II

_
gfg 'ipffiMffHir

Instructor s Solutions

Introduction to
Manual
'

^^B

anagemenH
[^jrf)
Sixth Edition

Bernard W. Taylor ill

 Digitized by the Internet Archive
in 2010

http://www.archive.org/details/instructorssolutOOtayl
Instructor's Solutions Manual

Introduction to

Management

Science
 Instructor's Solutions Manual

MaQagemeDi
Science

Bernard W. Taylor III

Prentice Hall, Upper Saddle River, New Jersey 07458

 Acquisitions Editor: Tom Tucker
Associate Editor: Kristen Imperatore
Production Editor: Joseph F. Tomasso
Formatter: GGS Information Srevices
Manufacturer: Technical Communication Services

© 1999 by Prentice Hall, Inc.

A Simon & Schuster Company
Upper Saddle River, NJ 07458

All rights reserved. No part of this book may

be reproduced, in any form or by any means,
without permission in writing from the publisher.

Printed in the United States of America

10 9876543
ISBN 0-13-975087-8

Prentice-Hall International (UK) Limited, London

Prentice-Hall of Australia Pty. Limited, Sydney
Prentice-Hall Canada Inc., Toronto
Prentice-Hall Hispanoamericana, S.A., Mexico
Prentice-Hall of India Private Limited, New Delhi
Prentice-Hall of Japan, Inc., Tokyo
Simon & Schuster Asia Pte. Ltd., Singapore
Editora Prentice-Hall do Brasil, Ltda., Rio de Janiero
 Contents

Chapter 1 page 1

Chapter 2 page 5

Chapter 3 page 16

Chapter 4 page 46

Chapter 5 page 64

Chapter 6 page 123

Chapter? page 166

Chapters page 188

Chapter 9 page 205

Chapter 10 page 209

Chapter 11 page 238

Chapter 12 page 259

Chapter 13 page 266

Chapter 14 page 289

Chapter 15 page 297

Chapter 16 page 308

Chapter 17 page 329

Chapter 18 page 344

Chapter 19 page 356

 Chapter One

ANSWERS TO QUESTIONS 12. maximize Z = S 1 OOjc - 40.c

subject to
1. Management science is the application of
= 5(X)
25.r
mathematical techniques to management
x = 20 tables
Solution:
problems in order to help managers make better
Z = $ 100(20) - 40(20) = $ ,200 1

decisions.
13. Because the results provide feedback to the
2. It is a philosophy of approaching a problem in
originalmodel which can cause the model to be
a logical, systematic, and consistent way.
updated and changed
3. a) Observation
14. The actual use of management science model
b) Problem definition
results. Fixed cost, variable cost, volume, price,
c) Model construction
and demand
d) Model solution
15. By setting the profit function equal to zero and
e) Implementation of results
solving for volume.
4. An individual skilled in the techniques of
management science and trained to identify
management problems.
PROBLEM SUMMARY
1. Total cost, revenue, profit, and break-even
5. The existence of a problem implies that the

objectives of a business firm are not being met 2. Total cost, revenue, profit, and break-even
in some way.
3. Total cost, revenue, profit, and break-even
6. An abstract representation of an existing
4. Break-even volume
problem situation

5. Graphical analysis (1-2)

7. Variables are represented in models by
symbols. These symbols can take on various 6. Graphical analysis (I^)
numerical values. In a mathematical equation
7. Break-even sales \olume
the value which a dependent variable equals
depends on what values are determined for 8. Break-even volume as a percentage of capacity
independent variables. A decision variable in a (1-2)
model is what is being solved for. It can be
9. Break-even volume as a percentage of capacity
either a dependent or an independent variable.
(1-3)
8. A mathematical equation which relates
10. Break-even volume as a percentage of capacity
variables and parameters
(1^)
9. Constant numerical values which are generally
11. Effect of price change (1-2)
coefficients of variables in an equation

12. Effect of price change 1—4)

10. An objective function is an equation which (

replicates the objective of a firm, such as profit. 13. Effect of variable cost change (1-12)
However, unlimited profit is not possible
14. Effect of fixed cost change (1-13)
because a firm faces limited available resources.
These limited resources are expressed in a model 15. Break-even analysis
by mathematical relationships called constraints.
16. Effect of fixed cost change (1-7)
The constraints and objective function consist
(in part) of decision variables, whose values 17. Effect of variable cost change (1-7)
achieve the objective subject to the constraints.
18. Break-even analysis
11. When model consists of a set of mathematical
a
19. Break-e\en analysis
relationships and the model is sol\ed to \ ield a
direct answer, the answer represents a decision.
However, some results simply describe the
system or situation being modeled (such as an
average), and the\ do not represent a potential
decision.
PROBLEM SOLUTIONS 8. Break-even volume as percentage of capacity

la. V = 300, c, = $8,000, c„ = $65 per table, p = $180; 3,750

= =
.469 46.9%
TC = c, + wcv = $8,000 + (300) (65) = $27,500; 8,000
TR = vp = (300) (180) = $54,000; Z = $54,000 -
27,500 = $26,500 per month 9. Break-even volume as percentage of capacity = —=
AC

8,000
b. v = —= 69.56 tables per month 24,705.88
=
- 65 98.8%
p - c. 180
25,000
= 12,000, c, = $60,000, c. = $9, p = $25; TC =
2a. V
10. Break-even volume as percentage of capacity = — =
c, + vcv = 60,000 + (12,000) (9) = $168,000; TR = A:

vp = (12,000) ($25) = $300,000; Z = $300,000 -

100,000
168,000 = $132,000 per year = .833 = 83.3%
120,000
C-,
= 60,000 ^^,„
b. V = 3,750
.

tires per year

60,000
- c. 25 - 9 Cf
11. 2,727.3 tires per year; it re-
p - c. 31-9
3a. V = 18,000, c, = $21,000, c, = $.45, p = $1.30;
duces the break-even volume from 3,750 tires to 2,727.3
TC = cr + vcv = $21,000 + (18,000) (.45) = $29,100;
TR = vp = (18,000) (1.30) = $23,400; Z = $23,400 -
tires per year.

29,100 = -$5,700 (loss) Cf 25,000

=
12. 55,555.55 lb per month; it

c, 21,000 p- c, .60 - .15

b. V = 24,705.88 yd per month
Cv 1.30 - .45 reduces the break-even volume from 100,000 lb per
month to 55,555.55 lb.

c, = $25,000, p = $.40, Cv = $.15, v = Cl 25,000

4,
p - c. 13. = 65,789.47 lb per month; it
p- c, .60 - .22
25,000
= 100,000 lb per month increases the break-even volume from 55,555.55 lb per
.40 - .15 month to 65,789.47 lb per month.

c, 39,000
14. = 102,631.57 lb per month; it

2501- .60 - .22

increases the break-even volume from 65,789.47 lb per

month to 102,631.57 lb per month.

15. Initial profit: Z -= vp - Cf - vc, = (9,000) (.75)

-
4,000 - (9,000) (.21) = 6,750 - 4,000 - 1,890 =
$860 per month; increase in price: Z = vp — C(
—
vcv = (5,700) (.95) - 4,000 - (5,700) (.21) = 5,415 -
4,000 - 1,197 = $218 per month; the dairy should not

raise its price.

1 ,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000

Volume, (v)

35.000
6. 16. = 1.750
/' - 'v
30- 10
150 r

125 The increase in fixed cost from $25,000 to

$35,000 will increase the break-even point
100
Total from 1.250 to 1.750 or 500 dolls, thus, he
Revenue should not spend the extra $10,000 for
£ 75 Break-even point
advertising.
50
17. Original break-even point (from problem)

25 7 = 1.250
New break-even point:

25 50 75 100 125 150 175 200 17.000

Volume, ($1,000$) = 1062.5
P - ', 30- 14

$25,000
= 1.250 dolls
/' - 'v 30 - 1
 2

18. a) Cf $27,000 e. With both options:

' " /TTT; =8.95-3.75 " 5,192.30 pizzas
Z= vp - Cj - VCy

b) 5,192.3
= 4,700(.99) - 2, 50 - 4,7()()(.25)
^ 259.6 days 1

20 = $1,328
c) Revenue for the first 30 days = 30(pv - vcy) She should purchase the new equipment but not
= 30[(8.95)(20) decrease prices.
(20)(3.75)]
= $3,120 CASE SOLUTION: OCOBEE RIVER
$27,000 - 20 = $23,880, portion of fixed
3,
RAFTING COMPANY
1

cost not recouped after 30 days.

$23,880
Alternative 1: 9= $3,000
New V =
p- c\.
= ^TTTi^ —
T^7?= 5,685.7 ^
7.95 - 3.75
pizzas
P = $20

cv = $12
Total break-even volume = 600 + 5,685.7 =
Cf 3.000
6,285.7 pizzas = 375 rafts
P- Cy 20 - 1

Total time to break-even = 30 + 5.685.7

20 Alternative = $10,000
2: ^7
= 314.3 days
^ = $20
19. a) Cost of Regular plan = $55 + (.33)(50 minutes)
= $71.50 Cy = $8
Cost of Executive plan = $75 + ct 10.000
(.25)(20 = 833.37
minutes) 20-8
= $80
Select regular plan. If demand is less than 375 rafts, the students

- should not start the business.

b) 55 -I- {X = 75 + (x - 100)(,25)
70)(.33)
31.9 + .33x = 50 + .25.v If demand is less than 833 rafts alternative 2
X = 226.25 minutes per month should not be selected, and alternative I should
be used if demand is expected to be between
375 and 833.33 rafts.
CASE SOLUTION: CLEAN CLOTHES
If demand is greater than 833.33 rafts which
CORNER LAUNDRY alternative is best? To determine the answer,
C-, 1,700 equate the two cost functions.
= 2,000 items ^per month
P-Cy 1.1 0-.25
3,000+ 12i = 10.000 + 81-

b. Solution depends on number of months: 36 used 4v = 7.000

here. $16,200 -;-

36 = $450 per month, thus monthly V = 1 .750

fixed cost is $2,150 This is referred to as the point of indifference

2,150 between the two alternatives. In general, for

= 2,529.4 items per month demand lower
P-Cy than this point (1,750) the
i.lO-.25
alternative should be selected with the lowest
529.4 additional items per month fixed cost: for demand greater than this point
the alternative with the lowest variable co.st
C. Z= VP - Cf - VCy
should be selected. (This general relationship
= 4.300( 1 . 10) - 2, 1 50 -4.300(.25)
can be observed by graphing the two cost
= $ ,505 per
1 month equations and seeing where they intersect.)
After 3 years. Z= $1,955 per month
Thus, for the Ocobee River Rafting Company,

't
1,700 the following guidelines should be followed:
d. v = = 2,297.3
P-c\ .99 -.25
demand < 375, do not start business 375 <
Z = Vp - Cf - VCy
demand < 1.750, select alternative I, demand >
= 3.800(.99) - 1.700 -3.800(.25) 1.750. select alternative 2

= $1,112 per month

Since Penny estimates demand will be
approximately 1,000 rafts, alternative I should
be selected.

Z. -7 - vcv

= ( ,000)(20) - 3,000 -
1 ( 1 ,000)( 2)
1

Z = $5,000
 Chapter Two

ANSWERS TO QUESTIONS c) Substitute the values of the decision variables

determined at each corner point to find the
1. A linear programming model includes only solution that results in the maximum value of Z.
linear relationships in the objective function
and constraints. 10. The objective function line will fall directly on
the constraint line, which indicates that all the
2. A linear programming problem has an points along the constraint line are optimal.
objective, it requires a choice between several
alternative courses of action; these choices are 11. It is limited to linear programming models with
represented by decision variables, and only two variables, it is cumbersome, and the

restrictions in the form of constraints exist

graph itself will not always be accurately
which make it impossible to have unlimited drawn.

achievement of the objective. The restrictions 12. It provides a picture of how the solution of a
as well as the objective must be definable by linear programming model is derived and thus
linear mathematical relationships.
helps one understand the mathematical solution
better.
3. First, define the decision variables: then,
develop the objective function; and finally, 13. For a constraint, let one variable equal zero and
develop the constraints. solve for the other variable, thus providing a

4. A feasible solution is one that does not violate point on one axis. Then let the other variable

any of the constraints, whereas an infeasible equal zero and determine a point on the other

solution violates at least one constraint.

axis; connect this point with the other with a
line. For the objective function, arbitraril\
5. An inequality constraint is used whenever there select a level of profit and plot a line in the
is no specific restriction requiring that the same way described for a constraint, or
constraint quantity equal the constraint acti\ ity. compute the slope of the objective function and
This provides the decision maker with plot a line arbitrarily that has the same slope.
additional flexibility.
14. The area that satisfies all of the constraints
6. A solution that achieves the objective function simultaneously
to the greatest extent possible without violating
15. It is the last corner point the objective function
a constraint
touches as it leaves the feasible solution space.
7. Yes; in a linear function the "rate of change" or
16. In a maximization problem the objective
slope of the function is constant, and thus a
change in the value of a decision variable will
function emanates from the origin and moves
change outward until it reaches the last corner point in
always result in a constant, proportional
the solution space, whereas in a minimization
in the functional values.
problem the objective function emanates in the
8. Discrete variables take on countable, integer solution space and moves toward the origin.
values. Airplanes and people are examples of
discrete items. Continuous variables can take
on any value and are divisible. For example, all
PROBLEM SUMMARY
weights and measures are continuous. 1. Maximization
9. a) Plot the model constraints as equations and 2. Maximization
then determine the feasible solution space.
b) Plot the objective function and move this 3. Minimization
line out from the origin until the optimal
4. Sensitivity analysis (2-3)
solution point is reached.
c) Solve simultaneous equations at the solution
5. Minimization

point to find the optimal solution values. 6. Maximization

Or:
7. Slack analysis (2-6)
b) Solve simultaneous equations at each corner
point. 8. Sensitivity analysis (2-6)
 9. Maximization, graphical solution PROBLEM SOLUTIONS
10. Slack analysis (2-9) 1.

11. Maximization, graphical solution

12. Minimization, graphical solution

13. Maximization, graphical solution

14. Sensitivity analysis (2-13)

15. Sensitivity analysis (2-13)

16. Maximization, graphical solution

17. Sensitivity analysis (2-16)

18. Maximization, graphical solution

19. Maximization, graphical solution

20. Maximization, graphical solution

21. Constraint analysis (2-20)

22. Minimization, graphical solution

23. Sensitivity analysis (2-22)

24. Sensitivity analysis (2-22)

25. Sensitivity analysis (2-22)

26. Minimization, graphical solution

27. Minimization, graphical solution

28. Sensitivity analysis (2-27)

29. Maximization, graphical solution

30. Maximization, graphical solution

31. Minimization, graphical solution

32. Maximization, graphical solution

M. Sensitivity analysis (2-32)

34. Minimization, graphical solution

35. Maximization, graphical solution

36. Maximization, graphical solution

37. Sensitivity analysis (2-36)

38. Maximization, graphical solution

39. Sensitivity analysis (2-38)

40. Maximization, graphical solution

41. Sensitivity analysis (2-^0)

42. Multiple optimal solutions

43. Int'easible problem

44. Unbounded problem

(b) •/4 : x, = (b)
 X2
(b)

10.

11.
14. The new objective function, Z= 300.ri +
fiOO.VT. is parallel to the constraint line for
platinum, which results in multiple optimal
solutions. Points B (A| = 2, xt = 4) and
C (.V| = 4, At = 3) are the alternate optimal
solutions, each with a profit of $3,000.

The feasible solution space will change. The

new constraint line, 3.r| + -Xx^ - 20. is
parallel to the existing objective function. Thus.
multiple optimal solutions will also be present
in this scenario. The alternate optimal solutions
are at point B (x\ = .33, xi - ^) 'ind
1

10 20 30 40 50 60 70
C (x\ = 5.33, .vt = 2), each with a profit of
$2,000.
17. The feasible solution space changes from the
area OABC to OAB'C. as shown on the
15.(a) Optimal solution: .V| = 4 necklaces, X2 = 3
following graph.
The maximum demand is not
bracelets.
X2
achieved by the amount of one bracelet.

(b) The solution point on the graph

which corresponds to no bracelets being
produced must be on the x\ axis where xi = 0.

This is point D on the graph. In order for point

D to be optimal, the objective function "slope"
must change such that it is equal to or greater
than the slope of the constraint line, 3x\ + 2x2 =
1 8. Transforming this constraint into the form y
=<:/ + hx enables us to compute the slope:
10 20 30 40 50 60 70
2.V2 =18- 3.V2
X2 = 9 - 3/2x2 The extreme points to e\aluate are now A. B\

From this equation the slope is -3/2. Thus, the and C.

slope of the objective function must be at least

A: X|=0
-3/2. Presently, the slope of the objective X2 = 30
function is -3/4: Z= 1.200

400.V2 = Z- 300x1 *BV X| = I5.8

.V2 = Z/400-3/4x, X2 = 20.5
The profit for a necklace would have to Z= 1.610

increase to $600 to result in a slope of -3/2: C: X| = 24

X2 =
400x2 = Z-600x,
= Z/400-3/2x|
.«:2
Z= 1.200

where both Point B' is optimal

However, this creates a situation

points C and D are optimal, ie.. multiple

optimal solutions, as are all points on the line
segment between C and D.

16.(a) maximize Z= 50x| + 40.V2 (profit. $)

subject to

3x| + 5.t2 < 150 (wool. yd2)

I Ox + 4.t2< 200
I
(labor, hr)

X|..t2 >0
18.

12

19.

X2
 30.

-6-4-2 2 4 6

27.

'
10 12

28. The problem becomes infeasible.

29. X2

12

10

2
 (b)

'A:x^=68.96 S.x, =96,77

X2 = 34.48 X2 =
Z= 410.35 Z= 47.59

Point A is optimal

100 120

33. No additional profit, freezer space

is not a bindina
"^
constraint.

34.(a) minimize Z= 200.^1 + 160x2 (cost, $)

subject to

6.V1 + 2x2 > 12 (high-grade ore, tons)

2x| + 2x2 > 8 (medium-grade ore, tons)
4x| -I- 12x2 > 24 (low-grade ore, tons)
-^:|,X2 >0

(b)

14
37. The slope of the original objective function is
(b)
computed as follows:

Z=30.V| +70.V2 18
70A:2 = Z-30;ti
X2 = Z/70 - 3/7^:,
slope - -3/7

The slope of the new objective function is

computed as follows:

2= 90.^1 +70x2

10x2 = Z-90x\
X2 = Z/70 - 9/7.V,
slope = -9/7

The change in the objective function not only

changes the Z values but also results in a new
solution point, C. The slope of the new
objective function is steeper and thus changes
the solution point.

.V|=0 C: .V, = 5.3

Jt2 = 8 X2 = 4.7
Z = 560 Z=806
A, =3.3 D: .V| = 8
X2 = 6.7 X2 =
Z = 766 Z= 720
41. The feasible solution space changes if the 43. X,

fertilizer constraint changes to 20x| + 20x2 -

80
800 tons. The new solution space \s A'B'C'D'.
Two of the constraints now have no effect. 70 Infeaslble Problem

60

50

40

30

20

10

10 20 30 40 50 60 70 80

20C''40 60 80 100 120 140

The new optimal solution is point C:

The graphical solution is displayed as follows. The graphical solution is shown as follows.

y
A

The optimal solution is .v = 1 , v = 1 .5. and

Z = 0.05. This means that a patrol sector is 1 .5

miles by 1 mile and the response time is 0.05

hr, or 3 min.
 Chapter Three

ANSWERS TO QUESTIONS
1. It is the model with the inequality constraints
converted to equations by adding slack
variables to < constraints and subtracting
surplus variables from > constraints.

2. Forxi =6.25, X2= 12.0:

9.6(6.25)+ 12.5(12.0) + .9, =210hrs.
s\ = hrs.
2 .7(6.25)
1 + 1 2.5( 1 2.0) + .vt = 500 lbs.
.92= 141.18 lbs.

At the optimal solution all of the labor hours

are being used but there are 141.18 extra
pounds of material unused.

3. If a resource is not being used it contributes

nothing to profit.

4. For.v, = 10,.X2 = 24:

7.2(10)+ 14.5(24) -,V| =420 A units

.S| =0 units
6.7( 10) + 3.8(24) - ,V2 = 150 B units
.92 = 8.2 units

At the optimal solution the requirement for A is

met exactly but the B requirement is exceeded
by 8.2 units..

110
 )

12. As the quantity value is increased the constraint 24. Computer solution; sensitivity analysis (3-22)
moves away from the origin altering the
25. Model formulation; standard form
solution point. When it reaches a new point
where the resource consumption changes this 26. Graphical solution; sensitivity analysis (3-25)
constitutes one limit. When the quantity value
27. Computer solution; sensitivity analysis (3-25)
is decreased the constraint moves inward
toward the origin. When it reaches a new point 28. Standard form
where the resource consumption changes this is
29. Model formulation; computer solution
the other limit.

30. Computer solution; sensitivity analysis

13. By solving the model again with the changes
included using the computer. 31. Model formulation; standard form

14. A shadow price indicates how much one 32. Graphical solution; sensitivity analysis (3-31)
additional unit of a constraint value will
33. Computer solution; sensitivity analysis (3-3 1
increase the objective function value in a
maximization problem. In a minimization 34. Model formulation; computer .solution

problem the shadow price indicates how much

35. Sensitivity analysis (3-34)
one less unit of a constraint value will decrease
the objective function value. 36. Model formulation; standard form

37. Computer solution; sensitivity analysis (3-36)

PROBLEM SUMMARY 38. Model formulation; standard form
1. Model formulation; standard form
39. Computer solution: sensitivity analysis (3-38)

2. Graphical solution; sensitivity analysis (3-1)

40. Standard form; computer solution
3. Sensitivity analysis (3-1)

4. Model formulation; standard form PROBLEM SOLUTIONS

5. Graphical solutions; sensitivity analysis (3^) 1. (a).vi = no. of basketballs
a:2 = no. of footballs
6. Sensitivity analysis (3—4)
maximize Z= I2x| + I6jc2

7. Model formulation; standard form subject to

8. Graphical solution; sensitivity analysis (3-7) 3x| + 2x2 ^ 500

4x| + 5jc2 < 800
9. Computer solution; sensitivity analysis (3-7)
JC|.X2>0
10. Model formulation; standard form

11. Graphical solution; sensitivity analysis (3-10) (b) maximize Z= 1 2.x\ + 1 6.V2 + O.V| + ().S2

subject to
12. Computer solution; sensitivity analysis (3-10)

3jc| + 2^2 + S] = 500

13. Model formulation; standard form
4jc| + 5.V2 + -V2 = 800
14. Graphical solution; sensitivity analysis (3-13) X|. ^2, .V|, 52 -

15. Computer solution; sensitivity analysis (3-13)

16. Model formulation; standard form

300
17. Graphical solution; sensitivity analysis (3-16)

18. Computer solution: sensitivity analysis (3-16)

19. Model formulation; standard form

20. Graphical solution; sensitivity analysis (3-19)

21. Computer solution; sensitivity analysis (3-19)

22. Model formulation; standard form

23. Graphical solution; sensitivity analysis (3-22)

 5

(a)/\;3(()) + 2(16()) + ,V| =300 3. (a)For c\ the upper limit is computed as

.V| = 180
-4/5=-r|/16
4(0) + 5(160) + .V2 = «00
5r| =64
.V2 =
f| = 12.8

fi.- 3(1 28.5) + 2(57.2) + 5:, =500

and the lower limit is unlimited.
jf| =0
4(128.5) + 2(57.2) + .V2 = 800 For ci the lower limit is,

.V2 = ^/5=-12/r2
C.-2(167) + 2(()) + .V| =500 4c2 = 60
.v|=0 c-2= 15

4(167) + 5(0) + .V2 = 800 and the upper limit is unlimited.

s^= 132
Summarizing.

(b)Z= + oo<C| < 12.8

\2xi 16.X-2

1 5 < c 2 < °°
and,
For f/i the upper limit is °o since no matter how
xi =Z/I6- 12.V|/16
much q\ increases the optimal solution point /4
The slope of the objective function, -12/16,
will not change.
would have to become steeper (i.e., greater)
than the slope of the constraint line The lower limit for q\ is at the point where the
4.V| + 5.V2 = 800, for the solution to change. constraint line 3x| + 2jC2 = q\ intersects with
point A where x\ = 0, jr2 = '60,
The profit. c\, for a basketball that would
change the solution point is, 3x| + 2.11:2 = q\
3(0) + 2(l60") = f/|
4/5=-C!/16
./I
=320
5ri =64
ri = 12.8 For qi the upper limit is at the point where the
rubber constraint line (3,V| + 2.V2 = 500)
Since $13 > 12.8 the solution point would
intersects with the leather constraint line
change to B where x\ - 128.5, X2 = 57.2. The
(4.V| + 5.V2 = 800) along the .V2 axis, i.e., .V| = 0,
new Z value is $2,585.70.
X2 = 250,
For a football,
4.V|+ 5.V2 = f/2
-4/5 = -12/0 4(0) + 5(250) = f/2
4f2 = 60 q2 = .250 1

('2 = 1

The lower limit is since that is the lowest

Thus, if the profit for a football decreased to point on the .V2 axis the constraint line can
$ 5 or
1 less, point B will also be optimal (i.e., decrease to.

multiple optimal solutions). The solution at B is

Suminarizing,
.V| = 128.5, .V2 = 57.2 and Z= $2,400.
320 < ^71 <oo
()<^/2< 1.250
(c)If the constraint line for rubber changes to
3.V| + 2.V2 = 1,000, it moves outward,
eliminating points B and C. However, since A is

the optimal point, it will not change and the

optimal solution remains the same,
A, = 0. .V2 = 160 and Z= 2,560. There will be
an increase in slack, .V|, to 680 lbs.

If the constraint line for leather changes to

4a| + 5.V2 = 1 ,300, point /\ will move to a new
location. .\ , = 0. .V2 = 250. Z = $4,000.

18
(b)

***** Input Data *****

Max. Z= llti + I6.r2

Subject to

t| 3.vi+2.V2<500
C2 4.x+ 5.V2 < 800
I

***** Program Output *****

Final Optimal Solution At Simplex Tableau : 1

Z= 2560.000

Variable
 . 5

4. (a).vi = no. of units ot ,4 (c)In order for the optimal solution point to
X2 = no. of units of fi change from B to y4 the slope of the objective
maximize Z= 9x\ + 7x2 function must be at least as flat as the slope of
subject to the constraint line, 4x| + 8,V2 - 40. which is

-1/2. Thus, the profit for product B would have

12.V| +4x2<60
to be,
4x| + 8x2 ^40
X|,X2>0 -9/r2 = -l/2
(-2=18

(b) maximize Z = 9x| + 1x2 + ()s\ + O^vt If the profit for product B is increased to $ 1

subject to the optimal solution point will not change,

although Z would change from $57 to $8
I2x| +4.V2 + .V| =60 1

If the profit for product B is increased to $20

4.V| + 8x2 + ''2 = 40
>
the solution point will change from B to A.
x\, X2, .V|, .V2
X, r=0,XT = 5,Z=$iOO.

5.

30
(b)

:::****
Input Data *****

Max.Z = 9.V| +7.V2

Subject to

(•| 12.V| +4.V2<60

('2 4.V|+8.V2<40

***** Program Output *****

Final Optimal Solution At Simplex Tableau : 2

Z= 57.000

Variable
7. (a)A| = no. of yards of denim
XT = no. of yards of corduroy
maximize Z = $2.25x1 +3.10x2
subject to

5.0x1 + 7.5x2 < 6,500 ^^

3.0x1 + 3.2x2 ^ 3,000 -^^
X2<510 ,

X|, X2 >

(b) maximize Z = $2.25xi + 3. 10x2 + O.vi +

+ ().V3
().V2

subject to

5.0x1 +7.5x2 + 51 =6,500

3.0x1 +3.2x2 + .92 = 3,000
.x-2 + .V3 = 5IO

X|, .V2, .S'|, .V2, .V3 >

***** Input Data *****

Max. Z= 2.25a I
+3.1.X2

Subject to

t| 5.V1 + 7.5.r2 < 6500

(-2 3.V| + 3.2.V2 < 3000
c I.V-. <5I0
***** Program Output *****

Final Optimal Solution .At Simplex Tableau : 2

Z= 2607.000

Variable
 (b) 0<ri< 2.906 6,105 <c/i <^ (a)6(4) + =
2(())-.vi 12
2.4<tT<oo 1,632 < 92 ^3,237 = .v| 12
< 93 < 692.308 2(4) + 2(0)-.v2 = 8
.V2 =
The demand for corduroy can decrease to zero
4(4)+ 10({))-.v:( = 5
or increase to 692.308 yds. without changing
V3 = 1 1

the current solution mix of denim and

corduroy. If the demand increases beyond
692.308 yds., then denim would no longer be (b)The slope of the objective function.
produced and only corduroy would be -6000/7,000 must become flatter (i.e.. less)

produced. than the slope of the constraint line,

2xi + 2x2 = 8, for the solution to change. The
cost of operating Mill 1. ci. that would change
lO.(a) A| = no. of days to operate mill 1

the .solution point is.

XT = no. of days to operate mill 2
minimize Z= 6,000.ti + 7,000jc2 -r|/7,000 = -l
subject to CI =7.000

6xi +2jc2> 12 Since $7,500 > ,$7,000, the solution point will
2x1 + 2x2 > 8 change to B where xi = I , .o := 3, Z= $28,500.
4xi + 10x2 > 5
X|,X2>0
(c) If the constraint line for high-grade aluminum
changes to 6x| + 2.o = 10, it moves inward but
(b)minimize Z= 6,000x1 + 7,000x2 + Os\ + does not change the optimal variable mix. B
0,V2 + 0.V3 remains optimal but moves to a new location,
subject to XI = 0.5. X2 = 3.5. Z= $27,500.

6.V I
+ - 12
2.V2 - v I

2.V I
+ = 8
2x-2 - .V2

4x| + 10x2 -.V3 = 5

X|, X2. .V|, si. XT, >

X2 = 6
Z= 42,000
12.

***** Input Data *****

Min.Z = 6()0().V| + 700(),V2

Subject to

t| 6.V| +2.V2> 12
2.V, + 2.V2 > 8
4x^+ Kh-T >5
***** Program Output *****

Final Optimal Solution At Simplex Tableau : 5

Z = 24000.000

Variable
 (b) ()<r| < 7.000 o°<qi <24
6.000 < o < oo 4 < q2 — °°
°°<qT,< 16

(c) There will be no change.

13.(a) .V| = no. of acres of corn

.V2 = no. of acres of tobacco
maximize Z= 300;f I
+ 520jc2
subject to

XI +JC2<4I0
I05.VI + 210x2 < 52,500
X2< 100
j:|,j:2^0

(b)maximize Z= 300.ri + 520x2 + O.V| + O.V2 + O53

subject to

.ri +X2 + .VI =410

I05,r| + 210.1-2 + .V2 = 52,500
X2 + .V3 = 00 1

X\,X2,S\, ,V2, .V3 >0

14.
15.

***** Input Data *****

Max. Z=300.r| + 520.r2

Subject to

ci l.V] + l.V2<4I0
o I05.VI +210;c2<52500
C2 I.V2 < 100

***** Program Output *****

Final Optimal Solution At Simplex Tableau : 3

Z= 142800.000

Variable
 X

(b)The shadow price for the budget is $2,095. (b)The slope of the objective function. -6/5. must
Thu.s, for every SI dollar borrowed they could become flatter (i.e.. le.ss) than the slope of the
expect a profit increase of S2.095. If they constraint line, .04xi + .04.r2 = 16. for the
borrowed $ ,000 it would not change the
1 solution to change. The profit for ham, r^. that

amount of corn and tobacco they plant since would change the solution point is,

the sensitivity range has a maximum allowable

-0.6/f2 = -l
increase of $1,050.
C2 = .60

Thus, an increase in profit for ham of 0.60 will

I6.(a) x\ = no. of sausage biscuits
create a second optimal solution point at C
X2 = no. of ham biscuits
where XI = 257, X2 = 143 and Z = $225.70.
maximize Z= .60xi + .50j:2
(Point D would also continue to be optimal,
subject to
i.e.. multiple optimal solutions.)
.10x1 <30
,15x2<30
(c) A change in the constraint line from.
.04x1 + .04x2< 16
.04x1+ 04x2 = 1 6 to ,04xi + .04x2 = 1

.01x1 +.024x2<6
would move the line outward, eliminating both
X|.X2>0
points C and D. The new solution point occurs
at the intersection of O.Ol.vi + .024.V2 = 6 and
(b)maximize Z = S0.60x| + 0.50x2 + Osj + 0^2 .lOx = 30. This point is xi = 300. .o = 125. and
+ OST, + 054 Z= $242.50.
subject to

.10x1 +.V1 =30 18.

+ .^2 = 30
.15.V2
***** Input Data *****
,04x1 + .04x2 + 53 = 16
.01x1 +0.024x2 + 54 = 6
Max. Z= .6x| + .5.V2
X[, X2, .Vj, ,V2, ST,, S4 >

Subject to

17.
C| .lxi<30
C2 . 1 5x2 ^ 30
C: X, = 257 ct, .04x| +.04v2< 16
X2= 143
C4 .01x| +.024.^2 <6
Z= 225.70
***** Program Output *****
'D: X, = 300
^2= 100
Z=230 Final Optimal Solution At Simplex Tableau : 2

Z= 230.000

Variable

100 200 300 400 500 600 700 800

(a)xi =300,X2= 100, Z= $230

.10(300) + VI
, =30
= left over sausage
Vj

.15(100) = 30
+ .V2

.V2 = 5 lbs. left over ham

1

.()1(3()0) + .024(I00) + ,V4 = 6

.V4 = 0.6 hr.
Objective Coefficient Ranges

Lower Current Upper Allowable Allowable

Variables Limit Values Limit Increase Decrease

A| ().5()() ().6()() No limit No limit O.IOO

X2 0.000 0.500 0.600 0.100 O.MM)

Rii!ht Hand Side Ranges

 (b)minimize Z= 50x| + TO.vt + 0s\ + Osj + Ost,

subject to

80x, +40x2-5| =3,000

80x| -S2= 1,000
40x2 - ^3 = 800

20.
-
Objective Coefficient Ranges
23. 24.

600
Objective Coefficient Ranges
Program; Linear Programming

Problem Title: 3-24c

='=**=^=*
Input Data =>=****

Max. Z= 3.V| +4x2

Subject to

(| I.V| + I.V2>400
ri .6x1 - .4x2 >0
(3 1x2 < 250
(4 lx| - 3.V-2 =
(•5 Ix| + lx2<500

::^**** Program Output *****

Final Optimal Solution At Simplex Tableau : 4

Z= 1625.()()()

Variable
25. (a).vi = $ amount invested in land 27.
X2 = S amount invested in cattle
maximize Z= l.20jri + 1.30.V2

subject to

.VI +.V2< 95,000

.I8.VI + < 20.000
.30.V2

.Vl..V2>0

(b)maximize Z= 1 .20.vi + 1 .30.V2 + O^i + 0.?2

subject to

.VI +-V2 + .y| =95.000

.18.VI +.3av2+ 52 = 20.000
.V|. .V2. .5|. 52 >

"z
26.
Objective Coefficient Ranges
Objective Coefficient Ranges

Lower Current Upper .Allowable Allowable

Variables Limit Values Limit Increase Decrease

•Vl
Objective Coefficient Ranges
Variable
31. (a) Maximize Z= $0.50x| + 0.75x2
subject to:

$0.i7x| + 0.25x2 < $4,000 (printing budget)

x\ + X2'^ 18,000 (total copies, rack
space)
x\ > 8,000 (entertainment
guide)
X2 S: 8,000 (real estate guide)
xi,X2^0

(b) Maximize
Z= 0.50x| + 0.75x2 + OSi + OS2 +
OSi, + OS4
 (c) Maximize Z = 7500.t| + 8.200.V2 + 39. xi3 = 350 tons
lO.5OO.r3 + OSi + OS2 + 053 + 054 X21 = 158.333 tons
subject to: .t22 = 296.667 tons
.2i.V| + + .18,V3 +
.24.V2 5| = 17 X23 = 75 tons
A] + X2 + .V3 + 52 = 80 .V31 =610 tons
12.r| + 14.5JC2 + 16x3 + 53 = 2.500 X42 = 240 tons
— — .V2 + 2jc3 + 54 =
.t|
Z= $77,910
.V|.A-2. .V3 >
Mine 1 = 350 tons
Mine 2 = 530 tons
37. .V| = 20
Mine 3 = 610 tons
X2 = 33.3334
Mine 4 = 240 tons
A-3 = 26.6667

Z = S703.333.40
Multiple optimal solutions exist

< (a)Mine 4 has 240 tons of "slack" capacity.

