CFD–DEM simulation of the gas–solid flow in a cyclone separator(不同喂入量耦合仿真重点参考)

CFD–DEM simulation of the gas–solid?ow in a cyclone separator

K.W.Chu a,B.Wang a,c,D.L.Xu b,Y.X.Chen b,A.B.Yu a,n

a Laboratory for Simulation and Modelling of Particulate Systems,School of Materials Science and Engineering,University of New South Wales,Sydney,NSW2052,Australia

b Institute of Powder Engineering,School of Materials Science and Engineering,Xi’an University of Architecture and Technology,Xi’an710055,PR China

c Key Laboratory of Western China’s Environmental Systems,College of Earth an

d Environmental Sciences,Lanzhou University,Lanzhou730000,PR China

a r t i c l e i n f o

Article history:

Received25August2010

Received in revised form

1November2010

Accepted11November2010

Available online26November2010

Keywords:

Cyclone

Gas–solid?ow

Computational?uid dynamics

Discrete element method

Separation

Granular dynamics

a b s t r a c t

In this work,a numerical study of the gas–solid?ow in a gas cyclone is carried out by use of the combined

discrete element method(DEM)and computational?uid dynamics(CFD)model where the motion of

discrete particles phase is obtained by DEM which applies Newton’s equations of motion to every

individual particle and the?ow of continuum?uid by the traditional CFD which solves the Navier–Stokes

equations at a computational cell scale.The model successfully captures the key?ow features in a gas

cyclone,such as the strands?ow pattern of particles,and the decrease of pressure drop and tangential

velocity after loading solids.The effect of solid loading ratio is studied and analysed in terms of gas and

solid?ow structures,and the particle–gas,particle–particle and particle–wall interaction forces.It is

found that the gas pressure drop increases?rst and then decreases when solids are loaded.The reaction

force of particles on gas?ow is mainly in the tangential direction and directs mainly upward in the axial

direction.The reaction force in the tangential direction will decelerate gas phase and the upward axial

force will prevent gas phase from?owing downward in the near wall region.The intensive particle–wall

collision regions mainly locate in the wall opposite to the cyclone inlet and the cone wall.Moreover,as the

solid loading ratio increases,number of turns travelled by solids in a cyclone decreases especially in the

apex region of the cyclone while the width of solid strands increases,the pressure drop and tangential

velocity decrease,the high axial velocity region moves upwards,and the radial?ow of gas phase is

signi?cantly dampened.

&2010Elsevier Ltd.All rights reserved.

1.Introduction

Gas cyclones are widely used in industries to separate dust from

gas or for product recovery because of its geometrical simplicity,

relative economy in power usage and?exibility.The most impor-

tant performance variables of a gas cyclone are usually gas pressure

drop and solid separation ef?ciency,both of which are known to be

signi?cantly affected by solid loading ratio or concentration(Yuu

et al.,1978;Hoffmann et al.,1992;Fassani and Goldstein,2000).

Yuu et al.(1978)reported that the presence of dust reduced the

pressure drop by as much as30%even at extremely low concen-

trations such as0.2g/m3,and in one case the tangential velocity

reduced as much as40%.Understanding and modelling the physics

underlying this phenomenon is however very challenging.

In the previous numerical models of gas cyclones for the

simulation of gas?ow,conventional computational?uid dynamics

(CFD)method was mainly used.Boysan et al.(1982)developed one

of the?rst CFD models for gas cyclones and showed that the

standard kàe turbulence model was inadequate for simulating

?ows with swirl since it led to excessive turbulence viscosities and

unrealistic tangential velocities.Recent studies suggested that

turbulence models like Reynolds Stress Model(RSM)and Large

Eddy Simulation(LES)that can address anisotropic turbulence

problems should be used for gas cyclones and the corresponding

numerical results obtained are comparable with experimental

measurements(Hoekstra et al.,1999;Slack et al.,2000;Hu

et al.,2005).For the simulation of solid?ow,except for a few

continuum models(Meier and Mori,1998;Qian et al.,2007),the so-

called Lagrangian particle tracking(LPT)method is mainly used,

where the effect of solids on gas?ow and particle–particle

interaction force are usually ignored(Yoshida,1996;Pant et al.,

2002;Derksen et al.,2006;Wang et al.,2006).The LPT approach is

able to qualitatively study the effect of parameters such as

geometry,gas?ow velocity and particle size.However,it cannot

satisfactorily describe the effect of solids on gas?ow and conse-

quently parameters such as solid concentration or loading ratio.

Efforts have been made to overcome this de?ciency by various

investigators.For example,Crowe and Pratt(1974)developed a

two-dimensional(2D)model to predict the increase of the overall

collection ef?ciency with solid loading ratio.Recently,Derksen

et al.(2008)developed a modi?ed LPT model and showed that the

presence of solid particles causes the cyclone to lose some swirl

intensity and the turbulence of the gas?ow is strongly damped.

They also pointed out that the modelling of particle–particle

Contents lists available at ScienceDirect

journal homepage:https://www.wendangku.net/doc/b212635693.html,/locate/ces

Chemical Engineering Science

0009-2509/$-see front matter&2010Elsevier Ltd.All rights reserved.

doi:10.1016/j.ces.2010.11.026

n Corresponding author.Tel.:+61293854429;fax:+61293855956.

E-mail address:a.yu@https://www.wendangku.net/doc/b212635693.html,.au(A.B.Yu).

Chemical Engineering Science66(2011)834–847

interaction in a gas cyclone is important.By means of a similar approach,Wan et al.(2008)successfully predicted the signi?cant change of gas?ow?eld as a result of the presence of solid particles and spiral dust strand.Nonetheless,in theory the calculation of the reaction force on gas phase acted by solid phase is mainly based on solid concentration and the relative velocity between solids and gas;the accurate prediction of solid concentration depends on the modelling of particle–particle interactions that is not shown in the LPT models.

