Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.

edu/em8719

PERFORMANCE EXCELLENCE EM 8719-E

IN THE WOOD PRODUCTS INDUSTRY October 1998
$2.50

Using the Graphical Method

to Solve Linear Programs
J. Reeb and S. Leavengood

A key problem faced by managers is how to

allocate scarce resources among activities or About this series
projects. Linear programming, or LP, is a method of According to the Operations
allocating resources in an optimal way. It is one of Research Society of America,
the most widely used operations research (OR) “Operations research [OR] is
tools. It has been used successfully as a decision- concerned with scientifically
making aid in almost all industries, and in financial deciding how to best design
and service organizations. and operate man–machine
Programming refers to mathematical program- systems, usually under con-
ming. In this context, it refers to a planning process ditions requiring the allocation
that allocates resources—labor, materials, machines, of scarce resources.”
and capital—in the best possible (optimal) way so This publication, part of a
that costs are minimized or profits are maximized. series, should be useful for
In LP, these resources are known as decision vari- supervisors, lead people,
ables. The criterion for selecting the best values of middle managers, and anyone
the decision variables (e.g., to maximize profits or who has planning responsibility
minimize costs) is known as the objective function. for either a single manufactur-
The limitations on resource availability form what ing facility or for corporate
is known as a constraint set. planning over multiple facilities.
For example, let’s say a furniture manufacturer Although managers and
produces wooden tables and chairs. Unit profit for planners in other industries can
tables is $6, and unit profit for chairs is $8. To learn about OR techniques
through this series, practical
simplify our discussion, let’s assume the only two
examples are geared toward the
resources the company uses to produce tables and
wood products industry.
chairs are wood (board feet) and labor (hours). It
takes 30 bf and 5 hours to make a table, and 20 bf
and 10 hours to make a chair. There are 300 bf of
wood available and 110 hours of labor available.
The company wishes to maximize profit, so profit
maximization becomes the objective function. The

James E. Reeb, Extension forest products manufac-

turing specialist; and Scott Leavengood, Extension
agent, Klamath County; Oregon State University.
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

resources (wood and labor) are the decision variables. The limita-
Decision variables. . . tions on resource availability (300 bf of wood and 110 hours of
labor) form the constraint set, or operating rules that govern the
“The resources process. Using LP, management can decide how to allocate the
available.” limited resources to maximize profits.
The “linear” part of the name refers to the following:
Constraint set. . .
• The objective function (i.e., maximization or minimization) can
“The limitations on be described by a linear function of the decision variables, that
resource availability.” is, a mathematical function involving only the first powers of the
variables with no cross products. For example, 23X2 and 4X16 are
Objective function. . .
valid decision variable terms, while 23X22, 4X163, and (4X1 * 2X1)
“The criterion for are not. The entire problem can be expressed as straight lines,
selecting the best planes, or similar geometrical figures.
values of the decision • The constraint set can be expressed as a set of linear equations.
variables.” In addition to the linear requirements, nonnegativity conditions
state that the variables cannot assume negative values. It is not
possible to have negative resources. Without these conditions, it
would be mathematically possible to use more resources than are
available.
This publication will introduce a small LP problem that can be
solved graphically. In other words, we’ll plot the appropriate
information on a graph, and then use the graph to find a solution to
the problem.
In EM 8720, Using the Simplex Method to Solve Linear Pro-
gramming Maximization Problems, we’ll build on the graphical
example and introduce an algebraic technique known as the sim-
plex method. This method lets us solve very large LP problems that
would be impossible to solve graphically or without the analytical
ability of a computer. In other publications, we’ll define several
special types of LP and use forest products examples to help
explain their functions.
For more complex problems, many software packages are
available for obtaining a solution. It’s just a matter of choosing the
one you wish to work with. In a later publication, we’ll use com-
mercial software to solve larger LP problems and to examine how
to use sensitivity analysis to gain more information than that
provided by a simple optimal solution.

2
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

Steps in solving an LP problem graphically

The steps in solving an LP problem graphically are introduced
briefly below. In the remainder of this publication, we’ll apply
these steps to a simple LP problem.

