Search NASA⌕ Search

Engineering topics

Lei, Bo

Publications and source records attributed to Lei, Bo.

Comparative study of machine learning techniques for post-combustion carbon capture systems

Computational analysis of countercurrent flows in packed absorption columns, often used in solvent-based post-combustion carbon capture systems (CCSs), is challenging. Typically, computational fluid dynamics (CFD) approaches are used to simulate the interactions between a solvent, gas, and column's packing geometry while accounting for the thermodynamics, kinetics, heat, and mass transfer effects of the absorption process. These simulations can then be used explain a column's hydrodynamic characteristics and evaluate its CO 2 -capture efficiency. However, these approaches are computationally expensive, making it difficult to evaluate numerous designs and operating conditions to improve efficiency at industrial scales. In this work, we comprehensively explore the application of statistical ML methods, convolutional neural networks (CNNs), and graph neural networks (GNNs) to aid and accelerate the scale-up and design optimization of solvent-based post-combustion CCSs. We apply these methods to CFD datasets of countercurrent flows in absorption columns with structured packings characterized by several geometric parameters. We train models to use these parameters, inlet velocity conditions, and other model-specific representations of the column to estimate key determinants of CO 2 -capture efficiency without having to simulate additional CFD datasets. We also evaluate the impact of different input types on the accuracy and generalizability of each model. We discuss the strengths and limitations of each approach to further elucidate the role of CNNs, GNNs, and other machine learning approaches for CO 2 -capture property prediction and design optimization.

97 MATHEMATICS AND COMPUTING↗

Graph Learning in Physical-informed Mesh-reduced Space for Real-world Dynamic Systems

This Git repository contains codes for the 'Graph Learning in Physical-informed Mesh-reduced Space for Real-world Dynamic Systems' paper that will be published in 2023 SIGKDD. This work uses physical-informed prior (PiP) information to learn and predict fluid dynamics in a reduced mesh space. We propose a two-stage graph-based model for fluid velocity field reconstruction and prediction. In the first stage, we learn a subgraph autoencoder to summarize the information in a mesh-reduced space using physical-informed priors. In the second stage, we learn a dynamics predictor to predict subgraph evolution. We demonstrate the effectiveness of our model on two fluid flow datasets: lid-driven cavity flow data and cylinder flow data. This code is also applicable for any other dynamic systems with corresponding data.

Lei, Bo↗

Numerical simulation of flow regions in red mud separation thickener's feedwell by analysis of residence-time distribution

The residence-time distribution (RTD) and the compartment model were applied to characterizing the flow regions in red mud separation thickener’s feedwells. Combined with the experimental work, validated mathematical model as well as three-dimensional computational fluid dynamics (CFD) model was established to analyze the flow regions of feedwells on an industrial scale. The concept of RTD, although a well-known method for the characterization of mixing behavior in conventional mixers and reactors, is still a novel measure for the characterization of mixing in feedwells. Numerical simulation results show that the inlet feed rate and the aspect ratio of feedwells are the most critical parameters which affect the RTD of feedwell. Further simulation experiments were then carried out. Under the optimal operation conditions, the volume fraction of dead zone can reduce by10.8% and an increasement of mixing flow volume fraction by 6.5% is also observed. There is an optimum feed inlet rate depending on the feedwell design. The CFD model in conjunction with the RTD analysis then can be used as an effective tool in the design, evaluation and optimization of thickener feedwell in the red mud separation.

42 ENGINEERING↗