DoMINO: A Decomposable Multi-scale Iterative Neural Operator for Modeling Large Scale Engineering Simulations

TL;DR

This paper introduces DoMINO, a novel neural operator architecture that efficiently models large-scale engineering simulations by leveraging multi-scale, decomposable, and iterative features, improving accuracy and scalability over traditional ML approaches.

Contribution

The paper presents DoMINO, a point cloud-based neural operator that enhances simulation accuracy and scalability for complex engineering problems, addressing limitations of existing ML models.

Findings

Demonstrates high accuracy and generalization on automotive aerodynamics data.

Shows improved scalability and performance over existing ML surrogate models.

Validates effectiveness using engineering-specific metrics.

Abstract

Numerical simulations play a critical role in design and development of engineering products and processes. Traditional computational methods, such as CFD, can provide accurate predictions but are computationally expensive, particularly for complex geometries. Several machine learning (ML) models have been proposed in the literature to significantly reduce computation time while maintaining acceptable accuracy. However, ML models often face limitations in terms of accuracy and scalability and depend on significant mesh downsampling, which can negatively affect prediction accuracy and generalization. In this work, we propose a novel ML model architecture, DoMINO (Decomposable Multi-scale Iterative Neural Operator) developed in NVIDIA Modulus to address the various challenges of machine learning based surrogate modeling of engineering simulations. DoMINO is a point cloudbased ML model that uses local geometric information to predict flow fields on discrete points. The DoMINO model is validated for the automotive aerodynamics use case using the DrivAerML dataset. Through our experiments we demonstrate the scalability, performance, accuracy and generalization of our model to both in-distribution and out-of-distribution testing samples. Moreover, the results are analyzed using a range of engineering specific metrics important for validating numerical simulations.