MISRIMAL NAVAJEE MUNOTH JAIN ENGINEERING

COLLEGE, CHENNAI - 97

DEPARTMENT OF MATHEMATICS

MATHEMATICS (MA2111)
FOR

FIRST SEMESTER ENGINEERING STUDENTS

ANNA UNIVERSITY SYLLABUS

This text contains some of the most important long answer questions (Part B) and
their answers. Each unit contains 15 university questions. Thus, a total of 75
questions and their solutions are given. A student who studies these model
problems will be able to get pass mark (hopefully!!).

Prepared by the faculty of Department of Mathematics

SEPTEMBER, 2008

www.engg-maths.com
Matrices 2

UNIT I MATRICES

Problem 1. Find the eigen values and eigen vectors of the matrix
  2 2  3
A   2 1  6
  1  2 0 
Solution:
The characteristic equation is | A - I | = 0.

2- 2 3
i.e., 2 1-  6  0
1 2 0-
i.e., (-2 - ) [-(1 - ) -12] - 2[-2 - 6] -3[-4 + 1 - ] = 0
i.e., (-2 - ) [2 -  -12] + 4 + 12 + 9 + 3 = 0
i.e., 3 + 2 - 21 - 45 = 0 (1)
3 2
Now, (-3) + (-3) - 21(-3) - 45 = -27 + 9 + 63 – 45 = 0
 -3 is a root of equation (1).
Dividing 3 + 2 - 21 - 45 by  + 3
3 1 1  21  45
0 3 6 45
1  2  15 0
Remaining roots are given by
2 - 2 - 15 = 0
i.e., ( + 3) ( - 5) = 0
i.e.,  = -3, 5.
 The eigen values are -3, -3, 5
 2  λ 2 - 3  x1  0
The eigen vectors of A are given by  2 1 - λ - 6   x 2   0
 - 1 - 2 - λ   x3  0
Case 1  = -3
 2  3 2 - 3  1 2 - 3

Now  2 1  3 - 6 ~  2 4 - 6

 - 1 -2 3  - 1 - 2 3 
1 2 - 3
~ 0 0 0 
0 0 0 
Matrices 3

 x1 + 2x2 - 3x3 = 0
Put x2 = k1, x3 = k2
Then x1 = 3k2 - 2k1
3k 2  2k 1 
 The general eigen vectors corresponding to  = -3 is  k 1 

 k 2 
3
When k1 = 0, k2 = 1, we get the eigen vector 0
1
 2
When k1 = 1, k2 = 0, we get the eigen vector  1 
 0 
 3  2
Hence the two eigen vectors corresponding to  = -3 are 0 and  1  .
 
1  0 
These two eigen vectors corresponding to  = -3 are linearly independent.
Case 2  = 5
  2  5 2 - 3   7 2 - 3
 2 1 - 5 - 6 ~  2 - 4 - 6


 - 1 - 2 - 5  - 1 - 2 - 5
 1  2  5 
~  0  8  16
 0 0 0 
 -x1 - 2x2 - 5x3 = 0
-8x2 - 16x3 = 0
A solution is x3 = 1, x2 = -2, x1 = -1
  1
 Eigen vector corresponding to  = 5 is  2 .
 1 

 1 1 2 
Problem 2. Find the characteristic equation of  2 1 3  and verify Cayley-
 3 2 3
Hamilton Theorem. Hence find the inverse of the matrix.
Matrices 4

 1 1 2 
Solution: Let A   2 1 3   Characteristic eqn. of A is
 3 2 3
   1  1  3    9  9  1  26  0
3 2

i.e  3   2  19  26  0
By Cayley-Hamilton theorem  A3  A2  19 A  26 I  0 .

