November/December AMP_Digital

A D V A N C E D M A T E R I A L S & P R O C E S S E S | N O V E M B E R / D E C E M B E R 2 0 1 8 8 0 3D PRINTSHOP MACHINE LEARNING AND 3D PRINTING JOIN FORCES High-value and intricate 3D-print- ed parts can require constant monitor- ing by experts to get them right. If any section of a part is below par, it can render the whole component unusable. That’s why Lockheed Martin and the Office of Naval Research are exploring how to apply artificial intelligence to train robots to independently over- see and optimize 3D printing of com- plex parts. The two-year, $5.8 million con- tract specifically studies multi-axis ro- bots that use laser beams to deposit material. The team led by Lockheed Martin’s Advanced Technology Center will develop software models and sen- sor modifications for the robots to build better components. Researchers will apply machine learning techniques to additive manufacturing so variables can be monitored and controlled by the robot during fabrication. High-quality production using robots will allow for printing replacement parts at sea or for new truck parts to be built in the desert, taking 3D printing to the next level. Lockheed Martin’s research will help machines make decisions about how to optimize structures during the 3D printing process based on previous- ly verified analysis. That verified anal- ysis and integration into a 3D printing robotic system is core to this new con- tract. Lockheed Martin and its team will vet common types of microstructures used in an additive build. The team will measure the performance attributes of the machine parameters and these microstructures, and align them to ma- terial properties before integrating this knowledge into a working system. With this information, machines will be able to make decisions about how to print a part that ensures reliable performance. lockheedmartin.com . 3D FLAWS CAUGHT ON CAMERA The DOE’s Argonne National Labo- ratory recently acquired a new infrared camera at its Advanced Photon Source (APS) facility that narrows the gap be- tween basic and applied research in ad- ditive manufacturing (AM). One of the largest challenges facing the 3D printing industry is how to ensure high-quality reproducibility of parts. The new cam- era provides some insights into the for- mation of defects by allowing research- ers to measure thermal signatures across surfaces in real time. The combined diagnosis tools of infrared and x-ray imaging enable in- dustrial practitioners and researchers to capture x-ray images at 1,000,000 frames per second and thermal images at 100,000 frames per second during the 3D printing process. This creates mo- vies of the formation of key defects caused by melt pool instability, pow- der spatter ejection, and inappropriate scan strategy. Insights provided by the camera can be used to reduce varia- tions in part design and improve the efficiency of AM for defense, medicine, automotive, and other applications. An- other key benefit of the cameras is their ability to be integrated into AM systems. Users of these systems could attach in- frared cameras to their machines to leverage insights found from coupling x-ray and infrared imaging, such as a thermal signature (found through infra- red imaging) correlated with the forma- tion of a defect (captured through x-ray imaging). anl.gov. The addition of a new infrared camera at Argonne’s Advanced Photon Source facility allows researchers to delve into the detailed dynamics of 3D printing by measuring thermal signatures across surfaces in real time. Courtesy of Shutterstock/MarinaGrigorivna. Lockheed Martin and the U.S. Navy are customizing robots to build complex 3D-printed components.