Step 1. Formulate the LP problem. Formulation refers

to translating the real-world problem into a format of
mathematical equations that represent the objective
function and the constraint set. Often, data gathering,
problem definition, and problem formulation are the
most important (and the most difficult, time-
consuming, and expensive) steps when using any OR
tool.
A thorough understanding of the problem is necessary
in order to formulate it correctly. During the formulation stage,
an OR specialist may discover new insights into the problem that
may change the scope of the original problem. For this reason, it’s
very important to obtain the help of those who work most closely
with the system being studied. They can help the OR specialist
collect the necessary data and correctly define the problem so that
he or she can create a valid model. In formulating an LP problem,
it often helps to put all of the relevant information into a table (for
example, see Table 1, page 4).

Step 2. Construct a graph and plot the constraint lines.

Constraint lines represent the limitations on available resources.
Usually, constraint lines are drawn by connecting the horizontal
and vertical intercepts found from each constraint equation.

Step 3. Determine the valid side of each constraint line. The

simplest way to start is to plug in the coordinates of the origin
(0,0) and see whether this point satisfies the constraint. If it does,
then all points on the origin side of the line are feasible (valid),
and all points on the other side of the line are infeasible (invalid). 110
=
If (0,0) does not satisfy the constraint, then all points on the other
0 X2
side and away from the origin are feasible (valid), and all points
) +1
on the origin side of the constraint line are infeasible (invalid). 5(4
There are two exceptions, which we will discuss later. 30(4) +
20X = 300
2
Step 4. Identify the feasible solution region. The feasible solution
region represents the area on the graph that is valid for all
4 ) - X = 0
constraints. Choosing any point in this area will result in a valid 4(2.4 4 2

solution.

3
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

Step 5. Plot two objective function lines to determine the direction

of improvement. Improvement is in the direction of greater value
when the objective is to maximize the objective function, and is in
the direction of lesser value when the objective is to minimize the
objective function. The objective function lines do not have to
include any of the feasible region to determine the desirable
direction to move.

Step 6. Find the most attractive corner. Optimal solutions always

occur at corners. The most attractive corner is the last point in the
feasible solution region touched by a line that is parallel to the two
objective function lines drawn in step 5 above. When more than
one corner corresponds to an optimal solution, each corner and all
points along the line connecting the corners correspond to optimal
solutions. We’ll use an example to illustrate optimal solutions later.

Step 7. Determine the optimal solution by algebraically

calculating coordinates of the most attractive corner.

Step 8. Determine the value of the objective function for the

optimal solution.

An LP example: Production of wooden

tables and chairs
Let’s look at the profit maximization problem that our furniture
manufacturer faces. The company uses wood and labor to produce
tables and chairs. Recall that unit profit for tables is $6, and unit
profit for chairs is $8. There are 300 board feet (bf) of wood
available, and 110 hours of labor available. It takes 30 bf and
5 hours to make a table, and 20 bf and 10 hours to make a chair.
Table 1 contains the information for the LP problem. We will go
through the step-by-step process of solving this problem graphi-
cally.

Table 1.—Information for the wooden tables and chairs linear programming
problem.
Resource Table (X1) Chair (X2) Available
Wood (bf) 30 20 300
Labor (hr) 5 10 110
Unit profit $6 $8

4
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

Step 1. Formulate the LP problem.

Table 1 will help us formulate the problem. The bottom row is Tip . .
used to formulate the objective function. Objective functions are In our example,
developed in such a way as to be either maximized or minimized. In X1 refers to tables,
this case, the company’s management wishes to maximize unit X2 refers to chairs, and
profit. The wood and labor rows are used to formulate the constraint Z refers to profit.
set. Finally, the nonnegativity conditions are stated.
Maximize: Z = 6X1 + 8X2 (objective function)

Subject to: 30X1 + 20X2 < 300 (wood constraint: 300 bf available)

5X1 + 10X2 < 110 (labor constraint: 110 hours available)

X1, X2 > 0 (nonnegativity conditions)

Since only two variables (wood and labor) exist in this problem,
it can be solved graphically. If there were more than two variables,
the graph would have to be more than two dimensions.

Step 2. Construct the graph and plot constraint lines.

