David C. (David Chandler) Robertson.

Evaluating the use of CAD systems in mechanical design engineering online

. (page 1 of 3)
Online LibraryDavid C. (David Chandler) RobertsonEvaluating the use of CAD systems in mechanical design engineering → online text (page 1 of 3)
Font size
QR-code for this ebook


no. Zi%-



Evaluating the Use of CAD Systems in
Mechanical Design Engineering

David C Robertson
Thomas J. Allen

WP # 3196-90-BPS

January 1990



Evaluating the Use of CAD Systems in
Mechanical Design Engineering

David C Robertson
Thomas J. Allen

WP # 3196-90-BPS January 1990


Despite the importance of and our long history with Computer- Aided Design
(CAD) systems, our understanding of the systems is limited. The academic and
trade literature provides little guidance on what organizational actions are necessary
to utilize CAD systems for maximum benefit.

To understand the role CAD systems are playing in the design engineering
processes of different companies, field interviews were conducted at twelve
companies. Design engineers, managers, CAD support personnel, and others were
interviewed to understand the use of CAD systems for mechanical design

The result of the field research is the conclusion that managers view CAD
technology in one of three ways: as physical capital, as supporting or extending
human capital, or as enabling improvements in social capital. Further, the value
received from the technology will depend directly on how managers view the
technology. Managers who see the systems as physical capital (i.e. electronic drafting
boards) will receive some benefit; managers who view the systems as enabling
improvements in social capital will receive the greatest benefit from the technology.

The characteristics of each of the three views of CAD technology will be
discussed, as well as the barriers that prevent companies from realizing the full
benefit of CAD technology.


In recent years, industry has focused a considerable effort on reducing product
development times.. Many believe that the appropriate use of Computer-Aided
Design (CAD) systems may aid significantly in achieving this goal. Many companies
have made large expenditures on CAD systems- expenditures have reached SI 00
million for a single company's hardware and software. The market for CAD/CAM
hardware and software was over 17 billion dollars in 1987 and is expected to grow to
39 bilUon by 1991 [28].

There has been some debate, however, over the benefits of CAD systems [13].
While many organizations have made large investments in CAD technology, some
of these organizations do not believe their money was well spent. These
organizations are not seeing the benefits they expected from their systems.

In the following sections, the results of an investigation into the uses of CAD
systems at twelve manufacturing companies are reported. The goals of the research
are to understand how the systems are used, what difficulties occurred in
implementing the systems, and what benefits resulted. In the next section, an
introduction to CAD systems is presented, followed by a brief review of the current
literature relating to Computer-Aided Design.

CAD Systems

CAD systems are defined in this paper to be those computer tools that support
the design and design engineering processes. This definition thus includes
computer tools normally classified under Computer-Aided Design, Computer-
Aided Drafting, and Computer-Aided Engineering (CAE). The greatest use of these
tools is in the support of mechanical design and engineering [20]; it is that area of

application that will be addressed in the present paper. More inforniation about the
features and functions of such software can be found in [4], [20].

CAD systems can potentially lead to many important benefits for
organizations. For example, CAD systems can potentially reduce the length of the
product development cycle and improve relationships with both vendors and
customers. The literature provides many descriptions of CAD applications that
have helped to reduce the product development time ([3],[5],[6],[14],[23],[25],[27]). A
shorter design cycle allows companies to respond more quickly to competitive
challenges, incorporate newer technology into products, and charge higher prices for
unique features.

CAD systems can also help improve a company's links with its vendors and
its customers [7], [11]. For example, one airframe manufacturer provides major
customers with a CAD terminal. If a problem occurs in the field, maintenance
personnel can access the CAD design file and either (1) communicate by phone with
the manufacturer's support personnel (who can access an identical CAD file) or (2)
make comments on a CAD file and send the file electronically to the manufacturer.
In either case, communication between the manufacturer and its customers is faster,
there is less ambiguity in the information communicated, and there is much less
travel required of the manufacturer's support personnel to customer sites.

Yet despite the potential benefits of CAD systems, our understanding of the
systems is limited. In the next section, a review of the literature on the use of CAD
systems in organizations is presented.

Research on the Impact of CAD Systems

The academic and trade literature provides little guidance on what
organizational actions are necessary to utilize CAD systems for maximum benefit.

The studies in the literature fall into three categories: "competition" studies, social
impact studies, and case studies.

In the "competition" category are studies which test the performance of CAD
systems against drafting boards. Such studies show productivity gains from 25% to
350%, depending on the complexity and repetitiveness of the task [19], and on the
type of task. Many managers interviewed for this study report, however, that such
gains do not always occur in their own companies.