Verification:

 1 1 2  1 1 2   9 2 7 
    
 A2  A. A   2 1 3  2 1 3    5 9 10 
 3 2 3  3 2 3   10 7 21 
    
 9 2 7  1 1 2   16 21 45 
    
 A  A .A   5
3 2
9 10  2 1 3    43 16 67 
 10 7 21  3 2 3   67 45 104 
   
Substituting in the characteristic equation
 16 21 45   9 2 7   19 19 38   26 0 0   0 0 0
         
 43 16 67    5 9 10    38 19 57    0 26 0    0 0 0
 67 45 104   10 7 21   57 38 57   0 0 26   0 0 0 

Hence verified.
Now to find the inverse of the matrix A, premultiply the characteristic equation by A1
 A2  A  19 I  26 A1  0

 A 1 
1
26
19I  A  A2 
19 0 0   1 1 2   9 2 7    9 5 5 
1       1  
 0 19 0    2 1 3    5 9 10     3 9 7 
26  26
 0 0 19   3 2 3   10 7 21    7 5 1
 

1 0 3
Problem 3. Given A  2 1  1 , use Cayley-Hamilton Theorem to find the inverse of A

1  1 1 
and also find A4
Solution:
The characteristic equation of A is
1 λ 0 3
2 1 λ 1  0
1 1 1 λ
i.e., (1-) [(1 - ) (1 - ) -1] + 3[-2 - (1 - )] = 0
Matrices 5

i.e., (1 - )3 - (1 - ) – 6 -3 + 3 = 0
i.e., 1 - 3 + 32 - 3 – 1 +  - 9 + 3 = 0
i.e., -3 + 32 +  - 9 = 0
i.e., 3 - 32 -  + 9 = 0
By Cayley-Hamilton theorem, A3-3A2 – A + 9I = 0
To find A , multiplying by A , A2-3A - I + 9A-1 = 0
-1 -1

1
 A-1 = [-A2 + 3A + I]
9
1 0 31 0 3  4  3 6
A  2 1
2
 1
2 1  1  3 2 4
   
1  1 1 
1  1 1  0  2 5
 4 3  6  3 0 9  1 0 0
1
1
A   3  2  4  6 3  3  0 1 0
 
9
 0 2  5 3  3 3  0 0 1
0 3 3 
 3 2  7 
1
9
3  1  1 
To find A4:
We have A3- 3A2 – A + 9I = 0
i.e., A3 = 3A2 + A - 9I (1)
Multiplying (1) by A, we get,
A4 = 3A3 + A2-9A 
= 3(3A2 + A - 9I) + A2 - 9A using (1)
= 10A2 - 6A - 27I
 4  3 6  1 0 3 1 0 0
 10 3 2 4  6 2 1  1  27 0 1 0
   
0  2 5 1  1 1  0 0 1
 7  30 42
  18  13 46
  6  14 17 

 0 0 2
Problem 4. . If A   2 1 0  express A6  25 A2  122 A as a single matrix
 1 1 3 
 
Solution: To avoid higher powers of A like A6 we use Cayley Hamilton Theorem.
Characteristic equation is  3  4 2  5  2  0
By Cayley Hamilton Theorem A3  4 A2  5 A  2 I  0
To find A6  25 A2  122 A we will express this in terms of smaller powers of A using the
characteristics equation. We know that (Divisor) X (Quotient) + Remainder = Dividend
Assuming A3  4 A2  5 A  2 I as the divisor we get,
Matrices 6

A3  4 A2  11A  22 I
A6  0 A5  0 A4  25 A2  122 A  0 I
A3  4 A2  5 A  2 I
A6  4 A5  5 A4  2 A3
4 A5  5 A4  2 A3  25 A2  122 A
4 A5  16 A4  20 A3  8 A2
11A4  22 A3  33 A2  122 A
11A4  44 A3  55 A2  22 A
22 A3  88 A2  100 A
22 A3  88 A2  110 A  44 I
10 A  44 I

