Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part.

WCN 02-200-203
 B
Review of Matrix Algebra
B.1 Definition of a Matrix
B.2 Types of Matrices
B.3 Matrix Operations
B.4 Solution of Simultaneous Equations by the Gauss-Jordan Method
Problems

In this appendix, some basic concepts of matrix algebra necessary for formu-
lating the computerized analysis of structures are briefly reviewed. A more
comprehensive and mathematilcally rigorous treatment of these concepts can
be found in any textbook on matrix algebra, such as [12] and [29].

B.1 Definition of a Matrix

A matrix is a rectangular array of quantities arranged in rows and columns.
A matrix containing m rows and n columns can be expressed as:

 A11 A12 ... ... A1n 

 
A21 A22 ... ... A2 n 
A 5 [ A] 5  ith row (B.1)
 ... ... Aij ... 
 
 Am1 Am 2 ... | ... Amn 
jth column m3n
As Eq. (B.1) indicates, matrices are usually denoted either by boldface letters
(e.g., A) or by italic letters enclosed within brackets (e.g., [ A]). The quanti-
ties that form a matrix are referred to as the elements of the matrix, and each
element is represented by a double-subscripted letter, with the first subscript
identifying the row and the second subscript identifying the column in which
the element is located. Thus in Eq. (B.1), A12 represents the element located
in the first row and the second column of the matrix A, and A21 represents the
element in the second row and the first column of A. In general, an element
located in the ith row and the jth column of matrix A is designated as Aij . It
is common practice to enclose the entire array of elements between brackets,
as shown in Eq. (B.1).
The size of a matrix is measured by its order, which refers to the number
of rows and columns of the matrix. Thus the matrix A in Eq. (B.1), which

749

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
750   APPENDIX B Review of Matrix Algebra

consists of m rows and n columns, is considered to be of order m 3 n (m by n).

As an example, consider a matrix B given by

 5 21 3 27 
B 5  40 26 19 23 
 
 28 12 50 22 
The order of this matrix is 3 3 4, and its elements can be symbolically
­represented by Bij, with i 51 to 3 and j 51 to 4; for example, B23 519, B31 528,
B34 5 22, etc.

B.2 Types of Matrices

Row Matrix
If all the elements of a matrix are arranged in a single row (i.e., m 51), then
the matrix is called a row matrix. An example of a row matrix is
C 5 [50 23 227 35]

Column Matrix
A matrix with only one column of elements (i.e., n 51) is called a column
matrix. For example,

210 
 33
D 5 {D} 5  
 26 
 15
Column matrices are also referred to as vectors and are sometimes denoted by
italic letters enclosed within braces (e.g., { D}).

Square Matrix
A matrix with the same number of rows and columns (m 5 n) is called a square
matrix. An example of a 3 3 3 square matrix is

 5 21 3 
 
A 5  40 26 19  (B.2)
 
 28 12 50 
 
Main diagonal
The elements with the same subscripts—that is, A11 , A22 ,…, Ann—form the
main diagonal of the square matrix A. These elements are referred to as the

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 Section B.2 Types of Matrices   751

diagonal elements. As shown in Eq. (B.2), the main diagonal extends from the
upper left corner to the lower right corner of the square matrix. The remaining
elements of the matrix (i.e., Aij with i ? j ) that are not along the main diagonal
are termed the off-diagonal elements.

Symmetric Matrix
If the elements of a square matrix are symmetric about its main diagonal
(i.e., Aij 5 Aji), the matrix is called a symmetric matrix. An example of a 4 3 4
symmetric matrix is

 212 26 13 5
 26 7 228 31 
A5 
 13 228 10 29 
 5 31 29 22 

Diagonal Matrix
If all the off-diagonal elements of a square matrix are zero (i.e., Aij 5 0 for
i ? j ), the matrix is referred to as a diagonal matrix. For example,

3 0 0
A 5  0 28 0 
 
0 0 14 

Unit or Identity Matrix

A diagonal matrix with all its diagonal elements equal to 1 (i.e., I ii 51 and
I ij 5 0 for i ? j ) is called a unit, or identity, matrix. Unit matrices usually are
denoted by I or [ I ]. An example of a 4 3 4 unit matrix is

1 0 0 0
0 1 0 0
I5 
0 0 1 0
0 0 0 1

Null Matrix
When all the elements of a matrix are zero (i.e., Oij 5 0 ), the matrix is
called a null matrix. Null matrices are commonly denoted by O or [O]. For
example,

