CFD simulation of gas–liquid flow in a hig(6)

[33]G.Y.Yang,K.Y.Guo,T.F.Wang,Numerical simulation of the bubble column at elevated pressure with a CFD-PBM coupled model,(2017).

[34],,Pressure and gas density effects on bubble breakup and gas hold-up in bubble columns,(8)(1990)2309–2315.

[35],,,A.Jain,,A correlation for gas holdup in turbulent coalescing bubble columns,(1986)705–717.

.

E-mail address:(H.Jin).

Bubble column reactor has simple structure,large capacity,easy operation,adequate heat and mass transfer,and small bed pressure drop[1–3].Therefore,bubble columns are widely used in industry including chemical engineering,petrochemical,bio-engineering,environmental energy etc.[4].Many scholars have applied the population balance model in studying atmospheric bubble columns[5–7].But bubble columns in chemical production are generally operated under high pressures and examples are hydrocracking of petroleum(P=5.0–21 MPa),Fischer-Tropsch synthesis(P=2.0–5.0 MPa)and benzene hydrogenation(P=5.0 MPa)[8–11].Although high-pressure bubble columns are widely used in chemical and biochemical processes,their fundamental hydrodynamic behaviors,which are essential for reactor scale-up and design,are still not fully effect of pressure on the hydrodynamic behaviors of bubble columns has been experimentally investigated by many gas holdup in high-pressure columns significantly increases due to the decreased bubble size[12–14].The gas–liquid mass transfer and reaction performance are enhanced as the pressure rises[10,15].With the development of computer technology,numerical simulation of gas–liquid two-phase flow has been greatly them,Krishna et al.[16]used a CFD model to simulate the high-pressure bubble column with the drag force between gas and liquid was considered only,and a density correction term ρ/ρ0due to pressure change was introduced into the drag force et al.[17]modified the gas density correction term in the drag model based on[16].Although the radial and axial velocity components were better predicted,the bubble diameter distribution was assumed the population balance model(PBM)can resolve the influence of bubble coalescence and breakup on bubble size distribution,the simulation of high-pressure bubble columns has been intensively conducted using the CFD-PBM coupled model in recent et al.[18]imposed the energy and capillary constraints in the bubble breakup model,and got a modified PBM to express the effect of bubble size distribution was then reasonably predicted by the modified PBM.Xing et al.[19]proposed a unified breakup model for both bubbles and droplets with the effect of pressure this unified breakup model gave good predictions of both the effect of pressure(or gas density)on the bubble breakup rate and the different daughter size distributions of bubbles and works have been reported on the effect of pressure in bubble columns,but the mechanism of pressure effect was little the influence of pressure on the hydrodynamics in bubble columns is pronounced and very important for the design and scale-up of reactors at high pressures,further studies need to be conducted on the effects of this paper,based on the Luo bubble coalescence kernel model,a correction coefficient Ceabout density ρ/ρ0is introduced in the bubble coalescence efficiency item.The modified CFDPBM coupled model is used to simulate the flow field in a high-pressure bubble effects of pressure on the gas–liquid two-phase flow in the high-pressure bubble column were investigated at 0.5–2.0 MPa.It is shown that the modified CFD-PBM coupled model can describe the effect of pressure on the hydrodynamic parameters in the high-pressure bubble column. fluid modelIn the present work,the main approach for simulating gas–liquid flow s in a bubble column is Euler–Euler model.In contrast with the Euler–Lagrange approach,the gas phase and the liquid phase in the bubble column were considered as continuous phases of mutual penetration with the Euler–Euler approach