CHAPTER FIVE

THE ENGINEERING
APPROACH TO
P R O B L E M S O LV I N G

Unlike problems usually assigned to engineering students, real-world prob-

lems are often unstructured and open-ended. At times, not all of the required
data are known or available. In other instances, it is necessary to sort through
a plethora of information and identify which parts of it are needed to solve the
problem under consideration.
Novice engineers are sometimes surprised to discover that a problem may not
have a single categorical solution. Often, the objective is to select a preferred
solution from among several alternatives. It may be necessary to weigh several
conflicting consequences of an engineering action and then select the solution
that best meets the needs and desires of an employer, a client, or the public.

T H E N AT U R E O F E N G I N E E R I N G D E S I G N

Engineering work often involves planning and analysis in the initial stages, but
the essence of engineering problem solving is design. Engineering design is as
varied as the engineering profession, and it is as broad as the problems facing
humankind. An engineer’s designs may be as small and intricate as a
microchip for a computer system or as large and complex as a space shuttle. To
perform engineering design is to conceive, imagine, devise, and plan a device,
a structure, a process, or a system that will benefit people.
As a profession, engineering is relatively young. In earlier times, the design
process was the domain of the craftsman, and the art and knowledge of design
was passed from master craftsman to apprentice. Since the birth of engineering as

111
112 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

a profession, the problems have grown exceedingly in complexity and diversity.

At the same time, advancements in technology have greatly expanded the abil-
ity of engineers to improve the comfort and well-being of the people they serve.
With advancements in technological knowledge have come more formal
institutions and procedures for the transfer of knowledge. Modern engineers
must not only gain experience under the tutelage of other engineers but must
also be educated in formal college- or university-level programs of study.
Engineering design has become more varied and challenging, requiring a
greater degree of specialization as well as a need for teamwork. It is not uncom-
mon, therefore, for many large engineering projects to be carried out by dozens
or even hundreds of engineering specialists.

Engineering Teams
Engineering teams may be used as simply a means of dividing up the work-
load. Similarly-trained engineers may be assigned specific design tasks by an
engineering manager. In other instances, several engineers contribute special-
ized knowledge and provide organizational competence to the solution of a
problem where the knowledge of any one engineer would be inadequate to
solve it.
When complex problems are faced, it may be desirable to establish a non-
traditional organizational structure in which team members have a dual report-
ing relationship to a functional manager or administrative supervisor as well as
to a team leader. The design of an industrial plant, for example, may require the
knowledge and skills of several types of engineers: industrial, chemical, civil,
mechanical, and electrical. Members from each of the specialties could be
assigned to a task force or a design team for the duration of the design process
while reporting to a functional manager in their respective specialty as well as
to the project engineer. Such design teams are often temporary and appointed
for a specific purpose, then dissolved when the work is completed. (1)

THE ENGINEERING METHOD

The nature of problems that must be solved by engineers varies both between
and among the various branches of engineering. Indeed, an individual engineer
may face a variety of problems during the course of his or her daily work activ-
ities. Because of the variability of engineering designs, there is no definitive
procedure or list of steps that will always fit the engineering problems at hand.
However, engineers tend to deal with problems in a special way. Certainly, the
engineering method of approaching and solving problems differs greatly from
that of most other professionals.1 Engineers are trained to think in analytical
and objective terms and to approach problems methodically and systematically.

1 It is, however, similar to the better-known “scientific method.”

 5.1 IDENTIFICATION OF THE PROBLEM 113

A number of engineering writers have set forth a list of steps or phases that
comprise the “engineering design method.” Typically, the list includes:

1. Identification of the problem.

2. Gathering needed information.
3. Searching for creative solutions.
4. Stepping from ideation to preliminary designs (including modeling).
5. Evaluation and selection of preferred solution.
6. Preparation of reports, plans, and specifications.
7. Implementation of the design.

As we describe these steps, it is important to keep in mind that in many

instances, one or more of the steps may not appear. In other cases, it may be
necessary to repeat the entire protocol several times in an attempt to converge
on a desired solution. Let us now examine the various steps or phases of engi-
neering design.

5.1 IDENTIFICATION OF THE PROBLEM

There is a tendency to think that this phase of the solution process is trivial
and unimportant. Such is not the case. An incorrect or improper definition of
the problem will cause the engineer to waste time and may lead to a solution
that is inappropriate or incorrect. Pearson ((2)) states: “A problem properly
defined is a problem partially solved. To state the problem correctly is a major
step toward its solution.”
It is important that the stated needs be real needs. A truly great design may
be worthless if it duplicates other known designs or if it solves a problem that
does not impact many people. If it is a product that is being designed, it may be
difficult to predict the mass appeal and the resulting marketability of a proposed
design. A preliminary market analysis will usually identify the prospective
users and statistics about comparable devices or methods and volume of sales.
The needs to be satisfied should be broadly defined and distinguished from
possible solutions. In this phase, care should be taken not to prejudice the solu-
tion by incorrectly defining the problem. Consider the following example.
For decades in the United States, deaths of drivers and passengers in motor
vehicle crashes were considered to be one of the nation’s most serious public
health problems. Traffic safety specialists defined the problem in terms of acci-
dent prevention rather than loss reduction. They spoke in terms of a need to pre-
vent accidents rather than a need to reduce losses from accidents. In so doing,
they presupposed the problem solution and focused their attention exclusively
on driver behavior. Predictably, their “solution” to the problem was driver edu-
cation, traffic enforcement, and “Drive Safely” campaigns. They overlooked pos-
sible benefits from more crashworthy vehicles and a safer roadway environment.
To the extent possible, the problem should be defined in objective terms.
Here is an example of a problem that is defined in objective terms.
114 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

Design an energy attenuation system that will control the energy of a crash
of a 2500-pound car traveling 60 miles per hour at impact. The device should
not be longer than 10 feet and should cost no more than $10,000 per unit. The
deceleration should not exceed 6 gs (193 ft/s2).
Contrast this definition with the following situation:
Design an energy attenuation system that will control the energy of a car
traveling at a fast speed at impact. The device should be short and inexpensive
to build. The deceleration should not be harmful to the driver.
Finally, the problem should not be unnecessarily constrained. If too many
constraints are placed on the problem, it may make its solution extremely dif-
ficult or even impossible. In fact, a careful examination of the example prob-
lem stated above in light of Newton’s second law of mechanics shows that it is
overconstrained. The device would need to be longer than 10 feet to meet the
other conditions of the problem.

