Digital Twin–Ready Acceleration of Production–Scale Wire Laser Additive Manufacturing Using NVIDIA Omniverse

• A unified NVIDIA Omniverse digital twin–ready framework that integrates robotic control and multiphysics simulation for wire laser additive manufacturing. • A kinematic and dynamic deposition model that predicts geometry deviations caused by robot speed changes, motion blending, acceleration effects and complex multi-axis motion. • A USD-based, GPU accelerated smoothed particle hydrodynamics and finite element multiphysics solvers that captures melt pool behavior, laser and material interaction, solidification, thermal evolution and residual stress. • Experimental validation showing close agreement between simulated and real bead geometry, including accurate identification of unstable deposition conditions. • A complete workflow that enables process parameter optimization, reduces trial-and-error, decreases production time and reducing overall manufacturing costs up to threefold. Predicting and optimizing process parameters for production-scale Wire Laser Additive Manufacturing (WLAM) remains challenging, as current WLAM workflows rely heavily on research setups and trial-and-error tuning. To overcome these limitations, we introduce a digital twin-ready simulation framework that combines multi-axis robotics with high-fidelity multiphysics modeling. Built on the NVIDIA Omniverse ecosystem, the GPU-accelerated framework integrates two complementary modules. The Kinematic and Dynamic Deposition Model (KDDM) provides fast, robotics-focused predictions of motion-related defects arising from motion blending and slicer settings, issues often missed by conventional simulation tools. The Multiphysics Deposition Model (MPDM) captures ray-traced laser–metal interaction and melt pool physics, modeling melt pool flow and solidification to predict layer-by-layer behavior, and includes a thermomechanical capability for forecasting elastoplastic distortion. Quantitative validation against physical builds shows prediction errors of 8.35%, 5.72%, 0.45%, 2.73%, and 16.10% in key geometric indicators. KDDM accurately predicted defects during layer transitions and speed mismatches. MPDM reproduced experimentally observed phenomena such as blobbing, balling, and discontinuous layers across varied process parameters. Applied to a production-scale part, the combined simulation framework enabled parameter optimization that reduced cost and build time by up to threefold relative to manufacturer guidelines, while distortion predictions informed improved part design. This framework provides a robust virtual commissioning and research environment for exploring WLAM strategies, defect mechanisms, and the relationships between geometry, process parameters, and thermal history, accelerating the shift from experimental WLAM setups to high-rate industrial manufacturing.