gives a possibility of lower computational cost and particle size control equations of the two- fluid model are generally based on the Reynolds-averaged method[20–22],assuming that the gas is incompressible,ignoring the heat transfer and mass transfer between the two a simplified form of control equations can be obtained:Continuity equation:Momentum conservation equation: equationsThe standard k-ε model is selected for turbulence is a classical representation of the Reynolds-averaged k and ε equations are:with C1ε=1.44,C2ε=1.92,C3ε=1.2,Cμ=0.99,σk=1.0,σε= viscosity is calculated by: forcesThe exact expression of the interphase forces is the key to simulating the gas–liquid two-phase flows,and many researches exist on the inter phase forces between gas and liquid[23,24].In this work,the drag force,transverse lift force,turbulent dispersion force and wall lubrication force are considered. drag forceIt is generally believed that the drag is the predominant force in modeling the gas–liquid flow s of bubble columns[25],as did in many simulation[14,26].Air bubbles are formed from the bottom of the tower with a certain gas the control volume formulation,all bubbles in the control volume suffer the total drag force as follow s:Liu et al.[23]introduced a modified drag coefficient CDin the bubble group drag the modified drag model was used in the numerical investigations of the flow characteristics of pressurized churn turbulent bubble column with the operation pressure varying from 0.5 MPa to 2.0 MPa,and superficial gas velocity from 0.20 m·s?1to 0.31 m·s?1.The simulation results can accurately reflect varioushydrodynamic parameters in the bubble refore,their drag coefficient CD[23]is adopted in this work: dispersion forceIn order to simulate the turbulence of the fluid in the high-pressure bubble column,it is necessary to introduce the turbulence diffusion force,which can help in making the gas holdup evenly turbulent diffusion force formula proposed by Lopez de Bertodano[27]is as follows:CTDisthe coefficient of turbulent diffusion force,and its default value is 1.If Eq.(8)is used directly in the simulation,the simulation is not easy to ,the limiting function fTD,limitingis introduced into the Fluent 15.0 platform to adjust the turbulent disperation force new expression of the turbulent diffusion force model becomes:w here εG,1and εG,2was set to be 0.3 and 0.7. lift forceWhen the bubble is moving in the liquid,the pressure distribution around the bubble is not balanced due to the asymmetry of the liquid in the direction of the moving air that a transverse lift force is generated perpendicular to the direction of bubble lift force of the discrete phase in the continuous phase given by Drew[20]is.The coefficients CLand CTDhave different values in the uncertainty reflects the complexity of gasbubble diffusion in turbulent multiphase media and the limitations of prior know [28]considered the parameter CL/CTDas a function of liquid holdup εLand the equation is as follow s: lubrication forceThe bubbles are subjected to a force toward the center,so the bubbles move in the direction away from the wall,and this force is called the wall lubrication velocity dependence on wall lubrication force is clearly shown by Nguyen et al.[29].So the wall lubrication force was introduced in the simulation of gas–liquid flow wall lubrication force model[30]is used in this work:where CWLis a value from Eq.