Feed-Forward Controls for Laser Powder Bed Fusion Based Metal Additive Manufacturing
Navy SBIR FY2018.1

Sol No.: Navy SBIR FY2018.1
Topic No.: N181-085
Topic Title: Feed-Forward Controls for Laser Powder Bed Fusion Based Metal Additive Manufacturing
Proposal No.: N181-085-0014
Firm: Applied Optimization, Inc.
3040 Presidential Drive, Suite 100
Fairborn, Ohio 45324
Contact: Anil Chaudhary
Phone: (937) 431-5100
Web Site: http://www.appliedO.com
Abstract: The research objective of the proposed work is to develop a concept for a feed-forward control (FFC) hardware, algorithms, and multi-physics-based models to allow real-time tracking of powder bed layer variability and corresponding laser processing parameter compensation to improve the quality of laser fusion-based metal additive manufacturing (AM) parts. The FFC concept is designed to compensate for the systematic physical property variabilities of the powder bed layer arising from the non-uniformity of thermal evolution for the AM build and recoat layer powder. This includes the characterization of systematic layer thickness variability that is caused by splatter, molten particle ejects, denudation zones, and the previous layer surface roughness. The proposed FFC compensation algorithms build upon the predictions of multi-physics models in order to minimize defects and the non-uniformity of microstructure, which arise from the inconsistency of the melt-pool shape and its solidification behavior. The parameter compensation is designed to predict the current, relevant information using off-the-shelf hardware and data-reduction procedures that have been demonstrated. The Phase I will refine these procedures and the hardware utilization for real-time FFC application. It will also specify a range of parameter variables for the purpose of model development, system verification, and technology validation.
Benefits: The benefits of the proposed concept for FFC are: (1) It utilizes off-the-shelf hardware and existing algorithms to demonstrate the characterization the systematic property variabilities of the powder bed layer in real-time; (2) It compensates the process parameters to optimize the confidence interval to minimize build defects; (3) It utilizes the dimensionality reduction on multi-scale, multi-spectral sensor data to perform FFC; and (4) It is designed to integrate the sensors and algorithms with the control system of a commercial laser powder bed equipment. The commercial application from the proposed concept will be a FFC module that consists of sensor hardware, its implementation design, data reduction procedures, and integrated algorithms to compensate process parameters to maximize the confidence level for a quality AM build.