Abstract:The intricate process of bubble growth dynamics involves a broad spectrum of physical phenomena from microscale mechanics of bubble formation to macroscale interplay between bubbles and surrounding thermo-hydrodynamics. Traditional bubble dynamics models including atomistic approaches and continuum-based methods segment the bubble dynamics into distinct scale-specific models. In order to bridge the gap between microscale stochastic fluid models and continuum-based fluid models for bubble dynamics, we develop a composite neural operator model to unify the analysis of nonlinear bubble dynamics across microscale and macroscale regimes by integrating a many-body dissipative particle dynamics (mDPD) model with a continuum-based Rayleigh-Plesset (RP) model through a novel neural network architecture, which consists of a deep operator network for learning the mean behavior of bubble growth subject to pressure variations and a long short-term memory network for learning the statistical features of correlated fluctuations in microscale bubble dynamics. Training and testing data are generated by conducting mDPD and RP simulations for nonlinear bubble dynamics with initial bubble radii ranging from 0.1 to 1.5 micrometers. Results show that the trained composite neural operator model can accurately predict bubble dynamics across scales, with a 99% accuracy for the time evaluation of the bubble radius under varying external pressure while containing correct size-dependent stochastic fluctuations in microscale bubble growth dynamics. The composite neural operator is the first deep learning surrogate for multiscale bubble growth dynamics that can capture correct stochastic fluctuations in microscopic fluid phenomena, which sets a new direction for future research in multiscale fluid dynamics modeling.

What problem does this paper attempt to address?

This paper focuses on the scale coupling problem in the multiscale bubble growth dynamics. Traditional bubble dynamics models usually separate the dynamic process into two independent models: microscopic mechanics and macroscopic thermal fluid dynamics. In order to bridge the gap between the microscopic stochastic fluid model and the continuum-based fluid model, researchers have developed a composite neural operator model by integrating many-body dissipative particle dynamics (mDPD) model and the continuum-based Rayleigh-Prandtl (RP) model. The model utilizes a novel neural network architecture that includes a deep operation network (for learning the average behavior of bubble growth under pressure changes) and a long short-term memory network (for learning the statistical characteristics of microscale bubble dynamics). The aim is to unify the analysis of nonlinear bubble dynamics at the microscale and macroscale. The paper points out that bubbles play a key role in various industrial processes such as water treatment, mineral flotation, and micro-robotics. Nonlinear bubble growth dynamics is a multiscale phenomenon that involves a wide range of physical processes from bubble formation and growth to the interaction with the surrounding thermal fluid dynamics. Under the continuum limit, the Rayleigh-Prandtl equation can describe bubble growth dynamics. However, at the microscale, bubble growth is affected by the competition of fluid dynamic fluctuations and surface tension, leading to strong stochastic processes. To address this issue, the researchers propose a composite neural operator model that can accurately predict bubble dynamics at different scales, including the time evolution of bubble radius under pressure changes, and capture the correct stochastic fluctuations of microscale fluid phenomena. By simulating and generating training and testing data using mDPD and RP, the model demonstrates 99% prediction accuracy on bubbles with different initial radii and reproduces size-dependent stochastic fluctuations of bubble growth dynamics at the microscale. In summary, this paper aims to establish a deep learning agent model that can capture the correct stochastic fluctuations in microscale fluid phenomena, paving the way for future multiscale fluid dynamics modeling.

Bridging scales in multiscale bubble growth dynamics with correlated fluctuations using neural operator learning

A seamless multiscale operator neural network for inferring bubble dynamics

Multiscale Fluid Mechanics and Modeling

BubbleNet: Inferring micro-bubble dynamics with semi-physics-informed deep learning

Bubbles in Turbulent Flows: Data-driven, kinematic models with memory terms

Multiscale modeling framework of a constrained fluid with complex boundaries using twin neural networks

Novel Models for Predicting the Shape and Motion of an Ascending Bubble in Newtonian Liquids Using Machine Learning

Bubble Detection in Multiphase Flows Through Computer Vision and Deep Learning for Applied Modeling

Interfacing finite elements with deep neural operators for fast multiscale modeling of mechanics problems

MRCNN: multi-input residual convolution neural network for three-dimensional reconstruction of bubble flows from light field images

Multiscale Neural Operator: Learning Fast and Grid-independent PDE Solvers

The Dynamics of Bubbles

Application of Two-Branch Deep Neural Network to Predict Bubble Migration Near Elastic Boundaries

Machine learning of hidden variables in multiscale fluid simulation

Application of Two Population Balance Models to Predicting Large Scale Multi-Size Bubbly Flow

Bayesian calibration of bubble size dynamics applied to CO2 gas fermenters

Interpreting neural operators: how nonlinear waves propagate in non-reciprocal solids

Computationally Efficient Multiscale Neural Networks Applied to Fluid Flow in Complex 3D Porous Media

Machine learning models for the secondary Bjerknes force between two insonated bubbles

Scaling field-theoretic simulation for multi-component mixtures with neural operators

Adaptive physics-informed neural operator for coarse-grained non-equilibrium flows