5.2 GATHERING NEEDED INFORMATION

Once the problem is identified and the needs properly defined, the engineer
then begins to gather information and data needed to solve it. The type of infor-
mation needed will, of course, depend on the nature of the problem to be
solved. It could be physical measurements, maps, results of laboratory experi-
ments, patents, results of opinion surveys, or any of a number of other types of
information. This phase of the problem-solving process involves gathering and
evaluating information that is already available. If the engineer is employed by
a large corporation or a public agency, it will probably be desirable to search
old files and interview other employees to see if others have undertaken simi-
lar work. Subsequently, it may be necessary to supplement this information by
making additional measurements or conducting more laboratory experiments,
opinion surveys, and the like.
In this phase of the process, engineers typically undertake a literature search
to determine what others have learned about related problems. They may visit
technical libraries and study textbooks, journal articles, and manufacturers’
catalogs. Librarians can be very helpful in locating textbooks and journal ref-
erences, and most libraries now have access to computer-aided searching ser-
vices that are fast and relatively inexpensive. Some libraries also maintain
manufacturers’ catalogs reduced onto microfiche film with a subject index to
help in locating already-manufactured components.
Finally, it is worthwhile to perform a patent search, especially if there is a
patent library nearby. Most “mouse traps” never make it to market or to press;
consequently, there are many great designs described in intricate legal detail,
which can be found in the chronological listing of U.S. patents.

5.3 SEARCHING FOR CREATIVE SOLUTIONS

After completing the preparatory steps in the design process, the engineer is
ready to begin identifying creative solutions. Actually, the development of new
 5.3 SEARCHING FOR CREATIVE SOLUTIONS 115

ideas, products, or devices may result from creativity,, a subconscious effort, or

from innovation, a conscious effort.
There are several operational techniques that may be used to help a group
or individual to produce original ideas. These techniques are designed to
enable the group or individual to overcome obstacles to creative thinking such
as those described in Chapter 4.

Brainstorming
One of the most popular techniques for group problem solving is brainstorm-
ing. Typically, a brainstorming session consists of 6 to 12 people who sponta-
neously introduce ideas designed to solve a specific problem. In these sessions,
all ideas are encouraged, including those that appear to be completely imprac-
tical. Efforts are made to generate as many ideas as possible. Participants are
encouraged to combine or improve on ideas of others. Judgment and evaluation
of the ideas are not permitted in the idea-producing session.
It is suggested that participants in brainstorming sessions be chosen from a
diversity of backgrounds and that people with little direct experience with the
problem be included. Brainstorming sessions usually last not longer than one
hour. Ideas produced by a brainstorming session are recorded and evaluated at
a later time by the brainstorming group or by another group or individual.
A form of the brainstorming technique can also be used by an individual.
The individual follows the same rules used for a group session: combination of
ideas, postponement of evaluation, and an emphasis on obtaining a large num-
ber of ideas. The individual brainstorming session need not take more than a
few minutes. Again, the ideas are recorded and evaluated at a later time.

Checklists
One of the simplest ideas for generating new ideas is to make a checklist. The
checklist encourages the user to examine various points, areas, and design pos-
sibilities. For example, suppose that you were attempting to improve a certain
device. You might make a checklist that includes:

Ways the device could be put to other uses.

Ways the device could be modified.
Ways the device could be rearranged.
Ways the device could be magnified.
Ways the device could be lessened, and so on.

Attribute Listing
Another technique that can be used by individuals to produce original problem-
solving ideas is attribute listing. With this technique, all of the major character-
istics or attributes of a product, object, or idea are isolated and listed. Then, for
each attribute, ideas are listed as to how each of the attributes could be changed.
Every idea is listed, no matter how unrealistic or impractical. After all of the
116 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

ideas have been listed, each of the various ideas is evaluated, bringing to light
possible improvements that can be made in the design of the product or system.
Consider the following example of how the attribute listing technique could
be used to improve the design of a telephone ((3).

Attribute Ideas
1. Color Could be any color
Could be transparent
Could utilize designs such as plaid
Could have a personalized design
2. Material Could be metal
Could be glass
Could be wood
Could be hard rubber
3. Dial Could be 10-push-button design
Could be a lever system
Could use abacus-type system
Could be push buttons arranged in a line
4. Handset and base Make it square
Make it round
Make it oval
Use higher base
Use lower base
Eliminate handset by using microphone and speaker

Forced Relationship Technique

Another group of operational techniques that individuals may use to generate
ideas is known as forced relationship techniques (3). ). That is, such techniques
force a relationship between two or more normally unrelated ideas or products
to begin the idea-generating process. One such technique consists of selecting
the fixed element in the forced relationship, that being the product or device
that is to be designed or an idea related to the problem statement. Next, atten-
tion is focused on some other element, chosen randomly. Then a forced or
unnatural relationship between the fixed element and the randomly chosen
element is established. This forms the basis for a free-flowing list of associa-
tions from which hopefully new and original ideas will emerge.
For example, suppose that we are interested in designing a weed-cutting
device. The device is the focused object. Suppose that we choose arbitrarily an
automobile wheel as the other element. Some of the ideas that might occur
based upon the characteristics of the automobile wheel include:

A round weed cutter.

A rubber weed cutter.
 5.3 SEARCHING FOR CREATIVE SOLUTIONS 117

A weed cutter that rolls.

A weed cutter that has spokes.
A weed cutter that has pneumatic tires.
A weed cutter that has brakes.
A weed cutter that will not break, and so on.

