Mathematical Modeling and Optimization of Parameters of the Mode for Tungsten-Inert Gas Remelting with Nanomodification of the Surface Layer

摘要

The object of the present paper is to model the process of pulse current Tungsten-Inert Gas (TIG) remelting of surface layers of details of structural steel S235JR in order to be nano-modified. Neural models are trained for determination of the parameters of the welding mode, in which the weld with desired geometric characteristics is obtained, as well for determination of these parameters of melting at pre-selected geometry of the weld. A regression model with real data for pulse current TIG welding overlay is presented. This model is solved and the optimal technological modes are found at pre-selected parameters of the weld and the minimal ratio I/v.