(14).and CWis defined as:The definition of Eo is: balance modelWhen the gas contacts with the liquid in the high-pressure bubble column,the dispersed bubbles exist in a broad range of size.A significant attribute of gas–liquid flows is that the bubbles of different sizes interact with each other through the mechanisms of breakup and present,the population balance model(PBM)is used to deal with this general form of the population balance equation is: breakup modelBecause the Luo model has a relatively simple form,high prediction accuracy,it is widely ,the Luo model is adopted in this paper.in which K,n,m,β,b can be expressed as coalescence modelThe expression of bubble coalescence model[31]is.The collision frequency between bubbles can be expressed as.Based on the Luo model and the correction coefficient Ceintroduced,the modified bubble coalescence model is.In this work,the coalescence resulted from turbulent eddies was eddies cause the bubbles to collide and coalesce with a certain mechanism of pressure on bubble coalescence is not yet clear.Up to now,models for bubble coalescence are mainly based on experimental phenomena by semi-theoretical semiempirical [6]introduced a constant correction coefficient Cein the bubble coalescence ,we find that the correction coefficient Ceis not a constant under different gas velocities and this work,according to the experimental data of Qin[14]from cold experiment,a density correction factor Cecan be obtained by simulation at the apparent gas velocity 0.199,0.233,0.275 m·s?1and the operating pressure at 0.5,1,1.5,2.0 data are shown in Table 1.The linear regression equation between Ceand the gas density is as follow s:Table 1Coalescence model correction coefficients and the simulated gas holdupsugP/MPa ρ/ρ0CeεexpεsimError of ε 0.199 0.5 5 1.16 0.3573 0.3603 0.83%0.199 1.0 10 1.45 0.4586 0.4593 0.15%0.199 1.5 15 1.50 0.4941 0.5016 1.52%0.199 2.0 20 1.62 0.5321 0.5313 ?0.15%0.233 0.5 5 1.16 0.3937 0.3961 0.61%0.233 1.0 10 1.45 0.4881 0.4910 0.59%0.233 1.5 15 1.50 0.5254 0.5300 0.88%0.233 2.0 20 1.62 0.5523 0.5527 0.07%0.275 0.5 5 1.16 0.4250 0.4333 1.94%0.275 1.0 10 1.45 0.505 0.5200 2.97%0.275 1.5 15 1.50 0.548 0.5500 0.36%0.275 2.0 20 1.62 0.571 0.5774 1.12% Model and Mesh GenerationThe experiment of Qin[14]mainly examined the impact of pressure on the hydrodynamic behaviorsin variousoperating conditionsat 25°C.Air wasused asthe gasphase and water asliquid phase,the gas velocity was varied from 0.119 to 0.312 m·s?1and pressures from 0.5 to 2.0 MPa.The geometry of the high-pressure bubble column is shown in Fig.1(a):height 6600 mm,diameter 300 mm,wall thickness5 mm.Tw o conductivity probes at the height of 2550 mm and 3050 mm are used to measure the bubble size in the radial direction and the velocity of rise of bubble electrodes of Electrical Resistance Tomography(ERT)are evenly mounted on the inner wall of the high-pressure bubble column at the height of 2600 mm and 3000 mm to measure the local air holdup and average gas holdup. diagram of the high-pressure bubble column and meshing with the 3D geometric model,the computational amount of the numerical simulation will be greatly reduced by the two-dimensional geometric et al.[18]succeeded in simulating gas–liquid flow behaviors in a bubble column by an axisymmetric refore,two-dimensional axisymmetric geometric model are used in this work.As shown in Fig.1(b):6.6 m high and 0.15 m wide.In the model,no gas distribution plate is arranged on the column bottom and the gas is directly fed from the central bottom of the column(r/R≤0.8).The heights of 2550,2600,3000 and 3050 mm were set as the sampling experimental and simulation results are compared and analyzed in the following mesh of 2D simulation structure is generated by the Mapped Face Meshing of mesh at the wall:wall inflation,Maximum Layers:5,Grow th Rate:1.2.As shown in Fig.1(c),the obtained mesh is between the two sections from 2000 mm to 3200 mm above the bottom,and this section of 1200 mm length is enlarged to show the refine meshing at the wall.In Fig.2,a grid independence verification was performed under the condition of 0.160 m·s?1and 0.5 MPa.The simulation results of radial gas holdup were compared and analyzed with four different grid figure indicates that the simulation results with other three grids are basically consistent except for the first ,to ensure a high computing accuracy and an acceptable computing time,the grid with 5940 cells was adopted. of grid on the simulation results of gas holdup. and holdup gas holdup holdup