 A6  25 A2  122 A   A3  4 A2  5 A  2 I  A3  4 A2  11A  22 I    10 A  44 I 

But A3  4 A2  5 A  2 I  0
A6  25 A2  122 A  0  10 A  44 I
  10 A  44 I 
 0 0 20   44 0 0  
   
   20 10 0    0 44 0  
 10 10 20   0 0 44  
 44 0 20 
 
   20 54 0
 10 10 74 

 44 0 20 
 
   20 54 0 
 10 10 74 


Problem 5. If i are the eigen values of the matrix A, then prove that
i k i are the eigen values of kA where ‘k’ is a nonzero scalar.
ii. im are the eigen value of Am and
1
iii. are the eigen values of A1 .
i
Solution: Let  i be the eigen values of matrix A and Xi be the corresponding eigen
vectors. Then by defn: AXi   iXi......( I ) ( i.e by defn. of eigen vectors)
i. Premultiply ( I ) with the scalar k. Then
k  AXi   k  iXi 
i.e.  kA  X i   k i  Xi
 k i are the eigen values of kA (comparing with ( I ) i.e by defn.)
Matrices 7

ii. Premultiply ( I ) with A, then

A  AXi   A  iXi 
i.e. A2 X i  i  AXi 
 i  i Xi  from (I)
  i  Xi
2

1y we can prove that A3 Xi   i  Xi and so on Am Xi   i  Xi

3 m

  i m are the eigen values of the Am (comparing with ( I ) i.e. by defn.)

iii. Premultiply ( I ) with A1 , then

A1  AXi   A1  iXi 
i.e.  A1 A  Xi  i  A1 Xi 
i.e. IXi   i  A1 Xi 
1
i.e. A1 Xi  Xi
i
1
are the eigen values of A1 (comparing with ( I ) ).

i
2 0 1
Problem 6. Find the characteristic vectors of  0 2 0  and verify that they are
1 0 2
 
mutually orthogonal.

2 0 1
 
Solution: A =  0 2 0  Characteristic equation is  3  6 2  11  6  0
1 0 2
 
Solving:   1, 2, 3
Consider the matrix equation  A   I  X  0
Case (i) when   1;
 1 0 1  x1   0  1x1  0 x2  1x3  0  1
    
 0 1 0  x2    0  i.e. 0 x1  1x2  0 x3  0   2  equation (1) & (3) are identical.
 1 0 1  x   0  1x1  0 x2  1x3  0   3 
  3   
Solving (1) and (2) using the rule of cross multiplication
 1
x1 x2 x3 x1 x2 x3  
  i.e.    X 1   0 
0 1 0 1 0 1 1 0 1 1
 
Case (ii) when   2;
Matrices 8

 1 0 1  x1   0  0 x1  0 x2  1x3  0 x3  0
    
 0 1 0  x2    0  i.e. 0 x1  0 x2  0 x3  0 i.e. x2 is arbitrary say k
 1 0 1  x   0  1x1  0 x2  0 x3  0 x1  0
  3   

 0  0
   
 X 2   k  i.e  1  .
 0  0
   
Case (ii) when   3;
 1 0 1  x1   0   x1  0 x2  1x3  0
    
 0 1 0  x2    0  i.e. 0 x1  1x2  0 x3  0 Solving (1) and (2)
 1 0 1 x   0  1x1  0 x2  1x3  0
  3   
1
x1 x2 x3  
   X3   0
1 0 1 1
 
Thus the eigen values are 1,2,3 and the correspondent eigen vectors are
 1  0  1
     
 0  ,  1  and  0  . To check orthogonallity, X 1 X 2  0
T

 1  0 1
     
X 2T X 3  0
X 1T X 3  0
 X1, X 2 , X 3
are mutually orthogonal.
 6 6 5 
Problem 7. Find the latent vectors of 14 13 10 
 7 6 4 
 
Solution: Characteristic equation is    1  0    1, 1, 1
3

When   1 (repeated 3 times)  we have to find 3 corresponding latent vectors.

 7 6 5   x1   0  7 x1  6 x2  5 x3  0
    
14 12 10   x2    0  i.e. 14 x1  12 x2  10 x3  0 All three equation are identical
 7 6 5   x   0  7 x1  6 x2  5 x3  0
  3   
.i.e. we get only one equation, but we have to find three vectors that are linearly
independent.
0
x2 x3  
 Assume x1  0  6 x2  5 x3  0 i.e.  6 x2  5 x3 i.e.   X 1   5 
5 6 6
 
Matrices 9

 5 
x1 x3  
Assume x2  0  7 x2  5 x3  0 i.e. 7 x1  5 x3i.e..   X 2   0 
5 7 7
 
6
x1 x2  
And assume x2  0  7 x2  6 x3  0 i.e. 7 x1  6 x2 0i.e..   X 3   7 
6 7 0
 
X1, X2 and X3 are linearly independent.

