
-L/2
0 ir?ll l=rzI
trzll ol
L 0 0 L-rl
L.e. ffl = 1, XZ=- L, xg=Z, )C4--2.

Soive the lbllou'ing equations with the help of determinants (1 to 4) :

1. r +y + z=4;x -y +e=0 ; 2x+y *z =5. (Os
2 " x + 3 y + 6 2 = 2 ; 3 x ! r 4 z = 9 i x - 4 y+ 2 2 = 7 .
S . J r+ y + e = 6 . 6 ; x - y * z = 2 . 2 ; x + 2 y + 3 z = L 5 ' 2 .
4. xzzS/y = uB ; y2z/, = ; xgy/24 = L.
"4
6. Zuut - LUa+ uu = Suuw ; \uw + 2wu + Auu = L9uuw ; 6uw + 1wu - ttu = l7uuw.
Solve the fcrllowing system of equations by matrix method (6 to 8) :
6. 11 -l-x21.lcg= 1, rt + 2x2+ }xg= 6, xt+ }xz+ 4xt= 6-
7, x -r 3, * z = 3 ; x + 2y + 3z = 4; x + 4y + 9z = 6. . (E-
8. 2x-- 3], + 42 - - 4, x + z = 0, - y + 4z =2. (w-
9 . 2 x * y + 3 z = 8 ; * - 2 y - z = - 4 ; 3 x + y - 4 z= 0 . (E'
lO. 2x1* )c2* 2xg+ x4 = 6, 4xt + 3x2+ 3"r - 3x4=- 1, 611- 6*z+ 613+ 72xa= 36, Zxt + 2x2 -
(u'
A)(=.B,
equation
thelinear
11. By frndingA-1, solve ^=
where ," = """ , =
|J il]
LJ li] [
12. In a given electrical network, the equations for the currents iy i2, i3 are
3 \ + i 2 + i , = I ; 2 i 1 - } i z - 2 i s = ' 5 ; 7 i 1 + 2 i 2 ' 5 i 3= g .
Calculate il and is by Cramer'srule.
 LINEARALGEBRA: DET€RMINANTS,
MAIRICES 49

rl3 121
"=;l_1 -3 -tl [ExpandbyCr]
2 1l
= ] t s 1 - e+ 2) + 3(1- 4) + 4(- 1+ 6)l = 1

r l3 3 2l , 1 3 - 31 - 331l
Similarly, ,=il? 4
-3 -Ll=Z
1l
and
" = *^1 1
21 24i
=*t

HenCe X=l,y=2,2=*1.

Note. The use of Cramer's rule involves a lot of labour when the number of equations exceeds fbur. In
such and other cases, the numerical methods given in $ 28.4 to 28.6 are preferable.
(ii) Sol. by Matrices :

ls L 2l l"r b1 rrl
Hereo = |, -B -11= l"r b2 ,rl (say).
11 z 11 1", h ,el
L.r Then At=-L,AZ':3,Ag -5;Bt=-3, B Z =1 , B g= 7 ; C t = 7 , C 2 - - 5 , C g * - 1 1 .
Also A - atAt + afuz+ a,sAt= g.

t: ^'1.ljl
fl=+Ei
^
L"l L"' cz Lo'J "'J
,[-1 B sl l- gl ,[-s-e+2ol l-rl
=tl-J - b
L/
Hence tc,= L, y = Z, z -- - I.
E x a m p l e 2 . 3 1. S o l u et h e e q u a . t i o n sx l - x 1 * f 3 * x4=2 i xt*?c2-fi1*x4=-4 ; x1 *x2*
xB - x4 =4 ; xt * !r2 * xg * xc4= 0, by fipd,ing the inuerse by eletnentary row operations.
Sol. Given system can be written as/X= B, where

1 1 1l t' zl
A=l [r
i i i-il'*=l:?[,"=l
ll
["'l

L1 1 1 lJ L*J L oJ
To find A- 1, we write
-1 1 1 1 o o ol
[r
1 0 L 0 0l lEz - Rr
n=l 1 1-1
r 1-1 0 o 1 ol [ft3 +.R1
lr
Ll 1 1 I 0 0 0 1 l [Ra + R1

[t -1 1 ]" 1 o o ol 1

IiR,
lo z -z o -1 1 0 0l
12 0 2 0 1 0 1 0l IF'
0 22 1 0 0 1J
12 Iln^
 50 MATHEMATICS
ENGINEERING
HiGHER

-1 ol l R 1 +R 2
flIr 0
1 1:

l.0: L/2
1 0
L -1 0: -L/2 L/2
0
0
0
7/2
0l

L1 0 L 1: L/2 0 0 t,3) lBq- Rs

Ir
lo
l1
0

0
0 1:
1 -1 0:-L/2
10:
l/2 L/2

L/2
L/2
0 t/2
0
0
ol
[Rt - Rs,
lR2 + Rs sl
0 01: 0 0 * t/2 trz
lo )
Ir
Ir
l1
000
100
010
l/2 L/2 + L/2 - L / 2

L/2
0 L/2
0
L/2
L/2
0
0 tE3-ftl
I *,-.,
Lo 001 o 0 -L/2 L/2
I
Ir 000 l/2 L/2 t/2 trzf
100 -L/2 0
lo 0 L/2 |
l0 010 0 -L/2 0 L/2 |
-L/2
Io 001" 0 0 L/2
)
r
I r/2 L/2 1/ z - ttzf
t L/2 0 0 t/2 1
Thus, A-1
t- 0 -L/2 o vzl
L 0 0 - 1 /2 1/2
)
"r L/2 L/2 L/z -rtzll zl I rl
t/2 0 0 L/2 lt-4 | l-11
H e n c ex,= A - ' r = [ -
0 -t/2 o trzll al=l zl
I

L 00 r/2 r/2)L o)L-2)

i.e. f1=1, x(Z=-1, *3=2, X4=-1,.

solve the lbllowing equationswith the help of determinants (1 to 4) :

1 . ) c+ y + z = 4 : x - y + z = 0 ; 2 x + y * z = 5 . (Osmania,2003)
2" x + 3y + 6z = 2 ; 3x - y + A = 9 ; x - 4y + 2z =7.
3 . x + y + z = 6 . 6 ; x - y * z = 2 . 2 ; x + 2 y+ 3 2 = 1 5 . 2 .
4. x2t3,/y = e8 ; yLz/x = e4 ; *3y/24 = 1.
5. Zuw - Luu+ uu = Suuw; Suw+ Zwu + 4uu = l9uuw ; 6uw+ 7wu - uu = r7uuw.
Solve the following system of equationsby matrix method (Gto 8) :
6 . t 1 - + . r g * J=r 31 , 1 1+ 2 x 2 + 3 r g = 6 ,1 1 + 3 r z + 4 x 3 = 6 . (P .7.U .,2006)
7. x -ry + z = 3 ; x+ 2y + 3z = 4; x+ 4y + 9z = 6. . (Bhopal,2002)
8 . 2 x - B l , + " 4 2 + - 4 , x + z = 0 ,- y + 4 2 = 2 . (w.8.7.U., 2005)
9 . 2 : r- y + . 3 2= 8 ; r - 2 y - z = - 4 ; 3 r + y - 4 z = 0 . (BombaY,2005)
l O. 2x 1+x z + 2x g+ x 4 =6 ,4 4 + 3 x 2 + 3 t3 -3 x 4 -- 1 ,6rr - 6x2+6* 3+ L2xn= 36,2x1+ 2x2-nB * x+ = L0.
@ .P .T.U .,2001)

I1. 83' IindingA-1 , sotve

thelinearequation
Ax--B, o =[i ; 3.l,* =lil *u =i-i l
where
L 51 1 1 V l " L 5 l
12. In a given electricalnetwork,the equationsfor the currents i1, i2,13are
3 i 1 + i r + i a - 8 ; 2 i 1 - 3 , i z -2 i 3= - 5 ; 7 i y + 2 i 2 - 5 i r = 9 .
Calculate i1 and 13by Cramer'srule.
 5l
- LINEAR ALGEBRA.DETERMINAMS. MATRICES
'
the following equations are obtained:
i'i. Urirrg theloop current method on a circuit,
- = Q'
lir- +ir- !2,- 4\+ t2i2- 6ie o,- 6i2+ 14is
By matrix method', solve for i1, i2 and i3'
the inverse by elementary row operations :
14. solve the following equations by calculating
+)c2-3ts- 4'4=- 1;2rr rx2* 5x3+ x4=5'
z x y + 1 4 o z + b B - * g ; o =2 ; 3 x 1 + 6 x 2 - 2 * s * x 4 = 8 ; r r I

2.11.(l)coNS|sTENcYoFLINEARsYsTEMoFEaUATIoNs
Consider the system of m linear equations
ar';ct* a12x2+"' + drilcn- ft1.|
a z frt+ a z * z + " ' + &znxn- k,l ...(t)

; ;;:;, ; : ; ;;:;;;= ;;)

";;;;;
whether the equations (i) are consistent
containing the n unknowrs f 1rfi2t ...t*rr. To determine
(i.e. possess a solution) or not, we consider the ranks of the matrices
an1 arz & Ln U tl
[ o1 aLz [o1
azz azz ::' &2n u'l
^ -l;;, ";:,1 and x =lo^
I I @m2
I amn 7t* ;:l
dm1 o*n) lo*t )
Lo*r
matrix of the equations (i).
augrnented,IIL
$e dugrnerltvlt
A is the co'fficient matri"tcirnd K is called t11e
and only if the coefficien't
e)Rouche,s theore m. The systemof equations (i) is consisten.tif
rank otherwise the system is inco.nsistent'
matrix A and. the augmented, matiix K are oitt " sa,nte
:
Proof. we consider the following two possible cases :
m and n)r The uqoutio"s (i) can,
I. Rank of A = ranh of K,=r (r ( the smaller of the numbers
by suitable row operations, be reduced to
b t p r + b r * z + . . . . .+. b v x n - 1 1 )
0.r1 + bZZrcZ+ .""' * b2nttn- lZl ...(tt)
r
o;;; o:;;;: : ; ;,;;,.;;l,i
= g'
and the remainin g nx -r equations being all of the form 0.r1 + 0'r2 + """ + 0'*n
- of ,the unknowns may be chosen
The equations (ii) will have a solution, though n r
only whert . = n. Hence the equations (l) are consistent'
arbitrarily. The solution will be unique
rank of K be r + 1' In this case' the
IL Rank of A (i.e. r) < ranh of K.In particular, let the
equations (i) will reduce, by suitable row operations, to
b n x t + Q t z x z +" " " + b Y r x , r - 1 1 ,
0.q /b22x2 * ......* b2nxn= 12,

0.r1 + 0.x2+ """ + b,'rxcr=1,',

0 . * 1+ O . x 2 + . . . . .+. O ' s r =
, I, +r, "r'
-
0'r1 + }'xz + """ + 0'r" 0'
and the remainin g nx -(r + 1) equations are of the form
by any set of values for the unknowns'
clearly the (r + 1)th equation cannot be satisfied
Hence the equations (i) are inconsistent'
 HIGHER MATHEMATICS
ENGINEERING

lProcedure to test the consistency of a system of equations'in n unknowns :

Find tlte ra.nksof the cofficient m,atrixA and the augmentedmatri.x,K, by reducingA to the
triangular form by eiementaiy row olic nn*-t egthe rank ofA be r and that ofK be r'.
(t) If r *.r', the equationsare inconsistent,i.e. there is no solution.
(ti) If r = r' = n, the equationsare consistentand thereis a uniquesolution.
(rtt) If r -- r' 1 rt,the equationsare consistentand,thereare infinite number of solutions.Giving
arbitrary values to n - r of the unknowns, we may expressthe other r unknowns in terms of
these.I
Example 2.32. Testfor consistencyand solue
5x + 3y + 7z =4, 3x + 26y + 2z =9, 7x + 2y + 1.02- 5.

Sol. We have

lrs e 21-l[.'llrzl
LT"3 lSll:l=L-gj
Operate3.R1,5h2,

Operate Rz- Rt, rz?L

['l iilp1=[*il
7
Operaie Rr, 5Rs,
; i*, [:lii ;3]E]=t;ll
1[sB
OperateRs-ftr + Rz, iBt, I O 11
'100
ilE]
Lil
The ranks of coeffrcient matrix and augmented matrix for the last set of equations,
2. Hence the eqgations are consistent. Also the given system is equivalent to
- are both

5 x + 3 y+ 7 2 = 4 , L L y- z = 3 , 3zaf16
Y=1T*rr ancl'r=11 -nz

whereeisaparameter.

3 ^-r
Hence *=! y. ' -= and zz = 00,. i is_a
s a oparticular
artict solution. I

n, 11 I

Obs. In the above solution, the coefficient matrix is reduced to an upper triangular matrix by
row-transforrnations.
Example 2.33. Inuestigatr"in-rlctly4_e!-?v
and,1tso that the equations
2x + 3y + 5z = 9, 7x + Jy - 2z = 8, Zx + Jy + 7s = F,
i
haue(i) na solwtron,(ii) a unique solution and. (iii) an infinite number.ol solutions.
(Ranclzi,2000)
lz s bl l-'l l-gl
Sol.Wehave
IZ s -zllyl=ls I
lz s rJL"jLpl
 MATRICE$
LINEARALGEBRA: DETERMINANTS.
53
3' This
The system admits of unique solution if, and only if, the coefficient matrix is of rank
requires that
bl
z B -21=
l',7
3 - r) * o
15(5
12 3 r l
Thus for a unique solution l, * 5 and p may ha'e any value. If l, = 5, the system will have'no
solution for those lrllo", of p for which the matrices
lz s - 2bl lz 3 o el
A =lx 3 | a n dK = | 7 . - -
_s -2 8l
lz s 5J Lz-ts D FIJ
= L = 5 and
are not of the same rank. ButA is of rank 2 andKis not of rank 2-unless F 9. Thus if
p * 9, the system will have no solution.
If l": 5 and F = 9, the system will have an infinite number of solutions.
(g) System of linear homogeneous equ.ations. Consider the homogeneous linear
equations
e L tX t + a t* Z + .....+ o P* ,, = 0' l
a 1 ttc l +a Z * Z + .....-+ tt2 n tc ' n= 01 ...liii\

drn1fl* am*Z+
t
. ' . . . .+ a * n x n = 0 )

Find the rank r of the coefficient matrixA by reducing it to the triangular form by elementary
row operations.
I. If r = n, the equations (iii) haue only a triuial zero solution
X l = 2 C 2 = . . . . . .= X n = 0

If , < n, the equations ftii) haue (n - r) Iinearly independent solutions.

