Chapter 5 Linear Programming

5 LINEAR
PROGRAMMING
Objectives
After studying this chapter you should

be able to formulate linear programming problems from

contextual problems;

be able to identify feasible regions for linear programming

problems;

be able to find solutions to linear programming problems using

graphical means;

be able to apply the simplex method using slack variables;

understand the simplex tableau procedure.

5.0

Introduction

The methods of linear programming were originally developed

between 1945 and 1955 by American mathematicians to solve
problems arising in industry and economic planning. Many such
problems involve constraints on the size of the workforce, the
quantities of raw materials available, the number of machines
available and so on. The problems that will be solved usually
have two variables in them and can be solved graphically, but
problems occurring in industry have many more variables and have
to be solved by computer. For example, in oil refineries, problems
arise with hundreds of variables and tens of thousands of
constraints.
Another application is in determining the best diet for farm animals
such as pigs. In order to maximise the profit a pig farmer needs to
ensure that the pigs are fed appropriate food and sufficient
quantities of it to produce lean meat. The pigs require a daily
allocation of carbohydrate, protein, amino acids, minerals and
vitamins. Each involves various components. For example, the
mineral content includes calcium, phosphorus, salt, potassium, iron,
magnesium, zinc, copper, manganese, iodine, and selenium. All
these dietary constituents should be present, in correct amounts.
A statistician at Exeter University has devised a computer program
for use by farmers and companies producing animal feeds which
enables them to provide the right diet for pigs at various stages of
development, such as the weaning, growing and finishing stages.
The program involves 20 variables and 10 equations!
83

Chapter 5 Linear Programming

Undoubtably linear programming is one of the most widespread

methods used to solve management and economic problems, and has
been applied in a wide variety of situations and contexts.

5.1

Formation of linear
programming problems

You are now in a position to use your knowledge of inequalities

from the previous chapter to illustrate linear programming with
the following case study.
Suppose a manufacturer of printed circuits has a stock of
200 resistors, 120 transistors and 150 capacitors
and is required to produce two types of circuits.
Type A requires 20 resistors, 10 transistors and 10 capacitors.
Type B requires 10 resistors, 20 transistors and 30 capacitors.
If the profit on type A circuits is 5 and that on type B circuits is
12, how many of each circuit should be produced in order to
maximise the profit?
You will not actually solve this problem yet, but show how it can be
formulated as a linear programming problem. There are three vital
stages in the formulation, namely
(a) What are the unknowns?
(b) What are the constraints?
(c) What is the profit/cost to be maximised/minimised?
For this problem,
(a) What are the unknowns?
Clearly the number of type A and type B circuits produced; so
we define
x = number of type A circuits produced

y = number of type B circuits produced

(b)

What are the constraints?

There are constraints associated with the total number of
resistors, transistors and capacitors available.
Resistors

Since each type A requires 20 resistors,

and each type B requires 10 resistors,
then
20x + 10y 200 ,
as there is a total of 200 resistors available.

84

Chapter 5 Linear Programming

Transistors Similarly
10x + 20y 120

Capacitors

Similarly

10x + 30y 150

Finally you must state the obvious (but nevertheless important)
inequalities

x 0, y 0 .
(c) What is the profit?
Since each type A gives 5 profit and each type B gives 12
profit, the total profit is P, where
P = 5x + 12y .

You can now summarise the problem as:

maximise

P = 5x + 12y

subject to

20x + 10y 200

10x + 20y 120
10x + 30y 150
x 0
y 0.

This is called a linear programming problem since both the

objective function P and the constraints are all linear in x and y.
In this particular example you should be aware that x and y can only
be integers since it is not sensible to consider fractions of a printed
circuit. In all linear programming problems you need to consider if
the variables are integers.

Activity 1

Feasible solutions

Show that x = 5, y = 3 satisfies all the constraints.

Find the associated profit for this solution, and compare this profit
with other possible solutions.

At this stage, you will not continue with finding the actual solutions
but will concentrate on further practice in formulating problems of
this type.
The key stage is the first one, namely that of identifying the
unknowns; so you must carefully read the problem through in order
to identify the basic unknowns. Once you have done this
successfully, it should be straight forward to express both the
constraints and the profit function in terms of the unknowns.
85

Chapter 5 Linear Programming

A further condition to note in many of the problems is that the

unknowns must be positive integers.