0 0 0 0
O5 0 0 0 0 
 
0 0 0 0

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
752   APPENDIX B Review of Matrix Algebra

B.3 Matrix Operations

Equality
Two matrices A and B are equal if they are of the same order and if their
corresponding elements are identical (i.e., Aij 5 Bij). Consider, for example,
the matrices

 23 5 6   23 5 6 
A 5  4 7 9  and B 5  4 7 9 
   
 12 0 1   12 0 1 
Since both A and B are of order 3 3 3 and since each element of A is equal
to the corresponding element of B, the matrices are considered to be equal to
each other; that is, A 5 B.

Addition and Subtraction

The addition (or subtraction) of two matrices A and B, which must be of
the same order, is carried out by adding (or subtracting) the corresponding
elements of the two matrices. Thus if A 1 B 5 C, then Cij 5 Aij 1 Bij ; and if
A 2 B 5 D, then Dij 5 Aij 2 Bij . For example, if

2 5  10 4 
A 5  3 0  and B 5  6 7 
   
 8 1  9 2
then

 12 9 
A1B5C 5  9 7 
 
 17 3 
and

 28 1
A 2 B 5 D 5  23 27 
 
 21 21 
Note that matrices C and D have the same order as matrices A and B.

Multiplication by a Scalar
To obtain the product of a scalar and a matrix, each element of the matrix must
be multiplied by the scalar. Thus, if

B 5 
7 3
and c 523
 21 4 
then

cB 5 
221 29 
 3 212 

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 Section B.3 Matrix Operations   753

Multiplication of Matrices
The multiplication of two matrices can be carried out only if the number of
­columns of the first matrix equals the number of rows of the second matrix.
Such matrices are referred to as being conformable for multiplication.
Consider, for example, the matrices

A 5 
5
B 5 
21 2 3 26 
and (B.3)
 7 23   4 28 9 
in which A is of order 2 3 2 and B is of order 2 3 3. Note that the product AB
of these matrices is defined, because the first matrix, A, of the sequence AB
has two columns and the second matrix, B, has two rows. However, if the
sequence of the matrices is reversed, the product BA does not exist, because
now the first matrix, B, has three columns and the second matrix, A, has two
rows. The product AB is usually referred to either as A postmultiplied by B
or as B premultiplied by A. Conversely, the product BA is referred to either
as B postmultiplied by A or as A premultiplied by B.
When two conformable matrices are multiplied, the product matrix thus
obtained will have the number of rows of the first matrix and the number of
columns of the second matrix. Thus, if a matrix A of order m 3 n is postmulti-
plied by a matrix B of order n 3 s, then the product matrix C will be of order
m 3 s; that is,

(B.4)

As illustrated in Eq. (B.4), any element Cij of the product matrix C can be eval-
uated by multiplying each element of the ith row of A by the corresponding
element of the jth column of B and by algebraically summing the resulting
products; that is,
Cij 5 Ai1 B1j 1 Ai 2 B2 j 1 1 Ain Bnj (B.5)
Equation (B.5) can be conveniently expressed as

n
Cij 5 ∑ Aik Bkj (B.6)
k 51

in which n represents the number of columns of the matrix A and the number
of rows of the matrix B. Note that Eq. (B.6) can be used to determine any
element of the product matrix C 5 AB.

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
754   APPENDIX B Review of Matrix Algebra

To illustrate the procedure of matrix multiplication, we compute the prod-

uct C 5 AB of the matrices A and B given in Eq. (B.3) as

C 5 AB 5 
21 5   2 3 26   18 243 51 
5
 7 23   4 28 9   2 45 269 
232 233 233
in which the element C11 of the product matrix C is obtained by multiplying
each element of the first row of A by the corresponding element of the first
column of B and summing the resulting products; that is,
C11 521(2) 1 5(4) 5 18
Similarly, the element C21 is determined by multiplying the elements of the
second row of A by the corresponding elements of the first column of B and
adding the resulting products; that is,
C21 5 7(2) 2 3(4) 5 2

The remaining elements of C are determined in a similar manner:

C12 521(3) 1 5(28) 5243
C22 5 7(3) 2 3(28) 5 45
C13 521(26) 1 5(9) 5 51
C23 5 7(26) 2 3(9) 5269
Note that the order of the product matrix C is 2 3 3, which equals the number
of rows of A and the number of columns of B.
A common application of matrix multiplication is to express simultaneous
equations in compact matrix form. Consider the system of simultaneous linear
equations:
A11 x1 1 A12 x 2 1 A13 x3 5 P1
A21 x1 1 A22 x 2 1 A23 x3 5 P2 (B.7)
A31 x1 1 A32 x 2 1 A33 x3 5 P3

in which x1 , x 2, and x3 are the unknowns and As and Ps represent the c­ oefficients
and constants, respectively. By using the definition of matrix multiplication, this
system of simultaneous equations can be written in matrix form as

 A11 A12 A13   x   P 

   1   1 
 21 22 23   x 2  5  P2 
A A A (B.8)
 A A A   x   P 
 31 32 33   3   3 

or, symbolically, as
Ax 5 P (B.9)
Even when two matrices A and B are of such orders that both products AB
and BA can be determined, the two products are generally not equal; that is,

AB ± BA(B.10)

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 Section B.3 Matrix Operations   755

It is, therefore, necessary to maintain the proper sequential order of matrices when
computing matrix products. Although matrix multiplication is generally not com-
mutative, as indicated by Eq. (B.10), it is associative and distributive, provided that
the sequential order in which the matrices are to be multiplied is maintained. Thus
ABC 5 ( AB)C 5 A(BC) (B.11)
and
A(B 1 C) 5 AB 1 AC (B.12)
Multiplication of any matrix A by a conformable null matrix O yields a
null matrix; that is,
OA 5 O and AO 5 O (B.13)
For example,
 0 0   5 27   0 0 
 0 0   9 2 5 0 0 
     
Multiplication of any matrix A by a conformable unit matrix I yields the
same matrix A, that is,
IA 5 A and AI 5 A (B.14)
For example,
 1 0   5 27   5 27 
 0 1   9 2 5 9 2 
     
and
 5 27   1 0   5 27 
9 2 0 1 5 9 2 
     
As Eqs. (B.13) and (B.14) indicate, the null and unit matrices serve the pur-
poses in matrix algebra that are analogous to those of the numbers 0 and 1,
respectively, in scalar algebra.

Inverse of a Square Matrix

The inverse of a square matrix A is defined as a matrix A21 with elements of
such magnitudes that the multiplication of the original matrix A by its inverse
A21 yields a unit matrix I; that is,

A −1 A 5 AA −1 5 I(B.15)

Consider, for example, the square matrix

 1 22 
A5 
 3 24 
The inverse of A is given by
 22 1 
A21 5  
 21.5 0.5 

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
756   APPENDIX B Review of Matrix Algebra

so that the products A21 A and AA21 satisfy Eq. (B.15):

 22 1   1 22 
A −1A 5  
 21.5 0.5   3 24 
 (22 1 3) (4 2 4)   1 0 
5  5 5I
 (21.5 1 1.5) (3 2 2)   0 1 

and

 1 22   22 1   (22 1 3) (1 2 1)   1 0 
AA21 5     5  (26 1 6) (3 2 2) 5 0 1 5I
 3 24   21.5 0.5     

The operation of inversion is defined only for square matrices. The inverse
of such a matrix is also a square matrix of the same order as the original
matrix. A procedure for determining inverses of matrices is presented in the
following section. The operation of matrix inversion serves the same purpose
as the operation of division in scalar algebra. Consider a system of simultane-
ous equations expressed in the matrix form as
Ax 5 P
in which A represents the square matrix of known coefficients, x represents the
vector of the unknowns, and P represents the vector of the constants. Since the
operation of division is not defined in matrix algebra, we cannot solve the fore-
going matrix equation for x by dividing P by A (i.e., x 5 PyA ). Instead, to deter-
mine the unknowns x, we premultiply both sides of the equation by A21 to obtain

A21 Ax 5 A21P
Since A21 A 5 I and Ix 5 x, we can write

x 5 A21P
which indicates that a system of simultaneous equations can be solved by pre-
multiplying the vector of the constants by the inverse of the coefficient matrix.
An important property of matrix inversion is that the inverse of a symmet-
ric matrix is always a symmetric matrix.

