Reconstruction of Regression Relationships for Determination of Strength Values for Alloys of Products in Diffusion Welding

Abstract

The article deals with the problem of obtaining the dependence of the product strength parameter on the welding time, welding temperature and pressure during mechanical tests, leak tests. The relevance of this work is due to the complexity of carrying out field experiments to identify dependencies. In particular, the complexity arises from the duration of diffusion welding and the high cost. Application of the method of regression analysis based on a non-compositional plan of the second order for three factors will allow to restore the dependence of the product strength parameter on the time during which welding was carried out, the temperature at which diffusion welding was carried out or could be carried out and on the applied pressure at which mechanical tests were carried out. In the current study, a non-compositional design of the second order for three factors was used - allowing to restore the dependence of the missing values of the strength of the product. The aim of the research is to improve the quality of mathematical modeling. Application of the proposed approach will make it possible to obtain the strength distribution function depending on time, temperature and pressure using the example of a product made of VT14 titanium alloy and 12X18H10T stainless steel. This will make it possible to obtain optimal parameters for the diffusion welding mode and to improve the quality of the resulting products.