(a)The sensitivity range for A2 is 7.500 q—
8.774.999. Since $7,600 is within this range the (b)The dual values for the 4 constraints
values for A|. A2. and xt, would not change, but representing the capacity at the 4 mines show-
the profit would decline to $683,333.30 (i.e., that mine I has the highest dual value of S61.
less the difference in profit. ($600)(a2 = so its capacity is the best one to increase.
33.3334)
(c)The sensitivity range for mine 1 is 242.8571 <
(b) One ton of grapes; the dual value is S23.333.35 C| < 414.2857. thus capacity could be
increa.sed by 64.2857 tons before the optimal
(c) Grapes: (0.5)($23.333.35) = S 11. 666.68
solution point would change.
Casks: (4)(S3.833.329) = $15,333.32
Production: SO (d) The effect of simultaneous changes in

Select the casks. objective function coefficients and constraint

quality values cannot be analyzed using the
(d) S6.799: slightly less than the lower band of the
sensitivity ranges provided by the computer
sensitivitv ranae for c,.
output. It is necessar)' to make both changes in

the model and solve it Doing so results

again.
38. (a)Minimize Z= S37.r|| + 37ai2 + 37a|3 + in a new solution with Z = S73.080, which is
46a2i + 46a22 + 46a23 + 50x3] + 5OJC32 + $4,830 less than the original solution, so

5O.V33 + 42a4| + 42.X-42 + 42a43 Exeter should make these changes.

subject to:
.7x11 + •6JC21 + ^-^'Sl + 3-'^4i = 400 tons 40. minimize Z= 8.Zr|+ 7.0.V2 +6.5.V3 + 9.0.1:4 +
.7x|2 + -6x22 + -5x32 + .3.r42 = 250 tons 0,v I
+ 0^2 + O.V3 + O54
.7x|3 + .6.r23 + .5x33 + -3-^43 = -90 tons subject to
+ ^12 + ^\3 — 350 tons
-'^ll
6x1 + + 5x3 + 7x4 - s\= 820
2.r2
^21 + '^22 + -*^23 — 530 tons
- - .3x3 ~ -3x4 - 52 =
.7x| .3.^2
^31 + ^32 + -^33 — 610 tons
-.2x1 + + -^3 ~ 2-'^4 + ^3 =
X41 + X42 + X43 < 490 tons
-'^2

.V3-X1 -.C4 — .V4 =

(b) Minimize Z= S37xii + 37x|2 + 37xi3 +
46x21 + 46x22 + 46x23 + 50x3 1 + 50x32 +
Program: Linear Programming
50x33 + 42x41 + 42x42 + 42x43 + OS| +
052 + 053 + 054 Problem Title: 3-31
subject to:
***** Input Data *****
.7x11 + -6^21 + -5x31 +400 .3x41 =
.7X12 + •6'Y22 +-5X32 .3X42 +250 =
Max. Z = 8.1v| + 7.O.V2+ 6.5.r3 + 9.0.t4
.7X13 + •6.V23 +-5X33 3.^43+290 .
=
^[l + -^12 + -'^13 + -^1 350 = Subject to
^21 + '^22 + -^23 + ^^2 =^ 530
X31 + .V32 + X33 + 53 = 610 c\ + 2x2 + 5.V3 + 7.V4 > 820
6x1

X4\ + Ml + ^43 + 54 = 490 ci - .3.r2 - .3.r3 - .3.>:4 >

-7x1

x,„ s, > C3 -.2x1 + ••'^2 + '-^3 - •2j^4 -

C4 -Ixi + l.r3- 1x4 >0

41
 ***** Program Output ***** (d)x| = 56.70. X2= 154.64

Infeasible Solution .30(56.7) + .25(1 54.64) + .V

I
=60
.V| = 4.33 hr. of molding time
because Artificial variables remain in the fmal .27(56.7) + .58(1 54.64) + .92= 105
.V2 = hr. of baking time
tableau.
.16(56.7) + .20(154.64) + .V3 =40
***** End of Output ***** .S3 = hr. of glazing time
32.8(56.7) + 20( 54.64) +
1 .V4 = 6,000
.V4= 1,047.42 lbs. of clay
CASE SOLUTION:
MOSAIC TILE COMPANY
(e)The optimal solution is at point B. For point C
(a)maximize Z = $190a:| + 240x2 to become optimal the profit for a large
subject to tile, X|, would have to become steeper, than the
constraint line for glazing, .
16xi + .20x2 = 40:
.30x| + .25x2 ^ 60 hr.- molding
.27x| + .58x2 < 105 hr. - baking |/240: ,16/. 20

.16x| + .20x2 - 40 hr.- glazing 192

32.8x, + 20x2 ^ 6,000 lb. - clay
This is the upper limit for c\. The lower limit is
X|,X2, >0
at point A which requires an objective function
slope flatter than the constraint line for baking.
(b)maximize Z= $190x| + 240x2 + 0.?l + ^^2 +
-C|/240 .277.58
0.^3 + O.s-4
111.72
subject to

+ .25.^2 ^-.^i =60 Thus, lll.72<r| < 192

.30x|
.27x1 + .58x2 + .y2= 105 The same logic is used to compute the
.16x| + .20x2 + ^3 = 40 sensitivity range for ci- The lower limit is

32.8x| + 20x2 + -^4 = 6,000 computed as,

X|, X2, .^|, .^2' ''3' ^ •S'4

-190/c2 = -l6/.20
(-2 = 237.5
(c)
The upper limit is,

D: X, = 136.36
181.03 Xj = 76.36 -190/f 2 = .277.58
Z= $43,447.20 Z= $44,234.80 r2 = 408.15
*e: X, = 56.70 E: X, = 1 82.93
=
The sensitivity ranges for the constraint
X2= 154.64 X2
Z= $47,886.60 Z= $34,756.70 quantity values are determined by observing

=
the graph and seeing where the new location of
C: X, 100
Xj = 120 the con.straint lines must be to change the
Z= $47,800 solution point.

For the molding constraint, the lower limit of

the range for (i\
is where the constraint line

intersects with point B.

.30(56.7) + .25(1 54.64) = ^/,

c/i =55.67

The upper limit is 0° since it can be seen that

this constraint can increase indefinitely without
changing the .solution point.

Thus,

55.67 <<y, <°o

42
 For the baking constraint the lower limit of the (g)
range for </2 is where point C becomes optimal,
and the upper limit is where the baking Program: Linear Programming
constraint intersects with the xi axis (xj = 200).
Problem Title: Case Problem: Mosaic Tile Company
AtC: .27(1 00) + = 92
.58(1 20)
***** Input Data *****
92 = 96.6
At X2 axis: .27(0) + .58(200) = 92 Max. Z= 190.V1 +240.V2
92= 116
Subject to
Thus,

96.6 < 92^ 116

C| .30.t| +.25x2<60
C2 .27x| +.58.t2< 105
For the glazing constraint the lower limit of the C3 .16x1 +.20.r2<40
range for 93 is at point A, and the upper limit is C4 32.8a: +20.^2^6000
I

where the glazing constraint line. 16.X| + .20^2 .

= 40, intersects with the baking and molding

constraints (i.e.. .V| = 80.28 and .vt = 143.68).

MA: .16(0) + .20(1 8 1.03) = 93

93 = 36.21

At intersection of constraints:

.16(80.28) + .20(1 43.68) = 93

93 = 41.58

Thus,

36.21 > 93 > 41.58

For the clay constraint the upper limit is °o

since the constraint can increase indefinitely.

The lower limit is at the point where the
constraint line intersects with point B:

AtB: 32.8(56.7) + 20(154.64) = 94

94 = 4,952.56

Thus,

4,952.56 < 94 < °o

(f) The slope of the objective must be flatter than

the slope of the constraint that intersects with
the xi axis at point -4, which is the baking
constraint,

-190/c2 = .27/.58
= $408.14

43
***** Program Output *****

Final Optimal Solution At Simplex Tableau : 2

Z = 47886.598

Variable
 hours the problem should be solved again using CASE SOLUTION: The Rowntown Cab
the computer with this change. This new Company
solution results in a profit of $5 ,736.98. an
1

increase in profit of only $3,490.39. The reason

(a)Maximize Z = XO.v, + .^OO.v: + 420x3 +
for this small increase can be observed in the
graphical solution; as the glazing constraint
300x4 + 270.V5 + 21().Vf,

subject to
increases it quickly becomes a "non-binding"
new .r| + X2 + XT, + .V4 + .C5 + X(y = 70
constraint with a solution point.
X6 + xi > 10
X| + XT s 12
(k)A reduction of 3 hours is within the sensitivity xj + .r^ > 20
range for kiln hours. However, the shadow x^ + X4 > 25
price for kiln hours is $1,170. 103 per hour. X4 + X5 > 32
Thus, a loss of 3 kiln hours will reduce profit X5 + Xf, > 18
by (3)( 1.1 70. 103) = $3.5 10.31. X, >
Where .r, = the number of drivers who start

their 8-hour shift in period /'.

( = 1.2 6
(period 1 = midnight to 4:00 a.m.: period 2
4:00-8:00 a.m.: period 3 = 8:00 a.m.-noon:
period 4= noon^:00 p.m.: period 5 =
4:00-8:00 p.m.: period 6 = 8:00
p.m.-midnight).

Solution:

xi = 0. X2 = 27, X3 = 1 : X4 = 24. .vs = 8,

X6 10, Z= $25,380

(b)Add the constraint. .Vf, -I- X| + .vt ^ 15

Solution:

xi = 7, X2= 5. X3 = 23, .V4 = 17. xs =

.V6 = 3. Z = $22,500

(c) Add the constraint. .V3 < 20

Solution:

XI = 7. X2 = 5, X3 = 20, .1-4 = 20, xs = 15.

X6 = 3, Z= $22,140

45
 ,

Chapter Four

ANSWERS TO QUESTIONS Further limiting the servings of the four hot and
cold cereals to two cups, X| + Jt? + xt, + ^4 < 2,
1. (a)Since the profit values would change, the results in the following solution:
shadow prices would no longer be effective.
X3 = 2 cups of oatmeal
Also, the sensitivity analysis provided in the
X(, - .750 slices of bacon
computer output does not provide ranges for
Xf^ - 2 cups of milk
constraint parameter changes. Thus, the model
.X9 = .5 cups of orange juice
1 1
would have to be resolved.
X|o = 4 slices of wheat toast
(b) The reformulated mode! would have unit costs Z= $0,925
increased by 10 percent. This same amount
The new requirement is that,
would be subtracted from unit profits. The
individual processing times would be reduced X, =X2 = Xt,=X4
by 10 percent. This would result in a new,
This can be achieved within the model by
lower solution of $43,3 1 0. Thus, the suggested
creating three additional constraints.
alternative should not be implemented.

2. Probably not. The t-shirts are a variable cost

X\ -Xl
purchased by Quik-
x\ =X3
and any additional t-shirts

which
X\ =X4
Screen would likely reduce unit profit,

would shadow price for

change the current If xj equals xi, x->, and X4 then xi, X3 and X4 must
blank t-shirts. The shadow price is effective also equal each other. These constraints are
only if the profit is based on costs that would changed to,

be incurred without regard to the acquisition of

X| -.1C2 =
additional resources.
xi -X3 =
3. With a minimum of 500 calories, the three food xi -X4 =
items remain the same, however the amount of
= 2.995 The new .solution is X| = .V2 = x^ =X4 = 1 1 2.5.
each and the total cost increases: xt,

cups of oatmeal, x^ = 1.352 cups of milk, x\q = 6. It would have no effect; the entire $70,000
1 .005 slices of toast, and Z= $0,586. would be invested anyway.

With a minimum of 600 calories, the food Since the upper limit of the sensitivity range for
items change to - 4.622 cups of oatmeal and
xt, the investment amount is "unlimited," an
1 .378 cups of milk and Z = $0,683. A change of increase of $10,000 will not affect the shadow
variables would be expected given that 600 price, which is $0,074. Thus, the total increase
calories is greater than the upper limit of the in return will be $740 (i.e., $10,000 x .074 =
sensitivity range for calories. $740).

4. Many different combinations of maximum No, the entire amount will not be invested in
servings of each of the 10 food items could be one The new solution is
alternative.
used. As an example limiting the four hot and X| = $22,363,636, X3 = $43,636,364, and
cold cereals, .V| , .ri, x^ and X4 to four cups, eggs .if4 = $14,000.
to three, bacon to three slices, oranges to two,
milk to two cups, orange juice to four cups and
The shadow price is 1 .00, thus for every $ I

increase in budget, up to the sensitivity range

wheat toast to four slices results in the
upper limit of $14,000, audience exposure will
following solution:
increase by 1.00. The total audience increase
x-x, = 2 cups of oatmeal for a $20,000 budget increase is 20,000.
.V4 = .464 cups of oat bran
1

.V5 - .065 eggs This new requirement results in two new model
constraints,
jcg = 1.033 cups of milk
JTIO = 4 slices of wheat toast 20.00Ov| = I2.00().V2
Z= $0,828 20,000x1 = y,0()()A

46
 )

or, increa.se of l,7(X) bbls. or 6,200

total bbls. of

20.000- I2,000.t: = component For example an increase of one

I.

- 9.OOO.V3 = bbl. of component from 4.500 to 4.501 results

I
20.000.vi
in a increase in total cost to $76,820.
The new solution is .V| = 3.068, xi = 5. 14. 1

15, Increasing the regular production capacity for

.x^ = 6.8 8 and Z =
1 84.090. This results in
1

each week to 165 results in a new solution that

approximately 61 .362 exposures per type of
reduces the total cost by S2.000 to S2 14.300.
advertising (with some slight differences due to
Increasing the regular production capacity for
computer rounding).
week 6 only to 190 units results in a new
10. The slack \'ariables for the three < warehouse solution that reduces the total cost by $2.4(X) to
constraints would be added to the constraints as $213,900. This could have been determined by
follows. multiplying the shadow price (-80.00) by 30
units.
-»^IA+-V|B + V|C + .V, =300
-'^2A + -^28 + ^2C + ^2 = 200 Yes, the second alternative results in a greater
•'^.^A + •'^38 + ^3C + ^^ = 200 cost reduction than the first.

These three slacks would then be added to the

objective function with the storage cost PROBLEM SUMMARY
coefficients of S9 for 5|, S6 for S2 and
S7 for St,.
1. Product mix (maximization)

2. Sensitivity analysis (4-

This change would not result in a new solution. 1

3. Diet (minimization)
11. The model must be reformulated with three
new variables reflecting the shipments from the 4. Product mix (minimization)
new warehouse at Memphis (4) to the three
5. Product mix (maximization)
stores. JC4A, ^^48' and ^4^. These variables must
be included in the objective function with the 6. Product mix (maximization)
cost coefficients of SI 8. 9 and 12 respectively.
7. Product mix (maximization)
A new supply constraint must be added.
8. Ingredients mix (minimization)
X4A + X4B + X4C < 200
9. Transportation (maximization)
The .solution to this reformulated model is.

10. Product mix (maximization)

.Vic = 200
11. Ingredients mix (minimization)
.V2B = 50

V3A= 150 12. Crop distribution (maximization)

X4B = 200
13. Monetary allocation (maximization)
Z = 6,550
14. Diet (minimization), sensitivity analysis
Yes, the warehouse should be leased.
15. Transportation (maximization)
12. The shadow price for the Atlanta warehouse
shows the greatest decrease in cost. S6 for 16. Transportation (minimization)
every additional set supplied from this source.
17. Warehouse scheduling (minimization)
However, the upper limit of the sensitivity
range is 200, the same as the current supply 18. School busing (minimization)
value. Thus, if the supply is increased at
19. Sensitivity analysis (4-16)
Atlanta by even one television set the shadow
price will change. 20. Ingredients mixture (minimization)

13. This change would not affect the solution at all 21. Interview scheduling (maximization)

since there is no surplus with any of the three 22. Investments mixture (maximization)
constraints.
23. Insurance policy mix (maximization)
14. Component I has the greatest dual price of S20.
24. Product mix (maximization)
For each barrel of component 1 the company
can acquire, profit will increase by S20. up to 25. Advertising mix (minimization),
the limit of the sensitivitv range which is an sensitivitv analysis

47
26. Blend (maximization) (b)This change requires a new variable, .V4. and
that the constraint for apples be changed from
27. Muitiperiod borrowing (minimization)
< to =. No, the Friendly's should not produce
28. Muitiperiod production scheduling (minimization) cider.The new solution would be .V| = 35, 1

X2 = 0, X3 = 0, X4 = 8.33 and Z = $26,475. 1

29. Blend (maximization), sensitivity analysis
This reduction in profit occurs because the
30. Assignment (minimization), sensitivity analysis
requirement that all 6,500 apples be used
31. Transportation (minimization) forces resources to be u.sed for cider that would
be more profitable to be used to produce the
32. Scheduling (minimization)
other products. If the final model constraint for
33. Production line scheduling (maximization) apples is, < rather than =. the previous solution
in Kb) results.
34. Network flow (minimization)

35. Investment example in chapter (maximization)

3. (a).V| = no. of eggs
36. Blend (maximization) X2 = no. of bacon strips

37. Trim loss (minimization) x->, - no. of cups of cereal

38. Muitiperiod investment (maximization) minimize Z= 4.V| + 3.V2 + 2.i[:3

subject to
39. Muitiperiod sales and inventory (maximization)
2x\ +4x2 +f.^ - 16
40. Muitiperiod production and inventory
(minimization)
3x| +2^2 + .V3 > 12

jri, ^2, .V3 >

41. Employee assignment (maximization)

42. Network flow (maximization) (b).V| =2

43. Muitiperiod workforce planning (minimization) JC2=3
Z = $0.17
44. Integer solution (4-43)

45. Machine scheduling (maximization),

4. (a).v, = number of boats of type /. / = I (bass boat),
sensitivity analysis
2 (ski boat), 3 (speed boat)
46. Cargo storage (maximization)
In order to break even total revenue must equal
47. Product mix/advertising (maximization) total cost:

48. Scheduling (minimization) 23,0()O.V| + I8,()0().V2 + 26,000.^3 = I2.500.r|

49. Truck purchasing/leasing (minimization) + 8,500.V2 + 13,7()O.X3 + 2,800.()()0

50. Muitiperiod workforce planning (minimization) or,

51. Muitiperiod workforce (4-50 10,50().V| + 9,50().V2 + i2.3OO.v3 = 2.800.000

minimize Z= I2.500a-| + 8.50().V2 + I3.700.V3

PROBLEM SOLUTIONS subject to

1. (a)ina\imizeZ = $l9().V| + 17().V2+ 155.V3 1(),500.V| + 9.500.V2 + I2,3()0.v:i = 2.800.000

suhject to v, > 70
.V2 > 50
3.5.V| + 5.2.V2 + 2.8.V3 < 500
.V3 > 50
I.2.V, +().S.V2+ 1.5.V3<24()
.V| < 120
40a| + 55.V2 + 2().V3 < 6,500
.V2 < 1 20
.V1..V2..V3 >
.V3 < 1 20
vi, .v->, .va >
(b)v| =41.27. A-2 = O..V3= 126.98,2 = $27,523.8!
,v| =,S2 = 0. .S3 = 2,309.52
(b).V| = 70.00
2. (a)It would not affect the model. The slack apples .V2 = 2().()0
1

arc multiplied by the revenue per apple of $.08 .V3 = 75.203

lo determine the extra total re\'enue, i.e.,
Z= $2,925,284,553

(2,309.52)($.()8) = $ 184.76.

48
 2
1

(a)A| = no. of clocks (b).ri = 5.000

.vt = no. of radios xi = 1 .000
.V3 = no. of toasters .V4 = 5.000
Z= S60.000
maximize Z= 8.X| + IO.vt + Ix-i,

subject to
8. (a).r, = ore/(/= 1.2.3.4.5.6)
7.V| + 10.V2 + 5.V3< 2.000
2x1 + "^^2 + 2x3 < 660 minimize Z= 27.V| + 25.vt + 32v3 + 22.V4 +
JC| < 200 2av5 + 24.vf,

JC2 < 300 subject to

.V3< 150
. 1 9.V1 + .43.^2 + 7.V3 + .2O.V4 +. 2a,s >
. 1 I .2
x\, .V2, .V3 >
. 1 5.V, + . 1 0.V2 + 2.V4 + .24.V5 + 8.Vf, <
. 1 . 1 . 1

. 1 2.r I + .25.x:2 + Oxs + bx^, < . 1 . I .07

(b).V| = 178.571 . 1 4.V| + .07x2 + 53.V3 + 8x4 + .2 .v, + .25.r6 >
. 1 1 .30
.V3 = 50.(M)
1 . 14x| + .07.V2 + .53.t3 + 8.V4 + .3 1x5 + .25.rf, <
. 1 .65
Z=S2.478.571 .60x| + .85x2 + .70x3 + -50x4 + .65x5 + .7 l.v^ = 1 .00

6. (a)x| = no. of gallons of Yodel (b).V2 = .1 153 ton -ore 2

X2 = no. of gallons of Shotz X3 = .8487 ton - ore 3
X3 = no. of gallons of Rainwater X4 = .0806 ton - ore 4
X5 = .4 6 ton - ore 5 1 1

maximize Z= 1.50.r| + 1.60x2+ '-25x3

Z = S40.05
subject to

X| +.V2+X3= 1.000
(a)Xy = number of trucks assigned to route from
1 .50x| + .90.V2 + .50x3 < 2.000
warehouse / to terminal 7. where / = I

X, < 400
(Charlotte). 2 (Memphis). 3 (Louisville) and
X2 < 500
y = a (St. Louis), b (.-Xtlanta). c (New York)
X3 < 300
X|. .V2. X3 > maximize Z= I .SOOx,,, + 2.100x|/, + 1.600x|..
+ I .OOO.r2„ + 700x2/, + 900x2, + -4(X)x3„ '

+ 800x3/, + 2.200x3,
(b)X| =400
subject to
X2 = 500
.V3= 100
Z =51.525.00 Xju = 30
+X2h + -V2,

X?,ci = 30
+ X3b + -'^.^r

X|u + X2,, + X3„<40

7. (a)X| = no. of lb of Two
Super at Fresno
X\h + X2h + X^h^^G
V2 = no. of lb of Super Two at Dearborn
X|< +X2, +X3, <50
X3 = no. of lb of Green Grow at Fresno
X., >
.V4 = no. of lb of Green Grou at Dearborn

maximize Z= 7x| + 5.V2 + 5x3 + 4x4

(b)X|,, = 30
subject to
X2„ = 30
lv| + 4.V2 + 2x3 + 3.V4 < 45.000 xl, = 30

xi +X2< 6.000 Z = S 59.000 1

+ X4 < 7.000
X3
+.V3< 5.000
X|
X2 + X4 < 6.000
X|,X2, X3, X4 >

49
10. (a)A| = no. of sofas subject to
X2 = no. of tables
.V3 = no. of chairs

maximize Z = 400ai + 275;c2 + 190jc3

subject to

7.V| + 5.V2 + 4x3 < 2,250

I 2a- +7a:3< 1,000
I

6a- I
+ 9a:2 + 5a:3 < 240
x\ + A2 + X3 < 650
A'l, A-2, A-3 >

(b)A, =40
Z = $16,000

II. (a) xij = lbs. of seed / used in mix j, where / = t

(tall fescue), m (mustang fescue), b (bluegrass)

and./= 1,2,3.

minimize Z= 1 .70 (a,| + a:,2 + x,t,) + 2.80

(-^'m I
+ -^»,2 + •'^/«3) + 3-25 (a:/,i + x/,2 + x/,3)
subject to

.5()a-,| -.50a,„| -.50x/,| <0

-.20a-,| +.80x,„|-.20x/,| >0
-.30a-,2 - .30x,„2 + •70x/,2 >O
-.30.\-,2 + .70x„,2 - .30x/,2 >
.80x,2 - .2a<:„,2 - .20x/,2 <
.50x,3 - .50.v„,3 - .50x/,3 >
.3()x,3-.70a-,„3-.70x/,3<0
-.I()v,3-.I0a-,„3 + .90x/,3>0
.v,| +x„,| +x/,| > 1,200
•^'/2 + +
-'^'/»2 - 900
•'^/)2

x,3 + A-„,3 + x/,3 > 2,400

X,y>0

(b)A,| =(-)()0

v,2= 180
A,^= 1,680
.v,„|=600
.v,„2 = 450
.v,„3 = 480
v/,|=0
.v/,2 = 270
V/,3 = 240
Z= $10,123.50

12. (a) xii = acres of crop / planted on plot /', where

/ = c (corn), p (peas), s (soybeans) and 7 = 1,2,3

maximize Z = 600(x,.| + .(r,.2 + .v<.3)

+ 450(.v,,| +.v,,2 + X/,3)

+ 3()()(v,| +.v,2+-Vv3)
14. (a) minimize Z= .80.ri + 3.70;c2 + 2.30.t3 + .90a4 16. (a) Minimize Z = 69x|i + 71x|2 + 71vi3 + 74x|4 +
+ .75x5 + -40x6 + .83.V7 76.r2i + 74x22 + 75.V23 + 79x->4 + 86x31 + 89x3-)
subject to + 8O.V33 + 82.V34
subject to
520.1-1 + 500^2 +
+ 6OOA4 + 86O.V3
+ 240.V7 > .500
5OJC5 + 460JC6
V|| + V|2+.V|3 +xi4<220
1

520.V| + 500x2 + 860x3 + 6OOX4 + -^2 1

+ -'^22 + ^2?. + -^24 ^ 70 1

50x5 + 460x6 + 24O.V7 < 2.000

-"3 1
+ ^^32 + ^33 + -V34 ^ 280
X|| +.V21 + V31 =110
4.4x| + 3.3x2 + -3^3 + 3.4.r4 +
X|2 +X22 +-t32 = 160
.5x5 + 2.2x6 + -2x7 > 5
X|3 +X23 +X33 = 90
30x| + 5.r2 + 75x3 + 3x4 + IOX7 > 20
X|4 + .V24+V34= 180
30x| + 5x2 + 75x3 + 3x4 + IOX7 < 60
7x| + 85x2 + 82x3 + 0x4 +
1 1

Ash: .03x, ,
- .OI.V21 - .02x3, <
6.V5+ 10x6+ 16.V7>30
.04x|2 + .av22-.OI.r32<0
30.x:4 + 70x6 + 22x7 ^ 40
.04x|3 + .Ox23-.OIx33<0
1 80x1 + 90x2 + 350x3 + 2OX7 < 30
.03x|4-.OI.r24-.02.V34<0
ri>0

Sulfur: .01x11 - 01x2 -02x31 <0 1

(b)x4 = .667
1
+ .01x22-.02.r32<0
.Olxi2
X6 = 0.304 -.01x|3 + •03.V23 - •O4.V33 <
X7= 1.500 .OX|4-.02.V24-.03X34<0
Z= $2,867

(c) The model becomes infeasible and cannot be (b)xii=34

solved. Limiting each food item to one-half Xi3= II
pound is too restrictive. In fact, X|4 = 45
experimentation with the model will show that X23 = 35
one food item in particular, dried beans, is X24 = 351

restrictive. All other food items can be limited X31 =76

except dried beans. X32= 160
X33 = 44
Z = S44.054
15. (a) Xij = number of units of products / (/ = 1 .2,3)

produced on machine 7 (j = 1 ,2,3,4)

maximize Z= S7.8.vii + 7. 8x12 + 8.ZV13 + 17. (a) x,j = space (ft-) rented in month / \orj months,
7.9x|4 + 6.7.V21 + 8.9x22 + 9.2.r23 + 6.3x24 + where / = 1.2....6. and / = 1.2 6
8.4x31 + 8. 1X32 + 9.0.V23 + 5.8x34
Minimize Z= 1 .70x| 1
+ 1 .40x|2 + 1 .20x|3 +
subject to
1.10x14+ 1.05x15+ 1.00x16+ + I.70.V2I

35x1 1 + 40.V21 + 38.1-31 < 9.000 1 .40X22 + 20X23 +11 0X24 + .05X25 +
I 1

41x|2 + 36x22 + 37x32 < 14.400 1 .70x3 1 + -40x32 + .20x33 +11 0x34 +
I 1

34xi3 + 32x2^ + 33x33 ^ 12.000 1.70x41 + 1.40x42+ i.20.t43+ 1-70x51 +

39x|4 + 43x24 + 40x34 < 15.000 1.40x52+ 1.70.V61

^11+^12 +^13 +-^14 = 400 subject to:

X2 + .V22 + ^23 + ^24 = 570
1

X3 + X32 + X33 + X34 = 320

1 + X12 + X|3 + X|4 + xi5 + X16 = 47.000
xi I

x„>0 •^1 2 + -^14 + -^'15 + -^'16 + -^'22 + -^2

^13 + ^

+ .V24+.V25 = 35.000
Xi3 + Xi4 + Xi5 + Xi6 + .V22 + -V23 + .V24
(b)xii = 15.385
+ .V25 + X3 + .V32 + X33 + .V34 = 52.000 1

xi4 = 384.615
X|4 + Xi5 + X16 + X23 + -^24 + X25 + .V32
.V22 = 400.00
+ X33 + X34 + X41 + X42 + X43 = 27.000
.1-23 = 70.00
1

Xi 5 + V 6 + X24 + X25 + X33 + X34 + .^42

X3, = 121.212
1

+ .V43 + .V51 +.V52= 19.0{X)

X33 = 198.788
-^16 + -^'25 + -^'34 + -^43 + + -^61 = 5.000 '^.'i2 '

Z = SI 1.089.73

51
(b)x|| = 12,000 (b)Change the 3 demand constraints in the
.r, 3 = 25,000 (a) formulation from < 1 .200 to = 1 ,000.

jci4 = 8,000
•fnc
jr|5 = 2.000
JC33 = 2,000
x^4= 15,000
Z = $80,200

(c).V|(s = 52.000
Z= $52,000
It is much cheaper to rent all the space for the
entire six month period in April and have
excess or "surplus" space.
 ,

22. (a) x,! = tlollar amount invested in alternative / in maximize Z = 2.2.5a| + 4,00a2 + 1 .5()v3
year 7. uhere /= p (product research and subject to
development), m (manufacturing operations
A + A2 +
I
> .V3 5,000
impro\ ements). a (advertising and sales
.01a| + .0.3a2 + <
.O2.V3 120
promotion) and / = 1.2.3.4 (denoting year):
.2a| + .5.V2 + .3.V3 < 3.000
.v^ = slack, or uninvested funds in yeary
Xi < 3.000
maximize Z = ,V4 + 1 .2.V(,4 + 1 .3.x,„3 + 1 .5x^3 .V2 < 2.000
subject to A3 < 6.000
x„| > 30.000 X[,X2. .V3 >
x,„i > 40.000
Xpl > 50.000
(b)A| = 3.000
^ti\ + "'ml +^p\ + -^1 = 500,000
.V2 = 2,()()()
.V3 = .500
1

.r„3 + .x-,,,3 + .r^3 = .92 + 1 .2.r„2 + 1 .3.v,„

Z = $17,000
-V„4 + .v^4 = .93 + 1 .2.v,,3 + 1 .3a„,2 +
\.5xpi
.x,j,sj>0 25. (a) A| = no. of television commercials
A2 = no. of newspaper ads
Note: Since it is assumed that any amount of
= no. of radio commercials
funds can be invested — in each alternative i.e.,
.V3

there is no minimum investment required — and minimize Z= 15.()00a| + 4.0()0.V2 + 6.000.^3

funds can always be invested in as short a subject to
period as one year yielding a positi\ e return, it
xji/x2>2/\
is apparent that the .v, variables for uninvested
25,0OO.V| + 10,000.V2 + i5.OOO.r3 > 100.000
funds will be driven to zero in every period.
I5,000.r| + 3.000.r2 + I2.0(K)a3/
Thus, these variables could be omitted from the
IO.OOOa, + 7.OOO.V2 + 3.()OOa3 > 2/1
model formulation for this problem.
15.()00a| + 4.OOO.V2 + 9.()()()A3/

25.000a| + 1().00().V2 + 15.0()().V3 > .30

(b).v„| =410,000 .v,„i =40.000 .V2 < 7

x„2 = 492.000 Xj,\ =50.000 A|.A2. .V3>0

.r„3 = 642,400 Z= $1,015,056
x„4 = 845,880
(b).V2 = 2.5
JC3= 5.0
23. (a) A I = no. of homeowner's policies Z = 40.000
V2 = no. of auto policies
XT, = no. of life policies
(c) This reformulation of the model would result
maximize Z= 35.vi + 20.^2 + 58.r3 in a fourth variable .V4, with model parameters

subject to inserted accordingly. However, it would have

no effect on the solution.
14v| + 1 2.V2 + 35.V3< 35.000
6a I
+3a2+ 12.V3< 20.000
A|. .V2. .V3 >0 26. (a) x,j = lbs. of coffee / used in blend / per week,
where ( =b (Brazilian), o (Mocha), c
(Colombian), m (mild) and 7 = s (special), d
(b)A| = 2.500
(dark), r (regular)
Z = S87.500
maximize Z = 4.5.\>,., + 3.75.Vos + 3.6().Vj.s
+ 4.8.Vm^ + 3.25Ah(i + 2.5.iod +
24. (a) A = no. of issues of Daily Life
I

2.35.r,.d + 3.55.Vnnl + •75.Vhr + I

A2 = no. of issues of Agriculture Today

1.00.Vo,+0.85.tcr+2.()5.Vn,r
x-i, = no. of issues of Surf's Up subject to

53
 •
4-^bs - -4X0^ - •4Xms
6-«cs - ^ 28. (a) Xjj= production in month to meet demand in /

+ .7xcs - .3.rcs - .3xms

-.3A-b.s > month J, where = ,2,... 7 and 7 = 4,5.6 and 7
/ 1

4xhd - .6xod - .6xcd - .6x,,,d > yj = overtime production in monthj where

-. Xbd - I'TocI - l"^cd + ^Xmd
1 • • ^ ;•= 4.5,6,7.
-.6xhr - •6-'^or - ^Xcr + •4Xmr <
Minimize Z = 50x|4 + 100.^24 + 50x^4 + 0.r44
1

.7xhr - .3Xor - .3Xcr - .3Xmr ^

+ 200x|5 + 150x25 + 100.^35 + 5().r45 + Oxss +
J^bs + -^bd + -^br ^ HO
50x56 + 250x|f, + 200.V26 + 50x35 + Oxm, + 1

Xos + ^od + -^or ^ 70

50x67 + 300x|7 + 250x27 + 2OOX37 + 0x77 +
^cs + -^cd + f c. ^ 80
400y4 + 40035 + 400y6 + 4()0y7
•^ms + -^md + -»^mr ^ 50•

Xy>0 Subject to

(b) Xos - 60 Special: Xqs + -V(-s + x^j = 200Ibs. X|4 + A-24 + .V34 + X44 + y4 = 60 + X45
Xj-s = 80 Dark: x^d + -'^md = 72 lbs. X| 5 + .^25 + X35 + .x-45 + X55 + y5 = 85 + A'56
-^ms = 60 Regular: x^r + Xor + Xm,- = 38 lbs. 1 "16 + -»'26 + f.Vi + J^.Sfi + -"^66 + .^•6 = 00 4- X67 I

Xhd = 64.8 •"'17 + -"'27 + •"37 + -"'67 + -"'77 + ."'7 - 20 '

-^md = 7-2 X|4 + X|5+X|c, +X|7<3()

Xh, = 45.2 .r24 + X25 + X26 + X27 < 30

x„,= I0 X34 + X35 + X36 + X37 < 30

= 82.8
x„i,. X44 < 40
Z= $1,296 .<:55 < 60

X66<90
27. (a) X| = $ amount borrowed for six months in July
X77 < 50
v, - $ amount borrowed in month (/ = / 2, ...,
1 ,

^•4 < 20
6) for one month
J.s ^ 20
c, = $ amount carried over from month / to / + 1
y6<20
6
."'7 ^ 20
minimize Z= . i lx| + .05 /^ v,
1=1

subject to (b)x|4=10 >-5 = 5

X34= 10 .V6=10
July: X| + y, + 20.000 - r, = 60,000
X44 = 40 .V7 = 20
August: r + y: + 30,000 - r^ = 60,000 + yi
I

X35 = 20 Z = $3 1.500
September: ct + y^ + 40,000 - c^ = 80,000 + yj
X55 = 60
October: (3 + y^ + 50,000 - 04 = 30,000 + yi,
X66 = 90
November ^4 + ys + 80,000 - c? = 30,000 + y4
X|7 = 20
December: c^ + y^ + 1 00,000 -C(,= 00,000 + y^
\

X27 = 30
End: x^ + yjs < c^
.V77 = 50
x,,y,,c-,>0

(b) Solution

X| =70.000
y^ = 40,000
y4 = 20,000
.^1 =.V2 =.V5 =.V6 =n
(•5 = 30,000
((,= 110,000
Z = $10,700

(c) Changing the six-month interest rate to 9%

results in the following new solution:

X| = 90,000
y^ = 20,000
c-| =50,000
(2 = 20,000
(5 = 50,000
Cf, = 130.000
Z= $9.1()0
 (c) .v:,4 = 20 The nev\ solution wduld be.
-V44 = 40
.r„ = 20
.V-.5 = 5

vis = 20
V=10
Var = 70
vss = 60
Abb = 75
V26=10
Xpb = 25
V66 = 90
=180
.tern
.v,7 = 40
= 225
JCbn,
X21 = 25
Vpm = I ' -5

.V77 = 50
^am = 330
y7 = 5 Z= SI. 477.50
Z= $26,000
30. (a) This is an asaianmeiit problem.
29. (a) xij= amount of ingredient / in wiener typey,
where / = c. b. p, a represent chicken, beet", A = operator
I I to drill press

pork, and additives, and 7 - r.b.m represent .vt = operator I to lathe

regular, beef, and all-meat, respectively .V3 = operator I to grinder

A4 = operator 2 to drill press

maximize Z= Jx^^ + .6.Vbr + + .S5x^
^.Vpr
A5 = operator 2 to lathe
+ 1.05.rcb + .95.Vbb + .75xpb A6 = operator 2 to grinder
+ .2avab + .55.Ven, + .45.Vbn,
1 1 1
.V7 = operator 3 to drill press
+ 1.25.Vpm+ 1.70.tan, Ag = operator 3 to lathe

subject to A9 = operator 3 to grinder

+ Jfcb + -Vcm ^
Xcr 200 minimize Z = 21x\ + S.vt + 35.V3 + 4 1.r4
I

+ -Vbb + -Vbm ^
-Vbr 300 + 30a5 + 28.V6 + 25.V7 + 36a-8
XpT + Xpb+'Cpm^ 150 + 8.V9 1

Xar + -^ab + ^am ^ 400 subject to

.90.tbr + .90.tpr - .lO^cr - lO.Var <

.V| + A3 =
+.V2
.SO.ter -.20.Vbr - .20Tpr - .20jrar >
.lf4+ A5 + .t6 =
.25.rbb + -75.Vcb - •75.rpb - .75.Vab >
-V7+.V8+.V9 =
*x*am =0
X\ +A4 + .V7 =
^pm S-Tam <
.Vt + -Vs + X^ =
A„ >
.V3+.V6 + A9=I
*Also feasible to delete A;,m from the problem. A| , A2. A3. .V4, .V5, .Vf,. .V7. As, A9 >

(b) A,n, = 200 (b)A| = I

-vbni = 300 AS = 1

Apn, = 150 A9 = 1

-Vn,m=400 Z = 70
Z=$1.6I2.50
(c) This would require the model to be
(c)This would require the addition of three reformulated with three new variables. A|o. A||.

constraints to the model. A|2, representing Kellys assignment to the

press, lathe, and grinder. The model would be

-Vcr+ -Vbr + Apr + Aar > lOO
reformulated as.
-fcb+ ^bb + -Vpb + -Vab ^ H)0
^cm + -^bm + -Vpm + Aam > 00 1 Minimize Z= 21v| + IS.vt + 35a3 +41.14 + 30.^5
+ 28a6 + 25.V7 + 36as + 8.^9 + 1

20a|o + 20a|| +20a,2

55
 subject to (b)X| = 30
A +.V2 +X3 < I
X3 = 1

A-4 + As + Af, < 1

.V4- 10

A7 + Ah + Xg < 1 xs = 40
X6 = 20
A| +A4 + A7+A|()= 1
.V7= 10

At +X5 + Ah + A| = I
1
.V9 - 40
XT. + A5 + Ac) + Ap = I X|o=IO
A,>() Z= 170

This new model would not change the solution.

Joe should not hire Kelly. 33. (a)A| = no. of hours molding
.V2 = no. of hours smoothing

31. (a) This is a transportation problem. A3 = no. of hours painting

A| = no. of tons of carpet shipped from maximize Z 175(7a|)= 1.225x|

St. Louis to Chicago subject to

XT = no. of tons of carpet shipped from

8x| + 5x2 + 6.5X3 ^ ^^^^^
St. Louis to Atlanta X| -l-.V2-t-X3< 120
A3 = no. of tons of carpet shipped from 90(7x|)< 10.000
Richmond to Chicago 7x| - 12x2= ")-^'3 7x, - 12x2 =
.V4 = no. of tons of carpet shipped from (The no. of units going 2a2 - 1 OV3 =
Richmond to Atlanta through each process 7a - 10.v^=()
I

minimize Z = 40xi + 65x2 + "70x3 + 30x4 must be equal. More X|..V2. .V3>()

bathtubs cannot be
subject to
smoothed than are first
= 250
X| +.^2 molded.)
.V3 + .V4 = 400

X| +.V3 = 300

X2 + X4 = 350 (b)x| =27.778

X|, XT, X3, X4 > X2= 16.204
X3= 19.444

(b).V2 = 250 Z= 34,027.778

X3 = 300
X4 = 400 34. (a) X|3 = tons shipped from 1 to 3
Z= 40,250 x\4- tons shipped from 1 to 4
X|2 = tons shipped from 1 to 2

32. (a) -v, = number of nurses that begin the H-hour X34 = tons shipped from 3 to 4

shift in period /, where / - 1,2.... 12 and period ;c4s = tons shipped from 4 to 5
= 2 1 AM-2 AM, period 2 = 2AM-4AM, etc. X2S = tons shipped from 2 to 5

minimize Z = X| + X2 + A3 + V4 + X5 + X(, + x-j minimize Z= 3xi3 + 4a|4 + 5Ap + 2.V34

+ .Vs+.V9 + A|()+X|| -t-A|2 + 7.V4S + 8a2s

subject to subject to

-^l()+ -^il + V|2 + X| >30 A|3 +A|4+Ap = 5

.<
1
+ V 2 + V + A2 ^ 20
1
1
I

X-X^ + =?
^-25

.V 2 + X
1
+ .V2 -I-X3 > 40I A|^ = A^4
A + V2 + V3 + .V4 > 50
I A- 4 + A-34 = A-4S
1

A2 + A3 + .V4 + .V5 > 60 X\l =V25

A-3 + V4 + V5 .V5 > 80 -(-
X\y- X|4. .Vp. .V34, .V4S. .\25 ^ ()

V4 + .vs + Vf, + A7 > 80

As + X(, + .V7 Ah > 70 -t-

>70 (b)x,4 = 5
.lYi + x-j + xh + -Vc)
.V4S = 5
.V7 +.Vx +.Vc) + A- 10 > 60
Z = $55,000
.VH+.Vy+Aio + Aii >50
.