Transpose of a Matrix
The transpose of a matrix is obtained by interchanging its corresponding rows
and columns. The transposed matrix is usually identified by the superscript T
placed on the symbol of the original matrix. Consider, for example, the 2 3 3
matrix

 6 22 4 
A5 
 1 8 23 
The transpose of A is given by

 6 1
AT 5  22 8 
 4 23 

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 Section B.3 Matrix Operations   757

Note that the first column of A becomes the first row of AT . Similarly, the
second and third columns of A become, respectively, the second and third rows
of AT . The order of AT thus obtained is 3 3 2.
As another example, consider the 3 3 3 matrix
 9 7 25 
B 5  7 23 2 
 25 2 6 
Since the elements of B are symmetric about the main diagonal (i.e., Bij 5 B ji ),
interchanging the rows and the columns of this matrix produces a matrix BT
that is identical to the matrix B itself; that is,
BT 5 B
Thus, the transpose of a symmetric matrix yields the same matrix.
Another useful property of matrix transposition is that the transpose of a
product of matrices equals the product of the transposes in reverse order; that is,

( AB)T 5 BT AT (B.16)

Similarly,
( ABC)T 5 C T BT AT (B.17)

Partitioning of Matrices
Partitioning is a process by which a matrix is subdivided into a number of
smaller matrices called submatrices. For example, a 3 3 4 matrix A is partitioned
into four submatrices by drawing horizontal and vertical dashed partition lines:
 3 5 21 2
   A11 A12 
A 5  22 4 7 9 5  (B.18)
   A 21 A 22 
 6 1 3 4 

in which the submatrices are
 3 5 21  2
A11 5   A12 5  
 22 4 7  9
A 21 5 [ 6 1 3 ] A 22 5 [4]
Matrix operations such as addition, subtraction, and multiplication can be
performed on partitioned matrices in the same manner as described previously by
treating the submatrices as elements, provided that the matrices are partitioned in
such a way that their corresponding submatrices are conformable for the particu-
lar operation. For example, suppose that we wish to postmultiply the 3 3 4 matrix
A of Eq. (B.18) by a 4 3 2 matrix B, which is partitioned into two submatrices as
 1 8
 25 2   B 
B5  5  11  (B.19)
 23 6   B21 
 7 21 
 

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
758   APPENDIX B Review of Matrix Algebra

The product AB is expressed in terms of the submatrices as

 A11 A12   B11   A11B11 1 A12 B21 

AB 5    5  (B.20)
 A 21 A 22   B21   A 21B11 1 A 22 B21 

Note that the matrices A and B have been partitioned in such a way that their cor-
responding submatrices are conformable for multiplication; that is, the orders of
the submatrices are such that the products A11B11 , A12 B21 , A 21B11, and A 22 B21 are
defined. As shown in Eqs. (B.18) and (B.19), this is achieved by partitioning the
rows of the second matrix B of the product AB in the same way that the columns
of the first matrix A are partitioned. The products of the submatrices are given by
 1 8
 3 5 21     219 28 
A11B11 5    25 2  5  
 22 4 7 
 23 6  
248 34 

 

2  14 22 
A12 B21 5   [ 7 21 ] 5  
9   63 29 
 1 8
 
A 21B11 5 [ 6 1 3 ]  25 2  5 [ 28 68 ]
 
 23 6 
A 22 B21 5 [ 4 ][ 7 21 ] 5 [ 28 24 ]

Substitution into Eq. (B.20) yields

  219 28   14 22    25 26 
  1   
AB 5   243 34   63 29   5  20 25 
   
 [28 68] 1 [28 24]   20 64 
 

B.4 Solution of Simultaneous Equations by the Gauss-Jordan Method

The Gauss-Jordan elimination method is one of the most commonly used
procedures for solving simultaneous linear algebraic equations. To illustrate
the method, consider the following system of three simultaneous equations:
2 x1 2 5 x 2 1 4 x3 5 44
3 x1 1 x 2 2 8 x3 5 235 (B.21a)
4 x1 27 x 2 2 x3 5 28
To solve for the unknowns x1 , x 2, and x3, we begin by dividing the first e­ quation
by the coefficient of its x1 term:
x1 2 2.5 x 2 1 2 x3 5 22
3 x1 1 x 2 2 8 x3 5 235 (B.21b)
4 x1 2 7 x 2 2 x3 5 28

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 Section B.4 Solution of Simultaneous Equations by the Gauss-Jordan Method   759

