Detached eddy simulation on the turbulent flow in a stirred tank

A detached eddy simulation (DES), a large‐eddy simulation (LES), and a k‐ε‐based Reynolds averaged Navier‐Stokes (RANS) calculation on the single phase turbulent flow in a fully baffled stirred tank, agitated by a Rushton turbine is presented. The DES used here is based on the Spalart‐Allmaras turbulence model solved on a grid containing about a million control volumes. The standard k‐ε and LES were considered here for comparison purposes. Predictions of the impeller‐angle‐resolved and time‐averaged turbulent flow have been evaluated and compared with data from laser doppler anemometry measurements. The effects of the turbulence model on the predictions of the mean velocity components and the turbulent kinetic energy are most pronounced in the (highly anisotropic) trailing vortex core region, with specifically DES performing well. The LES—that was performed on the same grid as the DES—appears to lack resolution in the boundary layers on the surface of the impeller. The findings suggest that DES provides a more accurate prediction of the features of the turbulent flows in a stirred tank compared with RANS‐based models and at the same time alleviates resolution requirements of LES close to walls. © 2011 American Institute of Chemical Engineers AIChE J, 58: 3224–3241, 2012     

Stereo‐PIV experiments and large eddy simulations of flow fields in stirred tanks with Rushton and curved‐Blade turbines

The flow characteristics in pilot‐scale stirred tanks with Rushton and curved‐blade turbines were investigated by using stereoscopic particle image velocimetry (SPIV) experiments and large eddy simulation (LES) methods. The velocity and turbulent kinetic energy (TKE) in the impeller discharge regions were carefully resolved with a high resolution SPIV system, and the detailed phase‐resolved velocity and TKE profiles were used to validate the LES results. The effects of Reynolds number and blade shape on the flow characteristics were discussed. The LES results of velocity, TKE, and the evolution of trailing vortices were compared with the SPIV experimental data, and good agreement was obtained at various phase angles. The effects of subgrid scale model and hybrid grid with different mesh resolutions on the LES results were investigated. LES is a computationally affordable method for the accurate predictions of the complex flow fields in pilot‐scale stirred tanks is presented. © 2013 American Institute of Chemical Engineers AIChE J, 59: 3986–4003, 2013     

Numerical simulation of turbulent flow in a baffled stirred tank with an explicit algebraic stress model

An explicit algebraic stress model (EASM) was used to simulate anisotropic turbulent flows in baffled stirred tanks equipped with a standard Rushton turbine. The quantitative predictions of velocity components, turbulence kinetic energy, Reynolds stresses and turbulence energy dissipation rate in the context of anisotropic turbulence were conducted to assess the comprehensive performance of the EASM. A lot of efforts have been made to ensure numerical stability during the calculations such as using a good initial flow field, manipulating source terms and adjusting under-relaxation factors. The predicted results were also compared with experimental data and other simulation results obtained using the standard k–ε model, algebraic stress model (ASM), Reynolds stress model (RSM) and large eddy simulation (LES). All the simulations were run with in-house codes. The simulation results show that agreement between the EASM predictions and experimental values is satisfactory. The EASM is consistently superior to the standard k–ε model when predicting both peak values and trend of variation in velocities and turbulence quantities. In comparison to the RSM, the EASM has almost the same predictive accuracy. The EASM is inferior to the LES on the prediction of turbulence kinetic energy. Nevertheless, the computational cost of the EASM is significantly lower than that of the LES, which is an obvious advantage in practical applications.     

