New data perspective for selective laser melting simulations

In selective laser melting (SLM), a powdered material is locally melted by and, after cooling, forms coherent solid structure that enables the production of complex geometries with various materials. The process involves extreme heating and cooling rates thus, large temperature gradients, which lead to anisotropic properties on macroscopic scale in worst case, reduced mechanical properties. order reliably predict final component properties, simulations can be performed at different time length scales. Enormous computational resources are often required perform such simulations. transform these into suitable surrogate models, generated data must compressed evaluated way. This paper shows first preliminary work possible new description