Notice that we began this process by merely taking the characteristics of the
wheel—roundness, rubber material, ability to roll, and so on—and applying
them to the weed cutter. Other ideas may then develop based upon simple ver-
bal relationships or similar-sounding words. For example, the word “brakes”
led to the idea of a weed cutter that does not “break.”
This technique takes little time. If the forced relationship being used does
not seem to be profitable, one can simply select a new random element and
repeat the process. As with other idea-generating techniques, many impracti-
cal, even foolish, ideas may be generated. Evaluation of these ideas should not
be undertaken during the idea-generation stage.

Morphological Analysis
An operational technique for idea generation attributed to Fritz Zwicky (3)
involves listing every conceivable theoretical solution. This technique consists
of first defining the problem in terms of its dimensions or parameters and
devising a model that enables one to visualize every possible solution. For a
solution with only two parameters, the model takes the form of a large square
divided into a group of smaller squares. The horizontal axis would show vari-
ous possible choices for one parameter; the vertical axis would show possible
choices for the other parameter. Such an arrangement allows the user to exam-
ine the combined effects of attributes of the variables.
For problems with more than two parameters, the model becomes a matrix
with each parameter assigned to one axis of a rectangular array. This is best
explained by an example.
Referring again to the design of a weed-cutting device, suppose that we wish
to examine three attributes or variables: the type of power supply, the type of
blade motion, and the type of material used in its construction. We list the pos-
sible methods of satisfying each attribute, for example:

Power: electric engine, gasoline engine, water pressure.

Blade action: rotary, reciprocating.
Material: steel, aluminum, plastic.

The combination of these conditions yields 18 possible solutions that may

now be subjected to scrutiny and evaluation. Figure 5.1 graphically illustrates
the various combinations that comprise possible solutions. Each of the small
cubes represents one possible solution. For example, cube A represents a steel
device with an electric engine and a rotary blade.
118 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

Figure 5.1 An illustration of

the morphological technique.

5.4 STEPPING FROM IDEATION TO PRELIMINARY DESIGNS

The engineer is now ready to move from ideation to preliminary designs. This
is the heart of the design process, and it is the phase that relies most on expe-
rience and engineering judgment. Here, unworkable ideas are discarded, and
promising ideas are molded and modified to form workable plans and designs.
This step may require many decisions to be made about alternative layouts,
configurations, materials, dimensions, and other specifications. Conceptual
sketches may need to be drawn; preliminary plans may need to be prepared;
and thought may need to be given to material specifications.
Preliminary designs may evolve through analysis or synthesis. Analysis
involves the separation of a whole into constituents for individual study.
Synthesis involves combining facts, principles, or laws into a whole idea that
will accomplish some desired result or solve a problem.
In this phase, it is necessary to subject possible solution ideas to careful
scrutiny. Possible solutions are carefully and critically examined and studied.
There are many ways that this can be done. In certain instances, preliminary
sketching of a device or casual analysis of a process will show that an idea is
not worthy of further consideration. In other cases, a component may need to
be examined by laboratory tests. In still other instances, a formal and compre-
hensive research program may need to be undertaken to examine the validity
of a hypothesis or the efficacy of a proposed solution.
To facilitate the design process, engineers often rely on models. Duderstadt,
et al. (4)) defines a model as “any simplified description of an engineering sys-
tem or process that can be used to aid in analysis or design.” Engineers use a
variety of models that may be tangible or intangible. In the broadest definition
of the term, sketches and graphs (discussed in Chapter 6) can be thought of as
types of models. In addition, three types of model are commonly used to facil-
itate the solution of engineering problems:
 5.4 STEPPING FROM IDEATION TO PRELIMINARY DESIGNS 119

1. Analytical or mathematical models.

2. Simulation models.
3. Physical models.

Mathematical Models
A mathematical model consists of one equation or a group of equations that
represents a physical system. For example, the following equation represents
the wind pressure on a structure:
p KV2

where p wind pressure expressed in pounds per square foot

K a factor that depends principally on the shape of the structure
V the wind velocity expressed in miles per hour

Many physical phenomena can be described by mathematical models. Such

models may be based on scientific theories or laws that have stood the test of
time. Others are empirically based, relying upon or derived from observations
and experiments.
Generally speaking, mathematical models can describe only relatively sim-
ple physical phenomena. Mathematical models designed to describe complex
phenomena tend to be so intractable as to be of little value.

Simulation Models
When studying complex systems, engineers often employ computer simulation
models. Such a model may incorporate empirically based mathematical mod-
els as components of the total model. For example, Thomasson and Wright ((5)
used computer simulation to study traffic at a two-way stop controlled inter-
section. They first made empirical studies of driver behavior at the intersec-
tions and found that the phenomenon could be separated into several
well-defined events:

1. The arrival of cars on the side and main streets.

2. Turning movement decision.
3. Acceptance or refusal of available gaps in the traffic on the main street.
4. Start time, the hesitation prior to moving into the intersection.
5. Service time, the time the intersection was occupied.

The description of these events was based in part on actual observations of

drivers or on empirical mathematical models that had been reported by other
researchers. For example, it was found that the time a vehicle occupies the
intersection could be described by a normal probability distribution (see
Figure 5.2). (A discussion of the normal probability distribution appears in
Chapter 7.)
120 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

Figure 5.2 The results of an

empirical study used for a
simulation model. The figure
shows distributions of times that
vehicles occupy an intersection.

A computer program (or model) was then developed to describe the behav-
ior of drivers that arrived at and passed through the intersection. Using a pro-
cedure described by the general flowchart shown as Figure 5.3, the researchers
modeled traffic flow through the intersection and estimated delays to motorists
under different conditions of traffic flow, as Figure 5.4 illustrates.