Next, the unknown x1 is eliminated from the remaining equations by succes-

sively subtracting from each remaining equation the product of the coefficient
of its x1 term and the first equation. Thus, to eliminate x1 from the second
equation, we multiply the first equation by 3 and subtract it from the
second equation. Similarly, we eliminate x1 from the third equation by multi-
plying the first equation by 4 and subtracting it from the third equation. The
system of equations thus obtained is

x1 2 2.5 x 2 1 2 x3 5 22
8.5 x 2 2 14 x3 5 2101 (B.21c)
3 x 2 2 9 x3 5 260

With x1 eliminated from all but the first equation, we now divide the second
equation by the coefficient of its x 2 term:

x1 2 2.5 x 2 1 2 x3 5 22
x 2 2 1.647 x3 5 211.882 (B.21d)
3x2 2 9 x3 5 260

Next, we eliminate x 2 from the first and the third equations, successively,
by multiplying the second equation by 22.5 and subtracting it from the first
equation, and then by multiplying the second equation by 3 and subtracting it
from the third equation. This yields

x1 2 2.118 x3 527.705
x 2 2 1.647 x3 5211.882 (B.21e)
2 4.059 x3 5224.354

By dividing the third equation by the coefficient of its x3 term, we obtain

x1 2 2.118 x3 5 27.705
x 2 2 1.647 x3 5 211.882 (B.21f)
x3 5 6

Finally, by multiplying the third equation by 22.118 and subtracting it from

the first equation, and by multiplying the third equation by 21.647 and sub-
tracting it from the second equation, we determine the solution of the given
system of equations (Eq. (B.21a)) to be

x1 5 5
x2 522 (B.21g)
x3 5 6

That is, x1 5 5, x 2 522, and x3 5 6 . To check that the solution is carried out
correctly, we substitute the numerical values of x1, x 2, and x3 back into the
original equations (Eq. (B.21a)):
2(5) 2 5(22) 1 4(6) 5 44 Checks
3(5) 2 2 2 8(6) 5235 Checks
 
4(5) 2 7(22) 2 6 5 28 Checks

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
760   APPENDIX B Review of Matrix Algebra

As the foregoing example illustrates, the Gauss-Jordan method essentially

involves successively eliminating each unknown from all but one of the equa-
tions of the system by performing the following operations: (1) dividing an
equation by a scalar and (2) multiplying an equation by a scalar and subtracting
the resulting equation from another equation. These operations, which do not
change the solution of the original system of equations, are applied repeatedly
until a system with each equation containing only one unknown is obtained.
The solution of simultaneous equations is usually carried out in matrix
form by operating on the rows of the coefficient matrix and the vector con-
taining the constant terms of the equations. The foregoing operations are then
referred to as elementary row operations. These operations are applied to both
the coefficient matrix and the vector of the constants simultaneously, until
the coefficient matrix is reduced to a unit matrix. The elements of the vector,
which initially contained the constant terms of the original equations, now
represent the solution of the original simultaneous equations. To illustrate this
procedure, consider again the system of three simultaneous equations given in
Eq. (B.21a). The system can be expressed in matrix form as
Ax 5 P
 2 25 4  x1   44 
     (B.22)

3 1 28   x 2  5  235 
 4 27 21   x3   28 
 

When applying the Gauss-Jordan method, it is usually convenient to write the

coefficient matrix A and the vector of constants P as submatrices of a parti-
tioned augmented matrix:

 2 25 4 44 
 
3 1 28 − 35  (B.23a)
 
 4 27 21 28 

To determine the solution, we begin by dividing row 1 of the augmented matrix
by A11 5 2:

 1 22.5 2 22 
 
 3 1 28 − 35  (B.23b)
 
 4 27 21 28 


Next, we multiply row 1 by A21 5 3 and subtract it from row 2 and then multiply
row 1 by A31 5 4 and subtract it from row 3. This yields

 1 22.5 2 22 
 
 0 8.5 214 2101  (B.23c)
 
 0 3 29 260 


Divide row 2 by A22 58.5, obtaining

 1 22.5 2 22 
 
 0 1 21.647 211.882  (B.23d)
 
 0 3 29 260 

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 Section B.4 Solution of Simultaneous Equations by the Gauss-Jordan Method   761

Multiply row 2 by A12 522.5 and subtract it from row 1; then multiply row 2
by A32 5 3 and subtract it from row 3. This yields

 1 0 22.118 27.705 
 
 0 1 21.647 − 11.882  (B.23e)
 
 0 0 24.059 224.354 

Divide row 3 by A33 524.059 :

 1 0 22.118 27.705 
 
 0 1 21.647 − 11.882  (B.23f)
 
 0 0 1 6 
Multiply row 3 by A13 522.118 and subtract it from row 1; then multiply row
3 by A23 521.647 and subtract it from row 2. This yields
 1 0 0 5 
 
 0 1 0 −2  (B.23g)
 
 0 0 1 6 

Thus x1 5 5, x 2 522, and x3 5 6 .

