Title: Statistical estimation of random failure probability based on stress levels

Authors: Hyun Su Sim; Jin Hyeong Jung; Yong Soo Kim

Addresses: Department of Industrial and Management Engineering, Kyonggi University Graduate School, 154-42, Gwanggyosan-Ro, Yeongtong-Gu, Suwon-Si, Gyeonggi-Do, South Korea ' Department of Industrial and Management Engineering, Kyonggi University Graduate School, 154-42, Gwanggyosan-Ro, Yeongtong-Gu, Suwon-Si, Gyeonggi-Do, South Korea ' Kyonggi University, 154-42, Gwanggyosan-Ro, Yeongtong-Gu, Suwon-Si, Gyeonggi-Do, South Korea

Abstract: We develop a new method for estimation of random failure probability; this will inform test design and statistical analysis. First, the random failure mechanism is characterised so that a suitable test can be designed. Second, the experimental results are statistically analysed. This enables the development of a Gompertz model of the failure characteristics, which are then extrapolated, and the failure probability is estimated. The probability of random brake disc failure in the real world was quantitatively estimated using a case study: a reliability test of random failure caused by hot judder designed by a user based on our model.

Keywords: reliability engineering; brake disc; failure mechanism; hot judder; brake torque variation; BTV; random failure; test design; statistical analysis; degradation model; log-normal distribution; failure probability.

DOI: 10.1504/IJQET.2022.123501

International Journal of Quality Engineering and Technology, 2022 Vol.8 No.4, pp.366 - 378

Received: 31 Mar 2021
Accepted: 22 Oct 2021
Published online: 23 Jun 2022 *