Draw the graph with the x axis representing the number of tables
and the y axis representing the number of chairs. Plot the two
constraint lines by finding the x and y intercepts for the two con-
straint equations in the following manner.
First, rewrite the constraint inequalities as equalities and solve to
obtain the intercepts:
Wood Labor
30X1 + 20X2 = 300 5X1 + 10X2 = 110

Set X2 = 0 and solve for X1 Set X2 = 0 and solve for X1

30X1 = 300 5X1 = 110

X1 = 300/30 X1 = 110/5

= 10 tables = 22 tables
(All the wood is used to make tables.) (All the labor is used to make tables.)

Next:
Set X1 = 0 and solve for X2 Set X1 = 0 and solve for X2

20X2 = 300 10X2 = 110

X2 = 300/20 X2 = 110/10

= 15 chairs = 11 chairs
(All the wood is used to make chairs.) (All the labor is used to make chairs.)

5
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

Chairs Now plot the wood

constraint line, using the
intercepts X1 = 10 and
X2 = 15. Plot the labor
constraint line, using the
intercepts X1 = 22 and
X2 = 11. See Figure 1.

Step 3. Determine the

valid side of each
constraint line.
Tables
We will use the origin
(0,0) to check the valid side
Figure 1.—Wood and labor constraint lines. for both constraint lines.
30(0) + 20(0) < 300 is
valid, so we know the side toward the origin (0,0) is the valid side
of the wood constraint line. 5(0) + 10(0) < 110 also is valid, so we
know the side toward the origin (0,0) is the valid side of the labor
constraint line. We can draw arrows indicating the valid side of
each constraint line. See Figure 2.
We could have chosen any point to test for the valid side of the
line. For example, setting X1 = 20 and X2 = 10 (clearly on the other
side, away from the origin) for the wood constraint line, we get
30(20) + 20(10) < 300, which is not valid. In other words, there
simply isn’t enough wood to make 20 tables and 10 chairs.

Chairs

Wood

Labor

Tables

Figure 2.—Identification of the feasible region.

6
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

Step 4. Identify the feasible region.

The feasible region is the area on the valid side of both con-
straint lines. Any point located on the invalid side of a constraint
line is infeasible. Because of the nonnegativity conditions, the
feasible region is restricted to the positive quadrant. See Figure 2.

Step 5. Plot two objective function lines to determine

the direction of improvement.
First, we’ll arbitrarily set profit, Z = 48, and then set profit,
Z = 72. We’ll find the x and y intercepts when Z = 48 and when
Z = 72, and plot the two lines.
Set Z = 48 Set Z = 72

Set X2 = 0 and solve for X1 Set X2 = 0 and solve for X1

48 = 6(X1) 72 = 6(X1)

48/6 = X1 72/6 = X1

X1 = 8 X1 = 12

Next:
Set X1 = 0 and solve for X2 Set X1 = 0 and solve for X2

48 = 8(X2) 72 = 8(X2)

48/8 = X2 72/8 = X2

X2 = 6 X2 = 9

Now plot the objective

function lines when Z = 48 Chairs
Wood
and Z = 72. See Figure 3.
We can see from the two
objective function lines
that as we move away Z = 72
from the origin (0,0), Labor

Z increases.
Z = 48

Tables

Figure 3.—Determining direction of increasing value.

7
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

Chairs

Wood

Labor

Tables

Figure 4.—Locating the most attractive corner.

Step 6. Find the most attractive corner.

Since we want to maximize Z, we will draw a line parallel to the
objective function lines that touches the last point in the feasible
region while moving away from the origin. This identifies the most
attractive corner, which gives us the amounts of wood and labor that
will result in the maximum profit (maximize Z). Thus, it represents
the optimal solution to the problem (Figure 4).

