Probability Evaluation On Offshore Platform Crack Detection Ability Based On ACFM Technique

With the increasing of offshore oil exploitation, the role of the offshore platform becomemore important. In order to ensure the safety and reliability of offshore operation, the analysisof offshore platform is crucial. If there is a crack in the key parts of platform, it may affect thesafety of platform seriously, so nondestructive testing of offshore platform is one of theimportant means of the security evaluation for platform. ACFM is a new developed type ofelectromagnetic nondestructive testing method in recent years, which has many advantagesfor detecting offshore platform. But for offshore platform, it is impossible to test each keyparts of the platform, and how to evaluate the possibility of the existence of crack with lesssample is an important problem, this paper proposed Bayesian probability evaluation modeland evaluation of offshore platform with small sample data the possibility of crack, whichhas important guiding significance.At present, the domestic and foreign scholars’ research mostly focus on the methods ofACFM applied in offshore platform engineering, and the probability evaluation of crackdetection is relatively rare, this article firstly through to the test board ACFM test analysis thesuperiority of the method, and then puts forward a crack exists probability model based onBayes theory. Finally, ACFM is applied to detect offshore platform model and evaluate thepossibility of the existence of crack.Firstly, this paper introduces several crack detection methods. For offshore platform withcomplex structure, the traditional methods have many shortcomings. According to theprinciple of ACFM method of detecting, cracks on the offshore platform using the detectionof ACFM is more simple, easy to operate than the other method.Secondly, by introducing ACFM detection principle and operating software, combiningwith crack of laboratory test plate for detecting, it is obtained a POD curve of ACFM and MTmethod. By analyzing the curve, it can be obtained: delivered ACFM detection accuracy is higher, which in practice proved the feasibility of ACFM in offshore platform crack detection, and provides a technical support for the probability evaluation on the existence of the crack ofMarine platform key points in this paper/thesis.Finally, in order to solve the small sample events based on the Bayes theory, it presents the probability evaluation model on the existence of the crack. One is no a priori information of point estimates and confidence lower limit estimation, the other is to have a priori information of point estimates and confidence lower limit estimation. Combined share prior distributionoptimization is proposed to estimate the posterior distribution, and then two cases can be achi eved, based on the laboratory testing information of offshore platform model. the estimated probability of crack, have further estimate the probability of crack than with no a priori information closer to the actual estimate of the probability of crack; On this basis, by estimating the possibility of the existence of the crack of some key parts, it can apply the result to estimate thepossibility of the existence of other crack of similar structure, according to the similarity in the structure of the offshore platform. This provides a theoretical support for the practical application in the future.