Ai) + A|() + A|l + Ap > 50

x,>0

56
35. maximize Z ,n85.V| + .IOOa-2 + .065JC3
+ . 1 3O.V4
subject to

x^< 14,000
Xl <X\ +X3 +X4
X2 + ;t3 > 2 1 ,000
-I.2.V1 +.V2+.V3— 1.2x4 >0
.V| + .V2 + .V3 + .V4 < 70,000

.V|, .vi. A3, .V4 >

.V, = 17,818.182

.V2 = 35,000

.^3 = 3,181.818

X4= 14.000
Z= $7,041.36

36. (a) .v,/ - barrels of component / used in gasoline

graded per day, where = 2, 3, 4 and / = R / 1 ,

(regular), P (premium), L (low lead)

Regular: A|r + xi^ + ^3^ + A4R

Premium: X|p + xy? + x-t,^ + A4P
Low lead: X| l + X2\^ + .V3L + a^l
maximize Z= 3.V|r + 5x2p + OX3R + 6x3p
+ 2x4p + Ix] L + 3x7L + 4x4L
1

-2-r3L
subject to

•^1 + ^m + -^'.^R + -^^R ^ 3,000

R
+ X2P + X3P + X4P > 3,000
X| p

X| 2 + X22 + Xyi + X42 > 3,000

X|R+X|p + xiL< 5,000

X2R + .V2p + X2L ^ 2,400
Xy^ + X3P + X3L < 4,000
X4R + X4P + X4L < .500 1

.6x1 R - .4.V2R - .4X3R - .4X4R >

-.2x|R + .8x2R — 2x3R — .2x4R <

-.3x|R - 3x2R + .7x3R - .3x4R >
.

-.4x| p - .4.X2P + .6x3p - .4.V4P >

-.5X| L + .5.V2L - .5x3L - .5X4L <

.9X|L - 1.Y2L - 1-X3L - •'^4L ^

. •

all Xi, >

(b)x|R = 2,000
X2R= too
X3R = 900
X2P = 2,300
X3p = 3,100
.V4P = 1 ,500
X|L = 3,000
Z = $71,400
38. (a)

Year
1
 subject to subject to

jc,< 2.000 (/= 1,2.3.4,5) < I

.v,<600(/= 1,2.3.4,5) + '^H> + ^2c <

^2a I

+v, -»v, = 1.200

.V|
^3a + ^?ih + ^?>< < I

At + y2 + H| - Ht = 2. 100 X4,, + X4/, + X4,. < I

.V3 + y3 + ht - H'3 = 2,400

^\a+ ^2(1 + ^3(1+^4(1 = I

.V4 + y4 + VV3 - H'4 = 3.000 X\b+X2h + XT,i,+X4i, = 1

,r5 + 3'5 + VV4 = 4.000 X|,. + X2, +X3,. +X4,

Xj, j„ Wj < X, >

X2(, = I

x^h = 1

(b).V| = 2.000 yi =300 u-| =1.100

X4, = I

A: = 2.000 \2 = 600 U2= 1,600

=
Z=660
A-3 = 2.000 yj 600 vv3= 1,800
^4 = 2.000 y4 = 600 H4= 1.400
xs = 2.000 y^ = 600 42. (a)Xy = amount shipped from city i to cil) /

Z= $152,300 maximize Z = x^r + x\q or maximize

Z = xep + Acp
subject to
41. (a)v,Y = employee / assigned to department y.
where / = 1.2.3,4 and 7 = a (lamps), b -Vab ^ 7
(sporting goods), c (linen) '<^AC^6

-VBD ^ 5
maxiiTiize Z= 130x|,, + 150a|^ + 90ai^ +
XBC<IO
275a:2„ + 300jt2fc + 100^2, +
J^CD ^ 3
18OA3,, + 225x3/, + 140x3, +

200x4^+ 120x4/,+ 160x4,.

XcF<9
subject to
-^DE^4
Xef^3
'^AB = -^BC + -f BD + -^BE
^2a + ^Ib + ^2c = '
-^AC +'^BC = "fCD
-^3^ + ^3b + X3c = I -'fBD+-^CD=-"fDE
X4,, + X4/, + X4,^. = 1
-'^BE + -^'DE = '^EF
i<(x,<, +X2<, +X3„ + X4,,)<2 X,y>0
I < (X|/, + X2/, + X3/, + X4/,) < 2
I < (X| ,. + X2t. + X3,. + X4,.) < 2
(b)xAB = 6
x,j>0
Xac = 6
X\u = I

•^BE = 2
X2b= I

-^BD= '

•^3/7 = I

^BC = 3
-V4< = 1

^CF = 9
Z=815
JfDE= 1

^EF = 3
(b)x,y = employee / assigned to department;, Z= 12

where i = 1 .2,3,4, andy = a (lamps), b

(sporting goods), c (linen)

maximize Z= 130x|(, + 150x|/, + 90xi, +

275x2« + 300x2/, + lOO-^lc- +
+
180x3„ 225x3/, + 140x3,. +
200x4„ + 1 20x4/, + 1 60x4,.

59
 1 )

43. (a) Minimize Z= .v subject to

subject to y = Ai (Component 1

= 650 + .V|
15n.v y = A2 (Component 2)
I5().v + y| =450 + V2 y = XT, (Component 3)
50x + V2 = 600 + y3
1

Note: The above constraints reflect the fact that

15O.r + .V3 = 500 + V4
the quantity of completely assembled product
50.V + y4 = 700 + 3'5
1

cannot exceed the quantities of components 1,

50a- + y5 ^ 650 + }(,
I

2, and 3 produced per day. i.e.. no in -process

50jr + Vf, = 750 + y-j
1

inventory.
5().v + y-j = 900 + ys
1

5()a- + ys = 800 + yt)

I The time available on one lathe is: 60 min/hr.
I50.v + yc, = 650+yio X 8 hns/day = 480 min. In order to reflect the

150A- + yi() = 700+yi| availability of two lathes (with the work load
15()A + yii >500 shared evenly between them) the total minutes
of lathe time available is: 480 min/day/machine
X 2 machines = 960 min.
(b)Z = A = 4.4545 lawyers
y, = 18.1818 Thus, the lathe use/a\ailability constraint is:

y2 = 236.3636
1()a| + 8.V2 + (rxT, < 960 (Lathe Time)
y3 = 304.5454
y4 = 472.7273 By the same rationale, the press u.se/

ys = 440.909 availability constraint (for three presses) is:

y6 = 459.09 9a| + 2 l.r2 + 5a3 < 1440

I (Press Time)
y-j = 377.2727
Assuming that the work loads are shared
y8= 145.4545
equally among the lathes, and similarly for
vc, = 13.6364
pres.ses. the individual machine utilizations, are:
V|o = 3 1.8182
( IOv| + 8.V2 + 6a- 0/2 = 5 a- + 4v2 + 3a-3
.vii=() I

(9a + 2 1.V2 + 5x^)B = 3a + 7.V2 +

I
1
|
5.V3

44. The original optimal solution has a = 4.4545 The balance condition is reflected by
lawyers being hired. To get an "optimal" specifying that the absolute difference between
integer solution you would have to round individual machine time consumed (on lathes

either down to a = 4 or up to a = 5. Adding a versus presses) n-iust be less than or equal to 60

new constraint, a= 4, to the original model minutes.
formulation results in an infeasible solution so
I(5a| + 4.v^ + 3A:t) - (3a + 7 XT + 5aOI < 60
rounding down is not possible. Adding the I

constraint, a = 5, instead (i.e., rounding up)

results in a feasible solution so (by logical
12a - 3.V2 - 2.V3I < 60
must be optimal. The original
I

deduction) it

solution had 168.1818 surplus hours while the Absolute value constraint conditions can be
new integer solution has 1 , 1 50 surplus hours, a reflected by:

large difference in excess capacity. However,

2a - 3.V2 - 2.V, < 60
number of lawyers
I

increasing the (to 6, for

-2a + 3.V2 + 2.V3 < 60
example) only increases the surplus so a = 5
I

has to be the optimal integer solution. The model is summarizeil as follows:

maximize Z - y
45. (a)y = quantity of assembled product produced subject to
per day
y-A| =0
.V| = quantity of component I produced per day
y-A2 = ()
.vt = cjuantity of component 2 produced per day
V - x-^ =
XT, = quantity of component 3 produced per day
1()a| + 8.V2 + 6.V3 < 960
maximize Z= y 9a +2I.V2+ 15.v:i< 1440
I

2a - 3.V2 - 2.V3 < 60

I

-2.V| + 3.V2 + IxT. < 60

.Vl. XT, XT,. \ >

60
 (b)A| =A2=.V3 = \=2n (b).V| =.S.45H.128
V2 = 2.735.63
1

yi = 152.709
machine balancing requirement is
(c) If the
y2= 1.747.126
relaxed such that the constraints 2.V| - 3.V2 - 2.^:3

< 60 and -2x\ + 3x1 + 2x3 < 60 are eliminated.

Z= $14,201.64
production output would increase:
48. (a)v, = no. of employees assigned to time period /,
.V| = X2 = -V3 = y = 32
where /= 1.2 6 (time period = 12:(X) 1

Alternatively, if the restriction that there be no midnight -4:00 A.M.:

excess component parts at the end of the day is period 2 = 4:()0-X:00 A.M.: etc.)
relaxed such that y< jt] y < .vt and . \ < .V3.
minimize Z= A| + xi + .V3 + A-4 + .vs + Xf,
production output would increase:
subject to
.V| =43.077
.v-,=.V3=y = 29.31

The best alternative is to remove the machine

balancing requirement. If both requirements are
relaxed the production output is a |
= X2 = .V3 =
y = 32.

46. (a)v,y - amount of commodity / put into hold/

where / = b (beef), g (grain) and / - f (fore).

a (aft)

maximize Z= .35(,v/,^ + X[,i,) + . 1

2(x^f+ x^.^,)

subject to

^hf+^Kt
50. (a).v = full-time operators Solution:
\-, = part-time operators hired in week /, where X| = 14 =53.556
yi
/= 1,2,. ..8 X2 =6 y2 = 51.111 Z= $361,456
8

Minimize Z = $4,880jc + $450 ^ >',

X3= 18.4 y3 = 43.615
/•= I X4 = 6.4 y4 = 41.(X)7
subject to X5 = 24.8 y, = 30.904
360A: + 270.y| = 19,500 y6= 17.57
300x -I- 270y2 = 21,000 .V7 = 7.2

360x + 270y3 = 25,600 y8 = 22.()15

360x + 270v4 = 27,200

360a- + 270v5 = 33,400
360x + 270v6 = 29,800
CASE SOLUTION:
360x + 270v7 = 27,000
SUMMER SPORTS CAMP AT STATE
360x + 270ys = 31,000 UNIVERSITY
l.lx-i-2.7y,< 160,/= 1,2 8 Model Formulation:
.V, y, >
Wj = new sheets purchased in week /

Note: Objective function (/= 1,2.. ..8)

coefficient of $4,880 for x is
X, = sheets cleaned at laundry at end of week /

computed as ($610/week)(8 = sheets cleaned by Mary's friends at end of

y,
weeks). week /

(b)Soiution
minimize Z= 10(vi'| + vi'2 + w-^ + H'4 -1- n^ -1- ny,+
= 69.6 full-time operators
w-j + Wft,) + 4(.V| + XT -I- -i"3 + xji + X5 +
v

=0 part-time X6)+20'| -I- \2 + y3 + y4 + y.s)

yi operators
subject to
y2 = part-time operators
yt, = 2.0148 part-time operators
»'l

y4 = 7.9407 part-time operators

y5 = 30.9037 part-time operators
)'(, = 17.5704 part-time operators
yy = 7.2000 part-time operators
yx = 22.0148 part-time operators

Z = $379,088.00

51. Minimize Z = 4880xi 4270.V2 + 3660^3 + -i-

3050x4 + 2440x5 + 1^30x6 1220x7 + 610x8 + -t-

450 2. V,

subject to

360x|+270yi = 19,500
+ 270y2 = 21.000 360(X|+.r2)
360(xi + X2 + X3) 270y3 = 25,600 -I-

360(X| + X2 + .r^ + .^4) + 270y4 = 27,200

360(x| + X2 .V3 + X4 + x^) + 270y5 = 33,400
-1-

36()(.vi + X2 + .V3 .V4 .V5 .V5) + 270y6 = 29,800 -I- -I- -I-

36()(.V| + X2 + X3 + X4 + X5 + X(, + X7) + 270y7 = 27,000

360(X| + X2 + X3 + X4 + X5 + X6 + X7 + xr) + 270yH = 3 ,000 1

l.lx| +2.7y| < 160

1.1 (X| +X2) + 2.7y2< 160

1,1 (xi +.V2-H.V3) + 2.7y3< 160
1 . 1
(.vi + X2 + X3 + X4) + 2.7\'4 < 160
1 . 1 {.V I -t- + -V3
.X2 + X4 + X5) + 2.7y'5 < 1 60
1 . 1(.V| + X2 + X3 + X4 + X5 + Xft) + 2.7y(, < 1 60
1 . 1 (.V| + .V2 + X3 + X4 + .V5 + X(, + .V7) + 2.7y7 < 1 60
I . I (.V| + .V2 + .V3 + .V4 + +.\'5 .V(s + .V7 + vx) + 2.7yx < 1 60
v„'y,>()
CASE SOLUTION: CASE SOLUTION:
SPRING GARDEN TOOLS SUSAN WONG'S PERSONAL
Model Formulation:
BUDGETING MODEL
Let / = (trowel). 2 (hoe), 3 (rake). 4 (shovel)
1
(b)Let Xjj = amount invested for / months in month
Rj = regular production of product / in stage 1
j where /= 1.3 and 7. andy =1.2 12.
12 12
Sj = subcontracted production of product / in 12

stage 1
Max Z = 2 0.005x|^ + 2 0.02x3^ + 2 O-Olxy
7=1 j=\ ;=i
Xj = overtime production of product / in
subject to
stage 1

A, = regular production of product / in Jan: -x,- X3 - X3 + 2.450 +

, , , 3.800 = 2.750
stage 2 Feb: - X|2 - X32 - X72 +
X| 1 2.450 = 2.860
V, = overtime production of product / in Mar: ^i 2 - X| 3 - X33 - X73 + 2.450 = 2.335
stage 2 Apr: X13-X14 + .V31 - X34 - X74 + 2.450 = 2, 1 20
May: ^u-xis + X32-.V35-X75 + 2.450 = 2.205
minimize Z=S6/?| + + 8/?3+ 10/?4 + 7.25i
10/?2
June: + X33-X36-X76 + 2.450 =
X|5-X|(j 1.600
+ 12^2 + 9.653 + 1254 + 6.2;c| +
July: - X|7 + X34 - X37 - X77 + 2.450 =
xi(, 3.050
10.7^2 +8.5x3 + 10.7JC4 + 3/4, + 5^2
Aug: X] 7 - X| 8 + X35 - X38 - X7 - X7S + 2.450 = 1 2.300
+ 4>43 + 5/44 + 3.1j| + 5.4j2 + 4.3y3
Sep: X| g - X| 9 + X36 - X39 - X72 - X79 + 2.450 = 1 .975
+ 5.4y4
Oct: X19-X110 + X37-X310-X73-X710 + 2.450= 1.670
subject to
Nov: xiio-xiii +X38-X311-X74-X711 +2.450 = 2.710
.04/? I
+ 2/?4 < 500 hrs.
+ . 1 7/?2 + .06/?3 . 1 Dec: X|ii -X112 + X39-X312-X75-X712 + 2.450 = 2.980
.04x| + .17x2 + 06x3 + ISau < 100 hrs. •

.05/? + 4/?2 + 4/?4 < 400 hrs. I

. 1 . 1

Investments (/= 1,3,7)

.05x1 + 14x2 + 14x4 < 100 hrs. .

1.2/?| + l.6/?2 + 2.1/?3 + 2.4/?4 Month (j) 1-month 3-month 7-month

+ .2x, + .6.V2 + 2. 1x3 + 2.4x4 < 10.000 ft2
1 1
1 . January ,V||=4I0 .V3, = 3.090
.06/4 + 3/42 + 05/43 + 10/44 < 600 hrs.
1 . 1 .

2. February
.()6yi + 3.V2 + .05y3 + 10y4 < 100 hrs. . 1 .

3. March xi3= 115

.05/4 + .2 1/42 + 02/43 +
1 0/44 < 550 hrs. . 1

< 100 hrs. 4. April x,4= 3.535

.05yi + .21y2 + .02y3 + 10y4 .

.03/4 + 5/42 + 04/43 +

1 5/44 < 500 hrs.
. 1 1
5. May X75 = 780

.03yi + 15y2 + .04 V3 + 15y4

< 100 hrs. . . 6. June X|6 = 600 X36 = 250
/?! +5| +X| =/4| +yi 7. July
/?2 + 52 +X2 =A2+\2 8. August x,8=l50
/?3 + 53 +X3 =/43 +y3
9. September .V39 = 875
R4 + S4+ X4 =Aj[+ \4
10. October V, ,0=780
y, +/4, = 1.800
y2+/42= 1.400 ll.Nobember X|||=520
y3+/43= 1.600 12. December X| 12 = 5.645
+ A4 = .800
y4 1

/?,. 5„ x„ A,, y, > Z= S844.60 earned in interest pa\ments

Solution:
(b)Using sensitivity analysis for the January
/?, = 1.691.954 S\ = .^2 = ^3 = constraint, the lower range for the right hand
/?2= 1.319.54 s^ = 866.6667 side is —410. Thus. Susan needs S7I0 out of her
Rj = 1 .600 yi=o Z= 585,472.60 original S3. 800 to make the model feasible

/?4 = 933.33 v^ = 315.1514 (i.e., avoid an infeasible solution).

/4 , = .800 1 yj =
A2= 1.084.849 y4 = 388. 1822
A^= 1.600 X| = 108.0457
/A4= 1.461.818 vo = 80.4599
X3=X4 =

63
 Chapter Five

ANSWERS TO QUESTIONS 10. a) Transform the model constraints into

equations.
1. The slack variables are added to < constraints b) Set up the initial tableau for the basic
to take on any value necessary to make the left- feasible solution at the origin.
hand side of the constraint equal the right-hand c) Determine the pivot column and entering
side, thus creating equations.
nonbasic variable.

2. A feasible solution is any solution that satisfies

d) Determine the pivot row and leaving basic
the constraints, while a basic feasible solution variable.

satisfies the constraints and contains as many e) Compute new pivot row values.
variables with nonnegative values as there are f) Compute all other row values using the

model constraints. pivot formula.

g) Compute new Zj andCj - Zj row values.

3. They represent unused resources or activity and h) See if any Cj - row values are positive.
Zj If
and thus contribute nothing to the achievement yes, go to step c; if no, the solution is
of the objective function. optimal.

4. The simplex method starts at the origin where 11. There are resources or some level of constraint
all decision variables equal zero and all
activity left unused in the optimal solution.
resources or constraint activity are unused.
Thus, only slack variables have nonzero values. 12. A basic variable is a variable in the current
basic solution of the simplex tableau, while a
5. The net increase per unit of entering a nonbasic
nonbasic variable is not.
variable into the basic solution.
13. When the constraint is either > or =; in both
6. The entering variable has the largest positive situations an initial simplex solution at the
Cj - Zj value, indicating that it will result in a
origin is not possible. In the case of a >
greater achievement of the objective function
constraint the addition of a surplus variable,
than the other variables. The leaving variable while resulting in a constraint equation, will
has the smallest nonnegative quotient of the have a negative value at the origin. In the case
quantity values divided by the pivot column of an = constraint, since neither a slack nor a
values. This indicates that this variable is the surplus is added, there is no basic solution
most constrained variable and hence feasible. variable associated with the constraint. In both

7. There will be three basic variables and five cases the artificial variable enables an initial

nonbasic variables (including slacks). basic solution to be established in the initial

simplex tableau.
8. Each tableau in the simplex method reflects a
corner point on the graph. Selecting the pivot 14. The simplex method computes a Cj - Zj row for

column (and the entering variable) in the initial a maximization problem and Zj - Cj for a

tableau indicates which axis to go out on first.

minimization problem. In a maximization
The pivot row (or leaving variable) determines problem the objective function coefficient for

the exact point on the axiswe go to. an artificial variable is a negative w, whereas it

Subsequent tableaus represent the movement is a positive m for a minimization problem.

from one corner point to another adjacent 15. given a large positive coefficient
It is in the
corner point on the boundary of the solution
objective function of a minimization problem
space until the optimal solution is reached. and a large negative objective function
The when - coefficient for a maximization problem. Both
9. solution is optimal all Cj row Zj

values are zero or negative. actions will ensure that an artificial variable
This means that no
variable remains which, if entered into the
will never be selected as the entering nonbasic

solution, would increase variable.

the value of the
objective function.

64
16. In the optimal solution tableau, a Cj - Zj (or 27. It enables a decision maker to see what the
Zj - Cj) row value of zero for a nonbasic effects of changes in model parameters (which
variable. may be uncertain) on the optimal solution
might be. It also indicates within what ranges
17. The nonbasic variable in the optimal tableau
of parameter values the dual values for
with a Cj - Zj (or Zj - cj) row value of zero is
resources are valid.
selected as the entering nonbasic variable and
the next tableau is constructed, which 28. The sensitivity range for a Cj value indicates
constitutes the alternate optimal solution. These the range over which the basic solution
two alternate solutions are actually only the variables and their values will remain the same.
corner points at each end of a constraint line.

All points along this line are alternative

29. A change in <7, within its sensitivity range will
maintain the optimal basic variables but the q,
solutions.
values and Zj value will change. If the change in

18. An infeasible problem has no feasible solution qi is outside the sensitivity range, the basic
area that satisfies all constraints value will no longer be feasible and a new basic
simultaneously. It is usually a result of an error feasible solution will result.
in model formulation. It is identified in the
simplex method by one or more artificial

variables being basic in the optimal simplex

PROBLEM SUMMARY
tableau. 1. Simplex short answer

19. It is a problem in which the feasible solution 2. Simplex discussion short answer
area is not closed, thus allowing the objective
3. Simplex short answer
function to increase indefinitely. It is identified
in the simplex method by the inability to select 4. Simplex short answer
a pivot row.
5. Simplex short answer
20. It is a tie for the pivot row which results in
6. Simplex short answer
zero quantity value for a basic variable in the

next tableau. 7. 4 tableaus

21. Ties for the pivot column and row are broken 8. 2 tableaus
arbitrarily.
9. 3 tableaus
22. <. add a slack variable: =. add an artificial
10. 3 tableaus
variable; >. subtract a surplus variable and add
an artificial variable. 11. 2 tableaus

23. ni variables and n constraints 12. 5 tableaus

24. The marginal value is the amount the decision 13. 5 tableaus
maker would be willing to pay for one
14. 5 tableaus
additional unit of the resource. In the simplex
tableau the marginal values of the resources are 15. 6 tableaus
the Cj - Zj row values.
16. 4 tableaus
25. It enables the decision maker to know (within
17. 3 tableaus
the limitations of the existing solution) the
maximum amount that should be paid for 18. 3 tableaus
additional resources. It also provides the decision
19. 3 tableaus
maker w ith some knowledge of the productivity
of resources — i.e., how much value they create. 20. Simplex short answer

26. No; it means that the slack variable is a basic 21. 3 tableaus

variable, with a positive value indicating that

22. 4 tableaus
there are resources (represented by the slack
variable) that are left over and not used. Since 23. graphical analysis
all the resources are not being used, additional
24. 2 tableaus
units are not needed. Thus, the Cj - Zj value

means that additional resources would not be 25. 2 tableaus

65
 26. 6 tableaus
PROBLEM SOLUTIONS
27. 5 tableaus
1. a) -fi = 10,^2 = 40, 53 = 30, z = 420
28. graphical analysis b) Yes; all Cj - Zj row values are zero
or
29. Mixed negative.
constraint model transformation
c) = 0; ^2 =
JC3
30. Mixed constraint model transformation d) Maximize Z= 10x^+2x2 + ^x^
31. 5 tableaus e) 3 constraints
f) Since there are three decision variables,
a
32. 3 tableaus
three-dimensional graph is required.
33. 3 tableaus

34. 4 tableaus 2. a) minimization; because -

zj cj is being
calculated on the bottom row and not
35. 3 tableaus, multiple optimal
Cj-Zj
36. 4 tableaus b) xi =
= 10, 5] = 10, z = 240
20, X2
- Cj values are negative or
c) Yes; all Zj
zero.
37. 3 tableaus, multiple optimal
d) Minimize Z= 6jci + 20^2 +
12^3
38. 2 tableaus, infeasible e) 3 constraints
f) Yes, one; because there are 3 constraints
39. 2 tableaus, unbounded but only 2 surplus variables. Since both <
40. 4 tableaus, pivot row and column and > constraints have slack or surplus
ties, multiple
optimal variables, and an equality does not, then one
of the three constraints was an =.
41. Infeasible problem g)X2 =
42. Dual formation and interpretation, sensitivity
analysis
3. a) maximization; because of
"c.-z,"
43. Dual formation and interpretation, b);ci = n,X2 = 0,X2 = {),x^= 15
sensitivity
analysis c) .^1 = 20, S2 = 0, J3 = 0, .y4 = 45
d) Ux2 is selected as the entering basic
44. Dual formation and interpretation, sensitivity variable,Z will increase by 20 for every unit
analysis
of j:2 entered into the solution.
45. Dual formation and interpretation, sensitivity e) This solution is not optimal
because there
analysis are positive values in the Cj - row.
zi

46. Dual formation and interpretation, sensitivity

analysis

47. Dual formation and interpretation, sensitivity

analysis

48. Dual formation and interpretation, sensitivity

analysis

49. Dual formation and interpretation, sensitivity

analysis

50. Dual formation and interpretation, sensitivity

analysis

51. Sensitivity analysis, Cj and

qj

52. Sensitivity analysis with duality

53. Sensitivity analysis with duality

54. Sensitivity analysis with duality

66
Cj
a
.05 .03 M
9
11. Maximize Z = .V] + S.yt (profit. S)
subject to
9
13.
0-
9
M
16.
17.

9
100 20 60
22.

9
23.

oV^iyV \12 16 20
First \ Fourth tableau
tableau Second tableau

24.

9
25.

9
 100 75 90 95
Basic
27.

9
30. Minimize Z= 40.ri + SS.vt + 3O.V3 + O^j + 0^2 +
0^3 + MAi + MAj + MAt,
subject to

x\ +2jr2 + 3jt3 + 5] = 60
2x\ + X2 + x-%, + A\ = AO
-Vi + 3.V2 + .V3 - 52 + ^2 = 50
5.V2 -3.V3 - 53 + /\3 = 100
.ri,.V2. .V3 >0

31.

-' Basic
40 60
9
34.

10 5 -Af -M
Variables
35.

9
36.
37,

J Basic 400 350 450

Variables Quantity xi X7 ^3 ^1 ^2 S3 S4

^1 120 2 3

52 160 4 3

J3 100 3 2 1

54 40 1 1 1

Cj-Zj 400 350 450

J Basic
400 350 450

Variables Quantity xj X2 x^ si ^2 *3 *4

51 70 1/2 2 -1/2

52 135 13/4 5/2 1 -1/4

450 xj, 25 3/4 1/2 1/4

54 15 -1/4 (1/2 -1/4

11.250 1,350/4 450/2 450 450/4

250/4 250/2 -450/4

400 350 450

J Basic
Variables Quantity Xi Xi Xj, H 52 53 54

51 10 -1/2 1/2 -4

52 60 2 1 1 -5

450 .V3 10 1/2 1 1/2 -1

350 x^ 30 1/2 1 -1/2 2

Zj 15.000

51 20 1

52 20 0-4
400 .V, 20 1 2

350 -ri 20 1 -1

15.000 400 350 450

38. (a)

Infeasible-oo common solution space

J I I I L

(b)
 (b)

9
^i Basic
41.

9
 c) C], basic:

'^j
Basic
 e) The marginal value of 1 hr of labor is $2.33. c)
The sensitivity range for qi, labor, is 45 < 72 -
1 80. Thus, the company would purchase up to

180 hr at the marginal value price.

43. a) Minimize Zj 500) , + 800>2

subject to

10>, +34v2>200
5(hi + 20>2 > 300

vj = $4.13 = the marginal value of one

additional lb of chili beans; V2 - $4.67 = the 10 20 C 30 40 50 60
marginal value of one additional lb of
ground beef. The constraint line for chili beans rotates,
creating a new, smaller solution space, and the
optimal solution shifts from point B to point B'
where X] = 21.43 and X2 = 3.57.

20 c 30

Point C must become the optimal solution for

X2 — 0; therefore the slope of the objective
function must be greater than the slope of the
constraint for ground beef, -34/20. Solving the
following for the profit, p, of Razorback chili

yields

-p/300 = -34/20
/? = $510

Thus, if the profit of Razorback chili is greater

than $510, no Longhom chili will be produced.

The new optimal solution will be X] = 23.5 and
X2 = 0.

Ml
 d) c\. basic:

^i Basic
 f) The marginal value of 1 lb of chili beans is
$4.13. The sensitivity range for q^. chili beans,
is 235.3 <q\< 2,000. Thus, the company
would purchase up to 2,000 lb at the marginal
value price.

g). Groundbeef

h) No effect

44. a) Minimize Zj = 60vi + 40v2

subject to

12vi+4y2>9
4>l + 8y2 ^ 7
yi..V2^0

b) yi = the marginal value of one additional hr of

process 1 : vt = the marginal value of one
additional hr of process 2

For the ^i column, the Cj - Zj value of S.55 is

the marginal value of 1 hr of process 1

production time. For the 51 column, the Cj - Zj

value of $.80 is the marginal value of 1 hr of

process 2 production time.

c) Cj. basic:

^J Basic
 A

Solving for the Cj - Zj inequalities: Since C2 =7+ A; A = C2 - 7. Thus

-1 1/20- A/10 <0 C2 - 7 > -4

-A/10< 11/20 C2>3
-A < 1 1/2

A> -1 1/2
Summarizino, 3 < o< 18.

Since c\ = 9 + A, A = C| - 9. Thus
d)9i:
c, -9>-ll/2
jc,. 4+ A/10>0 X2: 3-A/20>0
f, >7/2
A/10>^ -A/20 > -3
-12/20 + A/20 <0
A>-40 -A>-60
A/20< 12/20
A<60
A< 12
Therefore, -40 < A <60. Since
Since C| = 9 + A, A = t] - 9. Thus
^1 =60 +
C|-9< 12
A = ^i-60
c, <21
-40<9, -60<60
Summarizing, 7/2 < C| <2 1 .
20<gi< 120
c->, basic:

'^j
Basic
 e) The marginal value of 1 hr of process 1 Point A must become the optimal solution for
production time The sensitivity range
is $.55. .r| = 0; therefore the slope of the objective
for ^1, production hours, is 20 < ^| < 120. function must be less than the slope of the
Thus, the company would purchase up to 120 constraint for labor, -2/5. Solving the following
hr at the marginal value price. for the profit, p, of coffee tables gives

-20(yp = -2/5
45. a) Minimize Z<j= 180>i +135y2 /J = $500
subject to
Thus, if the profit for coffee tables is greater

2\i + 3.V2 > 200 than $500, no end tables will be produced. The
5yi + 3y2 ^ 300 new optimal solution will be .t] = 36 and x-y =
0.

b) vi =$33.33 = the marginal value of an

additional hr of labor, vt =$44.44 = the d)
marginal value of an additional lb of wood

The constraint line for wood moves outward,

creating a new solution space, and the optimal
10 20 30 40 50 60 70 80 90 1CX)
solution point shifts from pwint B to point B'
where .rj = 3 1 .67 and X2 = 23.33.

MS
 e) C], basic:

-' Basic
 A

g) The marginal value of 1 lb of wood is $44.44.

The sensitivity range for qi. wood, is 108 < qi
< 270. Thus, the company would purchase up
to 270 lb of wood at the marginal value price.

g) The marginal value of labor is S33.33 and the

marginal value of wood is $44.44; thus, wood
should be purchased.

46. a) Minimize Zj = 19vi +14y2+20j3

subject to

2yi +y2 + y3 -70

vi + \2 + 2x2 > 80
yi, y2, Xt, >

b) yi =$20 = the marginal value of an additional

hr of production time; vt =$0 - the marginal
value of an additional lb of steel; y^ = $30 = the
marginal value of an additional ft of wire

c) C]. basic:

70 + 80
Basic
Variables
B
47. a) Minimize Zj = 64v| +50v2 +120y3 +7v4 +7y5
subject to

^y\ + 5y2+ 15^3 +_V4 >9

8yi + 5.V2 + 8v3 + ys > 12
V], y2. .V3. y4. y5 >

b) y 1
=S.75 = the marginal value of one additional
hr of labor for process 1 ; vt =$1 .20 = the
marginal value of one additional hr of labor for
process 2: 33, y4, yj = SO; these resources have
no value since there were units available which
were not used.

c) c\. basic:

Solving for the Cj - Zj inequalities:

-3/4 + A/4 < -6/5 - 2A/5 <

ZV4 < 3/4 -2A/5 < 6/5
A< 3 A > -3
Since c\ -9 + A: A= c\ -9. Thus

-3 < A< 3
-3 < -9 <
ci 3
6<c, < 12
C2, basic:

Solving for the Cj - Zj inequalities:

-3/4 + A/4 < -6/5 + A/5 <

-A/4 < 3/4 A/5 < 6/5
A > -3 A< 6

Since 01= ^2 + A: A = c\ - 12. Thus

-3<C2- 12<6
9<o< 18

d)qi

A,: 4-A/4>0 J5: l-A/4>0 53: 12 + 7A/4>0

-A/4>-l -A/4>-l 7A/4>-12
A<16 A<4 A > -6.86

54: 3 + A/4 > .V2: 6 + A/4 >

A/4 > -3 A/4 > -6
A>-12 A>-24

109
 Summarizing,

-24<-12<-6.86<A<4< 16

and, therefore,

-6.86 <A<4
Since q\ = 64 + A; A= q\ - 64. Therefore,

-6.86<A9, -64<4
57.\4<qi <68
e) q3-

xi: 4 + 0A>0 55: 1+0A>0 .^3: 12 + A>0

0A>^ A<oo A>-12
A<oo
54. 3 + 0A>0 X2: 6 + 0A>0
A < 00 A < 00
-12<A<oo
Since ^3 = 120 + A; A= ^3 - 120. Therefore,

-12<A<oo
-12<93- 120 < 00
108 < ^3 < 00

Since 100 pounds is less than the lower limit

of the range, the optimal solution mix will
change. ^2 enters the solution and .^3 leaves.
The new solution is,

xi = 3.27
S2= 1.82
55 = 0.64
S4 = 3.73
X2 = 6.36
Z= 105.82

48. a) Minimize Zj = 209v, +I6OV2 +IOOV3 +40v4

subject to

2)'i+4>'2 + 3j3+}'4>40
3>'i+33;2 + 2j3 + >'4>35
2}'i + >'2 + 4j3 + y4 > 45
>'1.>'2>>'3..V4^0

= 0; J3 = $5 = the marginal value of 1 hr

">) >'l' 3'2

of operation 3 time; >'4 =$25 = the marginal

value of 1 ft^ of storage space

c) It does not have an effect. In the alternate

solution the dual values remain the same, i.e.,

}'3 =$5 and _V4 =$25.

110
d) ct. basic:

^> Basic
c) C], basic:

.03 +A .02

Variables
 :

91

A-.: 3.6 + 0J^>0 A,: 1 .6 + OA > 5,: 4.4 + A>0

A>-4.4
Therefore. A> -4.4. Since

qi=20-A
A = 20-q,
20-^1 >-4.4
qi < 24.4

^2- Solving for the Cj - Zj row inequalities:

X2: 3.6 + .2A>0 .vi: 1.6-.133A>0 -10-4A/3<0 C3=12 + A

.2A > -3.6 -.133A>-1.6 ^A/3<0 A = C3-12
A>-18 -A>-12 -A < 30/4 C3- 12 > -30/4
A< 12 A > -30/4. C3> 18/4

jj. 4.4 + .47A>0 -2 - A/3 < A = C3- 12

.47A > -4.4 -A/3<2 f3 12>-6
A > -9.36 -A<6 C3>6
Therefore, -9.36 < A< 12. Since
A>-6

92 = 30 - A
^ - A/3 < A = C3 12
-A/3<4 C3- i2>-i:
A = 30 - 92
-9.36 < 30 -92 ^12
-A< 12 c%>0
A>-12
18 < 92 ^39.36
The present solution will remain optimal as

long as C3 > 6.
a:2: 3.6-.8A>0 .ti: 1.6 + .2A>0
-.8A>-3.6 .2A>-1.6 Summarizing: cj < 16. C2 < 4. C3 > 6.

-A > -4.5 A > -8

A<4.5
si 4.4-3.2A>0
-3.2A > -;.4
-A > -1.375
A<-1.375
Therefore. -8 < A< 1.375. Since

93=12-A
A= 12-93
-8< 12-93 < 1-375
10.625 < 93 < 20
e) The marginal value of mg of carbohydrates
1 is

S.Ol. The sensitivity range for 93.

carbohydrates, is 10.625 < 93 < 20. Thus, the
dietician could lower the requirements for
carbohydrates to 10.625 at the marginal value
without the solution becoming infeasible.

113
50. a) Minimize Zj = l,200>i + 500j3
subject to

.50j, +>'2>1.25
1.2vi->'2+>'3>2.00
My
+>'2>1.75I

>'l,>'2,>'3^0

)1 = the marginal value of an additional budget

dollar
= $0
>'2 = the marginal value of increasing the
combined demand for cheese sandwiches
by one sandwich
= $1.75
J3 = the marginal value of producing an
additional ham salad sandwich
= $3.75

b) ci, non-basic:

-.5 + A<
A<.50
Since cj = 1.25 + A; A = c, - 1.25. Therefore,

ci-1.25<.50
c, < 1.75

Q, basic:

-3.75 -A<0
-A < 3.75
A > -3.75
Since C2 = 2 + A; A= C2 - 2. Therefore,

C2-2>-3.75
C2>-1.75

C3, basic:

-.5-A<0 -3.75- A <0

-A < .5 -A < 3.75
A>-.5 A > -3.75
Summarizing,

-3.75 < -.5 < A

and, therefore,

-.5<A
Since £-3 = 1.75 + A; A = C3 - 1.75. Therefore,

-.5<A
-.5<C3- 1.75
1.25 <C3

114
c) qy-
c\, basic:

3 + A
Variables
b) When determining sensitivity ranges for 9,
values, since artificial values are eliminated,
the surplus variable column coefficients must
be used. This corresponds to aq,- A. change.

