Quantitative Method Of Temperature Effect In Ultrasonic Testing Of Concrete Based On Vector Residual SVD

Due to its rich information and sensitivity to micro changes in media,ultrasonic coda waves have shown potential engineering application value in measuring and evaluating stress in concrete structures,and have received continuous attention from researchers at home and abroad in recent years.With the deepening of research,the impact of ambient temperature and humidity on coda signals has become a key issue that cannot be ignored.Currently,research has focused on solving this problem,the most typical of which include the thermal compensation method and the optimal baseline selection algorithm.Under complete test conditions or data sets,the two methods are still effective.Therefore,there is an urgent need for a mathematical or physical model to solve the problem theoretically.Establish a temperature effect quantification and removal method using singular value decomposition(SVD)of signal vector residual matrix based on the mathematical variation characteristics of coda signals,which provides theoretical support for subsequent research on ultrasonic stress testing techniques.The topic is derived from the National Natural Science Foundation project "Research on non-loading and non-unloading measurement of stress field in concrete structures based on Coda wave interferometry"(51478072).The main research work of this thesis is as follows:(1)Summarize and analyze the research progress on ultrasonic coda testing of concrete structure stress at home and abroad,focusing on investigating relevant literature on the influence of environmental temperature and humidity on coda signals in ultrasonic testing,summarizing and summarizing important research progress and achievements in this field,analyzing existing problems,and proposing the research content of this article.Based on the theory of SVD high dimensional spatial rotation transformation,a vector projection analysis method for temperature effect of coda signals was proposed by investigating algorithms in applied acoustics and digital signal processing.(2)An ultrasonic test of the dead weight state of two concrete beams was designed and carried out,using the normalized angle between signal vectors as an evaluation index of signal fluctuations to reflect the environmental impact of the coda signal.By comparing and analyzing the correlation between the two beam coda signal fluctuation indicators and their relationship with ambient temperature and humidity,it is determined that ambient temperature and humidity are the homologous factors of coda signal fluctuation.The data processing methods of "the same temperature" averaging and continuous moving average are proposed,and the variation rules of coda waveform and fluctuation index caused by temperature and humidity are analyzed and compared.(3)Focusing on the influence of ambient temperature on coda signals,a temperature effect quantization and removal method based on vector residual SVD is proposed.The total characteristics of temperature induced signal changes are obtained through the vector residual matrix,and the directional vector(temperature effect vector)of the temperature impact in the high-dimensional space is obtained through the residual matrix SVD.The projection relationship between the signal vector and the temperature effect vector is used to achieve effect quantification and removal.The relationship between effect quantization and removal is analyzed,and two practical methods for temperature effect quantization and removal,namely,effect vector substitution method and subsection iterative removal method,are proposed.By analyzing the ability of temperature effect vectors to remove environmental impacts from different samples,it is found that temperature effects dominate the overall environmental impact.Finally,methods for removing the overall environmental impact are further presented.