Ultrasonic flaw detection can only qualitatively analyze casting defects, but can not accurately describe defects. There are many scanning blind areas for cast steel joints with complex shapes, and it is difficult to 100% ultrasonic detection for large-scale cast steel joints. Therefore, on the basis of obtaining the overall distribution law of casting defects in cast steel joints, a four branch cast steel joint whose shape, form and defect distribution are typical cases in engineering is selected as the test piece for slice test, so as to verify the effect and accuracy of ultrasonic testing, and obtain more accurate defect types, distribution, size and other information in the joint, so as to provide a reference basis for more accurate understanding and Research on the casting defects of cast steel joints, It provides a basis for reliability analysis and residual life prediction of cast steel joints.
The slice test is divided into two steps. The first step is to roughly understand the concentrated area of casting defects through nondestructive testing, and the second step is to arrange the section according to the results of nondestructive testing.
(1) The ultrasonic waveform image of internal defects of a typical four branch cast steel joint is obtained by ultrasonic nondestructive testing technology. According to the waveform image, the defect types are judged, and the main defect types and distribution of the cast steel joint are found.
(2) The cast steel joints were cut by sawing machine, water knife and other cutting equipment to further clarify the nature of casting defects. The results of slice test are compared with the previous NDT results, which verifies the accuracy of ultrasonic flaw detection and provides suggestions for the NDT of cast steel joints.
(3) The distribution law of random variables of casting defect size is studied by using the methods of probability theory and mathematical statistics. The parameter estimation and fitting optimization of casting defect size are carried out by using exponential distribution function, which provides a theoretical basis and data support for the reliability research of cast steel joints and a certain basis for the establishment of defect random field of cast steel joints.