9i:
52. a)}i = 7/15 = $.467 d) Range for 95:

Range for <7i: 55: 8 + A>

A>-8
xi: 8 + A/15>0 xj: 16- A/30 >0
A/15 > -8 -A/30 > -16 Since ^5 = 40 + A,

A>-120 A<480 A = ^5 - 40
JC3: 3-A/40>0 (?5 - 40 > -8
A> 120 95^32
54: 36 A/30>0 55: 8-A/10>0 No, increasing ^5 from 40 to 50 will have no
-A/30 > - 36 -A/10 > -8 affect on the optimal solution.

A< 1,080 A<80

e) Since y\ =1/15 - $.467, pears should be
-120<A<80
secured.

Since q\ = 320 + A, 53. a) >'2 = $0, spruce has no marginal value

A = (71-320
-120>gi-320<80 sj: 70 + A>
200 <qi< 400
A>-70
As many as 400 pears can be purchased.
Since ^2 = 160 + A,

= $.067
A = 92- 160
b) V2= 1/15
92- 160 > -70
Range for q2: 92^90
xi: 8 + A/30>0 xy. 16 + A/15 >0
b) >'3 = $2, marginal value of cutting hours
A/30 > -8 A/15 > -16
A > -240 A > -240 93:

JC3: 3-A/10>0 S4: 36- A/30 >0 sf 80-4A/3>0 52: 70 + A/3 >
-A/10 > -3 -A/30 > -36 -^A/3 > -80 A/3 > -70
A<30 A< 1,080 A<60 A>-210
-240 < A < 30
x^: 20 + 2A/3>0 xj: 10-A/3>0
Since ^2 = 400 + A, 2A/3 > -20 -A/3 > -10
A>-30 A<30
A = ^2 - 400
-240 <q2- 400 < 30 Therefore, -30 < A < 30. Since 93 = 50 + A,
160 < 92 ^430
A = 93 - 50
Range over which the value of peaches is valid 20 < 93 < 80

c) Range
d) From part a, ^2 > 90; thus, a decrease from 160
to 100 lb of spruce will not affect the solution.

e) Compute the range for c\, a nonbasic Cj value.

c\ = 4+A
-2 + A<
A<2
Ci-4<2
ci<6
The unit profit from Western paneling would
have to be $6 or more before it would be
produced.

f) Compute the range for C3.

CASE SOLUTION:
'THE POSSIBILITY" RESTAURANT —
CONTINUED
a) First the linear programming model formulated for
the case problem must be converted to standard
form, as follows:

maximize Z= \2x\ + 16;c2

subject to
9
 Since ^i = 60 + A, A= ^, - 60. Therefore,
d) The final question concerns an increase in the
-2.86<^, -60< 12.72 coefficient for c^ from $12 to $14. This
57.14 < 9, < 72.72 requires the computation of the range
for c\.

Thus, an increase of 10 meals does not c\, basic:

affect
the shadow price for mean demand, which is
$8.00. An increase of 10 meals will result
-8 - 2A < -16 + 4A<0
in
increased profit of ($8)(10) = $80, -2A<8 4A< 16
which
exceeds the advertising expenditure
of $30. A<4
The ad should be purchased. -4<A<4
Since c, = 12 + A, A= -
c) The reduction in kitchen staff
from 20 to 15
c, 12. Therefore,
hours requires the computation of the
-4<ci - 12<4
sensitivity range for^2-
8<ci < 16
Smce ci = $14 is within this range the price
qi-
increase could be implemented
without affecting
Si- 20-10A>0 54: 14 + 4A>0 Pierre's meal plans.
-20A > -20 4A>-14
A<1 A>-3.5
xy. 40-40A>0 xi- 20 + 4A >
-40A > -40 4A > -20
A< 10 A>-5

Summarizing,

-5<-3.5<A< 1 < 10

and,

-3.5 <A< 1

Since ^2 = 20 + A, A= 92 - 20. Therefore,

-3.5 < 92 - 20 < 1

16/5 < 92 < 21

A reduction of 5 hours to 15 hours would

exceed the lower limit of the sensitivity
range.
This would result in a change in
the solution
mix and the shadow price, so the impact could
not be totally ascertained from
the optimal
simplex tableau. Solving the model
again with
^2=15 results in the following new solution,

:f 1 = 5.45
:?3 = 8i.82
x\ = 49.09
X2 = 5.45
Z= $676.36
Notice that simply using the shadow
price of
$16 for staff time (hr) would have
indicated a
loss in profit of only (5 hr)( = $80, or
1 6)
Z=
$720. The actual reduction in profit to
$676.36 is greater. The final question concerns
an mcrease in the coefficient for C|
from $12
to $14. This requires the computation of
the
range for C|.

122
 Chapter Six

ANSWERS TO QUESTIONS will result in the greatest net decrease in

cost (- ky).
1. MODI duplicates the steps of the stepping- d) Repeat steps b and c until all ky values are
stone method, except that individual cell cost positive or zero
changes are determined mathematically in

MODI. 10. The steps of the assignment method are as follows:

a) Perform row reductions.
2. In a balanced transportation problem supply b) Perform column reductions.
equals demand, while in an unbalanced c) In the completed opportunity cost table, cross
problem supply exceeds demand or vice versa. out all zeros using the minimum number of
horizontal and/or vertical lines.
3. The minimum cell cost method and VAM take
d) If there are fewer lines than rows or columns,
cell cost into consideration in selecting the

initial solution, while the northwest comer

subtract the minimum uncrossed value from

method does not.

all other uncrossed values and add this same
value to all places where two lines intersect.
4. An optimal solution is indicated and one or Repeat step c.
more of the individual cell cost change {k,j) e) If the lines required equal the number of
values will equal zero. The alternate optimal rows or columns, the optimal solution exists
solution is determined by allocating units to the and the assignments can be made.
cell with the cell cost value equal to zero
according to its stepping-stone path.
PROBLEM SUMMARY
5. A penalty cost is the difference between the
cost of making the correct decision and the 1. Balanced transportation

cost for an incorrect decision — i.e., the penalty 2. Balanced transportation

suffered for making a wrong decision.
3. Short answer, discussion
6. Degeneracy in a transportation tableau occurs
when m rows + n columns - 1 > the number of 4. Unbalanced transportation
cells with allocations. It is remedied by 5. Balanced transportation
arbitrarily creating a cell with an artificial

(nonexistent) allocation.
6. Unbalanced transportation

7. k,j values correspond to zj

— Cj values in a simplex
7. Unbalanced transportation
tableau. 8. Unbalanced transportation
8. The line test indicates whether there are a sufficient 9. Unbalanced transportation, multiple optimal
number of indep)endent assignments for an
10. Sensitivity analysis (6-9)
optimal solution.

9. The steps of the stepping-stone method are as

11. Unbalanced transportation, multiple optimal
follows: 12. Unbalanced transportation, degenerate
a) Determine the stepping-stone paths and cost
changes for each empty cell.
13. Unbalanced transportation, degenerate

b) Allocate as much as possible to the empty 14. Balanced transportation

cell with the greatest net decrease in cost.
15. Balanced transportation
c) Repeat steps a and b until all empty cells
have positive changes indicating an optimal 16. Sensitivity analysis (6-15)
solution.
The steps for MODI are as follows: 17. Unbalanced transportation, multiple optimal

a) Compute w, and vj values for each row and 18. Sensitivity analysis (6-17)
column by applying the formula ui + vy = Cy
19. Shortage costs (6-17)
to each cell with an allocation .

b) Compute k,j for each empty cell, where 20. Unbalanced transportation

c) Allocate according to the stepping-stone path

21. Unbalanced transportation, multiple optimal

as much as possible to the empty cell that 22. Balanced transportation

123
23.
2. a)

\. To
From ^\^
4. Initial solution using VAM:

From ^x^
6. a)

\^To
 b) Solution using the stepping-stone method; TC = $1,370

\. To
From \s„^
9. a) The initial solution should be found using VAM, since
it is the most efficient.

\. To
^^\ To
b) MODI solution:
12. a)

^\ To
b) The initial solution is degenerate. is allocated arbitrarily
to XiD-

\. To
From \v,^
c) Minimize Z = 7j:,a + &x,u + 5xic + 6x2a + 'W-f:B
+ 6x:( + lOxiA + 4jriH + 5jck
+ 3x4A + 9x4„ + Mxii
subject to

JCiA+ x,a + x,i < 5

JC2A+ JrjH + Jr:c- ^ 25
Jr,A + Jr,B + Jf.ic ^ 20
XiA + -if-iH + -f4( 25 ^
J^iA + J^:a + -f.lA + JfjA 10 =
J^IB + -^28 + -»^'H + A-,1, 20 =
Xn + Xn + Xu + Xu = 15
X,, >
14. a) Minimum cell cost method

"\^ To
From \^
 b) Stepping-stone method;

^\ To
From\v^
b)
18.

"\^^ To
From ^\^
19.

Froni\^
21. a)

"\. To
22.

^\
From
To
\s„,^
 Allocate 40 students to cell (South, C):

^\ To
From^\
 Allocate 40 students to cell (South. C):

\. To
From\^
25. The initial solution is determined using VAM, as
follows.

^\ To
 Alternate, allocate 18 cases to (E,4)

^\To
Froni\^
This new solution results in a greater profit than the
solution in problem 25 ($1,545 > $1,528). Thus, some
of the demand (specifically at store 3) should be met
from the distributors.

27. Solve the model as a linear programming

model to obtain the shadow prices. Among the
5 purchase locations, the store at Albany has
the highest shadow price of $3. The sensitivity

range for supply at Albany is 25 < <7i

< 43.
Thus, as much as 17 additional cases can be
purchased from Albany which would increase
profit by $51 for a total of $1,579.

28.

\. To
From ^\^^
 Allocate "0" to cell (4. dummy):

Frorn^^
 Allocate "0" to cell (2,B):

Froni\.
 31. Optimal transportation tableau (costs = SlOOs).
Multiple optimal solutions exist.

Froni^\

January

April

Mav

June
34.

^"^^ To
From ^\.^^
36. This is an assignment problem. 37.

Solu
Optimal solution:

1-C
Opportunity cost
45. This changes the solution and increases the total

mileage. The new assignments are;

Officials 3. 6 and 7 - Athens

Officials 1. 2 and 8 - Columbia
Officials 4. 5 and 9 - Nashville

Total mileaae. Z= 1,220

46. Solution Summary:

Al's - Parent's Brunch

Bon Apetit - Post-game Party
Bon Apetft - Lettermen's Dinner
Divine - Booster Club Luncheon
Epicurean - Contributor's Dinner
University - Alumni Brunch

Total Cost (Z) = S 103.600

157
47. Subtract all values from 100 and minimize the difference. 48.
"*9. Subract all sales values from highest sales, $630.

Home
Furnishings China Appliance Jewelr>

1
50.

\. To
51. Opportunity cost table:
GA
CASE SOLUTION:
THE DEPARTMENT OF MANAGEMENT
SCIENCE AND INFORMATION
TECHNOLOGY AT TECH
The problem is formulated as a transportation
model with the ten faculty as the source, each
with a supply of 2 sections, and with the eight
courses as the destinations with demand of
either 1, 2 or 3 sections.

The solution generated by QM for Windows

results in the following teaching schedule
(however there are multiple optimal solutions.)

Clayton
Multiple
1-B
 (65)

35 (80)

(105)

CASE SOLUTION: BURLINGHAM

TEXTILE COMPANY
This assignment problem has ten sources and
five destinations. Thus, if it is solved by hand it

would require five dummy columns. It is also a

maximization problem which would require all

the scores to be subtracted from the highest

score in the tableau (i.e. 102) and then
minimized.

The solution output for this assignment

problem obtained using QSB+ is as follows.

Applicant
 Chapter Seven

ANSWERS TO QUESTIONS 6. The upper bound is the Z value for the relaxed
simplex solution for the problem, the lower
1. The three basic types of integer programming bound is the Z value for the rounded-down
models are total integer, in which all decision integer solution.
variables must have integer values, 0-1
integer, in which the decision variables can 7. A relaxed solution is the simplex solution to

equal either or and, mixed integer, in

the linear programming problem with the
1 ,

integer restrictions removed.

which some decision variables are continuous
and some are integer. 8. The optimal solution is reached when a

2. If noninteger solution values are rounded up,

feasible integer solution is generated at a node
they may no longer be feasible; whereas if
and the upper bound at that node is greater

values are rounded down, a suboptimal than or equal to the upper bound at any other

solution may result.

ending node.

3. The steps of the branch and bound method are

as follows. PROBLEM SUMMARY
a) Determine the optimal simplex solution 1. Integer model, branch and bound solution, 5
with the integer restrictions relaxed. nodes

b) At node 1 let the relaxed simplex solution 2. Integer model, branch and bound solution, 3
be the upper bound and the rounded-down nodes
integer solution the lower bound.
3. Integer model, branch and bound solution, 3
c) Select the variable with the greatest nodes
fractional part and create a new < constraint
and a > constraint, reflecting the partitioned 4. Integer model, branch and bound solution, 3

integer values. nodes

d) Create two new nodes for the constraints

5. Integer model, branch and bound solution, 3

developed in step c.
nodes

e) Solve the relaxed LP model with the new 6. Integer model branch and bound solution, 5

constraint added at each node. nodes

f) The relaxed simplex solution is the upper 7. Integer model, branch and bound solution, 7

bound at each node, and the maximum nodes

integer solution (at any node) is the lower 8. Mixed integer model branch and bound
bound. solution, 3 nodes

g) If a feasible integer solution is derived with 9. 0-1 integer model, branch and bound solution,
the greatest upper bound of any ending 3 nodes
node, the solution has been reached;
otherwise, branch from the node with the 10. Integer model, branch and bound solution, 3

greatest upper bound. nodes

h) Return to step 11. 0-1 integer model, implicit enumeration

c.
solutions (Problem 9)
4. Implicit enumeration eliminates obviously
infeasible solutions and evaluates the 12. 0-1 integer model, implicit enumeration

remaining ones; in complete enumeration solution, discussion

all

solutions are evaluated. 13. Integer model, formulation and computer

5. For the variable with the greatest fractional solution

part, new < and > constraints arc created that 14. Integer model, formulation and computer
together restrict the variable to an integer solution
value. These constraints are subsequently
added to the linear programming model and 15. Integer model, formulation and computer

solved. solution

166
16. Integer model, formulation and computer PROBLEM SOLUTIONS
solution
1. a. Relaxed solution: .x\ - 3.33
17. Integer model, formulation and computer
solution ©UB-
LB=
16.65
15 (.t,
Ui = 3.33,
=:3,X2 = 0)
jr2 = 0)

18. Integer model, knapsack problem, formulation

Branch on .vj: < 3, xj > 4
x\
and computer solution
Addition of .V] s 3 constraint:
19. Integer model, formulation and computer maximize Z= 5x\ + 4x2
solution subject to

20. Integer model, formulation and computer

3x, -^4x2 ^ 10

solution xi < 3
X], X2 ^
21. Continuation of Problem 20, formulation
Relaxed solution: X2 = 2.3, xj = 3, z = 16
22. Continuation of Problem 20, formulation
Addition of jci S: 4 constraint:
23. Integer model, formulation and computer maximize Z= 5xi + 4x2
solution subject to
3;ci + 4.X2 > 10
24. 0-1 integer-model, computer solution
JTl > 4
25. Integer model, computer solution xi, X2 S:

Relaxed solution: Infeasible

26. Integer model, branch and bound solution, 7
nodes

27. Integer model with 0-1 restriction

28. Integer model, formulation and computer

solution

29. 0-1 integer model, computer solution

30. Integer model, formulation and computer

solution

31. Integer model, formulation and computer

solution (7-30)

167
 UB= 16.65 (Xi = 3.33, X2 = 0) UB= 16.65 (Xi = 3.33, X2 = 0)
LB = 15 (Xi =3, X2 = 0) LB = 1 5 (Xi - 3, X2 = 0)

UB = 16 (Xi =3, X2 = .25)

LB = 1 5 (Xi = 3, X2 = 0)

UB = 16 (Xi = 3, X2 = .25) Infeasible

LB = 1 5 (Xi = 3, X2 = 0)

Branch on xj t'roin node 2: jct = (not

necessary to attempt ^ since negative UB = 15 (xi = 3, X2 - 0) UB = 14 (xi = 2, X2 = 1)
values not possible), .x\ > 1 LB = 1 5 (Xi = 3, X2 = 0) LB = 1 5 (Xi = 3, X2 = 0)

Addition of X2 = constraint:
 2. Relaxed solution: .vi = 5.71, Z= 17.13 b. Relaxed solution: .v, = 10.5, X2 = 23.7, Z= 1,473

LB = 17.13 (x, = 5.71, JC2 = 0) UB = 1,473 (x, 10.5, X2 = 23.7)

UB = 18 (x, = 6,a:2 = 0) LB = 1,420 (jc, 10, x: = 23)

Branch on Xi: x, ^ 5, x, ^6 Branch on ^2:^2 ^ 23, X2 ^ 24

Addition of X| ^ 5 constraint: Addition of xi ^ 23 constraint:

minimize Z = 3x^ + 6x2 maximize Z = SOjti + 40^2

subject to subject to

7x, + 3^2 > 40 3xi + 5x2 < 150

^, ^5 lOx, + 4x2 ^ 200
Xi,X2 s X2 < 23

Relaxed solution: x\ = = 1.67, Z - 25 x,,X2 >

5, .vt

Addition of x, S: 6 constraint:
Relaxed solution: x\ = 10.8, xt = 23, ^i = 2.6
minimize Z = 3xi + 6x2
subject to Addition of X2 ^ 24 constraint:
7x, + 3x2 s 40
maximize Z = 50xi + 40x2
x, > 6
subject to

3xi + 5x2 ^ 150

Relaxed solution: xj = 6, s\ = 2, Z= II lOx, + 4x2 ^ 200
X2 ^ 24
X|,X2 S
LB = 17 (X, = 5.71,Xj = 0)
Relaxed solution: .vj = 10, xt = 24.5, S2 =4
UB = 18 (X, =6,X2 = 0)
UB = 1,473 (X, = 10.5, X2 = 23.7)
LB = 1,420 (X, = 10,X2 = 23)

LB = 25 (X, = 5, X2 = 1 .67) LB = 18 (x, = 6, x^ = 0)

UB= 18(x, =6,X2 = 0) UB= 18(x, =6, X2 = 0)

Optimal solution at node 3:

X, = 6 UB= 1.460 (X, = 10.8,X2 = 23) UB = 1,460 (x, = 10,Xj = 24)

X2 = LB = 1,460 (X, = 10,Xj = 24) LB = 1,460 (x, = 10,X2 = 24)
7=18
Optimal solution at node 3:

3 a. maximize Z = 50x, + 40x: (profit, $) X, = 10

X2 = 24
subject to
< Z= 1 .460
3xi + 5x2 150 (wood, yd")
lOx, + 4x: < 200 (labor, hr) The rounded down solution would not be
X|, Xj ^ and integer optimal.

4. a. maximize Z = 400x, + 100x2 (profit, $)

subject to

8xi + 10x2 ^ 80
2xi + 6x2 ^ 36
x, < 6
X|, X2 s and integer

b. Relaxed solution: xi = 6, X2 = 3-2, S2 = 4.

Z= 2,720

169
 UB = 2,720 (X, = 6, Xj = 3.2)
/T^ UB = 2,720 (x, = 6, X2 = 3.2)
LB = 2,700 (X, = 6, Xj = 3)
V^J29/ LB = 2,700 (x, = 6, .^2 = 3)

Branch on X2: JC2 ^ 3, j:2 — 4

Addition of j:2 ^ 3 constraint:

maximize Z = 400xi + 100a:2

subject to
8x, + 10x2 ^ 80
2a:, + 6x2 < 36
jc, < 6
X2 ^ 3 UB = 2,700 (X, = 6, X2 = 3) UB = 2,400 (x, = 5, X2 = 4)
X,,X2 ^ LB = 2,700 (X, = 6, X2 = 3) LB = 2,700 (x, = 6, Xj = 3)

Optimal solution at node 2:

X, = 6
X2 = 3
Z= 2,700
Relaxed solution: X\ = 6, X2 = 3, S\ = 2, S2 = 6,

Z= 2,700

Addition of ^2 ^ 4 constraint:

maximize Z = 400Ar, + IOO.X2

subject to
8x, + 10;t2 ^ 80
2jt, + 6JC2 ^ 36
j:, < 6
X2 ^ 4
a:i,a:2 2:

Relaxed solution: x\ = 5, X2 = 4, .1^2 = 2, ,^3 = 1,

Z= 2,400

5. a. maximize Z = SOjti + 10j:2

subject to

Xi + X2 ^ 15
4x, + jr2 :s 25
jTi, jTj > O and integer

5. b. /^T^ UB = 312.5 (x, = 6.25, X2 = 0) c. Relaxed solution:

\^n^ LB = 300 (a:, = 6, ;c2 = O)

Branch onjri:x, < 6,jC| > 7

Addition of X| ^6 constraint:

maximize Z = SOx, + IOjtt

subject to

Xi + X2^ 15
4a:, + a;2 :£ 25
a:, < 6
a:,,j:2 >

Relaxed solution: x\ = 6, xt = I , .9| = 8, Z= 3 10

Addition of x, ^ 7 constraint:

maximize Z = 50x, + 10x2

subject to
X, + X2 s 15
4x, + X2 :^ 25
X, >7
X|,X2 S
Relaxed solution: Infeasible

170
 UB = 312.5 (Xi =6.25, X2 = 0) Node 3:

LB = 300 (Xi = 6, X2 = 0)
25

20

15
Optimal integer nfeasible
solution: 10

5
-
X, = 1

5 10 15
^^^^^1
20 25 5 10 15
^^-^^^1
20 25
UB = 310(Xi =6, X2 = 1) Infeasible
LB = 310(Xi =6, X2 = 1) Optimal solution at node 2:

Xi =6
X2 = 1

Z=310
6. a. maximize Z = 600jC| + 540x: + 375 JC3
subject to

Xt + X2 + Xi^ 12
X| < 5
80jC| + 70x2 + 50x3 ^ 750
Xi, X2, Xj s and integer

Relaxed solution: .xt = 10.71, jt = 1-28, 53 = 5,

2=5.785.7

©UB =
LB =
5,785.7 (x,
5,400 (x, =
=
0,
0,

X2
X2
=
= 10.71, x,
10, x, = 0)
= 0)

Branch on X2: X2 ^ 10, X2 ^ 11

Addition of X2 ^ 10 constraint:

maximize Z = 600xi + 540x: + 375x3

subject to
X, + X2 + X3 < 12
X, <5
80x, + 70x2 + 50x3 s 750
X2 ^ 10
Xi,X2, X3 S

Relaxed solution: xj = 0.63, xj - 10, si - 1.39, sj = 4.37,

Z= 5,778

171
Addition of ^2 ^ 11 constraint:

maximize Z = 600jci + 540j:2 + 375x3

subject to

Xt + X2 + X) ^ 12
X, <5
80a:, + 70x2 + 50x3 < 750
X2 ^ 11

Relaxed solution is infeasible. (This can be determined by

inspection or by working out the simplex model.)

UB = 5,785.7 (X, = 0, Xj = 1 0.71 Xj = 0)

,

LB = 5,400 (X, = 0, Xj = 10, X3 = 0)

X2< 10

UB = 5,778 (X, = .63, x^ = 10, X3 = 0) Infeasible

LB = 5,400 (X, = 0, Xj = 10, X3 = 0)

Branch from node 2 on x,: x, = 0, X| > 1

Addition of X| =0 constraint:
maximize Z = 600xi + 540x2 + 375x3
subject to
X| + x: + X, < 12
X, <5
80x, + 70x2 + 50x3 s 750
X2 ^ 10
X, =
X|,X2, X3 ^

Relaxed solution: xj = 0, X2 = 10, X3 = 1, jj = 1,

52 = 5, Z= 5,775

Addition of x, > 1 constraint:

maximize Z = 600x, + 540x2 + 375x3

subject to
X, + X2 + X3 s 12
x, < 5
SOx, + 70x2 + 50x3 ^ 750
X2 ^ 10
Jt, ^ 1

Xi,X2, X3 ^

172
 Relaxed solution: .vi = 1. .V2 = 9.57. s\ = 1.43. 51 = 4.

J4 = . 43. Z= 5.767.8

UB = 5,785.7 (X, = 0. Xj = 10.7, X3 = 0)

LB = 5,400 (X, = 0, Xj = 10, X3 = 0)

X, < 10

UB = 5,778 (X, = .63, Xj = 10, X3 = 0)

LB = 5.400 (X, = 0, X2 = 10. X3 = 0)

Infeasible

= 10, X3 = 1) UB = 5,767.8 (X, = 1 Xj = 9.57, X3 = 0)

UB = 5,775 (X, = 0, Xj
,

= = 10,X3=1) LB = 5,775 (X, = 0. Xj = 10, X3 = 1)

LB = 5,775 (X, 0, Xj

Optimal solution at node 4:

X, =
Xj = 10

Z= 5.775

7. a. maximize Z = SOx, + 40^2

subject to
2x, + 5x2 ^ 35
3x, + 2x2 ^ 20
X,, X2 s and integer

UB = 372.9 (x, = 2.73, xi = 5.91)

LB = 300.0 (x, = 2, X2 = 5)

Branch on X2: X2 ^ 5, X2 ^ 6

Addition of X2 :^ 5 constraint:

maximize Z = 50xi + 40x2

subject to
2x, + 5x2 s 35
3x, + 2x2 ^ 20
X2^5
;c„X2^0

173
Relaxed solution: X] = 3.33, X2 = 5, ^j = 3.33,
Z= 366.5

Addition of Jt:2 ^ 6 constraint:

maximize Z = SOxi + 40jC2
subject to

2xi + 5;t2 < 35

Sjt, + 2x2 ^ 20
JC2 ^ 6
x,,Ar2 ^

Relaxed solution: x\ = 2.5, JC2 = 6, ^2 = 0.5,

Z=365

UB = 372.9 (X, = 2.73, Xj = 5.91)

LB = 300.0 (x, = 2, Xj = 5)

UB = 366.5 (X, =3.33,X2 = 5) UB = 365.0(x, = 2.5,X2 = 6)

LB = 300.0 (X, =2,X2 = 5) LB = 300.0 (X, = 2, X2 = 5)

Branch from node 2onX|:xi ^ 3, x^ >4

Addition of x, < 3 constraint:

maximize Z = SOx, + 40^2

subject to

2xi + Sat, s 35
3jr, + 2j«:: < 20
X: < 5
X| < 3
X|,X2 >

174
Relaxed solution: x\ = 3, .vt = 5, 5| = 4, jt = 1.

Z=350

Addition of j:, > 4 constraint:

maximize Z = 50xi + 40jc2

subject to
2x| + 5x: < 35
3xt + 2a:: < 20
JC; < 5
j:, > 4
jCi.jT: >

Relaxed solution: .V] = 4, .V2 = 4, jj = 7, ^3 = 1,

Z= 360

UB = 372.9 (x, = 2.73, X2 = 5.91)

LB = 300.0 (X, = 2, Xj = 5)

UB = 366.5 (X, = 3.33, Xj = 5) UB = 365.0 (X, = 2.5, x^ = 6)

LB = 300.0 (X, = 2, X2 = 5) KB = 300.0 (X, =2,X2 = 5)

UB = 350.0 (X, = 3, Xj = 2) UB = 360.0 (x, = 4, Xj = 4)

LB = 360.0 {X, = 4. Xj = 4) LB = 360.0 (x, = 4, Xj = 4)

Branch from node 3 (since it now has greatest UB) on x,:

JC, < 2, X| > 3

Addition of x, ^2 constraint:

maximize Z = SOx, + 40x2

subject to
2x, + 5x2 ^ 35
3x, + 2x2 ^ 20
X2 S: 6
x, <2
X,,X2 S

175
 Relaxed solution: a'i = 2, X2 = 6.25, S2 = 0.5, .Y3 - 0.25,
Z = 350

Addition of Xi s 3 constraint:

maximize Z = 50a: 1
+ 40;iC2

subject to
2x, + 5x2 ^ 35
3x, + 2x2 ^ 20
X2 s 6
X, > 3
X,,X2 2:

Relaxed solution is infeasible (which can be determined by

inspection or simplex solution).

UB = 372.9 (x, = 2.73, Xj = 5.91)

LB = 300.0 (X, = 2, X2 = 5)

UB = 366.5 (X, = 3.33, x^ = 5) UB = 365.0 (X, = 2.5,Xj = 6)

LB = 300.0 (X, = 2, Xj = 5) 300.0 (X, = 2, Xj = 5)

Infeasible

UB = 350.0 (X, = 3, X2 = 5) UB = 360.0 (x, = 4, Xj = 4) UB = 350.0 (x, = 2, Xj = 6.25)

LB = 360.0 (X, = 4, Xj = 4) LB = 360.0 (x, = 4, Xj = 4) LB = 360.0 (x, = 4, Xj = 4)

Optimal solution at node 5:

X, =4
X2 = 4
Z= 360.0

The rounded down solution would not be

optimal.

176
V
8. a. maximize Z = SOOOjr, + 6(X)0j:2 (annual return, $)
subject to
70 JT, + 30j:: < 500 (capital outlay, $ 1,000s)
X, + 2xi < 14 (annual maintenance budget, Sl,000s)
jc, s and integer
X: >
b. .V| = 5.27. A2 = 4.37, Z= 68,380

1 A UB = 68,380 (x, = 5.27, x. = 4.37)

[i3^ LB = 66,220 (x, =5,x. = 4.37)

Branch on x\ (since x\ is the only variable restricted to an

integer value): JCi ^ 5, JCi ^ 6

Addition of Xi ^ 5 constraint:

maximize Z = 8000a:, + 6000^::

subject to
70x, + 30a:: < 500
AT, + JTj :s 14
a:, < 5
a:i ^ and integer
Xj^O

Relaxed solution: .V] = 5. .V2 = 4.5. ^i = 15, Z = 67,000

Addition of a:, s 6 constraint:

maximize Z = SOOOx, + 6000Ar2

subject to
70a:, + 30a:2 < 500
a:, + 2a:2 -^ 14
a:, > 6
AT, s and integer
X2 ^

177
 Relaxed solution: x\ = 6, jto = 2.66, ^2 = 2.68,
Z= 63,960

UB = 68,380 (X, = 5.27, Xj = 4.37)

LB = 66,220 (x, = 5, x^ = 4.37)

UB = 67,000 (X, = 5, X2 = 4.5) UB = 63,960 (x, = 6, x^ = 2.66)

LB = 67.000 (X, = 5, Xj = 4.5) LB = 67,000 (x, = 5, Xj = 4.5)

Optimal solution at node 2:

X, = 5
Xj = 4.5
Z= 67,000
9. Relaxed solution: j:i
= 1, ^2 = 0.17, ;C3 = 1, 53 = 0.83,
Z= 1,916.67

UB = 1,916.67 (x, = 1, ;C2 = .17, x^ = 1)

LB = 1,800 (jt, = 1, X2 = 0, Xi = 1)

Branch on Xj: X2 = 0, jc; = 1

Addition of X2 = constraint:

maximize Z = l,000xi + 700x2 + SOOxj

subject to
5,000x, + 6,000x2 + 4,000x3 ^ 10,000
X, < 1

X2 = (not
necessary
to also have

X2^ 1)
X3^ 1

X|,X2, X3 = or 1

178
 Relaxed solution: .X] = 1. vt = 0. A3 = 1, 5] = 1,000,
Z= 1,800

Addition of jc: = 1 constraint:

maximize Z = l.OOOx, + 700x: + SOOx,

subject to
5,000x, + 6,000^2 + 4,000x5 < 10,000
< JT, 1

X2= 1

X3< 1

X\,X2,X} = or 1

Relaxed solution: xi - 0.8, X2 - 1, ^2 = 0.2, 53 = 1,

Z= 1,500

UB= 1,916.7 (X, = 1,Xj= .167, X3= 1)

LB = 1.800 (X, = 1,X2 = 0,X3 = 1)

UB= 1,800 (x, = 1.Xj = 0,X3= 1) UB= 1,500 (x, = .8,X2= 1,X3 = 0)
LB = 1,800 (X, = 1,X2 = 0.X3= 1) LB = 1.800 (x, = 1,Xj = 0, X3 = 1)

Optimal solution at node 2:

X, = 1

Xj =
X3 = 1

Z= 1,800

10. Relaxed solution: .^i

= 2, ^2 = 3.33, Z= 30

®UB =
LB =
30
28
(x,

(jc,
=
=
2,

2,
X2
X:
=
=
3.33,
3, x,
jc,

=
=
0)
0)

Branch on Xz- X2 ^ 3, Xi ^ '^

Addition of ^2 ^ 3 constraint:

maximize Z = 5x, + 6x2 + 4x3

subject to
5x, + 3x2 + 6x3 < 20
X| + 3x2 ^ 12
X2<3
x,,X3>0
X2 ^ and integer

Relaxed solution: X] - 2.2, X2 = 3, ^2 = 0-8, Z= 29

Addition of X2 ^ 4 constraint:

maximize Z = 5xi + 6x2 + 4x3

subject to
5x, + 3x2 + 6x3 < 20
X, + 3x2 ^ 12
X2 2: 4
X„X3^0
X2 ^ and integer

Relaxed solution: x\ = 0, X2 = 4. X3 = 1 .33, Z= 29.32

179
 UB = 30 (X, = 2, X2 = 3.33, X3 = 0)
LB = 28 (X, = 2. Xj = 3, X3 = 0)

UB = 29 (x, = 2.2, Xj = 3, X3 = 0) UB = 29.32 (x, = 0, X2 = 4, X3 = 1 .33)

LB = 28 (X, = 2, X2 = 3, X3 = 0) LB = 29.32 (x, = 0, X2 = 4, X3 = 1 .33)

Optimal solution at node 3:

X, =0
Xj = 4
X3 = 1 .33
Z= 29.32

11.

Solution JCi X2 X} Feasibility

12. a.
16. .vi = day contacts by phone =
18. Maximize Z 9Qxt + ISOx; -1-
30jc3
At = day contacts in person subject to
.V3 = night contacts by phone 2x, -t- 3x: -(- JCj < 5
= night contacts in person
.V4
Solution: Z = $240, x, = 1, x, = 1, x, =
Maximize Z = SI6.V1 + SS.vi + I7.V3 + 37.V4 19. X, = no. of salespeople to East, xi = no. of salespeople to

subject to:
Midwest, xj = no. of salespeople to West

maximize Z = 25,(X)0x, -1-

IS.OOOx; + Sl.OOOxj
+ .V4 < 575
.V2
subject to
6x1 + 13.V2^ 1,320
Xi + X2 = 100
+ Xi
7x3 + 19.r4 < 2,580
5,000x, -I- ll.OOOx: + ^ 700,000
7,000x3
xi,.r2,.V3,A4 s and integer X, > 10
X2 > 10
Integer solution:
Xj > 10
.VI = 220 x,,X2, X3 > and integer
XT, = 368 Solution: x, = 20, x. = 10, x, = 70, Z = 2,850,000
Z= $9,776 20.x,, = vehicles [1,000s shipped from plant 1 (/ =
1, 2, 3, 4, 5) to warehouse ; ( ;
= A, B, C, D), y, = plant i
The non-integer solution is:
= = Oor 1
(i 1,2,3,4,5)
jc, = 220 minimize Z = 2,100y, + SSOy^ + 1,800^3
X3 = 368.57 + 1,100^4 + 900>'5 + 56x,A
Z= $9,785.71 -1-
21xiB + 32xic + 65xiD
-t- 18x2A + 46x2B + 7x:c
The rounded down solution is only slighdy less
-I- 35X2D + 12X3A + 71X3B
(i.e.. 59.71)
-t- 41X3C + 52X3D + 30X4A
+ 24x48 + 28x4D + 45x5A

= tv ads + 50X58 + 26X5C + 31X5D

17. (a) .x\
subject to
X2 = newspaper ads
C\ ~ XjA ~ X18 ~ Xjc ~ XiD =
X2 = radio ads
C: — Xta ~ X28 ~ X2C ~ X2D =
minimize Z= $25,000.vi + 7,000.V2 + 9.OOO.V3 —Cj X3A ~ X38 ~ X3C ~ X3D =
—Ci XiA ~ X4B ~ X4C ~ X4D ==
subject to:
C5 " X5A ~ X5B — X5C — X5D =
53,OOO.vi + 30,000.V2 + 4i.OOO.v3 > 20.000 XlA + X2A + XjA + X4A + XsA = 6,000
X18 + X2B + X38 + X4B + XsB = 14,00C
14,000
32.00O.V, + 2O.OOO.v2 + l8.OOO.v3
> 1.5 Xic + X2C + Xjc + X4C + X5C = 8,000
(21.000.V1+ lO.OOO.v2 + 23.OOO.v3)
XiD + X2D + X3D + X4D + XjD = 10,000

+ 12,OOO.V2 + 23,OOO.V3 c, < 12,000>',

34.000.vi
> .60 C2 s 18,000>'2
(53,000.v, + 30,00O.V2 + 41,000.V3) C3 ^ 14,000^3
.vi,.V2,.V3,.V4 ^ and integer C4 s 10,000y4
cs < 16,000y5
Integer solution:

x,=4 Solution: y2, .V4. y^ = 1. .V2a = 6.000

.v-.= .V2B= 4,000. .V2C = 2.000. .V4B = 10.000.

.V3= X5C = 6.000. .V5D = 10.000. Z= $3,902,000

Z = $99,999.99
21. Add the constraint y2 + }4
'^ 1 •

(b) Non-integer solution:

Solution: y2. y3. ys = 1. .V2b - 14.000
.V, = 2.9275 .V2C = 2.006. .V3A = 6.000. .V5c = 6.000.
.V2 = .9713 = 10.000. Z = $4,786,000
.vJd
.r3 = .383
Z= $83,433.65 22. Add the constraint y, ^ \\.

Solution: yi, y?, ys = 1, = 12.000

-Vib

.V2A = 6,060, .V2B = 2.000. .V2C= 2.000.