at different apparent velocities and operating pressures(0.5,1.0,1.5,2.0 MPa).The simulation results of the average gas holdup between the heights 2600 and 3000 mm at different apparent gas velocities and pressure were compared with the experimental data[14]in Fig.3.It is indicated that although the gas holdup made by the CFD-PBM coupled model without Cedirectly increases with the increasing apparent gas velocity,the error is greater than that by the modified CFD-PBM model with is shown that the modified CFD-PBM coupled model can accurately simulate the average gas holdup under other experimental conditions of Qin[14]. gas holdupIn Fig.4,the simulated radial gas holdup profiles are compared with the experimental data under the operating conditions of 0.5,1.0,1.5,2.0 MPa and 0.160,0.215,0.253,0.317 m·s?1.The two diagrams at the same apparent velocity show the simulation results of the modified CFD-PBM coupled model and the CFD-PBM coupled model respectively compared with the experimental data.And the radial gas holdup showed a decreasing trend from the center to the column wall.It can be seen that the simulation results obtained by the CFD-PBM coupled model are generally smaller than the experimental data at elevated et al.[32]proposed a model for the bubble interaction time by energy conservation analysis,ignoring the effects of operating the Luo coalescence model is used directly to simulate high-pressure bubble column,as the simulation environment pressure is lower than the really pressure,the simulation result of gas holdup is can be seen that when the density correction coefficient Ceis introduced into the Luo coalescence model,the error between the simulation results and the experimental data of Qin[14]is reduced. diameter distribution of describes the effect of operating pressure on bubble size radial distribution at the apparent gas velocity of 0.160,0.215,0.253,0.317 m·s?1.It can be seen from Fig.5 that although the bubble size decreases from the center to the edge of the column,the results obtained by the CFD-PBM coupled model are far larger than the experimental data at four different operating pressures(0.5–2.0 MPa)per gas the other hand,the results obtained by the modified CFD-PBM coupled model are more consistent with experimental ,the effect of operating pressure on the bubble diameter in the gas–liquid flow s can be well simulated by adjusting the density correction coefficient Cein the bubble coalescence model. size shows the simulation results of the influence of pressure on the bubble sizedistribution at the apparent gas velocity of 0.160 m·s?1.The bubbles show a unimodal number of medium bubbles(3–8 mm)increased at elevated pressure,while the number of smaller bubbles(＜3 mm)almost was consistent with the simulation results of Yang et al.[33].This is fully showed that the bubble size became smaller and more uniform at elevated pressure. distribution axial shows the effect of operating pressures on radial distribution of gas rising results at four gas velocities show a gradual decrease from the center to the column wall,and the gas velocity increases with the increase in the operating pressure and apparent gas variation trend was consistent with the experiment reported by Wilkson et al.[34] holdup on radial distribution(uG=0.160,0.215,0.253,0.317 m·s?1). axial shows the effect of pressure on the radial profiles of water axial can be seen from Fig.8 that the axial liquid velocity gradually decreases from the center to the wall of the liquid velocity is positive in the center of the column and negative near the side wall of the indicates that the liquid phase appears to circulate in the liquid circulation is in favor of gas–liquid fully difference in axial velocity is not significant at different operating pressures,indicating that the liquid velocity is not greatly affected by the simulated results affected by pressure at other apparent gas velocities(0.215,0.253,0.317 m·s?1)are similar to result of 0.160 m·s?1. diameter in the radial direction(uG=0.160,0.215,0.253,0.317 m·s?1). of the Modified CFD-PBM Coupled ModelThe experiment of Wilkson and Dierendonck[34]mainly examined the influence of pressure on gas hold-up and bubble size in bubble column had a diameter of 0.16 m,the height of column was 2.0 m.The liquid was deionized water(20°C),and gas was the research of Reilly et al.