| Abstract Scope |
Additive manufacturing (AM) has substantial benefits versus conventional part manufacture, including the ability to rapidly iterate designs without requiring tooling, or incorporation of complex geometric features to enable multi-functionality. The AM process itself, however, is prone to the formation of defects, thus requiring extensive post-processing or intensive ex-situ inspection to ensure quality.
Here we present recent work that couples efficient physical process modeling tools with experimental data to rapidly generate digital twins of AM parts. These hybrid data and physics-informed models capture each printed part’s unique microstructure and defect characteristics, which can in turn be exercised to determine part performance. Using these tools, HRL is developing a framework for rapidly computing AM part quality, paving the way to high-fidelity, in-situ quality monitoring and unlocking the potential of AM for a broad range of quality-critical applications. |