.V5C = 6.000. .V5D = 10,000. Z = $4,822,000

181
23. Maximize Z = 12,100a:i + 8,700^2 + 10,500x3 b)xi =1 (Management I)
X2 = 1 (Principles of Accounting)
subject to:
X5 = (Marketing Management)
1

360x1 + 375x2 + 410x3 - 30,000 X7 = 1 (English Literature)

X] + X2 + X3 ^ 67 Z = 30 hours per week
14x1 + 10x2+ 18x3 < 2,200 Minimum grade point average = 2.50
X1/X2 ^ 2
X2/X1 > 1.5
25. a) maximize Z= l,650xi + 850x2 + 790x3

xi,X2,X3 > and interger subject to

6.3x1 + 3.9x2 + 3.1x3 < 125

Integer solution: 17x1 + 10^2 + 7x3 ^ 320
xi = 22 -*!, -^2,-^3, ^ and integer
X2 = 34
b)xi = 10
X3= 11
X3 =20
Z= $677,500
Z = 32,300
24. a) minimize Z= 5xi + 10x2 + 8x3 + 12x4 + 7x5 The relaxed, noninteger solution is,

+ 10x6+ 8x7 = 13.61

XI

subject to X3 = 12.67
Z = 32,460.46
9x1 + 6x2 + 6x3 + 1x4 + 6x5 + 1X6 + 9x7 > 2.00 The rounded down solution is xi = 13,
3(xi+ X2 + X3 + X4 + X5 + X6 + X7) X3 = 12, and Z = 30,930, which is not optimal.
3(xi + X2 + X3 + X4 + X5 + X6 + X7) > 12 maximize Z-
26. 575xi + 120x2
X2 + X3 + X4 + X6 ^ 2
xi + X2 + X6+ X7 > 3
subject to

X, = or 1
40x, + 1 5x2 < 600
30x1 + 1 8x2 < 480
4x1 - .^2 ^
xi, X2 ^ and integer

UB = 4,964.7 (x, = 4.71 X2 =

, 1 8.82)
LB = 4,460 (x, =4, X2= 18)

UB = 4,747.5 (X, = 4.5, Xj = 18) UB = 4,925 (Xi =4.6, X2 = 19)

LB = 4,460 (x, = 4, X2 = 1 LB = 4,460 (Xi =4, X2= 18)

UB = 4,460 (X, = 4, X2= 18) Infeasible

uB = 4,700 (x, = 4, X2 = 20) '"'«^^''^'«
LB = 4,460 (X, =4, X2= 18)
LB = 4,700 = = 20)
(x, 4, X2

Optimal solution:
X, = 4
X2= 20
2=4,700

182
 1

27. Maximize Z = $575.V| + llO.xi + 45.^3 29. a. maximize Z= I8.V14 + 20.vig + 2\x\c + 17.V|o
+ 9x2A + 5.V2fi + 22.V2C + 8.r2D + 20.V3,, +
1 1 1

subject to:
20.V3B + 17.V3C + I9.V3/) + 24.^4^ + 2I.V4fl +
40x| + 15x2 + 4x3 < 600 \6x4c + 23.V4P + 22v5^ + \9x^[) + 2\xsc +
30^1 + I8.r2 + 5.r3 < 480 21-^50
4.V| -.V2 ^
.V3 = 20 V,
subject to
.vi,.V2.-V3 > and interger
•3-Vu ^.ViB + .6a-|c + .4.r|£, + .8.V2,a + .5.V2b +
+
vi =0or 1
1
. + -^'^20 + '-^3,4 + -3X35 + .6x3(7 +
l.r2C 1 1

Or the last restriction that V| = or can be .8x3/) + .2.V4^ + .8x4g + .6x4c + 9X4£) +
1

included in the model as a constraint. \ |

< 1. 1 .0x5^ + .9x5a + .Oxsc + 0x5/)/ 8x1^ +
I ' 1

20xig + 21xic+ 17x|£)+ 19x24 + 1525 +

Solution: 22x2c + 8x20+ 20x3^ + 20x3B + 7x3c +
1 1

.V, = 3 19x3o + 24x4^ + 2l.r4B + I6.V4C + 23.^4^ +

.V2= 16 22x5^ + 9x5B + 2 Ixsc + 2 1.V50 < .04
1

.V3 = 20
VI = 1
-'^M +xifi + X|c + xio< I

Z = S4.745 X2A + -V2B + X2C + -^2D - •

X3A + Xi,B + X1.C + X3D < 1

They should produce the batch of 20 stools

MA + MB + MC + -r4o < 1
since the profit is slightly greater ($4,745 vs.
X5A + X^B + X5C + X^D ^ I

$4,700).
+X2A+ X^A + X4A + X5A =
X\A 1

XiB+ X2B + Xt,b + X4B + X^B= •

28. .V| = bass boat XlC + X2C + X^c + X4C + X5C = '

A'2 = ski boat X\D + X2D + X3D + X4D + Xf,D = •

JC3 = speed boat Xjj= or I

Maximize Z = 20.500.vi + i2.OOO.r2 + 22.300.r3 b) x,c-I

x?.D = 1

subject to:
X4fl= 1

1.3.V, + I.O.V2+ 1.5.V3<210 X^A = 1

-^'
Z = 83 parts
<2
(-V2 + .V3)

jt) +2.Y3 < 160 30. Maximize Z= 120x| + 75.r2

X|^2-^3 — 3nd integer subject to:

Solution: 220x| + I40.V2 >

jci = 110
.V2 = 3
x^ = 24
Z =53.162.200
31. Minimize Z= 120x, + 75x2 + 4 SOxj CASE SOLUTION: PM COMPUTER
subject to: SERVICES
220x1 + 140x2+ 12x3 - 6,300 minimize Z= l,280(«i + «2 + "3 + «4 + «5 +
8x1 + 8x2 + -^3 ^ 256 + 120i + Vz + >'3 + >'4 + }'5+>'6) + 200(/2i +
/?6)

.4x1 + •9-^2 + 16x3.

< 15 ^2 + f^ + /24 + /!5 + h(,) + 320(/'i +/2 +/3 +/4 +
xi,X2 ^ and integer /s +/6) + 15(/i + I2 + I3 + I4 + h + ^6)
X3>0
subject to
Solution:
xi y\ - I\
+ = 63
xi = 27 X2 + y2 + /i - /2 = 74
X2 =1 -<:3 + y3 + /2 - /3 = 95
X3 = 1 8.33 X4 + y4 + 73- /4 = 57
Z= $3,397.50 X5 + >'5 + 74-/5 = 68
X6 + y6 + h-k = 86
xi < 12.7/71

X2 < 12.7«2
X3 < I2.7/J3
X4 < 12.7«4
X5< 12.7«5
X6< 12.7«6
y\ < 0.6n\
}'2 ^ 0.6^2
}'3 < 0.6«3
>'4 < 0.6«4
^ 0.6n5
)'5

< 0.6/16
y6
"1 -'Jl +/l = 5
nj- hi+fi- n\ =0
«3 - /?3 +/3 - /!2 =
«4-^4+/4-"3 =
«5-^5 +/5-"4 =
«6 - ^6 +/6 -"5 =
Xi,yj,niJi,hi/i, >

The integer solution of this model was

obtained using QSB + as follows:

184
CASE SOLUTION: THE TENNESSEE
PTERODACTYLS
Xi - player /',
/ = 1,2 12

(a) Minimize Z= $8.2j;i + 6.5^2 + 5.2jc3 + 16.4^:4

+ 1.43x5 + 23.5.V6 + 4.7JC7 + T.Ug + 15.8x9 +
26.4x10+ 19.5x11 +8.6x12

subject to

X] + X2 + X3 + X4 + X5 + X6 + X7 + xg + X9 +
^10 + -^11 +-'^12 = 5

14.7x1+ 12.6x2 + 13.5x3 + 27.1x4 +

18.1x5+ 22.8x6 + 9.3x7 + 10.2x8 + 169x9 +
28.5x10 + 24.8x11 + 11.3xi2 - 80

4.4xi + 10.6.r2 + 8.7x3 + 7.1x4 + 7.5x5 +

9.5x6 + 12.2x7 + 12.6x8 + 2.5x9 + 6.5xio +
8.6x11 + 12.5x12 ^40
9.3xi + 2.1x2 + l-7jf3 + 4.5x4 + 5.1x5 +
2.4x6 + 3.5x7 + 1.8x8 + 11.4x9 + 1.3xio +
6.9x11 +3.2x12 > 25

40.3x1 + 34.5x2 + 29.3x3 + 42.5x4 +

41.0x5 + 38.5x6 + 31.5x7 + 44.4x8 +
42.7x9 + 38.1x10 + 42.6x11 + 39.5xi2 ^ 190

xi + X3 + X4 + X5 + xio — 2

X2 + X6 + X7 + Xg + X9 + xi 1 + X12 ^ 3

X, = or 1

Solution: xi (Mack Madonna) =1 Z= $55.3 million

X2 (Darrell Boards) = 1
X4 (Ramon Dion) = 1
X7 (Hiram Grant) = 1
xii (Barry Bird) = 1

(b) The team does not have enough cash ($50

million) to sign the projected players.
Reformulating the model with an objective
function of points per game and a constraint
for salary with a constraint value of ^ $50
million, results in the following solution.

xi (Mack Madonna) = 1

X3 (Silk Curry) = 1

X8 (Antoine Roadman) = 1

X9 (Fred Westcoast) = 1

X12 (Grant Hall) = 1

Z= 66.6 points per game

187
 )

Chapter Eight

ANSWERS TO QUESTIONS 9. Computer solution

1. Linear programming has a single objective, 10. Model formulation, computer solution
while goal programming has multiple objectives. 11. Model formulation, product mix. computer
solution
2. Deviational variables reflect the
underachievement or overachievement of a 12. Model formulation, product mix, computer
goal. The negative deviational variable is
solution

analogous to a slack variable, while the 13. Model formulation, clinic personnel selection,

positive deviational variable is analogous to a computer solution

surplus variable. 14. Model formulation, production scheduling,
3. Goals are prioritized by giving them an computer solution, sensitivity analysis

objective function coefficient Pj that reflects 15. Model formulation, employee scheduling,
the rank of the goal relative to the other goals. computer solution
4. When overachievement of a goal is specifically 16. Model formulation, R&D project selection

prohibited, it is like a strict constraint in a 17. AHP, company takeover

linear programming model; and the positive 18. Pairwise comparison (8-17)
deviational variable is eliminated. Negative 19. AHP. faculty raises
deviational variables cannot be eliminated, 20. Pairwise comparisons (S-19)
since they must exist to form the basic feasible 21. AHP mutual funds
solution in the modified simplex method. 22. AHP (8-2!)
5. In a satisfactory solution all of the goals, or 23. AHP, utility vehicles
multiple objectives, might not be achieved; 24. AHP, anchor persons
while in a linear programming model the single
25. AHP, hotel selection
objective is always maximized or minimized .

26. AHP, college selection

6. In a pairwise comparison the decision maker
27. AHP, dating service
compares two alternatives (i.e., a pair) according
28. AHP R&D projects
to one criteria and indicates a preference
29. AHP, student .selection
according to a preference scale that gives a
30. AHP, athletic facilities
numerical rating to a level of performance.
31. AHP, vacation locations
7. Answers depend on student preference.
32. Pairwise comparisons (8-3 !

8. The normalized matrix assigns a numerical

33. AHP, major options
preference (i.e., an average) for each alternative
for a particular criterion. It ranks the decision
34. AHP basketball players

alternatives with a criterion.

35. AHP school facilities

9. If maker has to provide

the decision PROBLEM SOLUTIONS
preferences for a number of alternatives (i.e., 3
1. Minimize P,d;. P^dl, P^d, , P^dt
or more) over a number of criteria he or she
subject to
can lose track of previous comparisons and
5x, + 2x2 + 4x, + d^ - dl = 240
thus give inconsistent responses.
3jC| + Sjc: + 2.V, + dl - d^ = 500
PROBLEM SUMMARY Ax^ + 6x2 + 3.V, + dy - d'y = 400

1. Model formulation, product mix 2. a) Minimize Z = P, (d, + rf. + d^), P:^. . P,d>,.

P, (t/g + du^ + t/n + </,:).

2. Model formulation, transportation, computer
P.dU, P.dl
solution
subject to
3. Model formulation, urban recreation facility JTiA + -TiH + atk + .V|„ + d^ = 420
allocation -"^:a + '^iti + X:v + d~2 =610
+ -"f.x

4. Model formulation, crop determination, ^.w + Xyo J^.K ^ Xh) + (iy = 340
*"

coniputer solution A-|A + X2^ + -v,A + d: = 520

v,,) + .t:H + -V,H + d: = 250

5 Model formulation, product mix, computer
Ak + Jt;c + »,H + '^6 = 400
.solution
All) + -V:i, + .r,„ + d^ = 380
6. Model formulation, OSHA safety compliance,
xm + - d^ = 80
d^
computer solution •'^lA + X2,\ + -Via + f/y - d^ = 416
7. Computer solution; graphical solution -tiR + .V.1R + A-,n + J,„ - dlfl = 200
8. Computer solution; graphical solution Xic + -V:r + .ifH + du - du ^ 320

188
 x,i, + -tro + x^o+ dn - d^i = 304 5. a)
22x,A + 17x,B + 30jc,c + 18jt,D + 15x:a
+ 3SxiB + 20x:c + 25^20 + 28xm + 21x.-,b
+ 16x.,t + 14x.,o + djy - d*y = S24,717
X2C - rfu =

Note that the negative deviational variables must be at the

highest-priority level to force all the supply to be used.

5) x,H = 249 (249.19), x^c = 171 (170.81), x.a = 496,

jc:b = 1 (.81). X20 = 113 (113.91), jc.,c = 149 (149.19),
x,D = 191 (190.81)
 b) X,
^2

X, + Xj = 800

0. 200 400 600 800 1,0001,200 '

Solution: x, = 500
X2 = 300

xi = 500, X2 = 300, di, = 500

satisfied

9. ;c| = 15, d\ = \2,dt = 10, d^ = b

satisfied

10. a) minimize P,d^ SP.di, IPid^, P^dt

,

subject to

8j:, + 6x2 + di - d] = 480 (production capacity, hr)

X, + ^2" = 40 (demand, 100 yd)

X2 + di = 50 (demand, 100 yd)
d* + di - dt = 2Q (overtime, hr)

X,, d;, d; >

b) X| = 60, X2 = 30, di = 140, ^4 = 120

satisfied

11. a) Minimize P^d;, P.dt, 7>Pydi + 2P^d~,, P,dl

subject to

X, + X2 + - dt = 80
d;
+ dz = 60
X,
x. + d; = 35
d! + d; - dt = 10

b) JC| = 60, X2 = 30, dt = 10, d^ = 5

satisfied

12. a) Minimize P^d^, P^di, 5P,dJ + 4P,d:, P,d;

subject to
5x, + 8a:2 + d; - d! = 4,800
.20x, + .25;c2 + di - dt = 300
Jt, + di = 500
X2 + d; = 400
d* + d; - di = 480

b) The solution is. V| =431.4,X2= 135.4,

^/f = 4X0,^2 = 19.2, ^4 = 1.6. I

13. a) Minimize P^di, P.d^, P^d;, P,d;

subject to
x^ + d^ - dt = 30
20;c, + 40a:2 + 150^:, + d^ - dt = 1,200
X2 + xy + dt - dt = 20
Xi + dt - dt = 6
15. a) minimize P^d^. Pjidi. d^), Pj,{dj), P^d^, d^, 16. minimize P^dl P^. P^d^. ^4(^4. d^l P<.d(;.
d5. de) P6di
subject to subject to
.vi + .vt + -V3 + JC4 + + d\-d\ = 60
x^ + X(, + x-]
.675.V1 + 1.05.V2+ .725x3 + .43O.V4 + I.24.V5 +
X] + X4 + x^ + X(, + X-J + di- di = 47
.890^6 + 1 .620x7 + 1 -20x8 + dj- d^ = 5.0
Ai + .V4 + .V5 + X(,+ xi + d^ - d-C = 22
6x| + 5x2+ 7x3 + 8x4 + 10x5 + 6x6 + 7x7 + 6xg
X\ + JC4 + X5 + X6 + X7 + d^- dl = 28
+ ^2-^2 =27
x\ + x^ + x^ + X(, + X-] + df - dt, = 35
.820x| + 1.75x2 + '•60.V3 + 1.90x4 + -930x5 +
X + I
.V4 + .V5 + .X6 + X-I + df -d^ = 34
1.70.^6 + 1 .30x7 + I -80x8 + d^- <3 = 6.5
JC2 + JC4 + Jt5 + X(,+ X-] + d-]~ - J7 = 43
xi + X3 + X4 + X(, + t/4 -dl = 2
X3 + JC4 + X5 + X(j + x-j + di- d^ = 52)
X2 + X5 + X7 + xg + c/5 -d^ -2
X, d], df
X2 + X3 + X5 + .rg + X-; + d(,-d^ = 3
b) Employees beginning vvoric on day / = Xj
X5 + Xf, + 6^7 - ^7 = 3

.V| = - Sunday b) Solving sequentially using Excel:

.vt = - Monday
X2 =
.V3 =13- Tuesday 1

.V4 =17- Wednesday

.r5 = - Thursday
Jt5 = 30 - Friday
X7 = - Saturday

Goal achievement:
achieved
achieved
achieved
not achieved: store understaffed on
Wednesday (^5 = 5) and on Thursday (d(, = 4)

17. Normalized Matrices:

Load
Criteria
25. Normalized matrices:

Ambiance
Row
Hotel Cheraton Milton Harriott A\era"es

Cheraton . i 250 .111! . 1 304 .1222

Milton .2500 .2222 .2174 .2229
Harriott .6250 .6667 .6522 .6479

Location
Row
Hotel Cheraton Milton Harriott Averajies

Cheraton .6522 .5000 .7059 .6194

Milton .1304 .1000 .05S8 .0964
Harriott .2174 .4000 .2353 .2K42

Cost
Row
Hotel Cheraton Milton Harriott A\erai:es

Cheraton .5882 .5714 .6250 .5949

Milton .2941 .2857 .2500 .2766
Harriott . 1 1 76 .1429 . 1 250 . I 2X5

Row
Criteria Ambiance Location Cost A\erai:es

Ambiance .5714 .6000 .5000 .5571

Location .2857 .3000 .3750 .3202
Cost .1429 .1000 . 1 250 .1226

Hotel Scon
Ambiance Location Cost Criteria
Cheraton .3393
Cheraton .1222 .6194 .5949 .5571
Milton .1929
Milton .2299 .0964 .2766 .3202
Harriott .4677
Harriott .6479 .2842 .1285 .1226
1 .0000

The Harriott Hotel should be selected for the meeting.

26. NoriiKilizecl matrices:

Academic
Row
College A B C .\\erages

.3000 .2857 .3750 .3202

.6000 .5714 .5000 .5571
, 1 000 .1429 . i 250 .1226

196
 1

Location
Row
College B A\erages

A .42X6 .4545 .4055

B .4286 .4545 .5556 .4796
C .1429 .0909 .1111 . I 1 50

Cost
Row
Colleee B Averases

.1429 .1429 .1429 .1429

.2857 .2857 .2X57 .2X57
.5714 .5714 .5714 5714

Academic .279 .3333 .2459 .4()()()

Location .0698 .0833 .0984 .0667

Cost .5581 .4167 .4918 .4000
Social .0930 .1667 .1639 .1333
27. Normalized matrices:
Cost
Usage:

(1)
Fun

(1) (2) (3)

1 1/2 51
34. The solution to this problem depends on how the student develops the pairwise comparisons tor the
individual criteria and between the criteria.

35. Normalized matrices:

 School White Black Total
CASE SOLUTION: OAKDALE COUNTY
North
SCHOOL BUSSING
* 1. X,, = no. of white students bused from district i

(/ = n,s, e,w) to district j j = n, s, e,w); y,i = no. of

{

black students bused from district (( = n, s, e, w) to dis- /

trict j ( j = n,s, e,w).

minimize Pi{di ,dT,d^ ,d4,df,dii,dT ,di^), P2{d^,dn„

du, dti), Pfdti, Pi(dii, dti, du„ dn)
subject to

JCnn+ X„, + X„c + X„^ + d; = 1,000

>'r,n + yn. + ^nc + ynw + ^2' = 300
^sn + X,, + X,, + x,^ + d; = 450
ysn + y.s + ys, + y.w + d; = 800
JC„ + Ales + JCec + ^e» + d^ = 1,050

>'en + y.. + ^cc + >'cw + ^6 = 400

ATvvn + X^, + X^, + X^^ + dl = 500

y^n + ywv + ywc + yww + d^ = 500
.4jCnn + •4jr,„ + •4jCr„ +

.4x,^n — .6y„„ — .by^„

~
•6yen - •6ywn + ^M - d^ =
.4jc„s .4jrss +
-I- ^JCcs +

.4xws ~ •6}'„s ~ (>y>,s

~
6y,s - -by^s + d'w - rfu, =
.4xnc + .4a:,c + Axcc +

•4jtwc — •63'„c — .6_y,c

—
6y„ - 6y»c + ^n - tin =
.4Xnw + •4x,w + •4jtcw +
Ax^y, — .6>'nw ~ (^ysv ~
.6>'ew - .6y>,« + dn - dn =

()j«:„„ + Oy„„ + 3()jc„, + 30>'„, + \lx^, + 12y„,

+ 20a-„, + 20y„„ + 30x,„ + 30y,„ + Ojc„
+ O.Vs. + ISjfs, + ISy,, + 26x,„ + Iby,^
+ 12jt:,„ + 12>\„ + 18jc,, + 18y,. + Oa,,
+ Oy,, + 24jc,„ + 24y,„ + 20jc„„ + 20y„„
+ Ibx^^ + 26>'„, + 24a;„, + 24y„, + Qx^^
+ 0>\., + ^, - dn = 30,000
.

X,m + X^n + X^„ + X„„ + y„„ +

ys. + >v„ + y-n + d;, - ^,'4 = 1,360

Jf„s + ^Nv + -tcx + J^ws + ynv +

y.. + y.s + y*. + - ^r? = 1,140

<ir>;

X„c + Jfsc + -ffc + ^wt + ync +

ysc + y.c + ywc + d;,. - dl,, = 1,1 40

JC,m + X^^ + jr,.w + Xww + ynw +

ysw + y.w + yww + d^-, - dt-, = 1,360

The first 8 constraints at the first priority ensure that all

students will be allocated to a school. The second 4 con-

straints reflect the first goal of achieving no more than
60% white students at each school. The 13th constraint
reflects the goal for busing mileage, and the last 4 con-
straints are the goals for avoiding overcrowding. Notice
that each school's normal capacity has been increased by
(approximately) 14%, the percentage by which the total

number of students (5,000) exceeds the total school ca-

pacity (4,400).

2. x„„ = 734 (733.73), x„, = 266 (266.27), x,^ = 418

(417.73), x.^ = 32 (32.27), x,„ = 82 (82.27), ;t,, = 684,
jc,, = 284 (283.73), x^^ = 500; y„„ = 300, y,, = 456,
y., = 56, y., = 288, y,, = 400, y„„ = 244, y^„ = 256;
all goals are achieved.
 Chapter Nine

ANSWERS TO QUESTIONS
1. The optimal volume is determined in profit

analysis by finding the derivative of the

nonlinear profit function, setting it equal to
zero and solving for the optimal price. This
value for price is then substituted into the
linear equation for volume.

2. A mathematical programming model in which

there is an objective function to be optimized
and one or more constraints

3. In the substitution method the constraint

equation is solved for one variable in terms of
another. This expression is then substituted into
the objective function, which eliminates the
constraint and creates a nonlinear objective
function. This function is differentiated, set
equal to zero, and solved for the remaining
variable value.

4. In the Lagrangian method, the constraints are

treated as multiples of the Lagrange multiplier,
X, and subtracted from the original objective
function. This altered version of the objective
function is the Lagrangian function.

5. The Lagrangian multiplier reflects the

approximate change in the objective function

resulting from a unit change in the quantity

value of the constraint equation. As such, it is

analogous to a dual variable in a linear

programming problem.

PROBLEM SUMMARY
 Z = 93,460p - l,153.8p^ - 1,350,000 30/7 = 1,055
= 93,460(40.5) - 1,153.8(40.5)^ - 1,350,000 p = $35.16
= $3,785,130 - 3,242,520 = $542,610 Z = 1,055/7 - 15/7^ - 25,600
= 1,055(35.16) - 15(35.16)' - 25,600
4. V = 17,000 - 5,666p, c, = 8,000, Cv = .35,
= 37,093.80 - 26,836.23 = $10,257.57
Z = vp - Cf — vCv
= (17,000 - - - 8. Maximize Z = lx\ — .3jC|' + 8x2 - Axi
5,666p)p c, (17,000
subject to
- 5,666/7)Cv
= - - - 4xi + 5^2 = 100
17,000p 5,666p' Cf 17,000cv
Solve the constraint for JC|:
+ 5,666/7Cv = 18,983p - 5,666p^
- 13,950 4;ci = 100 - 5jc2
dZ x, = 25 - 1.25;c2
—
dp
= 18,983 - ll,332p
^ Substituting,

= 18,983 - ll,332p Z = 7(25 - 1.25^:2) - .3(25 - 1.25;t2)^

ll,332p = 18,983 + 8;t2 - .4x2'

p = $1.68 = 175 - 8.75x2 - .3(625 - 62.5x2

V = 17,000 - 5,666/7 = 17,000 - 5,666(1.68) + 1.56x2^) + 8x2 - .4x2'
= 7,481 yd = 175 - 8.75x2 - 187.5 + 18.75x2
Z = 18,983p - 5,666/7^ - 13,950 - .468x2^ + 8x2 - .4x2'
= 18,983(1.68) - 5,666(1.68)^ - 13,950 = 18x2 - 12.5 - .868x2^
= $31,891.44 - 15,991.72 - 13,950 dZ
= $1,949.72 = 18 - 1.74x2
ax2
5. c, = $2,500, Cv = $9, V = 200 - 4.75/7, = 18 - 1.74x2
Z = V/7 - Cf - VCv =
1.74x2 18
= (200 - 4.75/>)p - Cf - (200 - 4.75/p)Cv X2 = 10.34
= 200/7 - 4.75/7^ - 2,500 - 200(9) + 4.75/7(9) 4xi + 5x2 = 100
= 242.75/7 - 4.75/7^ - 4,300 4x, + 5(10.34) = 100
4x, = 48.3
—
bp
= 242.75 - 9.5p X, = 12.08
Z = 7(12.08) - .3(12.08)^ + 8(10.34)
= 242.75 - 9.5/7 - .4(10.34)^
9.5/7 = 242.75 = - 43.78 + 82.72 -
84.56 42.77
p = $26.97 = 167.28 - 86.55 = $80.73
V = 200 - 4.75p 200 =
4.75(26.97) - = 72
Z = 242.75p - 4.75/7^ - 4,300
= 242.75(26.97) - 4.75(26.97)' - 4,300 9. Maximize Z =
= $6,546.97 - 3,455.06 - 4,300 = -$1,208.09
(loss)

6. Z= vp - Cf- vcy
= (4,000 - SOp)p - 25,000 - (4,000 - 80/7)10
= 4,000/7 - 80/72 _ 25,000 - 40,000 + 800/7
= 4,800/7 - 80/72 _ 65,000

dZ= 4,800- 160/7 =

dp
160/7 = 4,800
/7 = 30

V = 4,000 - 80(30)
= 1,600
Z = 4,800(30) - 80(30)2 - 65,000
= $7,000

1. Z = vp - 12,000 - 17v; v = 800 - 15/7; substituting,

Z = - \5p)p - 12,000 - 17(800 - 15p)

(800
= 800/7 - 15/7^ - 12,000 - 13,600 + 255p
= l,055p - 15/7^ - 25,600
37
— = = 1,055 - 30p
dp
10. Maximize Z = IOjc, - .{)2x,^ + 12^2 - .03x2^ dL
subject to
:— = 25 - 1x2 - 6A = (2)
dX2
.2jci + .1x2 = 40
Let .2jti = 40 - .1x2 and x, = 200 - .Sxj. —
dL
oA
= -3x, - 6x2 + 300 = (3)
Z = - .5x2) - .02(200 -
10(200 .5x2^
+ 12X2 - .03X2^ Multiply (1) by 2 and subtract (2) from it:

= 2,000 - 5x2 - .02(40.000 - 200x2 60 - 8x, - 6A =

+ .25x2^ + 12x2 - .03x2^ -25 + 1x2 + 6A =
= 2,000 - 5x2 - 800 + 4x2 - .005x2^ 35 - 8x, + 1x2 =0 (4)
+ 12x2 - -03X2^
Multiply (4) by 6 and add (3):
= 1,200 + 11x2 + .035x2^
-3xi - 6x2 = -300
—=
dZ
= 11- .07x2 -48xi + 6x2 = -210
-51x, = -515
= .07x2 11
X, = 10
X2 = 157
3x, + 6x2 - 300 =
.2x, + .1x2 = 40
3(10) + 6x2 = 300
.2x, + .1(157) = 40
6x2 = 270
.2x, = 24.3
X2 = 45
X, = 121.5
Z = $212.50
Z= lOx, - .02x,^ + 12x2 - -03x2^
= 10(121.5) - .02(121.5)^ + 12(157) 13. Maximize Z = lOx, - .02x,2 + 12x2 - 03x2^
- .03(157)2 subject to
= 1,215 - 295.25 + 1,884 - 739.47 .2x, + .1x2 = 40
= 3,099 - 1,034.72 = $2,064.28 Then .2x, + .1x2 - 40 = and L = lOx, - .02x,2 +
12x2 - 03x22 - A(.2x, + .1x2 - 40).
11. Maximize Z = 7xi - .3x1^ + 8x2 - .4x2^
dL
subject to -— = 10 - .04x, - .2A = (1)
4x, + 5x2 = 100 ax,
- 100 = = - +
Then, 4x,
8x2 - .4x2^
+ 5x2
- A(4x, + 5x2 -
and I
100).
7x, .3x1^
—=
dL
3X2
12 - .06x2 - .lA = (2)

—=
dL
ax,
7 - .6x, - 4A = (1)
dL
= -.2x, - .1x2 + 40 = (3)
dX

—=
dL
dX2
8 - .8x2 - 5A = (2)
Multiply (2) by 2 and subtract from (1):

10 - .04x, - .2A =
—
dL
= - = -24 + .12x2 + .2A =
-4x, 5X2 + 100 (3)
dA -14 - .04x, + .12x2 = (4)
Divide equation (2) by 1.25 and subtract from equa-
Multiply (3) by .2 and subtract (4):
tion (1):

- 4A = -.04x, = -8
- .02x2
7 - .6x,
.Q4xi .12x2 = -14
-
-6.4 + .64x2 + 4A =
-.14x2 = -22
.6 - .6x1 + .64x2 =
X2 = 157

Next, multiply equation (3) by .15 and subtract (4) from it:
.2x, + .1(157) = 40
.2x, = 24.3
-.6x, = -15 - .75x2
X, = 121.5
+ .6xi .64x; = + -
.6
Z = $2,064.28
-1.39x2 = -14.4
14. Maximize Z = $25xi - .8x1^ + 30x2 - 1-2x22
X2 = 10.35
subject to
4x, + 5(10.35) - 100 =
X, + 2x2 = 40
4x, = 48.25
Let x, + 2x2 - 40 = and L = 25x, - .8x,2 +
X, = 12.06
30x2 - 1.2x22 - A(x, + 2x2 - 40).
Z = $80.73
12. Maximize Z = 30x, - 2x,2 + 25x2 - .5x2^ —= dL
dx,
25 - 1.6x, - A = (1)
subject to

Then 3x,
3x, +
+
6x2
6x2
= 300
- 300 = and L = 30x, - 2x,2 +
—= dL
3X2
30 - 2.4x2 - 2A = (2)

25x2 - .5x2^ - A(3x, + 6x2 - 300).

—
dL
= -X, - 2x2 + 40 =
—=
dL
ax,
30 - 4x, - 3A = (1)
aA
(3)

207
 Multiply (1) by 2 and subtract (2):

50 - 3.2x, - 2A = 18. Minimize Z= ( .24 + — .15

-^1
\

/
\ +
/
i.31
\
+ — .21

X2
) +
-30 + 2.4x2 + 2A =

— + —
20 - 3.2x, + 2.4x2 = (4) .12 \ / .30
1.21 + + .48
Multiply (3) by 3.2 and subtract (4): X3 X4
; \
-3.2x, - = -128
6.4jt2 Subject to:

3.2xi 2.4x2 =
- 20 X] + X2 + X3 + X4 < 20
-8.8x2 = -108
X2 = 12.3
X, + 2x2 = 40
4+ —< 10

xi + 2(12.3) = 40
8
X, = 15.4 8 + —< 10
Z = 25(15.4) - .8(15.4)^ + 30(12.3)
Xt
-
= 385
1.2(12.3)^
- 190 + 369 - 182 = $382 6+ —<
10

^3
10

15. Using equation (1) in Problem 14:

3 + —9 < 10

3X|
= 25 - 1.6x, - A = M
25 - 1.6(15.454) - A =
A = .27 Solution:

Therefore, we would be willing to pay $.27 for one addi- xi = 5

tional hour or labor. X2 = 5
X4 = 4
16. Maximize
X5 = 6
Z= 1 .45 crimes per 1 ,000 population
9,000 15,000
Z= 15,000 + 24,000
X2
•^1
CASE SOLUTION: ADMISSIONS AT
5,300 \ /
-
7,600 STATE UNIVERSITY
+ ( 8,100 1 + ( 12,000
-^3 X4 Maximize Z = + X2/2
xi^i
Subject to:
12,500
+ 21,000
^5 xi + X2 ^ 14,000 freshmen
subject to X] + X2 S: 700 freshmen

x\ + X2 + X3 + X4 + X5 < 15 .55x1 + -72x2 ^ 800 dorm rooms

355x| + 540x2 + 290x3 + 275x4 + 490x5 ^ 6,500 960x, + 1,150x2
X, >
solution:

X, =3
X2 =4
X3 = 2
X4 = 3
X5 = 3
Z = $64,000

17. Maximize Z= - .06/?;, + .005pi)p,, +

(5.8 (3.0
- Mpi + .OOSp,,)pi
Subject to:

5.8 - .06;?/, + .005/7/ < 2.5

3.0-.ll/7/ + .008/j/,<2.5
PhPh ^
Solution:

Pi,
= $56.40 per kwh
Pl = $16.70 per kwh
Z= $167.94
 Chapter Ten

ANSWERS TO QUESTIONS d) Repeat steps a. b. and c until there are no more

paths with available tlow capacity.
1. The shortest route problem is to determine the
shortest distance between an originating point 7. This is the means used to compute net tlow
and several destination points. The minimal along a branch.
spanning tree problem is to connect all nodes
the network so that the total branch lengths
in
PROBLEM SUMMARY
are minimized.
1. Shortest route
2. The steps of the shortest route solution method
are as follows: 2. Shortest route

3. Shortest route
a) Select the node with the shortest direct route
from the origin . 4. Shortest route

b) Establish a permanent set with the origin node 5. Shortest route

and the node selected in step a.
6. Shortest route
c) Determine all nodes directly connected to the
7. Shortest route
permanent set nodes.
8. Shortest route
d) Select the node with the shortest branch from
9. Shortest route
the group of nodes directly connected to the
permanent set nodes. 10. Shortest route

e) Repeat steps c and d until all nodes have joined 11. Shortest route
the permanent set.
12. Shortest route
3. The steps of the minimal spanning tree solution
13. Shortest route
method are as follows.
14. Minimal spanning tree
a) Select any starting node.
15. Minimal spanning tree
b) Select the closest node to the starting node to
16. Minimal spanning tree
join the spanning tree.
17. Minimal spanning tree
c) Select the closest node not presently in the
spanning tree. 18. Minimal spanning tree

d) Repeat step c until all nodes have joined the 19. Minimal spanning tree

spanning tree.
20. Minimal spanning tree

4. A network is popular because it provides a 21. Minimal spanning tree

picture of the system under analysis and
22. Minimal spanning tree
because a large number of reallife systems can
be easily modeled as networks. 23. Minimal spanning tree

5. A directed branch will have no flow at one 24. Minimal spanning tree

node and flow at the other, while an undirected 25. Maximal tlow
branch will show flow at the nodes at both
ends of the branch.
26. Maximal flow
27. Maximal tlow
6. The steps of the maximal flow solution method
are as follows: 28. Maximal tlow

a) Arbitrarily select any path in the network from 29 Maximal tlow

origin to destination
Maximal tlow
.

30.

b) Adjust the capacities at each node by

31. Maximal tlow
subtracting the maximal flow for the path
selected in step a.
32. Maximal tlow

33. Maximal tlow

c) Add the maximal flow along the path to the
flow in the opposite direction at each node. 34. Maximal tlow

209
PROBLEM SOLUTIONS
1.

Permanent
Set Branch Distance

{1} 1-2 21
1-4 30
*
1-3 I 17 I

Permanent
Set Branch Distance
*
{1,3} 1-2 I
21 \

Set
The shortest route network;
 1

The shortest route network:

Branch Distance
{1, 2, 3} 2-5 116
3-5 114
3-7 170
3-4 118
1-4 I 88 *

Permanent
Set Branch Distance

{1} 1-2 85
1-3 *
I 53 \

1-4 88

Permanent
Set Branch Distance
{1,2,3,4} 2-5 116
3-5 rrm *

3-7 170
4-6 225

Permanent
Set Branch Distance
{1, 3} 1-2 fssl *
1-4 88
3-5 114
3-7 170
3-4 118

Permanent
Set Branch Distance
{1, 2, 3, 4, 5} 5-7 186
3-7 *
flTOl
4-6 225

212
 1

Permanent
80
Set Branch Distance
-
1-3
o-^^^-o {1,2}
1-4
I
13()|
140
2-5 ZOO
80

130

^ 130
Permanent
Set Branch Distance

{1.2.3,4,5,7} 4-6 225 80

*
7-6 194

O-^-O
1

The shortest route network:

80
140
210
140

40
i

i /
/
130

50
130

Permanent
Set Branch Distance

{1, 2. 3} 1-4
2-5 200
3-4 170
Permanent 3-6 180
80 Branch Distance
Set 3-7 340
{1} 1-2 noi *

1-3 130
1-4 140
80

0*
140

^
130

213
 1

Permanent Permanent
Set Branch Distance Set Branch Distance
2-5 *
{1,2,3,4} 200 {1, 2, 3, 4, 5, 6} 3-7 1
340 1

*
3-6 1
180 1
5-7 350
3-7 340 6-7 370
4-5 190
The shortest route network:

Permanent
Set Branch Distance

{1,2, 3, 4, 6} 2-5 200

3-7 340
*
4-5 1 190
6-7 370

214
Permanent
Permanent
 Permanent Permanent
Branch Time Set Branch Time
Set
5-8 75 (1, 2, 3, 4, 5, 6, 7, 8, 9} 5-10 \jT\'
{1, 2, 3, 4, 5, 6. 7}
7-8 72 8-10 8')

'
7-9 I 61 I

The shortest route network;

5-10 72

60

43
 1

Permanent
Set Branch Distance

{1, 2, 3, 4, 6, 7} 2-5 420 Branch Distance

6-5 437
{1, 2, 3, 4, 5, 6. 7, 9} 5-8
4-10 630
6-8 495
*
6-9 1 391
7-10 539
7-9 501

Permanent
Set Branch Distance
*
{1, 2, 3, 4, 6, 7, 9} 2-5 |420|
6-5 437
6-8 495
9-8 584
4-10 630
7-10 539
The shortest route network:
m
Permanent
Set Branch
Permanent
 7.5

4.5

3.5

H
0-
 17.

Communication system = 2,700 yd

16. 18.

160
Length of pipe = 25 miles
Length of route = 320 >d

19.

Length of pipe = 2,400 miles

20.

Length of path = 1.160 yd

225
 210

160

1 90 ^F 230
Total sidewalk = 1,086 ft.
Length of ductwork = 1,060 ft

Length of path = 22 miles

Minimal spanning distance = 41

226
 Maximal flow network:

16 •16

26.

227
 17

17

22

22

Maximal flow network.

-22

22

27.

228
31 •31

40' -40

43' 43

229
 Miiximal flow network:

19' 16,

15
H 00
10
27
20
20 ^^3 •47
12
12

12 11

28.

0.^.4
5 Oi
;^9

e° /o
\
10-
7^ 5,^%'

15-
17-

20- -20

21 -21

Maximal flow network:

21 21

231
29. 9,

IS

IS-

17

18

232
 Maximal flow network:

18

18

30.

