A Case Study of Reliability Testing Mixed with Experimental Design

Josh VerHage


New product development creates unique challenges to design teams in developing functional, robust products while also achieving reliability goals. There are often time and resource constraints within a program requiring a balance between iterative product development and final design reliability demonstration. One approach is to combine design of experiment (DOE) methodology with test sample quantities sufficient to demonstrate acceptable reliability. Selection of design and application factors believed to influence component life in a traditional DOE format, while testing sufficient sample size to demonstrate reliability, allows embedding of a designed experiment into a traditional reliability test. This presentation will demonstrate the proposed approach using a case study of a rubber isolator used in an unconventional floor cleaning machine application. Combining experimental design with increased test sample quantities resulted in an optimized design, demonstrated reliability, and reduced overall program resources and timeline.