Step 7. Determine the optimal solution by algebraically

calculating coordinates of the most attractive corner.
The most attractive corner lies at the intersection of the wood and
labor constraint lines. Therefore, coordinates for the most attractive
corner can be found by simultaneously solving the constraint equa-
tions (wood and labor):
30X1 + 20X2 = 300 (wood)

5X1 + 10X2 = 110 (labor)

To do so, multiply the labor equation by -2 and add it to the wood

equation so the X2 variable becomes zero and we can solve for X1.
30X1 + 20X2 = 300 (wood)
-2(5X1 + 10X2 = 110) (labor)

20X1 + 0 = 80

X1 = 4 tables

8
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

Next, substitute the number of tables calculated above into either

of the constraint equations to find the number of chairs. For this
example, we will substitute into both equations to illustrate that the
same value is found.
Wood constraint Labor constraint
30(4) + 20X2 = 300 5(4) + 10X2 = 110

120 + 20X2 = 300 20 + 10X2 = 110

20X2 = 300 - 120 10X2 = 110 - 20

20X2 = 180 10X2 = 90

X2 = 180/20 X2 = 90/10

X2 = 9 chairs X2 = 9 chairs

Thus, the company’s optimal solution is to make four tables and

nine chairs. In this case, you could read this solution off the graph
(Figure 4) by finding the values on the x and y axes corresponding
to the most attractive corner. However, when the most attractive
corner corresponds to an optimal solution with fractions, it is not
possible to read directly from the graph. For example, the optimal
solution to this problem might have been 3.8 tables and 9.2 chairs,
which we probably would not be able to read accurately from the
graph.

Step 8. Determine the value of the objective function for

the optimal solution.
Plug in the number of tables and chairs and solve for Z:
Z = $6(4) + $8(9) = $96

Thus, we find that maximum profit of $96 can be obtained by

producing four tables and nine chairs.

9
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

Other types of constraints

Mixture constraint. . .
“A special constraint Mixture constraints
occurring when a A constraint line that passes through the origin (0,0) represents a
resource must be used special kind of constraint known as a mixture constraint. Mixture
in a fixed ratio with constraints arise when a resource must be used in a fixed ratio with
other resources.” other resources.
We can add a mixture constraint to our example by saying that
tables and chairs are sold in sets of one table with four chairs. Some
extra chairs can be made. We’ll keep the other labor and wood
constraints and add the following new constraint:
4X1 < X2 (The number of chairs must be at least 4 times the number of tables.)

To find the x and y intercepts of this new constraint line, first

move all variables to the left side of the inequality:
4X1 - X2 < 0

Next, rewrite the inequality as an equality and solve to obtain the

intercepts:
4X1 - X2 = 0

When X1 = 0, then X2 also is 0, so we see that the constraint line

passes through the origin (0,0). In other words, if the company
doesn’t make any tables, it can’t make any chairs and vice versa.
Now we can arbitrarily choose other values and solve the equation
to find other points on the constraint line. If X1 = 3, then substituting
3 into the equation we find:
4(3) - X2 = 0
Chairs
Mixture Constraint 12 - X2 = 0

X2 = 12
A

Thus, if the company

B makes 3 tables, it needs to
make 12 chairs. Point A
represents where X1 = 3 and
X2 = 12 in Figure 5.

Tables

Figure 5.—Determining the mixture constraint line.

10
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

We will plot one more point for the constraint line by arbitrarily
setting X1 = 2 and solving for X2. We find:
4(2) - X2 = 0

8 - X2 = 0

X2 = 8

Point B represents where X1 = 2 and X2 = 8. So, if the company

makes two tables, it needs to make eight chairs.
We now can draw the mixture constraint line on the graph. See
Figure 5.
To evaluate the valid side of the mixture constraint line, evaluate
a point not on the line. We will arbitrarily choose X1 = 5 and
X2 = 10 (Figure 6a). Substitute the values into the equation:
4(5) - 10 < 0

This statement is false, Chairs

so this is not the valid

side of the mixture
constraint line.
When we substitute the
values for a point on the
opposite side of the
mixture constraint line, Wood
Labor

we get a true statement.

For example, let X1 = 2
and X2 = 15 (Figure 6b).
Tables
Substituting the values
into the equation:
4(2) - 15 < 0
Chairs

This statement is true,

thus signifying the valid
side of the constraint line.

Labor

Wood

Tables

Figure 6a and 6b.—To determine the valid side of the mixture constraint line,
evaluate a point not on the line.