10'

•10

233
 Maximal flow network:

10-

10

31.

Allocation
Branch in Which Total
Step Path Flow Amount Capacity Is Used
1-2-5-7-9 4 2-5
1-3-5-7-9 2 3-5
1-3-6-8-9 2 1-3, 3-6
1-4-6-8-9 4 4-6, 6-8, 8-9
Maximum flow = 12,000 cars

12- 12

17

234
 34. 1-2-6-10-12-13-15 = 25
Allocation
1-2-6-12-13-15 = 35
Step Path Flow Amount 1-2-9-12-15= 10
1-2-5-7-10 1-3-6-10-12-15 = 30
1-2-5-8-10 1-3-7-10-13-15 = 20
1-3-5-9-10
1^^7-10-13-15 = 20
1-3-5-2-6-8-10
1-^1-7-6-10-13-15= 10
1-3-5-2-6-9-10
1-3-6-^10 l^t-7-6-12-15 = 5
= 1-4-8-13-15 = 25
Maximum Flow 17,000 units
1-5-8-13-15 = 35
1-5-8-14-15 = 5
1-5-11-14-15 = 30
33.
maximum flow = 250

32- 32

Allocation
*CASE SOLUTION: THE PEARLSBURG
RESCUE SQUAD
The network for the Pearlsburg Rescue Squad follows.
 ) 85

CASE SOLUTION: BATTLE OF THE

BULGE
I) Verdun-(2)Stenay = 16
1 Verdun - (3) Montmedy - 23
)

1) Verdun -(5) Etain= 18

2) Stenay - (3) Montmedy = 9

2)Stenay-(ll)Bouillon = 7
3) Montmedy - (4) Longuyon = 7
3) Montmedy - (6) Virton = 10
3) Montmedy -(11) Bouillon = 1

4) Longuyon - (5) Etain = 6

4) Longuyon - (7) Longwy = 1

5)Etain-(8)Briey= 15
5) Etain - (9) Havange = 9
6) Virton - (7) Longwy = 4
6) Virton -(13) Tintigny = 6
7) Longwy - (9) Havange = 1

7) Longwy - (14) Arlon = 4

8)Briey-(10)Thionville= 15
9) Havange - 10) Thionville = 2
(

9) Havange - 15) Luxembourg = 8

(

10) Thionville - (15) Luxembourg = 17

1 Bouillon -(12) Florenville = 10
1

11) Bouillon- (16) Paliseul = 12

12) Florenville - (13) Tintigny = 8

2) Florenville -(16) Paliseul = 2

1

13) Tintigny - (17) Neufchateau = 14

14) Arlon -(18) Martelange = 8
15) Luxembourg - (19) Diekirch - 21

5) Luxembourg - ( 14) Arlon = 4

1

16) Paliseul - (20) Recogne = 14

17) Neufchateau - (18) Martelange = 12

7) Neufchateau - (20) Recogne = 2

1

18) Martelange - (21) Bastogne = 25

19) Diekirch - (2 1) Bastogne = 1

19) Diekirch - (18) Martelange = 3

20) Recogne - (21) Bastogne = 16

Total flow = 57.000 troops

237
 Chapter Eleven

ANSWERS TO QUESTIONS 3. Network construction and analysis

1. PERT/CPM is popular because it provides a 4. Network analysis

picture of the steps of a project, it is easy for
5. Network, earliest and latest event times, slack
managers and participants to understand, and,
(Problem 11-4)
it is easy to apply.
6. Network, earliest and latest event times, slack
2. To show the precedence relationship of
activities in a project 7. Network, earliest and latest event times, slack

3. A dummy activity is used to show a 8. Network, earliest and latest event times, slack

precedence relationship without the passage of 9. Network, earliest and latest event times, slack
time. It is used most frequently to complete a
precedence relationship so that two activities 10. Network construction and analysis
will not have the same start and end nodes.
Network analysis
11.

4. The critical path is the longest path in the

12. Network analysis
network. It can be computed by summing the
activity times along each path and then seeing 13. Network analysis (Problem 11-6)
which path is the longest. It also is the path
14. Network analysis (Problem 11-7)
with no slack available.
15. Probability analysis (Problem 11-13)
5. Slack is the amount of time an activity can be

delayed without affecting the overall project 16. Network analysis, probability analysis

duration. It is computed by subtracting the

17. Network analysis, probability analysis
earliest time at node / and the activity time for

activity / - / from the latest time of nodej. 18. Network analysis, probability analysis

6. Slack shared between adjacent activities such 19. Probability analysis (Problem 11-14)

that the two activities can be delayed jointly by

20. Network analysis, probability analysis
the amount of shared slack
21. Network construction, probability analysis
7. The mean activity time is computed as
t = {a + 4m + b)l 6, where a is the optimistic
22. Network construction, probability analysis
activity time, m is the most likely time, and b
23. Project crashing, linear programming model
is the pessimistic time. The variance is
formulation
computed as v = [{b - ayb]".
24. Project crashing, linear programming model
8. Total project variance is computed by summing formulation
the variances of the critical path activities.
25. General linear programming model formulation
9. The purpose of project crashing is to shorten
(Problem 11-4)
the project duration at the least possible cost.
26. General linear programming model formulation
10. See which activity on the critical path has the
(Figure 16.12)
minimum crash cost, and reduce this activity
duration by the maximum amount or until 27. Project crashing, linear programming model
another path becomes critical. If more than one formulation

path is critical, both paths must be reduced by

28. Project crashing, computer (Problem 11-12)
the same amount simultaneously. Repeat this

process until the crashing objective is reached.

29. Project crashing, computer (Problem 11-6)

PROBLEM SUMMARY
1. Network construction and analysis

2. Network construction and analysis

238
 PROBLEM SOLUTIONS 4. Paths: 1 ^ 3 -» 5 -* 6 = 10+ 4 + 2 = 16; 1^3^
4-*5-*6= 10 + 5 + 3+ 2 = 20*, critical path;
l-*2-^4-» 5 -^6 = 7 + 6 + 3 + 2= 18
1. Paths: 1-^2^4 = 5 + 3 = 8; 1-^3^4
4 + 6 = 10*; path 1 ^3^4 is critical.

5. The critical path activities have no slack.

2. Paths: 1^2->4^6 = 3 + 2+l=6; 1^3

4^6 = 7 + 5 + 1 = 13; 1^3 ^5
6 = 7 + 6 + 4=1 7* critical path

3. Paths: 1-^2^4^ 6, 4 + 8 + 2=14; 1-^2-* Activity

4->5^ 6, 4 + 8 + 5 + 6 = 23*; 1 -» 2 ^ 5 -^ 6,

4 + 3 + 6=13; 1^3^5^6, 7 + 9 + 6 = 22;

1^3^ 6, 7 + 5= 12
Activity Time ES EF LS LF Slack

i
-2
Activity
 S = 16
ET = 3 ET = 5
LT =12 LT =21

Activity Time ES EF LS LF Slack

1
-2
10.

Activity
 11.

Activity a ES EF LS LF Slack

1
-2
Activity
Activity
17.
Activity
 20

^^s = 7 50 ^i^

Activity
 Time (days)
u = 57.33 67
a = 5.77

Expected completion time = 57.33

a= 5.77

Z- 1^
a
= 5Zz2Zi! = ,68
5.77

P{x < 67) = .9535

Critical path = a-b-d-j-n-o

Expected project completion time = 45 weeks

= 4.10

Since probability .90. Z= 1.29

.V - 45
1.29 =
4.10

,v- 45 = 5.29

A- = 50.3

To be 907c certain of delivering the part on

time. RusTech should probably specify at least
50.3 weeks in the contract bid.

251
22.

The activity schedule, critical path, expected

project time and variance determined using

QSB + as follows
 Critical path for PM Computers 23.

Critical path: a -=> c ==> d ==> h ==> 1 ==> o ==> q.

Variance on this path = 367.6667

Done.

QSB+ will also determine the probability that

the marketing program will be completed
within 6 months, or 180 days, as follows.

Probability Analysis for PM Computers

The following probability calculation assumes

that activities are independent and so are all

paths. It also assumes that your network has a

large enough number of activities to assume
the normality. Therefore, when the activities are
not independent or the number of activities is
not large, the analysis may be biased.

Expected completion time (507c chance)

= 135.333

When do you expect to finish your project

(Otoend)'^ 180

Critical path: a ==> c ==> d ==> h ==> 1 ==> o ==> q.

Variance on this path = 367.6667

The probability to fmish in 180 is 99.00719%.

Do you want to enter another scheduled completion

time (Y/N)? n
(1) To manually crash, the first step is to select the d) The total cost of
activity on the critical path with the minimum
crash cost, which is activity 1 —* 2. Reduce this
activity to the point where another path becomes e)
critical. Activity 1 ^ 2 can be reduced by 5 weeks
before path 1
— 3 ^ 4 also becomes Our critical.
revised network is shown below. Crashing cost =
$200; critical paths: 1 ^ 2 ^ 4 and 1-^3-^4;
critical path time = 25.

$80
10(6)

(2) Next, both critical paths must be reduced simulta-

neously. The minimum crash cost activities on the
two critical paths are activity 1 -^ 2 and activity

3^4. The most these activities can be crashed is

/ week before the path 1 —> 4 becomes critical.

The revised network with these two activities re-

duced by one week is shown below. Crashing
cost = 30 + 40 = $70; critical paths: 1^2-^4,
1
—* 4, 1 -^ 3 —» 4; critical path time = 24.
24. a)

$80
10(6)

(3) Now all three paths are critical, and we want to b)

crash to 20 weeks, the maximum amount. The
minimum crash cost activities on each path are
1 -^ 2, 1 -^ 4, and 3 —» 4. Crashing each of these
activities by 4 weeks results in the following re-

vised network. Crashing cost = $160 + 200 +

120 = $480; critical paths: 1 ^2^ 4, 1^4,
1 —» 3 —» 4; critical path time = 20.
 $100 $100

c) Minimize Z == x,.
(1) To manually crash, the first step is to select the
subject to
activity on the critical path with the minimum
crash cost, which is activity 1 — > 2. This activity JT, - X, > 16

can be reduced to the point where another path JTj - x, > 14

becomes critical. Activity 1 —* 2 can be reduced Jf4 - X: > 8

3 weeks before path 1 ^ 3 —* 5 — * 6 becomes X; - x, > 5

critical. (Notice that reducing activity 1 -^ 2 re-

Xj — X, > 4

duces not only the critical path but also 1 -^ x^ - X, > 6

2 ^4^ 6. thus keeping it from becoming criti- Xb - X4 > 10

cal.) The revised network is shown below. Crash- X6 - X5 > 15

ing cost = S900; critical paths: 1-^2^5-^ x„ X, >

6, 1 ^3^5— » 6: critical path time = 33.
d) The minimum project duration is 22 weeks.
Minimize Z = 300y,: + 160y,3 -t- lOOy^^ + TOOyis

$100 -I- 750y,, + 400>'36 + 500y^

+ 600^5^
subject to
$300 y,, <8
13(8)
yn ^5
yu ^ 2
y25 ^ 1

>'35 ^ 2
^36 ^ 2
>-« ^3
y«,^5
X| -t- 16 -
< X: v'l:

X, -t- 14 -
< X:. >',,

(2) Next, both critical paths must be reduced simulta- X: + 8 - >'24 S X.

neously. The minimum crash cost activities on X: + 5 - vz! ^ X5
these two paths are 1 ^ 2 and 1^3. The most Xj + 4 — ^35 < X5
these activities can be crashed is 5 weeks, which X} + 6 - yik < Xft
results in the desired crashed network time of X4 10 - y^fi < Xft
-(-

28 weeks. The revised network with these two ac- X5 + 15 - >'5* < Xf,
tivities reduced by five weeks is shown below. x*<22
Crashing cost = $1,500 -t- S800 = S2,300; criti- ^< ^p y-i -
cal paths: 1^2-^5^ 6, 1-^3-^5-^6;
critical path time = 28. The total cost of crashing
is $3,200 ($900 + 52,300 = 53,200).

255
25. Minimize Z = Xt,

26.

27.

b)
 ET = 9.67
LT = 12.17

Hotel arrangements m
5.50
ET = 4.17 s = 2.50
s = 2.50
LT =4.17
Hire
Union stagehands
negotiations
s =
4.17

s = 250

ET
 Chapter Twelve

ANSWERS TO QUESTIONS 9. Additional information, if available, can

sometimes enable one to alter the marginal
1. An objective probability of a specific outcome
probabilities of the occurrence of an event.
is the ratio of the number of specific outcomes
to the total number of outcomes, whereas 10. The normal distribution is a continuous
subjective probability is not based on past probability distribution in the shape of a
evidence but is based on personal knowledge symmetrical bell. The expected value is an
of the situation. There are numerous possible average computed by multiplying each \alue of
examples of each type. a di.screte random \ariable by its probability of
occurrence.
2. a) Subjective

b) Subjective

c) Subjective and objective PROBLEM SUMMARY

d) Subjective
1. Determining event probabilities
e) Objective
2. Determining event probabilities
f) Subjective

g) Objective 3. Binomial distribution

h) Objective
4. Binomial distribution

3. a) 12/52 5. Binomial distribution

b) 4/52
6. Binomial distribution
c) 13/52 = 1/4

d) 1/52 7. Binomial distribution

8. Conditional probability
4. a) Rolling the dice, b) 1 1 events: a number
between 2 and 12 inclusive 9. Probability tree and marginal probability

5. With deterministic techniques the assumption 10. Probability tree

is that conditions of complete certainty exist,
11. Probability tree and marginal probability
while a probabilistic technique reflects
uncertain information and results in uncertain 12. Probability tree and marginal probability
solutions. The mathematical programming
techniques in Chapters 2 through 10 are 13. Baye's rule
generally deterministic, while the techniques
14. Baye's rule
in Chapters 1 1 through 16 are probabilistic.
15. Expected \alue
6. An experiment is an activity that results in one
of several possible outcomes. An event is an
16. Expected value
outcome of the experiment.
17. Expected value

7. If the occurrence of one event does not affect 18. Decisions and expected value
the probability of another, the events are
19. Decisions and expected value
independent; but if one does affect the other,
the events are dependent. 20. Expected value

8. The properties of a Bernoulli process are as 21. Expected value, cumulative distribution
follows:
22. Normal distribution
a) There are two possible outcomes for each trial.
23. Normal distribution
b) The probability of the outcome remains
constant over time. 24. Normal distribution

c) The outcomes of the trial are independent. 25. Normal distribution

d) The number of trials is discrete and inteser.

259
26. Normal distribution (CPM/PERT)

27. Normal distribution (inventory)

28. Normal distribution

29. Normal distribution (break-even analysis)

30. Normal distribution

31. Normal distribution

32. Normal (sampling) distribution

33. Normal (sampling) distribution

34. Normal (sampling) distribution

35. Chi-square test

36. Chi-square test

PROBLEM SOLUTIONS
1. a) Snowfall Probability

0-19
10.

P(D) =
 17. £(lead time) = 1(.2) + 2(.5) + 3(.2) + 4(.l) = .2 +
1.0 + .6 + .4 = 2.2 days
X

18.

19.

20.

21.
 670 u = 805
a =207
z = = 1.25; P(x > 19) = .1056; this is
4 4 X - M _ 670 - 805 _ 135
Z = -.50
the type of probabilistic analysis done in Chapter 21 on a ~ 207 ~ 207
CPM/PERT. From Table A.l: Z = .50; P = .1915; P(.r > 670) =

.1915 + .5000 = .6915

30. ^1 = 63
CT= 10
.52 = xg - 63 (for an area equal to .20. Z= .52)
10~"

5.2=^-63
Xg = 68.2
1 .28 = .v;i - 63 (for an area equal to .40. Z= 1 .28)
u = 4,500 6,000
10
a = 900
12.8=.r,, -63
6.000 - 4,500 1,500 xa = 75.8
Z = 1.67; P{x > 6,000) =
900 900 jcc = 63-5.2 = 57.8
.0475; this is the type of probabilistic analysis done in .ro = 63- 12.8 = 50.2
Chapter 19 on inventory control.
An "A" is a grade greater than or equal to 75.8
A "B" is a grade between 68.2 and 75.8
A "C" is a grade between 57.8 and 68.2
A "D" is a grade between 50.2 and 57.8
An "F' is a grade less than 50.2

31. ^= I 050
a= 120
Z= 1200- 1050
120
= 1.25

X - u. X - 180
P(x< 1200) = .3944
Z = = .3944 X 620 = 244.5
o- 60
- _ 207
A probability of .4000 has a Z value of 1.28: 32.
10
X - 180
1.28 =
60 (Z..,)=
76.8 = X - 180 S-vr-
X = 256.8
(= I 4569 - 4284.9
//- 1

or approximately 257 recorders must be ordered to meet

90% of customer demand. .v2 = 31.57

j = V^=V31.57 = 5.61

25 - 20.7
Z=
5.61

= .766

263
 P(x>25) = .279
The goal is P(x > 25) = . 10, thus the hotel is

not providing the desired level of customer

service.

33.

34.
36.
 Chapter Thirteen

ANSWERS TO QUESTIONS probability an event will occur; and posterior

probability is the revised prior probability,
1. Carrying an umbrella to work because the
altered due to additional information.
weather forecast predicted an 80% chance of
rain is one example of decision making with 10. The purchase of insurance and buying a lottery

probabilities, while deciding to get married or ticket are two examples.

taking a job are examples of decision making 11. The expected value of purchasing insurance is
without probabilities. always negative for the buyer. If it was not
this way, insurance companies would not make
2. A payoff table is a means of organizing and
illustrating the payoffs from different
as much money as they do. People purchase
insurance, nonetheless, because the utility of
decisions, given the various states of nature in
having insurance and averting a financial
a decision problem. States of nature are the
disaster offsets the negative expected value.
actual events that may occur in the future in a

decision making situation. 12. Defense expenditures, welfare, and Medicare

are three examples. The expecterd dollar values
3. A maximax criterion selects the decision that
of these expenditures are negative. This is
will result in the maximum of the maximum
obvious since they represent costs with no
payoffs; thus it is very optimistic. The maximin
financial return. However, the satisfaction of
criterion selects the decision that will result in
not fearing invasion by a foreign power, of
the maximum of the minimum payoffs, which
helping people less fortunate, and of seeing
is pessimistic.
older people healthy has more utility to society
4. Regret is the penalty to the decision maker for as a whole than the dollar expenditure.
making a wrong decision.
13. A two-person, zero-sum game consists of two
5. In the Hurwicz criterion the decision payoffs players, and the sum of their gains and losses
are weighted by a coefficient that is a measure equals zero. In a non-zero sum game the sum of
of the decision maker's optimism. the players gains and losses does not equal zero.

6. With expected value the decision outcomes 14. In a pure strategy game each player adopts a

(payoffs) are multiplied by their probabilities single, optimal strategy; while a mixed strategy

of occurrence, while with expected opportunity game occurs when each player adopts an

loss these probabilities are multiplied by the optimal strategy that does not result in an

regret for each outcome. equilibrium point when the maximin and
minimax criterion are applied, resulting in the
7. Expected value of perfect information (EVPI) players shifting to new strategies.
is the maximum amount a decision maker
15. According to the minimax criterion, each
would pay for additional information. It is
player in a game plays in order to minimize the
computed by subtracting the expected value
without perfect information from the expected
maximum possible losses.

value given perfect information. The expected 16. The expected gain and loss method develops a
value given perfect information is the plan of strategies wherein the expected gain of
maximum payoff for each state of nature the maximizing player will equal the expected
multiplied by their probabilities. The expected loss of the minimizing player.
value of sample information (EVSI) is

computed by subtracting the expected value PROBLEM SUMMARY

without information from the expected value
1. Decision-making criteria without probabilities
with information, based on Bayesian analysis.
EVSI is typically less than EVPI since perfect 2. Decision-making criteria without probabilities
information is rarely obtained.
3. Decision-making criteria without probabilities
8. A decision tree is the best method when a 4. Decision-making criteria without probabilities
series of decisions is required.
5. Decision-making criteria without probabilities
9. Conditional probability is the probability that
6. Decision-making criteria without probabilities
an event will occur given that another event
has already occurred; prior probability is the 7. Decision-making criteria without probabilities

266
 8. Decision-making criteria without probabilities 48. Mixed strategy game, expected gain and loss
method
9. Decision-making criteria without probabilities

10. Decision-making criteria without probabilities 49. Pure strategy game, dominance

11. Decision-making criteria without probabilities 50. Mixed strategy game, dominance, expected
gain and loss method
12. Decision-making criteria without probabilities
51. Mixed strategy game, dominance, expected
13. Decision-making criteria without probabilities
gain and loss method
14. Decision-making criteria without probabilities

15. Expected value

16. Expected value and opportunity loss PROBLEM SOLUTIONS

17. Expected value 1. a) Lease land: maximum payoff = S90,000

18. Expected value and opportunity loss, EVPI b) Savings certificate: maximum of minimum
19. Expected value payoffs = $10,000

20. Indifferent probability 2. a) Drive-in window: maximum payoff

= S20.000
21. Expected %alue (13-10)

Expected value (13-1

b) Breakfast: maximum of minimum payoffs=
22. 1 )
$4,000
23. Expected value. EVPI (13-12)
3. a)
24. Expected value. EVPI (13-13) Good Recession

25. Expected value (13-10) Bellhop 25,000

Management 35,000
26. Payoff table, expected value

27. Payoff table, expected value Choose bellhop job.

28. Payoff table, expected value b) Bellhop: 120,000(.4) -i- 60,000(.6) = 584,000:
29. Payoff table, expected value management: 85.000(.4) + 85.000(.6) =
$85,000: select management job.
30. Payoff table, decision making without
probabilities c) Bellhop: I20.000(.5) + 60,000(.5) = $90,000;
management: 85,()()0(.5) + 85,000(.5) = 85,000;
31. Expected value (13-14)
select bellhop job.
32. Decision tree (13-25)
4. a) Course III, maximax payoff = A
33. Sequential decision tree
b) Course 1, maximin payoff = D
34. Sequential decision tree
5. a) Plant corn; maximax payoff = $35,000
35. Sequential decision tree
b) Plant soybeans: maximin payoff = $20,000
36. Bayesian analysis, EVSI (13-13)
c)
37. Bayesian analysis, EVSI (13-18) Pass Fail

38. Bayesian analysis, EVSI ( 1 3-25) Corn 12,000

Peanuts 17,000 8.000
39. Bayesian analysis
Soybeans 13,000
40. Sequential decision tree
Plant corn; minimum regret = $12,000
41. Sequential decision tree ( 1 3^0)
d) Corn: $35,000(.3) + = $16,100:
8,000(.7)
42. Sequential decision tree (13-14)
peanuts: 18,000(.3)+ 12,000(.7) = $13,800:
43. Bayesian analysis, EVSI soybeans: 22.000(.3) + 20,000(.7) = $20,600;
plant soybeans.
44. Utility

e) Corn: 35,000(.5) + 8,()()0(.5) = $2I,5(X):

45. Expected value, utility
peanuts: i8,000(.5)+ 12,000(.5) = $15,000:
46. Pure strategy game soybeans: 22,000(.5) + 20,000(.5) = $21.(X)0;
plant corn.
47. Mixed strateav same

267
6. Note that this payoff table is for costs. 10. a) Risk fund, maximax payoff = $147,000

a) Product 3, minimin payoff = $3.00 b) Savings bonds maximin payoff = $30,000

b) Product 3, minimax payoff = $6.50 c) Money market: 2(.2) +3. (.20) + 4(.2) +
1

4.3(.2) + = 36,000; .stock growth: -3(.2) -

5(.2)
7. a) Build shopping center; maximax payoff = 2(.2) + 2.5(.2) + 4(.2) + 6(.2) = 15.000; bond:
$105,000 6(.2) + 5(.2) + 3(.2) + 3(.2) + 2(.2) = 38,000;
government: 4(.2) + 3.6(.2) + 3.2(.2) + 3(.2) +
b) Lea,se equipment; maximin payoff = $40,000
2.8(.2) = 33,200; risk: -9(.2) -4.5(.2) + .2(.2)
1

c) + 8.3(.2) + 14.7(.2) = 21,400; savings bonds:

Stable Increase 3(.2) + 3(.2) + 3.2(.2) + 3.4(.2) + 3.5(.2) =

Houses 35,000 10,000 32,200; select bond fund.

Shopping center 20,000
11.
Lease 65,000

Build shopping center.

d) Houses: $70,000(.2) + 30,000(.8) = $38,000;

shopping center: $105,000(.2) + 20,000(.8) =
$37,000; lease: $40,000(.2) + 40,000(.8) =
$40,000; lease equipment.

e) Houses: 70,000(.5) + 30.000(.5) = $50,000;

shopping center: 105.000(.5) + 20,000(.5) =
$62,500; lease: 4(),000(.5) + 40,000(.5) =
$40,000; build shopping center.

8. a) Purchase motel; maximax payoff = $20,000

b) Purchase theater; maximin payoff = $5,000

c)
Shortage Stable Surplus

Motel 14,000
Restaurant 4,000 7,000 14,000
Theater 9,000 15,000

Select either motel or restaurant (both have

minimum regret values of $14,000).

d) Motel: 20,000(.4)— 8,000(.6) = $3,200;

restaurant: 8.000(.4) + 2,000(.6) = $4,400;
theater: 6,000(.4) + 5,000(.6) = $5,400; select
theater.

e) Motel: - 8,000(.33) + 15.000(.33) +

20,000(.33) = $9,000; restaurant: 2,000(.33) +
8,0()()(.33) + 6,000(.33) = $5,333; theater:
6.000(.33) + 6.000(.33) + 5.000(.33) = $5,666;
select motel.

9. a) LaPlace criterion:
EV(AIA) = 10.2(.33) + 7.3(.33) + 5.4(.33) = 7.6
EV(GIGT) = 9.6(.33) + 8. 1(.33) + 4.8(.33) = 7.4
EV(AIN) = 12.5( 33) + 6.5(.33) + 3.2(.33) = 7.3
Select Alabama vs. Auburn.

b) Select Alabama vs. Auburn; maximin payoff

= 5.4

c) Select Army vs. Navy; maximax payoff = 12.5

 M

d) Equal likelihood: d) Equal likelihood

South Korea: 21.7(.33) + 19.l(.33) + 15.2(.33) =
Byrd = 4.4(.33) + + (-3.2)(.33) = -S.989M
(1.3)(.33)
18.48
O'Neil = 6.3(.33) + ( + (-5. )(.33) = +S.99M
1 .8)(.33) 1

China: 19.0(.33) + 18.5(.33)+ I7.6(.33)= 18.18

Johnson = 5.8(.33) + (0.7)(.33) + (-2.7){.33) = +$1.254M
Taiwan: 19.2(.33)+ 17.1(.33)+ 14.9(.33) =
Gordan = 9.6(.33) + r-1.6)(.33) + (-6.3)(.33) = $.56 1

16.90 <— minimum

Select Johnson
Philippines: 22.5(.33) + 16.8(.33)+ 13.8(.33) =
17.52 15. EV(press) = 40.000(.4) - 8.()(XK.6) = S II .200;
Mexico: 25.0(.33) + 21.2(.33) + I2.5(.33) = 19.37 EV(lathe) = 20.0OO(.4) + 4.000(.6) = S 10.400:
Select Taiwan EV(grinder)= 12.000(.4)+ 10.()()()(.6)

= 510,800; purchase press.

13. a) Maximax criteria:

Office park 4.5 <— maximum 16. a) EV(sunvisors) = -500(.3) -200(.I5) +

Office building 2.4 1500(.55) = S645: EV(umbrellas) = 2.000(.3) +
Warehouse 1.7 0(.I5) -900(.55) = $105: carry sunvisors.
Shopping center 3.6
b) Opportunity loss table:
Condominiums 3.2
Select office park Rain Overcast Sunshine

Sunvisors 2.500 200

b) Maximin criteria:
Umbrellas 2,400
Office park 0.5
Office building 1.5 <r- maximum EOL(sunvisors) = 2,500(.3) + 200(.15) + =
Warehouse 1.0 S780: EOL( umbrellas) = + + 2.400(.55) =
Shopping center 0.7 SI. 320; select sunvisors since it has the
Condominiums 0.6 minimum expected regret.
Select office building
17. EV (snow shoveler) = S30(.13) + 60(.I8) +
c) Equal likelihood 90(.26)+ 120(.23)+ 150(.10)+ 180(.07) +
Office park: 0.5(.33) + 1.7(.33) + 4.5(.33) 210(.03) = S99.60
= 2.21 <— maximum
Office building: 1.5(.33) + 1.9(.33) + 2.4(.33) The cost of the snow blower (S625) is much
= 1.91
more than the annual cost of the snow shoveler.
thus on the basis of one year the snow sho\eler
Warehouse: 1.7(.33)+ l.4(.33)+ 1.0(.33)
= should be used. However, the snow blower
1 .35
Shopping center: 0.7(.33) + 2.4(.33) + 3.6(.33) could be used for an extended period of time

= 2.21 <— maximum such that after approximately six years the cost

+ + of the snow blower would be recouped. Thus,

Condominiums: 3.2(.33) 1.5(.33) 0.6(.33)
= the decision hinges on weather or not the
1.75
decision maker thinks 6 years is too long to
Select office park or shopping center
wait to recoup the cost of the snow blower.
d) Hurwicz criteria (a = .3)
+ 0.5(.7) = 1.70 18. a) EV(widget) = 120,000(.2) + 70,000(.7)
Office park: 4.5(.3)
-30.000(.l = $70,000; EV(hummer) =
2.4(.3) + 1.5(.7) =
)
Office building:
1 .77 <r- maximum + 40,000(.7) + 20.00()(. ) =
60,0()0(.2) 1

$42,000; EV(nimnot) = 35,000(.2) + 30,000(.7)

Warehouse: 1.7{.3)+ 1.0(.7)= 1.21

+ 0.7(.7) = 1.57 + 30.000(.n = S3 1. 000: introduce wideet.

Shopping center: 3.6(.3)
Condommiums: 3.2(.3) + 0.6(.7) = 1.38 b)
Select office building Favorable Stable Unfavorable

Widget 60.000
14. a) Maximax = Gordan
Hummer 60,000 30,000 10,000
b) Maximin = Johnson Nimnot 85,000 40.000

c) Hurwicz (a = .60)
EOL( A) = + + 60.000(. = S6.000; 1 )

Byrd = 4.4(.6) + = $1.36M

(-3.2)(.4) EOL(B) = 60,000(.2) + 30,000(.7) + 10,000(.l)
O'Neil = 6.3{.6) + (-5.1)(.4) = $1.74M = $34,000; EOL(C) = 85,000(.2) + 40.000(.7)
Johnson = 5.8(.6) + (-2.7K.4) = S2.40M + = $45,000; select A (widget).
Gordan = 9.6(.6) + (-6.3)(.4) = S3.24M
Select Gordan

269
 +

c) Expected value given perfect information = b) EV(off tackle) = 3(. 10) - 2(. 10) + 9(. 10) +
+ 70,000(.7) + 30,0()()(.l) =
120,()()()(.2) 7(.10)- 1(.60)= l.l;EV(option)=-l(.l()) +
76,000; EVPI = 76,000— EV(widget) = 8(.10)-2(.10) + 9(.10)+ 12(.60) = 8.6;
76,000—70,000 = $6,000; the company would EV(toss sweep) = 6(. 10)-)- 16(.10) - 5(.l()) +
consider this a maximum, and since perfect 3(.IO)-l- I4(.60)= l().4;EV(draw) = -2(.10) +
information is rare, it would pay less than 4(. 10) + 3(. 10) + 1()(. 10) - 3(.60) = -.3:

$6,000 probably. EV(pass) = 8(.10) + 20(.10) + I2(.10) - 7(.10)

- 8(.6()) = -1 .5; EV(screen) = - 5(. 0) - 2(. 10)
EV(operate) =
1

19. 1 20,000(.4) + 40,000(.2) +

+ 8(. 1 0) + 3(. 1 0) + 1 6(.6()) = 1 0.0; select toss
(-4(),000)(.4) = $40,000; leasing = $40,000; if
sweep. Yes, it is likely Tech will make the first
conservative, the firm should lease. Although
down.
the expected value for operating is the same as
leasing, the lease agreement is not subject to 23. EV (South Korea) = 21.7(.40) + 19.1(8.50) +
uncertainty and thus does not contain the I5.2(.10) =19.75
potential $40,000 loss. However, the risk taker EV (China) = 19.()(.40) + 18.5(.50) + 17.6(.10)
might attempt the $120,000 gain. = 18.61

20. To be indifferent, the expected value for the

EV (Taiwan) = 19.2(.4()) + I7.i(.5()) +
investments would equal each other: EV(stocks)
14.9(.I0)= 17.72 <- minimum
= EV(bonds). Next, let the probability of good EV (Philippines) = 22.5(.40) + I6.8(.50) +
13.8(.I0)= 18.78
economic conditions equal p and the probability
of bad conditions equal - p: 1
EV (Mexico) = 25.0(.4()) + 21.2(.50) +
12.5(.I0) = 21.85
EV(stocks)= 10,000{/;>) + Select Taiwan
-4,000(1 -/;) Expected value of perfect information =
EV(bonds) = 7,000(p) + 19(.40)+ 16.8(.50)+ 12.5(.1())= 17.25
2,000( - p) 1
EVPI= 17.25- 17.72
EV(stocks) = EV(bonds) = $-0.47 million
l().()0()(yj) + (-4,000) (1 -p) = 7,000(/7) +
2,000( 1-/7) The EVPI is the maximum amount the cost of

- 4,000 + 4,000/7 + 2,000 the facility could be reduced ($0.47 million) if

l().00()/j =: 7,000/7
- 2,000/7 perfect information can be obtained.

= 6,000 9,000/7 24. a) EV (Office park) = .5(.50) + 1.7(.40) -t- 4.5(.10)

p = .667 = 1.38
Therefore, probability of good conditions = EV (Office building) = 1.5(.50) -i- l.9(.40) +
/; =.667, probability of bad conditions =1-/7 2.4(.10)= 1.75
= ,333. EV (Warehouse) = 1.7(.50) + 1.4(.4()) -i-

21. EV(money market) = 2(.2) + 3.1 (.3) + 4(.3) 1.0(.10)= 1.51

4.3(. + 5(. ) = 34,600; EV(stock growth) =
1 ) 1
EV (Shopping center) = ().7(.5()) -t- 2.4(.4()) +
-3(.2) - 2(.3) + 2.5(.3) + 4(. ) + 6(. = 5,500: 1 1 )
3.6(.10)= 1.67
EV(bond) = 6(.2) + 5(.3) + 3(.3) + 3(.l) + 2(.l) EV (Condominiums) = 3.2(.50) + 1.5(.40) +
= 41.000; EV(government) = 4(.2) + 3.6(.3) .06(.10) = 2.26 <— maximum

3.2(.3) + 3(. + 2.8(. = 34,200; EV(risk) =

1 ) I
)
Select Condominium project

-9(.2) - 1.5(.3) + 1.2(.3) + 8.3(.l) + 14.7(.l) =

b) EVPI = Expected value of perfect information
^9,000; EV(savings bonds) = 3(.2) + 3(.3) - expected value without perfect information =
3.2(.3) + 3.4(. ) + 3.5(. ) = 3 ,500; purchase
3.01 -2.26
i 1 1

bond fund. EVPI = $0.75 million

22. a) EV(off tackle) = + 9(.20) +
3(.4) - 2(.1())
25. Using expected value: EV(compacts) =
7(.2())- i(. 10) = 4.0;EV(option) = -l(.4) +
3()(),000(.6) += $24().()()0: EV(full-
!50.()()()(.4)
K(. 10) - 2(.20) + 9(.2) + 12(. 10) = 4.2; EV(to.ss
sized) = + 6()(),0()0(.4) =
-10().()00(.6)
sweep) = 6(.4) + 16(.10) - 5(.20) + 3(.20) +
$180,000: EV(trucks) = 120,000(6) +
14(. 10) = 6.4; EV(draw) = - 2(.4) + 4(. 10) +
170.()00(.4) = $14().()00: select the compact car
3(.20) + l()(.20) - 3(.1()) = 1.6; EV(pass) = 8(.4)
dealership.
+ 20(.I0) + 12(.2()) - 7(.20) - 8(. 10) = 4.6;
EV(screen) = - 5(.4) - 2(. 10) + 8(.20) + 3(.2())
+ 16(.I0) = 3.2; toss sweep is best, followed by
pass, option, off tackle, screen, and draw.

270
26. Pa\ off matrix
c) Opportunity cost table:
 34.
580.000,000

Power
outage
SI. 000.000

Not
Installed

S80.000

32. Select compact car. Shortage Since cost of installation ($80,000) is greater
|s240,ooo|
$300,000
(6) than expected value of not installing
($495,000). do not install an emergency power
generator
$150,000

$100,000

$600,000

$120 000

Surplus ^^^ 3170,000

(4)

33.

