ABSTRACT
Fatigue is one of the main diseases in orthotropic steel deck for long-span bridges. Fatigue cracks have been reported, and generated at stress concentration part during bridge service life inevitably. It will lead to serious structure damage if not being detected and repaired in time. Many detection methods have been developed to detect the damage of structure, while few of them could be applied well to the field damage detection, due to the complex detection environment inside orthotropic steel deck. Meanwhile, the geometric features of fatigue crack in orthotropic steel deck is complicated. In this paper, the detection methods to the real crack were proposed based on ultrasonic testing, and compared with real detection results after the specimen was cut out.
The previous research of detection method is introduced concerning different crack features (Xie et al., 2015). 1/3 measurement method are used to measure the crack length, and double traverse method is suggested to detect crack depth, and a method to judge the crack angle is proposed. However, these method are proposed on the artificial specimen with very small. The geometry feature of real fatigue crack is complex that will lower the detection precision. Therefore, the detection method need to be revised to improve the detection precision to the real fatigue crack.
A revised 1/3 measurement method is proposed regarding the flaw echo reflected from crack surface as standard (Fig. 1), and the signal is more stable and stronger for corner reflection of crack surface. Double traverse detection method is used to detect crack depth and crack angle. Several measurement points were arranged every 5 mm through the crack path. Then, using the ultrasonic devices to detect the fatigue crack geometric features, on the basis of revised detection method. Revised method for crack angle. https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/fig280_1.tif"/>
The specimen was cut out through crack propagation section, and some geometric features are measured by ruler. It is found that the revised detection method can fully recognize the real fatigue crack geometric feature precisely. A model is established on the basis of detection results. It turns out that the model describes the crack propagation details approximately, especially the the crack geometric model is similar with the real crack propagation section of out-of-plain gusset. And crack angle at welded joint. This similarity indicates that the detection method is applicability for real fatigue crack detection, and has good precision.