11
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

Chairs The boundary of the

feasible region is defined
by the mixture constraint
Optimal solution and the labor constraint
lines illustrated with the
dark line in Figure 7. As
before, plot two objective
function lines to determine
the direction of increase.
Draw a line parallel to the
two objective function lines
Tables that touches the last point
in the feasible region
Figure 7.—Locating the feasible region and the most attractive corner. In this (going in the direction of
case, the optimal solution is where the labor and mixture constraint lines increasing value). The most
intersect. attractive corner lies at the
intersection of the mixture
constraint and labor constraint lines (Figure 7).
The next step is to simultaneously solve the mixture and labor
equations by multiplying the mixture constraint equation by 10
and adding it to the labor constraint equation so the X2 variable
becomes 0 and we can solve for X1:
5X1 + 10X2 = 110 (labor constraint)
10(4X1 - X2 = 0) (mixture constraint)

45X1 + 0 = 110

X1 = 110/45

X1 = 2.444 (number of tables)

Now, solve for X2:

4X1 - X2 = 0

4(2.444) - X2 = 0

9.777 - X2 = 0

- X2 = - 9.777

X2 = 9.777 (number of chairs)

This results in a non-integer (fractional) solution. Non-integer

solutions are discussed later in this publication. Meanwhile, we can
view the 0.444 tables and 0.777 chairs as work in progress.
From observing the graph, we see that the wood constraint line
does not border the feasible region. This indicates that some wood
12
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

was not used in the optimal solution. The unused wood can be
calculated as:
30 bf(2.444) + 20(9.777) = 268.88 bf of wood used out of 300 bf available

300 bf - 268.88 bf = 31.12 bf of wood unused

The unused wood can be thought of as surplus or slack. In

EM 8720, Using the Simplex Method to Solve Linear Program-
ming Maximization Problems, we’ll use this concept in solving LP
problems with the simplex method.
Note that it would be difficult to read this fractional optimal
solution directly from the graph. Thus, this example is one where
simultaneously solving the equations representing the intersecting
lines at the most attractive corner is the most practical method for
determining the optimal solution.

Equality constraints
Constraints can take the form of an equality. For example, the
company may decide to sell all of its tables and chairs in sets (one
table and four chairs). In this case, the new constraint is:
4X1 = X2
or
4X1 - X2 = 0

In this case, there is no valid side of the mixture line. Valid

points lie exactly on the constraint line. The feasible region con-
sists of the segment of the constraint line that meets the conditions
of the other constraints, i.e., the part of the equality constraint line
from the origin (0,0) to the point where the equality constraint line
meets the labor constraint line (Figure 8).

Chairs

Equality constraint. . .
“A special constraint
that causes all valid
points to lie exactly on
the constraint line.”
Tables

Figure 8.—Determining the feasible region for an equality constraint problem.

13
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

Chairs We set Z = 48 and

Z = 72 in the same manner
as above to see that Z
Most attractive corner
increases in value as we
move away from the origin.
Drawing a line parallel to
these objective function
lines, we find the most
attractive corner is where
the equality constraint line
Tables
and the labor constraint line
Figure 9.—Identifying the most attractive corner for an equality constraint
intersect (Figure 9).
problem. In this case, by simulta-
neously solving the equal-
ity and labor constraint equations, we get the same solution (and
equation set) as in the mixture constraint example above.
If there is more than one equality constraint, the optimal corner
is where the equality constraint lines intersect. We can add another
equality constraint to our tables and chairs example by saying that
exactly two tables need to be produced. This constraint is graphed
with a vertical line intersecting the x axis at 2 (Figure 10). The
optimal solution is where the table equality constraint and the set
(four chairs for each table) constraint lines intersect. Substituting
2 into the equation gives:
4X1 - X2 = 0

4(2) = X2

X2 = 8 chairs

The optimal solution to the LP is:

Z = $6(2) + $8(8) = $76 unit profit (Figure 10)

Chairs

Tables

Figure 10.—Two equality constraints.

14
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

Multiple or alternative optimal solutions

Let’s consider another maximization problem, unrelated to the With multiple
tables and chairs problem, to illustrate multiple, sometimes called resources that can be
alternative, optimal solutions. combined in various
ways, it’s possible
Step 1. Let Z = profit, and formulate the LP problem as: that more than one
Maximize Z = 10X1 + 12X2 resource mix will lead
to equally optimal
Subject to 5X1 + 6X2 < 60 (resource one) solutions.
8X1 + 4X2 < 72 (resource two)

3X1 + 5X2 < 45 (resource three)

X1, X2 > 0

In this case, there are multiple resources that can be combined in

various ways. Our goal is to find the combination of resources that
will maximize profits. It’s possible that more than one resource
mix will lead to equally optimal solutions.

