Validation of transient heat models in Laser Powder Bed Fusion of Metals
Please login to view abstract download link
This presentation gives an overview over recent advancements in modeling laser based powder bed fusion of metals (PBF-LB/M) with a focus on non-linear transient heat diffusion models approximated by finite elements. Such models are rather basic, yet deliver very accurate results for cases where diffusion is the dominating physical phenomenon. This is the case for melt pools in conduction mode for which we present a thorough double blinded validation. Input to the model and the experiment are different laser shapes (i.e. laser power densities) and the scan strategy. The output, that is compared, is the geometry of the melt pool. We not only demonstrate that the melt pool geometries are accurately predicted for such cases. We also show that the model is able to indicate when it leaves its range of validity~\cite{Holla}. The validation above is carried out for stationary cases where determining the model-error w.r.t. the melt pool geometry is straightforward. This changes in transient cases for which the talk will present different measures together with their characteristics when assessing the accuracy of the model. The above investigations form a solid basis for answering the inverse question: How should the laser parameters be chosen to produce a desired melt pool shape? The developed numerical methodology, which is freely available~\cite{hollaLaserBeamShape2023}, is shortly presented. Finally, the talk will give an outlook on how to extend the inverse modeling to control the microstructural evolution.