Physical Models
Physical models have long been used by engineers to gain a better under-
standing of complex phenomena. Such models probably constitute the oldest
method of structural design. Physical models have also been used for many
years in the fields of hydraulics, hydrodynamics, and aerodynamics.
Examples of studies made with physical models include:

1. Dispersion of pollutants throughout a lake.

2. Behavior of waves within a harbor. (See Figure 5.5.)
3. The underwater performance of submarines of different shapes.

Full-scale models are sometimes built, but they are often built to a smaller
scale. Typical scales for physical models range from 1 : 4 to 1 : 48 ((6).
Perhaps the greatest value of physical models is that they allow the engineer
to study a device, structure, or system with little or no prior knowledge of its
behavior or need to make simplifying assumptions.
 5.4 STEPPING FROM IDEATION TO PRELIMINARY DESIGNS 121

Figure 5.3 A general flowchart for a simulation model.

122 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

Figure 5.4 Output from a computer simulation model. The

figure shows the average times cars wait to cross a main street.

An Example of Physical Models Some of the most beneficial research with

physical models has been accomplished with wind tunnels (see Figures 5.6
and 5.7). A brief description of wind tunnels, abstracted from Reference 6, is
given in the following paragraphs.
Wind tunnels are based on a fundamental law of fluid dynamics, namely,
that a body immersed in a moving fluid experiences the same forces as if the
body were moving and the fluid stationary, assuming that the relative speed of
the fluid and solid object is the same in both cases. This means, for example,
that the conditions surrounding an airplane in flight can be replicated by hold-
ing the plane stationary and moving the air past it at a velocity comparable to
flight speeds.
Advantages of wind tunnels over flight testing are economy, safety, and
research versatility. A model airplane can be tested in a wind tunnel at a frac-
tion of the cost of building and operating a full-scale prototype, and the air-
worthiness of new and experimental designs can be tested without risking a
 5.4 STEPPING FROM IDEATION TO PRELIMINARY DESIGNS 123

Figure 5.5 A physical model used to study the behavior of

waves in a harbor. (Courtesy of the U.S. Army Corps of
Engineers.)

Figure 5.6 Some of the most bene-

ficial research with physical models
has been accomplished with wind
tunnels. Here a ram jet model is
being studied in a 10 ft. ! 10 ft. test
section. (Courtesy of the National
Aeronautics and Space
Administration.)
124 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

Figure 5.7 This giant air pas-

sage houses a massive set of
stationary guide vanes in one
elbow of a wind tunnel which
force the air to make a right
angle smoothly. (Courtesy of
the National Aeronautics and
Space Administration.)

pilot’s life. Wind-tunnel testing can simulate flight under conditions more con-
trolled and measurable than would be possible in a flight test. Even before the
Wright brothers’ first flight, the wind tunnel was the principal tool of the aero-
nautical engineer.
All wind tunnels have common features that circumscribe their characteris-
tics and capabilities. All have a test section in which a model, component, or
system can be fixed or suspended. The cross section may be round, oval, rect-
angular, or polygonal. Test sections may vary in width and height from a few
inches up to 100 feet or more.
Wind tunnels may be either return or nonreturn. Nonreturn tunnels draw air
from the atmosphere, pass it through a tube that includes a test section, and
discharge it into the atmosphere. Such tunnels are simple and inexpensive to
build but are inefficient and limited in the types of flow they can generate.
Many sophisticated tunnels use a return-type circuit in which the same air is
moved around a closed loop.
A major advantage of closed tunnels is that they can be pressurized, a tech-
nique that allows objects to be studied that are a fraction of the full scale.
Comparability between conditions of wind-tunnel tests on models and condi-
tions experienced by a full-scale aircraft in flight depends on a dimensionless
mathematical quantity known as the Reynolds number (named for the nine-
teenth century British engineer Osborne Reynolds). The Reynolds number is a
 5.5 EVALUATION AND SELECTION OF PREFERRED SOLUTION 125

flow-similarity parameter that describes forces acting on a body in motion with

respect to the fluid in which it is immersed. The number is directly propor-
tional to the size of the body and the density and relative speed of the fluid,
and inversely proportional to the viscosity of the fluid. Other things being
equal, a model “moving” with respect to an airstream would have a smaller
Reynolds number than a full-scale plane in flight. The easiest way to equalize
the Reynolds numbers—and thus to obtain comparable flow conditions for the
plane and the model—is to increase the speed or density of the airstream in
which the model is immersed.
Almost all wind tunnels employ a complex array of balances and other mea-
suring devices designed specifically for the purpose. Most closed-circuit tun-
nels use tunnel vanes to guide the airflow smoothly around the corners in the
circuit. Most tunnels use complex arrays of settling chambers, screens, and
throat contractions to smooth and straighten the airstream as it accelerates into
the test section. A variety of model-support systems is used, depending on the
configuration of the test object. Some tunnels use smoke to help visualize air-
flow. Some are rigged with special photographic devices that record shock
waves produced at high speeds. Some tunnels are refrigerated to produce ice
on the models like that encountered under certain flight conditions. In fact,
wind tunnels have been designed to replicate nearly every condition encoun-
tered by airplanes in flight (6).

5.5 EVALUATION AND SELECTION OF PREFERRED SOLUTION

As the engineering design process evolves, the engineer may evaluate again
and again alternate ways of solving the problem at hand. Typically, the engi-
neer winnows the unpromising design choices, yielding a progressively
smaller set of options. Feedback, modification, and evaluation may occur
repetitively as the device or system evolves from concept to final design.
Depending on the nature of the problem to be solved, evaluation may be based
on any number of factors. If it involves a product, safety, cost, reliability, and
consumer acceptability are often of paramount importance.
Perhaps the most straightforward way to evaluate a product is to develop a
prototype and simply test it in operation. In some cases the prototype might
not work due to one or more components of the design. The designer should
try to identify all of the weak links of a prototype before accepting or discard-
ing the design idea. Many great ideas have been discarded prematurely, and
many working prototypes have failed to operate as expected when turned into
a product. No idea should be evaluated solely on the basis of one prototype or
one test.
There are many indirect methods for evaluating a proposed design. For
example, wind-tunnel testing a scale model of an aircraft design can reveal
good and bad features of the design at a small fraction of the cost and risk of a
full-scale assembly. Alternatively, the aerodynamics of the new aircraft design
can be evaluated using computer simulation of the expected flight conditions.
The associated mathematical equations used in the computer simulation may
126 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

