This lecture is intended to familiarize engineering students with the nature of engineering design and how it is done. We start by drawing the distinction between classical mathematics and science problem solving on the one hand and engineering design the other. We then examine a definition of engineering design and present a nine-step model of the engineering design process. We introduce the notion of engineering systems and address design issues at the system level and explore the evolution of design ideas from their inception to implementation.

Many students cite their interest in science and mathematics as one of their reasons for wanting to study engineering. They base this perception of engineering on their high-school experience in solving abstract mathematics problems. They also recall their experience with science based problems that require combining mathematical manipulations with applications of scientific principles. This perception that engineering is only "applied science and mathematics" is reinforced by traditional engineering curricula, which emphasize science and mathematics courses during the first two years and specialized applications of science and mathematics in what are typically called "engineering science" courses. Whatever the particular field of mathematics or science used in these types of problems, they tend to have four features in common:

• The problems are well-posed in a very compact form. By well-posed we mean that the statement of the problem is complete, unambiguous, and free from internal contradictions. If it didn't have these features, the students would complain vigorously and the teacher would apologize profusely for presenting a poorly stated problem.

• The solutions to each problem are unique and compact. There generally is a single correct answer available, i.e. a number, a set of numbers, or symbols.

• These problems have a readily identifiable closure. It is easy to recognize when the answer has been obtained.

• These problems require application of very specialized areas of knowledge and there is little doubt what the subject is for each problem.

Solving problems that have some or all of these four characteristics is an important part of engineering education. It develops and strengthens specific analytical skills that are essential in most engineering design situations. However, most real world engineering design problems do not share these characteristics. In particular, many real engineering design problems are poorly posed and do not have a unique solution or a readily identifiable closure.

Engineering science courses deal with applications of scientific principles and mathematical concepts for analyzing a wide variety of engineering problems. Students who have mastered the skills of solving traditional mathematics and science problems but have not had prior exposure to design may find it difficult to adjust to this less precisely defined world of real engineering design problems. Much of the emphasis in this lecture is on appreciating the less-precise nature of engineering design and on developing and utilizing new skills for being a successful design engineer.

The world of traditional mathematics and science problem solving can be compared to standing on a rock-solid surface at the top of a cliff. The firm foundation provided by the unambiguous and never changing laws of science and rules of mathematics is a comfortable place for most students about to embark on an engineering curriculum. In contrast, the world of engineering design involves many uncertainties, ambiguities, and inconsistencies; it is like a swamp at the base of the cliff. It is very difficult to get a firm footing in the swamp, and a completely different set of survival skills are needed.

Because subjective considerations tend to be much more prevalent down in the swamp of engineering design as compared to the objective nature of analytical life up on the cliff, the relationship of the engineering design instructor to engineering students is fundamentally different.

The mathematics, science, or engineering science instructor is an expert in their field, and education consists of a one-way transfer of some of that expertise to the student. While there are many modes for facilitating that transfer

(lectures, interactive problem solving, discussion sessions, textbooks) the dominant direction is that the instructor transmits, and the student receives, objective information. The instructor presumably knows the answers, and with luck by the end of the course the student will have learned enough of the right answers. But since design is much more subjective, there rarely is a single "correct" answer. Judgments as to whether one design alternative is superior to another may be highly dependent on the values and preferences of the evaluator.

The design instructor is not so much a transmitter of facts, but a facilitator of the design process and a partner with the students in searching for successful solutions of design problems. The design instructor is less like a basketball referee who determines whether the actions are consistent with the rules and more like a fishing guide whose experience can make a fishing trip more enjoyable and productive. The guide can point out the logs and boulders that are scattered throughout the swamp, and provide you with a pair of hip boots to make your journey more pleasant. Studying design will help ease entry into the swamp and make your experience not only survivable but enjoyable. It won't remove the subjective considerations and uncertainties associated with design problems, but it will help you adapt to this new environment and function effectively in it.

