Algorithmic Solutions for Third Party Low-Cost Sensing of LPBF

Most current in-situ sensing technologies seek to limit LPBF failures by directly evaluating part quality, or process monitoring. For example, monitoring the thermal properties of the build to predict part warping that may lead to a re-coater crash. However, many LPBF build failures can occur simply due to system maintenance, wear, errors, etc. Monitoring of the physical system to limit build failures is known as system health monitoring. In this program, machine learning algorithms will be demonstrated using data acquired from low-cost, non-invasive sensors to detect and report errors related to processes such as layering, positioning, inerting, etc. Such sensors complement detailed melt-pool and build monitoring for qualification and certification of machines and built parts.