Example
A small firm builds two types of garden shed.
Type A requires 2 hours of machine time and 5 hours of
craftsman time.
Type B requires 3 hours of machine time and 5 hours of
craftsman time.
Each day there are 30 hours of machine time available and 60
hours of craftsman time. The profit on each type A shed is 60
and on each type B shed is 84.
Formulate the appropriate linear programming problem.
Solution
(a)

Unknowns
Define
x = number of Type A sheds produced each day,

y = number of Type B sheds produced each day.

(b)

Constraints
Machine time:
Craftsman time:

5x + 5y 60

and
(c)

2x + 3y 30

x 0, y 0

Profit

P = 60x + 84y
So, in summary, the linear programming problem is
maximise P = 60x + 84y
subject to 2x + 3y
x+y
x
y

86

30
12
0
0

Chapter 5 Linear Programming

Exercise 5A
1. Ann and Margaret run a small business in which
they work together making blouses and skirts.
Each blouse takes 1 hour of Ann's time together
with 1 hour of Margaret's time. Each skirt
involves Ann for 1 hour and Margaret for half an
hour. Ann has 7 hours available each day and
Margaret has 5 hours each day.
They could just make blouses or they could just
make skirts or they could make some of each.
Their first thought was to make the same number
of each. But they get 8 profit on a blouse and
only 6 on a skirt.
(a) Formulate the problem as a linear
programming problem.
(b) Find three solutions which satisfy the
constraints.
2. A distribution firm has to transport 1200
packages using large vans which can take 200
packages each and small vans which can take 80
packages each. The cost of running each large
van is 40 and of each small van is 20. Not
more than 300 is to be spent on the job. The
number of large vans must not exceed the
number of small vans.
Formulate this problem as a linear programming
problem given that the objective is to minimise
costs.

5.2

3. A firm manufactures wood screws and metal

screws. All the screws have to pass through a
threading machine and a slotting machine. A
box of wood screws requires 3 minutes on the
slotting machine and 2 minutes on the threading
machine. A box of metal screws requires 2
minutes on the slotting machine and 8 minutes
on the threading machine. In a week, each
machine is available for 60 hours.
There is a profit of 10 per box on wood screws
and 17 per box on metal screws.
Formulate this problem as a linear programming
problem given that the objective is to maximise
profit.
4. A factory employs unskilled workers earning
135 per week and skilled workers earning 270
per week. It is required to keep the weekly wage
bill below 24 300.
The machines require a minimum of 110
operators, of whom at least 40 must be skilled.
Union regulations require that the number of
skilled workers should be at least half the
number of unskilled workers.
If x is the number of unskilled workers and y the
number of skilled workers, write down all the
constraints to be satisfied by x and y.

Graphical solution

In the previous section you worked through problems that led to

a linear programming problem in which a linear function of x
and y is to be maximised (or minimised) subject to a number of
linear inequalities to be satisfied.
Fortunately problems of this type with just two variables can
easily be solved using a graphical method. The method will
first be illustrated using the example from the text in Section
5.1. This resulted in the linear programming problem
maximise P = 5x + 12y
subject to 20x +10y 200
10x + 20y 120
10x + 30y 150
x 0
y 0

87

Chapter 5 Linear Programming

You can illustrate the feasible (i.e. allowable) region by

graphing all the inequalities and shading out the regions not
allowed . This is illustrated in the figure below.
y
20

20x + 10y = 200

16

12

10x + 30y = 150

4

10x + 20y = 120

Feasible region

0
0

12

16

20

Magnifying the feasible region, you can look at the family

of straight lines defined by
x=0
C = 5x + 12y

y
5

10x + 30y = 150

C = 45

where C takes various values.

C = 30

The figure shows, for example, the lines defined by

C = 15

10x + 20y = 120

B

C = 15, C = 30, and C = 45.

0
0

On each of these lines any point gives the same profit.

Activity 2
Check that the points

x = 1, y =

25
12

x = 2, y =

5
3

x = 4, y =

5
6

each lie on the line defined by C = 30. What profit does each of
these points give?

