Abstract:Physics-based simulations are often used to model and understand complex physical systems and processes in domains like fluid dynamics. Such simulations, although used frequently, have many limitations which could arise either due to the inability to accurately model a physical process owing to incomplete knowledge about certain facets of the process or due to the underlying process being too complex to accurately encode into a simulation model. In such situations, it is often useful to rely on machine learning methods to fill in the gap by learning a model of the complex physical process directly from simulation data. However, as data generation through simulations is costly, we need to develop models, being cognizant of data paucity issues. In such scenarios it is often helpful if the rich physical knowledge of the application domain is incorporated in the architectural design of machine learning models. Further, we can also use information from physics-based simulations to guide the learning process using aggregate supervision to favorably constrain the learning process. In this paper, we propose PhyDNN, a deep learning model using physics-guided structural priors and physics-guided aggregate supervision for modeling the drag forces acting on each particle in a Computational Fluid Dynamics-Discrete Element Method(CFD-DEM). We conduct extensive experiments in the context of drag force prediction and showcase the usefulness of including physics knowledge in our deep learning formulation both in the design and through learning process. Our proposed PhyDNN model has been compared to several state-of-the-art models and achieves a significant performance improvement of 8.46% on average across all baseline models. The source code has been made available and the dataset used is detailed in [1, 2].

Analysis and Development of Novel Data-Driven Drag Models Based on Direct Numerical Simulations of Fluidized Beds

Development and Validation of Filtered Drag Models for Fluidization of Low-Density Geldart A Particles

Experimental and numerical investigation of solid behavior in a gas–solid turbulent fluidized bed

Models of Interphase Drag Force from Direct Numerical Simulations of Upward Turbulent Particle-Laden Channel Flows

Data-driven Nonlinear Constitutive Relations for Rarefied Flow Computations

Drag Model from Interface-Resolved Simulations of Particle Sedimentation in a Periodic Domain and Vertical Turbulent Channel Flows

Two-Fluid Model Simulations of the National Energy Technology Laboratory Bubbling Fluidized Bed Challenge Problem

Novel Drag and Nusselt Number Models Based on Direct Numerical Simulations of a Bidisperse Gas–solid System

Homogeneous drag models in gas–solid fluidization: Big data analytics and conventional correlation

Experimental study and transient CFD/DEM simulation in a fluidized bed based on different drag models

Drag Model of Finite-Sized Particle in Turbulent Wall-Bound Flow over Sediment Bed

Particle‐resolved Direct Numerical Simulation of Gas–solid Dynamics in Experimental Fluidized Beds

Construction and Analysis of the Mesoscale Drag Force Model Based on Machine Learning Methods

Comparative analysis on gas–solid drag models in MFIX-DEM simulations of bubbling fluidized bed

Drag prediction of rough-wall turbulent flow using data-driven regression

A machine learning-based drag model for sand particles in transition flow aided by spherical harmonic analysis and resolved CFD-DEM

Conventional and data‐driven modeling of filtered drag, heat transfer, and reaction rate in gas–particle flows

Numerical Simulation and Experimental Study of the Gas–Solid Flow Behavior Inside a Full-Loop Circulating Fluidized Bed: Evaluation of Different Drag Models

Modeling of Drag with the Eulerian Approach and EMMS Theory for Heterogeneous Dense Gas–solid Two-Phase Flow

Physics-guided Design and Learning of Neural Networks for Predicting Drag Force on Particle Suspensions in Moving Fluids

Modelling of filtered drag force for clustered particle-laden flows based on interface-resolved simulation data