Multiphase Computational Fluid Dynamics Simulations Research Articles

Capturing mesoscale structures in computational fluid dynamics simulations of gas‐solid flows

AbstractMultiphase reactors' performance depends on the mesoscale structures formed due to multiphase hydrodynamics. Examples of mesoscale structures include gas bubbles in a fluidized bed and particle clusters in a riser. Experimental investigation of these mesoscale structures is challenging and expensive. To this end, computational fluid dynamics (CFD) simulations are extensively employed; however, post‐processing CFD data to capture mesoscale structures is challenging. This article develops a density‐based spatial clustering of applications with noise (DBSCAN)‐based methodology to capture and characterize mesoscale structures from multiphase CFD simulation data. DBSCAN is an unsupervised machine‐learning algorithm, which requires the value of two hyperparameters. A simple technique to calculate these hyperparameters is provided and the performance of DBSCAN is assessed on CFD‐DEM simulations of bubbling fluidized beds and particle clustering. We demonstrate the computational complexity of DBSCAN to be , lower than the existing techniques, by testing its scalability on highly resolved grids (up to 100 million grid points).

Unsupervised machine learning of virus dispersion indoors

This paper concerns analyses of virus droplet dynamics resulting from coughing events within a confined environment using, as an example, a typical cruiser's cabin. It is of paramount importance to be able to comprehend and predict droplet dispersion patterns within enclosed spaces under varying conditions. Numerical simulations are expensive and difficult to perform in real-time situations. Unsupervised machine learning methods are proposed to study droplet dispersion patterns. Data from multi-phase computational fluid dynamics simulations of coughing events at different flow rates are utilized with an unsupervised learning algorithm to identify prevailing trends based on the distance traveled by the droplets and their sizes. The algorithm determines optimal clustering by introducing novel metrics such as the Clustering Dominance Index and Uncertainty. Our analysis revealed the existence of three distinct stages for droplet dispersion during a coughing event, irrespective of the underlying flow rates. An initial stage where all droplets disperse homogeneously, an intermediate stage where larger droplets overtake the smaller ones, and a final stage where the smaller droplets overtake the larger ones. This is the first time computational fluid dynamics is coupled with unsupervised learning to study particles' dispersion and understand their dynamic behavior.

The effects of ventilation conditions on mitigating airborne virus transmission

The effects of ventilation strategies on mitigating airborne virus transmission in a generic indoor space representative of a lobby area or information desk found in a hotel, company, or cruise ship are presented. Multiphase computational fluid dynamics simulations are employed in conjunction with evaporation modeling. Four different ventilation flow rates are examined based on the most updated post-COVID-19 pandemic standards from health organizations and recent findings from research studies. Three air changes per hour provide the best option for minimizing droplet spreading at reasonable energy efficiency. Thus, a higher ventilation rate is not the best solution to avoid spreading airborne diseases. Simultaneous coughing of all occupants revealed that contagious droplets could be trapped in regions of low airflow and on furniture, significantly prolonging their evaporation time. Multiphase flow simulations can help define ventilation standards to reduce droplet spreading and mitigate virus transmission while maintaining adequate ventilation with lower energy consumption. The present work significantly impacts how heat, air-conditioning, and ventilation systems are designed and implemented.

Effect of NaX Zeolite Mass and Baffle Presence on Copper Sorption

Sorption is often conducted in stirred batch reactors without assessing the impact of mixing on yield and costs. Operating the mixing in the batch reactor at a state of complete suspension, <i>i.e.</i>, at just suspended impeller speed, <i>N</i><sub>JS</sub>, is a compromise between these conflicting goals. Therefore, copper sorption on NaX zeolite using a pitched blade turbine was conducted in both baffled and unbaffled batch reactors at <i>N</i><sub>JS</sub>. Maintaining constant impeller to reactor diameter ratio (<i>D</i>/<i>d</i><sub>T</sub> = 0.32), and impeller off-bottom clearance (<i>C/H</i> = 0.33), kinetic sorption experiments were performed. Besides kinetic sorption experiments and the kinetic analysis of data obtained, to gain insight into the hydrodynamic behaviour of the system, transient multiphase computational fluid dynamics simulations (CFD) were performed.<br /> The <i>N</i><sub>JS</sub> speed is related to the particle settling speed, wherein a higher zeolite suspension mass concentration requires a greater NJS speed, resulting in increased energy consumption. The differences in <i>N</i><sub>JS</sub> speeds for the investigated zeolite suspension mass concentrations were found to be insubstantial, and therefore, the increase in mixing intensity for the tested systems was not deemed significant. Regardless of the zeolite mass used, the values of <i>N</i><sub>JS</sub> and <i>P</i><sub>JS</sub> (mixing power consumption) were consistently higher in the unbaffled reactor. The hydrodynamic conditions employed were found to have no significant effect on the maximum amount of copper sorbed or process efficiency, but was significantly affected by zeolite mass. The kinetics were affected by both, and, generally, sorption occurred faster in the unbaffled reactor.