88

Feasible region

y=0

10

20x + 10y = 200

Chapter 5 Linear Programming

Where is the point representing maximum profit ?

As the profit line moves to the right, the profit increases and so the
maximum profit corresponds to the last point touched as the profit line
moves out of the feasible region. This is the point B, the intersection
of

10x + 30y = 150 and 10x + 20y = 120

Solving these equations gives 10y = 30, i. e. y = 3 and x = 6. So
maximum profit occurs at the point (6, 3) and the profit is given by
P = 5 6 + 12 3 = 66 .

Example
A farmer has 20 hectares for growing barley and swedes. The farmer
has to decide how much of each to grow. The cost per hectare for
barley is 30 and for swedes is 20. The farmer has budgeted 480.
Barley requires 1 man-day per hectare and swedes require 2 man-days
per hectare. There are 36 man-days available.
The profit on barley is 100 per hectare and on swedes is 120 per
hectare.
Find the number of hectares of each crop the farmer should sow to
maximise profits.
Solution
The problem is formulated as a linear programming problem:
(a) Unknowns
x = number of hectares of barley

y = number of hectares of swedes

(b) Constraints
x + y 20
Land
30x + 20y 480
Cost
Manpower x + 2y 36

(c) Profit
P = 100x + 120y

To summarise, maximise P = 100x + 120y

subject to x + y
30x + 20y
x + 2y
x
y

20
480
36
0
0
89

Chapter 5 Linear Programming

The feasible region is identified by the region enclosed by the five

inequalities, as shown below. The profit lines are given by

C = 100x + 120y
and again you can see that C increases as the line (shown dotted)
moves to the right. Continuing in this way, the maximum profit
will occur at the intersection of

x + 2y = 36 and x + y = 20
y
24

x=0

30x + 20y = 480

20

16

c = 1800
x + 2y = 36
12

c = 1200
8

feasible
region

c = 600
4

x + y = 20
x

0
0

12

16

20

24

28

32

36

y=0

At this point x = 4 and y = 16, and the corresponding maximum

profit is given by
P = 100 4 + 120 16 = 2320

The farmer should sow 4 hectares with barley and 16 with swedes.

Exercise 5B
1. Solve the linear programming problem defined in
Question 1 of Exercise 5A.

At least 200 m 2 must be available for each

caravan and 90 m 2 for each tent.

2. Solve the linear programming problem defined in

Question 2 of Exercise 5A.

The nightly charges are 2 for a caravan and 1

for a tent.

3. A camp site for caravans and tents has an area of

1800m 2 and is subject to the following regulations:

Find the greatest possible nightly takings.

The number of caravans must not exceed 6.

Reckoning on 4 persons per caravan and 3 per
tent, the total number of persons must
not
exceed 48.

90

How many caravans and tents should be admitted

if the site owner wants to make the maximum
profit and have
(a) as many caravans as possible,
(b) as many tents as possible?

Chapter 5 Linear Programming

4. The annual subscription for a tennis club is 20 for 5. The numbers of units of vitamins A, B and C in a
adults and 8 for juniors. The club needs to raise at
kilogram of foods X and Y are as follows:
least 800 in subscriptions to cover its expenses.
The total number of members is restricted to 50.
The number of junior members is to be between one
quarter and one third of the number of adult
members.

Food

Vitamin A Vitamin B Vitamin C

Represent the information graphically and find the

numbers of adult and junior members which will
bring in the largest amount of money in
subscriptions.

A mixture of the two foods is made which has to

contain at least 20 units of vitamin A, at least 24
units of vitamin B and at least 12 units of
vitamin C.

Find also the least total membership which will

satisfy the conditions.

Find the smallest total amount of X and Y to

satisfy these constraints.

Food Y is three times as expensive as Food X.

Find the amounts of each to minimise the cost
and satisfy the constraints.

5.3

Simplex method

Where will a linear programming solution always occur?

Looking back at the second example in Section 5.2, the slope of the
5
profit line was 6 . (The slope is actually negative but it is sufficient
to just consider the magnitude of the slopes of the lines.) This is

more than the slope of the line x + 2y = 36 (namely

1
2

), but less

than the slope of the other two lines, x + y = 20 (i.e. 1) and

30x + 20y = 480 (i.e.

3
2

).

