Reliability centered additive manufacturing computational design framework

Analatom, Applied Research Laboratory Penn State University, and University of California Irvine propose to apply modern concepts from simulation, corrosion modeling, and control theory along with Prognostics and Health Management (PHM) in the development of an end-to-end, prognostics-based Additive Manufacturing (AM) materials assessment architecture. The project team will consider four potential levels of implementation that includes molecular dynamics simulation of material corrosion and cracking defects, AM advanced statistical process control and monitoring, enhancing end-use items with corrosion, strain sensors and operational environment assessment techniques. Selection of appropriate applied levels architecture will be dependent on the desired target application. Emphasis will be placed on developing a flexible computational framework that incrementally learns optimal parameters and discerns parameter sets that lead to degradation profiles. An approach for a Reliability-Centered AM Computational Design Framework (RCAM CDF) focused on selective laser melt (SLM) AM processes for metallic components has been developed. This design framework initially includes minimum part geometry for a load bearing coupon structure, with defect types and densities induced by design of the AM build fabrication process, followed by laser Computed Tomography (CT) inspection of the components, with subsequent in situ structural monitoring sensors (corrosion, strain) to record the accelerated lifetime and fatigue testing. Existing designs, associated design tools appropriate for design, materials, processes, standards and data management systems already exist. These designs already have associated failure modes, fault analysis tools and fault isolation manuals referenced to actual designs. Introducing new tools that attempt to unilaterally replace existing design tools will inhibit RCAM CDF adoption rate. An alternate approach that Analatom recommends uses associative memory linking data exported from design tools, materials, processes, standards, and data management systems across the entire US Navy enterprise creating a library of reference Additive Manufacturing designs. The designs would be based on original designs modified for Additive Manufacturing. Preliminary proof has been demonstrated that this approach can capture and find “seeded” print defects correlating to scan images taken during build. The associative memory image based “defect” scanner has been able similar “defect regions” corresponding to small simulated corrosive pits.