11. Some results of strength calculations relying on analytical and FEM approaches. Trends of using contemporary machine learning strategies

Annotation: This article analyzes the results of studies, which are based on numerical methods of analysis, of the stress-strain state of thin-walled shell structures. This article also discusses analytical solutions that apply asymptotic approaches and a method of initial parameters in a matrix form for solving a problem of equal stability of reinforced compartments of combined shell systems of the rocket and space technology within the scope of the research being carried out jointly by teams of Yuzhnoye State Design Office and Zaporizhzhya National University. The primary attention is paid to the use of FEM-based direct numerical methods and the research results for which analytical methods can be useful for making a preliminary assessment of the bearing capacity of load-bearing structures, and in some cases for their rational design. This article does not contrast numerical and analytical approaches but about the possibility of using them effectively. The article talks about possible ways of using the up-to-date technique of machine learning (Machine Learning Technology) in the calculation and experimental methods for determining the characteristics of the rocket and space technology.