Drifting Models for Surrogate Flow Modeling
2026-06-05 • Machine Learning
Machine Learning
AI summaryⓘ
The authors address the problem that traditional Computational Fluid Dynamics (CFD) models are accurate but slow, making it hard to quickly test different indoor environments. They use a new method called generative drifting, which creates flow simulations much faster by working in a simpler mathematical space and conditioning results on boundary conditions. Their approach achieves similar accuracy to slower methods but runs about 100 times faster. They also explore a version that might work for new shapes, aiming to generalize better. Overall, their method offers a faster alternative for generating fluid flow simulations when speed is very important.
Computational Fluid Dynamics (CFD)Generative SurrogatesGenerative DriftingVariational Autoencoder (VAE)Latent SpaceBoundary ConditionsDiffusion ModelsFlow SimulationSpatial ConditioningSurrogate Modeling
Authors
Chris R. Jung, Markus Dörr, Natalie Jüngling, Jennifer Niessner, Adam T. Müller, Nicolaj C. Stache
Abstract
While Computational Fluid Dynamics (CFD) provides high-fidelity flow fields for optimizing indoor environments, its computational cost limits rapid exploration. To solve this problem generative surrogates offer better distribution modeling than deterministic networks, but iterative sampling is slow. To enable high-quality, single-pass generation, we adapt the novel generative drifting framework to fluid mechanics. We introduce a conditional architecture that performs drifting in a learned VAE latent space and uses label-aware masking to align generated samples with their boundary conditions. Our label-conditioned model matches iterative diffusion in accuracy and flow consistency while running two orders of magnitude faster. Additionally, we propose a spatial-conditioning variant that establishes a promising path towards generalization to unseen geometries. Ultimately, conditional drifting serves as a highly efficient alternative to diffusion based approaches, unlocking real-time CFD surrogates where inference speed is critical.