{"title":"绝热压缩试验的统计考虑","authors":"Barry Newton, T. Steinberg","doi":"10.1520/STP159620150077","DOIUrl":null,"url":null,"abstract":"ASTM G74 has been used for many years to evaluate nonmetallic materials and components for oxygen service. When originally published in 1982, this standard considered a “passing” result to be zero ignitions of a material out of 20 samples tested. However, researchers have recognized that the originally prescribed methodology results in a cumulative binomial confidence of about 36 % for a passing result. As a result, the low confidence for a passing result could be potentially misleading when results are used to qualify materials or components for oxygen service, unless the data is analyzed through available statistical approaches. This paper summarizes research performed to evaluate the statistical aspects of gaseous fluid impact testing so that ignition probabilities can be considered in the test methodology. Data derived by the test method are evaluated by a logistic regression approach in order to describe the behavior of the materials being tested and to compare different materials or test conditions. Therefore, the statistical aspects of the test are shown to be crucial to understanding and applying the data obtained. This paper demonstrates that the ASTM G74 test and all international tests of a similar nature because all use the same test embodiment and are inherently probabilistic and subject to variability that seems random without application of appropriate statistical analysis. However, meaningful results can be developed when the appropriate statistical tools are utilized. Logistic regression analysis is only one available method to analyze binomial data (ignition/no-ignition); but it is a powerful tool that can help to bring clarity to the trends in data that are obscured by sometimes seemingly random behavior.","PeriodicalId":21486,"journal":{"name":"Science & Engineering Faculty","volume":"33 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2016-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Statistical Considerations for Adiabatic Compression Testing\",\"authors\":\"Barry Newton, T. Steinberg\",\"doi\":\"10.1520/STP159620150077\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"ASTM G74 has been used for many years to evaluate nonmetallic materials and components for oxygen service. When originally published in 1982, this standard considered a “passing” result to be zero ignitions of a material out of 20 samples tested. However, researchers have recognized that the originally prescribed methodology results in a cumulative binomial confidence of about 36 % for a passing result. As a result, the low confidence for a passing result could be potentially misleading when results are used to qualify materials or components for oxygen service, unless the data is analyzed through available statistical approaches. This paper summarizes research performed to evaluate the statistical aspects of gaseous fluid impact testing so that ignition probabilities can be considered in the test methodology. Data derived by the test method are evaluated by a logistic regression approach in order to describe the behavior of the materials being tested and to compare different materials or test conditions. Therefore, the statistical aspects of the test are shown to be crucial to understanding and applying the data obtained. This paper demonstrates that the ASTM G74 test and all international tests of a similar nature because all use the same test embodiment and are inherently probabilistic and subject to variability that seems random without application of appropriate statistical analysis. However, meaningful results can be developed when the appropriate statistical tools are utilized. Logistic regression analysis is only one available method to analyze binomial data (ignition/no-ignition); but it is a powerful tool that can help to bring clarity to the trends in data that are obscured by sometimes seemingly random behavior.\",\"PeriodicalId\":21486,\"journal\":{\"name\":\"Science & Engineering Faculty\",\"volume\":\"33 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2016-07-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Science & Engineering Faculty\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1520/STP159620150077\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Science & Engineering Faculty","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1520/STP159620150077","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Statistical Considerations for Adiabatic Compression Testing
ASTM G74 has been used for many years to evaluate nonmetallic materials and components for oxygen service. When originally published in 1982, this standard considered a “passing” result to be zero ignitions of a material out of 20 samples tested. However, researchers have recognized that the originally prescribed methodology results in a cumulative binomial confidence of about 36 % for a passing result. As a result, the low confidence for a passing result could be potentially misleading when results are used to qualify materials or components for oxygen service, unless the data is analyzed through available statistical approaches. This paper summarizes research performed to evaluate the statistical aspects of gaseous fluid impact testing so that ignition probabilities can be considered in the test methodology. Data derived by the test method are evaluated by a logistic regression approach in order to describe the behavior of the materials being tested and to compare different materials or test conditions. Therefore, the statistical aspects of the test are shown to be crucial to understanding and applying the data obtained. This paper demonstrates that the ASTM G74 test and all international tests of a similar nature because all use the same test embodiment and are inherently probabilistic and subject to variability that seems random without application of appropriate statistical analysis. However, meaningful results can be developed when the appropriate statistical tools are utilized. Logistic regression analysis is only one available method to analyze binomial data (ignition/no-ignition); but it is a powerful tool that can help to bring clarity to the trends in data that are obscured by sometimes seemingly random behavior.