The number of linearly independent solutions is (n * r) means, if arbitrary value_s are
assigned to (rr.- r) of th" nu"i^bl"r, th* values of the remaining variables can be uniquely found.
Thus the equations (ill) will have an infinite number of solutions.
IL When nt,<n (i.e. the number of equations is less than the number nf variahles), the
solution is always other than xj.= xz= ...... = x.n= 0. The number of iolutions is infinite-
IIl. When trl =tx,(i.e. the number of equatio[s = the number of variables), tlw necessary and
sufficient cond,ition'foi" solutions other than )cL= xz3 ...... = xtt= 0, is that the deterrminanttof the
coefficient matrix is zero.,In this case the equations are said to be consistent and such a solution
is called non-trir4ial so,Iution.The determinant is called the eliminant of the eguations.
Exarnple 2'34.Soluethe equations
(i) x + 2y,+ 3z = 0, 3x + 4y + 4z = 0, 7x + L0y + 722= 0 (Osmania, 1999\
(ii) 4x + 2y + z + \rD = 0, 6x+3y + 4z + 7w =0, 2x +y + w = 0.
Sol. (i ) Rank of the coefficientmatrix
fr z ft s'l z Bl
13 4 4l-lo -2 -51 [Operating ^83- 3R1

[z 10 12] L? 10 12J
- 1[rz
0 -2 3l
1l
[Operating fts - 7R1- ?.R2
100
is 3 which - the number of variablOs(;.n.r -- n)
54 HIGHER MATHEMATICS
ENGINEERING

.'. The equations have only a trivial solution : x =! =z =0.

(ii) Rank of the coefficient matrix

l+ z 1B'l l+ z 1 Bl R 1
l9 '1 0 1ll-19 9 5/2 E/?l [operatinsaz-"zar,.R3-i^t
-L/2 -L/21
L2 Lo 0

lo o o ol
is 2 which < the number of variable G.e.r <n)
"'. Number of independentsolutions - 4-2=2. Given system is equivalent to
Lr+2y+z+3w=0, z+w=0.
.'.We havez =-rp andy * - 2.x-w
which give an infinite number of non-trivial solutions,r and ru being the parameters.
Example 2.35. Find the ualues of k for which the system of equations
( 3 k - 8 ) r + 3 y + 3 2 = 0 , 3 x + ( 3 k* S ) y + 3 2 = 0 , 3 x + 3 y + ( 3 h - 8 ) z - 0 h a s u n o n - t r i u i a l s o l u t i o n , .
(U.P.T.U.,2006)
Sol. For the given systemof equations to have a non-trivial solution, the determinant of the
coefficientmatrix should be zero.
3.&-8 3 3 | l3k-2 3 3 |
L.e. 3 3ft-8 3 l=OorlSk-Z 3k-8 3 l=0
3 3 3h-81 3 3,k-Bl ,
l3h-2
[Operating C1 + ( C 2 +cs)
Ir B s I 11 B B I
(3e-2)113&-8 3 l=oor(3k-2) lO lk-tL 0 l=0
11 s Bh-sl lo o Be-111
loperatin g Rz - .R1,Ra R 1
(3,t* 2) (Sk- 11)2= O whence k = 2/3, LL/g, LL/g.
Example 2'36. If the following system has non-triuial solution, proue that a + b + c = 0 o r
a=b =c : q,x+ by + cz = 0, bx + cy + az = O, cx + q,y+ bz = 0. (Bombay,2006)
'
Sol. For the given system of equations to have non-trivial solution,'the determinant of the
coefficient matrix is zero.
abc a+b+c a+b+c a+b+c
LC, bca. =0 or bca, -0
c ab cab
lOperating ftr + R2 + Rs
111 L 0
(a+b+c) bca =O or (a+b+c) b c*b a-bl=0
cab c a-c b-c
[OperatineCz- C1,Cs- C1
or ' (a +b +c) [(c-b) (b -c) - (a-c)(a-b)] =O
or s ( a + b + c ) ( - o 2- b 2 - 1 2 + a b + b c * c o ) = Q
i.e. a+b*c=0 or a2+bz +12-ab-bc-ca=0

Dr a+b*c=0 o,rlllr*-b)z+(b-c)2+(c-ayzl=g
or a+b +c=0; a=b, b:c, c=a.
Hencethegiven systemhasanon-trivial solutionifa +b +c=0 or a=b.=c.
LINEARALGEBRA: DETERMINANTS,
MATRICES 55

Example 2.37. Find the valuesof )ufor which the equa'tions

(1,- l)r + (3),+ l)y + 27,2- 0
( 1 , - / ) x + ( 4 ) , - 2 ) y+ ( 1 , +3 ) z - 0
2x + (37u+ l)y + 3(1"- 1) z - 0
areconsistent,a.ndfindtheratiosofx:y:zwhen\,"!tasthesmallestoftheseualues.Whathappens
when )yhas th,egreater of these ualues. (Kurukshetra,2006 ; Del.lt'i,2002\
Sol. The given equations will be consistent, if
I-1 31,+1 2 7 , "I
l"-1 4L-2 l.+3 [=O [Operate Rz- Rt
2 31"+1 3(I-1)l
I-1 31"+1 2 7 , "I
or if, o r.-3 3-r. l=o [Operate Cs+ C2
2 3)"+1 3(I-1)l
l"-1 31.+1 5i.+11
orin 0 l"-3 0 l=0 lExpand by Rz
I 2 31"+1 6),-21
r^ '
or if, ) [n: t
( t - g -' 5i,+1 I ^
-2(31, , 1 = 0u or
? * * +11)l o r rr' 2 ( t "-'r)
i f , z\tL t ( 1 . -1 X 3 ] , -1 ) * ( 5 I + t ) l = 0
- g ) t(^-
I 2
or if, 6t(I-3)2=0 orif, L=0 or3.
(a) When l. = 0, the equations become - s + J = 0 ...(r)
,
. -tc-2y+32-0 ...$i)
2x+y-32=0 ...uii)
Solving (ii) and Qiil, we g.t = -J-. Hence tc = y = z.
#E #T
(D When tr = 3, equations becomes identical.

l. Investigate for consistency of the following equations and if possible find the solutions :
4*-Zy +62=8,x+y -Bz= - 1, 15r- 3y + 9z=21. ( O s t n a n i . oL, 9 9 9 )
2 . F o r w h a t v a l u e s o f f t t h e e q u a t i o n sx , + y + z = 1 , 2 x + y + 4 2 = k , 4x+y+1:0z=lf haveasolutionand
solve them completely in each case. (V.f.(I., 2006)
p
3. Investigate for what values of l" and the simultaneous equations
ic+y+z=6, x+2y+32= 10,* +2y+)rz=1t,
have (i) no solution, (ii) a unique solution , (iii) an infinite number of solutions.
(U.P.:f .U., 2006 ; P.T.U., 2002 ; Sarnhalpul, 2002')
4. Test for consistency and solve
( i ) 2 x - 3 y + 7 z = 5 , 3 r + y - J z = 1 3 , 2 x + 1 9 y- 4 7 2 = 3 2 . (Kuruleslwtra,2005;Raipur,2005')
(ii) x+2y +z=3,2)c + 3y+ 2z=5,3x-W+52=2,3x+9y -z=4. (Bhillai,2005;Madras,2002,\
( i i i ) Z x + 6 y + 1 1 = 0 , 6 x + 2 0 y- B z + 3 = 0 , 6 y * I 8 z + 1 = 0 . (Ra.jasthan, 2005)
5. Find the values of o and b for which the equations
)c+ ay + z = 3, x + 2y + 2z =b, x + 5y + 3z = 9
are consistent. When will these equations have a unique solution ? (Kurukshetra, 2AA5 : M&drfts, 20A3)
6. Show that if l, * - 5, the system of equations
3)c- y + 4z = 3, x + 2y - 3z = - 2, 6x+ 5y + 7,2= - 3,
have a unique solution. If l, = * 5, show that the equations are consistent. Deterrnine the solutions in
each case
56 HIGHER
ENGINEERING
MATHEMATICS

7. Show that the equations

3x + 4y + 5z = a,,4x,+ 5y + 6z = b,5x+ 6y + 7z = c
do not have a solution unless a * c =2b. (Raipur, 200*; Nagput, 2001)
8. Provethattheequations5r+3y +22=L2,2x+4y +52=2,39r+ 43y+452=careincompatibleunless
c = 7 4 ; a n d i n t h a t c a s e t h e e q u a t i o n s a r e s a t i s f i e d b yx = 2 + t , ! =2-3t,2=-2 +2f, where f is any
arbitrary quantity.
9 . F i n d t h e v a l u e s o f l , f o r w h i c h . t h e e q u a t i o n s ( z - - ? r ) t c + Z y+ 3 = 0 , ? . x + ( 4 - 1 , ) y + 7 = 0 ,
2x + 5y + (6 - l) = 0 are consistent and find the values ofr andy corresponding to each of these values
of 1..
10. Show that there are three real values of l, for which the equations (o - tr) rc+ by + cz = 0,
bx+(c -I)y*az=0,9x+a,y+(q-7r)z = 0 . a r e s i m u l t a n e o u s l yt r u e a n d t h a t t h e p r o d u c t o f t h e s e
lo b rl
valuesof l, is D = lb t ol
l' & bl
11. Determine the values of & for *lti"h the foilowing system of equations has non-trivial solutions and
find them:
(ft - 1) x + (4le- 2) y + (k + 3) z = 0, (k - 1) r + (Sfr+ 1) y + 2hz = 0, 2x+ (3& + 1) y + 3 (k - 1) z = 0.
(Bombay,2005)
12. Show that the system of equations 2q-2xz*t3= \ut,bt-3*z*2xg=7ux2,--14*kz=k3 can
possess a non-trivial solution only if l, : 1, l, = - 3. Obtain the general solution in each case.
13. Determine the values of l, for which the following set of equations may possess non-trivial solution:
3r1 + xz- Ms = 0, 4tr1* zxz- $re = 0, 2lrt + 4*Z+ Lr3 = Q.
For each permissible value of 1,,determine the general solution. (Kurukshetra, 2006)
14. Solve completely the system of equations
U) x + y - 2z + 3u.r= 0 i rc- 2y + z - w = 0 i 4x + y - 5z + 8ur = 0 ; 5r - 7y + ?a - w = 0.
( i i ) 3 r + 4 y - z - 6 w = O ; 2 x + 3 y + 2 z - B w = 0 ; 2 x + y - L 4 z - 9 w = 0 ; r + 3 y + L B z + 3 r u= 0
Q.N.T,U.,2002 S)

2.r2. (r) UNEARTRANSFORMATTONS

Let (x, y) be the co-ordinates of a point P referred to set of rectangular axes OX, OY. Then
its co-ordinates (x',y') referred to OX,OY, obtained by rotatingthe former axes through,an
angle 0 are given by
t c ' = J c c o s 0 + y s i n 0 ,I
y' = -r sin 0 iy .os'0J "'(t)
A rnore general transformation than (j) is

;"=::;:ir,I ...(rr)
whichin matrixnotationis [";l=["t ?tlkl
L", urlLY
LY'J- l
Such transformations as (i) and (ii), are called linear transformations in two dimensions.
simirarly
therelations
ofthetype
i,,lrrl.:w,:::"1 ...uii)
z'=lgx+mgy+ngz)
give a linear transformation from (x, y, z) to (x', !' , z') in three dimensional problems.

lrtl l"t bL c1 nrl ["tl

In general,
therelation y=lrrl,o=1", b2 c2
y=,Axwhere url,*=l*rl
...,tr,
l."r r"' :"1 l"'l
bn cn 1r")
/. Lr'J Lo" L""j
LINEAR
ALGEBRA: DETERMINANTS,
MATRICES 57
i

grveUinear transformation from n variables 11, x2t...,xu to the variables yl, !2,...,y,, i.e. t]5'1e
transformation of the vector X to the vector Y.
This transformation is called linear because the linear relations A(X1 + Xz) = AXt+AX2 and
A(bn =bA)(, hold for this transformation.
If the transformation matrix A is singulir, the transformation also is said to be singular
otherwise non-singular. For a non-singulartransformation Y = AX,we can also write the inverse
transformationX = A-1 Y. A non-singular transformation is also called a regulartransformation.
Cor. If a transformation from (x1, x2,xs)to (yt, !z,le) is given by Y = AX and another transforrnation
of (yr,!2, !s) to (21,22,zg)is given by Z =BY, then the transformation from (rr, x2, x3)to (zt, zz,zs) is given
z = BY = B(AX)= (BA)X.
(2) Orthogonal transformation. The linear transftrmation (iu), i.e. Y - AX, is said to be
orth,ogortal if it transforms
ytz +yz2+... + y,rz into *t2 +*22 + ...+)cn2.
The matrix of an orthogonal transformation is called. an orthogonal matrix.

we havexx = txp2......x,r ::l=

,. | *t2 + *22+ ... * xnZ
t...1

and similarly
1",,-l
YY = ytz +yzz * ... * !,r2
.'. If Y = A)( is an orthogonal transformation, then
)(X =YY = 1AX)' (AJ0 = XA'AX which is possible only if A'A = I.
But A-1 A = I, therefore , A' =,{L for an orthogonal transformation.
Hence a squdre matrir A is said ta be orthogonal if AA' -- A'A = I.
Obs. 1. If A is orthogonal,A' and A- I are alsoorthogonal. (Assam, Iggg)
SinceA is orthogonal,A'=A-1.
(A')' = (A-1)'= (A')-r i.e. B, = B-L where B = A,
Hence B (i.e.A') is orthogonal.As A' =A-1, A-r is also orthogonal.
Obs.2.IfAisorthogonal,then lA | =11.
Since 111\'=/'!=!. lAl A' = l / l (Bom.bay,2006)
But lA'l=lAl, lA'l A =11
or lAlz=L i.e. A =*1.
Example 2.38. Shaw that the transformation
y l = 2 x l * x 2 * x g , ! 2 = t c 1 + x 2 * 2 t c g ,! S = x 1 - 2 x g
is regular. Write down the inuerse transformation.
Sol. The given transformation may be written as

Y =AX

where
[,'l ["]
X=lxzl,Y=llzl,a=1f L2l
lz 11-]

*2)
L"J Lr'J Ll 0

1l
Now I0 -21
,o,=l? 2l - -'t'
11
Thus the matrixA is non-singular and hencethe transfgrmation is regular.
 MATHEMATICS
HIGHERENIGINEERTNG

X=dr Y
Iz -z -1.l
where a-r=l_4 b gl
-lJ
L1-1
Thus xt=ZYt- ZYZ-)3 i xz= - 4Yr+ 5yg+ tsys; r g = J t ^ t z - l g
is the inverse transformation.

l-zts t /s 2nl
1 2/3 | 2/3 1/3 (Kurukshetra, 2005)
I tts -z/s zrcJ
sor, 1tur,-l?r?
wehave LtZ ir?1-l'rrl 2/3/3
2/3 L/3
-2/3 I
-z/s I vs z/sJI ue L/3
/3 2/3
- 4/9 + 2 / 9 + i2/9t -
l2 / s - 2 / 9 + 4 / 9 1
l- 4/9 + L/9 + 4/9
=l - 4/9 + 2/9 + 2/9 4/9 + 4 / 9 + 'L/9 2/9 - 4/9 + 2/91=I.
- z/9-
l- zts z/e+4/e 4 / 9 + i2/9 L/9+4/9+4/91
Hencethe matrix is orthogonal.
,11 2 ol
Example 2.4O.If A = il 2 I b orthogonal,find. &, b, c and,A- 1. (Bombay; 2006)
" L2 -2 cJlis
Sol. As A is orthogonal, AA' = I

1l
l-r

s l ;l - 2i ;l+flbi 3l=[l
? sl
o t]
.
'- "J'1" "l [o
2+2+oh 2-4+a"l Ig 0 ol
or | ;l ti* 4+\+bz 4-2+bcl=l 0I ol
4-Z+bc 4+4*t'l Io o 9l
12-4+at
5 + a 2 = g , 5 + b 2 = 9 , 8 + 12= 9 i ,e. a,2= 4, b2 = 4, c2 = L
Thuse = 2 , b = 2 , c = 1 .

