水工混凝土结构缺陷成像诊断的试验研究

采用超声波成像识别技术，对含有不同类型和性质缺陷混凝土试件进行成像诊断试验。试验中为了提高成像诊断的准确性，引入小波变换理论对测试信号进行分析处理，并就缺陷性质、尺寸、分布方位、充水、信号处理方法、计算模型以及图像滤波处理方法等对成像试验结果的影响进行了分析。试验结果表明，单一缺陷成像结果反映较敏感，缺陷边界清晰，范围准确；而含有多个缺陷的成像结果分辨率相对较低。水和缺陷尺寸对成像结果的分辨率影响较大，当缺陷很小同时又被水充填时，成像分辨率将显著降低。缺陷分布方位对成像结果影响明显，垂直于缺陷方向的边界成像要好于平行于缺陷方向的结果。在分析处理方法上，采用小波处理可以明显提高成像分辨率；而计算模型和图像滤波处理方法对提高成像诊断解释的准确性具有决定性的作用。

Experimental study on tomographic diagnosing of defects inhydraulic structure concrete

The ultrasonic tomographic technology is applied to diagnose the defects in concrete. In order to improve the precision of diagnoses, the wavelet transformation is used in the processing of ultrasonic signals. The influences of water, scale and orientation of defect, processing methods and theoretical model on image resolution are investigated. The experimental results indicate that the tomographic diagnosing result of a single defect is sensitive and the boundary can be clearly determined. But the image resolution of multiple defects is not satisfactory. The water content and the scale of defect may significantly affect the imaging resolution. Defects with the orientation perpendicular to the diagnosing direction may have higher precision in diagnosing. The wavelet transformation technology can elevate the imaging resolution and the calculation model to be applied plays very important role in improving the accuracy of detection.