ВИСОКОПРОДУКТИВНІ МЕТОДИ АНАЛІЗУ СТАТИСТИЧНОЇ МІЦНОСТІ ЗВАРНИХ ТРУБОПРОВОДІВ ТА СУДИН ТИСКУ ЗА МЕТОДОМ МОНТЕ–КАРЛО

• Date: 2020
• Main Authors: Velikoivanenko, E.A., Milenin, A.S., Popov, A.V., Sydoruk, V.A., Khimich, A.N.
• Format: Article
• Language: English
• Published: V.M. Glushkov Institute of Cybernetics of NAS of Ukraine 2020
• Subjects: статистичне моделювання; метод Монте–Карло; високопродуктивні обчислення; напружено-деформований стан; в'язке руйнування; трубопровідні елементи
• Online Access: https://jais.net.ua/index.php/files/article/view/503
• Journal Title: Problems of Control and Informatics

The use of modern approaches to numerical analysis of the state of responsible structures in performing typical design tasks, assessing their strength, performance and residual life can take into account a wide range of initial parameters and increase the validity of relevant expert conclusions. Part of such expert analysis involves conducting numerical studies aimed at minimally conservative determination of the limiting state of the structure, which uses different analytical or numerical approaches. One of the promising directions in this area is the development of probabilistic methods for assessing the performance of complex structures, which allow taking into account the uncertainty or incompleteness of the initial data. There are a number of standard algorithms that allow you to estimate the level of reduction of the strength of the defective pipe element (РE) with detected in the process of defectoscopy corrosion-erosion loss of metal. The used limit state criteria require a certain set of initial data, such as physical and mechanical properties of the material, the actual geometry of the structure, and the resistance characteristics of a certain type of failure; in addition, it is usually assumed that the material is homogeneous and isotropic. To reduce the conservatism of the estimate of the ultimate state of PE with corrosion damage, taking into account the stochastic spatial distribution of material properties, the methods of statistical strength theory can be used, the simplest of which is the Monte Carlo method. However, the implementation of this approach requires an accurate description of the subcritical and critical destruction of PE under the action of operating load (internal pressure), as well as the appropriate high-performance software for its implementation. An integrated approach is proposed for the numerical prediction of the strength and reliability of pipelines with identified defects in corrosion loss of metal with spatially inhomogeneous properties using the Monte Carlo method to implement a statistical analysis of strength alongside a finite element description of the state of a specific structure. This work is devoted to the development and implementation of high-performance methods of statistical analysis of the residual strength of welded pipelines and pressure vessels with identified defects of local wall thinning, taking into account the spatial scatter of the initial data.