SinceA is orthogonal,A-t =A'

*[ii ?]
2.13.(l) vEcToRs
Any quantity having n-componentsis called a uectorof order n. Therefore, the coefficients
in linear equation or the elements in a row or column matrix will form a vector. Thus any n
a
numberg x11*2t......,rn written in a particular order, constitute a vector r.
(2) Linear dependence. The uectorsx1r x21......,\r are said to be linearly dependent' f
there existr numbers Ir, h......, I" not all zero,such that
)tfxf + )r,2x2+ ... + I,:t,. = O. ...(t)
If no such nu,rnbers,other tha.nzero, exist, the uectorsare said to be linearly independent.
If l,t * 0, transposing ltxr to the other side and dividing by - Ir, we write (l) in the form
xl = It2x2+ lrgxg+ ......+ p#,-
Then the vector xl is Eaid to be a linear combination of the vectors x2, x3, ..., xr.
 59
MATRICES
ALGEBRA'DETERMINANTS,
LINEAR
(3, - 5, 2' 2) and'xg= (2' - 7' 3' 2)linearly
Exampt e2'4L.Are the uectorsxl = (/, 3, 4, 2), x2=
of the others'
d.epend.ent? If so etcpressone of these as a linear combity'd'tion
Sol. The relation l'1x1 + Mxz+ l'3x3 = 0'
- =
i.e., f,1(l,3, 4, 2) -; I2(3,- 5,2,2) + Ls(Z' 1' 3' 2) O
is equivalent to 7i1+ 314+ Z?ug= 0, 3It - 5?'2- 13= 0'
4L1+ zt"z+3l'3= 0' 2I1 + 27'"2+ 2l'3= Q
'
= = 13- - 2 which are not zero. the given
As these are satisfied by the values 1"1 t' f,z 1'
Yectorsare linearly dependent.Also we have the relation,
x1*X2-2x3-0
of the others'
by means of which any of the given vectors can be expressed as a linear combination
x1, x2' x3' w€ seethat the natrices
Obs. Apptyrng elementary row operations to the vectors
[",.l
A =lxzl B =l
[ *'
x2
I|
"ttd -
L*rJ L"t+x2 2xsJ
is also 2. Moreover x1, Y are linearly independent
have the same rank. The rank of B being z, the rank ofA
x1 and x2 = (*, +.-d] Similar results will hold
and x3 can be expressed as a linear combination of [ "' ",
i

general, we have the following results :

for column operations and for any matrix. In
(rows or columns) and the remoining uectors are
If o girin matrix has r linearly ind.ependent uectors r' it
matrix is r- conuersely, if a matrie is of rank
Iinear combinations of these r uectors, tien ranh of th" be expressed as a lineor
uectors $f ong can
contains r tincarty ind.epend.ent uectors and. remaiiiis
combination of these uectors.

1. Represent each of the ttansformations

x1=3!1+2y2, !I=zt+bzand 1 2= - ! t + 4 I z , ! z = 1 z t

by the use of matrices and find the composite transformation which expresses xr,'-tz in terms of
2L,zz.
that
2. If q=rcoscr-ysin6x,I=rsincr+ycosct, write the matrix A of transformation and prove
A-l = A'. Hence write the inverse transformation'
yt' !z'!s is given by Y-A){' and another
8. A transformation from the variables 11' !tu')cg to
= where
transformation from y1, !2, ll to 21,z2,23 is given by Z B{
lz 1 o-l ft 1 1l x2,xtto z1,z2tzs'
I =l 0 L -;1, U =l i i a OUtainthe transformationfromrl,
l.
l-r z 1l lt 3 5l 24'
..i*^"^-:-+i^- ^r^,] - r. + 2.r^+ l-rro: v.) = 2x, * 4x,)a 1 1 r g i ! 3 = - x 2 +
4 . F i n d t h e i n v e r s e t r a n s f o r m a t i o n o f y t = x 1 + 2 t c 2 + 5 1 3i l z = b t + M z +
5. Verify that the following matrix is orthogonal :
Itrt 2/3 2/3f
6)12/3 t/3 -2/31
-2/3 t/3
lzrt )
fcos0 0 sinel (Kurukshetra,2005l
(tt)l 0 I 0 I
e o cosoJ
lsin zb
[o
- c"ll i s o r t h o g o n a l ? (Bom,baY.2005
S)
6. Findthe$ues ofa,b,cffA=la
:1,; - b? cl
lo
60 HIGHER MATHEMATICS
ENGINEERING

II t tn n ol
7 . P r o v et h a t l
o o o -tl.
n t _ n r . j l i s o r t h o g o n a l w h eI =
n2/7,m=3/7,n=6/7.
l r
-I
L-", n 0J
8. If A and B are orthogonalmatrices,prove thatAB is also orthogonal. (Anna, 20051
9. Are the following vectorslinearly dependent.If so,find the relation betweenthem :
( i ) ( 3 ,2 , 7 ) , ( 2 ,4 , 1 ) ,( 1 ,* 2 , 6 ) .
( i i ) ( 1 , 1 , 1 , 3 ) ,( 1 , 2 , 3 ,4 ) ,( 2 , 3 ,4 , 9 ) .
( i j i ) x r = ( 1 , 2 , 4 ) , x z = ( 2 , - 1 ,s ) , x : r = ( o 1, , 2 ) , x + = ( - 3 , 7 , 2 ) . ( u . p . T . U . , z 0 0 a ; N a g p r2e0r0, 1 )

2.r4.(l) EreEN
VALUES
IfA is any square matrix of order ,?,we can form the matrixA - l.f, where / is the nth order
unit matrix, The determinant of this matrix equated to zero, i.e.
- 1 arz dLn
lott I
azz-)" dzn
lA-trIl=l"rt l=o
| ;;, inz ::: o,,,- xl
is called tb.e characteristic equation of A.On expanding the determinant, the characteristic
equation takes the form
(-I)''X"+hr)!-r +krV"-2 + ...*kn=e,
where ft's are expressible in terms of the elements ay. The roots of this equation are called the
eigen-ualuesor latent roots or characteristic roots of the matrix A.
(2) Eigen vectors

Y =AX ...(t)

carries the column vector X into the column vector Y by means of the square-matrix A. In
practice, it is often required to find such vectors which transform into themselves or to a scalar
multiple of themselves.
'
Let X be such a vector which transforms into lX by means of the transformation (i).
Then )uX-A)( orAX-^lX=OorlA-MX=O ...Ui)
This matrix equation represents n homogeneous linear equations
( a n - l " ) r 1+ a I Z x 2 + . . . + a 6 t c r = 0 )
c ,2 Ix L +(a Z Z -)' )x Z + ... + a 2,r:c,r= Ol
,,.uii)
. . . . . . . .I. . . 1 ,
artLxI* an*Z+ ... + (ann- )u)x,- 0)
which will have a non-trivial solution only if the coefficient matrix is singular, i.e. if
lA-M l=0. \
This is called the characteristic equation of the transformation and is same as the
characteristic equation of the matrix A, It has n roots and corresponding to each root, the
equation (ii) lor (iil)] wilt have a non-zero solution
X - ku x2t ......trrrJ',which is known as the eigen uector ot latent vector.
Obs. 1. Correspondingto n distinct eigenvalues,we get n independenteigenvectors.But when two or
more eigen values are equal, it may or may not be possible to get linearly independent eigen vectors
correspondingto the repeatedroots.

\
 MATRICES
6l
ALGEBRA: DETIRMlt''lANTS,
LINEAR
from (ii) that cX; is also a solution, where
obs" z. If X; is a solution for a eigen value l.; then it follows
to a.eigen ualue is not un'ique but may be any
c is an arfiprary constant. Thus the eigen uector correspond'ing
one of the vectors cX;.
I
te,'o'n, 1999)
Exampl e 2.42. Fird, the eigen ualues and, eigen uectorsof the motrf*1', fl
--t
L^
-
Sol. The characteristic equation is [A M = 0
ls-r 4l=o or )"2-77"*6=o
L.e.
I r z-?"1-
or (t-6)(I-1)=0 l.=6,1'
Thus the eigen values are 6 and t'
value l" then
If x, y be the components of an eigen ve9t9r corresponding to the eigen
^
tA-Mr=[u
i ,:sF]='
to l. = 6,weh"""
corresponding - i] =t
[- 1 b] g
- x-+ 4y -
which gives only one independent eqiation
' tc - I vector (4, 1).
E ' r ' r ^ - b the eigen
giving
4- t

tol"= 1,wen.""
corresponding
[t t] F]= .
which gives only one independent equation r *y = 0'
the eigenvector(1,- 1)'
i=+giving
I l1 1 3l
of
Exarnple 2.48.Find.theeigenualuesand.elgenuectors the matrixll
1
rI
'L. 13 1 1)
(Raipu.r,2005)
^
is I A - r/ I =ft i
equation
sor.Thecharacteristic s 1 l' ? lrr. ^t - 7 X 2 + 3 6 = 0
1 1 - 1"1 I
since )v= -Zsatisfiesit, we can write.this equa,Lr":.,
( 1 " + 2 )Q " 2- 9 1 . + 1 8 ) = 0 o r ( 1 " + 2 )( r - 3 ) ( ) ' - 6 ; = g '
Thus the eigen values of A are }, = - 2,3,6.
have
If r, y, z be the components of an eigen vector corresponding to the eigen value )', we
[r-r. 1 slkl
tA-MX=l 1 5-1. 1 llyl=o "'(t)
L3 1 l-til"J
P u t t i n gX = - Z , w e h a v e3 x + y + 3 2 - 0 , x + 7 y + z = 0 , 3 x + y + 3 e = 0 .
The frrst and third equationsbeing the same,we have from the first two

*-t=h or+="0=t
is an eigen
Hence the eigen vector is (-1, 0, 1). Also every non-zero multiple of this vector
vector corresponding to 7v= -2.
= non-zero
Similarly the eigen vectors corresponding to l. = 3 and l. 6 are the arbitrary
from (i)'
multiples ofihe vectors (1, -1, 1) and (7,2,1) which are obtained
-1, 1), (1, 2, t)'
Hence the three eigen vectors may be taken as (-L, 0, 1), (1,
62 MATHEMATICS
HIGHERENGINEERING

[e 1 41
of the matrix e --l O 2
Exampl e 2.44. Find, the eigenualuesand,eigen uector:s 6 I
L0 0 5J
(u.P,7.U.,2A05)
Sol. The characteristic equation is
[s-r r 4l
tA-xil=Oi.e.l 0 2-L 6 l=O
L 0 o 5-r.l
or (3-I)(2-1,)(5-h)=0
Thus the eigenvaluesofA are 2,3,5.
lf x, y,e be the components of an eigen vector corresponding to the eigen value I, we have
[s-r ]. 4 l['l
tA-MX=l 0 2-|," 6 llyl=o
L o o 5*r,ll,l
L=2'*" nT= 3z=0i'e'x*v =0 andz=o'
:.""t"* ::?::.',::,t'
1 -1=0="r(saY)
Hence the eigen vector corresponding to ), = 2 is k1 (1, - 1, 0).
P ut t ing ) v = 3, w e h a v e y + 4 2 = 0 , - y + 6 z =0,22 = 0 i .e.y = 0, z = 0

!-+-?
- -6= Rz
T o
to l' = 3 is &2(1,0, 0).
Hencethe eigenvectorcorresponding
to l. = 5 is ft3(3,2, 1).
Similarythe eigenvectorcorresponding

OF EIGENVALUES
2.I5. PROPERTIES
I. Any squa.re matrix A and its transpose A' haue the same eigen ualues.
We have (A' ^I)' = A' - |tI' = A' - ^I
lra-^I)'l=lA'-f,Il
lA-^Il=lA'-lJl l'.'lB' I = lB I
lA-^Il =0isandonlyif lA'-Ml =0
i.e, 7"is an eigen value ofA if and only if it is an eigenvalue of A'.
II. The eigenualuesof a triangular matrix arejust the diagonal elementsof the rnatrix.
att dty dLn

o azz ::: azn

A-- be a triangular matrix of order z.

0 0 dn,,

Then lA - ^I | = (orr - 7r)(azz - 7").....(ar,,- I).