S300.000

S60 000

$70 000

Si 05 000

$40,000

$80,000

273
 P(flc)F(c)
P(clf) =
35. P(flc)P(c) + P(f|n)P(n)

(.70) (.40)
= .70
(.70) (.40) + (.20) (.60)

P(f) = P(flc)P(c) + P(fln)P(n) = (.70) (.40)

+ (.20) (.60) = .40

P(fln)P(n)
P(nlf) =
P(fln)P(n) + P(flc)P(c)

(20) (.60)
.30
(.20) (.60) + (.70) (.40)

P(uln)P(n)
P(nlu) =
P(uln)P(n) + P(ulc)P(c)

(.80) (.60)
.80
(.80) (.60) + (.30) (.40)

P(u) = P(uln)P(n) + P(u|c)P(c) = (.80) (.60)

+ (.30) (.40) = .60

P(ulc)P(c)
P(clu)
Did not P(ulc)P(c) + P(u|n)P(n)
bid
(.30) (.40)
= .20
(.30) (.40) + (.80) (.60)

36. />(c) = probability of contract = .40; P(n) = probability

of no contract = .60; P(flc) = .70; P(ulc) = .30;

P(uln) = .80;P(fln) = .20

$40,000

$25,600

Favorable
report

Unfavorable
report

$4,000
$10,400
I $12,000

P(nlu) = m^ ^^0,000

274
 Decision strategy: If report is favorable, purchase a Decision strategy: Produce the widget regardless of
lathe. If report is unfavorable, purchase a grinder. EV the report. EV(strategy) = $69,966; EVSI =
(strategy) = = EV.„h ,„,„,„„,„„ -
$16,480; EVSI EV«„h,„,„,„3„„„ - EV,„h„„, ,„fo,m„,on = $69,966 -
EV. = $16,480 - 11.200 = $5,280
,„.,„, ,„'f„,„,3„„„ $70,000 = 0. Additional information has no value, since
37. P(f) = favorable market conditions = .2; P(s) = stable the owner will produce the widget in either case.

market conditions = .7; P(u) = unfavorable market con- 38. Let s" = shortage; s* = surplus; Pis') = .6; P(s*) =
ditions = .1; P{p\{) = .60; F(nlf) = .40; P(p|s) = .30; .4. = report of shortage;
Let S" S* = report of surplus;
P(nls) = .70; P(plu) = .10; P(nlu) = .90 P(S"ls") = .90; P(S*ls") = .10; P(S*ls') = .70;
P(S-|s*) = .30.
Posterior probability table for a positive report:

(1)
 1

P(S Is-)P(s')
P(s"IS") =
P(S'ls-)P(s") + P(S"ls^)P(s*)
(.90) (.60)

(.90) (.60) + (.30) (.40)

P(S") = .66 = .818

P(s^lS") = 1 - .818 = .182

P(S*ls')P(s^)
.
P(s^lS^)
+
=
P(S*ls^)P(s^) + P(S*ls')P(s-)

(.70) (.40)

(.70) (.40) + (.10) (.60)

PiS*) = .34 = .824

P(s"lS^) = 1 - .824 = .176

$300,000

$150,000

$272,700 -$100,000

Surplus
report

$600,000
$476,800
$120,000

$170,000

Decision strategy: If the report indicates a gas shortage,

stock compacts. If the report indicates a surplus,
stock full-sized cars. EV(strategy) = $342,094; EVSI =
EV.„h,„,„rma,ior " E V„ „ho., , nformauon = $342,094 -
$240,000 = $102,094

Expected value given perfect information =

300.000(.6) + 60().()0(4) = $420,(KK); EVPI =
- 240,000 = $ 80,000; EVSI =
$42().()()() 1

$102,094; efficiency - EVSI/EVPI =

$ 02,094/$ 80,000
1 1 = .51 or 5 7r

276
39. P(s) = .10
P(f) = .90
G = good review
B = bad review
P(Gls) = .70

P(Bls) = .30

P(Glf) = .20

P(Blt) = .80

P(Gls) P(s)
P(slG) =
P{G\s) Pis) + P{G\f) P(f)

(.70)(.10)
~ (.70)(.10) + (.20)(.90)
"
P(t"IG) = .72
P(slB) = .04 P(G) = .25
P(tlB) = .96 P(B) = .75

$25M

Bad Review
P(B) = .75

Don't
Produce

t''^'^' — t; * with information tV^^/p information

= $0.31M-(-4.7M)
= $5.01M
Hire Sickel: if good review produce, if bad
review don't produce

277
40.
6 ) $1,200

$500

(2) Oil
sample
($20)
$120,

$62.80

(3) Oil
$111.20^ change
($14.80)

(4) Oil
change and
sample
($34.80)

Sample negative,
no maintenance

278
41
6 ) $15,000

9 $2,000
J

(1)No Sample positive,

maintenance maintenance required

(SO)

Sample negative,
emergency maintenance 1^10)515.000

12 )
SI. 200

20^.,---''''''^Sample positive.
. .. r unnecessary maintenance

.80

Sample negative^^-r^j^i
so

(4) Oil
change and
sample
17 $2,000
(S130)

Sample positive,
maintenance required

Sample negative,
emergency maintenance \^18)S15.000

20 ) SI .200

20^,,--'^ Sample positive,

..q T- unnecessary maintenance

80^

Sample negative,
21 ) SO
no maintenance

279
42.
$1.69M
5

Loser
P(L) = .214

$3.75M

Make
playoffs
P(P) = .430

$7.91 M

280
42. Continued P(l) = .21 P(LI1) = .75 P(Llc) = .10 P(Llp) = .05
P(c) = .35 P(cll) = .15 P(Clc) = .80 P(Clp) = .IO
P(p) = .44 P(PI1) = .I() P(Plc) = .10 P(Plp) = .85

P{L\\)P(\) (.75)(.21) .1575

A IIL) = = = = .734
P{L\\) P{\) + P(Llc) P(c) + P(Llp) P(p) (.75)r.2n + nO)(.35) + f.n5K.44) .2145

(.IO)(.35)
P(clL) = = .163
.2145

(.05)(.44)
P(plL)= = .103
^ .2145

P(liC) = .089
P(clC) = .788
P(plC) = .124
P( IIP) = .049
P(clP) = .081
P(plP) = .870
P(L) = P(Lin P(l + P(Llc) P(c) + P(Llp) P(p)
)

= (.75)(.21) + (.10)(.35) + (.05)(.44) = .2145

P(C) = P(CI )P( + P(Clc) P(c) + P(Clp) P(p)
1 1 )

= (.15)(.21) + (.8)(.35) + (.10)(.44) = .3555

P(P) = .430
tVCil = bV^yjth information "" t'
' w/o information
= $3.75M - 3.08M = $670,000
43. P(d) = .50
P(s) = 40
P(i) = .10
P(Dld) = .80 P(Sld) = .10 P(Ild) = .I0
P(Sls) = .70 P(Dls) = 10. P(Ils) = .20
P(Ili) = .90 P(Dli) = .02 P(Sli) = .08

Joint probabilities:

P(Dd) = .40 P(Sd) = .05 P(Id) = .05

P(Ds) = .04 P(Ss) = .28 P(Is) = .12
P(Di = .002
) P(Si ) = .008 P(Ii) = .09
Marginal probabilities:

P(D) = .442 P(S) = .338 P(I) = 260

Posterior probabilities:

P(dlD) = .905 P(dlS )= 48 . 1 P(dll) = 92. 1

FfsID) = .090 P(slS) = .828 P(sll) = .462

P(ilD) = .005 P(ilS) =.024 P(ill) = .346
f V.Uh .nfornK-non = 3.03(.442) + 2.18(.338) +
2.49(.260)
= 2.72

t.\ ol = t. V^jii, information ~^ ' wjihoul information

= 2.72-2.26
EVSI = S.464 million

281
 efficiency = EVSI
EVPl
= .464
.750
= .619 or 61.9%

Place-Plus would be willing to pay the

consulting firm up to $.464 million and the
information is 61.9% as efficient as perfect
information.

44. Because the couple had limited funds, the

utility of money was greater than it might be

The $5,000 had as much

for a wealthier person.
them as perhaps $20,000 or 30,000.
utility to

This makes the couple risk averters.

45. a) EV(money market) = $162,000(1) = $162,000

EV (oil investment) = 0(.5) + 400,000(.5) =

$200,000

According to expected value Annie should

invest in the oil exploration project, however,
most people would invest in the "sure" money
market.

b) EV (money market) = .80(.5) + .80(.5) = .80

EV (oil investment) = 0(.5) + 1(.5) = .50

Thus, according to utility, the money market is

the best investment.

46.
 Reduced payoff table:

7
~ Tech would employ a shuffle 73% of
an overload 27% of the time.
the game time and

State:
1 10 3
2 4 5
shuffle; lip + 60(1 - p) = 60 + 12/?
overload: 58/? + 91(1 - p) = 91 - 33p
No pure strategy; therefore, mixed strategies must be de- 60 + 12p = 91 - 33n
termined.
45^ = 31
Smoothie: p= .69

strategy A: i0/> + 4(1 - = 4 + 6/7 1 - P = 31

p)
strategy C: ^p + 5(1 - p) = 5 - 2p EL(State) = 72(69) + 60(.31) = 68.4 points given
where p = probability strategy 1 will be used and t>y state

I - p = probability strategy 2 will be used. state would employ a zone defense 69% of the time and
6/7 + 4 = 5 - 2p man-to-man 31% of the time.

8p= 1

p= .125
1 - /? = .875

EG(Smoothie) = 10(125) + 4(.875) = 4.75 percent gain in

market share

Cooler Cola:

strategy 1: lOp + 3(1 - p) = 3 + Ip

strategy 2: 4/7 + 5(1 - ;?) = 5 - />
5 - p= 3 + Ip
%p = 1
p = .25
1 - p = .75

EL(Cooler Cola) = 4(.25) + 5(.75) = 4.75 percent loss

in market share

51.
 *CASE SOLUTION: STEELEY ASSOCIATES VS.
CONCORD FALLS
1. The decision tree for Steeley Associates is as follows.

The decision tree analysis provided by the AB:QM package is

as follows.

284
Program:
 This output indicates that the recommended
decision is, "Request a permit tor the
apartment," with an expected value of
$1,767,000.

2. The expected value is twice as much as the

certain gain for selling the property, so a

decision maker might make the conservative
decision to sell.

CASE SOLUTION: TRANSFORMER

REPLACEMENT AT MOUNTAIN STATE
ELECTRIC SERVICE
The decision tree solution for this problem is as

follows.

$90,000,000

Replace
Transformer

$8,000,000

Retain
Transformer
$90,000,000

Low Incident
Likelihood $8,000,000
.50

The decision should be to retain the existing

transformer; the cost of replacement ($85,000)

is greater than the cost of retention ($61 ,000).

The decision tree analysis provided by AB:QM

is as follows.

286
Pr
CASE SOLUTION: THE CAROLINA COUGARS

$112.5
 Chapter Fourteen

ANSWERS TO QUESTIONS 12. Steady-state analysis ( 1 3-7)

1. The properties of a Markov process are as

13. Absorbing state, debt example

follows. 14. Absorbing state, debt example (13-13)

a) The transition probabiUties for a given 15. Steady-state determination, 3x3 matrix
beginning state of the system sum to .0.
Steady-state determination 3x3 matrix
1
16.
b) The probabilities apply to all participants in the
17. Steady-state determination, 3x3 matrix
system.
18. Steady-state determination, 3x3 matrix
c) The transition probabilities are constant over
19. Steady-state determination. 3x3 matrix
time.

The states are independent over time. 20. Steady-state analysis (13-19)
d)

2. Steady-state probabilities are average 21. Steady-state analysis, 3x3 matrix

probabilities that the system will be in a certain 22. Steady-state analysis, 3x3 matrix
state after a large number of transition periods.
23. Steady-state determination, 3x3 matrix
They are computed by multiplying the matrix
of state probabilities by itself a large number of 24. Steady-state determination (14-23)

times. 25. Absorbing state, maintenance example

3. Probabilistic information about the state of a 26. Steady-state determination, 3x3 matrix (13-1)
system in the future
27. .Absorbing state, 4x4 matrix
4. Transition probabilities are the probabilities of
28. Steady-state determination. 4x4 matrix
the system moving from one state to another in
one time period. The transition matri.x includes 29. Steady-state determination, 4x4 matrix

the transition probabilities for the states

all all
PROBLEM SOLUTIONS
in the system.
1,
5. A transient state will never be returned to once Month 1 Month 2 Months
the system has moved out of it; whereas once
the system achieves an absorbing state, the
system cannot move out of it.

6. .A. descriptive technique provides probabilistic

information about future events, while an
optimization technique provides a single,
optimal recommended decision.

PROBLEM SUMMARY
1. Decision trees, computing future state

probabilities

2. Markov properties, discussion

3. Decision trees, computing future state

probabilities

4. Matrix multiplication ( 13-1)

5. Matrix multiplication (13-31

6. Matrix multiplication, steady-state

determination

7. .Steady-state determination

8. Steady-state determination
Probability in month 3:
9. Stead) -state analysis

10. Steady -state analysis ( 13-3) Starting State Petroco National Gascorp Sum
National .14 .54 .32 1.00
11. Steadv-state analvsis

289
2. a) The transition probabilities for a given beginning state
of the system sum to 1.0.
b) The probabilities apply to all participants in the system.
c) The transition probabilities are constant over time.

3.

Month 1 Month 2 Month 4

.4 .6
P(Cheesedale) = .147 + .126 + .036 + .108 = .417 b. [0„ B„] = [0„ Bo]
.8 .2
.2"
.5 .3
Oo = .40o + .8B„; B„ = .6O0 + .2B„; Oo +
4. [Pp(3) Np(3) Gp(3)] = (.5 .3 .2] .1 .7 .2
Bo = 1.0; substituting B„ = 1.0 - Oo into the first
.1 .1 .8 equation above gives
= [.30 .38 .32] Oo = .40„ + .8(1.0 - Oo) = .40o + .8 - .80„
5. Notation: Let A = Creamwood and B = Cheesedale. 1.40„ = .8

0„ = .8/1.4 = .571
[A.(3) B,(3)] = [.7 .3] [.61 .39]
and

Oo + Bo = 1.0
IA3(4) 63(4)] = [.61 .39] 583 .417] .571 + B„ = 1.0

Ba(4) = probability of Cheesedale at month 4 Bo = .429

= .417 [O B] = [.571 .429]

= =
6. a. week 2: [0„(2) B„(2)] [.4
4:: :'l
7. [T, D,] [T,•
-4- f^
= [-64 .36]
T, = .65T, + .45D,, D, = .35T, + .55D,, and
T, + D, = 1.0; substituting D, = 1.0 - T, into the first
week 3: [0„(3) Bo(3)] = [.64 equation above gives

= [.544 .456]
T, = .65T, + .45(1.0 - T,)
= .65T, + .45 - .45T,
r.4 .6
week 4: [Oo(4) Bo(4)] = [.544 .456] .8T, = .45

=
[-8 .2
T, = .45/80 = .563
[-582 .418]
T, + D, = 1.0

week 5: = 4 .6 .563 + D, = 1.0

[Oo(5) Bo(5)] [.582 .418]
8 .2 D, = .437
= [.567 .433] [T, D,] = [.563 .437]

4 .6 These are the probabilities of a customer purchasing a

week 6: [0„(6) B„(6)] = [.567 .433]
8 .2 Tribune and Daily News in the long-run future regardless
= [.573 .427] of which paper he or she purchases in week 1.

290
8. a.
13, T

Q=
(A)
(I) -.7V + .6N + .4M =
(2) .5V - .8N + .IM =

(3)
24. Ofrah Josie Barney 27.

.3821 .4870 .1308

Number of Josie viewers = (27,000)(.4870)

= 13.149

New Josie viewers = 13.149 - 8.907 = 4.242 per

day

4-month commercial revenue = (80 weekdays)

(SO. 12)(4,242) = $40,723.20

Since S40.723.20 is greater than the cost of the ad

campaign (S25.000) the station should
undertake the campaign.

25. Compute the steady-state conditions from the following

equations: pi = p4, P2 = + .6p2, p3 = .Ipi +
.8p)
.2p2 + .5p3, p4 = .Ipi + .2p2 + .5p3, pi + p2 +
P3 + P4 = 10. Solving simultaneously results in the fol-

lowing steady-state probabilities: [pi p2 pj P4]

=
[.2 .4 .2 .2]. The expected costs are EC = SOpi +
l(X)p2 + 4(X)p3 + 800p4 = 0(.2) + 100(.4) -t- 400(.2) +
800(.2) = $280 per day.

26.
.5 .3 .2

[P N G] = [P N G] .1 .7 .2

.1 .1 .8

=P .5P + .IN + .IG (1)

N= .3P + .7N + .IG (2)
G = .2P + .2N + .8G ^ eliminate
P -(- N -I- G = 1.0 (3)

One of the first three equations is redundant. If p is

known and n is known, then g must be known from
p + n + g = 1.0. Thus, we will eliminate the third
equation above and solve the remaining three simulta-
neously.

(1) -.5P + -.5P + .IN + .IG =

(2) -» -.3P + .3N - .IG =

(3)

(2)

(A)
(B)
28. a) rDry Wall Trim Framing Roofing

.3438 .2521 .2918 .1124

b) Number of days:

Dry wall = 85.95

Trim = 63.03
Framing = 72.95
Roofing = 28.10

29. a) .steady-state 54 63 Nickel Blitz

.2476 .4233 .1756 .1036

b) Number of defensive plays:

54 = 34.22
63 = 48.68
Nickel = 20.19
Blitz = 11.91

c) (57.5)(. 1036) = 5.96 or approximately 6 plays

CASE SOLUTION:
THE FRIENDLY CAR FARM

The steady state probabilities computed using

AB:QM are:

2 3

.284 .284 .265 167]

Expected number of cars in stock = .284(0)

+ .284(1) + .265(2) + .167(3) = 1.315 cars

Average inventory holding cost = .284(75)

+ .265( 175) + 167(3
. 10) = $ 1 19.64

CASE SOLUTION: DAVIDSON'S

DEPARTMENT STORE
p
 Chapter Fifteen

ANSWERS TO QUESTIONS j) Mainframe computer: multiple-server: tlrst-

come. first-served (or priority level); infinite

1. Examples might include a number of retail
calling population
stores, banks, movie theaters, service stations,
university operations and functions, computer 8. The addition of a new counter created two
access, telephone calls, etc. queues. The multiple-server model is for a
single queue with more than one server.
2. The ratio of the arrival rate to the service rate
must be less than one. which means that the 9. a) False. The operating characteristic values may
service rate must be greater than the arrival be higher or lower depending on the magnitude
rate. If customers arrive faster than they can be of the standard deviation compared to the mean
served, the system eventually develops an of the exponentially distributed service time.
infinite queue.
b) True. Since there is no variability the operating
3. Examples might include a doctor's office, a characteristics would always be lower.
faculty advisor, a machine shop, a number of
10. When arrivals are random, in the short run
breakdown/repair systems in which the
more customers may arrive than the serving
operating units are finite, etc.
system can accommodate.
4. Queue discipline, calling population, arrival
11. When customers are served according to a
rate, and service rate
prearranged schedule or alphabetically, or are
5. Waiting lines are an integral part of a multitude picked at random.
of business operations, and their efficient
12. When \i = a
design and management are an important part
of these businesses. 13. Automated equipment or robots.

6. The mean effective service rate is the number

of servers multiplied by the service rate, and it
PROBLEM SUMMARY
must exceed the arrival rate.
1. Single-server model analysis
7. a) Hair salon: multiple-server; first-come, first-

served or appointment: calling population can 2. Single-server model analysis

be finite (appointments only) or infinite (off-

3. Single-server model analysis
the-street business)
4. Single-server model analysis
b) Bank: multiple-server: first-come, first-served:
infinite calling population 5. Single-server model analysis

c) Laundromat: multiple-server: first-come, first- 6. Single-server model, decision analysis

served; infinite calling population

7. Single-server model, decision analysis

d) Doctor's office: single- (or multiple-) server;

8. Single-server model, decision analysis
appointment (usually); finite calling population
9. Single-server model, decision analysis
e) Advisor's office: single-server; first-come,
first-served or appointment; finite calling 10. Single-server model. (Problem 9), P„ analysis

population
11. Multiple-server model. (Problem 8), decision

f) Airport runway: single-server; first-come, tlrst- analysis

served; finite calling population

12. Multiple-server model analysis

g) Service station: multiple-server: first-come, 13. Single server, finite calling population
first-served: infinite calling population
14. Single-server model analysis
h) Copy center: single- or multiple-server: first-
come, first-served; infinite calling population 15. Single-server, constant service time

i) Team trainer: single-server: tlrst-come. first- 16. Single server, constant service time

served or appointment; finite calling population

17. Single server, constant service time

297
18. Single server, constant service time 4. The arrival rate must be on an hourly basis; A = —60
7.5
19. Single server, finite calling population *-
8 per hour; ^t = 10 per hour; /,,,

20. Single server, finite calling population m(m ~ A)

—A = — =
(8)' 8
-^^- = 3.2 parts; t7 = .80; / - U
21. Multiple server model ^
10(2) M 10

22. Multiple-server model (Problem 12), decision / - .80 = .20

analysis
5. (7 = — ;
M = 10 per hour; U = .90; therefore, .90 =
23. Multiple-server model, decision analysis
A
•, or A = 9 per hour, or 1 part every 6.67 min.
24. Multiple-server model, decision analysis 10

25. Multiple-server model, decision analysis 60

6. A = 4 per hour; fx = — = 5 per hour
26. Multiple-server model, decision analysis
A- (4)-
a. L,| = ._
3.2
27. Single-server model analysis Mm - A) 5(1)
A 4
28. Single-server model, decision analysis b. W, = .80 hr (48 min)
ix{fx - A) 5(1)
29. Single-server, finite queue
c. W = hr (60 min)
30. Multiple-server model analysis
45
31. Multiple-server model analysis
d. 45 min per
^ hour is a — , or .75, utilization factor. V
60

32. Multiple-server model analysis cannot exceed .75. V = —A = —4 = .80 presently,

/i, 5
33. Single server model, finite queue Therefore, a new air traffic controller must be hired.

Multiple-server model, decision analysis 60

34.
7. A = 12 per hour; /i, = —= 15 per hour. One window:
35. Multiple-server model, decision analysis
W = = = .26 hr (16 min). Two win-
36. Multiple-server model, decision analysis "
M/^ - A) 15(3)
dows: /Lt = 15 per hour (does not change). However, the

PROBLEM SOLUTIONS W,
^
= —
arrival rate for each

M(M -
A
A)
=
window

15(9)
6
=
is now split.

.044 hr (2.67 min); 16

A = 6 per hour;

2.67 = 13.33 min, reduction in waiting time; 13.33 X

1. A = 16 per hour; ^ = 24 per hour; P» = \
=
$2,000 = $26,660; cost of window = $20,000;
16
= =
/aV
— = ,
=
/" $26,660 > $20,000; therefore, the second drive-in win-
I .33; P, •
P„ (.67)' •
33 .099;
24 Vm/ dow should be installed.
A 16 A- _ (16)-
~_ 60
~ ~
_
~
^
^' ~ ~ 8. A = 28 per hour; /u. = — = 30 per hour
M - -
' ' "'
A 8 m(a<- A) 24(8)
28
1.33; W = -
1

125 hr (7.5 min); W, = a. L =

— A
— = 14; L, =
A-

-
/A A 8 /A 2 ^(/j, A)

= - = —
A 16 A16 (28)^

-
= = .083 hr (5 min); (7 = 13.1; W
1
— = —
1

= .5 hr (30 min);
/x(M A) 24(8) /x 24 30(2) /i, - A 2
.67 A 28
W., = = = .47 hr (28.2 min); V =
10 M(M - A) 30(2)
2. A = 10; = 12; W., = .41 hr 28
IX
Mm - A) 12(2) -A = — = .93 = 93%
A ^ 30
(24.6 min); U = — = —=
10
.833
fx 12 b. W = let W = 10 min = .167 hr; .167
/i. — A
A- (6)-
A - u = 10; = = ^—^ = car; - = =
M -
3. 6; /., .9
" - A)
, so (/x 28) (.167) 1; 28 6, so
10(4)
/x(M
M - 28

W =
I

= —=
1

.25 hr (15 min); W„ = = 34 students per hour.

60
— = 1.76 min required
/x - A 4
' /I to

A 6 approve a schedule in order to meet the dean's goal.

= .15 hr (9 min);if the arrival rate is
M(/x - A) 10(4) Since each assistant will reduce the .service time by
increased to 12 per hour, the arrival rate would exceed the .25 min, then 1 assistant is all that is needed (i.e., 2.00
service rate; thus, an infinite queue length would result. min - .25 = 1.75 min).

298
 A = 200 per day; /n = 220 per day 12. A = 100 per hour; /x = 60 per hour; c = 2
^L'' (200)'
L.= 9.09 trucks; W= Pn
1

m(m A) 220(20)
100\"
^1 KK) :(60)

—
= —= .05 day; 8 hr/day X 60 min/hr =
[?K 60 2! \ 60 120 - 100
/A A 20
480 min; .05 X 480 = 24 min; so W= 24 min;
1
—=
1
.09
A 200 (1 + 1.67) + 8.34 11
W„ = = = .045 day (21.6
^ min)'
'
^L{^JL - A) 220(20) ^

b. 24 - 15 = 9 min; 9 min X S10,000 = $90,000 per

year loss presently. With a new set of scales the arrival L = -^-^
;
•
P, + —
(c- 1)!(CM - A)- M
rate would be split. A = 100 per day per scales; VV =
(100) (60) (1.67)-
1
= 1 (.09) + 1.67 = 5.44
= .008 day
^ (4 min). The entire 1!(20)-
pt - A 220 - 100
$90,000 would be saved. Since the scales cost $50,000 L^ = L = 5.44 - 1.67 = 3.77
M
10. P„
per year, they should be installed.

= the probability of n customers in the system; W=—=

L
A
—
5

100
44
= .054 hr(3.26 min)

Pi — the probability of exactly 4 in the system. However,

L„ 3.77
we need the probability of 4 or more. Therefore, compute W,
'
= —= = .038 hr(2.26
^ min)'
A 100
P„, Pi, P:, and P, and subtract from 1.0.

p„ =

11.
 .

13. No. of machines: 5

Breakdowns: X-- per day

Repairs: |i = 1 per day

No. of repairmen: 1

a) P„= !

n =
|_(5-«)! \4/ J

1 + 1.25+ 1.25 + .94 +.47 + .12

= -p-TyT = .2 (repairmen idle 20% of the time)

^ -^
b) L,, = 5 - -^^ '
(1 - .2) = 5 - (-L^)(.8) = 5-4=1 machine
'

.25 ^ .25 ^

Mean No. Machines waiting for Repairman = 1 Machine

c) W,
(N - L)X

Where. L = La + (\ - Pq) = \ + (\ - .2) = \

W,, = ! = 1= 1 .25 days

'
(5-I.8).25 .8

Mean time waiting for repair = 1.25 days

= 5\
d) P„ (lY(.2) = (60)f-LY.2) = .1825

18% chance of three machines in the system

(being repaired or waiting for repair).

60 60
14. \ = = 13.33 customers per hour 15. \ = = 20.7 per hour
4.5 2.9 *

60 60
ix = = 21 .43 customers per
^ hour LL = = 22.22 per hour, constant service
^ 2.8 2.7 .. ^
time
\2 (13.33)2
L L^i = 6.3434 customers
" ^JL(|JL-\) (21.43)(21.43- 13.33)

= 1.02 customers waiting W^ = .3064 hr or 18.39 minutes

\2 (13.33)
W.
''
^JL(|JL-\) (21.43)(21.43- 13.33)

= .077 hr or 4.61 minutes waiting in line

This is not \'ery good service for a fast food

drive-through- window.

300
16.
20. X = 7.5/hour

|i = 3.0/hour

(=4
c|i= 12.0

Po = .07

\-Po = .93

^^0' X
^7.5)(3.0)(||)^
7.5
^= r. - ....... .^^ Pn
- 1)2 /^O +77
+ ^ - (4-l)![(4)(3.0)-7.5]2^0^^ + = ^"^"^^^
(c- I )!(c^ 3.0

L, = L-^= 3.03- Z|.. 533

V'V =
Y = y- = .404 hour = 24.24 minutes

^ .533
~
W^q = T- ~ TT -^"^^ ^^"'' ~ '^•26 minutes

21. X^ 4()/hour
)i = 1 5/hour

c =4

t|i = 60

Po = .06\

\ -Pi) = .9390

^~{c- \)\(c\i-X)'-^^^\i-{4- l)![(4)(15)-40]2(06')+ ,5

= 3.42

L, = L-^= 3.42- ^=.756

L 3 42

,.. ^'1 -756

v^d = T- = T7r= 019 hour =1.13 minutes

302
 I

^•PV^=6l = <"

idle time = 1 - .67 = .33, thus the postal

workers are idle 33% of the time which seems

excessive.

b) M/q = and Lq = .756 customers,

1.13 min.
neither which seems excessive

c) A customer can expect to walk in and get

served without waiting approximately 6% (i.e.,

Po = .061) of the time.

Overall, the system seems very satisfactory

from a customer service perspective, however,
the post office might want to analyze the
system with 3 stations instead of 4 because of
the large percentage of idle time.

22. A = 100; M = 60; c = 3

Po
^ 1 AooV" 1 /looVV 180 \
"^
:r n! \ 60 / J 3! V 60 / \180 - 100/

(1 + 1.67 + 1.39) + — (1.67)-(2.25)

6
1 1
.17
1 + 1.67 + 1.39 + 1.75 5.81

L = ^-^-^ ; •
P„ + —
(c- 1)!(CM- A)- M

(100) (60)
^
mV
^V60/ 100
= +
^
^(17)' + .37 1.67
(2)! (180 - 100)- 60
2.04

A
L 2.04 - 1.67 = .37

With 2 presses: i, = 3.77; with 3 presses: Lq = .37;

£,(2) - Z.,(3) = 3.77 - 0.37 = 3.4 reduction; S50 X

3.4 = $170; therefore, the third press should be added,
since it costs onlv S150.
 (6) (4) (1.5)^
29.
(.21) + 1.5 = 1.74
2! (6)^

L^ = L = 1,74 - 1.5 = ,24

TA = —
L„ .24
M^q = = .04 hr(2.4 min)
6

Hire a third doctor.

25. A = 10; jii = 6. For c = 2: P,, = .090909 and Pw =

CfJ. (2) (6)

c\\n Cfl 2! \ 6 (2) (6) - 10

(.0909) = .75758. Thus, 2 salespeople are not enough.

For c = 3: P„ .172662 and Pw=-( —

3! \ 6
(3) (6)
(.172662) = .29976. Thus, 3 salespeo-
(3) (6)- 10_
pie are sufficient to meet the company policy that a cus-
tomer should have to wait no more than 30% of the time.

26. A = 40; M= 15; c = 3; P„ = .028037; L =

A/Li,(A/u)'

- \y.(cfi-\y
p ^A= (40) (15) (40/ 15)'
{c IX 2! [(3) (15) - 40]-

(.028037) + — = 9.0467; L^ = L -- = 9.0467 -

15 p,
40
— = 6.38; ly, =
Z,,
-a =
6.38
-— = .1595 hr = 9.57 min.
1-^ A 40
Thus, 3 servers should be sufficient.

27. Determine the required value of l/yu, (average repair time) if

A, the average arrival rate, is 6 per day and if W, the

average time in the system, is assumed to be I day.

^J
 A 32.8 minute waiting time may seem long, but 34. a) A =5; 2;c = 3
actually restaurant customers sometimes
perceive a wainting line and a reasonably long
waiting time as an indicator of "quality."
i\\ii]\iiy ,1 !y (3)(2)

31. L,^ - 2.2469 manuscripts fe( 2/ 1!V2; 2!V2/J .l!V2/(3)(2) - 5

= .045
L = 12.2469 manuscripts
(5) (2) (5/2)-- 5
= ^'^^
W,i .32\ weeks -(3-l)![(3)(2)-5]^^°'^>"l =
VV^ = 1 .7496 weeks Lq = 6 = W„
3^
3.5; 0.70 hr(42 min);
5
t/=.8333 6.0
32. a) By testing several different numbers of servers
W = 1.20 hr (72 min)

(teams) for the multiple server model in QM for b) A = 5; l/fi = 25 min; therefore, ^l = 2.4; c = 3;
Windows, it is determined that at least 7 teams P„ = .0982;

are required to be within the two week waiting (5) (2.4) (5/2.4)-- 5

period. The operating characteristics are,

' '
(3 ^ .)l|,3),2.4) - 5|^
<"^^^> *
T.
' '"
L(^ = 2.315 jobs L^ = 3.18; - —=
2.4
5
1.1; W = ^5— =
1.1
.22 hr

L = 8.0292 jobs (13.2 min waiting time). The improvement in average

waiting time per truck is 42 - 13.2 = 28.8 min. The
estimated value of this time savings is ($750) (28.8) =
W = 2.0073 weeks $21,600. Since the cost of achieving the improved ser-
vice is only $18,000, the firm should implement the

p = .8163 improved system, yielding an expected savings of

$21,600 - $18,000 = $3,600.
b) This can be determined in two ways. First, if
c) Alternative 1: Add a fourth loading/unloading loca-
the average number of jobs arriving each week tion at the dock, yielding four locations, where each
is 4. at SI. 700 apiece they will generate S6.800
location has a mean service rate of /x = 2 per hour.
in revenue per week whereas 7 painting teams Alternative 2: Add extra employees and equipment at

will cost $3,500 per week for a difference of the existing three dock locations to reduce loading/
$3,300 per week. Alternatively, if there are unloading times from 30 min to 23 min per truck,
approximately 8 jobs in the system (L = 8.0292) yielding ft = 2.6 per hour. Decision analysis: The ten-

over a two week period (W = 2.0073) then that dency of the student will be to compute the waiting
time for both alternatives. However, this is not required,
will result in SI 3.600 in revenue for two weeks
since the alternatives can be evaluated by using the
with the 7 teams paid S7.000 for two weeks or
concept of "effective service rate," which is determined
approximately S3, 300 per week.
by multiplying the number of servers by the mean ser-

a. Arrival rate: A = 40 units per hour. Processing times: vice rate. The purpose of this part of the problem is to

(1) without additional employees; 1//Xi = 1.2 min per introduce this concept; thus, the instructor may wish
unit; thus, /x, = 50 units per hour; (2) with additional em- to give the student a hint before assigning this prob-

ployees; 1//J.: = .9 min per unit; thus, /xi = 66.67 units lem. Computing the effective service rate for each al-

per hour. In-process inventory = number in process and ternative gives, for alternative 1, (no. of servers) (mean
waiting to be processed = L = number in the sys- service rate) = (4) (2) = 8 trucks per hour, and for
tem = \/ifJ. — A); (1) without additional employees; alternative 2, (no. of servers) (mean service rate) =
/., = 40/(50 - 40) = 40/10 = 4 units in system; (2) (3) (2.6) = 7.8 trucks per hour. Since the cost of each

with additional employees: Lj = 40/(66.67 — 40) = alternative is approximately equal, alternative 1, to add
40/26.67 =1.5 units in system. Decision analysis: Cost a fourth dock location, is superior because it increases

of in-process inventory: (1) without additional employees: the effective service rate to 8 trucks per day; whereas
(4)(S31) = $124.00/day; (2) with additional employees: adding extra resources to the existing dock increases

(1.5)($31) = $46.50/day. Difference = $124.00

- the effective service rate to only 7.8 trucks per day.

$46.50 = 77.50/day. Thus, the optimal decision is to add 35. 7 :00 to 9:00 a.m.: A = 10, ft = 2.5; c must equal at
additional employees at a cost of $52.00 per day, yielding least 5 for the mean effective service rate to exceed the
a net expected savings of S77.50 - $52.00 = $25.50/day.

305
 .

arrival rate; W^ = .222 min with 5 cashiers; therefore, 5 CASE SOLUTION: NORTHWOODS
9:00 a. m. to noon: A = 4, /a = 2.5; c must
is sufficient.

equal at least 2; W^ = .711 min with 2 cashiers; there-

BACKPACKER
fore, 2 is sufficient. Noon to 2:00 p.m.: A = 14, /A = 2.5; There are four system configurations to be
c must equal at least 6 cashiers; W^ = .823 min with 6 considered, as follows.
cashiers; therefore, 6 is sufficient. 2:00 to 5:00 p.m.:
A = 8, = 2.5; c must equal at least 4 cashiers;
/i.
1 5-day, 8-hour per day service
W, = .298 min with 4 cashiers; therefore, 4 is sufficient.
2. 7-day, 8-hour per day service
36. A = 5, /i = 1.714
a. c = 4, Pn = .043, L = 4.197, L, = 1.28, W
= 3. 5-day, 16-hour per day service
.839 hr = 50.34 min, W^ = .256 hr = 15.36 min
= 4. 7-day, 16-hour per day service
b. c = 5, Po = .051, L = 3.220, L, = .303, W

.644 hr = 38.64 min, W^ = .061 hr = 3.66 min. Al- In each case the first step is to determine the
though the customer waiting time is reduced from
number of servers that are required to make
15.36 to 3.66 min, 15 min does not seem excessive for the system feasible, i.e., c|l > X. Remember,
a hairstylist; thus, the impact of adding a fifth stylist
the current system has 5 operators (servers),
is probably not significant.
and, \x = 60/3.6 = 16.67 custoiners per hour.

*CASE SOLUTION: THE COLLEGE OF BUSINESS 5-day, 8-hour service: k = 175, |i = 16.67: c > }J\i
COPY CENTER or c> 175/16.67= 10.49. Thus,

A multiple-server queuing model must be evaluated for a center at least 1 1 total operators are

with 2 copiers and 3 copiers for the normal academic year and required tor this (the

the summer. Normal academic year: 2 copiers: A = 7.5, pi = 5, current) system to be feasible.

c = 2; W= .475 hr = 27.42 min. In an 8-hr day there are Since the cunent physical facil-

60 jobs (8 X 7.5 = 60). 60 jobs X .475 hr = 27.42 hr of to-

ity can only accommodate a
secretarial time in the college spent on copying jobs.
tal maximum of 10 work .stations,
$8.50/hr X 27.42 hr = $233.07 in secretarial wages spent
this alternative is eliininated.
daily for copying in the college. 177 days in the normal aca-
demic year X = $41,253.39 per year. 3 copiers: A =
$233.07 7-day, 8-hour service: X = 125. |a = 16.67; c > }J\l

7.5, /A = 5, c = 3; W = .232 hr = 13.92 min. In an 8-hr or c > 125/16.67 = 7.49.

day there are 60 jobs (8 X 7.5 = 60). 60 jobs X .475 hr
=
Thus, at least 8 operators are
13.92 hr of total secretarial time in the college spent on copy- required for this system to be
ing jobs. $8.50/hr X 13.92 hr = $118.32 in secretarial wages
feasible.
spent daily for copying in the college. 177 days in the normal
academic year X $118.32 = $20,942.64 per year. Summer 5-day, 16-hour service: X = 87.5: p. = 16.67; c > A/|i
months: 2 copiers: A = 3.75, /a = 5, c = 2; W = .233 hr = or c > 87.5/16.67 = 5.24.
13.98 min. jobs/day = 30; 30 jobs X .233 hr = 6.99 hr; Thus, at least 6 operators are
$8.50/hr X 6.99 = $59.42/day; 70 days X 59.42 = required for this systein to be
$4,159.05. 3 copiers: A = 3.75, fi = 5, c = 3; W = .204
feasible.
hr = 12.24 min. jobs/day = 30; 30 jobs X .204 hr = 6.12 hr;
$8.50/hr X 6.12 = $52.02; 70 days X 52.02 = $3,641.40. 7-day, 16-hour service: X - 62.5. |i = 16.67: c > A/|i

Current copy center total wage copying costs = $41,253.39 -l-

or c > 62.5/16.67 = 3.74.
4,159.05 = $45,412.44. Copy center with 3 copiers total wage Thus, at least 4 operators are
copying cost = $20,942.64 + 3,641.40 = $24,584.04. Total required for this system to be
annual copying wage savings by adding a third machine
=
feasible.
$45,412.44 - $24,584.00 = $20,828.40 per year. Since a configuration (the
Therefore, only the first
copying machine costs $36,000 and has maintenance costs of
current one) is not feasible and is eliminated.
$8,000 per year, the total cost over the life of the copier will be
Next the costs of the remaining 3 alternatives
$84,000 (with no present value discounting). Over the same
are evaluated.
6-year period, adding a third copier would save $124,970.40 in
copying wage costs, which outweighs the cost of a new copier.
7-day, 8-hour service
However, Dr. Moore may still not be able to convince Dr. Bur-

ris. The wage savings are not really savings to the college but a X^ 125. n= 16.67
measure of secretarial time that could be reallocated to other Cost for 7 day service = $3,600
tasks within the departments. The college would not save any Cost per extra operator = $3,800
money; it would simply incur the cost of the copier. The de-

partments could argue that other tasks the secretaries might Recall that at least 8 operators are required for

perform instead of copying would be a more efficient use of this configuration to be feasible. Thus, starting
$20,828.40 in annual wages, but Dr. Burris would probably be at this point we must compute liic waiting
a hard sell with this argument.

306
 times for different numbers of operators until
the goal of one-half minute waiting time is

achieved.

8 operators: Uq = 5.76 minutes

9 operators: W^= 1.20 minutes
10 operators: W'q = 0.42 minutes
(Probability of waiting = .18)

Since 5 extra operators are required to reach

the waiting time goal the cost of this alternative
is S3,600 + (5)(3.800) = S22.600.

5-day, 16-hour service

X= 87.5. ^= 16.67
Cost for 1 6 hour ser\ ice = S 1 1 .500
Cost per extra operator = $4,700

At least 6 operators are required for this configuration

to be feasible.