Studies of the social impact of CAD systems have produced mixed results,
some finding for enrichment of jobs [1], and others finding that CAD leads to "de-
skilling" [10]. Some studies even show both effects occurring [18], [26]. These mixed
results were also seen in field interviews- different companies report different
changes in jobs. Some reported that the work had become more repetitive, others
stated the opposite.

In the third category of research, case study research, companies report the
productivity gains from implementing a certain system or application. For example,
Swerling [24] reports that the use of computer tools cut the development costs for
the IBM 3081 computer by 65%. Bull [5] reported that CAD/CAM improved
productivity for product development by 150%. Chrysler estimates that computer
tools will cut the development cycle for new cars from five years to four or less [14].
Kodak [12] linked product designers and tool designers through a common CAD
system and were able to develop the "Fling 35" camera from project start to shipping
approval in 38 weeks. They estimated that this CAD system helped them reduce
development costs by 25%.

Unfortunately, with few exceptions, this third category of research provides
managers little insight into how to achieve similar gains in their own
organizations. The studies in this stream of research focus largely on the specific
characteristics of the new technology, and ignore the other technological and

organizarional changes that must be made to utilize the technology to full
advantage. With few exceptions (e.g. [15], [23], [26]), it is not clear how the results
achieved in these studies could be repeated in other organizations.

It is important to note that the large gains reported in the case study literature
are hardly guaranteed- some managers believe that the introduction of CAD systems
has hurt productivity (cf. [13]). If the productivity gains of CAD systems are mixed, it
is less likely that the negative impacts would ever be reported in the literature.
Thus we cannot count on the literature to accurately reflect the average companv's
experiences with CAD systems.

The reason for the lack of convergence in the social impact studies and
managers' disagreement with case study results is that CAD systems do not
necessarily cause any changes to occur to the structure or processes of an
organization- they only enable changes. The eventual use of a CAD system is as
much a result of managerial decisions, individual predispositions, and
orgaruzational environments as it is of the features of the system. Thus different
organizations can (and do) use the systems quite differently. CAD systems are a
complex technology, and their application in organizations tends to reflect the
characteristics of the organization as much as the features of the technology [2], [IS].

Different Perspectives on the CAD Investment

The concepts of physical, human and social capital that are defined and
developed in this section will be used to analyze the observational data from the
field. Physical, human, and social capital can be seen as resources in the product
development process. A good manager will develop and use each to its maximum
advantage. Physical Capital comprises the machines and equipment that are used to
add value to a product, or allow a product to be developed more efficiently (i.e. more
quickly or less expensively). The cost and value of this type of capital are relatively

well understood by managers. Human Capital comprises the skills and knowledge
of a company's workers that allow them to add value to a product during its
development, or to develop a product more efficiently. Managers are usually
willing to pay the cost of developing this capital, even if they cannot quantify the
value to be received. For example, managers and companies will release employees
for weeks, months, or even years so that they can attend seminars, classes, or degree
programs. This is done with the implicit assumption that the employee is worth
more to the company upon returning from such as experience.

The third type of resource is Social Capital [9]. Social capital resides in the
relationships between individuals that allow them to add value to a product during
its development or develop a product more efficiently. Social capital is not a
property of the individuals in an organization- it is a property of the relations
among individuals. As with physical and human capital, social capital is not
completely fungible- the development of a certain type of social capital may be
productive for certain tasks, but have no effect or be harmful for others. Managers
often do understand that the ability' of their employees to work together to
accomplish goals has a great deal of value, and that developing social capital can
return large benefits. Many organizational actions such as forming ad hoc teams,
creating liaison roles, or adopting a matrix structure are undertaken to improve
social capital.

Research Method

To understand how organizations are using CAD technology, field interviews
were conducted at twelve manufacturing companies. The twelve companies
produced a wide range of different products, including college rings, plastic bottles,
jet engines, airframes, copiers, and automobiles. The goal of the interviews was to

• The nature of the design engineering process,

• The coordination demands of the design engineering process,

• The features and capabilities of CAD systems,

• The changes in the process and structure of the organization that have
occurred since the introduction of CAD technology,

• The nature of the companies' management and their attitudes toward CAD
systems, and

• The potential future changes that are enabled by the systems.
Interviews were conducted with a broad cross section of roles, including

design engineers, managers, CAD support personnel, and others. A total of 46
design engineers, 32 managers, and 22 CAD support personnel were interviewed in
the 12 companies. In addition, 39 individuals from other groups which assisted the
design engineer in his or her work (such as analysis or manufacturability groups)
were also interviewed. An average of two days was spent in each company.