.'. Roots of l4 - l'Il = 0 are 7u=cr11,ct22,..., Q'nn;
Hence the eigen values ofA are the cliagonal elements ofA i.e..a11,dz5, ... Qnn.
Con The eigen ualuesof a d,iagonalmatrix are just the diagonal elementsof the matrix.
III, The eigen ualues of an id,empotentmatrix are either zet'o or unity.
Let A be an idempotent matrix so that A2 = A.If l, be an eigen value of A, then there exists
a non-zerovectorXsuch thatAX= lX ...(1)
 LINEARALGEBRA: DETERMINANTS,
MATRICES

.'.A@n =A (lX) i.e.Azx =L(AX)

i.e. AX = l.(lX) 'A2 =A & Nt= lX
N{ =}"2X ..,e)
From ( 1) and (2), we get )"2X - lX or Q'2 - lr,)X= 0
.r2
l\' ?u=0whencel,=0or1.
Hence the result.
IV. The,sumof the eigen valuesof a matri^xis the sum of the elements of theprincipal d,iagonal.
lThispropertywill be provedfor a matrix of order3, but the methodwill be capableof easyextension
to matrices of any order.l
Consider the square matrix
atz
lo"
-l
^ =
"rrl
a2z ortl
lort asz
Loer "r?j
I - ^
lott
atz arl
I
so that lA-IIl=l azz-L o'zs I (O" expanding)
"rt
I ost agz offi-rl

= - ?r3+ 1.21a11
* azz+ agg) - ...(tt
)
If 1,1,M,Trrbe the eigenvalues ofA, then I A - lJ | = (-1)s (I- IIXX- ?rd(lr- Xg)
= - 13+ 1,211,1
+ Lz+ ls) - ..,... ...(ttr)
Equating the right hand sides of (ii) and (ili) and comparing coefficients of ?u2,we get
l,t+M* 1,3=ott * azz* o33.Hence the result.
V. The product of the eigen ualues of a matri* A is equal to its determinant. /
Putting l. = 0 in (iri), we get the result.
vI. If ?vis an eigen ualue of a matrix A, then L/?uis the eigen ualue of A-L
If X be the eigen vector correspondingto 1.,then AX = 7J(
Prernultiplying both sidesbyA-r, we get A-LAX=A-rlJ(
L.e. IX=lA-k or X= i.@-lK) i.e. A-1X= (L/t)x
This being of the same form as (i), shows that L/t' is an eigen value of the inverse matrix
A-r
VII. If 7vis an.eigenualue of an ortlrcgonal matrix, then !/?v is also its eigen value.
We know that if l' is an eigen value of a matrixA, then 1/I is an eigen value ofA-l. [Property
vl. since A is an orthogonal matrix , A'r is same as its transpose A'.
,'. 1,/Lis an eigen value of A'.
But the matrices A and A' have the same eigen values, since the determinants A - II
I I
and I A' - ^I I are the same.
Hence \/L is also an eigen value ofA.
Wil. If ?q,M, ,.....,?r,,
arg the,eigenualuesof a matrix A, then Am has the eigen ualues
LL'n, M'n, ...,..,Ln'n(m being a positiueinteger).
64 MATHEMATICS
ENGINEERING
HIGHER

LetA; be the eigen value of A and X; the corresponding eigen vector. Then
AX, - X;9, ...u)
We have AzXi = A(AX,)= A(?"#i)- ?v(AX)= MI,X,) - )"i2Xi
tX, -\"rtX,
Srmilarly, A' In general, A'nXi-\'nXiwhich is of the same form as (i).
Hence I;'' ir an eigen value of A''.
The corresponding eigen vector is the same X;.

LTONTHEOREM"
2.16. CAYLEY-HAMI
Euery squ.a.rematrix satisfies its own characteristic equatr.on ; i.e. if the characteristic
equation for the nth order square rnatrix A is
1 + . . . . .*.k , , = Q
lA-f,I | =(-l)"I" +k{- ...U)
then (*l)"A" + k1*- 1 + ......+ len* o.
Let the adjoint of the matrixA - l"I be P. Clearly the elements ofP will be polynomials of the
(n * 1)th degree in 1.,for the cofactors of the elements in I A - l,I I will be such polynomials.
P can be split up into a number of matrices, containing terms with the same powers of
1,.such that
- -z
P = PtN' r + P2N' +...... + P,, - 1)v+ P,, ...u.i\
where Py P2, ,.....,P,, are all the square matrices of order n whose elements are functions of the
elements ofA.
Since the product of a matrix by its adjoint = determinant of the matrix x unit matrix.
W - M } P = l A - ? l 1| x I
.'. by (i) and (ll), w - )'Aw
r?""-'
: i::-;:;..'u,*l!: ;':. :: +k,)r
17u
Equating the coefficients of various powers of 1.,we get
- Pr= (-l)'T I'.' IPt - P,

f,:_ii=I:J',
APn -t - Pn= k, - t!,
AP,r = k,rl
7,
Now pre-multiplying the equations by An, A"- ......,A,.[ respectively and adding, we get
( - L ) " A n + k 1 A n - 1 + . . . . .*. k r - 1 A + k r l - 0 , ...Uii)
for the terms on the left cancel in pairs. This proves the theorem.
Cor. Another method of finding the inverse.
Multiplyin g (iii) by A-1, we get
(-L)"A"- r * 7r4, -2 + .....* kn - { + k/-L = g

A - L= - * t
whence f f - t l ' t n n - + h 1 A " - 2+ . . . . . . + k n - f l .
R,,.

This result giues the inuerse of A in term,s of n - L powers of A and is considet'ed as a practical method
for the computation of the iruuerseof th.elarge matrices. As a by-product of the computation, the characteristic
equation and the determinant of the matrix are also obtained.

* Seefootnoteon p. 19. Williarn Rawan Hamilton (1805-1865)aq Irish mathematicianwho is known

for his work in dynamics. /'
 LINEAR
ALGEBRA: DEIERMINANTS,
MATRICES

Example 2'45. Verify Capttey-Hamilton theorem for the matrin t]""O

find iits
"=U
inuerse. AIso expressA5 -- 4( - 7A3 + 11A2-A - l0I as a linear polynomiat in A.
SoI. The characteristic equation ofA is
lr-r 1 I =o or t"2-+t'-5=o
^
l-r'" g]?,'1 ...(t)
By Cayley-Hamilton theorem,A must satisfr its characteristicequation (l), so that
A2-4A-br=o ...(tt)
NowA2-4A-'=[; f]t; 3]-,[l f]-,[l l]
-[l ls]-[s g]=[3 3]=o
rhis
verines,r"
;I",,"_ig]
Multiplying (ii) by A- l, w€ get A - A - EA- 1 = 0

or A-L_*,o_4r)=*[;
,1_-[l l]] =*[-; f]
Now dividing the polynomial ?'5 -4114-?l,3+ 11],2-1"- l0/ by the polynomial
]rz -_4]r-5,
we obtain
x5- +t( - 7?\3-r - 10/= (* -41- sxrs _ 27u+3)
+ t, + b
=1,+5 hv (r)
HenceA5 - 4A4-7A3+ LlAz -A-10/= A+Ewhich is a linearpolynomialinA.
t,
Exampl e2'4l.Find,thecharacteristicequationofthematrixA=f lt
i - 4 - 4 )I and.hence
fi.nd,itsinuerse. l-z
l-t 1 3 |
Sol. The characteristic equation is 1 3-I - 3 l = 0 i . e . ,I 3 - e O l + g = 0 .
-2 -4 -4-Ll
By cayley-Hamilton -_ZaA+g/=0,whenceA--L
theorem,AB -? -t",

b[l 9 gl r [-.1 :l _Lzj I s

=;13 L s/21
I ?l-;l 'g ,3 _i:,i_"ii) kfEx223
JL
,g)=L:ti
I 11
equationof the matrix,o=l l zo i
Example 2'47"Find the charucteristi,c 0 land. hence
computeA- r. AIso the
LI I 2)
find. matrix represented.by
A8-5A7 +T46 -gA5 +Aa -dAs+ gAz -ZA+1. (Rajasthan,2005;U.p.T.U.,z00g)
sol. The characteristic equation of the matrixA is
2 *1. 1 1
0 1 -r. 0 =0 or 1,3-5I2+?1,-B=0.
1 t 2 -i,
According to cayley-Hamilton theorem, we fave A3 - EA2+ 7A - B.I= 0
66 HIGHER MATHEMATICS
ENGINEERING

1,
Multiplyittg (;) by A- we get
A 2 - E A + 7 1- A A - 1 = 0 o r A - t = * W 2 - 5 A + 7 I 1 ...(tt)

lz 1 rl[z 1 1l [+*0+1 2+1+1 2+0+2.l f s 4 4l

B u t A z = lo 1 o l l o 1 0 l = l o * 0 + 0 0 + 1 + 0 0 + 0 + 0 1 = l0 1 0 1
L+L+2 1+0*4J L4 4 5J
It t zJlt 1 z) Lt*0+2
'll
A2-5A+"=ll i -i l]
il] i l].'[li l]=L -l
H e n c e f r o m ( i i )o, A -;
L - t *- 1
' = t ll
zl
N o w A s - 5 A ?+ 7 A 6* 3 A 5 + A a - 5 4 3 + 8 A s- 2 . A + I
=A5 (A3- 5A,2 +7A- 3I) +A (AB- 5A2+7A- 3O +A2 + A + I
=A2+A+I f . . A s- 5 A , 2 + 7 A - 3 l = 0 1
4 +l 1 tl o ol 5 5l
=l[ so I o l*ll zo 1 o l+[
[t [a
o 1 I l=lI I5 q I
l+ 4 sJ It 1 z] lo o tJ ls sJ

1. Find the sum andproductof the eigenrn"frr",off3 I tl (Madras,2006)

5J
tt ,F 3-'
2. Find the product of the eigenvalues of l- e - 1 Zl

B. Findtheeigenvaluesandeigen ,,l"*"r"tJ'
"-**l;"
(") 3] (w.B.r.u
[l
, zoos\ (b)
[l t
(Bhopal,2A02S\

4. Find the latent roots and the latent vectors of the matrices :
Is -6 zl
(o)l-6 7 -4',
-4 3J
L2
lz o r'l
( b ) 1 02 0 l
Ll 02)
l*z z -el
G)l 2 1-61
-2 0l
L-1
,' [e -z z] lz 1 -1-]
@l-7 _i *,1_l _; -?.J
(Kurukshetra,20a6) (Madras,2006\
ll
5. If ir,be an eigen valrre of a uon-singular matrirA, show tlrat I A l/Lis an eigen value-of the matrix
adj A. (U.P.T.U',2001\
LINEAR : DEIERMINANTS,
ALGEBRA U4[8!9E!

lz z 1l
6. TwoeigenvaluesofthematrixA=l 1 I I lare=1each'FindtheeigenvaluesofA-l'
It z z)
the latent roots
that if 2rr,lrz,..., trn are the latent roots of a matrix A, then A2 has
'7. Show
^A,...74- (P.T.U.,2005)
tr?, unequal eigen
g. For a s5rmmetrical square matrix, show that the eigen vectors corresponding to hro
values are orthogonai.
g.UsingCayley-Hamiltontheorem'findtheirrverseof
ftosl
('Dlt (osmania'
20:s'
(D ;] I ll
[3
fr 1 ql .^^^A . . . f ot -?I ? l
(iii)l 1 3 4l s)d \
(Blnpat,2002 (;u)
l 9l
u'P'r'u" 2006)
-4J
Lt -4 r--Lo^
9.',3J
11 3 7l
10. Find the characteristicequationof the *rtri*A =li s showthat the equationis satisfiedby
? l.
Ll 2 lJ (Anno' 20':05; KeraIa',2A05)
A and hence obtain the inverse of the given matrix'
-i _-1.1
11. Verify Caytey-Hamilton theorem for the matrixA and find its inverse
;" :i -Ll(Mad.ras,2006; u.P.T.u., 2005) (tt)l-6 -1 (coimbatore'
2007\
(r) l-1 2
-r z) le 2 -?l JI
lr
2 4l
( i i i l[la4 s 2 l (P.T.u.,2006\
lz 4 3l
12. Using caytey-Himittontheorem,findA8, oo = -i] (Aino,20a3,
[l
I a 6 ql r,{2 andA-3'
rs. If A =l i s i l,
-4 -31 "t"toateA-
L-1
I z -1 zl
t4.tfA=l-1 2 -rl,rnaa4. Modras,2006)

L1-1 2J

TODIAGONAIFORM
2"r7.(l) REDUCTION
a nwtri'x P can be foun'd'
If a square matrix A of ord,ern has n tineqrly ind.epend.entei6en vectors, then
I AP is a diagonal motrix.
swch that F

to matrices of any order.l

a square matrix of order 3. Let 7n,)u2, ?t3be.its eigen values and \
LetAbe
.["'l
l.r1 ['rl. vectors.
eigen 1,,
xL=lr, l, *r=lrrf ana xs=lr,luuthecorresponding
L"'J l',) L'uJ
f"t x2 ,rl-by P, we have
- = lrr
Denoting the square matrix IXLX#sl h Js I
P, 22 ttJ
AP - AV\X#sl = l4)(1, A){2, /DteI = 0'1X1, ? /12,}'3X31
68 MATHEMATICS
HIGHERENGINEERING

[Ir", Wz lg"sl lrt x2 "rl h, 0 0l

=llut b t z i . g y g l = l y ry z r s l x l o ?", 0 l= pO. whereD is the diagonalmatrix.
lLr", )"*z M4) pr 22
"t) L
0 o 1rJ
.'. P-LAP = Fr PD =D, which provesthe theorem.
Obs. 1. The matrix P which diagonalises A is called thc mod,al matris of A and. the resulting diagonal
matrix D is hnown as the spectral motri,x of A.
2. The diagonal matrix has the eigen ualues of A as its diagonal elemnnts.
8. The rnatrir P, which diagonalise A, consitutes tlrc eigen uectors of A.
(2) Similarity of matrices.A squ(ffe matrixlof ord.er n is called.similar to a square rnatrix
A of order n if
A = F IAP fo, sor-nenon-singular n x n matri.x P.
This transformation of a matrix A by a non-singular matrix P to A is called a similarity
transformation.
Obs. If the matrix A is similur to the matrixA, thenA hasthe sonrceigenvaluesosA.
If x is an eigen uector of A, then ! = F tx is an eigen uector of A correspond.ing to the sanrc eigen ualue.
(3) Powers of a matrix. Diogonalisation of a matri,xis quite usefulfor obtaining powersof
a matrix.
Let A be the square matrix. Then a non-singularmatrix P can be found such that
D _ P' I A P
-
D2 = (P-LAP) (P-lail _ P-rAzP t'.' PFL=I
Similarty D3 = P_LA?Pand in general, Dn = P-l*P ...(t)
To obtain An, premultiply (i) by P and post-multiply by P-1.
Then PDryFL - PFLA"PP-L =An which gives An.
lit o 9l
Thus, An l
=pDnF =l
where,
D,
B \ ^l .J
Working procedure :
1. Find the eigenvalues of the squarematrixA.
2. Find the correspondingeigen vectors and write the modal matrix P.
3" Find the diagonal matrix D from D = P- r AI,
4. ObtainA" from An =PD'FL.

