形状参数
威布尔分布
可靠性(半导体)
加速寿命试验
离群值
可靠性工程
失效模式及影响分析
压力(语言学)
领域(数学)
数学
计算机科学
统计
结构工程
工程类
功率(物理)
物理
哲学
量子力学
语言学
纯数学
作者
Sarath Jayatilleka,James A. McLinn
出处
期刊:Proceedings, annual Reliability and Maintainability Symposium
[Institute of Electrical and Electronics Engineers]
日期:2021-05-24
卷期号:: 1-6
被引量:6
标识
DOI:10.1109/rams48097.2021.9605732
摘要
Among many shape parameters of various distributions such as Gamma, normal and log normal, there is more accumulated knowledge around the Weibull shape parameter over the past decades. When all failures modes are pooled, for example, in the first attempt of analyzing field reliability data, the resultant shape parameter would be different from the shape parameters resulted from specific failure mode focused field data analysis (and accelerated life tests). In addition, accelerated life tests run at stressed levels on test fixtures provide shape parameters to uncover failures modes and verify effective corrective action across the product design life cycle. In such applications many shape parameters and their shifts provide important information to strategize the reliability test administration before and during the product development phases.Weibull shape parameter can be a characteristic of a specific failure mode during reliability analysis. However, shape parameters during the three phases of the bathtub model can lead to confusion. For example, late life failures with shape parameter values less than unity and early life shape parameters with larger than unity may exist. These issues will be discussed with examples and reasons behind the outliers.When accelerated life tests (ALT) run, several different levels may be chosen for acceleration. If the stress levels rise to the highly accelerated life test (HALT) levels, one needs to be careful. One successful method is provided by monitoring shape parameter change. In addition, optimum values of the stress levels acceleration can be determined using the deviations of slopes of the shape parameters. Optimum stress levels can be used to quickly qualify different design options, suppliers, and material options often encountered during new product development. Examples are provided from appliance industries.In summary, this paper presents how to correctly interpret the practically encountered Weibull shape parameters and help planning next design iteration across the New Product Development (NPD) cycle. It also helps planning ALT and HALT for choosing between options, designs and suppliers involved in NPD verification cycle and to finally demonstrate the reliability for design validation per customer’s requirements.
科研通智能强力驱动
Strongly Powered by AbleSci AI