当前位置: X-MOL 学术J. Mater. Eng. Perform. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
The Efficiency of Normal Distribution in Statistical Characterization of the Experimentally Measured Strength for Ceramics
Journal of Materials Engineering and Performance ( IF 2.2 ) Pub Date : 2020-11-25 , DOI: 10.1007/s11665-020-05352-1
Jianghong Gong , Bin Deng , Danyu Jiang

It has been assumed that the measured strength of ceramics follows a Weibull distribution. However, there is still no sound evidence to confirm this assumption. On the contrary, some studies have shown that other distributions such as normal distribution may describe more appropriately the measured strength data than Weibull distribution. In this paper, an extensive comparison between the efficiencies of normal and Weibull distributions in describing the strength variations was performed based on the analyses of 27 strength datasets, each containing 30 data measured on different ceramics. It was shown based on Anderson–Darling (A–D) test that, in most cases, normal distribution may give a satisfactory description for the data. The analysis results reveal that, at least for the small datasets generally used in laboratory evaluation, it seems to be unnecessary to perform a Weibull analysis because a simple normal distribution analysis is accurate enough for the statistical characterization of strength for the examined materials.



中文翻译:

陶瓷实验测量强度的统计表征中的正态分布效率

假定陶瓷的测量强度遵循韦布尔分布。但是,仍然没有确凿的证据可以证实这一假设。相反,一些研究表明,其他分布(例如正态分布)可能比威布尔分布更恰当地描述了测得的强度数据。在本文中,基于对27个强度数据集的分析,对描述强度变化的正态分布和Weibull分布的效率进行了广泛的比较,每个数据集包含在不同陶瓷上测得的30个数据。根据安德森–达林(A–D)检验表明,在大多数情况下,正态分布可能会对数据给出令人满意的描述。分析结果表明,至少对于实验室评估中通常使用的小型数据集,

更新日期:2020-11-25
down
wechat
bug