Novel Statistical Techniques for Conducting Accelerated Life Test to Demonstrate Product Reliability

Defence science journal

article

Mustaq Basha¹  Nliesh R Ware² 

[1] DRDO-Electronics and Radar Development Establishment;DIAT, Deemed to be University

关键词: Confidence level;    Significance level;    Producer risk;    Consumer risk;    Accelerated life testing;    Reliability demonstration testing;    Sample size;    Mean time to failure;   

DOI  :  10.14429/dsj.72.17838

学科分类：社会科学、人文和艺术（综合）

来源: Defence Scientific Information & Documentation Centre

【 摘 要 】

In Reliability Demonstration Testing (RDT), finding the right sample size is very important since the cost of the prototypes is high and difficult to make. If the sample size for the RDT is test is less, the amount of information obtained from the test will be insufficient, and the conclusion will be meaningless; on contrary, if the sample size is big/huge, the amount of information obtained from the test will be in excess of what is required, resulting in unnecessary costs. Most of the time, the required sample size and test time are decided based on the RDT test design. Resources required for RDT in terms of batch size and long testing-time is practically not feasible, due to limitation of the project schedule and budget. The reliability engineers must have a sound knowledge of type challenge/risk that is allowed for conducting RDT. The research paper with a case study provides the required information about the modern techniques adopted in reducing the sample-size and testing time with the help of accelerated test models such as Arrhenius, Erying etc., for conducting accelerated life test to demonstrate the product reliability.

【 授权许可】

All Rights reserved