Realistic fuel spray modeling for gasoline direct injection engine applications

Fuel spray modeling plays a critical role during modern gasoline direct injection (GDI) engine development due to fuel injection’s dominant impact on engine performance and emissions as well as the complex physical processes involved. In engineering three-dimensional (3D) computational fluid dynamics (CFD) simulations, the liquid-phase fuel atomization, evaporation, and mixing are usually modeled with the discrete droplet model (DDM) adopting a Lagrangian approach for multiphase CFD simulations. To this end, general practices heavily depend on the reduced order characterization of the injector nozzle flow. However, such simplified injector modeling may lead to insufficient representations of the complex spray dynamics. To tackle this problem, this study proposes a novel workflow to numerically evaluate GDI sub-cooled and flash-boiling sprays under engine-relevant conditions using a side-mounted GDI injector together with real gasoline fuel properties. The workflow introduces a one-way coupling (OWC) method leveraging high-fidelity nozzle flow simulations to provide realistic boundary conditions to the Lagrangian injector model. The proposed workflow was first verified in a constant volume chamber (CVC) environment and then implemented in a practical GDI engine setup to study spray morphology, fuel-air mixing, and wall-wetting propensity. In addition, detailed comparison was performed between the OWC method and the conventional rate of injection (ROI) routine. Quantitative analysis of spray characteristics was conducted to highlight possible source of discrepancies of the conventional ROI method.

Enhancing TiO2 Precipitation Process through the Utilization of Solution-Gas-Solid Multiphase CFD Simulation and Experiments

Ensuring uniform particle size distribution is a crucial role in the precipitation process of manufacturing white pigment. This study presents a comprehensive investigation that combines multiphase computational fluid dynamics (CFD) simulations with experimental research to effectively address the challenge of achieving uniform particle distribution during TiO2 precipitation. The objective of this study was to enhance three-phase CFD simulations involving the mixing process of TIOSO₄ solution, steam as a gas phase, and solid seed particles. By analyzing the trajectories of the seed particles using CFD, the optimal injection position for the seed particles within the mixing process was determined. Subsequently, a lab scale test and real field test were conducted based on the insights gained from the CFD simulations. The particle size distribution of two different types of seed inlets was analyzed and compared using Transmission Electron Microscopy (TEM) and Scanning Electron Microscope (SEM). The findings of this study demonstrate that the developed multiphase CFD simulation can effectively enhance the precipitation process for the production of anatase titanium dioxide particles. Additionally, using the developed multiphase CFD solver, the real physics involved in the precipitation process were identified, leading to a better understanding of the process itself. Furthermore, TiO2 particles with uniform particle size had a positive impact on the washing and bleaching processes following the precipitation process, resulting in a significant reduction in the annual defect production rate.

Optimizing the Disinfection Inactivation Efficiency in Wastewater Treatment: A Computational Fluid Dynamics Investigation of a Full‐Scale Ozonation Contactor

AbstractThe inactivation kinetics of total coliforms to increase the pathogenic removal efficiency in a full‐scale ozonation contactor in a wastewater treatment plant in Algeria is investigated. An enhanced ozone contactor design is proposed and 3D multiphase computational fluid dynamics simulations were conducted to optimize the operating parameters, including flow rates, ozone concentrations considering the treatment performance, and total operating cost. The existing design has several limitations, including poor mass‐transfer efficiency and uneven distribution of dissolved ozone. To address these issues, the optimized design includes new injection‐point settings and an increased number of diffusers. The optimized design achieved a significant improvement in mass transfer efficiency. The ozone treatment effectively reduced the total coliform counts in the wastewater samples compared to the existing design. The Chick‐Watson model predicted inactivation kinetics, with a reduction of up to 99.997 %. The practical implications of this research can significantly improve the inactivation kinetics of ozone treatments.