 1 1 1
Problem 8. Find the eigen vectors of the matrix A   0 2 1
  4 4 3

Solution:
1 -  1 1 
The characteristic equation of A is  0 2- 1   0
  4 4 3 -  

i.e., (1 - ) [(2 - ) (3 - ) - 4] -1[0 + 4] +1[0 + 4(2 - )] =0

i.e., (1 - )(2 - 5 + 6 - 4) – 4 + 8 - 4 =0
i.e., (1 - )(2 - 5 + 2) + 4 - 4 =0
i.e., (1 - )(2 - 5 + 2 + 4) =0
i.e., ( -1)(2 - 5 + 6) =0
i.e., ( -1)( - 2)( - 3) =0
 The eigen values of A are  = 1, 2, 3.

1 - λ 1 1   x 1  0 

The eigen vectors are given by  0 2 - λ 1   x 2   0
  4 4 3 - λ   x 3  0

Case 1 =1
 0 1 1   4 4 2
 0 1 1 ~  0 1 1
   
 4 4 2  0 0 0
-4x1 + 4x2 + 2x3 = 0
x2 + x3 = 0
A solution is, x3 = 2, x2 = -2, x1= -1
  1
 Eigen vector X1 =  2
 2 
Matrices 10

Case 2 =2
  1 1 1  1 1 1
 0 0 1 ~  0 0 1
   
 4 4 1  0 0 0
-x1 + x2 + x3 = 0
x3 = 0
A solution is, x3 = 0, x2 = 1, x1 = 1
1
 Eigen vector X2 = 1
0

Case 3=3
 2 1 1  2 1 1 
 0  1 1 ~  0  1 1
   
 4 4 0  0 0 0
-2x1 + x2 + x3 = 0
-x2 + x3 = 0
A solution is, x3 = 1, x2 = 1, x1 = 1

1
 Eigen vector X3 = 1
1
2 2 0
Problem 9. Diagonalise the matrix  2 5 0  using orthogonal transformation.
 0 0 3
 
Solution: Characteristic equation is   10  27  18  0
3 2

Solving we get the eigen value as   1, 3, 6

 2  0 1
     
When   1, X 1   1  ; When   3, X 2   0  ; When   6, X 3   2 
0 1 0
     
 2   1 
 5  0  5
 1    2 
Normalizing each vector, we get 
5   
, 0 and 
5
  1  
 0   0 
   
Matrices 11

 2 0 1   2 1 0
 5 5  5 5 
 2  . N  NT   0
 Normalized Modal Matrix, N   1 1,
5
0 0
5  
   1 2 0 
 0 1 0 
 5 5 
 
Then by the orthogonal transformation,
 2 
 2 1 0  0 1
 5 5  2 2 0 5 5
  
N AN   0 0 1 2 5 0 0 0 2  . On simplifying, we get
 
 1   5
2 0   0 0 3   1
 5 5   1 0 
 5 
N AN  D  1 , 2 , 3 
1 0 0
  which is diagonal matrix with eigen values along the
 D 1,3, 6    0 3 0 
0 0 6
 
diagonal (in order).

 6 2 2 
Problem 10. Reduce  2 3 1 to a diagonal matrix by orthogonal reduction.
 2 1 3 
 
Solution: Characteristic equation is   12 2  36  32  0    8, 2, 2
3

When   8
 2 2 2   x1   0 
    
 2 5 1   x2    0 
 2 1 5   x   0 
  3   
i.e 2 x1  2 x2  2 x3  0
2 x1  5 x2  1x3  0
2 x1  1x2  5 x3  0
2
x1 x2 x3  
Solving any two equations    X 1   1 
2 1 1 1
 
When   2 (repeated twice)
 4 2 2  x1   0 
    
 2 1 1 x2    0  i.e 2 x1  2 x2  2 x3  0 . All the equations are identical.
 2 1 1  x   0 
  3   
Matrices 12

