This paper presents a grammar-based designer assistance tool for Epicyclic Gear Trains (EGTs) that helps designers generate concepts of EGT structures. The tool is comprised of three parts. The first is a grammar-based graph generation engine, the second is a functional schematic sketching module and the third is a concept selection module. Here we highlight the implementation of the first two modules. This is an initial exploration of using grammar-based design to build a practical designer assistance tool (DAT).

A patented double crank ring-plate-type cycloid drive is introduced in this paper. Its working principles, advantages and design issues are discussed. Some promising results of testing are also presented in this paper.

Grammars generate design options through the application of predefined rules that transform collections of symbols into more meaningful expressions. Research on the nature of grammars tells us that writing the rules is where the fundamental design activity occurs. Using the grammar rules allows us to explicitly articulate one design at a time. We can exploit the design power of grammars further by modifying a grammar to describe new languages of designs. Here we examine an existing grammar to demonstrate how modifying its rule base to relax an assumption can expand the space of solutions it generates significantly. We show that investing our design attention on the grammar itself can yield dramatic results.

Teams are employed in capstone mechanical design courses, as well as in other undergraduate courses. This puts students in a setting that they will encounter as professionals, which is beneficial to their development. For instructors, however, teams make it difficult to evaluate each individual’s understanding of the design process. While specialization is not as expected in undergraduate work, the level of variation between members of a team is not fully known. In this study design journals were kept by students during their capstone projects and entries were categorized using a coding scheme. This data revealed what a typical student recorded as activity during a capstone design project. Additionally the amount of variation within a team and among students as a whole was examined.

Successful group work is a critical facet of engineering design. In addition to simply increasing the manpower available to solve a particular problem, team settings also leverage the variety of thinking offered by each individual. There are any number of functional behaviors that one might offer a team. This paper explores the idea that design journal records can provide key insight into the functional activities taken on by team members. The main questions asked by this study include (1) Can design journals provide consistent information about an individual’s functional behavior? (2) Are individuals likely to shift behaviors over the course of the project’s timeline? (3) Do individuals adapt to compensate for design behavior changes in other team members? The data in this study was collected from five volunteer teams of five to six undergraduate students each (total N = 29). At the time of their participation in the study, the volunteers were enrolled in the University of Maryland’s mechanical engineering capstone design course, Integrated Product and Process Development. Four of the teams participated in fall 2015 and one team participated in spring 2014. Participants maintained individual journals throughout their semester-long design projects, after receiving a set of guidelines to steer but not restrict journal usage. After the projects were completed, journals were coded using a previously validated coding scheme which associates each journal entry with a design behavior class and activity. Findings for this pilot study were very encouraging. Design journals proved to be a viable source of information about the functional behaviors of individuals. The majority of the undergraduate participants in this study were found to regularly shift their focuses to different behaviors as their projects progressed. When these shifts occurred, other team members regularly shifted to accommodate the change and maintain overall similar levels of activity in each area of functional behavior for the team as a whole throughout the project’s timeline.

Nature is a powerful resource for engineering designers. The natural world provides numerous cases for analogy and inspiration in engineering design. Transferring the valuable knowledge and inspiration gained from the biology domain to the engineering domain during concept generation is a somewhat disorganized process and relies heavily on the designers’ insight and background knowledge of many fields to make the necessary leaps between the domains. Furthermore, the novice designer approaching biology for inspiration tends to focus heavily on copying the visual attributes of a biological system to develop a solution that looks like the biological system rather than explore at deeper levels to uncover relationships that lead to the development of true analogies. There are now well-known methods for teaching bioinspired design in engineering and the majority of methods prescribe the use of analogies in order to facilitate knowledge transfer, however, guidance in analogy formulation to foster the creative leaps is missing or ill defined. Thus little is known about how students use biological systems for design inspiration. This paper proposes categories for analogical knowledge transfer in bio-inspired design to foster and characterize diverse analogical knowledge transfer. The proposed analogy categories are used to describe the behavior seen in an engineering class. Results indicate that (1) single biological system provides multiple analogies that result in different engineering inspiration for design; (2) biological information from multiple categories is transferred during concept generation; and (3) non-physical characteristics may inspire more sophisticated engineering inspiration than those based on physical characteristics alone. Overall, the analogy data classification has resulted in a better understanding of analogical knowledge transfer during bio-inspired design and leads to best practices for teaching bio-inspired design to engineering students.

Designers and design researchers both agree that developing many feasible alternatives at the conceptual design stage is useful. In this paper we introduce generative configuration design (GCD) for conceptual design. We provide a partition of knowledge accessed during GCD and use the partitioned knowledge foundation to compare design tool architectures so that computational improvements can be made. We present an improved architecture for a GCD algorithm and implement it as a tool for office chair design. Subsequent examples show tradeoffs between computational load and design variety when applying constraints for behavior testing.