Infrared Radiation Characteristics And Characterization Of Near Surface Delamination Defects Of Tunnel Concrete Lining

During the operation of the tunnel,the concrete lining will be affected by many factors,such as water erosion,internal reinforcement corrosion,continuous freeze-thaw cycle,in-situ stress and so on,resulting in delamination diseases.In the initial stage,it is usually hidden and cannot be observed by human eyes,but it will gradually gather and expand with the accumulation of time.Finally,accidents such as lining peeling and peeling may occur,which seriously threatens the safety of vehicle driving and the bearing performance of tunnel structure.In order to avoid the gradual deterioration of the strength and stability of the tunnel lining over time and ensure the operation safety of the tunnel,it is very important to timely detect the defects of the lining structure with potential deterioration in the daily inspection and maintenance of the tunnel.Infrared thermal wave testing technology is a non-destructive testing method to detect internal defects according to the difference of surface temperature field of the tested object.According to the source of thermal excitation,infrared thermal wave method can be divided into passive excitation thermal wave detection method(natural excitation method)and active excitation thermal wave detection method.For the detection of delamination defects on the near surface of tunnel concrete lining,the active infrared thermal wave detection technology can meet the requirements of engineering in terms of detection depth,accuracy and applicability.At the same time,it has the advantages of simple and fast detection process.It has important theoretical and practical value in the fields of tunnel maintenance and evaluation,concrete health diagnosis and so on.However,it is a very challenging task to quantitatively identify the size and depth of delamination defects in concrete.In view of this,this thesis takes the tunnel engineering as the background to study the infrared radiation characteristics and characterization of delamination defects in the near surface layer of tunnel concrete lining under active thermal excitation,so as to realize the positioning and quantitative characterization of delamination defects in the near surface layer area(0～5 cm).The specific research work and research results of this subject are as follows:(1)Analysis of key factors affecting temperature field of concrete with defectsThe key factors affecting the temperature field of defective concrete are studied by numerical simulation,and the effects of defect depth,defect area and defect thickness on the temperature field of concrete are analyzed in detail.The research shows that when analyzing the concrete temperature field with temperature,absolute contrast and relative contrast as evaluation variables,it is found that the defect depth has the greatest influence on the temperature field of the defect area on the concrete surface,the defect area has the second influence,and the defect thickness has the least influence.Due to the non-uniformity of thermal excitation,the non-uniformity of concrete surface and interior,noise and acquisition error,it is difficult to mine and express the defect thickness information in the infrared thermal wave detection data.(2)Study on the detectability of concrete internal defects under active infrared thermal wave detection technologyBased on the active infrared thermal wave test,the effects of defect area and defect depth on the detectability of defects in concrete are studied.The research shows that the defect area and defect depth will affect the detectability of concrete defects.When the aspect ratio r＞1.50,the delamination defects in concrete can be clearly detected;When the aspect ratio is 1.25 ≤r ≤1.50,the defect is between detectability and non detectability,so it is not easy to be observed;When the aspect ratio r＜1.25,the delamination defect in concrete can not be detected.Because the influence of noise and uneven thermal excitation in the image is eliminated,the defect detectability of the infrared image extracted by principal component analysis has been further improved.The delamination defects with r≥1.25 in concrete can be clearly detected,but the key factor affecting the defect detectability is the width depth ratio of the defect,that is the detectability variable r.(3)Modification of heat conduction theory of defective concrete under long pulse thermographyThe heat conduction formula of concrete with delamination defects under long pulse thermography is modified.The two physical quantities of maximum temperature difference and maximum temperature difference time are used to evaluate the rationality of the modified theory.The results show that the modified theoretical formula can accurately describe the heat conduction law in concrete with defects after thermal excitation.Based on the theoretical data of heat conduction of concrete with defects under long pulse thermography,the relationship between the maximum temperature difference time and the defect depth is analyzed by data fitting method.The results show that in long pulse thermography,the time corresponding to the maximum temperature difference on the concrete surface has a quadratic relationship with the defect depth.(4)Characterization of depth information of concrete defectsBased on the test data of active infrared thermal wave detection of concrete with defects,the partial least squares method and the principal component analysis + BP neural network method were used to establish concrete defect depth prediction models.The research shows that the concrete defect depth prediction model established by the partial least squares method has a maximum error of 0.32 cm for the training set and0.47 cm for the test set,and the overall prediction error of the model is controlled within0.5 cm.The concrete defect depth prediction model established by principal component analysis + BP neural network method,the maximum error of the training set is 0.19 cm,the maximum error of the test set is 0.14 cm,and the overall prediction error of the model is controlled within 0.2 cm.The defect depth prediction model established by principal component analysis + BP neural network method has better prediction effect than partial least square method.(5)Characterization of concrete defect area informationBased on the experimental data of active infrared thermal wave detection of defective concrete,the defect area is segmented by using the spatial information and time sequence information of infrared image.The research shows that the defect area segmentation method using infrared image spatial information will be affected by uneven thermal excitation,uneven internal and surface of concrete,noise and other factors.It is difficult to effectively distinguish normal area and defect area in image segmentation.The defect area segmentation method based on infrared image timing information has good defect area segmentation effect,without artificial subjectivity,and the error between the defect area segmentation results of different defect depths and the actual defect area is 5.1～26.9%.The defect area segmentation method based on temporal information is obviously better than the defect area segmentation method based on spatial information.