Dynamic delayed detached eddy simulation of a multi‐inlet vortex reactor

The multi‐inlet vortex reactor (MIVR) is used for flash nanoprecipitation to manufacture functional nanoparticles. A validated computational fluid dynamics model is needed for the design, scale‐up, and optimization of the MIVR. Unfortunately, available Reynolds‐averaged Navier‐Stokes methods are unable to accurately model the highly swirling flow in the MIVR. Large‐eddy simulations (LES) are also problematic, as excessively fine grids are required to accurately model this flow. These dilemmas led to the application of the dynamic delayed detached eddy simulation (DDES) method to the MIVR. In the dynamic DDES model, the eddy viscosity has a form similar to the Smagorinsky sub‐grid viscosity in LES, which allows the implementation of a dynamic procedure to determine its model coefficient. Simulation results using the dynamic DDES model are found to match well with experimental data in terms of mean velocity and turbulence intensity, suggesting that the dynamic DDES model is a good option for modeling the turbulent swirling flow in the MIVR. © 2016 American Institute of Chemical Engineers AIChE J, 62: 2570–2578, 2016     

Numerical and experimental evaluation of heat transfer in helically corrugated tubes

The enhancement of convective heat transfer in single‐phase heat transfer through the use of helicoidally corrugated tubes has been studied numerically. By comparing the large eddy simulation (LES) results with detailed Stereo‐PIV and Liquid Crystal Thermography measurements obtained at the von Karman Institute for Fluid Dynamics (VKI), a validated numerical framework was obtained. Heat transfer enhancements of 83–119% were seen, at the cost of pressure losses that were approximately 5.6 to 6.7 times higher than for a bare tube. To extrapolate the results to industrial Reynolds numbers at which experimental data is scarce, the simulation data was used to develop an improved near‐wall Reynolds stress transport model (RSTM) that more accurately describes the heat flux vector. Comparison of both global and local flow characteristics at different Reynolds numbers confirms that the approach allows more accurate predictions over a wider range of design and operating parameters than using two‐equation turbulence models, while the computational cost is still significantly lower than LES. © 2017 American Institute of Chemical Engineers AIChE J, 64: 1702–1713, 2018     

A numerical comparison of precipitating turbulent flows between large‐eddy simulation and one‐dimensional turbulence

This study presents the results of computational fluid dynamics simulations of a multiphase, reacting, turbulent mixing layer in an idealized geometry. The purpose is to compare large‐eddy simulation (LES) to one‐dimensional turbulence (ODT) and examine the trends of the flow under differing mixing conditions. Aqueous streams are mixed together to precipitate polymorphs of calcium carbonate. The polymorphs of calcium carbonate are tracked numerically using population balance equations (PBE). Each PBE contains all of the relevant physical models to track the particle evolution including nucleation, growth, and aggregation. A simple subgrid mixing model that is convenient for use with PBEs was implemented in the LES code. The higher spatial resolution achievable with ODT allowed an investigation on the effect of resolution on the mixing‐model error. The Reynolds number of the flow is varied and is shown to cause a decrease in average particle sizes with higher mixing rates. © 2015 American Institute of Chemical Engineers AIChE J, 61: 3185–3197, 2015     

CFD simulation of stirred tanks: Comparison of turbulence models. Part I: Radial flow impellers

A critical review of the published literature regarding the computational fluid dynamics (CFD) modelling of single‐phase turbulent flow in stirred tank reactors is presented. In this part of review, CFD simulations of radial flow impellers (mainly disc turbine (DT)) in a fully baffled vessel operating in a turbulent regime have been presented. Simulated results obtained with different impeller modelling approaches (impeller boundary condition, multiple reference frame, computational snap shot and the sliding mesh approaches) and different turbulence models (standard k ? ε model, RNG k ? ε model, the Reynolds stress model (RSM) and large eddy simulation) have been compared with the in‐house laser Doppler anemometry (LDA) experimental data. In addition, recently proposed modifications to the standard k ? ε models were also evaluated. The model predictions (of all the mean velocities, turbulent kinetic energy and its dissipation rate) have been compared with the experimental measurements at various locations in the tank. A discussion is presented to highlight strengths and weaknesses of currently used CFD models. A preliminary analysis of sensitivity of modelling assumptions in the k ? ε models and RSM has been carried out using LES database. The quantitative comparison of exact and modelled turbulence production, transport and dissipation terms has highlighted the reasons behind the partial success of various modifications of standard k ? ε model as well as RSM. The volume integral of predicted energy dissipation rate is compared with the energy input rate. Based on these results, suggestions have been made for the future work in this area.     