6 operators: W^ = 3.24 minutes

7 operators: U'q = 0.78 minutes
8 operators: Wq = 0.27 minutes
(Probability of waiting = .21)

Since 3 extra operators are required to reach

the waiting time goal the cost of this alternative
is S 1 1 .500 -I- (3)(4,700) = S25.600.

7-day. 1 6-hour service

>. = 62.5. |i= 16.67

Cost for 6-hour service = $11 .500
1

Cost for 7-day service = S7.200

Cost per extra operator = S6,300

At least 4 operators are required for this configuration

to be feasible, however, since 5 operator stations
already exist, the starting point is 5 operators.

5 operators: Wq = 1.32 minutes

6 operators: Wq = 0.36 minutes
(Probability of waiting = .10)

Since only 1 extra operator is required to reach

the waiting time goal the cost of this alternative
is S 1 1 .500 + 7,200 + ( 1 )(6.300) = S25.000.

The 7-day. 8-hour service configuration has the

lowest cost. However, all three alternatives are
very close according to cost. All three also
meet the goal of a customer getting immediate
service at least 80 percent of the time. Thus,
other factors may be taken into consideration.
For example, both of the 16-hour service
alternatives might be more convenient for
customers who work durins the day.

307
 Chapter Sixteen

ANSWERS TO QUESTIONS PROBLEM SUMMARY

1. The Monte Carlo technique is a technique for 1. Rescue squad emergency calls

selecting numbers randomly from a probability

2. Car arrivals at a service station
distribution for use in a trial run of a
simulation. 3. Machine breakdowns

2. Random numbers are organized into ranges to 4. Inventory analysis

reflect the probabilities of the different possible
5. Decision analysis
occurrences of a random variable. When the
random numbers are selected in the simulation, 6. Machine repair time ( 1 5-3)
they represent the value of a random variable.
7. Product demand and order receipt
Real time is an actual time unit such as an
hour, week, day, or month; while simulated 8. Bank drive-in window arrivals and service
time is a simulated activity that replicates real
9. Loading dock arrivals and service
time.
10. Product demand and order receipt
By repeating the simulation many times

The random numbers must be uniformly

11. Football game
distributed, must be generated in an efficient 12. Student advising, arrival and service
manner, and must not reflect a pattern.
13. Markov process
In general, simulation provides descriptive,
probabilistic information about a system. It can 14. Inventory analysis

be used to find an optimal set of operating

15. Rental car agency
characteristics by searching through all

possible sets of operating characteristics until 16. CPM/PERT network analysis

the best set is found.

17. Store robbery and getaway
Optimal solutions can sometimes be achieved
18. Stock price movement
in simulation by using search techniques
wherein sets of operating characteristics are 19. Hospital emergency room staffing
.searched until the best set is found.
20. Break-even analysis
Simulation results can be validated by
21. Rating dates
comparing the results with manually derived
results, by comparing the results with similar 22. Model analysis (16-21)
real-world outcomes, and by making sure that
23. Production capacity
the starting conditions are accurate

Examples might include preregistration, an 24. Baseball game

athletic event, class scheduling, campus traffic

and parking, building maintenance, etc. PROBLEM SOLUTIONS

10. Some factors to consider are circling time, 1.
Time Between
number of air traffic controllers, landing time,
taxiing time for runway to terminal, unloading
time (passengers and luggage per flight),

refueling and maintenance time, loading time

(passengers and luggage per flight), taxi time

to runway, takeoff time, terminal trafflc. ticket

sales, employees, passenger terminal parking,
passenger facilities (restaurants, toilets,
magazines, etc.), waiting areas. In general, a
simulation model would be constructed to yield
the average traffic flows not only of planes but
also of passengers, employees, autos, etc. From
these values the planner could design the size
of the facility based on the number of people
expected to use it.
 a) Time Between Cumulative a)
Calls Clock Time Between
Arrival Arrivals (rain)
65
71 1
20 11
15 13
48 17
89 22
18 25
83 30
08 32
90 37
05 38
89 43
18 46
08 48
26 51
47 55
94 60
06 62
72 66
40 69
62 73

b) M = 27
= 3.48 hr between calls; EV = 1(.05) +
2(.10) + 3(.30) + 4(.30) + 5(.20) + 6(.05) = 3.65.
The results are different because there were not enough
simulations to enable the simulated average to approach
the analytical result.

c) 21 calls; no, this is not the average number of calls per

3 days. In order to determine this average, this simula-
tion would have to be repeated a number of times in
order to get enough observations of calls per 3-day pe-
riod to compute an average.

2.0
Time Between
Arrivals Cumulative Random
(min) Probability Numbers
.15 01-15
.45 16^5
.85 46-85
1.00 86-99,00
 Machine
Breakdowns/ Cumulative Random
Week Probability Numbers
.10 01-10
.20 11-20
.40 21-^M)
.65 41-65
.95 66-95
1.00 96-99, 00

a)
Week Breakdowns
1
 6. a) The first three columns are from Problem 3. Select as
many r:*s as there are breakdowns from a different
random number stream.

Week Ti Breakdowns Repair Time

1
Week Breakdowns Repair Time

13
8. One-teller system:
Ties for queue length: a)

Line Random Numbers

1 0(M9
2 50-99

Customer
Reorder
12.

Time Between
Arrivals
14. Order size = s demand: profit =
16 cases. Order size
DP - QC - = 16D - 164. Demand > order
(G/2)Cc
size: profit = QP - QC - (,Q/2)C, - C,{D - Q) =
108 - D.

Demand Random Numbers

15
Activity 1-3:
 Monte Carlo Simulation (using 16th row of random num-
bers from Table 15.3)

Considering the city as a grid with an x and y axis with

the store at point (0,0) each random number selected indi-
cates a movement of 1 unit (block) in either an x or v

direction.

18.
 Doctor Nurse Both

Cummulative RN Cummulative RN Cummulative RN

Time Probability Ranges Time Probability Ranges Time Probability Ranges

10
20.
Sales
Date
\\eek
Inning 1: Inning 4:

Team
Inning 7:

Team
This function is integrated to develop the cummulative
probability distribution.

F(
.=j:f*= (0
_ z _ _^
""6 3"

Letting F(z) - r.

= £- 1
6 3

and solving for z,

(-4)
Z. = 6r + 2

Monte Carlo Simulation:

 Chapter Seventeen

ANSWERS TO QUESTIONS 11. This question requires an opinion of the

1. Production planning such as scheduling, student but in general, the appropriate model is

inventory, process, facility layout and design, determined primarily by the extent of any trend
workforce, material purchasing: financial pattern.

planning including the development of budgets

12. A linear trend model will not adjust to a
and capital expenditures; and various
change in trend as the adjusted exponential
marketing functions are dependent on
smoothing model will, thus limiting the trend
forecasting demand.
line method to a shorter time frame.
2. Qualitative forecasting methods are subjective
13. This question requires a subjective estimate on
estimates based on judgment, opinion, past
the part of the student. A particular method
experience, etc.. whereas quantitative methods
might be viewed as being superior because it is
are mathematical, based on formulas.
easier to use (compared to the other methods),
3. Short-range forecasting methods typically it is easier to interpret, it makes more sense, it

encompass the immediate future, i.e.. several can be used alone rather than in comparison, or
months, and are concerned with daily it seems to tit the data better.

medium-range forecasts encompass

operations:
14. Linear regression relates demand to one other
anywhere from several months up to se\eral
independent variable whereas multiple
years and are used for annual budgets and
regression reflects the relationship between a
production plans or the development of a
dependent variable and two or more
project or program: long-range forecasts usually
independent variables.
are for periods longer than one or two years and
are used for strategic planning such as new 15. V is the dependent variable, x is the

product development, new programs, etc. independent variable, a is the intercept, and
b is the slope of the line.
4. A trend is a gradual long-term, up or down
movement of demand; a cycle is an undulating PROBLEM SUMMARY
up and down movement that repeats itself over 1. Moving average, MAD
a lengthy time span: a seasonal pattern is an
oscillating movement in demand that occurs
2. Moving average, MAD
periodically and is repetitive. 3. Moving average. MAD
5. Exponential smoothing is a moving average 4. Discussion
that weights the most recent past data more
5. Exponential smoothed, and moving average
strongly than more distant past data.

6. Other ways used to obtain initial forecasts

6. Exponential smoothed, adjusted. MAPD
include taking an average of demand in 7. Exponential smoothed, adjusted, linear trend
preceding periods or making a subjective line
estimate. If forecasting has been a continual
8. Exponential smoothed, adjusted, linear trend
process than preceding forecasts might be used.
line
7. The higher the smoothing constant, the more
sensitive the forecast will be to changes in
9. Adjusted exponential smoothed. E and E
recent demand. 10. Seasonally adjusted linear trend line (16-9)

8. Adjusted exponential smoothing is the simple 11. Seasonally adjusted linear trend line (16-3)
exponential smoothing forecast with a trend
adjustment factor added to it.

9. It is a judgmental choice, but in general, a high

smoothing constant. /3. reflects trend changes
more than a low (3.

10. In a linear trend line equation, the independent

variable, x. is alwavs time.

329
 ) ) )

12. Seasonally adjusted forecast c) Three month MAD = 5-month MAD

1.89, =
2.43. The dealer should use the 3-month
13. Adjusted exponential smoothed, linear trend
forecast for January because the smaller MAD
line, MAD indicates a more accurate forecast.

14. Seasonally adjusted linear trend line

15. MAD, MAPD and cumulative error

16. Discussion, forecast accuracy

17. Exponential smoothed (16-1)

18. MAD and cumulative error

19. Linear regression

20. Linear regression, correlation

21. Linear regression, correlation

22. Coefficient of determination ( 1 6-2 1

23. Linear regression, correlation, discussion

24. Linear trend line (16-23)

25. Discussion ( 1 6-2 1

26. Linear regression

27. Forecast model selection

28. Model comparison

29. Exponential smoothing (17-28)

30. Seasonally adjusted forecast; linear regression

31. Adjusted exponential smoothing ( 1 7-3 1

32. Model selection

33. Multiple regression. Excel

34. Multiple regression. Excel (17-28)

35. Multiple regression. Excel

PROBLEM SOLUTIONS
1. a) and b)
5. a) and b)
 Adjusted
Exponentially Exponentially
Smoothed Forecast Smoothed Forecast
Quarter Sales (a = .50) (a = .50, p = .50) Error
350 350.00
2 510 350.00 350.00 160.00
3 750 430.00 470.00 280.00
4 420 590.00 690.00 -270.00
5 370 505.00 512.50 -142.50
6 480 437.50 407.50 72.50
7 860 458.75 454.37 405.62
8 500 659.37 757.50 -257.50
9 450 579.69 588.91 -138.91
10 550 514.84 487.03 62.97
II 820 532.42 527.30 292.69
12 570 676.21 745.55 -175.55
13 631.22

E = 26.30 12. Dai Daily Demand

E= 289.336 1 212
The forecast seems to be biased low. 2 182
3 215
10. Seasonal factors: Quarter 1: 1 =
170/6630 .18
4 201
Quarter 2: 1540/6630 = .23
5 158
Quarter 3: 2430/6630 = .37
6 176
Quarter 4: 1490/6630 = .22
7 212
Forecast for 1999: v = 1850 + 180x = 2570 188
SD 544
Seasonally adjusted forecasts:
Quarter 1:2570(.18) = 462.6
389
Quarter 2: 2570(.23) = 591.1 Si(10AM-3PM) = .25
Quarter 3: 2570(.37) = 950.9
SD 1544

Quarter 4: 2570(.22) = 565.4 Dt 567

S2(3PM-7PM) = = .37
The seasonality factor seems to provide a more
SD 1544

accurate forecast. 320

S3(7PM-11PM) = = .21

11. Seasonal factors: Quarter 1: 395/1594 = .25

SD 1544

Quarter 2: 490/1594 = .31 268

S4(11PM-2AM) = ^ ,17
Quarter 3: 308/1594 = .19
^ ^-
SD 1544
Quarter 4: 401/1594 = .25 Linear Trend Forecast for Day 9:

V = 202.54- 2. 12.t = 183.46

Forecast for year 4: 440.33 + 45.5x =
622.33 Day 9 Forecast for
lOAM-3 PM: = 45.87
183.45(.25)
Seasonally adjusted forecasts:
3 PM-7 PM: = 67.88
183.46(.37)
Quarter 1: 622.33(.25) = 155.6
7 PM-I PM: I83.46(.21) = 38.52
1

Quarter 2: 622.33(.31) = 192.9

-
11 PM-2AM: I83.46(.17) = 31.18
Quarter 3: 622.33(. 19) 118.2
Quarter 4: 622.33(.25) = 155.6

334
13.
c)

15.
Cumulative Error
 17. a)

Month

MAD
22. Coefficient of determination = (.929)2 = .863.
indicating that 86.3% of the variation of ice
cream sales can be attributed to the
temperature.

23. a)y = 6541.66 - .448x

where v — applications
.V = tuition

If tuition is S9.000. forecast is 6541.66 -

.448(9000) = 2.503.2. applications

If tuition is S7.000. forecast is 6541.66 -

.448(7000) = 3.400.6 applications

b) Correlation coefficient = -.754. indicating a

fairly strong linear relationship between tuition

costs and number of applicants.

c) Number of class sections, number of dormitory

rooms, number of persons per class, plus

budgeting decisions.

24.
27.
 29.

Year Demand Forecast

I 526
2 510 326.0
3 296 381.2
4 478 355.6
5 305 392.3
6 506 366.1
7 612 408.1
8 560 469.3
9 590 4%.4
to 676 524.5
II 569.9

The exponential smoothing model appears to be

less accurate than the linear trend line forecast

developed in 28(a).

30. la) Seasonally adjusted forecast.

 33. a. v = 745.91 - 2.226x| + .163x2 Internal forecasts for classroom space,
facilities, dormitory space, dining, etc. would
b. /-^ = .992
enhance the planning process. Times series

c. V = 7.186.91 methods would probably be sufficient for this

type of forecasting.
34. a. y = 608.795 - 0.215x| - 0.985x2
The university might consider using a
b. r^- = 0.766 forecasting model to determine future funding

= 608.795 + 0.215(1,500) - 0.985(300) from the state. Several models such as a

c. y
= multiple regression and perhaps a qualitative
635.79
technique like the Delphi method might be
35. a. y = 1 17.128 + .072x| + 19.148x2 combined. Forecasts for other sources of
/-2 = .822
funding such as endowments and tuition
b.
increases could be forecast using more
c. y= 1 17. 128 + .072(1 0,000)+ 19.148(5) = conventional methods such as regression or
$932.87 time series.

The university's TQM approach requires a

forecast of what customers perceive
CASE SOLUTION: FORECASTING AT educational quality to be in the future, i.e., a

STATE UNIVERSITY definition of quality according to students,

Foreca.sting would be appropriate in a number parents and legislators. In-house forecasting

of diiferent areas. The university needs to be using key administrators, faculty and students
able to forecast future applications and might be appropriate. Surveys and market
enrollments both in the short and long term. A research techniques of alumni, students and
forecast of the college age population that will parents might be useful in determining what
apply to State is very important for planning quality factors will be important in the future.
purposes. A multiple regression model that
related applications to variables such as
population, tuition levels and entrance
requirements would probably be most
appropriate for this purpose.

CASE PROBLEM: THE UNIVERSITY BOOKSTORE STUDENT COMPUTER

PURCHASE PROGRAM
The following table shows several different
forecast models developed using QM for
Windows and selected measures of forecast
accuracy.

Forecast Year 15
Method Forecast MAD £ (bias)
Moving average (a; = 3) 1,004.66
Linear trend line
Exponential smoothing (a = 0.3)
Exponential smoothing (a=0.5)
Exponential smoothing (a=0.4, ^=0.4)
Exponential smoothing (q:=0.4, /3=0.5)
 investigate which different majors and classes
might be moving to more extensive computer
usage in the future, thus driving up long run
student demand. Additionally forecasts for
other products would help the bookstore plan
their inventor), warehouse usage and
distribution better.

CASE SOLUTION: VALLEY SWIM CLUB

Attendance
 Chapter Eighteen

ANSWERS TO QUESTIONS that they eliminate all uncertainty and potential

variation in the model.
1. In general, independent demand items are final

or finished products that are not dependent 7. In a continuous inventory system, the reorder
upon internal production activity; i.e., it is
point is the inventory level at which a new
usually external and beyond the direct control
order is placed, and, lead time is the time
of the organization. Alternatively, dependent required to receive an order after it has been
demand is usually a component part or material
placed.
used to produce a fmal product. An example
of independent demand for a pizza restaurant 8. In a noninstantaneous receipt model, the order
would be a final product such as a pizza while
quantity is received gradually over time and
dependent demand would be any of the the inventory level is depleted at the same time
ingredients (cheese, tomato sauce, dough, etc.)
it is being replenished, while in the basic EOQ
and perhaps complementary items such as model orders are received all at once.
drinks.

9. The EOQ model with shortages converges

2. In a fixed-order quantity system an order is
toward the basic EOQ model without
placed for the same constant amount whenever shortages, and shortages are eliminated.
the inventory on hand decreases to a certain

level, while in a fixed-time period system an

10. The total purchase price of all items demanded
order is placed for a variable amount after an must be included in the model since the
established passage of time.
difference in prices for different order sizes
must be reflected in the model.
3. The two basic inventory decisions are how
much to order and when to order items for
11. Price has no real inpact on the optimal order
inventory. In a continuous order system would not
size; it is a constant value that alter
whenever inventory decreases to a specific
the basic shape of the EOQ total cost curve.
level (referred to as the reorder point), a new
order is placed for a fixed amount. 12. The noninstantaneous receipt EOQ model
Alternatively, in a periodic inventory system,
would approach the basic EOQ model with
inventory on hand is counted at specific time
instantaneous receipt.
intervals and an order is placed for an amount
that will bring inventory back to a desired

level.

4. Ordering cost, carrying cost and shortage costs. PROBLEM SUMMARY

As the order size increases, ordering costs and
shortage costs decrease while carrying costs
1. EOQ model
increase.

2. EOQ cost analysis (problem I)

5. The optimal order quantity occurs when the
ordering cost equals carrying cost, thus, the 3. EOQ model
order quantity can be determined by equating
these two cost functions and solving for the
4. EOQ model

optimal value. Noninstantaneous receipt model

5.

6. Demand is known with certainty, shortages are 6. Shortage model

not allowed, lead time for order receipts is

constant, and, orders are received all at once. 7. EOQ model and reorder point
The.se assumptions are limiting to the extent

344
 8. EOQ model and reorder point PROBLEM SOLUTIONS
9. Noninstantaneous receipt model I. D = 1.200
10. Noninstantaneous receipt model
Co =$450
11. EOQ model and reorder point Cc = $170
12. Noninstantaneous receipt model

I) = /
2QD ^ / 2(450)(I.2(K)) ^ ^^
13. Shortage model
V Q V
DO/
170
14. Shortage model
1.200
i--i)"\
15. Shortage model b) TC = C„-^C, 4 = 450
(^^gyj^ I 70

16. Quantity discount model (2^) =513.550

17. EOQ model analysis

18. Noninstantaneous receipt model analysis c) — = = 13.03 orders

19. Carrying cost determination

20. Quantity discount model

'"i^ = ^^"""'>-
21. Quantity discount model

22. Quantity discount model

Cases TC
23. Quantity discount model (problem 19)

24. Quantity discount model

25. Quantity discount model (problem 22)

26. Quantity discount model

27. Quantity discount model

28. Service level

29. Service level (problem 28)

30. Reorder point with variable lead time

31. Safety stock and reorder point with variable

demand and lead time

32. Reorder point (problem 3 1 ) with variable

demand

33. Reorder point with variable demand

34. Reorder point and service level with variable

demand and lead time

35. Reorder point (problem 12) with variable

demand

36. Fixed period model, variable demand

37. Fixed period model

38. Fixed period model

i. D
T n ( ' 1 A'' 1
> 1 " 1 1 ^ -10
 b) TC^C,-^ +C ^

,/,_K5oq\
V 2,000 /

= 149,138+ 149,138
4<-'(^)('-tS)]
= $298,276
= 6,324.5 + 6,324.5
12. Operates 360 days/year
= $12,649
12 converters

5 tons coal/day/converter
D 1 0.000
c) A' - ~Q = ~ "^'^^ orders/year
D = (5 tons)( 12 converters )( 360 days) = 2 529 K
2 ,600 tons/year
1

250
C„ = $80 n = -,
gQ = 63.3 days

C, = 20% of average inventory level

C, = (.20)($12) = $2.40 d) Q = 2,529.8. /? = 60

>q,D The number of operating days to receive the

'2(80X21,600)
a) Q q. y 2.4
entire order is

Q 2,529.8
V ,440,000 = 1.200
1 tons J =-^;^= 303.7 days
14. D =3.700
b) TC=Q^ +Q^ C„ = $420

Q. = $ 1 .75
(1^99.^+ (80)(21,600) \
/'
4)
V 2 / V 1,200 / C =$4

= 1

= $2,880
,440 + 1 ,440
^^JW^
2(420)(3,7000) / .75 + 4
(^)
1

<^>
„
^
3oU
,
= ^ tzt: =
D — —
5(21,600)
TZK
3oO
-^„^
= 300 tons
A/ 1.75

= 1 ,598 tires

13. D = 10,000 logs/year

C
T = 250 days/year

R =60(250)= 15, OOO/y ear

Q, = $ ,600 = (1,598)
1

V 1.75 +4/
C, =$15
= 486.3
CD2 2(1,600)( 10,000)
a) =
-|,_m / (,5)(i-lMQ0\ —
'V /?/ V A 15,000/ rC= + + C„-rr
2Q 2Q " Q
= VmOOOOO = 2,529.8 logs

348
 1

(4)(486.3)-

2(1.598)
(l.T.^X 1.598 -486.3)2

2(1.598) „ , .yi^^^)
/2(650)(4(X)) /6() + 45\
= ^—j5—(-^;^) =142.21 sets

= 295.98 + 676.72 + 972.47

= $1,945.17

15. D =270.000 = 60.95 sets

C„ = $105
Q-^- Ctg-*^)- Q,D ^ 60(60.95)2
C< = $.25
22 ""
2^ ""
C^ 2(142.21)
C, = $.70

45(142.21 -60.95)- _.,. / 400 \

+ 650
2(142.21) V 142.21/

= $3,656.70
/2(105X270.000) / .70 + .25 \
V -25 \ .70 /
17. a. D = 17.400

C, = $3.75
17.544
C, = S2.600
2C^D /2(2.600)( 17.400)
Q ^ .1—^—=
Q.
J—
V
r^i
3.75
= 4.912.03
"V

C^jD C^Q _ (2.600)( 17.400)

= (17.544) TC =
v. 25 + .70/ e "^
2 ~ 4.912.03

(3.75)(4,9 12.03)
= 4.616.84

QS'- QiQ-SV- D =$18,420.11

2Q 2Q Q
b. C„ = $ .900
1

(.70)(4.562. )- (.25)( 17544 - 4.562. 1 )2

2(17.544) 2(17,544) Q = S4.50

270.000 \
/ 270.0U0 Q
[^ _ / 2(1.900)(17.400) .,„^^^^^
+ (105) V C^ \ 4.50
\ 17.544 /
C„D C,.Q ( 1. 900 )( 17.400)

Q 2 3.833.19
= 425.24+ 1.190.66+ 1.615.94
(4.50)(3,833.19)

= $3,231.84

= $17,249.36
16. D =400
C„ = S650 Select the new location.

C, = $45

C, = S60

349
18. a. D = 270,000 20. D = 10,000

C,. =$0.12/lb. Q, = $150

C„ = $620 Q. = $0.75

P =305,000

^CqD 2(620)(270,000)
Q =
270 \
(0.12) 1

305/

= 155,928.59

D
maximum level = e( •
]
= (155,928.59)(.1148)

= 17,900.6

—C„D CM
-— + — — I D\ = (620)(270,000)
TC = -
1

Q 2 \
I pj
p 155,928.59

(0.12X155,928.59)
(.1148)

=$2,148.07

b. P =360,000

2(620)(270,000)
Q = -2Jo\ = 105,640.90
(0.12) (
1 1

360/

(620)(270,000) (0.12)( 105,640.90)

TC = (0.25)
105,640.90

= $3,169.23

19. D =900 21.

C„ = $7,600

C, = ?

Q = 120

'2CD
Q =
C,

/ 2(7,600)(900)
120

2(7,600)(900)
(120)2 ^

C. = $950
 Without discount: (1.700) = $66,568.76

'2C„D Q =300:
Q =
C.P

// 2(160)(900) \
37.24 (1.700) = $65,384.80
^V (.I2)(205) /

= 108.2
Q = 500:

36.48 (1.700) = $64,704

= (.60)
(^). (.,2X205)
(^) Q = 800:

+ (205)(900)

36. 10 (1.700) = $65,235

- 1.330.86+ 1.330.86+ 184,500

Select Q = 500. TC = S64.704

= $187,161.72

With Discount:
23. Q= -"^°'l'-^"" =226
Q= 300

P= 190

TC=C„-^ +QP -^ +PD = (160) 38 (1,700) = $66,406.65

Q =300:

(IIJII-)
+(.12)(190) (^) + (190X900)
= $174,900
37.24 (1.700) = $65,188

Take the discount. Q= 300

2 = 500:

22. D = 1.700 rc=i2o(i^) +

8(^)
Co = 201
= $64,424
36.48 (1.700)
a=(.25X$38) = S9.50
2 =800:

'2C„D
Q =
2(120X1.700) =
(9.5)
207 rc = 120 fl^VsfM.)
V 800 / V / 2
C,P
36.10 (1.700) = $64,825

,C=,20(^).(9.50,(f).(38,
Select Q= 500. TC = $64,424

24. D =6.500

Co = $28

Cc = $3

351
 2C,D 2(28)(6,500) 26. D = 2.300,(J()0/10() = 23,000 boxes
Q =. = 348.32 = 348
C, V 3 Co = $320

^^ .n /6,500\ ,/348\ Cc = $ 1 .90

2C,D 2(320X23,000)
16 (6,500) = $105,045 e = /
= /
= 2,783.4
.90

Q = 1,000:
= 2,784 boxes

^^-28(^ /6.500\
+ 3(^ /1,000\
+
^^ ..«/23,000\ /2,784\
14 (6,500) = $92,682

47 (23.000) = $1,086,288.5
Q =3.000:

/6,500\ /3,000\ Q =7,000:

^^ = 28(3^)+3(^ +
^ .^^/23,000\ /7,000\
13 (6,500) = $89,(J60.67

= 6,000: 43 (23,000) = $996,701.43

C^

T^ o^ /6^500\ . /6,000\ Q = 12,000:

/23,000\ /1 2,000 \
12 (6,500) = $87,030.33 ^c-Ht27)oo)^'<-^)-
41 (23.000) = $955,013.33
Select Q = 6,000 TC = $87,030.33

Q = 20,000:
>(28)(6,500)
25. Q = 337.26 = 337 boxes
3.20

38 (23.000) = $893,368

16 (6.500) = $105,079.20 Select Q= 20.000 boxes. TC = $893,368

Q = 1 .()()():
/ 7f320)(7
"^
3 000)
27. Q = / .;^~;.' ' = 2,502.76 = 2.503

boxes
14 (6.500) = $92,582
/23.0()0\ /2,503\

Q = 3. ()()():

47 (23,000) = $1,086,881.50
rr^ .„ / 6.5()0\ ..,, /3,000\

Q = 7.000:
13 (6,500) = $88,460.67

Q = 6.()()():

43 (23,000) = $997,576.43

^^^-H^;:oooh--^H-^;
C? = I 2.000:
12 (6.500) = $85,230.33

Select (J = 6.000 TC = $85,230.33

47 (23,000) = $955,913.33

352
 Q = 20.000: 35. R = dL + Zo,i\L
/23.0O0\ /20.000\ R = 2.5(8) + .29( 1 1.2) (\8) = 24.38
^^^-^^Hlooooj^'-^n-T-)^ Decisions would be based on inventor) holding
cost, desire for low inventory, importance of
38 (23.000) = S893.368 reliable delivery, cost of the monitors from
each source, etc.
Select Q= 20.000 boxes. TC = S893.368
36. 5 = 200
28. J = 3.000 = 30
f/,

L = 6 L=4
o,/ = 600
a,i = 80

R = dL + Zoj yJL / = 60

R = 3.000(6) + .64(600) 1
\'6"= 20.425

= 2.425 yards
Q = d(ti, + L) + ZOj \ti, + L-l
Safety stock
Q = 200(30 + 4) + .65(80) \30 + 4-601

29. It" safety sjock = 2.000 = 7.509.69 oz.

Z(600)(\6 = 2.000 )
37. 5= 8
Z = .3608. which coiresponds
1 to a 9 1 %
service level
L= 3
30. d = 8.000 Od = 2.5
L = l / =
Gi = \.b
Q = d(ti, + L) + ZCd Ml, + L- I
R = dL + ZdGi -
= 8.000(7) + 2.06
e = 8(10 + 3) + 2.33(2.5) \ 10 + 3
(8.000)(1.6)
= 125 pizzas
R = 82.368 lbs. -5 =
125 120 pizzas
31. d= 18
38. 5= 18
a^ = 4 = 30
r^
L = 3
L=2
Oz.= .8
a,/ = 4
R = dL + Z ^o-dL + a-Ld^
/ = 25
/?= (18)(3)+ 1.29 V(4)2(3) + (.8)2(1 8)2

= 74.61 gallons Q = d{th + L} + Z<5d ^h + -Z^-/

Q = 18(30 + 2) + 2.05(4) ^30 + 2 -
Safety stock = 20.6 gallons 1

/?='l8(3)+ 1.65 V(4)2(3) + (.8)2(1 8)2

= 597.4 bottles

= 80.37 gallons
Saftey stock would increase to 26.37 gallons

32. R =7iL+ zo,i vr SOLUTION:

CfiiSE
R= 18(3)+ 1.29(4) (\^) = 62.94 THE NORTHWOODS GENERAL STORE
R drops to 62.94 gallons: less saftey stock is

neccessary This case requires a combination of classic

EOQ common sense. Assuming
analysis and
33. R = dL + Za,i \^ that thedemand of 7,500 gallons per year is
R ^ 26(9) + .68(10) (\9) = 254.4 gallons
constant, then monthh demand is 675 gallons.
for a 95% service level. __ During the 4-month maple syrup season the
Safety stock = Zg,i Vl = .65(10) (\9) = 49.5
total demand is 2.500 gallons which leaves
1

The reorder point increases to 283.5 gallons

demand of 5.(X)0 gallons for the remaining 8
months of the year.
34. R = dL + Z\ a-jL + a-id
/? = 2.5(25)+ 1.29 >i(1.2)2(25) + (10)2(2.5)2
First, determine an optimal order quantity for
=95.67 monitors
the maple syrup season:
Safetv stock = 33.17 monitors

2( 450 2.500)
)(

Q = = 435 gallons
15(1 -.205)

353
 The total production from December 3 1 to
The length of a production run is — or 4.35 February 28, results in approximately the extra

days and 65.75 runs are needed 5,000 gallons plus the normal production to

to meet demand during the 4-month period. meet demand during that period. Therefore,

This also means a run needs to start about the store should start producing syrup on a full

every 21 days. Using this information the time basis at about the first of the new year.

following schedule can be developed:

November 1 - start run

CASE SOLUTION: THE TEXAS
STADIUM STORE
November 5 - end run
The objective of this case problem is to

November 26 - start run determine the reorder point with variable

demand and lead time. The first step is to
December 1 - end run
complete the average demand and standard
December 21 - start run deviation, and, average lead time and standard
deviation from the data provided in the
December 26 - end run
problem. This a good opportunity allow
is to

January 16 - start run students to use a statistical software package (if

they have access to one) to compute these
January 20 - end run
statistics.

February 10 - start run

d - 42.57 hats per week
February 15 - end run o,/= 10.41 hats per week
L= 1 8.50 days

This schedule, producing approximately 435 O^ = 4.67 days

gallons each run, will meet demand in the 4-
Note that since demand is in terms of hats per
month season. weeks
week, lead time must be changed to

also,
Next we will assume that for the remaining
5,000 gallons, it is desired to produce them as L = 2.64 weeks
close to the end of the season as possible in G/ = .67 weeks
order to minimize storage costs. Working from =
The first question is. if/? 150, what level of
February 28, backwards, the following
service does this correspond to. Thus, we are
schedule can be developed producing the
seeking Z as follows,
maximum 100 gallons per day until 5,000
gallons is produced. \ 50 = 11 + Z^O-,iL + o-id~
150 = (42. 57)(2.64)
February 18 to 28, produce 1,000 gallons + Z a/( 10.41)2(2.64) + (.67)2(42.57)2
150= 112.38-1-2(33.16)
February 14 to 18, a normal production run so Z= 1.13
only 65 extra gallons (500-435) are produced.
This Z value corresponds to a normal
probability value of .7416, thus, the service
January 28 to February 14, produce 1,700
level is approximately 74.2 percent.
gallons
The desired service level is 99 percent
(Z = 2.58). The reorder point and safety stock
January 24 to 28, a normal production run so
for this service level is determined as follows.
only 65 extra gallons are produced.
R = dL + Z^o]iL + o-,7i-
January 7 to 24, produce 1,700 gallons. R = 42.57)(2.64) + 2.58
(

V( 10.41)2(2.64) -I- (.67)2(42.57)2

January 3 to 7, a normal production run so only /?= 112.38 -t- 85.55
65 extra gallons are produced. /?= 197.9 or 198 hats

December 31 to January 3, produce 400

gallons.

354
 \

Ms. Jones could determine the order size with Quantity Discount Analysis
EOQ analysis by using the average demand, d,
as D in the EOQ formula. However, she would If Q > $500,000: points = 2%
also need the ordering and carrying costs. It is

likely that the ordering cost is relatively high as Since Q is unaffected by points, and Q was
compared to carrying cost since the hats are $371,842; we know we must set Q= $500,000
shipped from Jamaica while it would probably for this alternate option.
not be very expensive to store hats (given their
small size and weight).
TC = q,^ + Q^+ .02 D
CASE SOLUTION: THE A TO Z
OFFICE SUPPLY COMPANY .r.^ /5.1 85.000 /37l,842\

D =$\ 7.000/day = S5, 1 85,000/year (305 day

= .,
(1,200)
(^yy^^j^ (.09) (-^)
year) + (.02)(5, 1 85.000)

C, = $.09/dollar/year = $.09
= 12.444 + 22,500+ 103,700
C„ = $\ ,200/loan + .0225 Q
= $138,644
L - \5 days

Since this option yields a lower TC of $ 1 1 ,484

Optimal loan amount per loan:
(150,128 - 138,644); it should be accepted.

'2q^ /(2)(1. 200X5,185.000)

Q =
C .09

- $37 1 ,842.26 loan amount per loan

Memo:

The .0225 Q cost per loan is not included in the

calculation of Q since it is paid on the entire
dollar amount of the loan regardless of loan
size, and thus it is simply an annual cost, i.e.,

.0225 X D.

TC = C„ — + C % + .0225 D

/5,185.000\ ^^ /371,842\
. (1,200)
(3yy;^) + (.09) (^T-j
+ (.0225)(5, 85,000) 1

= $ 50, 28.30 total cost of borrowing

1 1

D 5.185,000
N = = 13.944 loans/year
Q 371,842

= 14 loans/year for about $371,000 per loan

/• = (1 5)( 1 7.000) = $225,000 reorder point.

When cash balance gets down to $255,000

initiate another loan.

355
 Chapter Nineteen

ANSWERS TO QUESTIONS particular type of problem and a mechanism

for reasoning such that inferences can be made
1. A vehicle for accumulating, organizing and from the knowledge base
distributing information; typically, it is

computerized. 10. A DSS assists the decision maker, an expert

system replicates an advisor; a DSS
2. A data base is an organized collection of recommends a decision, an expert system
information that provides the basis for the makes it; the major focus of a DSS is decision
distribution of information.
making, in an expert system, transfer of

3. Because of the vast quantities of data expertise; in a DSS the decision maker queries
employed in a management information the computer, and vice versa in an expert

system, only a computer could accumulate, system; data manipulation is numeric in a DSS,
store and organize it efficiently. symbolic in an expert system; DSS problems
are complex and broad, expert system
4. A decision support system is a computer-based problems are narrow; DSS treats unique
information system that integrates the decision problems, expert system problems are
maker into the system so that the system and repetitive; a DSS data base is factual, an expert
the decision maker function interactively to
system data base is procedural and factual; a
reach a decision. A DSS typically encompasses
DSS has no reasoning and limited explanation
one or more management science models that
capabilities, whereas an expert system has
assist the decision-making process. The
both.
difference between an MIS and a DSS is that

the manager is typically external to the MIS. In 11. A heuristic reflects intuition, arguments, and

other words, the MIS supplies information but rules of thumb to make decisions, and as such,

does not interact with the decision maker to it reflects human reasoning. An algorithm is a

assist in the decision-making process. step-by-step solution approach that guarantees

a solution.
5. A form of experimentation using the computer
where the manager proposes a change in the 12. The application of information generated from
system and the computer provides the results management science models
of this change 13. This answer will be an opinion of the student

6. A type of computer program that determines but should be from the list provided in this

necessary input from managers in the form of chapter.

questions 14. Implementation does not encompass only the

7. The decisions of management and their results final decisions but must also reflect the steps of

provide additional information for the data- the management science process.

base on which subsequent decisions will be 15. Behavioral change on the part of the manager
based. often viewed as being a necessary ingredient
is

8. The student should develop a system in the in order for implementation to be successful.

form of Figure 19.1 in which the manager 16. Costs include personnel requirements,
interacts with one or more management science disruption of normal operating conditions,
models in order to reach a decision. The model
computer cost, and the cost of any outside
of the system should contain modules for the specialists.
management science model and for the

decisionmaking function, as well as various 17. The location of the management science staff

information flows and channels of interaction. within the organizational structure and its

An example is a system that determines the political power can have a significant effect on
optimal order quantity with the EOQ model the success or failure of implementation .

queried on an interactive basis given various

operating conditions.

9. A computer program that includes a knowledge

base containing an expert's knowledge for a

356
CASE SOLUTION: INFORMATION this information for tuition increases or submit

SYSTEMS AND DSS WITHIN A financial requests from an endowment or the

state. Forecasts of future cash flows are
UNIVERSITY
needed to make long run capital budgeting
This case problem can be addressed either decisions. Such decisions can be made using
individually or in student groups. There is no linear programming models.
specific solution. The student responses will
Universities always have a number of projects
be based to a large extent on the size and type
underway, such as constructing a new building
of school you are at and the effort devoted to
or facility or developing a new administrative
this case.
system. Project management techniques such as
Examples of IS and DSS that the student might CPM/PERT can be used as part of a DSS to
describe could include the following. establish schedules for project completion.

• Universities typically use school-age Fund-raising campaigns require information on

population demographic data for their state to the location of alumni and other potential
predict future enrollment trends. Forecasting donors. Financial data about donors allow
models can be employed to create a DSS to university administrators to develop target
predict future applications data. groups or individuals for fund raising.

• Data from applications are used to predict These are just a few of the man\ examples of
future academic year enrollments and class the potential for IS and DSS in a university

demand which subsequently determines class environment, which students might identif\ and
schedules and offerings, classroom assignments describe.
and teaching assignment

• The university budget office uses forecasts of

future applications and enrollments to predict
cash flows from tuition. They subsequently use

357
I
 —I
I
ISBN 0-13-^7S0fl7-a
9999 1

'rentice Hall

upper Saddle River, New Jersey 07458

http://www.prenhall.com 9 780139"750878