Matrix Inversion
The Gauss-Jordan elimination method can also be used to determine the
inverses of square matrices. The procedure is similar to that described previ-
ously for solving simultaneous equations, except that in the augmented matrix,
the coefficient matrix is now replaced by the matrix A that is to be inverted and
the vector of constants P is replaced by a unit matrix I of the same order as the
matrix A. Elementary row operations are then performed on the augmented
matrix to reduce the matrix A to a unit matrix. The matrix I, which was initially
the unit matrix, now represents the inverse of the original matrix A.
To illustrate the foregoing procedure, let us compute the inverse of the
2 3 2 matrix
 1 22 
A5  (B.24)
 3 24 

The augmented matrix is given by

 1 22 1 0 
  (B.25a)
 3 24 0 1 

By multiplying row 1 by A21 5 3 and subtracting it from row 2, we obtain
 1 22 1 0 
  (B.25b)
 0 2 23 1 

Next, by dividing row 2 by A22 5 2, we obtain
 1 22 0 1
  (B.25c)
 0 1 21.5 0.5 


Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
762   APPENDIX B Review of Matrix Algebra

Finally, by multiplying row 2 by 22 and subtracting it from row 1, we obtain

1 0 22 1 
  (B.25d)
 0 1 21.5 0.5 

Thus,

 22 1 
A21 5  
 21.5 0.5 

The computations can be checked by using the relationship A21 A 5 I . We

showed in Section B.3 that the matrix A21 , as computed here, does indeed
satisfy this relationship.

Problems

Section B.3
B.7 Solve the following system of simultaneous equations by
B.1 Determine the matrix C 5 A 1 3B if the Gauss-Jordan method.
 12 28 15   2 21 1  −12 x1 2 3 x 2 1 6 x3 5 45
A 5  28 7 10  B 5  21 4 6  5 x1 1 2 x 2 2 4 x3 5 29
   
 15 10 25   1 6 3  10 x1 1 x 2 2 7 x3 5232
B.2 Determine the matrix C 5 2B 2 A if B.8 Solve the following system of simultaneous equations by
the Gauss-Jordan method.
 3 7   21 6  5 x1 2 2 x 2 1 6 x3 5 0
A5 8 4  B5 5 1 
    22 x1 1 4 x 2 1 x3 1 3 x 4 5 18
 2 22   3 24 
6 x1 1 x 2 1 6 x3 1 8 x 4 5 229
B.3 Determine the products C 5 AB and D 5 BA if 3 x 2 1 8 x3 1 7 x 4 5 11
 6  B.9 Determine the inverse of the matrix shown using the
A 5  24  B 5 [ 23 1 25 ] Gauss-Jordan method.
 
 2   4 23 21 
A 5  22 5 1 
B.4 Determine the products C 5 AB and D 5 BA if  
 6 24 25 
 23 2   6 24 
A5  B5  B.10 Determine the inverse of the matrix shown using the
 2 5   24 1  Gauss-Jordan method.
B.5 Show that (AB)T 5 BT AT by using the matrices A and B  4 2 0 23 
given here.  2 3 24 0 
A5 
 8 22 5   1 25   0 24 2 21 
A 5  1 24 3  B5 7 0   23 0 21 5 
   
 2 0 6   0 23 
Section B.4
B.6 Solve the following system of simultaneous equations by
the Gauss-Jordan method.
2 x1 1 5 x 2 2 x3 5 15
5 x1 2 x 2 1 3 x3 5 27
2x1 1 3 x 2 1 4 x3 5 14

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 Answers to Selected Problems

