Colloquia

Summary

This thesis investigates thermal modelling methods and variable parameter optimization strategies to mitigate overheating in Metal Laser Powder Bed Fusion (M-LPBF). The focus lies on the development of an Analytical Melt Pool Model (AMPM) for optimizing laser power and scan speed at the layer-scale. Key findings show that maintaining the calculated melt pool width within defined limits using optimization algorithms effectively reduces thermal gradients, thereby preventing defects such as keyholing and lack-of-fusion. Laser power emerges as the more effective control parameter, while scan speed adjustment introduces added complexity due to thermal accumulation effects.

Step-wise and Newton-Raphson optimization algorithms are applied to selected layers, with the Newton-Raphson method resulting in faster and better convergence. Used independently or in combination with the existing Finite Element (FEM) model from NLR, the AMPM leads to improved microstructural homogeneity and reduced porosity, as shown by experimental results. A multi-scale modelling approach shows strong potential to capture a wide range of overheating mechanisms.

By advancing thermal modelling and thermal control in M-LPBF, this research contributes to improved part quality and supports the development of more efficient certification pathways, facilitating the broader industrial adoption of M-LPBF in the aerospace sector.