Uncertainty Quantification for Multiphase Computational Fluid Dynamics Closure Relations with a Physics-Informed Bayesian Approach

Multiphase Computational Fluid Dynamics (MCFD) based on the two-fluid model is considered a promising tool to model complex two-phase flow systems. MCFD simulation can predict local flow features without resolving interfacial information. As a result, the MCFD solver relies on closure relations to describe the interaction between the two phases. Those empirical or semi-mechanistic closure relations constitute a major source of uncertainty for MCFD predictions. In this paper, we leverage a physics-informed uncertainty quantification (UQ) approach to inversely quantify the closure relations’ model form uncertainty in a physically consistent manner. This proposed approach considers the model form uncertainty terms as stochastic fields that are additive to the closure relation outputs. Combining dimensionality reduction and Gaussian processes, the posterior distribution of the stochastic fields can be effectively quantified within the Bayesian framework with the support of experimental measurements. As this UQ approach is fully integrated into the MCFD solving process, the physical constraints of the system can be naturally preserved in the UQ results. In a case study of adiabatic bubbly flow, we demonstrate that this UQ approach can quantify the model form uncertainty of the MCFD interfacial force closure relations, thus effectively improving the simulation results with relatively sparse data support.

Two-phase stratified flow in horizontal pipes: A CFD study to improve prediction of pressure gradient and void fraction

Improved understanding of flow regime effects on design-influencing engineering quantities is of primary importance. This work is focused on the numerical prediction of pressure gradient and void fraction in a horizontal pipe where gas-liquid stratified flow is present in different operating conditions. The problem was modeled with unsteady, multiphase CFD (computational fluid dynamics) simulations. Volume Of Fluid (VOF) method was used as multiphase model. To define the numerical methodology, this study provides details on the influence of discretization grid and turbulence model on the simulation accuracy. It shows that mesh density on pipe cross-section is the most important grid parameter to focus on. Different turbulence models are required depending on the gas velocity and on its turbulence flow regime. Transition SST is able to model all the operating conditions but Realizable k-ε is adopted to further increase the accuracy of the results. A general underestimation of the pressure gradient is reported with an average error of − 6.43 % and − 16.21 % for a liquid superficial velocity of 0.04 m/s and 0.06 m/s respectively. Comparison with two-fluid 1D models shows that CFD simulations are the most accurate tools for predicting the pressure gradient at gas superficial velocities lower than 1.3 m/s. The implementation of a drift flux model shows a good agreement between experimental results and CFD simulations concerning void fraction estimation. CFD results are also used to underline the physical phenomena limiting the performance of 1D models.

High-pressure gas release from subsea pipelines: Multiphase modelling and CFD simulation for consequences analysis in risk assessment

The aim of this work is to assess the time evolution of natural gas release from high-pressure subsea pipelines (117 and 137 barg) to both assess the consequences of it on the sea surface and delimit the hazard areas for offshore structures, vessels, ships and workers. Such a hazard analysis is applied for the first time to risk assessment by a quantitative determination of the area of the different hazard zones on the sea surface as a function of the following parameters: i) water depth of release, ii) release flow rate, iii) sea current and iv) wind velocity. In this way, safety maps are developed from the obtained results based on different scenarios of accidental events. In detail, when a fire scenario is developed, the hazard area is delimited by the inflammable limits of natural gas in air. To perform this investigation, multiphase computational fluid dynamic simulations involving three phases – i.e., released gas, seawater and air – are carried out using the open-source software OpenFOAM®. The methodology introduced in this work goes beyond the current state-of-the-art simulations, as it takes into account the mutual interactions among all the three phases present in the system, for which there is no analytical model able to describe the behaviour of the three phases with a certain degree of precision. In particular, a multiphase hybrid model pairing Euler-Euler and VOF methods is implemented. As main results, the shape of the release gas subsea cone and superficial plume are precisely evaluated as a function of the above-mentioned operating conditions, determining the evolution of the gas dispersion also in the seawater, which is crucial in the description of accidental scenarios resulting from the consequences analysis. In particular, it is found that, at a water depth of release of 55 m, the cone diameter at the sea level is larger than that at 15 m due to dispersion of gas in the sea. On the other side, wind velocity does not affect significantly the cone diameter.

