Brian Giera1,Brian Au1,Brian Weston1,Kyle DeVlugt1,Haicho Miao1,Seth Watts1
Lawrence Livermore National Laboratory1
Brian Giera1,Brian Au1,Brian Weston1,Kyle DeVlugt1,Haicho Miao1,Seth Watts1
Lawrence Livermore National Laboratory1
A digital twin is an amalgam of physics-based (high fidelity or reduced order) and/or data driven models that describe a physical system, as shown below. Digital twins exist in a variety of settings ranging from manufacturing and inspection equipment, supply chains, city planning, and so on. In the context of advanced manufacturing (AM), the inputs/outputs of the digital twin and its real-world counterpart are indistinguishable, i.e., both operate on identical machine instructions and produce identical data structures. Like many, LLNL’s approach requires continually refining the digital twin with real data to better capture behavior of its real counterpart via advanced analytical techniques. As such, a digital twin’s evolving parameter set can inform of machine health and aging behavior, providing actionable insights on lifetime performance. Digital twins offer an inexpensive and risk-free environment for machine operator training and troubleshooting complex toolpaths. A suite of digital twins that capture all fabrication and inspection platforms of a given AM process can accelerate production for “born-qualified” parts at scale with minimized and quantified defects. This talk will walk through examples of how we are leveraging data from integrated pairs of real and digital twins of inspection and fabrication platforms to become more flexible and agile.