Reliability Engineering and Services

5Accelerated Stress Testing and Economics

5.1 Introduction

In accelerated stress testing (AST), the product undergoes higher‐than‐normal operating conditions in an effort to extrapolate the product reliability in use conditions by precipitating faults in a compressed period. Typical accelerating stresses include environmental, electrical, mechanical, and chemical factors. The choice of accelerating stress levels and the allocation of available test units to these levels are the key considerations in executing accelerated test experiments. This chapter begins with the introduction of the concept of AST that include a highly accelerated life test (HALT) and highly accelerated stress screening (HASS). The reliability extrapolation is often justified on the basis of physical and chimerical laws, or a combination with statistical models derived from the lifetime data. The chapter describes one‐stress Arrhenius law and multistress Eyring models, both of which are physics‐experimental‐based models. We also present three types of statistics‐based AST models: a scale and use rate acceleration model, a non‐parametric model, and semi‐parametric models that include a well‐known proportional hazard model (PHM). Both HALT and HASS are costly to implement due to the consumption of equipment, materials, and labor. Economic models are developed to guide the product manufacturer to realize the cost savings in the deployment of HASS. This chapter concludes by presenting a case study of implementing ...