Multiphase CFD simulation and experimental investigation of friction stir welded high strength shipbuilding steel and aluminum alloy

This study executed the volume of fluid (VOF) based multiphase computational fluid dynamics (CFD) simulation by incorporating a modified analytical model for dissimilar friction stir welding of DH36 shipbuilding steel and 6061-T6 aluminum alloy. The impact of rotational speed on the temperature and plasticized material flow associated with the tool–material interaction was investigated. The simulation results revealed that the variation in rotational speed was significantly affected the temperature and material flow around the FSW tool. The maximum velocity and strain rate were obtained at the outer edge of the shoulder due to the lowest dynamic viscosity on the Al6061 side. The velocity vectors revealed that the material flow pattern was discontinuous in the advancing side at low heat input, whereas it was improved and became uniform at higher heat input conditions. Transmission electron microscopy (TEM) analysis revealed the formation of a dual-phased intermetallic compound (IMC) layer of FeAl 3 and Fe 2 Al 5 at the interface. Its thickness increased with an increase in rotational speed due to the enhanced intermetallic reaction at a higher temperature and strain rate. The rotational speed of 875 rpm produced the maximum joint strength~207.4 MPa at the IMC layer thickness of 4.83 ± 0.65 µm. The grain structure was refined under dynamic recrystallization, and the hardness improved significantly as the heat generation and strain rate increased. The intercalated structures acted as the hardest zone because of extensive IMCs formation, i.e., Fe 3 Al at 450 rpm and 600 rpm, and Fe 3 Al + FeAl at 875 rpm and 1200 rpm.

Computational fluid dynamic simulations of granular flows: Insights on the flow-wall interaction dynamics

Dry volcanic granular flows are gravity-driven currents composed of solid particles where particle-particle interactions dominate the motion. The interaction with topography is a relevant factor controlling the propagation of such flows. In this paper we investigate the dynamics of channelised volcanic granular flows by comparing large-scale experiments with multiphase computational fluid dynamic simulations using the Two-Fluid Model approach, with an emphasis on the dynamics regulating the flow-wall interactions. We use the software MFIX to carry out sensitivity analysis of the boundary conditions for the solid phase implemented in the numerical code. The sensitivity analysis shows how the choice of the boundary condition and of the relevant parameters controlling the boundary conditions highly affect the dynamics of the whole flow. Finally, a preliminary benchmark of the MFIX boundary conditions with one large-scale experiment is presented, showing good agreement between the simulated and experimental flow front velocities.

Three-dimensional multi-phase simulation of different flow fields with cooling channel in proton exchange membrane fuel cell

The overall performance of PEMFC (proton exchange membrane fuel cell) is affected by the flow field structure, especially the cathode flow field design can effectively solve the uneven distribution of gas concentration in the traditional flow channel and the cathode flooding phenomenon. In order to solve the above problems, a PEMFC single cell model with waveform staggered flow field of cooling flow field and small cathode channel was established in this study. Three-dimensional (3D) multi-phase CFD (computational fluid dynamics) simulation method is used to compare with gas concentration, liquid water distribution, pressure drop, and net power density of three different cases, and the influence of different cooling velocity on the temperature of cooling flow field is considered. The results show that the overall performance of the proposed flow field is the best, in which the maximum current density is 1.391 A⋅cm−2 and increases by 14.9%. The cathode and anode waveform staggered flow field makes the proton exchange membrane (PEM) water distribution more uniform, at the same time, the small size of the cathode flow channel facilitates the discharge of heat, and the convective heat transfer effect is enhanced. The electrochemical reaction rate is fast, which accelerates the temperature reduction in the fuel cell under the action of the cooling flow field, and the temperature uniformity of the cooling flow field is better. In addition, net power density is improved by 39.7%, and the output performance is significantly improved.

Effects of the filter microstructure and ambient air condition on the aerodynamic dispersion of sneezing droplets: A multiscale and multiphysics simulation study.

Concerns have been ramping up with regard to the propagation of infectious droplets due to the recent COVID-19 pandemic. The effects of filter microstructures and ambient air flows on droplet dispersion by sneezing are investigated by a fully coupled Eulerian–Lagrangian computational modeling with a micro-to-macroscale bridging approach. Materials that are commonly applied to face masks are modeled to generate two different virtual masks with various levels of filtration efficiency, and the leakage percentages through the unsealed nose and cheek areas were set to 11% and 25%, respectively. The droplet propagation distance was simulated with and without mask wearing in still and windy conditions involving head wind, tail wind, and side wind. The results demonstrate that wearing a face mask reduces the transmittance distance of droplets by about 90%–95% depending on the mask type; nonetheless, the droplets can be transmitted to distances of 20–25 cm in the forward direction even with mask-wearing. Thus, a social distance of at least 20 cm between people would help to prevent them from becoming exposed to ejected droplets. This study is significant in that important aspects of mask materials, in this case the porous microstructure-dependent filtration efficiency and permeability under varied ambient flow conditions, were considered for the first time in an evaluation of the barrier performance against droplet transmittance through a multiphase computational fluid dynamics simulation of air-droplet interaction and turbulence flow dynamics.