Sol. The characteristic equation ofA is

[-r-i" z -21
I - t1 2-L
-1
tl=oor?r3-L2-51.+5=0.
-l, l
|
Solving,we gettr, = i, b = {5, Ig = - {5 ur'theeigenvaluesofA.
WhenL = 1, the corresponding eigenvectoris givenby
- 2"+x 2 y- h = 0 , x + y + z = A , - x ,- y - z = 0
Solvingthe first two equations,
we get the eigenvector(1,0, - 1)
t=t=Sgiving
LINEARALGEBRA: DETERMINANTS,
MATRICES 69
When l. = {5, the correspondingeigen vector is given by
( - 1 - r / S ; r + 2 t - 2 2 = 0 , x + ( Z - { S ly + z - 0 , - x - y - , , l i l z_ 0 .
Solving 2nd and 3rd equations, we get
x y z ='16
= or
6 - 2\B 1:6
grving the eigen vector t{S - l, t, - 1).
similarly the eigen vector correspondingto r. = - r/b, is 1r/b+ 1, - 1, 1).
writing tbe three eigen,vectorsas the three columns, we get the transformation (modal)
I r . / s - i { s- 1+ r l
m a t r i x a s P = l0 1 f.
Hence the diagonal matrix is

fl, o ol fr o o I
a = l o L z o l = lo . / s o I
l o o I ' a Jl o o - { b J
lt I 3l
Exampl e2.4g.Find, a matrix p which transformsthe *otri*e =l 5 ' 1 l t o t h ediagonal
L3 I 1 l
form. Hence calculate Aa.
S,ol.The eigen values ofA (found in Ex. 2.36) are -2,3, 6 and the eigen vectors are (-1, 0,
1), (1,-1, 1), (L,z, r). writing these eigen vectors as the three .jrr*rrr, the required
transformation matrix (rnodal matri.r) is
[-r _I
1 rl
P=LI
N
J-r t rl 1", br I
rofindFr, fpl=lo -i i l = | " , o i "r ( s a y )
";l
|1 I 11 1", bs ,rl
At = 4, Bt= 2, Ct= l,AZ= 0, BZ- -2, CZ= Z,Ag='8,BB= 2, Cg= L
Also I P | =atAr+b181+c1C1-B

pr'L=r"-,Fi
i: fr]=*[-?
i il
Thus D=p_,*
=[-l
; l]
=[';* J, ll=['r'J, BI
D4
_ Lo _9 e'Jlo o rzeeJ
HenceAI = PDAPI =*l-ii;ll'f
"L 1 1 tJLo {, ilf ; -s;l
o rzoeJl_r
z iJ
[-r r rlf -a o
sl lzst 4Bs zssl
=| 0 -1 2ll 27 -27 ZZl=l +AS 1051 aSSI
L1 1 ljfzro Etz zroJ lzss 48b zsr)
 HIGHER MATHEMATICS
ENGINEERING

OF OUADRATICFORM TO CANONICAL FORM

2.18. REDUCTION
A homogeneousexpression qf the second degree in any number of variables is called a
quadratic form.

la h sl t"l y zl,then
F o r i n s t a n cief A
, =lh b fl,X=lylandX=[r
ls r cJ VJ
XA){ = ax2 + byz * + Zfyz + 4zr + 2hry ...(r)
"rz
which is a quadratic form,
Let 1.1,M,Trsbe the eigen values of the matrixA and

l-"'l krlX"=["'l
Xr=ltt I, Xz=ltzl, ltr I
L"'J L"rJ L"'J
be its correspondingeigenvectorsin the normalized form (i.e. eachelementis divided by square
root of sum of the squares of all the three elements in the eigen vector).

- [11 o ol ["t xc2

"t1
F'AP -l O Lz 0 lwhereP =ltr
Thenby$ 2.17(l), tz tr I
rrj z2
[o o l"t "t)
Hence the quadratic form (i) is reduced to @canonical form (or sum of squares form or
Principal axes form.)
Llxz+ \'gf +t szz
and P is the matrix of transformation which is an orthogonal matrix.
Note. Congruent (or orthogonal) transformation. The diagonalmatrix D and the matrix A are
called congruent matrices and the aboue mcthod of reduction is called congruent (or orthogonal)
transformation
Remember that the matrixA coresponding to the quadratic form
.*i::.::!: .{::l:
rc.etr
";i
is "*n
If "rr i"* jrl i.o"r. :1,,""1,iii clj
"i,
orr" coeff.
ofry | l!
-
f *.o"n I lo,"r.ot", I
Example 2.50. Red.ucethe quad.raticform 3f + 5y2 + 3* - 2yz + Zzx - Zxy to the canonical
form and specifythe matrix of transforrnation. (Kurukshetra, 2006 ; Madras, 2002)

Is -1 1l
Sol. The matrix of the given quadratic form is A = | -i ; -1
I
-1 3J
L1
lA-lJ | =0i.e.lt;^ u-1?' lt l=o
Irscharacteristicequarionis
-1 B-?'l
I 1
which gives 7v=2,3,6 as its eigen values. Hence the given quadratic form reduces to the
canonical form
?u1x2
+tvyz*W2 i.e. ?-rz+ 3y2+ 622.
LINEARALGEBRA: DETERMINANTS,
MATRICES 7l
To ftnd the rnatrix af transfoirmation
From ta - ]J 1X.= 0, we obtain the equations
( 3- I ) x , - y * z = 0 i - x + ( 5- ? ' ) y- z = 0 ; x - y + ( 3- ? r . )=e0 .
Now corresponding to I= 2, we getx -y + z =0,-.r + 3y - z= 0, andx,- y + z = O,
x _L_ z
whence
1-0--1
.'. The eigen vector is & (1, 0, - 1) and its normalispdform is 1ltl2, O,- l/^lD.
Similarly correspondingto l, = 3, the eigen vector is X2 (1, 1, 1) and its normalised form is
$/\b, L/\8, L/rE).
Finally, correspondingto l, = 6, the eigen vector is Xs (L, -2,1) and its normalised form is
$/16, -2/46,1/{6).
t'13 L/$ V./6.l
Hencethe matrix of transformationis p = [ L/ls - zrlel
l- trlz L/ls tl'{,el
2.19.NATURE
OF A OUADRATIC
FORM
Let Q =X A)(be a quadratic form in n variables 11, x2t ,.. lcn.
Index. The number of positiue terms in its canoni.cal form is called, the ind.ex of the quadratic
form.
Signature (S) of the quadratic form is the difference of positive and negative terms in the
cananical forrn. If the rank of the matrix A i,s r and the signature of the quadratic form Q is S,
then the quadratic form is said to be
(i) positiue defi,nite if r = n and s = rl
(ii) negative d.efinite if r= D &Dd s = 0
(iii) positive semid,efi.niteif r < n and and s = r
(iu) negatiue semidefinite if r <n and s = 0
(u) indefinite in all other cases.
In other words a real quadratic form XN( in n variables is said.to be
(i) positive definite if aII the eigen values of A> 0.
(ii) negative definite if all the eigen ualaes of A < 0.
(lil) positive semidefinite if aII the eigen ualues of A> 0 and at least one eigen ualue = 0.
(lu) negative semidefinite if all the eigen ualues of A < 0 and at least one eigen aalue = 0.
(u) indefinite if some of th,e eigen ualues of A are positivb and others negatiue.
Exaiaple 2'51. Reduce the quadratic forrn 2xp2 + 2r7rg - 2x2xg to a canonical form by an
orthogonalreductionand discussits nature. (Madras,2006)
AIsofind the mod.almatrix.
I 1l
(z)The matrix of the givenquadraticform ,, O = 0 -11
[ I -1 0J
L1
[-l' 1 1l
Its characteristic equation is [A - M = O i.e.l 1 - l" -11=o
I t -1 -rJ
 HIGHER
ENGINEERING
MATHEMATICS

which gives 1,3- 31,+ 2 = 0

Solving, we get l, = 1, 1, - 2 as the eigen values. Hence the given quadratic form reduces to
the canonical form
7,"p2 +t"zyz*Lgzz =0 i.e. *2 +y2 -2a2 =0
(ii) Since some of the eigen values of A are positive and others are negative, tbe given
quadratic form is indeftnite.
(iii) To find the matrix of transformation
From [A - M X = 0, we get the equations
7 o + y + z = A , f r - l g t+ z = 0 , l c - y - 7 , 2 . - 0
Whenf, - - 2, we get ?.x+y + z = O,x + 2y - z = 0, x -! + ?.2=0.
Solving first and secondequations, we get

+=tT=i
The corresponding eigen vector Xt = (- 1, 1, 1) and its normalised form is
eL/\E, t/\8, L/lq
W h e n l v =7 , w e g e t - x ,+ y + z = 0 , f ,- y - z = 0 , x - y - z = 0 .
Theseequations are same.Take t = 0 so that x = 2,
.'. The correspondingeigenvector Xz= (1, 0, 1) and its normalisedform is 1L/r12,O,l/l1)
To ft.nd the eigen uectorXs= (1,m, n) (say)
SinceX3 is orthogonal to Xl, ... - I + m * fl. - 0
SinceX3 is orthogonal to Xz, :. I + n = 0

These give
equations
+=T= *,
.'. The eigenvectorXg = (L,2,- 1) and normalisedform is 11/{G,Z/ftG,- f Z{q.
Hence the modal matrix is
- v,le v,lz t/161
p =l| vls o u,la I
I rt,,ls r/lz - vta l

| -ttz -lsrz ol
I trz ,lgrz
r. IfA= ,ro, r =lrvz tro, 31,,n"* thatFlAP is a diasonatmatrix.
f-./szz B.l""u 1J
2. Show that the linear transformation
=l-:fr =
" 3 :* 3],where0 * tan-l*,,changes thematrix
n1
c=l? ilto ttt* diagonalform D = HCH'.
V
8. Reducethe matrixA = : the diagonalform. (8.P.7.U.,2005)
[ ll l, ]"
4. By diagonalisingthe matrixA =[ _
I I ] *uot (Madras,2002)
 MATRICES
LINEARALGEBRA: DETERMINANTS,
73
-l rl
5.If A =f-l -l'1,*rrorateAa. (coirnbatore,200l) 6.If A =| I ; 11,frndAs.
[r
-1 sJ L - 4 4 3l
^ (Madras,2006)
7. Find the indexandsignatureof the quadraticformrl + 2xl- 3*3.
-3
8. Find the eigenvectorsof the *rr"* -?1""u hencereduce
f-3 -1
l2 3J
6f + By, + k2 -2yz + 4zx - 4xy to a'sum of squares'. Also write the nature of the matrix.
(Calicui,2005)

9. Reduce the quadratic form |ny + Zyz + |axinto canonical form.

(Kurukshetra, 2006 ; Bombay, 2003 ; Madras, 2002)
[rool -|
lO. Find the eigen values, eigen veetors and the mod,al matrix of the matrix | 0 and hence
-1I 3J|
L0
reduce the quadratic form x12+3x22 + 3rg2 -bfisto a canonical form. (Andhra,2000)
'sum of squares'!v orthogonal transformation and
11. Reducetfe following quadratic forms into a 11
give the matrix of transformation. Also state the nature of each of these'
(i) 3rr2 + 3x22+ Sxsz+ ?^xg2* ?ags - Zxzrs.
(Anna,2002S)
Oi)A2 + 7y2+ 322- lzlrt - \yz + 4zr
lz. ShoJ that the form br12+ 2&r,22 + 10rs2+ Mzrra+ !4xsxt + *xfizis a positive semi-definite and frnd a
non-zero set of values of tr1, xz,xgwhich make the form zero. (P.T.U.,2003)

MATRICES
2,2O,COMPTEX
So far, we have consid,eredmatrices whose elements were real numbers. The elements of a
matrix can, however, be complex numbers also.
(1) Cor$ugate of a matrix. If the elenrcnts of a matrix A = [4."] q,re cornplex numbers
or" * i 9r", crr" and p* being real, then the ma,trix
A * Wrrl = [crr, - ipr"] is called the coniugate matrix of A.

The transpose of a conjugate of a matrix A is denoted by A* or A0 i.e. (A)- = A* .

(2) Hermitian matrix. A square matrix A such that A'=A is said to be a Hermitian
matrix.* The elements of the leading diagonal of a Hermitian matrix are evidently real, while
every is the complex conjugate of the element in the transpo-sed position. For
" +other element |
.instance l - z-
A =l - - ,
3 +'--'a l lis* ^'tr.'r''ur.'
-5 I" a *
.Hermitian , - r . " , sinceA' =l q 2 . 3 - 4
1- , " , . , r * ,matrix,
'rsr
:l=a
-5J
ya 4i + 4i
J li
(B) Skew-Hermitian matrix. A square matrix A such that A' - -A is said to be a
skew-Hermitian matrix. This implies that the leading diagonal elements of a skew-Hermitian
matrix are either all zeros or all purely imaginary.
Obs.A Hermitian matrixis a generalisationof a,real symmetric matrix as every real symmetric matrix
is Hermitian. Similarly, a skew-Hermitianmatrix is a generalisationof a real skew-symmetricmatrix.
Properties.
L Any square matrix A can be written as the sum of a Hermitian and shew'Hermitian
matrices.

Take B= + 1 A + A ) acn=df , r e - l , t
* Named alter the Frenchmathemati cianCharlesHermite (1822-f 901),known for his contributionsto
algebra.a;a theory.
"rr*1""
74 MATHEMATICS
HIGHERENGINEERING

1-1-
Then B' =; (A + A')' =; (A' + A\
1- 1_
and B=i(A+T)=;6+A"s=3'
i.e. B is a Hermitian matrix.

