Hybrid CFD-neural networks technique to predict circulating fluidized bed reactor riser hydrodynamics

Upadhyay, Mukesh; Nagulapati, Vijay Mohan; Lim, Hankwon

doi:10.1016/j.jclepro.2022.130490

Cited by 22 publications

(5 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The RNN model can be adapted to predict chemical reaction rates under various temperatures, reactant concentrations, and other parameters. Moreover, ML has been widely applied in computational fluid dynamics to improve simulation efficiency and accuracy 66 , 67 . The multi-timestep training method proposed in this study has the potential to be applied in such algorithms to improve parameter optimisation.…”

Section: Discussionmentioning

confidence: 99%

Neural network based prediction of the efficacy of ball milling to separate cable waste materials

Han

Kumagai

et al. 2023

Commun Eng

View full text Add to dashboard Cite

Material recycling technologies are essential for achieving a circular economy while reducing greenhouse gas emissions. However, most of them remain in laboratory development. Machine learning (ML) can promote industrial application while maximising yield and environmental performance. Herein, an asynchronous-parallel recurrent neural network was developed to predict the dynamic behaviour when separating copper and poly(vinyl chloride) components from the cable waste. The model was trained with six datasets (treatment conditions) at 3600 epochs. High accuracy was confirmed based on a mean-square error of 0.0015–0.0145 between the prediction and experimental results. The quantitative relationship between the input features and the separation yield was identified using sensitivity analysis. The charged weight of cables and impact energy were determined as the critical factors affecting the separation efficiency. The ML framework can be widely applied to recycling technologies to reveal the process mechanism and establish a quantitative relationship between process variables and treatment outputs.

show abstract

Section: Discussionmentioning

confidence: 99%

Neural network based prediction of the efficacy of ball milling to separate cable waste materials

Han

Kumagai

et al. 2023

Commun Eng

View full text Add to dashboard Cite

show abstract

“…K-means clustering algorithm can be adopted to classify an unlabeled data set into coherent subsets in an unsupervised way and has been proven to be superior in identifying particle clusters in fluidized beds. − Particularly, neural network (NN)-based approaches such as artificial neural networks (ANNs), convolutional neural networks (CNNs), and recurrent neural networks (RNNs) were immensely employed to predict hydrodynamic parameters or develop surrogate models for multiphase flows due to their excellent nonlinear fitting capability. ANN and CNN with multilayer neural networks were typically used in multiphase closure modeling, ,,,− particle/bubble/cluster feature extraction, − flow regime recognition, and hydrodynamic parameter prediction, , and so forth. RNN methods have special circular structures with gating mechanisms that can be applied to predict the temporal evolution of variables by processing sequence-to-sequence flow data …”

Section: Introductionmentioning

confidence: 99%

Machine Learning Assisted Characterization of Local Bubble Properties and Its Coupling with the EMMS Bubbling Drag

Jin,

Hu,

Liu

et al. 2024

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

Empirical correlations for bubble diameter and velocity are incapable of predicting the local bubble behaviors fairly because the impact of local hydrodynamics on bubbles in fluidized beds. Based on image processing, a novel bubble identification method with an adaptive threshold was proposed to distinguish and characterize bubbles in fluidized beds. The information regarding bubble properties and local hydrodynamics can thus be extracted using the big data from highly resolved simulations. Accordingly, the deep neural network was trained to accurately predict local bubble properties, where the inputs were determined by performing correlation analysis and a random forest algorithm. We found Reynolds number, voidage, and relative coordinates are the dominant factors, and a four-variable choice was demonstrated to output satisfactory performance for predicting local bubble diameter and velocity. The model was preliminarily validated by coupling with the EMMS drag into CFD codes, which showed that the accuracy of coarse-grid simulations can be significantly improved.

show abstract

“…[ 6 ] In many cases, CFD may require the incorporation of additional mathematical models to achieve realistic system hydrodynamics. Upadhyay et al [ 7 ] applied artificial neural network (ANN) for predicting solid volume fraction along the riser height. It was developed by data sets from the experimental and computational biomass fast pyrolysis fluidized bed reactor.…”

Section: Introductionmentioning

confidence: 99%

CFD‐ANN coupling model simulation of gas–solid feeding design for ternary biomass mixtures in bubbling fluidized bed gasifier

Soanuch,

Pareek,

Piumsomboon

et al. 2024

Can J Chem Eng

View full text Add to dashboard Cite

The performance of continuous feeding fluidized bed reactors is significantly influenced by their design. These reactors can effectively operate with a wide range of biomass mixtures. Therefore, it is imperative to carefully design the gas distributor plate and solid tube inlet to ensure stable fluidization and uniform distribution of fluidizing gas and solid particles within the reactor. This study investigated the impact of gas–solid feeder design in bubbling fluidized bed gasifier for biomass mixtures on system hydrodynamics, employing a computational fluid dynamics–artificial neural network (CFD‐ANN) coupling model to achieve more realistic simulations. A 2k factorial experimental design was adopted to inquire the impact of gas and solid feeding systems. The responses under investigation included the gas–solid mixing index and the solid residence time, both of which hold pivotal roles in specific chemical processes related to biomass utilization in fluidized bed technology. All cases were successfully simulated, and the results uncovered that the position and length of the solid inlet tube wielded a significant influence on reactor performance, particularly concerning solid residence time. Furthermore, the designs of the gas distributor were identified as critical factors capable of enhancing system turbulence and mixing. In summary, the results showed the potential for enhancing reactor performance through the optimization of gas–solid feeding systems and underscored the efficacy of the ANN drag model in simulating continuous biomass gasification systems.

show abstract

Hybrid CFD-neural networks technique to predict circulating fluidized bed reactor riser hydrodynamics

Cited by 22 publications

References 45 publications

Neural network based prediction of the efficacy of ball milling to separate cable waste materials

Neural network based prediction of the efficacy of ball milling to separate cable waste materials

Machine Learning Assisted Characterization of Local Bubble Properties and Its Coupling with the EMMS Bubbling Drag

CFD‐ANN coupling model simulation of gas–solid feeding design for ternary biomass mixtures in bubbling fluidized bed gasifier

Contact Info

Product

Resources

About