Luminary has introduced SHIFT-Battery, a Physics AI surrogate model designed to support cooling channel design for electric vehicle battery pack cold plates. Trained on conjugate heat transfer simulation data using NVIDIA‘s GeoTransolver architecture, the model aims to predict module surface temperature, channel velocity, and pressure drop for parameterized cold plate geometries and operating conditions.
The SHIFT-Battery, is designed to bridge the gap between detailed CFD studies and simplified lumped-parameter tools, helping thermal, safety, and pack reliability engineers evaluate channel layout trade-offs earlier in the design process while reducing the need to perform dedicated CFD simulations for every design iteration.
Image generated by: ChatGPT