not accurately account for some of the complicating factors such as the com-
ponent interference and turbulence. However, the computer simulation can
indicate approximate features of the design, which makes it easier to design the
first scale model for wind-tunnel testing.
The optimization scheme can become very difficult when the design
requires a human operator and a man-machine interface. This difficulty exists
because no two human beings are the same. The basic anatomical and physio-
logical differences between humans makes the human factors of the design dif-
ficult to quantify. One human user may find a design very acceptable and
efficient, while another may consider it to be intolerable; therefore, the opti-
mization of human factors becomes a matter of statistical comparisons. Hence,
the user population must be identified and characterized before attempting to
optimize any designs that involve human operators.
In addition to the routine judgments that engineers make about a device or
system, more formal and structured evaluations are often needed. This is espe-
cially true of public works projects, which must be judged from the viewpoints
of competing and often conflicting groups. Such evaluations have traditionally
relied on economic analysis, but recent concerns with social and environmen-
tal impacts of public projects have produced much broader evaluation tech-
niques. Let us now examine some of these formal evaluation techniques.

Economic Analysis
For at least 50 years, economic analyses have been used by engineers as a
decision-making tool for the building of dams, bridges, highways, and other pub-
lic works. Conceptually, such analyses attempt to compare the public benefits
from such projects with the costs of providing them.
Economic studies may be used:

1. To determine the feasibility of a project.

2. To compare alternate designs.
3. To determine the priority of construction of a group of projects.
4. To evaluate specific features of design.

In economic analyses, it is important to recognize the time value of money.

Because of the existence of interest, a quantity of money is worth more now
than the prospect of receiving the same quantity at a later date. Therefore, in
order to compare costs and benefits of an engineering project on a sound basis,
they must be converted to equivalent values at some common date. This pro-
cedure, which is known as discounting, is accomplished by using a suitable
interest rate in accordance with established economic principles. Such con-
cepts are more fully described in textbooks on economic analysis and facility
design (7, 8).
One approach is to compare benefits and costs on the basis of present worth
or present value. For example, the present worth, P, of some future single pay-
ment, F, can be calculated by the following equation:
 5.5 EVALUATION AND SELECTION OF PREFERRED SOLUTION 127

F
P (5.1)
(1  i )n

where i the interest rate or discount rate per period

n the number of interest periods, usually years

Example 5.1

Suppose that a bridge is to be constructed over a river and that it will cost
$385,000 to dismantle it at the end of its expected life of 40 years. Determine
the present value of dismantling the bridge using an interest rate of 10 percent.
Solution By Equation (5.1),
$385,000
present value, P $8500
(1  0.10)40
Economic equivalencies such as these are often illustrated in the following
manner:
$385,000

$8500

" " " n 40 years " " "

The present worth, P, of a series of uniform annual end-of-period pay-

ments, A, can be calculated by the following equation (7).

(1  i )n  1
Ps A (5.2)
i(1  i )n

Example 5.2

Suppose that it is estimated that the annual user benefits (in excess of any
maintenance costs) of the bridge described in Example 5.1 are a uniform
$94,500 per year. Determine the present value of these benefits assuming an
interest rate of 10 percent and an analysis period (bridge life) of 40 years.
Solution By Equation (5.2),

(1  0.1)40  1
present value, Ps $94,500 $924,000
0.1(1  0.1)40
128 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

Economic analyses may be carried out by one of several methods. Two pop-
ular methods are the net present value method and the benefit-cost ratio
method.
The net present value method involves reducing all benefits and costs to
their present value in accordance with the principles previously described.

$924,000

$94,500 per year

" " " n 40 years " " "

The net present value is defined as the difference between the present values
of the benefits and the costs of the project.

Example 5.3

Suppose that the initial costs of building the bridge described in Examples 5.1
and 5.2 was $800,000. Determine by the net present value method whether the
bridge is economically justified.

Solution As previously calculated, the present value of the benefits is

$924,000. The present value of the costs would be the sum of the initial cost
and the present value of the dismantling cost or $800,000  $8500 $808,500.
The difference between these amounts is $115,000. Since the benefits exceed
the costs, the project is economically feasible.
The benefit-cost ratio method, as the term suggests, is simply the ratio of the
benefits to the costs, each expressed as present values. A benefit-cost ratio
greater than 1.0 would indicate that a project is economically justified. A value
less than 1.0 indicates that it is not. In the previous example, the benefit-cost
ratio is $924,000 divided by $808,500 or 1.14, again showing that the project is
economically viable.

Other Evaluation Techniques

Over the past several years, there has been increasing awareness of the impact
that engineering works may have on people and the environment. Such pro-
jects may cause families and businesses to be relocated and subject citizens to
noise and water and air pollution. Many of these impacts cannot be simply
 5.5 EVALUATION AND SELECTION OF PREFERRED SOLUTION 129

reduced to a dollar amount. Techniques are available, however, to help the

engineer, to some extent at least, quantify such impacts. These techniques gen-
erally involve ranking alternate projects on a scale based upon some predeter-
mined criteria. The rankings may be made by a panel of experts or citizens or
based on attitude surveys. This approach is illustrated by the following exam-
ple, which was abstracted from Reference 9.
Suppose that a city considering a new mass transit system has established
the following objectives for the system:

1. The system should provide economy.

2. There should be minimum disruption of individuals by relocation.
3. The system should provide a high level of comfort and convenience.
4. The central area should be highly accessible.
5. The system should be accessible to low-income areas.

The following criteria were selected to provide a measure of each objective:

1. Benefit-cost ratio.
2. Number of persons relocated.
3. Load factor on transit vehicles in peak hour.
4. Accessibility index of core areas.
5. Transit accessibility index to low-income traffic zones.

