MTL member Manolis Stathatos successfully defends his PhD Thesis
MTL member Manolis Stathatos has successfully defended his PhD thesis titled “Optimisation of selective laser sintering and melting process (SLS/SLM) based on numerical simulation and machine learning”.
The novel contributions of his work can be summarized as:
- A numerical finite element thermal model (Parent) for predicting temperature range and final material density in SLS / SLM, optimized for efficient production of large volumes of training data.
- An ANN metamodel for long random trajectories (> 100 m) based on an original descriptor with memory and a repetitive fed scheme of parallel simulations in whole trajectories, achieving simulation times shorter than the actual SLS/SLM processing time.
- A Power Model (ANN) that implements an adaptive control law, designed for parallel application on the whole power profile along the trajectory, achieving an extremely smooth temperature front after a small number of repetitions.
Two excellent papers have already been published from this work and a third one is under evaluation.