0
x2 x3  
To get one of the vectors, assume x1  0  x2  x3  0 i.e.   X 2   1 
1 1 1
 
a
 
X 1 X 2  0 . Therefore X 1 and X 2 are orthogonal. Now assume X 3   b  to be mutually
T

c
 
orthogonal with X1 and X2.
a 
  
X 1 X 3  0 i.e.  2 1 1  b   0 i.e.2a  b  c  0 
T

c  a
  b c
 i.e  
a  2 2 2
 
and X 2T X 3  0 i.e.  0 1 1  b   0 i.e.0a  b  c  0 
c 
  
1
 
 X3   1  .
 1 
 
After normalizing these 3 mutually orthogonal vectors, we get the normalized Modal
 2 0 1 
 6 3
 
Matrix N   1 1 1 
 6 2 3
 1 1 1 

 6 2 3
Diagonalizing we get

2 1 1   2 1 1 
 6 6 6   6 2 2   6 6 3
    
D  N T AN   0 1 1  2 3 1  1 1 1 
 2 2    6 2 3
 2 1 3  
 1
 1 1   1 1 1 
 3 3 3  6 3 3
on simplifying we get D  D  1 , 2 , 3 
8 0 0
 
0 2 0
0 0 2
 
 D  8, 2, 2 
Matrices 13

3 1 1 
Problem 11. Diagonalise the matrix A  1 3 -1
1 -1 3 

Solution:
3- 1 1 
The characteristic equation of A is  1 3- -1   0

 1 -1 3- 

i.e., (-1)(2 - 8 + 16) = 0

 The eigen values of A are  = 1, 4, 4.

3-λ 1 1   x1   0 
The eigen vectors are given by  1 3-λ -1   x 2   0 

 1 -1 3-λ   x 3  0 

Case 1 =1
 1
Eigen vector X1 =  1 
 1 
Case 2  = 4
0
Eigen vector X2 =  1
 1 
a
 
Now assume X 3   b  to be mutually orthogonal with X1 and X2.
c
 
X 1 X 3  0 i.e.  a  b  c  0 
T
a b c
 i.e  
and X 2 X 3  0 i.e.  b  c  0 
T
2 1 1
 2
 
 X3   1 .
1
 
 1 0 2 
Hence the modal matrix M   1 1 1 
 1 1 1 
Matrices 14

 1 0 2 
 3 6
 
The Normalized Modal Matrix is N   1 1 1 
 3 2 6
 1 1 1 

 3 2 6
Diagonalizing, we get

 1 1 1   1 0 2 
 3 3 3 3 1 1 
 3 6
   
D  N AN   0
T 1 1   1 3 1  1 1 1 
 2 2   
 3 2 6
 1 1 3  
 2
 1 1  
 1 1 1 
 6 6 6  3 2 6

1 0 0 
 0 4 0  = D(1, 4, 4)
0 0 4 

Problem 12. Reduce the Quadratic From 10 x12  2 x22  5 x32  6 x2 x3  10 x3 x1  4 x1 x2 into
canonical form by orthogonal reduction. Hence find the nature, rank, index and the
signature of the Q.F. Find also a nonzero set of values of X which will make the Q.F.
vanish.

 10 2 5 
 
Solution: Matrix of the given Q.F. is A   2 2 3  , which is a real and symmetric
 5 3 5 
 
matrix. The characteristic equation is   17  42  0
3 2

Solving, we get   0, 3, 14
1 1  3 
     
When   0, X 1   5  ; When   3, X 2  1 ; When   14, X 3   1 
4 1 2
     
and X 1 , X 2 , X 3 are mutually orthogonal since X 1 , X 2  0, X 2 X 3  0 andX 3 X 1  0
T T T