[35]it is known that the influence of column diameter on gas-holdup can be we choose the experimental data of gas-holdup versus superficial gas velocity for the water/nitrogen system[34]to validate the modified CFD-PBM coupled shows the comparison between simulation results by the modified CFD-PBM coupled model and Wilkson's experimental data[34]under each pressure(0.5,1.0,1.5 MPa)at different apparent gas can be seen,the simulation results of gasholdup are basically in agreement with the experimental data.So it shows that the modified CFD-PBM coupled model according to the experimental data of Qin[14]can be applied to the simulation under other experimental conditions. of operating pressure on bubble size distribution at 0.160 m·s?1.The work focused on simulating the effects of operating pressure on hydrodynamic CFD-PBM coupled model was employed to investigate the gas–liquid flow in a high-pressure bubble the work,the following conclusions can be draw n: water velocity at different pressure simulated with Ce(uG=0.160 m·s?1).(1)From the comparison between experimental data and simulation results by two models(the CFD-PBM coupled model and the modified CFD-PBM coupled model),it can be seen that the latter model offers good agreement with experimental data.So the effects of operating pressure on the hydrodynamic parameters can be well predicted by the modified CFD-PBM coupled model.(2)The effects of operating pressure on the bubble size distribution were predicted by the modified CFD-PBM coupled model in gas–liquid flow.The bubble size became smaller and more uniform at elevated pressure.(3)Through the validation of the modified CFD-PBM coupled model with the water-nitrogen system of Wilkson and Dierendonck[34],it can be seen that the simulation results are in good agreement with the experimental the modified model may be applied to other experimental systems. axial velocity at different gas velocities and pressures. holdup under different apparent gas coefficientCD,∞ideal state drag coefficientCLtransverse lift coefficientCTDturbulent dispersion coefficientCWLwall lubrication coefficientdBdiameter of bubble,mE0parameter E0Fexinterphase forcesFDdrag force,N·m?3FLtransverse lift,NFTD,LFTD,Gturbulent dispersion force,NfTD,limitingturbulent diffusion force model limiting functionFWLwall lubrication force,NGkturbulence energy generationGbk term of turbulent kinetic energyg gravity acceleration,m·s?2k,kLturbulent kinetic energy,m2·s?2P operating pressure,PaPoatmospheric pressure under standard conditions,PaP(ViVj) bubble coalescence efficiencyUivelocity,m·s?1,i=1:gas phase,i=2:liquid phaseuGgas velocity,m·s?1uLliquid velocity,m·s?1uijcharacteristic velocity of bubble collision\Ωbr(V,V′)bubble breakage rateΩag(ViVj)bubble coalescence rate,εiphase holdup,i=1:gas phase,i=2:liquid phaseεGaverage gas phase holdupεGgas phase holdupεLliquid phase holdupεG,1εG,2turbulent diffusion force limit function constantζijrelative diameter of bubbleζminminimum relative diameter of bubbleμtturbulent viscosity,Pa·sρ0Gas density under standard conditions,kg·m?3ρLliquid density,kg·m?3ρGgas density,kg·m?3σ surface tension,N·s?1ω(ViVj) collision frequency between bubbles of size diand dj,m3·s?1References[1],Bubble Column Reactor,John Wiley&Sons Ltd,Chichester,1992.[2]B.Jager,,Advances in low temperature Fischer-Tropsch synthesis, 23(1)(1995)17–28.[3]B.Gandhi,,,Hydrodynamic behavior of slurry bubble column high solids concentrations,Powder (2)(1999)80–94.[4],,Bubble column reactors,Process (7)(2005)2263–2283.[5]P.Chen,J.Sanyal,,CFD modeling of bubble columns flows implementation of population balance,(22)(2004)5201–5207.