A panel of citizens and engineers then established the following ratios of rel-
ative importance for these objectives: 40 percent, 20 percent, 20 percent, 10
percent, and 10 percent. They then evaluated these objectives for each of three
alternate projects, with the following results:

Possible
Effectiveness Score for Score for Score for
Evaluation Matrix: Score Plan A Plan B Plan C
1. Benefit-cost ratio 40 35 25 30
2. Persons relocated 20 10 20 5
3. Transit load factor 20 10 15 3
4. Core accessibility 10 3 5 10
5. Low-income transit availability 10 2 10 8
Total effectiveness score 100 60 75 56

Plan B was selected as being the most responsive to the goal statements that
the transportation plan was designed to achieve.
130 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

5.6 PREPARATION OF REPORTS, PLANS, AND SPECIFICATIONS

After the preferred design has been selected, it must be communicated to those
who must approve it, support it, and translate it into reality. This communica-
tion may take the form of an engineering report or a set of plans and specifica-
tions. Engineering reports are usually directed to a client or to a supervisor
(e.g., if the engineer is employed by a large corporation). Plans and specifica-
tions are the engineer’s means of describing to a manufacturing division or to
a contractor sufficient detail about a design so that it can be produced or con-
structed. This important phase of the design process is treated more fully in
Chapter 6.

5.7 IMPLEMENTATION OF THE DESIGN

It could be argued that once the plans, specifications, and engineering reports
have been completed, the design process is finished. Actually, however, the
final phase of the design process is implementation, the process of producing
or constructing a physical device, product, or system. Engineers must plan and
oversee the production of the devices or products and supervise the construc-
tion of the engineered projects. Different engineers may, of course, be involved
in this final phase. This is the culmination of the design process; to the design
engineer, it is the most satisfying phase of all.

Patenting
It is not uncommon for the work of an engineer to be of such value that it
should be protected from exploitation by others. This may be accomplished by
patenting.
A patent for an invention grants a property right by the government to the
inventor or his or her heirs. It excludes others from making, using, or selling
the invention. In the United States, patents are administered by the Patent and
Trademark Office, a division of the U.S. Department of Commerce.
The patent law classifies subject matter that can be patented as “any new or
useful process, machine, manufacture, or composition of matter, or any new
and useful improvements thereof” ((10). This means that patents can be
obtained for practically everything that is made by humans as well as the pro-
cess for making them.
In order for an invention to be patentable it must be new as defined in the
patent law. The law states that an invention cannot be patented if:

1. The invention was known or used by others in this country, or

patented or described in a printed publication of this or a foreign
country, before the invention thereof by the applicant for patent, or
2. The invention was patented or described in a printed publication in
this or a foreign country or in public use or on sale in this country
 5.7 IMPLEMENTATION OF THE DESIGN 131

more than one year prior to the application for patent in the United
States. (10)
To apply for a patent, the inventor must submit an application which
includes:

1. A specification, a clear, concise, and exact description of the invention

presented in such manner as to distinguish it from other older inventions.
2. An oath or declaration made by the inventor asserting the belief that he or
she is the original and first inventor of the subject matter of the application.
3. A filing fee. Reference 10 gives a schedule of fees, beginning with a basic
fee of $630.
4. A drawing prepared to Patent Office specifications whenever the nature of
the case requires such to understand the invention.

The Scientific and Technical Information Center of the Patent and Trade-
mark Office in Arlington, Virginia has a Search Room where the public may
search and examine U.S. patents granted since 1836. There, patents are
arranged in over 400 classes and 120,000 subclasses. By searching these clas-
sified patents, it is possible to determine, before actually filing an application,
whether an invention has been anticipated by a U.S. patent.
Since a patent is not always granted when an application is filed, many
inventors attempt to make their own search before applying for a patent. This
may be done in the Search Room of the Patent and Trademark Office or at
libraries located throughout the United States which have been designated as
Patent Depository Libraries.
The preparation and processing of an application for patent is an undertak-
ing requiring the knowledge of patent law as well as knowledge of the scien-
tific or technical matters involved in the particular invention. Most inventors,
therefore, employ the services of registered patent attorneys or patent agents.
The Patent and Trademark Office maintains a register of attorneys and agents
qualified to help inventors with the patenting process (10).

Computer-Aided Design
Although digitial computers first appeared in the 1940s, most of the growth in
computer technology has occurred within the past two decades. In this rela-
tively brief span of time, the use of computer-based tools for engineering prob-
lem solving has become commonplace. These tools are variously referred to as
computer-aided engineering (CAE), computer-aided design (CAD), or computer-
aided design and manufacturing (CAD/CAM). These systems make the engi-
neering problem solving process much more efficient and free engineers from
monotonous and unimaginative tasks, allowing them more time for technical
ingenuity. Here we will focus primarily on CAD systems, which are becoming
increasingly powerful and pervasive design tools.
132 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

With CAD systems, interactive graphics allow the user to communicate eas-
ily with the computer in display-screen pictures. With the speed of modern
computers, this communication is carried out in real time; that is, the com-
puter’s response is practically instantaneous. Furthermore, little knowledge of
computers is required to operate these user-friendly systems.
Earlier CAD systems were used primarily to perform automated drafting
or to facilitate the solution of simple two-dimensional problems. Interactive
graphic systems now allow the user to develop three-dimensional models
and to perform a wide range of geometric manipulations and sophisticated
analyses.
Assisting the designer in constructing a model are thousands of soft-
ware aids that automate many of the tedious tasks consuming so much
time in traditional manual methods. With the stroke of the pen or the
push of a button, the user can move, magnify, rotate, flip, copy, or other-
wise manipulate the entire design or any part of it. . . .
In working with a complex model, the user can temporarily “erase”
portions of it from the screen to see the area under construction more
clearly. Then the deleted area is recalled later to complete the model.
Likewise, portions of the model may be enlarged to view and add minute
details accurately. And the model may be moved and rotated on the
screen for the user to view it at any angle. Furthermore, mechanisms such
as linkages and gears in the assembly may be animated on the screen to
ensure proper operation and check for interferences. When the design is
complete, the system may automatically add dimensions and labels. And
automated drafting features may be used to produce detailed engineering
drawings. (11)

Learning from Failures

Despite the best efforts of engineering designers, their designs occasionally fail.
Bridges collapse, roofs of buildings cave in, and earthen dams fail, threatening
human lives and causing millions of dollars of damage to property.
Engineering failures may be attributed to a wide variety of causes, including:

1. Mistakes made by inept or careless designers.

2. Imperfections in building or manufacturing materials and uncertainties as
to their variability.
3. Careless workmanship by technicians or craftsmen who implement the
design.
4. Poor communications between the managers, engineers, technicians, and
craftsmen who produce or construct the engineered design.

