摘要
针对目前声波透射法检测桩身混凝土质量缺陷所采用分析方法的不足,提出了一种采用多参量信息反映桩身混凝土质量缺陷的新方法———多参量综合分析法。该方法假定混凝土的缺陷信息量是与缺陷有关的表征量的线性组合,将突出混凝土缺陷的问题,归结为计算表征量的相关矩阵的最大特征值和相应的特征向量;通过趋势分析,除去缺陷信息量背景值,得到缺陷信息量的异常值。结合工程实例,证实本方法的可行性,并能取得很好的效果。研究表明,采用不同参量间的权系数计算混凝土缺陷信息量,得到的异常曲线比单因素分析法和多因数概率分析法能更合理地反映桩身混凝土质量情况;避免单因素分析法引起的误判、漏判;缺陷信息量异常值越大,混凝土的缺陷越严重。
In view of the disadvantages of the present analysis methods of cross-hole sonic logging to detect extent and location of pile defects,A new method called comprehensive analysis of multi-parameter was put forward to assess the concrete quality using several parameters associated with defects.Assume that the defect information is linear combination of the parameters,emphasizing defects boils down to calculating the largest eigenvalue and corresponding eigenvectors of the correlation matrix of the parameters.After the trend analysis,the background values were removed from the defect information to get abnormal values of the defect information.Based on the practical examples,the feasibility of this method was proved and very good results can be achieved.As a result,the abnormal curves of the information of concrete defects,which calculated with different weights of the parameters,can be more reasonable to assess the concrete quality than single factor analysis and numerous factor probability method(NFP),and avoid misjudgment caused by single factor analysis.In addition,the greater the abnormal value,the more serious the concrete defect is.
出处
《施工技术》
CAS
北大核心
2012年第4期71-75,80,共6页
Construction Technology
基金
国家自然科学基金项目(60672042)
湖南省高校科技创新团队项目(2008-244)
关键词
桩基础
检测
声波透射法
特征向量
趋势分析
piles
detection
cross-hole sonic logging
eigenvector
trend analysis