Steps 2, 3, and 4. Construct the graph and plot the constraint lines.
Determine the valid side of each constraint line. Identify the
feasible region (Figure 11).

Chairs

Resource 2

Resource 3
Feasible
region Resource 1

Tables

Figure 11.—Identifying the feasible region with multiple possible resource mixes.

15
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

Chairs Steps 5 and 6. Plot two

objective function lines and
determine the direction of
increasing value (if a
minimization problem, we
would determine the
direction of decreasing
value). Identify the most
Z
Z
attractive corner. For this
example, we find that there
Tables actually are two most
attractive corners, where
Figure 12.—Determining the most attractive corner with multiple optimal
solutions. resource one’s constraint
line intersects resource
two’s constraint line, and where resource one’s constraint line
intersects resource three’s constraint line (Figure 12).

Step 7. Determine the optimal solution by algebraically calculating

coordinates of the most attractive corners. By simultaneously
solving both equation pairs for both points, we can choose the
corner that corresponds to the highest objective function value
(since this is a maximization problem). Solve for X1:
5X1 + 6X2 = 60 (resource one)

-6/4(8X1 + 4X2 = 72) (resource two)

-7X1 + 0 = -48

1/7(-7X1 + 0) = -48

-X1 = -48/7

X1 = 6.86

Solve for X2 by substituting into a constraint equation:

5(6.86) + 6X2 = 60

34.28 + 6X2 = 60

6X2 = 60 - 34.28

6X2 = 25.71

X2 = 25.71/6

X2 = 4.29

16
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

Solve for Z in the objective function:

Z = 10(6.86) + 12(4.29) = 120

Solve for X1:

5X1 + 6X2 = 60 (resource one)

-6/5(3X1 + 5X2 = 45) (resource three)

1.4X1 + 0 =6

X1 = 6/1.4

X1 = 4.29

Solve for X2 by substituting into a constraint equation:

5(4.29) + 6X2 = 60

21.45 + 6X2 = 60
6X2 = 60 - 21.45
6X2 = 38.55

X2 = 38.55/6
X2 = 6.425

Solve for Z in the objective function:

Z = 10(4.29) + 12(6.425) = 120

Both points are equally attractive, so there are multiple optimal

solutions. As stated earlier, when more than one corner corre-
sponds to an optimal solution, each corner and all points along the
connecting line segment will correspond to optimal solutions.
Whenever all coefficients of one equation are the same multiple
of another equation’s coefficients, the lines are parallel. When the
objective function line is parallel to a constraint line, there are
multiple optimal solutions. In our example:
Z = 10X1 + 12X2 (objective function equation)

5X1 + 6X2 = 60 (constraint one equation)

The ratio of X1 terms (10 and 5) is 10/5 or 2. The ratio of X2

terms (12 and 6) is 12/6, also 2. Thus, the two lines are
parallel and multiple optimal solutions exist. Multiple
optimal solutions mean that more than one mix of
resources leads to an optimal solution. In real-world
problems, there may be reasons why one mix of
resources would be easier to obtain than another, or there
may be no reason to choose one solution over another.
17
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

Unbounded solutions
An unbounded solution occurs when there are unlimited
resources. A good way to explain unbounded solutions is to use a
small problem that can be solved graphically. Consider the follow-
ing LP problem:
Maximize: Z = X1 + 2X2

Subject to: X1 + X2 > 2

X2 < 4

X1, X2 > 0

We can graph this problem as in Figure 13.

The constraint line intersecting the X2 axis at 4 can increase
infinitely along the X1 axis. Thus, the feasible region is unbounded.
The objective function can be improved infinitely by adding
infinitely more resources. For real-life problems, of course, it is
impossible to have unlimited resources. Thus, an unbounded
solution signifies that the LP problem has been formulated incor-
rectly. Usually, one or more constraints has been omitted.