Engineering design is a more advanced version of a problem solving technique that many people use routinely. The general procedure for solving real everyday problems is straightforward: A problem is encountered, information about the problem is obtained, alternate solutions are formulated, and the best alternative is adopted. Some problems are so straightforward and solutions so obvious that people solve those problems without being consciously aware of the specific steps in the process. When the problem is more complex (as most engineering problems are), an organized and methodical approach is needed. Engineering design is a methodical approach to solving a particular class of large and complex problems.

The Accreditation Board for Engineering and Technology (ABET) defines engineering design as follows:

The engineering design component of a curriculum must include at least some of the following features: development of student creativity, use of open-ended problems, development and use of design methodology, formulation of design problem statements and specifications, consideration of alternative solutions, feasibility considerations, and detailed system descriptions; Further, it is essential to include a variety of realistic constraints such as economic factors, safety, reliability, aesthetics, ethics, and social impact.

ABET identifies the goal of engineering design as, ". . . devising a system, component, or process to meet desired needs." Note also that the result of design might not be a physical piece of hardware; it can be a process. This latter kind of design is of particular interest to chemical engineers, materials engineers, industrial engineers, and computer software engineers.

In very simple terms, "Engineering design is a swamp".

In this section we briefly outline a nine-step model of the engineering design process. Before discussing each of the nine steps in this model, a few general comments are in order. First, it is important to recognize that any model is a simplified description of a more complicated reality. The value of a model lies in its ability to help us organize our thoughts and gain insight into important aspects of reality. So keep in mind while we discuss these nine steps that actual designs do not necessarily evolve in a linear, orderly progression from step one through step nine. Not every step will be used to the same extent in every design, and some steps may be performed out of order.

Many engineering designs are performed by teams of engineers and not every team member participates in every step of the process. Some team members may be specialists in one or more of the nine steps. In many situations, design engineers unconsciously blend some of these steps together. Also, each step may be revisited several times during the evolution of a design. However, even experienced engineers will regularly step back from their immersion in design details and rely on such a model to assure themselves that they haven't overlooked key elements in their search for a de-

sign solution.

The first step in the design process establishes the ultimate purpose of the project via a general statement of the client's dissatisfaction with a current situation.

Once the need has been established, the next step is to translate the "needs" statement into one that addresses how we propose to satisfy the need. A problem statement generally consists of three components:

• Goals
• Objectives
• Constraints

After the problem has been defined, an overall plan should be developed. A good project plan will help you identify what tasks need to be accomplished and in what order. On the other hand, there may be very little information available if the design task is in an area which has never been considered before.

This step is where possible design solutions are first envisioned. The principal effort is to generate a wide range of design options. This is where you call upon your creativity and imagination to develop design approaches that have the potential to satisfy the objectives and constraints defined in Step 2. This activity involves a mode of thinking that is quite different from the analytical mindset so widely used by engineers.

Each design problem requires a unique combination of information sources. When designing in an area which is well developed, there is most likely a wealth of information on similar designs that have already been completed, and on what constraints are imposed on design possibilities by codes and standards.

This step is where possible design solutions are first envisioned. The principal effort is to generate a wide range of options. This is where you call upon your creativity and imagination to develop design approaches that have the potential to satisfy the objectives and constraints. This activity involves a mode of thinking that is quite different from the analytical mindset so widely used by engineers.

Once alternatives have been conceived, they must be evaluated to determine the extent to which they satisfy the design objectives and constraints. This is done by using mathematics and the engineering sciences to evaluate the performance characteristics of each design alternative. This is the only firm ground within the design engineering swamp. Other analytical techniques that are important parts of the engineering design process, but which are not included in many engineering curricula, include estimating the cost of production and determining the probability of the design suffering a catastrophic failure.

Back to the design swamp! Now it is time to select the best alternative. At first glance this might seem to be trivial; after all, won't the results of Step 6 tell us which alternative performs the best? Well, maybe and maybe not!

