Emergent Transfer of a Physics Foundation Model from Simulation to Laboratory Turbulence

2026-05-31

AI Summary

This work addresses the long-standing discrepancy in the mixing growth rate α observed between simulations and experiments in Rayleigh–Taylor instability (RTI). Leveraging Walrus—a physics-based fluid dynamics foundation model fine-tuned solely on a limited set of direct numerical simulations (DNS)—the study achieves zero-shot transfer to real laboratory RTI experiments with a sliding barrier, without incorporating any experimental data. The model accurately predicts the experimentally observed mixing growth rate (α ≈ 0.06–0.07) and generalizes to stable stratified buoyancy configurations unseen during training, effectively suppressing unphysical mixing growth. This represents the first demonstration that a physics-informed foundation model can generalize zero-shot to real turbulent experiments, highlighting initial conditions as pivotal in bridging the simulation–experiment gap and offering new evidence for data-driven approaches in tackling century-old challenges in fluid dynamics.

Abstract

Whether physics foundation models can be usefully deployed on laboratory experiments remains an open question for scientific machine learning (ML). We test this question on the Rayleigh-Taylor instability (RTI), a ubiquitous and demanding fluid instability seen from tabletop flows to supernova explosions, in which small perturbations at a density interface grow into chaotic, multiscale mixing as a lighter fluid accelerates into a heavier one. Standard ML models struggle with RTI, and despite over a century of theoretical, numerical, and experimental work, it carries an unresolved discrepancy between simulation and experiment: the late-time mixing growth rate, $α$, measured in most laboratory experiments ($\sim$ 0.06-0.07), is roughly three times the value from idealized direct numerical simulations (DNS, $\sim$ 0.02). The gap's origin remains debated. These properties make RTI a stringent test for a question that matters well beyond RTI: can foundation models trained only on simulations generalise to sparse, messy, and noisy laboratory settings? We finetune Walrus, a foundation model for continuum dynamics, on three or fewer DNS realizations and recover key RTI physics over long rollouts. Applied zero-shot to sliding-barrier laboratory data, the finetuned model leaves the DNS-like regime and enters the observed growth band, having never seen a single experimental sample. These results provide independent, data-driven evidence that initial conditions play a crucial role in the longstanding sim-experiment gap in $α$. The model also generalises zero-shot to stable stratification, a buoyancy regime absent from training, correctly slowing mixing-layer growth. Together, our results show that foundation models can generalise well beyond their training data, predicting laboratory behavior and unseen physical regimes, opening new ways to probe longstanding simulation-experiment gaps.

Problem

• Rayleigh-Taylor instability
• simulation-experiment gap
• mixing growth rate
• foundation models
• laboratory turbulence

Innovation

• foundation model
• Rayleigh-Taylor instability
• simulation-to-experiment transfer
• zero-shot generalization
• turbulent mixing