The goal of the interviews was to understand the design engineering process
and the role computer tools play in it. Investigating the use oi computer tools
throughout the product development process was outside the scope of the research;
it was decided to focus solely on the design engineering phase of the process.

It was also decided that the research should focus on the design engineering
of the ten companies developing complex products. Significant differences were
noted between simple and complex product development processes. Thus the
conclusions reached in this paper will not necessarily apply to the design
engineering of all products.

The Design Engineering Process

Many authors have characterized the product development process as passing
through a number of stages. Myers and Marquis [21], for example, describe five

stages: recognition, idea formulation, problem solving, solution, and utilization
and diffusion. Roberts and Frohman [22] call for six by designating Myers and
Marquis' "solution" "prototype solution" and adding "commercial development" as
an additional, and we might add, a very important phase. Clark and Fujimoto [8]
return to five phases, based on the automobile development process: concept
generation, product planning, design engineering, process engineering, and
production. Building on these earlier formulations, we will present another
characterization of the product development tailored to the current research. The
goal is to build a general model of product development which corresponds well to
the many different complex product development processes studied.

In the products studied, the product development process can be described by
six phases. In the Recognition phase, the company identifies a need for a new
product or a potential application of a new technology. In the Idea Formulation
phase, a design specification for a new product is developed. In the Design
Engineering phase, this specification is translated into a set of detailed drawings so
that a prototype product can be built. In the Prototype Refinement phase, defects are
removed and additional improvements may be made to the prototype. After the
final version of the design is decided upon, the Process Engineering phase begins. In
this phase, the detailed design specifications are used to create a process design,
which may include flow charts, plant layouts, tool and die designs, etc. When this
phase has been completed (as judged by the success of a pilot production run), the
Utilization and Diffusion phase begins.

This research focuses on the Design Engineering phase of the product
development process. In the next section, the observations from these interviews
are presented.

The Field Interviews
The Design Engineer's Tob

Engineers in the design engineering phase of product development are often
organized functionally around either parts of the product or types of analysis (cf.
[17]). Design engineers may be assigned a part or group of parts and are then
responsible for completing the design of those parts (and are often responsible for
the parts during the later phases of product develpment). The organization of a
typical work process (for the companies studied) can best be described by an example.

The design engineers responsible for the engineering of a gas turbine engine
are usually organized around the different parts of the engine- the compressors
(which compress and heat air), the turbines (which drive the compressors), the
burners (which mix compressed air with fuel and ignite the mixture), the shaft and
bearings (on which all rotating parts turn), the static structures, and the overall
systems (for circulating air, fuel and oil around the engine). The engineer
responsible for the engineering of a turbine blade would report to the manager of a
Turbine group, which is responsible for all parts in that area of the engine. This
engineer would work with the engineers responsible for the parts adjacent to his:
engineers in the Static Structures group, the Shaft and Bearings group, the Systems
group, and the group responsible for the next and previous stages in the engine
(possibly a Compressor group and the Burner group).

In addition, this engineer may be required to work with many other groups.
For example, for the design of a turbine blade, the design engineer must work with
the Aerodynamics group, who provide airfoil shapes to the engineer; the
Aeromechanics group, who test for vibration properties; the Heat Transfer group,
who test the temperature gradients to ensure that the blade is cooled properly; the
Stress and Life Analysis groups, who check the stresses on and wear of the blades
and predict how often they would need to be replaced in the field; the Drafting

group, who add dimensional information and additional views to the design to
prepare the design information for manufacturing; one or more testing groups, who
are responsible for completing the necessary engine certification or qualification
tests; a performance group, which tests the output of the engine to ensure that
specifications are met; and a Manufacturing representative, who ensures that the
part can be built for a reasonable cost in a reasonable period of time. Other groups,
such as a materials research laboratory, may occasionally be involved.

The engineer responsible for a part must balance the demands of each group,
as well as work within his own constraints. The Aerodynamics group prefers thin
airfoils for optimal airflow characteristics, while the Aeromechanics group prefers
thicker, stiffer blades to reduce vibration. The Heat Transfer group may require
cooling channels within the blade, which may conflict with both the Aerodynamics
and Aeromechanics groups' goals. Against each of these groups' demands must be
balanced the cost and weight of each design alternative. Finally, a design which
works well may not be durable, maintainable, or manufacturable, and thus may
have to be redesigned, necessitating changes undesirable to any or all groups.

The overall job of the design engineer is one of balancing conflicting
requirements. Unfortunately, this is often done by incorporating different
perspectives sequentially and iteratively converging on a solution (cf. [6]). Many of
the design engineers, when asked to describe their job, used the term "project
manager." They describe their job as one of coordinating a group of people who
carry out the bulk of the design and analysis work. Most of the design engineers'
time is spent coordinating efforts between the different groups.