Communication is essential in all aspects of design in order to inform interested parties of project progress and status. Communication also provides a record for the design engineer to reconstruct and justify what he/she did and why they did it. This is especially important after a prolonged period of inactivity on a project (a common occurrence for engineers with several simultaneous projects). Good communication clears the fog and allows other engineers to readily follow the route you forged through the swamp. Communication also serves as a bridge between the design engineers and the rest of the world; it helps clients understand what comes out of the design swamp without having tp get muddy themselves. Without communication, everyone would be stuck in the same swamp; each reinventing the wheel in their own way

Some people consider design and implementation to be separate activities. But design activities are intended to solve a problem and to do so, action must occur. Implementation refers to this translation of design concepts into actions; it is a natural step in the design process. The ABET definition of engineering design in refers to implementation using the words construction and testing.

Traditionally, designers tossed their designs "over the wall" to manufacturing or construction engineers who were responsible for implementation, but had no influence on the design. As a result, costly designs were implemented when less costly approaches might have been available. Modern approaches that integrate implementation considerations into the design process include "design for manufacturing" and "design for assembly."

A system can be defined as a collection of elements that interact with each other to fulfill a function. As mentioned, the object of any engineering design effort can be envisioned as part of a very large complicated system. However, for practical reasons we must recognize that we cannot take on the whole universe; we must limit our attention to those system elements that are the most important factors for the specific design activity. We do this by establishing a system boundary or set of boundaries that separates the universe into two parts: the system under consideration and the external environment. These boundaries either isolate the system from the environment or provide the mechanism by which the system interacts with the environment. Engineers and their clients define systems and their boundaries as the object of their design efforts during the problem formulation stage of design.

The smallest identifiable element of a system is called a component. Collections of components are called subsystems when not all elements of the system are included in the collection.

Another aspect of system design does not directly involve the design of any particular element; that is the design of the spatial relationships of the elements to each other. Take a look under the hood of your automobile. How did they ever decide where to put the battery, the windshield washer, or the alternator? And how did they ever get everything to fit? We refer to this aspect of system design as configuration design. While the immediate focus of configuration design is the arrangement of system elements, clearly decisions about the location, size, and shape of the space available for an element could affect the design of that element.

Depending on the complexity of the system, this progression can happen hundreds or thousands of times. Sometimes the design activity reverses direction as decisions made at a lower level in the system hierarchy require that prior decisions made at higher levels be reconsidered. Design of complex engineering systems requires a great deal of patience to wait for the overall solution to emerge from solutions to many individual pieces. Another important ingredient of system design is detailed documentation of ongoing modifications to system components and continuous communication among team members as these modifications are being considered and implemented.

The eight-step description of the life cycle of engineering designs includes:

1. Problem Definition
2. Feasibility Study
3. Preliminary Design
4. Detailed Design
5. Production
6. Distribution
7. Consumption
8. Retirement

This model identifies a natural progression of a design through stages, although we may need to retrace steps at any stage to modify decisions made at earlier stages. The sequential perspective emphasizes the advantages of systematically studying, understanding, and focusing our energies on the early stages of design since the cost of retracing steps increases dramatically in the latter stages The proverb that "an ounce of prevention is worth a pound of cure" applies as much to engineering design as it does to anything.

In this phase, the engineer addresses the fundamental questions of whether to proceed with the project. Just because someone has articulated a need doesn't mean that it is reasonable, desirable, or possible to fulfill that need. There may be technological and economic limitations:

Other feasibility considerations include assessing the market. If the object to be designed is envisioned as a mass-produced product, a market study of the sales potential for the product may be warranted before justifying the cost of establishing a manufacturing capability. On the other hand, the feasibility of a large public works project, such as a highway bridge or a sewage treatment plant involves assessing the reaction of the community in which the facility is to be located.

