The effectiveness of damage evaluation methods was surveyed among the master curve method, the parametrical statistics method and machine learning method by applying those methods to damage evaluation using KAM(Kernel Average Misorientation) parameters obtained from EBSD(Electron BackScateer Diffraction pattern) observations for the interrupted creep and creep fatigue tests of SUS304HTB equivalent heat resistant steel for boiler tube use. As for the parametric statistics method, the log-normal distribution was judged as the best fit distribution type among normal, lognormal and Weibul distributions. Being the algorithm of machine learning effective for pattern recognition, neural network was adopted for the analysis. As a result, it was found that the accuracy was higher in the order of neural network method, parametic statistics method, and master curve method. The reason why the neural network method was more accurate than the parametric statistics method was the latter method could not approximate the frequency distribution shape of KAM accurately. If the frequency distribution profile is unknown as in this case, the method like neural network independent on the distribution profile is considered to be very effective.
An artificial intelligence evaluation system using neural network was developed for upgrading the creep damage assessment methodology through image analysis of EBSD(Electron BackScateer Diffraction pattern) maps. KAM(Kernel Average Misorientation) maps were obtained for creep damaged austenitic stainless steel SUS 304HTB and the stratified data were manipulated as the representatives of damage degrees. The system consists of an input layer, intermediate layers and an output layer. As the activation function, ReLU(Rectified Linear Unit) function is used for the intermediate layers and Softmax function is used for the output layer. The evaluation results of the proposed system were compared with the results of the conventional quantitative damage evaluation method. As a result, the estimated damage accuracy of the artificial intelligence evaluation system developed in this research was proved to be improved by about 3.3% compared with the estimated damage accuracy using the conventional evaluation method. Furthermore, the accuracy was improved by about 6.7% after the optimization of the neural network compared with the conventional evaluation method. Moreover, it was proved this system had sufficient robustness through the check tests for the case of extremely missing EBSD image. Thus machine learning utilizing neural network was expected to be a potential method for versatile data analysis applicable to various sort of metallographic study.
Bivariate log-normal distribution analyses coupled with the cumulative hazard function method were conducted on the specific output class of 8 steam turbine units with components such as rotors, moving blades, nozzles, casings and other auxiliary equipment of high-, intermediate-and low-pressure turbines. The damage phenomena were classified into erosion, crack, deformation, corrosion, creep void formation and material degradation with corresponding components. Operation time and start up cycles for damage incidence in respective units were collected and statistically analyzed adding the non-failed data as well as failed data. After applying the bivariate log-normal distribution regression to those data sets, the prescribed failure probability was imposed to construct the equal probability ellipse contours as the quadratic function of operation time and startup cycles. To determine whether the events were time dependent or cycle dependent, the shape and inclination of the contours were utilized. The order of event incidence was determined by using the lower end values of the major axis of equal probability contours. Although the order of event incidence could show variations according to the prescribed failure probability values, the examples for 90% probability ellipse contours were demonstrated here. The assessment results showed that the statistical analyses were effective for investigating the damage incidental scenario making and maintenance planning for actual plants.
EBSD observations were conducted on the damaged materials obtained by interrupted creep tests and interrupted creep-fatigue tests for 304 austenitic stainless steel for boiler tube use in fossil power plants, and the shapes of crystal grains extracted from KAM maps and GOS maps were approximated by ellipses. Furthermore, a damage evaluation system has been developed with a neural network, which uses the information obtained by elliptic approximation as parameters. As a result, it was quantitatively found that as creep and creep-fatigue damage progress, crystal grains become elongated toward the load axis direction. Ensemble learning showed the best classification accuracy using the 20 learners obtained by changing the rank of the relative frequency of KAM. The damage evaluation system in this study was able to estimate the damage rates with a classification accuracy of 98.33% for creep test materials and 97.50% for creep-fatigue test materials using information from one of crystal grains in the EBSD image. Therefore, the system with the neural network developed in this study is effective for evaluating creep and creep-fatigue damage for 304 austenitic stainless steel.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.