Asain =i(A -T)'=lA' -A>

"'
and e =ro -fr,l=| A -A)=-c
:. ( = -e i.e. C is a skew-Hermitian matrix.
Thus A=Lz(A+A).|a-l.:)=B+c
Hencethe result.
II. If A is q,Hermitian ma.tria, then (iN is a skew-Hermitian matrix
TVehave 1il1'= fA-1'= (- r:I)' = - iX
=-itA f.' A'=A
Thus (rA) is a skew-Hermitianmatrix.
matrix thcn (i,N is a Hermitian matrix.
Similarly if A is a skew-Hermitia.n
III. The eigen ualues of a Hermitian matria are real. (seeFig. 2.1)
Let I be the eigen value andXthe cornespondingeigen vector of a Hermitian matrixA, so that
Alt =)J(
7 m=.Flx= rFx
t'=X NUXX
S inc eV X = i1r1 + I,c rz + ...+ L n x n = l rr1 2 + l rzl 2* ...+ l rol 2 i s real and non-zero.A l so FaX i s' A
Hermitian form which is always real.
.'. ?r,,the eigen value of a Hermitian rnatrix is real.
IV, The eigen ualues of a sk:ew-Hermitian matrir are purely imaginary or zero.
Let ?.be the eigen value and X the correspsnding ergenvector of a skew-Hermitian matrix B so that
BX=7J(.
7' BX=xlX= fFX or l'=.F BX/-XX
SinceXXis real andnon-zero.AlsoEAX is a skew-Hermitianformwhich is purelyimaginar5ror zdro.
.'. 1,,the eigenvalue of a skew-Hermitianmatrix is purely imaginaryor zero.
(4) UnitarXl natrix. A square matrix U such that U' = V 1 is called a unitary matrix. For
a unitary matrix (J, (J . t - U . U = I.
This is a generalisation of the orthogonal matrix in the complex field.
Properties
I. Inuerseof a unitary matrix is unitany
If U is a unitary matrix, then
T l= A = ( r - r
or U =U-'
l ( u - 1 ) - ' Y- F
Writing II-L =% we have
[ Y - 1 ] ' =7 o r V - L = 7
Thus V (=(J- 11is also unitary.
Con Inuerseof an orthogonal matrid is orthogonal.
II. Transpose of a unitary m.atri* is unitary
If U is a unitary matrix, O'
!-'
(u')= u- I
 MATRICES
ALGEBRA: DEIERMINANTS,
LINEAR
'y = tul- |
or
;y
writing(J'=v,*" n..,l9u]1
Thus V (i.e. U') is also unitary'
Cot. Tra'nsposeof an ofihogonal mntrix is orthogonal'
III. Prod.uct of two unitary matrices is a unitary tnatrir.
If Uand V arcunitary matrices then
(J'=t'L, V'=T-l
1 = ( U 1 4 t- = V - r U - L
Now, 1@- (J,V are unitary.
_VU
_(w,
lhus UV is a unitarY matrix'
matrir'
co* Prod,uetof two orthogonal matrixesis an orthogonal
value 7.
IV. The eigen ualue of a unitary matrix has absolute
= ...(1)
If U is a unitarY matrix then UX 7J(
Taking conjugate transPoseof (1),
- (W'=(Vb'=73'=TiT-r
'
Also (m)' = (]t X)' = ?'"X
i-e. V u-r =T7
Post-$ultiplyrng (2) by (1)' we get
tT v- t) (ux)= 1i'"Fltrx)
7 tu-t tnx= (-),r)&n 'wL(J=I
V x=(i,rlxx
Thus l.1r'=lLl2 = 1. t'.'fr*o
Hencethe result.
ualue 7'
Cor. The eigen ualue of an orthogonol motrix has absolute
t, - that,M* i,sa Hermitian*il?:where A*
Example 2.62.,f o =l' :; i:ti\'o"*
is the conjugate transPose of A.

I z+i
Sol. We have A' =l 3
l-1+3i
i\]
Img.l" Skew- Herrnitian ( skew'symmetric )
Unitary brthogonal)
Hennitian (sYmm'etric)

Real L

Fig.2.1.Eigenvaluesof variousmatrices'
76 HIGHER
ENGINEERING
MATHEMATICS

'; '
and A*=[ -u,I
[-r-e; a + zi)
-1+s;'[ 2-i -',
M * - i t - t ti 1
4 - n l l - a-:i, -. - i I
1-b
:f 4- i2+e+r'5
-L-10+5r+3r-fO-fO;, ii'-,I,- ro+loiI
2 E - t 2 + 1 6 -4 i ' J
- 20 + 2i
= |- 24- *hi.h
" ---- is a Hermitian matrix.
L- 20 2i 46 l, l'
| * u * a I' t* ;l I
Exampte2.53.prouethat tt'ce ^ =li(/
ma'trix unitaa o"oA-1'
+,) u-u ",:o
L z t - ' - ' 2 \ -i Ilis (Bombay,zooil)

L.ofna,
Sol. Conjugate r = flrt-rl
A i.e.A *nt-,)l
o,
I i
tr-;r |o *,1I
Lf l
[r - r - I
tl tl 1
* (1 f tt
.'.rranspose
orr i.e. As=
|
1-,)i,,.,,
L;,- J
-4 *n'."I
Now
AeA=fi(1 it'-"lfftr*;r
1-') r'*,,.J1|
L*. rr*;r | t,'-, .J

|=[]1i;::;11;:,*
L " ;l:;:;l:;','
l + + +'
I

l]='
Similarly AAs- I.
HenceA is a matrix.
""tt?l
-t, tt -t,
-Ao=I *:t
ArsoA-L i l
L-;(1+i)7tt*01
Example 2.64.GiuenthatA =l- ' I is a unitary
(t - A)(r + A1-
f* zi tr"), showthat
matrix.

sor. tit'],[+Al =1'(-1-4)=6.

I*A=[-11r,
 77
MATRICES
ALGEBRA: DffERMINANTS,
LINEAR

...(r)

= ' ...ui)
Its conjugate-transpose * t- il n, : t')
- 2 - 4 t^l _
t'-4 z+4i1 1 [s6
(i)
.'. Productof (ii) = 1
and
- [ - 4
86lz --+t - 4
)l- z + +r
-4 l=tol o ,3]='
Hencethe result.

A is real and symmetric and B is

1, prove that every Hermitian matrix san be written as A + iB, where P'T'U'' 1999\
real and skew-symmetric :
as P + iQ, where P and Q are Hermitian matrices'
2. Show that every square matrix can be expressed (BhoPaI,2002S)
g. Show that a Hermitian matrix remains Hermitian when transformed by an orthogonal matrix'
f cr+ ir - p + * 62= 1. (u'P'T'rl', 2006)
4. show that the matrrx *;d d- -t^:u
tr_ l t, " unitary matrix, if ae F' f+
L P J
I s i-+t -2+5i1
b. Show thatl 7 + 4i -2 3 + i lis a Hermitian matrix.
l-z-si 3-t 4 l
-L

f - f -2 + i 5 - a t l
6. IfA =l z-i Ei l,showthatAisaHermitianmatrixandiAisaskew-Hermitianmatrix'
7
-5i
2 I (Sambalpur,2oo2)
[s+s;
I l+il
----..^-.1[ 1
7. Show that the following matrix is unitary 15 f -; - t (u.P.T.U.,2oo2)
| J'
'jr' trr'lasp+iewherepisrealandskewsymmetricandQisrealand
8. Expresro=f -3'*,
3t o J
l--L-i (Bornbay,2006)
symmetric
fr 1 1l
o. rrs=li o2 a l , w h e r ae , = e 2 i n / s , p r o v e t thta=t* S '
It a zaa6; J.N.T.u.,z00r)
eturukshetra,
",)

TYPEOF AUESTIONS
2.21.OBJECTIVE

Choosethe correct answer or fitl up the blanks in the following problems:

1. To multiply a matrix by scalark, multiply
(o) any row by & (b) everyelementby & (c)any columnby ft'

z.ra=[t -i]thenA'is_
,",I t lt" ,:;"1rur ,, ',-jX] '"1',:T
,-Xl
I i.t l-o]
00-l
lil?
ofthe*^t*l
3. Theinverse 4 0[is
I 0lj
78 MATHEMATICS
ENGINEERING
HIGHER

[o.s o ol [o.s o ol l-z o ol lz o ol

(o)lo -4 ol @ lo -4 ol ( c ) l 0 0 . 2 5o l rallo -0'25 ol
-lJ
[o o lJ Lo o lJ Lo o lJ Lo o
lzo
e .I f A = 1 0 z 31""u =ll ? 3l ,0.' thedeterminantAB
hasthevalue
100 z ) " loo z )
(d4 (b) 8 (c) 16 (il 32.
5. fire system of equations r + 2y + z = 9, 2x + y + 3z= 7 can be expressedas

o,[] ?tl=t?l|i] ru,

ll i lll?l=
|i]
=tlt
(.,[; ?ltL] (d) none of the above.

r. -3] 'n""xequars
"[i i]"=[l
- G)[l ',] *,[-l l] ro[t _lt]
",[-l lt]
z.ra =[? ?]tn"", (adiA)equals

(") ,3]
[13 *,l$ '3] (") ,f]
[1]
(d) none of the above.

8. If 3r +IA +z=0,x+ 4y *z =0t2r+y +4,2=0, be a systemof equations, then

(a) it is inconsistent
(b) it has only the trivial solution x = O,! = 0, z = 0
(c) it can be reduced to a single equation and so a solution doesnot exist.
(d determinant of the matrix of coeffrcients is zero.
=ll l],"=l_li]""u :fi3ffi3]'n""
e.rrA "=f
(o)C=Acos0-8sin0 (b) C=Asin0+BcosO
(c)C=Asin0-Bcos0 (d) C=Acos0+Bsin0.
[rool frool
lO.LetA=lo 1 0landB=lZ 1 Ol.ttren
LP | 1l Ls 4 lJ
(d A is nrw equivalenttoB only when s.= 2, F = 3, and y= 4
(b) Ais row eqrrivalentto B only when o * 0, p * 0 and 1= Q
(c) A is not row equivalent to B
(il A is row equivalent to B for all value of a, $,1.

lLrfA _l]=[_
? l]*n*"o=fr 3],n""o
*
[3
r",[3l] (e) l]
[l
(.)[_i l] <atl-ln-l,r)
12. Matrix has a value. This staternent
(o) is always true (b) dependsupon the matrices (c) is false.
13. IfA is a square matrix such that AA' =1, then value ofAA is
(d 42 (6)r (c) A- I
 79
ALGEBRA
LINEAR MATRICES
: DEIERMIN4NI!,
of A is
14. If every minor of order r of a matrix A is aero,then rank
(o) greater than r (b) equal to r
(c) less than or equal to r (d) less than r'
15. A square matrixA is called orthogonal if
(a) A --Az ( b )A ' = A - r k) AA-r = I.
lz -1 -2
3 1l
16. The rank of matrix | 1 4 1 lit """
Ls 2 4 3J
elements of the principal diagonal'
l?. The sum of the eigen values of a matrix is the ....... of the
18. Rank of a unit (identity) matrix of order 4 is """'
l-r 2-21 fsz 6l
19.Inverseofl-1 3 olisll 1 frlthen&is"""
-2 lJ Lz 2 5J
lo (trueorFalse)
20. Theequationst+2'y=1,7x+l4y=12-ateconsistent
I a -76 - 4
+ l*" B and 18, then the third eigenvalue is .....
21, If two eigen values orl- e
lz -4 3J
22. Aquadratic form is positive semi-defrnite when
2g. Am*o and Bo *o are two matrices' lThen will .
(d A. B exist (*",
b )tL
A+rB x i s.to?
n e,i.r.rD,,.
fr 0 or
-
24. fire product of the eigenvatuesor[|
_I l.|.'......
fs 4 5 6 ?l
2E.Therankorthe*,**l
I [ ? I 3Ii'
L1o tL Lz 13 141
26. An example of a 3 x 3 matrix of rank one is
2?. The quadratic form correspondingto the symmetri. t tt i* - ?]is """"
[]
s = . . . . . . . 1/ = . . . . ' ' . . ' ,
,":1:::..:n" e q u a t i o n sf + 2 y + 3 2 = a , 3 x + 4 y + & = 0 , 7 x + L a y + L 2 . 2 = o ,

l-e _z z]
3 -
29. The eigen values of the matrix l- Z 1l*" """"
-1 3j
L2
if ......
30. A matrix Ais id'ernpatent
ft 1 -1-]
31. The rank of the matrix | 2 - 3 4 li" ..""'
-2 3J
L3
[-r 2 E the eigen values of Az are ""
s2. If A = | 03 s Ut""
- 2 l,J
Lo o (True or FalsP)
33. If rank (A) =2, rauk (B) = 3, then rank ('48) = 6'
34. The maximum value of the rank of a 4 x 5 matrix is """' (True or False)
s5. Any set of vectors which includes the zero vector is linearly independent' (True or False)
86. If l, is an eigen value of a symmetric matrix, then )uis real.
BZ. The eigen values f]*"
"r*"tt*[V2z
88. The eigen values of a triangular matrix ane """
80 HIGHER
ENGINEERING
MATHEMATICS

-
39. If the product of two eigen values of the **r* - ?l is 16, then the third eigen value is
i-u,
-1 3l
L2
"
40' ff tr;, i = L,2, ...... n are the eigen values of a square matrix A, then the eigen values of AT are ......
41. By applyrng elementary transformations to a matrix, its rank
(o) increases (b) decreases (c) does not change.
42. rf l. is an eigen value ofA, then it is an eigen value ofB, only if B = ......
[r 2z sl
ag. I f A = 10 5 l , th e n e i g e nv a l u e so f A- 1 a r" ......
lL o o 3J l
I q 1n Kl
equati", -; l3
44.rhe characteristic - ll-
"rj 3 7)
L
45. Every square matrix does not satisfy its own characteristic equation. (Thle or False)
46. If l, is an eigen value of an orthogonal matrix, then 1,21,is also itq eigen value. (Thre or False)
47. Every Hermitian matrix can be written asA + iB, whereA is real and ...... and B is real and ......

48.Thesumandproduct oftheeigen rrurrr",

orf? N ur" .......
and........
Ll 22)
49. If the rank of a matrixA is B, then the ranh of BA?
" ir 1. (True or False)
50. The product of the eigen values of a matrix is equal to ......