Design engineers also perform design and analysis work on their own. The
analysis performed by the engineers was largely described as "quick and dirty"-
analyses to understand the feasibility of new ideas, or to check the accuracy of results
generated by other analysis groups. The engineers do perform some design work.


which involves generating new design possibilities. This (in most companies) is the
smallest part of the design engineers' work.

CAD Systems

CAD technology has been applied in many different ways to design
engineering work, and design engineering work was changed in many different
ways with the introduction of CAD systems. The divergence in the experience of
different companies was due in part to the different types of products being
engineered and to variations in the capabilities of different CAD systems. Yet
similar systems were sometimes applied to similar tasks in very different ways.
Two different groups in the same company doing similar tasks sometimes have
significantly different experiences with the same CAD system. One cause of this
variance, the design engineers believed, is the variance in managerial attitudes.
Managers structure the work for design engineers. They allocate resources to and set
deadlines for engineers. They have a large voice in determining how engineering
work is carried out. Managers' attitudes about how CAD systems should be applied
to engineering work varied widely.

CAD Systems as Electronic Drafting Boards

Some managers saw CAD systems as electronic drafting boards. This is
understandable, as many managers gained their engineering experience before CAD
systems were developed and may not have had the time to learn the new
technology. Managers of this type often view CAD as a drafting board with some
additional (and, to them, mysterious) features. They see their subordinates'
performance improving on some tasks, but declining on others. Many adopt an
attitude similar to that of one manager interviewed: they view the CAD system as a
drafting board with a "rr,agic button." For some tasks, the productivity advantages


of the CAD system were undeniable. The production of a slightly different version
of a previously designed product was accomplished by the CAD system very quickly.
But this magic button did not work for other tasks, such as the design of new and
complex parts. For tasks such as this, this manager will require that the task be done
his way (which often means on the drafting board).

CAD Svstems as an Engineering Support Tool

Other managers believed that designing on a CAD system is a process very
different than designing on a drafting board, especially if the CAD design is done in
three dimensions. Such managers would require that all design engineering be
done in three dimensions; they stated that, while designing in three dimensions is a
more difficult and more time-consuming process, it provides significant benefits.
Designing in three dimensions requires greater mental involvement with the part
to be designed- it requires that the entire part be considered simultaneously and
completely. Design in two dimensions allows some cheating- some important
design details can be ignored and left for a downstream process to determine.

For example, designing the shape of a turbine blade in two dimensions is
done by specifying the cross-sectional dimensions of the blade at various points
along the blade. During construction of a prototype, a blade is produced by
connecting the cross sections with straight lines. The design engineer will then
decide how to fill or smooth this prototype to produce the final surface dimensions-
a process in which errors can be introduced. When designing in three dimensions,
the exact shape of the turbine blade must be specified, which requires a good deal
more effort on the part of the design engineer. The result of designing in three
dimensions is that (1) the time to produce the first set of engineering drawings is
-lengthened and (2) the downstream processes (e.g. the production and testing of
prototypes) are shortened.


Some managers also stated that the analysis features of the systems allowed
for a better simulation and testing of the design before any hardware was built. For
example, an engineer in one automobile company built a model of half of an axle
and suspension system. He was able to simulate the performance of an actual axle
travelling over a rough road. The simulation results corresponded well to test track
results. The simulation provided the engineer information about the areas of the
axle and suspension that were likely to fail. In the field, the failure of a part,
production of a new redesigned part, and retest of the part would take at least six
months to complete. The use of a CAD simulation allowed the same cycle to be
completed in one to two days. The result was that the design engineer had a better
understanding of the design, and a better first prototype was completed. Because of
the complexity of the analysis, however, the first prototype was completed behind

A similar experience was reported in a gas turbine engine manufacturer. To
certify an engine for commercial use, the engine must survive the ingestion of birds
(as this can occur during flight). The intersection of a simulated bird with a
simulated turbine blade was used to test the performance of the engine in this
respect. Again, the results of this simulation improved the engineers'
understanding of the design and the quality of the first prototype.

CAD Systems and Communication

CAD systems in many companies were used to improve communication.
Some engineers, when trying to explain a design concept to others or resolve a
design conflict with others, will often coordinate with others in front of a CAD
terminal. The rich representation of the design available on the CAD terminal
helps ground conversations and minimizes misunderstandings. CAD systems can

1 3

Online LibraryDavid C. (David Chandler) RobertsonEvaluating the use of CAD systems in mechanical design engineering → online text (page 1 of 3)