Accelerating Transient CFD through Machine Learning-Based Flow Initialization

2025-03-20

AI Summary

To address the excessive computational time required for statistical steady-state convergence in transient CFD simulations caused by suboptimal initial flow fields, this paper proposes a physics-informed machine learning (PIML) rapid initialization method. The approach constructs a geometry-generalizable deep neural network that embeds physical constraints and is tightly coupled with an unsteady RANS solver; it further introduces a sliding-time-averaged robust convergence criterion to eliminate reliance on subjective thresholds. Key contributions include: (1) the first physics-guided hybrid initialization strategy; (2) end-to-end deployable ML-based initial field generation without modifying the underlying solver; and (3) demonstrated performance on a 16.7-million-cell automotive aerodynamic simulation—achieving a 50% reduction in convergence time, initialization in seconds, and accuracy comparable to high-cost steady-state RANS initial fields.

Technology Category

• Machine Learning: Hardware-aware ML
• Search and Optimization: Sampling/Simulation-based Search
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Transient computational fluid dynamics (CFD) simulations are essential for many industrial applications, but a significant portion of their computational cost stems from the time needed to reach statistical steadiness from initial conditions. We present a novel machine learning-based initialization method that reduces the cost of this subsequent transient solve substantially, achieving a 50% reduction in time-to-convergence compared to traditional uniform and potential flow-based initializations. Through a case study in automotive aerodynamics using a 16.7M-cell unsteady RANS simulation, we evaluate three ML-based initialization strategies. Two of these strategies are recommended for general use: (1) a physics-informed hybrid method combining ML predictions with potential flow solutions, and (2) a more versatile approach integrating ML predictions with uniform flow. Both strategies enable CFD solvers to achieve convergence times comparable to computationally expensive steady RANS initializations, while requiring only seconds of computation. We develop a robust statistical convergence metric based on windowed time-averaging for performance comparison between initialization strategies. Notably, these improvements are achieved using an ML model trained on a different dataset of automotive geometries, demonstrating strong generalization capabilities. The proposed methods integrate seamlessly with existing CFD workflows without requiring modifications to the underlying flow solver, providing a practical approach to accelerating industrial CFD simulations through improved ML-based initialization strategies.

Problem

• Reduces computational cost of transient CFD simulations
• Improves initialization using machine learning strategies
• Enhances convergence speed without modifying CFD solvers

Innovation

• Machine learning-based initialization reduces CFD convergence time.
• Physics-informed hybrid method combines ML with potential flow.
• ML model generalizes across different automotive geometries.