Shock-Aware Physics-Guided Fusion-DeepONet Operator for Rarefied Micro-Nozzle Flows
This work addresses the need for efficient simulation tools in micro-nozzle flow analysis, but it appears incremental as it builds on existing operator learning methods with domain-specific adaptations.
The authors tackled the problem of constructing fast and accurate surrogate models for rarefied micro-nozzle flows with shocks by developing a physics-aware deep learning framework, achieving validation on the canonical viscous Burgers equation.
We present a comprehensive, physics aware deep learning framework for constructing fast and accurate surrogate models of rarefied, shock containing micro nozzle flows. The framework integrates three key components, a Fusion DeepONet operator learning architecture for capturing parameter dependencies, a physics-guided feature space that embeds a shock-aligned coordinate system, and a two-phase curriculum strategy emphasizing high-gradient regions. To demonstrate the generality and inductive bias of the proposed framework, we first validate it on the canonical viscous Burgers equation, which exhibits advective steepening and shock like gradients.