51. The eigen values *o =fl ll*" the roots of the equation ......
LZ OJ
52. A system of linear non-homogeneous equations is consistent, if and only if the rank of coefficient
matrix is equal to rank of ......
53. Thevectors[1, 1,- 1, 1], [1, -L,2,- U, [9, 1,0, 1] arelinearlydependent. (True or False)
54. If ltl, ?"2,1.3are the eigen values of a matrixA, then AB has the eigen values ......
55. If l, is an eigen value of a non-singular matrix A, then the eigen value ofA- 1 i, ....
56'The matrixcorrespondingtothe quadraticform
*2 +2y2 -722 -4ry+Bxz+Eyzis....
57, Cayley Hamilton theorem states that ........
58. If the rank of a matrix A is 2, then the rank ofA, is ........
59. The eigen values of a skew-symmetric matrix are real. (Tnre or False)
60. Inverse of a unitary matrix is a unitary matrix. (True or False)
61. A is a non-zero column matrix andB is a non-zero row mirtrix, then rank of AB is one. (True or False)
az. rra = f l I l,
O
,n"r,eigenvaluesofA- 1ar" .........
LJ J
es.uatrixI f forr = ...........
singular
lr x-?.,fis
^
-L' J
64.rfA=|_:n3 . ..
:il3],thenA3=

67. The index and signatureof the quadratic formfi + zxl- 3rAare respectively..j,"......
and .......
68. The matrix of the quadratic form g = 4t2 - 2y2+ zz - Zry + 6zx is ...........
 Vector Algebra & Solid Geometry
l. Vectors. 2. Space coordinates, Direction cosines. 3. Section formulae. 4-6. Products
of two
vectors. 7. Physical applications. 8-10. Products of three or more vectors. ll. Equation
of a plane.
12. Equations of a straight line. 13. Condition for a line to lie in a plane. la. Coplanar
lines. 18.
S.D. between two lines. 16. Intersection of thre_e planes. 1?. Equaiion of a sphere. lg.
Tangent
P-l-aneto a sphere'-19. Cone.2O. Cylinder. 21. Qularic surfaces. 22. Surfaces of Revolution. ZS.
Objective Tlpe of Questions.

\MCTOR ALGEBRA

3.r.(t) vEcToRs
A quantity which is completely specified by
its magnitude only is called a scalar. Length,
time, mass' volume, temperature, work, electric charge and numerical data in Statistics are all
examples of scalar quantities.

Weight, displacement, velocity, acceleration and electri..orrunt density are all vector quantities
for each involves magnitude and direction
A vector is represented by a directed line segment. Thur Pa represents a vector whose
magnitude is thelength PQ and direction is from p (starting pointl to
e (end point). we denote
a vector by a single letter in capital bold type (or with urro* on it) uni itt mlgnitude gength)
by the corresponding small letter in italics type. Thus".tif v is the
velocity vector, its magnitude is u, the speed. ny'
A vector of unit magnitude is called a unit uector. The idea of
unit vectors is often used to represent concisely the direction of
anyovector. Unit vector corresponding to the vector A is written
as A.
A vector of zero magnitude (which can have no direction
associated with it) is called a zero (or nulr) uector and is denoted
by 0-a thick zero.
,/" /t
The vecto, Q? represents the negative of Fe i.e. -A.
Two uectors AandB hauing the same magnitud,e and, the sdme Fig.3.1.
(or parallel) directions are said to be equal and we write
-) -) A = B.
clearly the vectorsAB, LM andpQ are all equal. (Fig. 3.1).
(2) Addition of vectors. vectors are added according
to the
triangle law of addition, which is a matter of common knoivledge.
-+ -+
Let A and B be represented by two vectors OP and PQ
-)
respectively then OQ = C is called the sum or resultant of
A and B. Symbolically, we write, A
C=A+B Fig.3.2
8l
 HIGHER
ENGINEERING
MATHEMATICS

(3) Subtraction of vectors. The subtraction of a uectorB from Ais taken,to be the addition
of -B to A and we write
A+(-B)=A-B
(4)Multiplicati.on of vectors by scalars.
Wehavejust seenthat A + A = 2A
and - A + ( -A)= -2A;
where both 2^A'and - 2A denote vectors of magnitude twice that of A ; the former having the
sarne direction as A and the latter the opposite direction.
In general, the product mA of a uector A and a scolar m is a uector whose magnitude is m
times that of A and direction is the same or opposite to A according a,sm is positiue or negatiue.
Thus
Example 3'1. If A and B are the uectors determined by twa
adj acent sides of a regular hexagon. What are the uectors
represented by the other sides taken in order ?
Sol. Let ABCDEF be the given hexagon, such that
-)
AB =AandBC=B
-) -) -t
AC = AB + BC=A+B
-) +
Also AD - 2BC -- 28
-+ -) -)
CD = AD - AC=2g--(A+B)_B-A
-+ + A
Now DE =-AB--A Fig.3.3.
t'.' AB = ard ll gOl
-+ -+
EF _ _BC _ -B I ' . ' B C= a n dl l r n l
--) ->
and FA--CD =-(B-A):A-B t...CD=andllAn
3'2. (1) Space co-ordinates. Let X'OX andY'OY,Z'OZ be three mutually perpendicular
lines which intersect at O. Then O is called the origin.
XOX is called the x-axis, YOY the y-axis, Z'OZ the z-axis and taken together these are
called the co-ordinate axes.
The plane YOZ is called the yz-plane, the plane ZOX
L the zx-plane, the plane XOY the xy-plane and taken
together these are called the ca-ordinate planes.
) Let P be any point in space. Draw PL, PM, PN Is to
the yz, zx and ry-planes. Then IP, MP, NP are
respectively called the coordinates ofP (Fig. 3.4). Thus the
co-ordinates of any point in space a,re the perpendicular
distances from the yz, zx and xy planes respectiuely. In
practice, to find the co-ordinates of a point P draw PN L
to the ry-plane ; from N draw NA I to the r-axis (Fig.
3 4). If OA = x, AN = y, NP = z, thett (x, y, z) are the
Fig.3 4. co-ordinates of P.

The x, y, z co-ordinates are positiue along OX, OY, OZ respectiuely and negatiue along
ox, oY,oz'.
The three co-ordinate planes divide the space into eight compartments called octants. The
octant OWZ in which all the co-ordinates are positive is called the positive or first octant.
 vEcToRRLoeeRR
& soLtDGEoMETRv 83
Note. As a right-handed screw moving in the direction of OX rotates from OY towards OZ, therefore,
the above system of axes is called a right-handed system. The left-handed system will, however, be
obtained by interchangmg the axes of r andy.
In general,three non-coplanar uectors A, B, C are said, ta
form, a right-handed (or a left-handed) sysfem accord,ing as
a right thread,ed screat rotated. through an angle less than
180' frpm Ato B will aduance along (or opposite to) C as shown
in Fig. 3.5.
An area of a closed curue described in a giuen manner is
represented by a uector whose magnitude is the giuen area and
direction normal to the plane of the area.
Thus vector A representing the area is taken to be positive
or negative according as the direction of description of the
boundary of the curve and the sense of A correspond to a Fig. 3 5.
right-handed or a lelt-handed system.
We have explained the most commonly used system of
co-ordinates namely t}eeRectbngular cartesian Co-ordinates. The C A'

I
a
other two systems of co-ordinates often used to locate a point in
space are the Polar Spherical co-ordinates and Cylindrical
co-ordinates, which are explained in $ S.21 and 8.20.
(2) B,esolution of vectors. Let I, d, K denote unit
vectors along OX, OY, OZ rcspectively. Let P(*, y,z)be a
point in space. On OP as diagonal, construct a rectangular
B'
Id,_t'
P(x,y,z)

parallelopiped with edges OA, OB, OC along the axes so

that
-+ -) B
+ "{
O A = xI, OB - yJ, OC = zK
-)
'a
-) -+
Then R - - OP = OC' + c'P A
-) -) -) +-+ C,
+
OA + AC' OC -O A + OB +oc Fig.3.6.

H e n c e R = r I + y J + z K is called the posltion uector of P

relative to origin O and
r= | R | ={(*2 +y2 +"21
l' .' rz = OP2= OC2 + C'Pfr= OA2+ Ag2 + C'PZ
(3) Direction cosines. Let any line,L or its parallel Op,
make angles cr, B, y with OX, OY, OZ respectively, then
cas d, cos B, cosy are eq,Iled the direction cosines of this line
which are usually denoted by I, rrt, n.
Fig.3.7.
If I, m, n q,redirection cosinesof a uector R, then
( t ) f t ,= I l + m d + n ] K , ( i i ) P + m 2 + ' n 2 =1
Proof.Let D be the footof the perpendicular fromP(x,y,e) on OI1 Then
l = O D = t c o s 9 = m r . S i m i l a r l y z= n r a n d x = l r .
R = rI +yJ + zK = r(II + mJ + nKi)
nr.r
or R=I=lI+mJ+nK
r
which expresses a unit vector in terms of its direction cosines.
Also 1 = 1 f i ,l = { ( 1 2+ * 2 + n z ) T h u s 1 2 + m r + n 2 = l
t,e. c o s 2 c + . o r 2 B + c oTs=2L
 HIGHERENGINEERINGMATHEMATICS

Obs. Direction ratios. If the direction cosines of a line be proportional to a., b, q then these are called
proportional direction cosines or direction ratios of the line.
If the direction cosines be l, m, n, then
! =ry -L -"1(12-+'n2+ nz)= -1-
a b-c ^ l @ 2 + b 2 + c \{ G o r )
I=-3 ,,*=-1,r=-4-
./(I4')
V(In") ./(I0')
(4) Distance between two points P(x1, yt,zL) and Q ()c2,yz,z2) is

./t(xz - xr)z + (Vs- yr)2 + (zz- ,r)21

P (xpypzl Q k2,y2'z)
and. direction ratios iA cr.r€x2- Xl, Vz - Vrr zz- zr
"f
We have
-?
OP = trll +ylJ + z1I(
-+
and oQ = xzl + yZJ+ z2K
+ '-+ -)
P8 - o Q - o P
- (xz-rr)I + (!z-yr)J + (zz-eilK 'X
Fig.3
Thus d= | Fq t =l[@z-xiz +(rz-yiz + (zz-"izJ

and directioncosin",of ,fr are proportionalto x2-xL,lz-!t,zz-zI.

Example 3.2. Show that the pointsA(-4,9,6),8(- 1,6,6) and C(A,7, '" 1A) form a
right
angled isoscelestriangle. Also find the direction cosinesof ,48.
Sol. We have AB =.lt(-l + 4)2+ (6 - 9)2+ (6 - 6)21= 3r/2
BC -^/t(0 + 1)2+ (7 - qz + (10- 6)21=3112
and CA ={(- + - 0)2+ (9 - 7)2+(6 - Lo)21
=e
Since AB2+BC2=CAz and,AB=BC, itfollows that AABC is a right-angled isosceles tri-
angle. The direction ratio, of.& are - 1 + 4, 6 - 9, 6 - 6.
-1 - 1 ^
.'. Its direction cosines
"r. lt, 15' u'
3.3.SECTION
FORMULAE
The point R (x, y, z) diuiding the join of the points A(rr, !r, zt) and B(x2, !2, zz) in the ratio
m 1 : r n 2i s
m t' B + m 2
-A (^r*r* m2xc1
mflz+ mzlt mFz+ m,p)
R-
m1tm2
r.e.
-f-, ' mL+mz ' mL+m, ...(r)
| )
Let P(A) and Q(B) be the given poinis referred to origin O. Let R(R) be the point dividing
the line joining P and Q in the ratio uL1: trl2 so that

PR tn1
i . e .m 2 - P R = m r . R Q
M=d
.'.we have *zfil = *rF,Q
-+-)J->
m2 (OR - OP) = trlt (OQ - On)
 85
ALGEBRA& SOLIDGEOMETRY
VECTOR

or mz(R- A) = rn1(B - R)

' + rn2A
mtB
whence R-
WL1* nn2

SinceA = rrl +yrJ t z1K,B = xzl + yzl + z2I(

and R=rI+YJ+zK
m {x2l + yzl + z2K)+ m L@1I+ YLJ + z tI()
xI+yJ +zK
o
(l). Fig. 3 9.
Equating coeffi.cientsof I, J, K, we get the desiredresults
Cor. 1. Mid'pointofP(A)and'Q(B);' j tn * nl'
m1B - m2A
join in the ratio m1 : rn'zexternally is Ft,-- '
2. Point.R dividing the of P(A) and Q(B) nf n,
the sum of the coefficients of A, B and
obs. Rewriting G) as mzL+ myB - (mr+ m2)'R= 0, we note that
ra R is zero.
B and C are colinear if
Hence it follows that any three points with position uectors Aa
^{
lA + pB + TC = 0, where l. + P + = 0'
Example 3.3. In a trapezium, prove that the straight line C B-C B
joining thi mid,-points of the diagonals is parallel to the
parallel sides and half their difference'
sol. consider a trapezium oABC with parallel sides
oA and BC. Take o as the origin and let the other vertices
be A(A), B(B), C(C). OA
'
Since CB is parallel to OA, therefore, Fig.3'10.
B-c =CB-?roa=l.A.
The mid-points of the diagonalsoB and AC arc D(B/z) and E(A + C)/2'
- dE - Ci)-*,o+c)-f,n=*to-(B-c)l
D'E " '(')

= - r.)A ...(ii)
*tt
=|foe-CB)'
From (iil, itis clear that iE is parallel to da; from (i), it follows that DE
Hencethe result.
Exarnple 3.4, Show that the line ioining one vertex of a
parallelogiam to the mid,-point of an oppasite side trisects the
diagonal and is itself trisected there at-
sol. consider a parallelogram oABc. Take o as the origin
and let the other vertices be A(A), B(B) and C(C)'
The mid-point D of OA is A/2.
DA
Now since OA is equal to and parallel to CB'
F i g .3 1 1 .
C A- C n ,i . r .A = B - c
which maybewritten ^rryp|]&= = P sothat P trisectsDC andoB'
?
86 HIGHER
ENGINEERING
MATHEMATICS

1. Given Rr = 5I - 2J + 4K and Rz = I + 3J + TK,Iind the magnitude and direction cosines of the vectors
R1 + R2 and 2R1 - R2.
2. Show that the points (0, 4, 1) ; (2,3, - 1) ; (4,5,0) and (2, 6,2) are the vertices of a square.
(Osmania, 1999 S)
3. A straight line is inclined to the axes of r and y at angles of 30" and 60o. Find the inclination of the
line to the z-axis (Madras,2003)
4. If a line makes angles d, F, y with the. axes, prove that
(i) sin2 o +' sin2 B + sinz ^{= 2. (V. T.U., 2000 ; Osmania, 1999)
(ll) cos 2cr+ cos 2B + cos 2^{= - L.
5. If A andB are non-colinearvectorsand P =(?,rc+3y-2)A+ (3r+ ?.y+5)B
and Q = (- r + 4y -2)A+ (3x - 4y + 7)8, find r,y such that 7P = 3Q.
6. Prove that the line joining the mid-points of the two sides of a triangle is parallel to the third side
and half of it.
7. Prove that (l) the diagonals of a parallelogram bisect each other ;
(il) a quadrilateral whose diagonals bisect each other is a parallelogram.
8. In a skew quadrilateral, prove that : .
(i) the figure formed by joining the mid-points of the adjacent sides is a parallelogram.
(ii) the joins of the mid-points of opposite sides bisect each other.
9. In a trapeziurn, prove that the straight line joining the mid-points of the non-parallel sides is parallel
to the parallel sides and half their sum.
1 0 . P r o v e t h a t t h e v e c t o r sA = 3 f + J - 2 K , 8 - -I+3J+ 4 I d .C
, -4I-2J-BIdcan form the sides of a
triangle. Also find the length of the median bisecting the vector C. @.N.T.U., 19e5 S)
11. Find the ratio in which XOY plane divides the join of the points (- 3, 4, - 8) and (5, - 6, 4) and thus
write the co-ordinates of the point of division.
'(2,
12. Find the ratio in which the line joining 4, 16) and (3, 5, - 4) is divided by the plane
2x-3y *z *6=0. ( M y s o r e ,1 g g 5 )
13. Show that the three points | - 2J + 3K, 2I + 3J - 4K, - 7J + 10K are colinear.
14. If A' B, C be the position vectors of the vertices A, B, C of the triangle ABC, show that the three
(l) medians concur at the point
*fA * B + C), called tinecentroid,.

