Function-Failure Design Method Research Articles

Linking product functionality to historic failures to improve failure analysis in design

Prior research has shown that similar failure modes occur within products (or components) with similar functionality. To capitalize on this finding, a knowledge base-driven failure analysis tool, the function-failure design method (FFDM), has been developed to allow designers to perform failure mode and effect analysis (FMEA) during conceptual design. The FFDM can offer substantial improvements to the design process since it enhances failure analysis thus giving it the ability to reduce the number of necessary redesigns. The FFDM, however, is only as good as the knowledge base that it draws from, and one fundamental question that arises in using the FFDM is: at what level of detail should functional descriptions of components be encoded? This paper explores two approaches to populating a knowledge base with actual failure occurrence information from Bell 206 helicopters. Results indicate that encoding failure data using more detailed functional models allows for a more robust knowledge base. Interestingly however, when applying FFDM, high level descriptions continue to produce useful results when using the knowledge base generated from the detailed functional models.

The Function-Failure Design Method

To succeed in the product development market today, firms must quickly and accurately satisfy customer needs while designing products that adequately accomplish their desired functions with a minimum number of failures. When failure analysis and prevention are coupled with a product’s design from its conception, potentially shorter design times and fewer redesigns are necessary to arrive at a final product design. In this article, we explore the utility of a novel design methodology that allows failure modes and effects analysis (FMEA)-style failure analysis to be conducted during conceptual design. The function-failure design method (FFDM) guides designers towards improved designs by predicting likely failure modes based on intended product functionality.