Numerical methodology and CFD simulations of a rotary vane energy recovery device for seawater reverse osmosis desalination systems

• Novel analytical grid generation methodology for rotary vane machines. • Removal of backflows and torque peaks through flow port optimisation. • Sensitivity analysis to reduce cavitation and improve volumetric efficiency. • Volumetric efficiency with optimal speed and blade design equal to 91.6%. Energy recovery devices in Seawater Reverse Osmosis Systems (SWRO) reduce energy consumption and may facilitate the large-scale deployment of desalination systems. In this paper, a Rotary Vane Energy Recovery Device (RVERD) is analysed and optimised by aiming at weakening cavitation and improving the volumetric performance of the machine. An innovative analytical methodology based on user defined nodal displacement is proposed to address the need to discretise the rotating and deforming computational domain of double-acting vane machines. The generated grids are interfaced with the ANSYS FLUENT solver for multi-phase computational fluid dynamics simulations. The flow topology is analysed to reveal the flow and cavitation features especially in the blade tip regions. A port optimisation is then carried out followed by a sensitivity analysis on the design parameters to improve RVERD performance. The results show that delaying the discharge angle at the high-pressure outlet port by 3° and an optimal port to stator length ratio of 70% helped to prevent backflows and eliminate torque peaks. The sensitivity analysis has identified the rotational speed and the blade tip clearance as the two most influential factors affecting cavitation and, in turn, the volumetric efficiency of the machine. With respect to the baseline design configuration, at the optimal rotational speed of 1000 RPM and with a tip clearance gap of 50 μm, the volume-averaged vapour volume fraction in the core decreased from 20.6 × 10 −3 to 0.6 × 10 −3 while the volumetric efficiency increased from 85.7% to 91.6%. The axial clearance gap of 70 μm contributed to 2.9% of the volumetric losses.

Multi-objective optimization of operating conditions of an enhanced parallel flow field proton exchange membrane fuel cell

Proton Exchange Membrane Fuel Cell (PEMFC) is a high-efficiency energy conversion device. Operating conditions play a vital role to obtain maximum power. In the present work, a modified parallel PEMFC is modeled by 3D multiphase computational fluid dynamics simulation. Water management and reactants transport are studied in solitary changes of temperature, pressure, mass flow rate, and humidity ratio to get an interpretation in the PEMFC operation. Mutual effects of operating parameters including cathode stoichiometry, anode stoichiometry, temperature, pressure, cathode relative humidity, and anode relative humidity are studied to obtain an optimal power density in all operating voltages. Response Surface Method (RSM) and Non-dominated Sorting Genetic Algorithm ii (NSGA ii) were combined to consummate a multi-objective optimization. Technique of Order Preference Similarity to the Ideal Solution (TOPSIS) was utilized to select the ideal case from the Pareto front. The voltage and amount of maximum power density of each operating condition are selected as objects of the optimization. The validated results show each operating parameter can play a positive or negative role in the performance due to the other operating conditions. The increase of temperature and pressure can have a beneficial effect only with appropriate stoichiometries of fuel and air and adequate relative humidities. High stoichiometries of cathode and anode result in the high amount of the maximum power density and contrary, the low range of stoichiometries leads to the high voltage of the maximum power density.

