Calibrated closed-loop control to reduce the effect of geometry on mechanical behaviour in directed energy deposition

Abstract

Directed energy deposition (DED) is an emerging technology with significant industrial potential in the repair of critical aerospace components, however its adoption has been limited by concerns about geometry-driven microstructural and mechanical property variation. These could be resolved by controlling the local temperature field, which would result in a consistent and predictable cooling profile. Closed-loop control approaches have been investigated previously, but with limited assessment of mechanical properties and only on small builds. In this work, we confirm that using fixed build parameters results in a statistically significant, geometry-driven variation in the bulk mechanical properties of DED-built 316 L steel. To address this issue, we have developed an industrially-suitable control algorithm using a low-cost coaxial camera, applying statistical process control techniques to identify representative melt pool images from the livestream. This has been tested on long builds, maintaining a control adjustment frequency of 1 Hz on build durations of >1 hour. Performance has been quantified through bulk mechanical testing, which confirmed that the control algorithm successfully eliminated the component-scale trends in melt pool size, and achieved a geometry-agnostic process with improved mechanical homogeneity. • Component geometry affects bulk mechanical properties in directed energy deposition • Geometry effects can be eliminated by closed-loop control of laser power • Statistical filtering was successful at selecting representative images for control • Control adjustments were maintained at 1 Hz over >1 hour build durations • Control yielded more predictable bulk mechanical properties with lower variation