SBIR/STTR Award attributes
We have demonstrated in phase I a methodology for establishing optimized processing parameters using melt pool characteristics as recorded from in-situ co-axial sensors linked to our machine learning tools. These methodologies are designed for developing data-driven models from sparse experimental and modeling data and for multi-objective optimization. In phase II of this effort, we will continue to develop the integration of sensors, models, and fatigue testing for a beta version of the new app, Smart-iCAAM, to be integrated with our ICME toolset iCAAM (integrated computation adaptive additive manufacturing). The first goal is to continue to expand a library of material specific “healthy” melt pool characteristics using an integration of sensor measurements and 3D simulations. This library of simulation data will enable us to compress the needed data to only the relevant features that can be scaled up to full parts. The second goal during phase II will be to establish corrective actions if the streaming and analyzed melt pool characteristics deviate from the optimum. This corrective action requires feedback to the AM machine via application programming interfaces (API) that MRL is developing on a machine inhouse. These two steps will allow smart-iCAAM to become the software that converts AM machines from passive machines to “self-driving” machines (i.e., smart machines). This will pave the path for sensor-driven smart-iCAAM part certification methodology that is based on location specific fatigue predictions with expected wide technology transition (DoD) and commercialization (OEM).
