Explicit correlation between surface integrity and fatigue limit of surface cold worked chromium-nickel austenitic stainless steels.

Explicit correlations were established between surface integrity (SI) characteristics (roughness, microhardness, residual stresses) obtained via surface cold working (SCW) and bending fatigue limit of chromium-nickel austenitic stainless steels (CNASS). The SCW was implemented via hardening DB, and the treated CNASS was AISI 304. Based on the correlations obtained, a simplified optimization approach to achieving the maximum fatigue limit without carrying out fatigue tests was developed. Five roughness parameters that are functionally significant indicators of fatigue behavior, microhardness, and surface residual axial and hoop stresses form the vector of objective functions. The governing factors are the burnishing force and the number of passes. The fatigue limit was obtained by the accelerated Locati's method. The SI models and the fatigue limit were obtained by a planned experiment and regression analyses. The correlations between SI characteristics and the fatigue limit were obtained by eliminating the variables in the corresponding pairs of models. The change trends of the microhardness and fatigue limit were found to be identical, i.e., the fatigue limit of CNASS can be controlled via inexpensively measured SI characteristics such as roughness parameters and surface microhardness. In other words, the essence of the proposed approach is that any static cold working process can be optimized without carrying out fatigue tests by choosing these SI characteristics as objective functions and seeking a compromise optimal solution that simultaneously achieves high values of the microhardness and skewness and a low value of the roughness parameter Ra. Given the explicit correlation between the microhardness and the fatigue limit, the compromise optimal solution maximizes the fatigue limit while satisfying the particular roughness requirements. The multi-objective optimization performed and comparisons made prove the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]