Improving Accelerated Reliability Testing by Using Optimized Signals
Abdelkhalak El Hami, David Delaux, Henri Grzeskowiak
Abstract
Abdelkhalak El Hami, David Delaux, Henri Grzeskowiak
Abstract
Accelerated life tests are often used where excitation levels are increased to reduce test duration, but which require large testing facilities. This chapter presents the use of optimized signals as an alternative to conventional random signals. It presents the optimized signals with low peak values for which their root mean square is increased without altering their power spectral density. The chapter introduces a method for controlling kurtosis and highlights the duality between kurtosis and crest factor (CF). It then discusses the use of optimized signals for real environmental tests with respect to applicable standards. The chapter conducts a comparative damage assessment by calculating the fatigue damage spectrum between conventional Gaussian signals, optimized signals with low peak values and signals with high kurtosis. Reducing the CF of a signal allows the signal to be amplified further without exceeding the limit value of a testing facility and allowing for more excitation to be introduced into a structure.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Accelerated life tests are often used where excitation levels are increased to reduce test duration, but which require large testing facilities. This chapter presents the use of optimized signals as an alternative to conventional random signals. It presents the optimized signals with low peak values for which their root mean square is increased without altering their power spectral density. The chapter introduces a method for controlling kurtosis and highlights the duality between kurtosis and crest factor (CF). It then discusses the use of optimized signals for real environmental tests with respect to applicable standards. The chapter conducts a comparative damage assessment by calculating the fatigue damage spectrum between conventional Gaussian signals, optimized signals with low peak values and signals with high kurtosis. Reducing the CF of a signal allows the signal to be amplified further without exceeding the limit value of a testing facility and allowing for more excitation to be introduced into a structure.
Key concepts: Kurtosis, Crest factor, Spectral density, Reliability (semiconductor), Gaussian, SIGNAL (programming language), Computer science, Root mean square