Chairs

Unbounded solutions. . .
“A solution to a prob-
lem with unlimited
resources.” In the real
world, such a situation Tables

is impossible.
Figure 13.—Example of an unbounded solution, in which X1 can be increased
infinitely to increase the objective function.

18
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

Limitations to linear programming

As we saw in an earlier example, there is no guarantee of an
integer (whole number) solution. The furniture company can’t sell
0.444 of a table. We thought of this as work in progress for this
particular problem. An optimal solution for another problem might
be to buy 6.75 trucks. Obviously, you can buy either 6 or 7 trucks,
but not 0.75 of a truck. In this case, rounding up (and for many
non-integer answers, rounding either up or down) offers a practical
solution. However, if you’re studying the possibility of building a
new plant addition, a variable can take on the value of 0 or 1, so a
fractional answer is useless. Either you build the addition or you
don’t. Fortunately, methods called integer programming can
handle these types of problems.
When using LP, there is no way to deal with uncertainty. We
formulate the problem with the assumption that we know all of the
values for costs, profits, constraints, etc. In real-life problems,
these variables often are unknowns. Methods such as chance
constrained programming or linear programming under uncer-
tainty can be used in these situations.
Another limitation is the assumption of linearity. Often, the
objective and constraints are not related linearly to the variables. In
some problems, a nonlinear relationship can be forced to be a
linear relationship without losing too much integrity in the original
problem. For other problems, nonlinear programming techniques
are available.
Integer programming, linear programming under uncertainty,
and nonlinear programming are beyond the scope of these publica-

??
tions. If you wish to learn more about these topics, see the
resources listed in the “For more information” section.

?
19
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

OPERATIONS RESEARCH

For more information

Bierman, H., C.P. Bonini, and W.H. Hausman. Quantitative Analy-
sis for Business Decisions (Richard D. Irwin, Inc., Homewood,
IL, 1977). 642 pp.
Dykstra, D.P. Mathematical Programming for Natural Resource
Management (McGraw-Hill, Inc., New York, 1984). 318 pp.
Hillier, F.S., and G.J. Lieberman. Introduction to Operations
Research, sixth edition (McGraw-Hill, Inc., New York, 1995).
998 pp.
Ignizio, J.P., J.N.D. Gupta, and G.R. McNichols. Operations
Research in Decision Making (Crane, Russak & Company, Inc.,
New York, 1975). 343 pp.
Lapin, L.L. Quantitative Methods for Business Decisions with
Cases, third edition (Harcourt Brace, Jovanovich, Publishers,
San Diego, 1985). 780 pp.
Ravindran, A., D.T. Phillips, and J.J. Solberg. Operations
Research: Principles and Practice, second edition (John Wiley
& Sons, New York, 1987). 637 pp.

20
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

GRAPHICAL METHOD

PERFORMANCE EXCELLENCE
IN THE WOOD PRODUCTS INDUSTRY

ABOUT THIS SERIES

This publication is part of a series, Performance
Excellence in the Wood Products Industry. The various
publications address topics under the headings of wood
technology, marketing and business management,
production management, quality and process control,
and operations research.
To view and download any of the other titles in the
series, visit the OSU Extension Web site at http://
eesc.oregonstate.edu/ then “Publications & Videos” then
“Forestry” then “Wood Processing” and “Business
Management”. Or, visit the OSU Wood Products Extension
Web site at http://wood.oregonstate.edu/

21
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719
Archival copy. For current information, see the OSU Extension Catalog: https://catalog.extension.oregonstate.edu/em8719

© 1998 Oregon State University

This publication was produced and distributed in furtherance of the Acts of Congress of May 8 and June 30, 1914. Extension
work is a cooperative program of Oregon State University, the U.S. Department of Agriculture, and Oregon counties. Oregon
State University Extension Service offers educational programs, activities, and materials—without regard to race, color, religion,
sex, sexual orientation, national origin, age, marital status, disability, and disabled veteran or Vietnam-era veteran status—as
required by Title VI of the Civil Rights Act of 1964, Title IX of the Education Amendments of 1972, and Section 504 of the
Rehabilitation Act of 1973. Oregon State University Extension Service is an Equal Opportunity Employer.
Published October 1998.