Digital Feedforward Control of Laser Powder Bed Fusion Additive Manufacturing

Abstract

This work concerns the laser powder bed fusion (LPBF) additive manufacturing process. We developed and applied a physics-based model predictive process control approach to regulate the spatiotemporal temperature distribution (thermal history) of an LPBF part.   The approach mitigated within-part variation in critical-to-quality aspects, such as grain size, surface finish, and geometric integrity. Currently, to obtain desired properties, the processing parameters for an LPBF part are optimized through exhaustive empirical build-and-test procedures. Once optimized, the processing parameters are maintained constant for all layers of the part. However, in complex parts, the differences in thermal history between layers causes significant within-part variation in properties. By contrast, the model predictive control approach maintains the thermal history within a desired window by changing the processing parameters between layers. Given a part geometry, material properties and a desired thermal history, the controller autonomously optimizes the processing parameters layer-by-layer before the part is printed ‒ a form of digital feedforward control. To demonstrate the approach, five thermal history control strategies were tested on four unique part geometries (20 total parts) made from stainless steel 316L alloy. Post-process analysis showed that, compared to uncontrolled processing, layer-wise model predictive control of the thermal history significantly reduced variations in grain size, and improved geometric accuracy and surface finish. Thus, this work takes a critical step toward physics-based control of part properties in LPBF.