Although the object of engineering design is to preclude failure, truly fail-

proof design cannot be achieved. An engineering structure or machine can fail
 5.7 IMPLEMENTATION OF THE DESIGN 133

in many ways, the results ranging from blemishes to catastrophes. It is a curi-

ous fact that we often learn more from our failures than from our successes.
One of the ironies of engineering history is that success ultimately leads to
failure. With each engineering success, questions arise from taxpayers, from
boards of directors, and from engineers themselves as to how much larger, or
how much lighter, or how much more economical the next structure of
machine can be made. Thus, in the interests of aesthetics or economy, there is
a tendency to be more daring in design and to take greater risks in workman-
ship and style.
On the other hand, after a failure, there is usually pressure to increase fac-
tors of safety, revise building codes, and to generally engage in a more conser-
vative engineering practice ((12).
When an engineering failure occurs, it is usually important that an investi-
gation be conducted to determine the cause of the failure and to identify reme-
dial actions that can be taken to prevent or lessen the likelihood of a
reoccurrence. Such investigations may be carried out by:

1. Government agencies or boards.

2. Private or professional associations.
3. Ad hoc committees or commissions.

For example, the National Transportation Safety Board (NTSB) has a staff of
about 300 specialists who investigate hundreds of civil aviation and surface
transportation accidents each year. The Board does not have regulatory powers,
but it recommends changes for improved safety that other agencies of govern-
ment may implement.
The National Fire Protection Association (NFPA), a private (but government-
assisted) organization, empanels teams of experts to study and report on the
causes of major fires. Great benefits have accrued to building owners and to the
public from recommendations of the NFPA to improve the nation’s fire codes.
The Architectural and Engineering Performance Information Center (AEPIC)
was organized at the University of Maryland in 1982 (13). AEPIC serves as a
focus for activities to collect and disseminate knowledge and experience
gained from structural failures. By establishing patterns of structural failures,
AEPIC seeks to help engineers and architects to prevent similar future failures
and avoid litigation (13).
When an especially dramatic or costly engineering failure occurs, a special
committee or commission may be empaneled to study the circumstances of
the failure and to make appropriate recommendations. Such groups may be
convened under the auspices of one or more professional organizations or
may be named by a governor or the president of the United States. A notable
example of the work of such a commission is given in the Report of the
Presidential Commission on the Space Shuttle Challenger Accident (14). The
Space Shuttle Challenger accident is described as a case study in Chapter 9 of
this book.
134 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

THE HYATT REGENCY HOTEL WALKWAYS COLLAPSE (15, 16)

On July 17, 1981, during a weekly tea dance, two suspended walkways
in the atrium area of the Hyatt Regency Hotel in Kansas City, Missouri
collapsed, killing 114 people and injuring 185. The collapse was one of
the most serious structural failures in the history of the United States.
The hotel included a 40-story guest room tower, a four-story wing con-
taining restaurants and meeting rooms, and a large, open atrium. The
atrium contained three suspended walkways at the second-, third-, and
fourth-floor levels which connected the tower section with the restau-
rants and meeting rooms. Each walkway was 120 feet long and approxi-
mately 81/2 feet wide.
The second- and fourth-floor walkways were constructed one above
another along the west wall of the atrium. The third-floor level walkway
was independently suspended from the atrium roof trusses and was built
about 13 feet toward the center of the room. The third-floor walkway was
not involved in the collapse. The second-floor walkway was suspended
from the fourth-floor walkway, which in turn was suspended from the
roof framing.
In the aftermath of the tragedy, the mayor of Kansas City requested that
the National Bureau of Standards (NBS) conduct an independent inves-
tigation to determine the cause of the collapse (16). There were, of course,
other investigations and hearings as well as at least 150 lawsuits seeking
total damages of $1.5 billion.
Each of the walkways consisted of four spans of about 30 feet.
Intermediate supports consisted of three pairs of 11/4-inch diameter steel
hanger rods. Each walkway was supported from underneath longitudi-
nally by 16-inch deep wide flange steel beams and laterally by 8-inch
deep steel box beams. The box beams were made of pairs of 8-inch steel
channels that were welded together at their open ends. The walkway
deck consisted of a formed steel deck overlain by a 31/4-inch lightweight
concrete slab which acted compositely with the 16-inch deep stringers.
Figure 5.8 shows a cross-section of the walkways as well as the hanger
rod connection.
The National Bureau of Standards discovered that a significant design
change had been made after the completion of the original design draw-
ings in the way in which the walkways were connected to the hanger
rods. The original drawings called for both the fourth- and second-floor
walkways to be hung from the same continuous steel rods. As originally
conceived, pairs of single rods about 45 feet long would have been passed
through the box beams at the fourth-floor walkway and continued down
to pass through the box beams at the second-floor level. The walkways
were to have been supported by a nut-and-washer connection under the
box beams at each level.
Figure 5.8 Cross section of walkways and detail of walkway hanger rod-
box beam connections. (Source: Neil Schlager, ed., When Technology Fails,
Gale Research, Inc., Detroit, 1994.) (Reprinted by permission of AP/Wide
World Photos.)

