Daniel D. Frey

Engineering design thinking, teaching, and learning

This paper is based on the premises that the purpose of engineering education is to graduate engineers who can design, and that design thinking is complex. The paper begins by briefly reviewing the history and role of design in the engineering curriculum. Several dimensions of design thinking are then detailed, explaining why design is hard to learn and harder still to teach, and outlining the research available on how well design thinking skills are learned. The currently most-favored pedagogical model for teaching design, project-based learning (PBL), is explored next, along with available assessment data on its success. Two contexts for PBL are emphasized: first-year cornerstone courses and globally dispersed PBL courses. Finally, the paper lists some of the open research questions that must be answered to identify the best pedagogical practices of improving design learning, after which it closes by making recommendations for research aimed at enhancing design learning.

Regularities in Data from Factorial Experiments

This article documents a meta-analysis of 113 data sets from published factorial experiments. The study quantifies regularities observed among factor effects and multifactor interactions. Such regularities are known to be critical to efficient planning and analysis of experiments and to robust design of engineering systems. Three previously observed properties are analyzed: effect sparsity, hierarchy, and heredity. A new regularity is introduced and shown to be statistically significant. It is shown that a preponderance of active two-factor interaction effects are synergistic, meaning that when main effects are used to increase the system response, the interaction provides an additional increase and that when main effects are used to decrease the response, the interactions generally counteract the main effects.

Adaptive One-Factor-at-a-Time Experimentation and Expected Value of Improvement

This article concerns adaptive experimentation as a means for making improvements in design of engineering systems. A simple method for experimentation, called “adaptive one-factor-at-a-time,” is described. A mathematical model is proposed and theorems are proven concerning the expected value of the improvement provided and the probability that factor effects will be exploited. It is shown that adaptive one-factor-at-a-time provides a large fraction of the potential improvements if experimental error is not large compared with the main effects and that this degree of improvement is more than that provided by resolution III fractional factorial designs if interactions are not small compared with main effects. The theorems also establish that the method exploits two-factor interactions when they are large and exploits main effects if interactions are small. A case study on design of electric-powered aircraft supports these results.

Toward a taxonomy of concept designs for improved robustness

The term ‘robust design’ denotes various engineering methods intended to make a products function more consistent despite variations in the environment, manufacturing, deterioration, and customer use patterns. Most robust design methods are employed at the detailed design stage, but the benefits derived may be significantly higher if efforts are made earlier in the design process so that the design concept itself is inherently capable of being made robust. To make progress toward these ends, we studied a large number of inventions documented by US patents that claimed robustness as a key advantage over the prior art. We grouped these patents on the basis of the general strategies they employed. We found that the strategies can be usefully organized via the P-diagram. This paper will describe a few of these strategies by means of examples and explain the relationship of the strategies to the P-diagram.

Evaluating Process Capability During the Design of Manufacturing Systems

This paper introduces the concept of a process capability matrix-an ordered set of dimensionless parameters that capture information on a manufacturing systems response to disturbances. The matrix is similar to the process capability indices C p and C pk , but is extended to multiple acceptance criteria and multiple causes of variation. Equations are presented that use the matrix to estimate yield in manufacture of products with multiple acceptance criteria. The surface mount of large body electronic packages serves as an example of the effectiveness of the process capability matrix as a tool for design decision making.

An Evaluation of the Pugh Controlled Convergence Method

This paper evaluates a method known as Pugh Controlled Convergence and its relationship to recent developments in design theory. Computer executable models are proposed simulating a team of people involved in iterated cycles of evaluation, ideation, and investigation. The models suggest that: 1) convergence of the set of design concepts is facilitated by the selection of a strong datum concept; 2) iterated use of an evaluation matrix can facilitate convergence of expert opinion, especially if used to plan investigations conducted between matrix runs; and 3) ideation stimulated by the Pugh matrices can provide large benefits both by improving the set of alternatives and by facilitating convergence. As a basis of comparison, alternatives to Pugh’s methods were assessed such as using a single summary criterion or using a Borda count. The models we developed suggest that Pugh’s method, under a substantial range of assumptions, results in better design outcomes than those from these alternative procedures.© 2007 ASME

Validating Robust-Parameter-Design Methods

This paper discusses validation of design methods. The challenges and opportunities in validation are illustrated by means of an analogy with medical research and development. Inspired by this analogy, a model-based process is presented for validating robust-parameter-design methods. A hierarchical probability model was used to create a large number of polynomial response surfaces with hierarchy and inheritance properties similar to those of an ensemble of 62 engineering responses from published experiments. Various robust-parameter-design methods were applied to each response surface and the robustness of the resulting designs was assessed. It was shown that crossed array methods are more effective than combined array methods if there is a reasonable probability of three-factor interactions. It is also shown that one-factor-at-time methods provide even greater effectiveness than crossed arrays methods.Copyright

The Mechanisms by Which Adaptive One-factor-at-a-time Experimentation Leads to Improvement

This paper examines mechanisms underlying the phenomenon that, under some conditions, adaptive one-factor-at-a-time experiments outperform fractional factorial experiments in improving the performance of mechanical engineering systems. Five case studies are presented, each based on data from previously published full factorial physical experiments at two levels. Computer simulations of adaptive one-factor-at-a-time and fractional factorial experimental were carried out with varying degrees of pseudo-random error For each of the five case studies, the average outcomes are plotted for both approaches as a function of the strength of the pseudo-random error. The main effects and interactions of the experimental factors in each system are presented and analyzed to illustrate how the observed simulation results arise. The case studies show that, for certain arrangements of main effects and interactions, adaptive one-factor-at-a-time experiments exploit interactions with high probability despite the fact that these designs lack the resolution to estimate interactions. Generalizing from the case studies, four mechanisms are described and the conditions are stipulated under which these mechanisms act.

Evaluating Three DOE Methodologies: Optimization of a Composite Laminate under Fabrication Error

ABSTRACT This article aims at systematically comparing three different designs of experiments (DOE) methodologies used in the optimization of engineering structures. The selected methods are a response surface methodology, an adaptive one-factor-at-a-time search methodology, and a Bayesian DOE-based methodology. Six evaluative criteria are defined for the comparison. To perform the study, a simulation-based layout optimization of a four-layer composite laminate is used under a fiber misalignment fabrication error. It is found that each solution method satisfies a particular aspect of the six evaluative criteria and, thus, an application of a multiple criteria decision aid model is suggested to assist the analyst.

An Adaptive One-Factor-at-a-Time Method for Robust Parameter Design: Comparison With Crossed Arrays via Case Studies

This paper presents a conceptually simple and resource efficient method for robust parameter design. The proposed method varies control factors according to an adaptive one-factor-at-a-time plan while varying noise factors using a two-level resolution III fractional factorial array. This method is compared with crossed arrays by analyzing a set of four case studies to which both approaches were applied. The proposed method improves system robustness effectively, attaining more than 80% of the potential improvement on average if experimental error is low. This figure improves to about 90% if prior knowledge of the system is used to define a promising starting point for the search. The results vary across the case studies, but, in general, both the average amount of improvement and the consistency of the results are better than those provided by crossed arrays if experimental error is low or if the system contains some large interactions involving two or more control factors. This is true despite the fact that the proposed method generally uses fewer experiments than crossed arrays. The case studies reveal that the proposed method provides these benefits by exploiting, with high probability, both control by noise interactions and also higher order effects involving two control factors and a noise factor. The overall conclusion is that adaptive one-factor-at-a-time, used in concert with factorial outer arrays, is demonstrated to be an effective approach to robust parameter design providing significant practical advantages as compared to commonly used alternatives.