North America

Green Room - Session 8

1:00 to 2:00 p.m. Thursday, June 15, 2017

Estimating a Distributed Reliability over a Range of Use Conditions with the Help of an Accelerated Stress Design of Experiment Approach

Many products operate over a wide range of use conditions. Often, reliability estimates for these products are based on some average or standardized "use-case" condition. However, this can result in reliability estimates that may be overly optimistic of what actually occurs in the field. The way around this is to gain knowledge about the reliability over a range of use stresses. This presentation gives an example of the use of accelerated stress testing using a design of experiments (DOE) approach to build a model to estimate a "distributed" reliability over a range of expected product use conditions. This real life example is based on the testing of blowers used in CPAP Sleep Apnea devices. An estimate of the distributed reliability in the situation of multiple use cases is made with the help of Monte Carlo simulations. This analyses is compared with a more traditional approach where only the nominal and extreme cases are considered. The results provide a more accurate estimate of the reliability performance that may be expected from products in the field over a range of actual use conditions.

Key Words: Design of Experiments, Accelerated Stress Tests, Monte Carlo, Distributed Reliability, Blowers, Medical Devices

Ray Gibson

Philips Respironics

Monroeville, Pennsylvania