工程结构试验中小体量测试数据误差来源与影响分析  

Analysis on the error source of small volume test data in engineering structure test and its influence

作　　者：黄勇[1] 赵晓倩[1] 刘辉[1] HUANG Yong;ZHAO Xiaoqian;LIU Hui(Science and Technology Division,Technology Department,Jiangsu Vocational Institute of Architectural Technology,Xuzhou,Jiangsu 221116,China)

机构地区：[1]江苏建筑职业技术学院科技处,江苏徐州221116

出　　处：《江苏建筑职业技术学院学报》2021年第4期25-29,共5页Journal Of Jiangsu Vocational Institute of Architectural Technology

基　　金：江苏省住建厅科技计划项目(2018ZD109);徐州市科技计划项目(KC18136)。

摘　　要：为进一步提高工程结构试验结论可靠性,简易有效分析小体量试验数据主要误差来源,量化评价误差影响,以项目组近十年进行的Q235钢材拉伸试验数据为基础,基于数理统计理论,提出了样本偏差结合示值误差分析法.经过可疑数据筛选与奇异数据剔除后,识别了数据中主要误差来源,评价了各类误差影响.研究表明,小体量试验数据中,误差主要来源于试件取样不均匀、人员操作水平不稳定与材料质量随年份变化;试件取样不均匀及试验人员操作水平不稳定引起的随机误差影响大于系统误差,不可忽视;多次间接测试取得的数据存在误差传递与累积,误差影响突出,数据离散性大;试验人员操作水平发挥稳定性是导致数据存在大偏差的首要原因,可通过提高试验人员操作水和增加平行试样等方法降低操作误差的影响.In order to further improve the reliability of engineering structure test conclusions,simply and effectively analyze the main error sources of small volume test data and quantitatively evaluate the impact of errors,a sample deviation combined with indication error analysis method was proposed,based on the tensile test data of Q235 steel conducted by the project team in recent ten years and based on the theory of mathematical statistics.After suspicious data screening and singular data elimination,the main error sources in the data are identified and the impact of various errors was evaluated.The results show that the errors in the small volume test data are mainly caused by the uneven sampling of specimens,the unstable operation level of personnel and the change of material quality with years;the random error caused by uneven sampling of test pieces and unstable operation level of test personnel is greater than that caused by the systematic error,so it can not be ignored;the test data obtained indirectly for many times has error transmission and accumulation,which l ead to prominent error influence and great data dispersion;the stability of the tester’s operation level is the primary reason for the large deviation of the data.The impact of operation error could be enhanced act structuraltest erro by improving the tester’s operation level and increasing parallel.

关 键 词：结构试验 误差来源 误差影响 量化评价 

分 类 号：TB9[一般工业技术—计量学] TU312[机械工程—测试计量技术及仪器]