Hi Romesh, there are several approaches to creating a digital twin of an AM process. First you must consider the interpretation of a digital twin to define its scope, and then you have to consider what AM process you want it to virtualise. This project focused on creating a digital twin of a LB-PBF machine using machine-focused sensors and inferring process performance from there. More details on how we did this are available from the project website: digitbrain.eu/1st-wave-of-digitbrain-experiments/experiment-4-digital-twin-for-additive-manufacturing-am-ensuring-compliance-across-multiple-machines/ If you are interested in looking at a different approach, I suggest reviewing the approach taken by experiment 7 of the Digitbrain project: digitbrain.eu/1st-wave-of-digitbrain-experiments/experiment-7-data-driven-modeling-of-powder-bed-fusion-technology-to-improve-product-quality/. In this experiment, they reviewed emission data from the laser and melt pool and used a digital twin algorithm to score and report on the process. Both experiments are valid approaches and provide great inside and value to users of such equipment.