Large Eddy simulation for lean premixed combustion

In comparison to previous numerical studies interested in the ORACLES benchmark (One Rig for Accurate Comparisons with Large Eddy Simulations), the present study demonstrates the advantages of LES‐WALE model in both inert and reacting flows using the Fluent‐CFD. So, the confirmation is based on the experimental research effort that was involved in the European Union‐funded research program MOLECULES (Modelling of Low Emissions Combustors Using Large Eddy Simulations), for three parameters: longitudinal velocity, longitudinal velocity fluctuation, and length of recirculation zone. In line with what was observed by the experimental reference study, the dynamic model (LES‐WALE) predicts, respectively, as well as the asymmetry and the symmetry, for both inert and reacting flows. In addition, the simulation succeeds to predict the zones of recirculation and shows the differences between the two cases, inert and reacting flows. Moreover, results have been compared with those of the k–ε model performed by Kurenkov and Obserlack [Kurenkov and Obserlack, Flow Turbulence Combustion 74, 387–407 (2005)] study. © 2012 Canadian Society for Chemical Engineering     

Simulation of dry powder inhalers: Combining micro‐scale,meso‐scale and macro‐scale modeling

The flow of carrier particles, coated with active drug particles, is studied in a prototype dry powder inhaler. A novel, multiscale approach consisting of a discrete element model (DEM) to describe the particles coupled with a dynamic large eddy simulation (LES) model to describe the dynamic nature of the flow is applied. The model consists of three different scales: the micro‐scale, the meso‐scale, and the macro‐scale. At the micro‐scale, the interactions of the small active drug particles with larger carrier particles, with the wall, with the air flow, and with each other is thoroughly studied using discrete element modeling and detailed computational fluid dynamics (CFD), i.e., resolving the flow structures around the particles. This has led to the development of coarse‐grained models, describing the interaction of the small active drug particles at the larger scales. At the meso‐scale the larger carrier particles, and all of their interactions are modeled individually using DEM and CFD‐LES. Collisions are modeled using a visco‐elastic model to describe the local deformation at each point of particle‐particle contact in conjunction with a model to account for cohesion. At the macro‐scale, simulations of a complete prototype inhaler are carried out. By combining the relevant information of each of the scales, simulations of the inhalation of one dose from a prototype inhaler using a patient relevant air flow profile show that fines leave the inhaler faster than the carrier particles. The results also show that collisions are not important for particle‐particle momentum exchange initially but become more important as the particles accelerate. It is shown that for the studied prototype inhaler the total release efficiency of the fine particles is between 10 and 30%, depending on the Hamaker constant, using typical settings for the properties of both particles. The results are also used to study regions of recirculation, where carrier particles can become trapped, and regions where fines adhere to the wall of the device. © 2016 American Institute of Chemical Engineers AIChE J, 63: 501–516, 2017     

LES–Lagrangian‐particles‐simulation of turbulent reactive flows at high Sc number using approximate deconvolution model

Large eddy simulation (LES) with the approximate deconvolution model is combined with Lagrangian particles simulation (LPS) for simulating turbulent reactive flows at high Schmidt numbers Sc. The LES is used to simulate velocity and nonreactive scalar while reactive scalars are simulated by the LPS using the mixing volume model for molecular diffusion. The LES–LPS is applied to turbulent scalar mixing layers with a second‐order isothermal irreversible reaction at Sc = 600. The mixing volume model is implemented with the IEM, Curl's, and modified Curl's mixing schemes. The mixing volume model provides a correct decay rate of nonreactive scalar variance at high Sc independently of the number of particles. The statistics in the LES–LPS with the IEM or modified Curl's mixing scheme agree well with the experiments for both moderately‐fast and rapid reactions. However, the LPS with the Curl's mixing scheme overpredicts the effects of the rapid reaction. © 2016 American Institute of Chemical Engineers AIChE J, 62: 2912–2922, 2016     

