Additive manufacturing (AM) increases the speed and flexibility of production and enables traditional part concatenation for advanced manufacturing capabilities. The ability for U.S. manufacturers to 3D-print advanced components in-house reduces reliance on traditional subtractive supply chains and bolsters national security readiness. While AM affords unique flexibility in design for manufacturability, its variance in lack of repeatability and reproducibility introduced various defects, including lack of fusion, gas entrapment, powder agglomeration, balling, internal cracks, and thermal stress, that degrades mechanical properties of final parts. The cost associated with performing post process inspection is an economic limiter and its efficacy is limited by material and material geometry, that is solved by in process nondestructive inspection methodologies. Kitware, in collaboration with Sigma Additive Solutions, proposes to bring the latest advances in deep neural network artificial intelligence and signal fusion to optimize and extend PrintRite3D for the Navys unique needs. PrintRite3D is a platform-independent, interactive, in-process quality assurance system that combines inspection, feedback, data collection and critical analysis. Optimizing PrintRite3D defect detection accuracy will improve confidence in and reduce part-rejection false-alarm rates. Our proposed method builds on an existing proof of concept for in-situ defect detection and extends our capabilities to cover a wider range of builds, printers, locations, and sensors.
Benefit: The Navy has a critical need to rapidly construct ships at reduced cost to develop and maintain a large fleet. To achieve this will require significant innovation in manufacturing processes. U.S. President Joe Bidens new initiative aimed at improving domestic supply chain resilience by focusing on AM highlights the importance of this need. In fact, Metal AM reported in May 2022 that Lockheed Martin will work with its suppliers to conduct research to improve the performance of techniques that are specifically focused on the use of AM as an alternative to castings and forgings. Programs like the one that began in 2020 to build Columbia-class submarines, for example, needed to be fast-tracked due to the U.S. imperative to remain competitive with Chinas large Navy investments. The global pandemic has also severely impacted supply chains thereby highlighting the need for a 3D printing-enabled supply chain. Among the top risks to the critical Columbia-class ballistic missile submarine program is fragility in key parts of the industrial base. While AM can accelerate production schedules, it requires changing how nondestructive testing is done. Our proposed work will provide commercial routes for expansion of this needed capability.
Keywords: Machine Learning, Machine Learning, Artificial Intelligence, additive manufacturing, Nondestructive Evaluation, defects, AM, AI/ML