Experiment and multiphase CFD simulation of gas-solid flow in a CFB reactor at various operating conditions: Assessing the performance of 2D and 3D simulations

Accurate prediction of gas-solid flow hydrodynamics is key for the design, optimization, and scale-up of a circulating fluidized bed (CFB) reactor. Computational fluid dynamics (CFD) simulation with two-dimensional (2D) domain has been routinely used, considering the computational costs involved in three-dimensional (3D) simulations. This work evaluated the prediction capability of 2D and 3D gas-solid flow simulation in the lab-scale CFB riser section. The difference between 2D and 3D CFD simulation predictions was assessed and discussed in detail, considering several flow variables (superficial gas velocity, solid circulation rate, and secondary air injection). The transient Eulerian-Eulerian multiphase model was used. CFD simulation results were validated through an in-house experiment. The comparison between the experimental data and both computational domains shows that the 3D simulation can accurately predict the axial solid holdup profile. The CFD simulation comparison considering several flow conditions clearly indicated the limitation of the 2D simulation to accurately predict key hydrodynamic features, such as high solid holdup near the riser exit and riser bottom dense region. The accuracy of 2D and 3D simulations was further assessed using root-mean-square error calculation. Results indicated that the 3D simulation predicts flow behavior with higher accuracy than the 2D simulation.

CFD simulation of the multiphase heat transfer during the quenching process

The paper presents the results of the CFD (Computational Fluid Dynamics) simulation of quenching process for the ring shape sample. The aim of the work is to develop and validate the methodology for multiphase CFD simulation including the boiling during the quenching process. CFD simulation is provided in ANSYS Fluent. The Lee model is used for modelling the phase change during the quenching process. The first step consider the simulation of cooling of the ring sample when the correct model parameters will be found. Validation of results is performed by comparison with experimental data. Experimental was realized inside the own designed quenching bath device filled with quenching polymer. The general description of the experimental setup is included in the paper. The CFD results are cooling curves, i.e. variation of solid temperature on time. The Lee model parameters especially the evaporation frequency was tuned. Thin polymer film on the solid surface was considered to bring the results closer to the experimental data. The comparison between experiment and CFD shows very good agreement for higher temperatures, which covers the boiling stage. On the other hand for lower temperatures worse match of results was found caused probably by the sensitivity on the inlet velocity profile settings. Some recommendations for future work were defined.

Single and multiphase CFD simulations for designing cavitating venturi

Hydrodynamic cavitation (HC) is becoming a popular process intensification tool for many industrial processes. It has its applications in emulsification, advanced oxidation processes, nano-material synthesis etc. Venturi is one of the most commonly used device to generate hydrodynamic cavitation at laboratory as well as industrial scale. Recently researchers have taken interest to find the effect of the geometric design parameters of the venturi on cavitating flows.In the present work, cavitating flow in a venturi is simulated using Computational Fluid Dynamics (CFD). Optimum mesh size and suitable turbulence model for this such a flow were selected using a preliminary study. Single phase as well as multiphase simulations were carried out and the results were validated using experimental data of pressure drop and flow rate.Further, discrete phase model (DPM) was used to predict the possible path taken by multiple cavities, and the pressure and turbulence data along these paths was extracted to solve the Keller–Miksis equation for cavity dynamics simulations. Parameters like bubble radius, collapse pressure and collapse temperature of the cavity during its dynamic behavior were estimated from this model. A comparison between the results for single phase and multiphase models is made to propose a numerically inexpensive and reasonably accurate method to model cavitating flows in a venturi.

Modeling Average Pressure and Volume Fraction of a Fluidized Bed Using Data-Driven Smart Proxy

Simulations can reduce the time and cost to develop and deploy advanced technologies and enable their rapid scale-up for fossil fuel-based energy systems. However, to ensure their usefulness in practice, the credibility of the simulations needs to be established with uncertainty quantification (UQ) methods. The National Energy Technology Laboratory (NETL) has been applying non-intrusive UQ methodologies to categorize and quantify uncertainties in computational fluid dynamics (CFD) simulations of gas-solid multiphase flows. To reduce the computational cost associated with gas-solid flow simulations required for UQ analysis, techniques commonly used in the area of artificial intelligence (AI) and data mining are used to construct smart proxy models, which can reduce the computational cost of conducting large numbers of multiphase CFD simulations. The feasibility of using AI and machine learning to construct a smart proxy for a gas-solid multiphase flow has been investigated by looking at the flow and particle behavior in a non-reacting rectangular fluidized bed. The NETL’s in house multiphase solver, Multiphase Flow with Interphase eXchanges (MFiX), was used to generate simulation data for the rectangular fluidized bed. The artificial neural network (ANN) was used to construct a CFD smart proxy, which is able to reproduce the CFD results with reasonable error (about 10%). Several blind cases were used to validate this technology. The results show a good agreement with CFD runs while the approach is less computationally expensive. The developed model can be used to generate the time averaged results of any given fluidized bed with the same geometry with different inlet velocity in couple of minutes.