So the solution will occur at the intersection of the two lines with
slopes

1
2

and 1.

Activity 3
Check the slopes of the constraints and profit function in the first
example in the text in Section 5.2.

Another point worth noting here is that the solution of a linear

programming problem will occur, in general, at one of the vertices
of the feasible region if either non-integer solutions are acceptable
or the vertex in question happens to have integral coordinates.
So an alternative to the graphical method of solution would be to
(a) find all the vertices of the feasible region;
(b) find the value of the profit function at each of these vertices;
(c) choose the one which gives maximum value to the profit
function.
91

Chapter 5 Linear Programming

(We have not considered the possibility of non-integer vertices,

which may not make sense in terms of the original problem.)
You can see how this method works with the second example in
Section 5.2. The vertices are given by
(a) 0 (0, 0)
(b) A (0, 18)
(c) B (4, 16)
(d) C (8, 12)
(e) D (16, 0)
and the corresponding profits in are
Point

Profit

2160

2320

2240

1600

As you can see, as you move round the feasible region, the
profit increases from 0 to A to B, but then decreases to C to D
and back to 0.
In more complicated problems, it is helpful to introduce the idea of
slack variables. For the problem above, with the three inequalities

x + y 20
30x + 20y 480
x + 2y 36
three new variables are defined by

r = 20 x y
s = 480 30x 20y
t = 36 x 2y
The three inequalities can now be written as
r 0, s 0, t 0
as well as x 0, y 0. The variables r, s and t are called the slack
variables as they represent the amount of slack between the total
quantity available and how much is being used.
The importance of the slack variables is that you can now define
each vertex in terms of two of the variables, x, y, r, s, or t, being
zero; for example, at A,
x = t = 0.

92

y
22
increasing
20 A p = 2160
profit
18
B p = 2320
16
14
C
p = 2240
12
10
8
6
4
1600
p = 2320
0
p = 2320
2
D x
0
0 2 4 6 8 10 12 14 16 18

Chapter 5 Linear Programming

Activity 4
Complete the table below, defining each vertex
0

x=y=0

t=x=0

................

................

................

The procedure of increasing the profit from one vertex to the

next will be again followed. Starting at the origin

P = 100x + 120y = 0 at x = y = 0
Clearly P will increase in either direction from the origin moving up the y axis means that x is held at zero whilst y
increases. Although you know from the diagram that the next
vertex reached will be A, how could you work that out without a
picture? As x is being kept at zero and y is increasing the next
vertex met will either be where
x = r = 0 or where x = s = 0 or where x = t = 0 . But note that
x = r = 0 y = 20
x = s = 0 y = 24
x = t = 0 y = 18 smallest

and so as y increases the first of these three points it reaches is

the one where y is the smallest, namely y = 18 and the vertex is
where x = t = 0 (which is the point A, thus confirming without a
picture what has already been seen). You can now express the
profit P in terms of x and t:
P = 100x + 120y
= 100x + 120

(36 x t )

2
= 40x 60t + 2160

How can you tell if P will increase as x increases?

So you now increase x, keeping t at zero. You will next meet a
vertex where either t = s = 0, t = r = 0 or t = y = 0 . As x is
increasing the vertex will be the one of those three points where
x is the smallest:

93

Chapter 5 Linear Programming

t=s=0x=6
t = r = 0 x = 4 smallest
t = y = 0 x = 36
So the next vertex reached is where t = r = 0, (which is B) and P
is now expressed in terms of t and r. To do this eliminate y
from the first and third of the original equations by noting that
2r t = 2( 20 x y ) (36 x 2y ) = 4 x
Hence
P = 40x 60t + 2160
= 40( 4 2r + t ) 60t + 2160
= 80r 20t + 2320

How can you tell that P has reached its maximum?

Throughout the feasible region you know that r 0 and t 0
and so it is clear from the negative coefficients in the above
expression for P that P reaches its maximum value of 2320
when r and t are both zero. This happens when
x = 4 and y = 16 .

y
20
A
16

t=0
B

12

s=0

x=0

Activity 5

r=0

y=0

Now travel round the feasible region from 0 to D to C to B. At

each vertex, express P in terms of the defining variables, and
check that P will continue to increase until B is reached.