Clearly, other professionals besides engineers are likely to be involved in feasibility studies. Some of the important skills needed in this phase of design are financial analysis, marketing, and community relations. Design engineers involved in feasibility studies have to interact with these other professionals, so good people-to-people skills and communications skills are extremely important.

Many times the question to ask is not whether a project is feasible, but whether it is feasible for the firm to take on the project at a specific time. It may involve an unfamiliar technical area, or other projects to which the firm is currently committed may preclude allocating sufficient resources to the proposed project. The company may decide that it doesn't want to deal with the specific uncertainties associated with the project. All these considerations should be dealt with during the feasibility study before a commitment is made to pursue extensive design activities.

Design alternatives begin to take shape in the minds of engineers, or as part of conversations among engineers exploring ways to satisfy the stated needs. A key step in the early phases of design is capturing these ideas in the form of free-hand sketches. These sketches, which at first, rnay be not much more than doodles, may contain suggestions of key features, such as overall form and ways to decompose the system into components. They may also indicate aspects of the loading and other operational conditions. The design engineer's thoughts about possible materials, size, and other features may begin to crystallize and be jotted down along with the pictorial representation. Generally these sketches and supporting notations precede any formal analysis of the design.

Strive to generate a large number of preliminary designs. If you are successful, more design concepts will emerge than we havetime to formally analyze or to begin detailed design. An important part of the preliminary design phase is to winnow down the design alternatives to a relatively small number. In this process of narrowing down the options, new options are bound to arise as we refine the concepts. This temporary enlargement of the number of options is followed by another contraction phase until we converge on a single approach or small number of approaches. This process of alternating between concept generation and elimination is called controlled convergence.

Once a single design concept or a small number of design options are selected, we move to the phase of refining it (them) in greater detail. Formal analysis helps to define sizes. choose materials, estimate costs, and plan procurement and fabrication requirements. These decisions are made at the individual component level and overall system level. Generally this phase results in a set of formal engineering drawings (either hard copy or computer files) for each component. These drawings provide complete unambiguous instructions for making that component. Also, assembly drawings are produced that display how the components fit together to form subsystems and how the subsystems fit together to form the system.

Models and prototypes: As the main features of the design are defined, it will be helpful to construct either a scale model or a full-scale prototype. Models and prototypes provide a three-dimensional treatment to designs so that they can be examined from any perspective to inspect for conformity to expectations, and so the design can be viewed in the context of its surroundings. Since these models and prototypes are intended primarily to display shape, they usually are made of an inexpensive material that is easy to work with. Sophisticated computer-based three-dimensional modeling techniques now allow engineers to create images of designs that can be maneuvered and rotated so that the design can be viewed from any angle and against any backdrop. These effects include shading from light sources located virtually anywhere relative to the object and animated "walk-throughs" to examine designs from the inside. These latter effects are particularly useful for designs of buildings and other large structures. The availability of these techniques substantially reduce, but do not totally eliminate, the need for prototypes. In particular, clients may need a physical object that they can inspect at their convenience, and they may not be content with images on a computer screen.

Rapid prototyping. Several new technologies for rapidly producing prototypes gained popularity within the engineering design community in the early 1990s.These techniques produce three-dimensional scale models in hours instead of days or weeks. Most rapid prototyping methods automatically produce models directly from computer-aided design (CAD) files of the object. One of the advantages of these rapid prototyping technologies is that the prototypes are constructed by automated machinery that doesn't require continuous human attention. Depending on the particular technique and materials used in rapid prototyping, surface finish or tolerances may be of low quality.

Production prototypes. Once design details have been established. Another type of prototype may be built. This may be a full-scale prototype fabricated from the same material as the anticipated production version. Its role may be to verify fabrication requirements or to be tested for strength or other performance characteristics.