Normalizing these vectors we get the normalized model matrix

 1 1 3 
 42 3 14 
 
N   5 1 1 
 42 3 14 
 4 1 2 
 
 42 3 14 
Matrices 15

Diagonalising we get D  N T AN
 D  12 , 3  in order
 D  0, 3, 14 
0 0 0 
 
i.e D   0 3 0  (i.e. the eigen values in order along the principal
 0 0 14 
 
diagonal).
Now to reduce the Q.F to C.F (.i.e Canonical form)
 y1 
 
Consider the orthogonal transformation X = NY where Y   y2 
y 
 3
Then the Q.F. X T AX becomes  NY  A  NY   Y T  N T AN  Y
T

= Y T DY since N T AN  D
0 0 0  y1 
  
  y1 y2 y3   0 3 0  y2 
0 0 14  
  y3 
 0 y1  3 y2  14 y3
2 2 2

Thus = 0 y1  3 y2  14 y3 is the Canonical form of the given Q.F. And the equations of
2 2 2

this transformation are got from X= NY.

 1 1 3 
 x1   42 3 14   y1 
    
x  NY   5 1 1   y2 
 2
x   42 3 14  
 3  4 1 2   y3 
 
 42 3 14 

1 1 3
 x1  y1  y2  y3
42 3 14
5 1 3
x2   y1  y2  y3
42 3 14
4 1 3
x3  y1  y2  y3
42 3 14

To get the non-zero set of values of x which make the Q.F zero we assume values
for y1 , y2 and y3 such that the C.F. vanishes.
Matrices 16

i.e 0 y1  3 y2  14 y3 will vanish if y2  0, y3  0 and y1 is any arbitrary value (for

2 2 2

simplicity sake, assume y1 as the denominator of the coeff. of y1 in the equations) let
y1  42

 x1 
1
42

42 
1
3
0 
3
14
(0)

i.e. x1  1  0  0  1
III 1 y x2  5  0  0  5
and x3  4  0  0  4
Thus the set of values of x i.e 1, 5, 4  will reduce the given Q.F. to zero.
To find the rank, index, signature and nature using canonical form:
C.F. is 0 y1  3 y2  14 y3
2 2 2

 rank is 2 (no. of terms in C.F)

Index is 2 (no. of positive terms)
Signature of Q.F. = ( no. of positive terms) – (no. of negative terms) = 2
Nature of the Q.F. is positive semi definite.

Problem 13. Reduce the Q.F. 2 xy  2 yz  2 zx into a form of sum of squares. Find the
rank, index and signature of it. Find also the nature of the Q.F.
0 1 1
 
Solution: Matrix of the Q.F. is A   1 0 1 
1 1 0
 
Characteristic equation is   3  2  0 solving   2, 1, 1
3

1
 
When   2, X 1  1
1
 
When   1 (repeated twice) we get identical equations as x1  x2  x3  0
x x
x1  0  x2  x3  0 i.e. x2   x3 i.e. 2  3
1 1
Assume 0
 
 X 2  1
1
 
which is orthogonal with X 1.
a
 
Now to find X 3 orthogonal with both X 1 and X 2 assume X 3   b 
c
 
Matrices 17

if X 2T X 3  0, a  b  c  0 

if X 2T X 3  0, 0a  b  c  0 
a b c
i.e.  
2 1 1
2  2 
   
 X 3   1 i.e.  1 
 1  1
   
which is orthogonal with X 1 and X 2 .
 1 0 3 
 3 2 6
 
Normalising these vectors we get N   1 1 1  and D  N AN
 3 2 6
 1 1 2 
 
 3 2 6 
2 0 0 
 
= D  1 , 2 , 3    0 1 0  .Consider the orthonormal transformation X = NY
 0 0 1
 
such that Q.F.is reduced to C.F.

The Q.F. is reduced as

X T AX   NY  A  NY 
T

 Y T  N T AN  Y
 Y T DY
 2 0 0  y1 
  
  y1 , y2 , y3 ,   0 1 0  y2 
 0 0 1 y 
  3 

 The C.F. is 2 y12  y2 2  y32

rank of Q.F.is = no. of terms in C.F=3
index of Q.F. = no. of positive terms in C.F. = 1
signature of Q.F. = ( no. of positive terms) – (no. of negative terms)
= 1-2 = -1
Nature of the Q.F. is indefinite.