12

16

Although this probably looks a much more complicated way of

solving linear programming problems, its real application is to
problems of more than 2 variables. These cannot be solved
graphically, but can be solved using a procedure using slack
variables called the simplex method.

Exercise 5C
1. (a) Solve the linear programming problem
maximise P = 2 x + 4 y
subject to x + 5y
4x + y
x
y

10
8
0
0

by a graphical method.
(b) Introduce slack variables r and s, and solve
the problem by the simplex method.

2. (a) Determine the vertices of the feasible region

for the linear programming problem
maximise P = x + y
subject to x + 4 y
2 x + 3y
3x + y
x
y

8
12
9
0
0

Hence find the solution.

(b) Verify this solution by using the simplex method.

94

Chapter 5 Linear Programming

3. Use the simplex method to solve the linear

programming problem
maximise P = 10 x + 15y
subject to 4y +10x 40
10y + 3x 30
5y + 4x 20
x 0
y 0

*5.4 Simplex tableau

The way this method works will be illustrated with the
example.
Maximise

P = x + 2y

subject to x + 4y 20
x+y 8
5x + y 32
x 0
y 0
As usual introduce slack variables r, s and t defined by
x + 4y + r = 20
x+y+s =8
5x + y + t = 32

and write the equations in the matrix form

P x 2y
x + 4y +r
x + y

+s

5x + y

=0
= 20
=8
+t = 32

0
0

P 0

x 20
y =
8
r 32

s

t

95

Chapter 5 Linear Programming

The augmented matrix with the extra right hand column will be used.
P

1
0
0
0

1
1
1
5

2
4
1
1

0
1
0
0

0
0
1
0

0
0
0
1

0
20
8
32

1
0
0
0

1
1
1
1

2
4
1

0
0
1
0

0
0
0

0
20
8

1
5

0
1
0
0

1
5

32
5

R4 / 5

9
5

0
0

0
1

19
5
4
5
1
5

0
0

1
0

32
5
68
5
8
5
32
5

R1 + R 4

1
5
1
5
1
5
1
5

9
5

19
5

32
5
68
5

5
4

1
5

1
5
1
5
1
4
1
5

1
4

10

R1 + 9 R 3
5

R 2 19 R 3
5

3
4
1
4
1
4

9
4
19
4
5
4
1
4

1
4

10

4
3

1
4

1
4

12

R3 + 1 R2
4

R4 1 R2
4

96

9
4
19
3
5
4
1
4

1
3
4
3
1
3
1
3

2
3
19
3
1
3
4
3

Comments

Increase x first (y could have been chosen) and

compare values of x where y = r = 0, y = s = 0,
and y = t = 0 , namely 20 1 , 8 1 , 32 5 * . These
values are easily spotted from the matrix as the
R.H. figures divided by the corresponding
coefficients of x.
(*This term has the smallest positive value so
now manipulate the matrix to express P in
terms of y and t)

R2 R4
R3 R 4

From the first row express P in terms of y and

t, with a positive coefficient of y. Increase y
and compare the values obtained from the
figures in the R.H. column divided by the
corresponding coefficients of y, namely
( 68 / 5 ) / (19 / 5 ), ( 8 / 5 ) / ( 4 / 5 )*,
( 32 / 5 ) / (1 / 5 ) .
(*This term has the smallest positive value
(where s = t = 0 ) so now manipulate the matrix
to express P in terms of s and t)

R 3 / ( 4 5)

32
5

2
6

R 4 1 R3
5

From the first row P could now be expressed

in terms of s and t, with a positive coefficient
of t. So now increase t and compare the
values 6 / ( 3 / 4 ) *, 2 / ( 1 / 4 ), 6 / (1 / 4 )
(*This term has the smallest positive value so
now manipulate the matrix to express P in
terms of r and s)

3
R2 / ( 4 )

R1 + 1 R 2
4

From the first row P could now be expressed

in terms of r and s, with a negative coefficient
of each, so now stop; i.e. since the top row
has all positive coefficients you can see that
the maximum value of P is 12 and that it is
reached when r = s = 0 (which happens when
x = 4 and y = 4 ).