[6]T.F.Wang,Experimental Study and Numerical Simulation on the Hydrodynamics in Gas-Liquid(Slurry)Reactors Ph D Thesis Tsinghua University,Beijing,2004(in Chinese).[7]C.T.Xing,T.F.Wang,J.F.Wang,Experimental study and numerical simulation with a coupled CFD PBM model of the effect of liquid viscosity in a bubble column,(1)(2013)313–322.[8]L.S.Fan,G.Q.Yang,D.J.Lee,,X.Luo,Some aspects of high-pressure phenomena of bubbles in liquids and liquid-solid suspensions,(21)(1999)4681–4709.[9],,,,,Influence of elevated pressure and particle lyophobicity on hydrodynamics and gas–liquid mass transfer in slurry bubble columns,AIChE J 56(3)(2010)584–596.[10],,,,Gas holdup and mass transfer in bubble column reactors operated at elevated pressure,(13)(1999)2237–2246.[11],A.Macchi,Hydrodynamics of a high pressure three-phase fluidized bed subject to foaming,(3)(2010)293–301.[12]X.Luo,D.J.Lee,R.Lau,L.S.Fan,Maximum stable bubble size and gas holdup in high pressure slurry bubble columns,AIChE J 45(4)(1999)665–680.13 R.Zou,X.Jiang,B.Li,Y.Zu,L.Zhang,Studies on gas holdup in a bubble column operated at elevated temperatures,(10)(1990)927–928.[14]Y.J.Qin,Numerical Simulation and Measurement of CFD Liquid Bubble Column the Pressurized Fluid Mechanics Parameter Ph D Thesis Beijing University of Chemical Technology,Beijing,2012(in Chinese).[15]G.Q.Yang,L.S.Fan,Axial liquid mixing in high-pressure bubble columns,AIChE J 49(8)(2003)1995–2008.[16],,Eulerian simulations of bubble columns operating at elevated pressures in the churn turbulent flow regime,(21)(2001)6249–6258.[17]P.Chen,,J.Sanyal,Three-dimensional simulation of bubble column flows with bubble coalescence and breakup,AIChE J 51(3)(2005)696–712.[18]T.F.Wang,J.F.Wang,Y.Jin,A CFD-PBM coupled model for gas?liquid flows,AIChE J 52(1)(2006)125–140.[19]C.T.Xing,T.F.Wang,K.Y.Guo,J.F.Wang,A unified theoretical model for breakup of bubbles and droplets in turbulent flow s,AIChE J 61(4)(2015)1391–1403.[20]D.A.DREW,Mathematical modeling of two-phase flow, Mech.15(1)(1983)261–291.[21],P.Chen,?,Computational modeling of gas-liquid flow in bubble column,(3)(2004)225–375.[22]D.Drew,,Theory of multicomponent fluids,Applied mathematical sciences,16(2),,Springer,New York,1999 viii.[23]X.Liu,Y.Zhang,L.Zhang,H.B.Jin,Hydrodynamic Simulation of Pressurized Bubble Column Based on Interphase Interaction Model,CIESC J.68(1)(2017)87–96(in Chinese).[24],,Numerical investigation of the interphase forces and turbulence closure in 3D square bubble columns,(17)(2014)154–168.[25]P.Chen,Modeling the Fluid Dynamics of Bubble Column Flows,Washing University Sever Institute of Technology,Saint Louis,2004.[26]Y.Xu,H.Dong,X.Tian,X.Zhang,S.Zhang,CFD-PBM coupled simulation of ionic liquid-air two-phase flow in bubble column,(10)(2011)2699–2706.[27]M.Lopez de Bertodano,Turbulent Bubbly Tw o-phase Flow in a Triangular Duct,Rensselaer Rolytechnic Institute,New York,1992.[28]Y.Zhang,Hydrodynamics of Turbulent Bubble Column with Internals in Well-developed Flow Region Ph D Thesis Zhejiang University,Hangzhou,2011(in Chinese).[29],C.H.Song,B.U.Bae,D.J.Euh,The dependence of wall lubrication force on liquid velocity in turbulent bubbly two-phase flows,(8)(2013)781–798.[30],Struggle with Computational Bubble Dynamics,Third International Conference on Multiphase Flow,ICMF,Lyon,France,1998.[31]H.Luo,,Theoretical model for drop and bubble breakup in turbulent dispersions,AIChE J 42(5)(1996)1225–1233.[32]H.Luo,,Modeling and simulation of binary approach by energy conservation analysis,(1)(1996)145–153.[33]G.Y.Yang,K.Y.Guo,T.F.Wang,Numerical simulation of the bubble column at elevated pressure with a CFD-PBM coupled model,(2017).[34],,Pressure and gas density effects on bubble breakup and gas hold-up in bubble columns,(8)(1990)2309–2315.[35],,,A.Jain,,A correlation for gas holdup in turbulent coalescing bubble columns,(1986)705–717.

文章来源：《人口与经济》 网址: http://www.rkyjj.cn/qikandaodu/2020/1013/558.html