135
136 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

Instead of using continuous rods, however, the contractor hung the

walkways from two separate sets of rods. Each fourth-floor box beam had
two holes at each end, one 2 inches from the end and the other 6 inches
from the end. The upper hanger rod passed through the outer hole, while
the lower hanger rod, onto which hung the second-floor walkway, passed
through the inner hole. A nut-and-washer connection was used on the
ends of the rods.
The effect of this change was to essentially double the load on the box-
beam hanger connection at the fourth floor walkway. In fact, it was con-
cluded that the collapse was initiated when the welding of one of the
fourth-floor beams split, and the nut-and-washer connection supporting
that part of the walkway slipped through the hole. The load of both walk-
ways was then transferred to the remainder of the fourth floor connec-
tions, which also failed. The fourth-floor walkway then collapsed onto
the one below it, and both walkways crashed onto the lobby floor.
It was reported that the failed connection was designed by the project’s
steel erector and fabricator, a common practice in the construction indus-
try. It was noted, however, that the shop drawings submitted by the fabri-
cator had been signed by the chief structural engineer and the project
engineer. Missouri law states that: “Any registered engineer who affixes
his signature and personal seal to any such plans . . . shall be personally
and professionally responsible therefore.” The Board of Architects,
Professional Engineers, and Land Surveyors voted unanimously to revoke
the licenses of these two engineers permanently (17) and to revoke the
certificate of authority of their firm. The firms’ assets were sold to another
civil and structural engineering firm.
An expert witness appearing before the Missouri Administrative
Hearing Commission stated that the engineer of record usually has to
delegate his duties because there is too much work. “But he cannot del-
egate his responsibility. . . . There has to be a point where the buck
stops.” (18)

REFERENCES
1. CLELAND, DAVID I., AND HAROLD KERZNER, Engineering Team Management, Van
Nostrand Reinhold Company, New York, 1986.
2. PEARSON, DONALD S., Creativeness for Engineers, 3rd Edition, Pennsylvania State
University, University Park, PA, 1960.
3. WHITING, CHARLES S., Creative Thinking, Rheinhold Publishing Corporation, New
York, 1958.
4. DUDERSTADT, JAMES J., GLENN F. KNOLL, AND GEORGE F. SPRINGER, Principles of
Engineering, John Wiley & Sons, Inc., New York, 1982.
5. WRIGHT, PAUL H., AND J. N. THOMASSON, JR., “Simulation of Traffic at a Two-Way
Stop Intersection,” Traffic Engineering, Vol. 37, No. 11, August 1967.
 EXERCISES 137

6. ROLAND, ALEX, Model Research, The National Advisory Committee for Aeronautics
1915–1958, Vol. 2, National Aeronautics and Space Administration, Washington,
DC, 1985.
7. GRANT, EUGENE L., W. GRANT IRESON, AND RICHARD S. LEAVENWORTH, Principles of
Engineering Economy, 8th Edition, John Wiley & Sons, Inc., New York, 1990.
8. WRIGHT, PAUL, H. Highway Engineering, 6th Edition, John Wiley & Sons, Inc., New
York, 1996.
9. WRIGHT, PAUL H., NORMAN J. ASHFORD, AND ROBERT STAMMER, JR., Transportation
Engineering: Planning and Design, 4th Edition, John Wiley & Sons, Inc., New York,
1998.
10. General Information About Patents, U.S. Department of Commerce, Patent and
Trademark Office, Washington, DC, Revised December, 1990.
11. KRAUSE, JOHN K., Computer-Aided Design and Computer-Aided Manufacturing,
Marcel Dekker, Inc., New York, 1982, p. 12. Reprinted by courtesy of Marcel
Dekker, Inc.
12. PETROSKI, HENRY, To Engineer Is Human, St. Martin’s Press, New York, 1985.
13. FITZSIMONS, NEAL, AND DONALD VANNOY, “Establishing Patterns of Building
Failures,” Civil Engineering, Vol. 54, No. 1, American Society of Civil Engineers,
January 1984.
14. Report of the Presidential Commission on the Space Shuttle Challenger Accident.
Report to the President, Government Printing Office, 1986.
15. SCHLAGER, NEIL, ed., When Technology Fails, Gale Research, Inc., Detroit, 1994.
16. PFRANG, EDWARD O., AND RICHARD MARSHALL, “Collapse of the Kansas City Hyatt
Regency Walkways,” Civil Engineering, July, 1982, pp. 65–68.
17. ALM, RICK, “Hyatt Engineers Lose Licenses in Missouri,” Engineering News
Record, January 30, 1986, p. 11.
18. “Hyatt Hearing Traces Design Chain,” Engineering News Record, July 26, 1984, pp.
12–13.

EXERCISES
5.1 Write a report comparing the scientific method and the engineering method of
problem solving.
5.2 List three ideas for overcoming the problems of drivers locking their keys in their
automobiles.
5.3 Describe three ways for reducing or eliminating flying insects from a patio area.
5.4 Give an example of a technological development that was delayed because tech-
nologists failed to properly identify the problem.
5.5 Prepare a report describing the Delphi technique for developing ideas for solving
problems. Describe the background of the development of this technique. How
does it differ from brainstorming? Indicate how it can be used to facilitate the
solution of engineering problems.
5.6 Prepare a scheme for developing:
a. A system for measuring or monitoring heat loss from a residential house.
138 CHAPTER 5/THE ENGINEERING APPROACH TO PROBLEM SOLVING

b. A device to measure the distance a jogger runs.

c. A device mounted on the instrument panel of a car to indicate the distance
from the edge of a curb.
5.7 List two types of problems that could suitably be modeled by:
a. Mathematical models.
b. Simulation models.
c. Physical models.
5.8 Suppose that you wished to improve the design of bathroom scales. Prepare an
attribute list showing how each of the attributes might be improved.
5.9 Using the forced relationship technique, develop a list of ideas for designing a
better mousetrap. For the random element, use a mail box.
5.10 Using morphological analysis, prepare a list of possible attributes of a device to
remove dust from a warehouse floor.
5.11 Write a report on a significant and well-publicized engineering failure. Discuss
any lessons that were learned, and describe any changes in engineering codes
and design procedures that resulted from that failure. For ideas, consult
Reference 15.