Chapter 2 3.9   Ax 5 0 ; A y 5 220 kN ↑; M A 5 650 kN ⋅ m

2.1 Beam BE: w 5 3.84 kNym; Girder AC : PA 5 PC 5 5.76 kN; 3.11 A y 5 146 kN ↑; Bx 5 0; By 5 272.2 kN ↑
PB 5 11.52 kN
3.13 Ax 5 37.5 kN →; A y 5100 kN ↑; RB 5 62.5 kN 
2.3 Beam AF : w 5 8.25 kNym ; Beams BG and CH: w 5
16.5 kNym ; Girder AC : PA 5 PC 5 49.5 kN; PB 5 99 kN ; 3.15 For 0 # x # 20 m: A y 5 45 2 2 x kN ↑; By 5 5 1 2 x kN ↑
Girder FH: PF 5 PH 5 99 kN; PG 5 198 kN For 20 m # x # 25 m: Ay 5 (25 2 x )2 y5 kN ↑ ;
By 5 (625 2 x 2 )y5 kN ↑
2.5 Beam CD: w 5 9.4 kNy m ; Girder AE: w 5 1.62 kNym ;
PC 5 35.25 kN ; PA 5 PE 5 19.34 kN 3.17 Ax 5 0 ; A y 585 kN ↑; By 5 280 kN ↑

2.7 Beam BF : w 5 16.04 kNym ; Girder AD: w 5 1.97 kNym ; 3.19 Ax 5 200 kN ←; A y 5125 kN ↓; By 5 475 kN ↑
PB 5 PC 580.2 kN; PA 5 PD 5 41.85 kN 3.21 Ax 5100 kN ←; A y 5 216.11 kN ↑; By 5183.89 kN ↑
2.9 Beam CD : w 5 6.91 kNy m; Girder AE: PC 5 25.1 kN ; 3.23 A y 5 244.07 kN ↑; Bx 5 240 kN ←; By 585.93 kN ↑
PA 5 PE 5 12.96 kN
3.25 Ax 5 0 ; A y 5 225 kN ↑; By 5 1050 kN ↑; C y 5 225 kN ↓
2.11 Beam EF: w 5 3 kNym ; Girder AG: PC 5 PE 5 9 kN;
PA 5 PG 5 4.5 kN; Column A: P 5 13.5 kN 3.27 Ax 5 168 kN →; A y 5 187.5 kN ↑; Bx 5 102 kN → ;
By 5 15 kN ↑
2.13 Windward side: 292.2 Nym 2 and 230.4 Nym 2 ; Leeward
side: 2552.9 Nym 2 3.29 Ax 5 0; A y 5 Dy 5 25 kN ↑ ; By 5 C y 5 650 kN ↑

2.15 Windward wall: 878.2 Nym 2 for 0 # z # 4.6 m ; 3.31 Ax 5 170 kN →; A y 5 310 kN ↑; Bx 5 170 kN ←;
1033.2 Nym 2 for z 5 10 m ; Leeward wall: 2598.6 Nym 2 By 5 370 kN ↑

2.17 0.7 kNym 2 3.33 Ax 5 55 kN ←; A y 5 216.11 kN ↑; Bx 5 45 kN ←;

By 5183.89 kN ↑
3.35 Ax 5 35.76 kN ←; A y 5 79.45 kN ↑ ; Bx 5 53.24 kN ←;
Chapter 3 By 5 168.46 kN ↑
3.1 (a) Unstable; (b) Determinate; (c) Indeterminate, ie 5 2;
3.37 Ax 5 0 ; A y 5 560.62 kN ↑; M A 5 1,462.4 kN ⋅ m ;
(d) Indeterminate, ie 5 1
By 5 585 kN ↑ ; C y 5 24.38 kN ↑
3.3 (a) Unstable; (b) Determinate; (c) Indeterminate, ie 5 3;
3.39 Ay 5 50 kN ↓ ; By 5 475 kN ↑; C x 5 0 ; C y 5 225 kN ↑;
(d) Indeterminate, ie 5 1
M C 5 900 kN ⋅ m
3.5 Ax 5 0 ; A y 5 209 kN ↑; By 5 161 kN ↑
3.41 Ax 5 62.5 kN ←; A y 5 62.5 kN ↓; M A 5187.5 kN ⋅ m ;
3.7 Ax 5100 kN →; A y 5 0 ; M A 5 333.3 kN ⋅ m Bx 5 62.5 kN ←; By 5 62.5 kN ↑; M B 5187.5 kN ⋅ m