Chapter 5 Linear Programming

The advantage of this method is that it can be readily extended

to problems with more than two variables, as shown below

Example
Maximise P = 4x + 5y + 3z

subject to 8x + 5y + 2z 3
3x + 6y + 9z 2
x, y, z 0
Solution
As usual slack variables r and s are introduced;

8x + 5y + 2z + r = 3
3x + 6y + 9z + s = 2
Now x, y, z, r, s 0 and the simplex tableau is shown below.
P

5
8

2
8

1
8

3
8

5
2

5
8
33
8

2
8
33
4

1
2
1
8
3
8

3
2
3
8
7
8

5
2

5
8

2
8

1
2
1
8
1
11

8
33

3
2
3
8
7
33

3
11
2
11
1
11

20
33
5
33
8
33

67
33
8
33
7
33

0
1

Comments
Increase x initially and compare 3 / 8*, 2 / 3 .
This smaller positive value of x occurs where

y = z = r = 0 and so now manipulate the

matrix to express P in terms of y, z and r.

R2 / 8

R1 + 4R 2
R 3 3R 2

R3 /

Increase y and compare ( 3 / 8 ) / ( 5 / 8 ),

( 7 / 8 ) / ( 33 / 8 ) * . This smaller positive value
occurs when z = r =
in terms of z, r and s.

s=0

and so express P

( 33 )
8

5
R1 + 2 R 3

5
R2 8 R3

The first row now has positive coefficients,

showing that there is a maximum of 67 / 33
when z = r = s = 0 (which happens when
x = 8 / 33, y = 7 / 33 and z = 0 ).

97

Chapter 5 Linear Programming

Exercise 5D
Use the simplex algorithm to solve the following
problems.
1. Maximise P = 4 x + 6 y

3. Maximise P = 3x + 8y 5z
subject to 2 x 3y + z 3
2 x + 5y + 6z
x
y
z

subject to x + y 8
7x + 4 y 14
x 0
y 0

2. Maximise P = 10 x + 12 y + 8z

4. Maximise 3x + 6 y + 2z
subject to 3x + 4 y + 2z 2
x + 3y + 2z 1

subject to 2 x + 2 y 5
5x + 3y + 4z
x
y
z

5.5

5
0
0
0

x 0

15
0
0
0

y 0
z 0

Miscellaneous Exercises

1. Find the solution to Question 3 of Exercise 5A.

2. A firm manufactures two types of box, each
requiring the same amount of material.
They both go through a folding machine and a
stapling machine.
Type A boxes require 4 seconds on the folding
machine and 3 seconds on the stapling machine.
Type B boxes require 2 seconds on the folding
machine and 7 seconds on the stapling machine.
Each machine is available for 1 hour.
There is a profit of 40p on Type A boxes and 30p
on Type B boxes.
How many of each type should be made to
maximise the profit?
3. A small firm which produces radios employs both
skilled workers and apprentices. Its workforce
must not exceed 30 people and it must make at
least 360 radios per week to satisfy demand. On
average a skilled worker can assemble 24 radios
and an apprentice 10 radios per week.
Union regulations state that the number of
apprentices must be less than the number of
skilled workers but more than half of the number
of skilled workers.
What is the greatest number of skilled workers
than can be employed?

Skilled workers are paid 300 a week, and

apprentices 100 a week.
How many of each should be employed to keep
the wage bill as low as possible?
4. In laying out a car park it is decided, in the hope
of making the best use of the available parking
space (7200 sq.ft.), to have some spaces for
small cars, the rest for large cars. For each small
space 90 sq.ft. is allowed, for each large space
120 sq.ft. Every car must occupy a space of the
appropriate size. It is reliably estimated that, of
the cars wishing to park at any given time, the
ratio of small to large will be neither less that
2:3 nor greater than 2:1.
Find the number of spaces of each type in order
to maximise the number of cars that can be
parked.
*5. A contractor hiring earth moving equipment has
the choice of two machines.
Type A costs 25 per day to hire, needs one man
to operate it and moves 30 tonnes of earth per
day.
Type B costs 10 per day to hire, needs four men
to operate it and moves 70 tonnes of earth per
day.
The contractor can spend up to 500 per day, has
a labour force of 64 men available and can use a
maximum of 25 machines on the site.
Find the maximum weight of earth that the
contractor can move in one day.