Assessment of large eddy and RANS stirred tank simulations by means of LDA

Large eddy simulations (LES) and Reynolds-averaged Navier-Stokes (RANS) calculations were performed on the flow in a baffled stirred tank, driven by a Rushton turbine at Re=7300. The LES methodology provides detailed flow information as velocity fluctuations are resolved down to the scale of the numerical grid. The Smagorinsky and Voke subgrid-scale models used in the LES were embedded in a numerical lattice-Boltzmann scheme for discretizing the Navier-Stokes equations, and an adaptive force-field technique was used for modeling the geometry. The uniform, cubic computational grid had a size of 240³ grid nodes. The RANS calculations were performed using the computational fluid dynamics code CFX 5.5.1. A transient sliding mesh procedure was applied in combination with the shear-stress-transport (SST) turbulence closure model. The mesh used for the RANS calculation consisted of 241464 nodes and 228096 elements (hexahedrons). Phase-averaged and phase-resolved flow field data, as well as turbulence characteristics, based on the LES and RANS results, are compared both mutually and with a single set of experimental data.     

Robust on‐line measurement of conversion and molecular weight using NIR spectroscopy during solution polymerization

Fiber‐optic near‐infrared (NIR) spectroscopy was used to monitor the monomer conversion and the weight‐average molecular weight of the polymer produced during solution polymerization of methyl methacrylate (MMA) carried out in a lab‐scale reactor. NIR spectra were recorded during batch and semi‐continuous reactions using an in situ transmission probe. Off‐line gravimetry and GPC were used as reference methods to provide the conversion and the average molecular weight data set required for the calibration procedure. A statistical model was generated using partial least‐squares regression (PLS) to relate the NIR spectral data to the two polymerization variables of interest. The measurements were then validated for various operating conditions (i.e., different solvent, initiator, MMA, and chain‐transfer agent concentrations) and for both batch and semi‐continuous modes. The conversion was predicted during three validation experiments with an average standard error of prediction (SEP) of 2.1%. The on‐line evaluation of M?_w was obtained with an average relative SEP of 5.7%; such on‐line NIR measurement was thus demonstrated to be robust and accurate, even in the case of versatile use of the polymerization plant. © 2002 Wiley Periodicals, Inc. J Appl Polym Sci 85: 2510–2520, 2002     

Large eddy simulation of dense two-phase flows: Comment on DEM-LES study of 3-D bubbling fluidized bed with immersed tubes

In a recent paper, Gui et al. (2008. DEM-LES study of 3-D bubbling fluidized bed with immersed tubes. Chemical Engineering Science 63, 3654–3663) reported the results of numerical simulations of a bubbling fluidized bed containing internal tubes using discrete element model (DEM) coupled to a large eddy simulation (LES). We comment on the correctness of the numerical set-up regarding the use of LES as well as the worthiness of using a turbulence model for the case of dense two-phase flows.     

Computational Investigation of an Industrial Cyclone Separator with Helical‐Roof Inlet

An industrial cyclone separator with helical‐roof inlet TsN‐11 has been numerically investigated as to pressure and flow field, pressure drop, fractional efficiency, and particle trajectories inside the cyclone. The turbulence was modeled with Reynolds stresses and large eddy simulations (LES) based on three different subgrid‐scales (SGS). The results with the different setups were compared to experimental data from previous studies. For a proper calculation of the flow field, LES combined with a dynamic SGS model was used for predicting cyclone performance. Individual particles were tracked through the unsteady flow field using the Lagrangian approach. The results of the numerical calculations of the tangential and axial velocity, pressure drop, and cut size are in good agreement with experimental measurements.     