Testing: Part of the detailed design phase may also involve a testing program. Many different kinds of tests can be carried out as part of this phase of design. Testing scale models of airplanes in wind tunnels help airplane designers decide on wing shape and placement of engines. Tests on production prototypes may be used to verify the performance analyses. In some cases, applicable codes and standards may require that production prototypes be tested to failure.

Once the design has been finalized, it is time to implement it by producing the product. Very large one-of-a-kind systems, such as buildings, tunnels, and petroleum refineries, are constructed on-site, although some components and subassemblies may be fabricated off-site and assembled. on-site. Very small products intended for mass production are produced in a manufacturing facility and shipped to the customer. In between these two extremes are large, mass-produced products such as automobiles and airplanes that involve many components and subassemblies produced in dispersed locations and then shipped to a central assembly facility. For example, the engines, landing gear, and major segments of the fuselage and wings of airplanes may be built by subcontractors and assembled by the prime manufacturer.

Design for manufacturing and assembly. We can think of the latter phases of detailed design in which the design starts to "move off the paper" as the beginning of the implementation step of the design process model. Consider a design process that does not include implementation.

How does one plan for implementation? It may be helpful to ask, "How would I implement this?" and, "Is there an easier way to do this?" Another approach is to seriously consider implementation during Step 2 of the design process, and explicitly incorporate these requirements into the Problem Statement. It also might be helpful to assume that you were responsible for monitoring and supporting your design solution over its entire life. Then before committing to a solution, consider its implications. Foresight will help you identify the benefits and shortcomings of alternative designs. You will find more elegant solutions to problems, and you will be able to avoid implementation quagmires before they arise. Of course, no matter how thoroughly implementation is planned, minor adjustments and refinements will still be necessary at Step 9. But if implementation requirements were anticipated, this will be a time of pride and accomplishment.

Techniques for incorporating these production considerations into early phases of design are labeled "design for manufacturing and assembly." Design and production can be integrated even more tightly so that certain production activities can begin even before all design details have been completed. This approach is known as "concurrent design."

Taguchi method. When size and other features of designs are specified, we recognize that it is virtually impossible to make the product so that these specifications are exactly satisfied. The traditional approach to dealing with these deviations is to incorporate tolerances into the detailed design. Product quality is maintained by an inspection procedure during production which rejects brackets in which the hole diameters fall outside of this acceptable range. A more sophisticated approach using statistical concepts to incorporate product quality into the early stages of design is known as the Taguchi method. More generally, the Taguchi method allows engineers to control design parameters so that the design is less sensitive to variations that can degrade product quality.

Many mass-produced products are delivered to customers through complicated distribution channels. The distribution process may include shipment by several transportation modes and temporary storage in warehouses or other distribution facilities. The products themselves and/or their packaging must be designed to withstand the sometimes very rough handling and extreme environments they experience during distribution.

Many complex engineered systems require regular sophisticated maintenance programs after they have been delivered to the customer. Examples include commercial airplanes, bridges, electric power generating plants, and petroleum refineries. For such systems, ease of maintenance may be an important design objective. Design techniques and approaches that focus on this issue are labeled "design for maintainability." During their operational phase, many engineering systems consume resources and generate waste products. Many state and federal environmental regulations constrain the nature and amount of adverse environmental impact associated with operation of these systems. Keeping environmental performance as a major focal point during design comes under the heading of "design for the environment."

As an engineered system approaches the end of its useful lifetime, consideration must be given to its ultimate disposition. The importance of reusing and recycling parts of the system has increased as we have become more sensitive to problems associated with solid waste disposal. This leads to design philosophies and procedures that focus on the ease of dismantling the system and recycling its components. The terms "design for disassembly" and "design for sustainability" have been adopted to refer to this emphasis.

In this lecture we have introduced the concept of engineering design. We saw that many aspects of design, including definitions of what it is and how it is done, are amorphous. This lack of solid physical laws and rigid rules regarding engineering design activities may lead many engineering students to adjust their conceptions of this most fundamental of engineering activities.