Abstract
We present an integral multiscale computational framework for the optimization of the microstructure in metal additive manufacturing. It consists of four modules: (i) the optimization solver that systematically generates feasible designs, (ii) the macroscale module for determining temperature evolution along the deposition process and further cooling, (iii) the microscale module that computes the evolution of microstructure in the deposited part for the temperature histories computed in the previous step, and (iv) the assessment module that quantifies how good the microstructure of the as-deposited part is for each design. The macroscale module uses ABAQUS for the finite element analysis of nonlinear transient heat transfer. The microscale module is a fast metamodel of the multiphase-field model for multicomponent alloys from the microstructure simulation software MICRESS. This metamodel is based on the Johnson–Mehl–Avrami–Kolmogorov law for isothermal transformations, calibrated using MICRESS’ results, and extended to non-isothermal transformations by approximating them as a series of isothermal steps. The whole workflow is implemented into ISIGHT, a user-friendly software that provides a suite of visual tools to create simulation process flows. Finally, the laser directed energy deposition of duplex stainless steels, whose mechanical properties are highly dependent on the ferrite–austenite ratio, is taken as case study. Results show that the microstructure of the as-deposited part can be significantly improved.