Phase-field study on microstructure formation in Mar-M247 during electron beam welding and correlation to hot cracking susceptibility

Abstract

Abstract Ni-base superalloys are frequently used for cast components in the aero-engine and power generation industries. For joining and repair of these components, beam welding is often the method of choice in industrial praxis. However, precipitation-strengthened nickel alloys generally present poor weldability as a consequence of their high weld cracking susceptibility, with high segregating alloys like Mar-M247 even being considered unweldable. Therefore, strong efforts are taken on optimizing techniques and parameters to reduce crack formation during welding of these alloys. Optimization of welding parameters can be assisted by virtual modelling methods through different scales. To be able to focus onto the factors which eventually are responsible for crack formation during welding, comprehensive modelling of the whole process chain is required, starting from a realistic model of the base material and a simulation of the heat source on the macro-scale, and including melting and microstructure formation during welding on the micro-scale. Then, based on the thermal history and the exact microstructure, cracking susceptibilities during solidification can be deduced by hot cracking models adapted to the specific conditions. In this paper, results of microstructure simulations are presented for the technical superalloy MAR-M247 using the phase-field software MICRESS with coupling to Calphad databases. Based on prior phase-field simulations of equiaxed and columnar microstructures of the base material as well as results of a macroscopic simulation of the heat source, melting and subsequent solidification of MAR-M247 has been simulated for two different welding parameter sets. As-weld microstructures are compared to experimental welds, and the virtual hot cracking susceptibility, obtained from the simulation results using a modified Rappaz–Drezet–Gremaud (RDG) hot cracking criterion, is discussed against experimental crack observations.