Multiscale modeling of the mechanical behavior of brazed Ni-based superalloy sheet metals

Abstract

The elastic–plastic behavior of brazed Ni-based superalloys used in abradable turbomachinery sealing systems is analyzed by means of numerical simulations. A sequential multiscale modeling approach is employed to analyze the relevant effects on the mechanical behavior of a layered composite consisting of the braze metal and the joining partners. The focus of the investigations lies on the role of the multiphase microstructure within the brazing layer and the significance of microscopic parameters such as, for example, the volume fraction of the phases compared to macroscopic parameters (e.g., the brazing layer thickness). A representative volume element is employed on the microscale, and a layered composite is modeled on the macroscale to capture all relevant effects on both length scales during mechanical loading of the layered material composite. Virtual tensile tests at different temperatures and strain rates are chosen as a controlled testing environment that captures the characteristic loading conditions during a rubbing event in a turbomachinery. The parameter on the microscale with the highest influence on the mechanical behavior is the volume ratio of the brittle and ductile phases. It also has a significant effect on the macroscopic mechanical behavior. To reduce the risk of damage, it is advised to minimize the fraction of brittle phases in the brazing layer. According to the simulation results, this can be even more effective than reducing the overall thickness of the brazing layer. Additionally, an improvement of an existing analytical model for the estimation of the flow stress in the ductile phase of a dual-phase microstructure is proposed. By increasing the order of the root function in the analytic model, the effect of stress distribution between brittle and ductile phases can be incorporated in an empirical manner. This reduces the deviation between the analytical and the numerical approaches significantly.