Study on flow fields in a bioreactor with perforated blades

Flow fields of a gas–liquid fluid in a bioreactor with perforated blades were evaluated with the computational fluid dynamics and particle image velocimetry (PIV). The flow pattern of the fluid was the complex turbulence, and the fluid had preferable flow fields for fermentations. Using a rotating coordinate system and a sliding mesh method, a mesh of 24 × 16 × 80 (X × Y × Z) was selected due to its high accuracy and appropriate workload. The numerical results of the two‐phase flow model agreed well with the experimental results of the PIV, such as flow field characteristics and data curves. The RNG k–ε model was adopted, which could precisely simulate anisotropic flow fields around higher‐speed blades. Whole deviations of X‐, Y‐ and Z‐direction velocities ranged only from 5.4% to 10.5%, and continuous turbulence models could effectively simulate transfer procedures of the fluid. Because of its low cost, simplicity and easy of scale‐up, the bioreactor with perforated blades and its models mechanism of fluid microelements are expected to be efficient for its industrial amplification and application in food industry. © 2012 Canadian Society for Chemical Engineering     

双层涡轮桨搅拌反应器内混合时间的大涡模拟

采用计算流体力学(CFD)方法对直径为0.476m双层涡轮桨搅拌反应器内的流动及混合进行了数值模拟,并实验测试了混合过程。利用大涡模拟(LES)及Smagorinsky-Lilly亚格子模型求解湍流流动与示踪剂传递过程,桨叶区域采用滑移网格技术。研究结果表明,大涡模拟得到的示踪剂响应曲线和混合时间与实验结果吻合良好,其预测精度明显优于基于雷诺平均(Reynolds-averaged Navier-Stokes,RANS)的标准k-ε模型的模拟结果。大涡模拟是研究搅拌反应器内非稳态及周期性湍流流动的有效方法。     

Large Eddy Simulation of mixing effects on the course of parallel chemical reactions and comparison with k-? modeling

The way in which reagents are mixed can have a large influence on the product distribution of chemical reactions. To model effects of mixing on various scales on the course of chemical reactions the method of Large Eddy Simulation (LES) of non-premixed, turbulent reactive flows of incompressible fluids is considered in this work. The subgrid modeling of chemical reaction is based on a beta distribution of the mixture fraction in combination with a conditional moment closure based on linear interpolation of local instantaneous reactant concentration values. The predictions obtained with LES are compared with experimental data for fast parallel chemical reactions, the fluid velocity measured using Particle Image Velocity (PIV) technique and the passive tracer concentration measured using the Planar Laser Induced Fluorescence (PLIF) technique. Predictions of the model based on LES are compared as well with results obtained using the non-equilibrium multiple-time-scale mixing model combined with a standard k-? model and employing similar conditional moment closure as LES, applied, however, at larger scale. All comparisons show a very good performance of the model based on LES.     

PIV experiments and large eddy simulations of single-loop flow fields in Rushton turbine stirred tanks

The single-loop flow fields in Rushton turbine stirred tanks with clearance C=0.15T (T is tank diameter) were investigated by using particle image velocimetry (PIV) experiments and large eddy simulation (LES) methods. The velocity and turbulent kinetic energy (TKE) were carefully measured and resolved with high resolution camera. The regions with high TKE are affected by the movement of the trailing vortices generated behind the impeller blades. The effects of both geometrical configuration and Reynolds number were discussed. It is found that the Reynolds number has little effect on the mean flow for the configuration of impeller diameter D=T/3, C=0.15T. However, the single-loop flow pattern is changed into a double-loop one if D is increased from T/3 to T/2. The LES results were compared with the PIV experiments and the laser Doppler anemometry (LDA) data in the literature. The effect of the grid was validated, and the levels of local anisotropy of turbulence near the impeller discharge regions were investigated. Both the phase-averaged and phase-resolved LES results are in good agreement with the PIV experimental data, and are better than the predictions of the k–ε model. The agreement shows that the LES method can be used to simulate the complex flow fields in stirred tanks.