WSEAS Transactions on Environment and Development
Print ISSN: 1790-5079, E-ISSN: 2224-3496
Volume 15, 2019
Deep Learning in Simulation of Nickel-based Superalloys Ultimate Tensile Strength: Accounting the Role of Alloying Elements
Authors: Andrey Tyagunov, Oleg Milder, Dmitry Tarasov
Abstract: Nickel-based superalloys are unique high temperature materials with complex doping used for the gas turbine engines parts and other heat-resistant devices manufacturing, which makes them extremely important for the industry. The alloys exhibit excellent resistance to mechanical and chemical degradation under high loads and long-term isothermal exposures. One of the main service property of the superalloys is the ultimate tensile strength (UTS) or creep to rupture that is measured after the alloy sample is heated to a certain temperature and is held for a certain (prolonged) time. The common design of new alloys is time and money consuming. In particular, given the circumstances of the lack of information on the complete set of studied times and temperatures. Computational modeling would significantly simplify this process. In our work, we have applied a method of deep learning for modeling the superalloys UTS based on preliminary knowledge of their compositions, information on the role of alloying elements, solidification type, crystallographic direction of crystallization, test temperatures and exposure times, and the known UTS values. The simulation shows good agreement with the experimental data.
Search Articles
Keywords: nickel-based superalloys, artificial neural network, ultimate tensile strength, heat resistance, deep learning
Pages: 340-345
WSEAS Transactions on Environment and Development, ISSN / E-ISSN: 1790-5079 / 2224-3496, Volume 15, 2019, Art. #38