A hybrid multi-physics and multi-zones simulation framework for laser-based manufacturing processes

During the modelling of laser-based manufacturing processes (welding, additive manufacturing), thermal-fluid models solving heat transfer, free surface evolution, and transport phenomena within the melt pool, are highly predictive but computationally expensive. To reduce the computational cost, an efficient framework, using Arbitrary Lagrangian-Eulerian (ALE) based finite element method, is developed to accelerate the thermal-fluid simulations in part-scale geometry. Instead of conducting thermal-fluid simulation in entire modeling domain, the framework propose an optimised solution to reduce the total degrees of freedom (DOF) consequently computational time by: (1) solving thermal-fluid problem (mass/energy/momentum conservation equations) only in the small moving fluidic zone containing the melting pool; (2) solving heat transfer problem (energy conservation equation) in a part of zone outside fluidic zone; (3) making the elements far from heating source (smaller than user-defined active temperature) to be non-active and fast evaluating the temperature evolution by an explicit time integration schema. Finally, the proposed framework has been successfully applied to laser welding and laser-based Direct Energy Deposition (DED) simulations, demonstrating good agreement with experimental and numerical benchmarks. For the given examples, the proposed framework is about 5 times more efficient than the classical ALE model.