Problem 14. Reduce the quadratic form 8 x12  7 x 22  3 x32  12 x1 x 2  4 x1 x3  8 x 2 x3 to the

canonical form by an orthogonal transformation. Find also the rank, index, signature and
the nature of the quadratic form.
Matrices 18

Solution:
 8 6 2 
The matrix of the quadratic form is A   6 7  4
 2  4 3 
The eigen values of this matrix are 0, 3 and 15 and the corresponding eigen vectors are
1  2 2
 
X 1  2  , 
X2   1  , X 3   2 , which are mutually orthogonal.
2  2  1 

1/3 2/3 2/3 

The normalized modal matrix is N  2/3 1/3  2/3

 2/3  2/3 1/3 
0 0 0 
and N AN = D  0 3 0 
T

0 0 15
Now the orthogonal transformation X = NY will reduce the given quadratic form to the
canonical form 0y12  3y 22  15y 32 .
Also rank = 2, index = 2, signature = 2. The quadratic form is positive semi definite.

Problem 15. Find the orthogonal transformation which reduces the quadratic form
2 x12 
2 x 22  2 x1 x 2  2 x 2 x3  2 x1 x3 into the canonical form. Determine the rank, index,
2 x32
signature and the nature of the quadratic form.

Solution:
 2 1 1 
The matrix of the quadratic form is A   1 2  1
 1  1 2 
2- -1 1
The characteristic equation of A is -1 2- -1  0
1 -1 2-
Expanding  - 6 + 9 - 4 = 0
3 2

 = 1 is a root
Dividing 3 - 62 + 9 - 4 by  -1,
1 6 9 4
0 1 5 4
1 5 4 0
The remaining roots are given by 2-5 + 4 = 0
2 - 5 + 4 = ( - 1) ( - 4) = 0
i.e.,  = 1, 4
Matrices 19

The eigen values of A are  = 4, 1, 1

Case 1 =4
2 - 4 - 1 1   x 1  0 
The eigen vectors are given by  - 1 2 - 4 - 1 
  x   0 
 2  
 1 - 1 2 - 4  x 3  0
 2  1 1  1 - 1 - 2
  1  2  1  ~ 0 - 3 - 3 
   
 1  1  2 0 0 0 
 x1 - x2 - 2x3 = 0
-3x2 - 3x3 = 0
A solution is x3 = 1, x2 = -1, x1 = 1.

1
 The corresponding eigen vector is X1 =  1
 1 
Case 2 =1
2 - 1 - 1 1   x1  0
The eigen vectors are given by  - 1 2 - 1 - 1   x 2   0

 1 - 1 2 - 1  x3  0
 1 - 1 1  1  1 1
- 1 1 - 1 ~ 0 0 0
   
 1 - 1 1  0 0 0
 x1- x2 + x 3 = 0
Put x3 = 0. We get x1 = x2 = 1. Let x1 = x2 = 1
1
 The eigen vector corresponding to  = 1 is X2 = 1
0
X1 and X2 are orthogonal as X 1T X 2 = 10 + (-1) 1 + 11 = 0.
a 
To find another vector X3 = b corresponding to  =1 such that it is orthogonal to both
 c 
X1 and X2 and satisfies x1- x2 + x3 = 0
i.e., X1.X3 = 0, X2.X3 = 0 and a – b + c = 0
i.e., 1.a -1.b + 1.c = 0, 1.a + 1.b + 0.c = 0 and a – b + c = 0.
i.e., a – b + c = 0 and a+b=0
i.e., a = -b and c = 2b
Put b =1, so that a = -1, c = 2
Matrices 20

  1
 X 3   1 
 2 
 1 1  1
The modal matrix is  1 1 1 
 1 0 2 
 1/ 3 1/ 2  1/ 6
 
Hence the normalized modal matrix is N   1 / 3 1/ 2 1/ 6 
 1/ 0 2/ 
 3 6

 The required orthogonal transformation is X = NY will reduce the given quadratic
form to the canonical form.
C.F= 4 y12  y 22  y 32
Rank of the quadratic form = 3, index = 3, signature = 3. The quadratic form is positive
definite.