Management framework for Additive Manufacturing

Additive manufacturing is a recent trend in production processes that can be defined as the process of producing parts through the deposition of material in a layer-by-layer fashion. The advantages of this technique are the material saving and the ability to produce pieces with complex geometric shapes that cannot be easily achieved with other techniques, such as CNC machining (subtractive manufacturing). However, AM processes present complications due to lower accuracy and often longer build times compared to their subtractive counterparts. AM exposes a larger number of process parameters, and the interactions between material properties and those parameters play a critical role in determining the final part's quality. There are different types of AM, but in the case under study only Selective Laser Melting (SLM) was considered. The idea behind this thesis is to improve an already existent AM platform in order to manage the entire production life cycle: selection of process parameters, process simulation and real-time communication with the machine for data acquisition. In this work, an MQTT connector was developed to enable communication between the AM Platform and SLM printers. In doing so, it is possible to collect in real time information about the printing jobs and their parameters; the data produced by the machines are recorded into the platform and presented through interactive dashboards at both printer and job level, allowing users to monitor current and past jobs. In this implementation, the platform also embeds existing modules of Physics-Informed Neural Networks (PINNs) for 3D spatiotemporal heat transfer modeling of a laser powder bed fusion (LPBF) process. Those simulations are available on the platform in a user-friendly web interface; the user could tweak the process parameters like scan velocity and laser power in order to estimate the temperatures reached in a selected region of the material in a given time interval.