Official ASME Group

Journal of Mechanical Design

The ASME Journal of Mechanical Design (JMD) serves the broad design community as a venue for scholarly, archival research in all aspects of the design activity.
New Blog Post

Enter the date that this post should be published.

Are you sure?

Are you sure you want to delete this file? This action is irreversible.

News & Updates

  • Robust Design for Multivariate Quality Characteristics Using Extreme Value Distribution

    Authors: Changming Yang, Xiaoping Du
    J. Mech. Des. 136(10), 101405 (2014) (8 pages) Paper No: MD-13-1534; 
    doi: 10.1115/1.4028016 

    Robust design makes product performance stable under variations and noises in the environment. So the product can work robustly even in harsh conditions. This work explores a way to measure the robustness of a product when it has multiple performance variables, such as strength, efficiency, and cost. These performance variables are dependent and oftentimes conflicting, meaning that improving one performance variable may make others worse. The robustness of the worst-case performance variable is used as an indicator of the robustness of the entire product. Analytical and numerical algorithms are developed to calculate the robustness. The work makes it easy to model the robust design optimization with multiple performance variables as the single-objective optimization, thereby increasing the effectiveness of the robustness design process.
  • Featured Article: Bio-Inspired Design: An Overview Investigation Open Questions From the Broader Field of Design-By-Analogy

    Authors: Katherine Fu, DIana Moreno, Maria Yang, and Kristen L. Wood

    J. Mech. Des. 136(11), 111102 (2014) (18 pages)
    doi: 10.1115/1.4028289

    Bio-inspired design is a cutting edge field of inquiry and practice, founded by thinkers such as Steele (bionics, 1950s), Schmitt (biomimetics, 1950s), and French (biologically inspired design, 1988).  Many successful products have resulted from this approach or way of designing, dating back to the 19th century, including barbed wire, Tiffany lamps, the Wright glider, the design of Central Park in Manhattan, and many more.  Based on these and other bio-inspired designs, foundational questions arise, such as: how can we go about finding these elegant analogies without being a biology expert or without counting on isolated experiences or chance? To answer this question, researchers have worked to understand the cognitive mechanisms that underlie bio-inspired design, as well as developed tools and methods to support it.  In this paper, we examine seminal methods for supporting bio-inspired design (including the work of Benyus/Deldin et al., Chakrabarti et al., Shu and Cheong et al., Nagel et al., Vattam et al., and Vincent et al.)and review the existing literature on bio-inspired design cognition, highlighting the areas well aligned with current findings in design-by-analogy cognition work and noting important areas for future research identified by the investigators responsible for these seminal tools and methods.  Supplemental to the visions of these experts in bio-inspired design, we suggest additional projections for the future of the field, posing intriguing research questions to further unify the bio-inspired design field with its broader resident field of design-by-analogy.

  • Featured Article: Machine Learning Algorithms For Recommending Design Methods

    Authors: Mark Fuge, Bud Peters and Alice Agogino

    Designers use specific methods to discern people’s needs and how to best create products or services that meet those needs. Choosing precisely the right method for a given problem is extremely difficult: it requires a deep understanding of the nature of the problem, knowledge of the vast array of design methods, and years of experience. This paper demonstrates that by collecting expert experience in the form of case studies, machine learning algorithms can help new designers pick better design methods and understand how methods are related to one another. Specifically, we show that looking at which methods designers use together can be more informative than just looking at the content of the method itself. In addition, you can use counts of which methods are used together to automatically cluster methods into groups that agree with human ratings; this means that you can study many more methods than could be done manually.