791

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
798  Answers to Selected Problems

15.21 and 16.21 M AC 5 58.6 kN ⋅ m ; M C A 5 286 kN ⋅ m ;  116.3 kN   78.5 kN   183.7 kN 

M C D 5 286 kN ⋅ m ; M DC 5 0      
 21.5 kN   116.3 kN   78.5 kN 
15.23 and 16.23 M AC 5 0; M DE 5 100 kN ⋅ m ; 108 kN ⋅ m   221 kN ⋅ m   222 kN ⋅ m 
M C A 5 69.6 kN ⋅ m   ; M BC 5 301.4 kN ⋅ m ; 17.9 Q1 5   Q2 5   Q3 5  
2116.3 kN  278.5 kN  2183.7 kN
M CB 5 37.2 kN ⋅ m ; M C D 5 32.5 kN ⋅ m ;  221.5 kN   183.7 kN   278.5 kN 
M DC 5 100 kN ⋅ m         
 21 kN ⋅ m  2249 kN ⋅ m  249 kN ⋅ m 
15.25 and 16.25 M AC 5 107.8 kN ⋅ m ; M C A 5 20.8 kN ⋅ m ;
M BD 5 222 kN ⋅ m ; M DB 5 249.2 kN ⋅ m ;
M CD 5 20.8 kN ⋅ m   ; M DC 5 249.2 kN ⋅ m   Appendix B
 18 211 18 
15.27 and 16.27 M AB 5 127 kN ⋅ m ; M BA 5 103.4 kN ⋅ m ; B.1 C 5  211 19 28 
M BC 5 103.4 kN ⋅ m    ; M C B 5 0  
 18 28 4 
15.29 and 16.29 M AC 511.7 kN ⋅ m ; M C A 5 43.9 kN ⋅ m   ;  218 6 230 
M CD 5 43.9 kN ⋅ m ; M DC 514.7 kN ⋅ m    ; M DB 514.7 kN ⋅ m ; B.3 C 5  12 24 20 ; D 5232
M BD 5 0  
 26 2 210 
15.31 and 16.31 M AC 5 M BD 5 119 kN ⋅ m ;
B.5 ( AB)T 5 BT A T 5 
26 227 2 
M C A 5 M DB 5 83.5 kN ⋅ m ; M C E 5 M DF 5 23.3 kN ⋅ m ;
 255 214 228 
M EC 5 M FD 5 44.2 kN ⋅ m ; M CD 5 M DC 5 106.8 kN ⋅ m ;
M EF 5 M FE 5 44.2 kN ⋅ m B.7 x1 527; x 2 5 3; x3 525
 0.42 0.22 20.04 
Chapter 17 B.9 A21 5  0.08 0.28 0.04 
 
17.1 Q1 5 268.33 kN (T); Q2 5 240 kN (C)  0.44 0.04 20.28 
17.3 Q1 5 102.8 kN (T); Q2 5 28.6 kN (C); Q3 5 145.4 kN (C)
Appendix C
 104.4 kN   245.6 kN  C.1 (a) 3,975 mm 2; (b) 2,836 mm 4 ; (c) 3,800 mm 2
   
17.5 Q1 5 
394 kN ⋅ m  Q 5 2232 kN ⋅ m C.3 528(10 6 ) mm 4
2104.4 kN 2  45.6 kN 
   
232 kN ⋅ m  2178 kN ⋅ m Appendix D
 47.47 kN   75.36 kN  D.1 A y 5 E y 5 62.5 kN ↑; C y 5 275 kN ↑
   
 74.64 kN   47.47 kN  D.3 A y 5108.75 kN ↑ ; By 5 357.5 kN ↑; Dy 583.75 kN ↑
 63.8 kN ⋅ m   65.59 kN ⋅ m 
17.7 Q1 5   Q2 5   D.5 A y 5 0 ; M A 5 120 kN ⋅ m ; By 5 220 kN ↑
 247.47 kN   275.36 kN  D.7 A y 5 125.6 kN ↑; By 5 354.9 kN ↑ ; C y 5 298.4 kN ↑;
 75.36 kN   52.53 kN 
    E y 5 71.1 kN ↑
265.59 kN ⋅ m 280.74 kN ⋅ m
D.9 A y 5 57 kN ↑ ; By 5 477 kN ↑; C y 5 216 kN ↑

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-203
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.