(fi) internal bisectorc orrtr" ."*,tr. concur rn" point 4##, called theincentre.
"r
15. Show that the co-ordinates of ttre centroid of the triangle
whose vertices are (r1, !t, zt), (x2,yZ, zZ), (xg, yB,zg) are

lx1+ x2*xs !t+ lz+ lz z1+z2+ zs]

B '----s-' s
L J
16. Show that the co-ordinates of the centroid of the tetrahedron
whose vertices are (r,' , !, , zr) i r = 1,2,3, 4, ate
[ t , + ) c 2 x+s * x 4 )L, l,a t + 1. .l
ILz= t' ***tl l z + t z + y q ) ;n @ t +z 2 *z s * r o ) 1 .
[Def. A tetrahedron is a solid bounded by four triangular
faces. Thus the tetrahedron ABCD has four faces-the As
ABC, ACD, ADB, BCD. (FiS. 3'12).
It has four vertices A, B, C, D and three pairs of opposite F i g .3 1 2 .
edges AB, CD ; BC, AD ; CA, BD.
Th,ecentroid of the tetrahedron diuides the join of each uertex to the centroid of the opposite triangular
face in the ratio 3 : 11.
L7. M and N are the mid-points of the diagonals AC and BD respectively of a quadrilateral A.BCD. Show
that the resultantof the vectorsfB, fD,C?, & is adN. (Cochin,rggg)
VECTORALGEBRA& SOLIDGEOMETRY 87
3.4.PRODUCTS
OF TWOVECTORS
Unlike the product of two scalars or that of a vector by a scalar, the product of two vectors
is sometimes seen to result in a scalar quantity and sometimes in a vector. As such, we are led
to define two types of such products, called the scalar product and the uectorproduct respectively.
The scalar and vector products of two vectors A and B are usually written as A. B and
A x B respectively and are read as A dot B and A cross B. In view of this notation, the former
is sometimes called the dot producf and the latter the crossproduct.
In vector algebra, the division of a vector by another vector is not defined.

3,5.SCA!.AROR DOTPRODUCT
(1) Definition. The scq,lar or dot praduct of two uectors
A and B is defined as the scalar ob cos 0, where 0 is the angle
betweenA and B.
Thus A . B = o b c o s0 .
(2) Geometrical interpretation. A.B ls tlrc product of
the length of one uector and the length of the projection of the
other in the direction of the former.

Let o-L - A. o? = B then Fig.3.13.

A. B =a6 cos0 - a(OMcos0) =a(Olto- I A I Proj.of I B I in the directionof A.

Similarly, A. B= I B I Proj. of I A I in the directionofB.

(3) Properties and other results.
I. Scalar product of two uectorsis commutatiue
A . B : B . A f o r A . B = a b c o s0 = b a c o s( - 0 )= B . A
II. The necessaryand sufficient condition for two uectors to be perpendicular is that their
scalar product should be zera.
When the vectors A and B are perpendicular, A. B = ob cos g0" = 0.

C o n v e r s e l y w h e n A . B = a , a b c o s 0 = 0 i . e . c o s0 = 0 . ('.' a*o,b*0), or0=90'.

-
m. L. A o2 which is written as ,4.2.Thus the square of a uector is a scalar which stand.s
for the squa,re of its magnitude.
IV. For the mutually perpendicular unit vectors f, J, K, we have the relations
I.J=J.K=f{.I=0
and 12=JZ
which are of great utility.
V. Scalar product of two uectorsis distributiuei.e.
(A-iB).C=A.C+B.C
F i g .3 1 2 1 .
VI. Schwarz inequality* : I A. B I s I A | | B I

lA.Bl=lAl lBl lcoselslAl lBl t'.'lcos0lsll

VtrI. Sealar product of two uectors is equal to the sum of the products oi their corcesponding
components.
* Named after the German mathematician HermannAmandus Schwarz (1843-1921) who is known for
his work in conformal mapping, calculus of variations and differential geometry. He succeededWeierstrass
in Berlin University.
88 IJIGHER MATHEMATICS
ENGINEERING

For ifA = aLI + a2J * asK,,B= brl + bzJ+ bBK

then by the distributive law, A . B = atbr + a2b2+ asbs
- otz + a22+ ag3.
In particular, .AtZ
VAL Angle between'"t#o lines whose direction eosinesare l, m, n and l',n'L',n'is
cos-t (u'*mm'+nn').
The unit vectors in the directions of the given lines are LJ - ll + mJ + nK and
U'=l'I+m'J+n'K.
If 0 be the angle between the lines, then
" U . tJ' = (lI + mf, + nK) . (I'l + m'J + n'K)
or 1 . 1 . c o s0 - l l ' + m n 1 . ' + n n '
Hence cos0=II'+mm'+nn' ...(t)
-
Cor. 1. sin2 0 = | cos2e = L - (ll' + mm' + nn'\2
* (12+ ^2 + 12111'2 + m'2 + n'2) - (Il' + ntm' + nn')z
- (mn' - nnx)z+ (nl' - ln')z,+(lrn' - mf)z
sin 0 = + .,lz(mn'- nrn)z. ...Gi)
Cor. 2. The condition that the lines whose direction cosinesare l, ffi, ft and l',m',n' shouldbe
perpendicular is
l I' + mm' + n n ' = 0 ...Qii)
and parallel is l=l'r11l=r1'rn=n' ...Gu)
These conditions easily follow from (i) and (ij).
Cor. 3. The angle 0 between two lines whose direction ratios dre d, b, c and &', b', c' is giuen by
a-a'+-bb'+cg'
cosg=
tr(Io") !(h''\
^l[(bc' - cb')z+ (ca' - ac')Z+ (ab''- ba')z
or sin0=
{(ro,).,1(>,'r)
Theselines are (i) perpendicular if aa' + bb' + cc' = 0, ftil parallel if a/a' = b/b' = c/c'.
IX. Proiection of the line joining two points (xr, yt, z) and (x2,!2, z2) on a line whose
direction cosines are l, m, n is
'/
l(xz-rr) +m!z-yr) +n(zz-zi.

Let dp = r1I + yrJ + de= xzr+ yzJ + z2]f-

-+ "tr.,
PQ = @z-r1)I + (yz- y r)J + (zz- zr)K
Also unit vector Lf along the given line is II + mJ + nK.

.'. Projection ofPQ on the given line - ,i .U

= l(xz- f r) + m$z-yr) + n(zz- zi.
Example 3'5. Find the sides and angles of the triangle whose uertices
arel-2J +ZIf.,2I+J -IK,and 3I-.J +2I{.
l-+-)
Sol.Let OA =l-ZJ +2K, OB=21+J-If-,OC = 3I-J+2K
-)
Then BC=I-2J+3K
-+
CA - _zT-J
+
and AB = I+3J-BK
BC - rlt4, CA - {5,.4.8 =.h9. Fig.3'15.
 89
VECTORALGEBRA& SOLIDGEOMETI?Y

Now d.c.'sof AB and AC being

L/lLg,3/^lrg, - S/'lLg and2/"15,l/^15,0,
wehavecosA=#+.# +.# o=r/(szre)
I
i.e. ZA=.or- {.ftt}). Again d.c.'sof BC andBA being
L/4L4, - 2/\1L4,ehtu and - l/lLg, - 3/llg,3/.llg ;

wehave cosB =# = "'l!+tte)i'e'tB ='o'- 1 l}+ttg)

fi; . #,' #.# #
- - g{la
Finally d.c.'sof CAandCB being- 2/^15,- L/15,0 and L/\G ,2/{G, ;

wehave cosc =-G =0 i'e'tc =e0"'

#. # h.0 !*
Example 3.6. Proue that the right-bisectors of the sides of a
triangle concur at its circumcentre.
sol. LetA(A), B(B), C(C) be the vertices of any triangleABC. The
mid-points of the sides BC, CA and AB are

n(u4s],"[q#],"(o+ B)
(-'-)
"
U'!,, \ )
Let thd perpendiculars at D and E to BC andCA resPectivelY
intersect at the point P(R). Then ip . iC = 0 Fig.3.16.

...(t)
i.e. ["-t#1.(c-B)=o
\ -, 'o, q#].
= r.n.f"- (A- c)=o . . . ( i)i
and EP . CA - -, ,.o.,1.^-
z,
)
Adding (i) and (ii), we s" - (A- B) = 0
[R T].
which shows that FP is p"rpenhicular to AB. Hencethe result.
Further PA=PBif lA-R l= lB-R I
or if, (A-R)2 = (B - R)2 or if, A2 -2A. f,,= 82 -28 . R

or if, - ^*l . (A - B) = 0, which is true.

t'"
\)
ExamplL g.7. If the d.istancebetweentwo points P and Q is d and the lengths of the
projectionsof PQ on the coard.inateplanes are d,1,dz, ds, showthat 2d'2= d'1+ a?2+ a?s'
Sol. Let P be (xr, yt,e1) and I b* (xz,y2, z2),then
d 2 - ( x t - x z ) z + ( y r - y z ) z+ ( z t - " 2 ) 2 .
The feet of the perperdiculars drawn from P and Q on the Xts-plane are the projectionsof
P and Q on this plane. If theseare I, and M, then L is (r1,3lr,0) andM is (r2,j,'2,0)'
d1 = projectionof PQ onXf-plane i-e-LM
or d'21-(x1* xz)z+ (Yt- Yz)z
-
Similarly d?,=(yt-yz)z + @t-"2)2 and d! -@t-"2)z + (rr xz)z
d? * aZ + aZ - 2 [(xt - *z)z + (vr - vz)z+ (zr - zz)2]= 2d'2'
Example 3.8.A tine mukesangles0, F, % 6 with diagonalsof a cub|, prouethat
, o r 2a . + c o s ? y+ror2 6-4/3. (V.T.U.,2006; Osmania,200AS)
B+cos2
90 HIGHER MATHEMATICS
ENGINEERING

Sol. Take O, a corner of the cube as origin and OA, OB,

OC the three edges through it, as the axes. z
C(o,o,a)
Let OA = OB = OC - a. Then the co-ordinates of the A,(n o. a)
corners are as shown in Fig. 3.L7. /
The four diagonals are B' Y(a, a)
(a,a,a
a)
OP, AA', BB' and CC'.
Clearly direction cosines of OP are
a-0 a-0 s_0 L 1 1
'lltu\'^/G"5''e'!E'!5'm' /,
)(o,o,o) |t . /^/B(o,o,o
{Gd1' V
&ro, a) C'(a,a,o)
Similarly direction cosinesof AA'are - +, I.iB'! ''
lu' {g F i g .3 1 7 .
111
BB' are -*.
vrj' r/g' r/B '

and " " " CC'atg#r#,-+

Let I, ffi, fl be the direction cosines of the given line which makes angles o, F, y,6 with
OP, AA', BB' , CC' respectively. Then
L., 1
cosu,=tr (I + m + n) ;cosp = (- I + m + n)
fr
!., -
cosT=fr (I m+ n); cos6=tr1 . (I
,
+ m - n)
Squaring and adding, we get

y*.or2 6 =+ [( + m+ n)z+ el +nx+ n)z+ (t -m + n ) 2 + ( t + * - n ) 2 1

,or2s *.or2 B + cos2
L= + 4
*2 + n211
wa' *2
a)
=?. f.'.'P+*2+n2-ll
a)

Example 3.9. If the edges of a rectangular parallelopiped are a, b, c, show that the angle

between
thefour d.iasonats 1( -
s.re
"or- "4#21 { )
-+0-+c- ) z
[
Sol. Let OA-a,OB=b, OC =c be the edges of the C(n.n..l A'(o,h,r)
rectangular parallelopiped. Then the coordinates of the corners
are as shown in Fig. 3.18. The four diagonals taken in pairs are B, Pt",r,/
(i) (OP, AA'), (ii) (OP,BB'), (iii) (OP,CC'), (iu) (,4g|',BB'), (u) "(a'o'c c
(AA', CC') and (ui) (BB', CC'). b Y
B(o,
Let the angles between these pairs of diagonals be ,,/oto.r,ot
0 1 ,02, . . . 05 r es pecti v e l y .C l e a rl y d .r.' s O P a rc a ,b, c; d.r.' s,of
(a,o,o) C'@,b,o)
AA' ate - a, b, c, d.r.'s of BB' are d.,- b, c and d.r.'s of CC' are x
a',b,-c'
Fig.3.18.
.'. For the pair (i) i.e. (OP, AA');
-o2+b2+c2 -o2+b2+c2
cos 01 =
^ l@2' + b 2 + r 2 1 1 1 o+2b z + c 2 1 o- 22+ b, 22+ c2
^ a 2- b 2 + c 2 ^ o 2+ b 2 - c 2
Similariy c o su 2 = c o su 3=
o 4 U z* ; ; d\b2 +cz;
o2 -b2 + 12 a2+b2-12
cosu4= cosu5=
ol*rz*;; oz*U4S;