98

Chapter 5 Linear Programming

6. A landscape designer has 200 to spend on

planting trees and shrubs to landscape an area of
1000 m 2 . For a tree he plans to allow 25 m 2 and
for a shrub 10 m 2. Planting a tree will cost 2
and a shrub 5.
If he plants 30 shrubs what is the maximum
number of trees he can plant?
If he plants 3 shrubs for every tree, what is the
maximum number of trees he can plant?
7. A small mine works two coal seams and
produces three grades of coal. It costs 10 an
hour to work the upper seam, obtaining in that
time 1 tonne of anthracite, 5 tonnes of best
quality coal and 2 tonnes of ordinary coal. The
lower seam is more expensive to work, at a cost
of 15 per hour, but it yields in that time
4 tonnes of anthracite, 6 tonnes of best coal and
1 tonne of ordinary coal. Faced with just one
order, for 8 tonnes of anthracite, 30 tonnes of
best coal and 8 tonnes of ordinary coal, how
many hours should each seam be worked so as to
fill this order as cheaply as possible?
8. A cycle manufacturer produces two types of
mountain-bike: a basic Model X and a Super
Model Y. Model X takes 6 man-hours to make
per unit, while Model Y takes 10 man-hours per
unit. There is a total of 450 man-hours available
per week for the manufacture of the two models.
Due to the difference in demand for the two
models, handling and marketing costs are 20
per unit for Model X, but only 10 per unit for
Model Y. The total funds available for these
purposes are 800 per week.
Profits per unit for Models X and Y are 20 and
30 respectively. The objective is to maximise
weekly profits by optimising the numbers of
each model produced.
(a) The weekly profit is P. The numbers of
units of Model X and Model Y produced each
week are x and y. Express P in terms of x
and y. Also write down inequalities
representing the constraints on production.
(b) By graphical means or by the simplex
method, find the maximum obtainable profit
and the numbers of each model manufactured
which give this profit.
(c) If competition forces the manufacturer to
give a 5 discount on the price of Model X,
resulting in a 5 reduction in profit, how are
weekly profits now maximised?
(AEB)

Tablet

Iron

Calcium Vitamin

Xtravit

Yeastalife

(a) By taking x tablets of Xtravit and y tablets of

Yeastalife the person expects to receive at
least 18 milligrams of iron, 21 milligrams of
calcium and 16 milligrams of vitamins.
Write these conditions down as three
inequalities in x and y.
(b) In a coordinate plane illustrate the region of
those points (x,y) which simultaneously
satisfy x 0, y 0, and the three inequalities
in (a).
(c) If the Xtravit tablets cost 10p each and the
Yeastalife tablets cost 5p each, how many
tablets of each should the person take in
order to satisfy the above requirements at the
minimum cost?
(AEB)
10. A maker of wooden furniture can produce three
different types of furniture: sideboards, tables
and chairs. Two machines are used in the
production - a jigsaw and a lathe.
The manufacture of a sideboard requires 1 hour
on the jigsaw and 2 hours on the lathe; a table
requires 4 hours on the jigsaw and none on the
lathe; a chair requires 2 hours on the jigsaw and
8 hours on the lathe.
The jigsaw can only operate 100 hours per week
and the lathe for 40 hours per week. The profit
made on a sideboard is 100, 40 on a table and
10 on a chair. In order to determine how best
to use the two machines so as to maximise
profits, formulate the problem as a linear
programming problem, and solve it using the
simplex tableau.
11. A diet-conscious housewife wishes to ensure her
family's daily intake of vitamins A, B and C
does not fall below certain levels, say 24 units,
30 units and 18 units, respectively. For this she
relies on two fresh foods which, respectively,
provide 8, 5 and 2 units of vitamins per ounce of
foodstuff and 3, 6 and 9 units per ounce. If the
first foodstuff costs 3p per ounce and the second
only 2p per ounce, use a graphical method to
find how many ounces of each foodstuff should
be bought by the housewife daily in order to
keep her food bill as low as possible.

9. In order to supplement his daily diet someone

wishes to take some Xtravit and some Yeastalife
tablets. Their contents of iron, calcium and
vitamins (in milligrams per tablet) are shown in
the table.

99

Chapter 5 Linear Programming

100