Week 5 - Conceptual Design

Reading

Announcements

  1. Two machine shop safety classes are offered every week.

    • First class lasts 30 minutes and is required
    • Second class is offered twice per week and is limited to 5 people

    Sign up now and take the safety classes this term. If you haven't finished the safety classes, you won't be allowed to work in the shop.

  2. Workbench assignments are listed on a table on the project page

  3. Next week (Feb 12) we have a Guest Lecture by William Nesbit, a professional technical writer, speaking on "Writing with Clarity". Mr. Nesbit's presentation begins at 9:00 AM

  4. In two week (Feb 19) we have a Guest Lecture by Jill Murfin, PSU BSME Alumna and Engineering Manager at Nike. Jill give a presentation titled "Thoughts & Tips for Teamwork".

Four Key Activities of Conceptual Design

If the conceptual design process had a recipe, it would look like this:

  1. Generate ideas: creative, generative, expanding possibilities
  2. Evaluate ideas: analytical, reductive, eliminating or combining possibilities
  3. Identify subsystems: logical and functional organization
  4. Set performance metrics for subsystems: recursive refinement of system-level metrics

It is helpful to list these activities as if they are separate steps because these activities have distinct objectives and engage different modes of working. But the four steps are not completely separate, and the four steps are not executed in a perfectly linear order. It is natural that your mind, at least subconsciously, works on other steps of the process while you are primarily engaged in one of the steps. For example, while working to identifying subsystems (step 3) or setting performance metrics for subsystems (step 4) you are likely to come up with new ideas (back to step 1). While setting performance metrics for subsystems (step 4), you may discover a better way to decompose the subsystems (step 3).

The four steps listed above are not a literal recipe to be followed rigidly. However, these four step are a useful guide that allows an engineering team to focus on a specific task. By engaging in a specific task (one of the four steps), the team advances the design. For example, the team may decide to have a meeting focused entirely (or at least mostly) on the task of generating new ideas.

Within the conceptual design process, the four key activities are iterated and revisited, sometimes in non-sequential order. While these activities are underway the team is also engaged in updating the project plan, and possibly doing preparatory work for later stages. For example, if subsystem engineering will require fabrication (as is very likely), the team may need to initiate purchase of materials or reserve time in the machine shop or seek out vendors with specialized manufacturing capabilities.

Although the entire design process is described in a series of phases, and the activities within each phase are described in a series of steps, the entire design process is more messy that our phase and step descriptions suggest.

Inexperienced and undisciplined designers may argue that because the process is messy, it is useless or even counterproductive to attempt to impose structure or to use systematic processes. While rigid adherence to plans and structure can be problematic, the other extreme – a total lack of planning – is certain to cause wasted effort and lower productivity.

1. Generating Ideas

Activities to generate ideas encourage and rely upon creativity. You want to generate many ideas without too much worry about probability for success or potential for positive impact. That evaluation happens after you have generated the ideas.

Idea generation can be described as consisting of internal and external searches.

  • Internal search is what the team does to create ideas within their own sphere of knowledge and experience: brainstorming, brain-writing, mind-mapping, analogs, …

  • External search is what the team does to find ideas from outside their own direct knowledge and experience: analogs, competing products, catalog search for components that can be purchased and incorporated into the design.

Last week we discussed several tools for internal search and we practiced the Crawford Slip-writing technique.

Case Study: Design of hot glue pot

2. Evaluating Ideas

Evaluating ideas is the reductive complement to generating ideas.

To evaluate ideas the design team can use screening techniques to compare competing options based on estimates of performance. Obviously skill at estimating performance will be easier with accumulated engineering experience. For specific techniques, refer to the Controlled Convergence, Scoring Matrix, and Screening Matrix method in part 2 of the textbook by Mattson and Sorenson.

Ullman describes a multi-stage screening process for identifying the most promising concepts. Dieter and Schmidt describe these methods in greater detail.

  1. Absolute scale rating
    • Feasibility screening: Using "gut feel" whether idea is likely to work
    • Technology readiness: Is the technology for the solution easy to obtain or implement?
    • Go/no-go screening: Does or can the solution meet the customer requirements
  2. Relative rating: Pugh Matrix

Review lecture notes and lecture slides from ME 491 on Concept Selection

Engineering Analysis

Simple engineering analysis can also be used at this stage. Like all activities in the concept evaluation stage, engineering analysis should be as efficient as possible. The precision of the analysis should be sufficient to decide whether an option is feasible, which is information that can be used in screening methods described above. During concept evaluation at the conceptual design phase, the team should not invest effort in optimizing any option.

Engineering analysis at the conceptual design state should also focus setting boundaries on expected performance of the system as a whole or of potential subsystems. For example, you can think about total energy budget (electrical input, motor torque, total load), weight budget (total and distributed to subsystems), cost (components, materials), time resources (build or buy), and which design goals need to be compromised or balanced. This information will provide concrete and rational guidance to the screening methods described above.

Low Fidelity Prototypes

Low fidelity prototypes are simple mockups that are helpful in evaluating ideas. "Low fidelity" expresses the idea that the mockup is not intended to be a close representation of the final device or system. Rather, the goal is work quickly to generate ideas and expose potential conceptual weaknesses in a design idea.

Low fidelity prototypes usually have limited functionality, or in the case of a cardboard and tape mock-up, may have no practical function other than providing a physical representation of an idea. Questions at this stage could be, "Is this too big to handle?", "How do you hold this when you use it?", "Should there be a button or switch to control the operation?", "Can this be made to fit into the available space?", "Should the user be able to separate these parts or should it be a single sealed unit?", "Does this look like something else I've seen before?" "What is the first thing you think about when I hand this to you?" etc.

Although the evaluation process is aimed at vetting and reducing possibilities by eliminating unlikely paths to success, it is also very likely to be the source of new ideas.

3. Identify Subsystems

Use functional decomposition or other strategies to identify subsystems of the solution. Refer to Decomposition in the second half of the textbook by Mattson and Sorenson.

Functional decomposition involves identifying the physical actions or behaviors of the parts of the system to be designed without regard to the physical components that might be used to create those actions or behaviors. This decomposition focuses on logic (e.g. this causes that) and transformations (e.g. electrical energy in, movement out). Functional decomposition describes what the system and its components must do without respect to the mechanism or physical structure that would embody the doing.

There are other ways to decompose a system. Perhaps the simplest idea is to figure our which parts of the system can be purchased (e.g. motors, control systems, heater assemblies,…) and which have to be produced through a custom design. This decision should probably be applied iteratively throughout the design process. For example, to do testing on an early prototype (either of the whole system or a subsystem), it may be advantageous to substitute readily available commercial components (think oversized motor, or simple switches instead of control systems) so that the team can quickly learn from the prototype.

4. Set Performance Metrics for Subsystems

The subsystem performance metrics must be consistent with the performance metrics for the entire design as specified in the Requirements-Measurements matrix created in the Opportunity Development phase. Performance metrics for subsystems will be more numerous and more detailed than the system level metrics. The purpose of performance metrics at any stage in the design process is to help the design team make decisions.

References

  1. Christopher A. Mattson and Carl D. Sorenson, Fundamentals of Product Development, 4th ed., 2016, Brigham Young University.
  2. George E. Dieter and Linda C. Schmidt, Engineering Design, 5th ed. 2013, McGraw-Hill, New York.
  3. David G. Ullman, The Mechanical Design Process, 1992, McGraw-Hill, New York

Document updated 2018-02-05.

Go back to the main Lectures page.