On the other hand,in recent years,the so-called combined approach of discrete element method(DEM)and CFD(CFD–DEM) has been developed(Tsuji et al.,1992;Xu and Yu,1997;Zhou et al., 2010)and accounts for both particle–particle and particle–?uid interactions.The CFD–DEM approach has proven to be effective in modelling various particle–?uid?ow systems(Tsuji et al.,1992;Xu and Yu,1997;Li et al.,1999;Xu et al.,2000;Rhodes et al.,2001; Kafui et al.,2002;Yu and Xu,2003;Limtrakul et al.,2004;Di Renzo and Di Maio,2007;Tsuji,2007;Kuang et al.,2008;Malone and Xu, 2008).In particular,efforts have been made to extend the CFD–DEM approach to study complex particle–?uid?ow systems (Kawaguchi et al.,1998;Rong and Horio,2001;Ibsen et al., 2004;Chu and Yu,2008a;Chu et al.,2009a;Chu et al.,2009b; Gui et al.,2009;Zhao et al.,2009).To date,to the authors’knowledge,few studies have been made on the gas–solid?ow in a gas cyclone by means of CFD–DEM approach.

In this work,a CFD–DEM model is developed to describe the gas–solid?ow in a gas cyclone.The effect of solid loading ratio is then investigated,aiming to generate a comprehensive under-standing of this complicated gas–solid?ow system.

2.Mathematical model

The CFD–DEM model used for the present work has been well documented in the literature(Xu and Yu,1997;Zhou et al.,2010). For brevity,therefore we only give a brief description of the method here.

The solid phase is treated as a discrete phase and described by the so-called discrete element method(Cundall and Strack,1979). According to the model,the translational and rotational motions of a particle at any time,t,can be described by Newton’s law of motion:

m i d v i

dt

?f pàf,it

X k i

j?1

ef c,ijtf d,ijTtm i ge1Tand

I i d x i

d t

?

X k i

j?1

eT ijtM ijTe2T

where m i,I i,v i and x i are,respectively,the mass,moment of inertia, translational and rotational velocities of particle i.The forces acting on solids are the gas–solids interaction force,f pàf,i,inter-particle forces between particles i and j,which include the contact forces, f c,ij,and viscous damping forces,f d,ij,and the gravitational force, m i g.In this model,the gas–solid interaction force includes the viscous drag force(f D,i)and pressure gradient force(f pg,i).The inter-particle forces are summed over the k i particles in contact with particle i.Torques,T ij,are generated by the tangential forces and cause particle i to rotate because the inter-particle forces act at the contact point between particles i and j and not at the particle centre. M ij is the rolling friction torque that is in opposition to the rotation of the i th particle.

The governing equations of gas phase are the same as those used in the two-?uid models(TFM)(Anderson and Jackson,1967; Gidaspow,1994;Enwald et al.,1996).There are three sets of governing equations in TFM,developed by Anderson and Jackson (1967).According to Zhou et al.(2010),Set II and in particular Set I can be used generally,and Set III can only be used conditionally.In this work,Set I is used.Thus,the conservations of mass and momentum in terms of the local mean variables over a computa-tional cell are given by

@er f eT

@t

tr Uer f e uT?0e3Tand

@er f e uT

tr Uer f e uuT?àr PàF pàftr Uee sTtr f e gtr Ueàr f u u u uT

e4Twhere e,u,u0,t,r f,P,F pàf,s and g are,respectively,porosity,mean and?uctuating?uid velocity,time,?uid density,pressure,volu-metric?uid–particle interaction force,?uid viscous stress tensor, and acceleration due to gravity.F pàf?e1=V cellT

P k c

i?1

f pàf,i,where f pàf,i is the total?uid force on particle i and k c is the number of particles in a CFD cell.àr f u u u u is the Reynolds stress term due to turbulence and solved by the Reynolds Stress Model(RSM) provided in commercial CFD software Fluent.The equations used to calculate the forces in Eqs.(1)–(4)are listed in Table1.They are very much standardised now(Zhu et al.,2007).

The?uid?ow?eld can be obtained by solving Eqs.(3)and(4)by use of a standard CFD method,and the solid?ow can be produced by solving Eqs.(1)and(2)by an explicit time integration method, facilitated by initial and boundary conditions for a given?ow.The modelling of the solids?ow by DEM is at the individual particle level,whilst the?uid?ow by CFD is at the computational cell level. Their two-way coupling(?uid forces act on particles and particles react on?uid)is numerically achieved as follows.At each time step, DEM will give information such as the positions and velocities of individual particles,for the evaluation of porosity and volumetric ?uid–particles interaction force in a computational cell.CFD will then use these data to determine the gas?ow?eld which then yields the?uid forces acting on individual particles.Incorporation of the resulting forces into DEM will produce information about the motion of individual particles for the next time step.The?uid force acting on individual particles will react on the?uid phase from the

Table1

Components of forces and torques acting on particle i.

Forces and torques Symbols Equations

Normal forces

Contact f cn,ijàE

3e1àvT

???????

2R i

p

d3=2

n

n

Damping f dn,ij

àc n

3m i E

??????????????????

2e1àv2T

p????????

R d n

p

!1=2

v n,ij

Tangential forces

Contact f ct,ij

à

m

s

f cn,ij

9d t9

1à1à

min9d t9,d t,max

èé

d t,max

3=2

"#

d t

Damping f dt,ij

àc t6m i m s f cn,ij

?????????????????????????

1àd t=d t,max

p

d t,max

!1=2

v t,ij Torque

Rolling T ij R i?(f ct,ij+f dt,ij)

Friction M ijàm r f cn,ij^x ij

Body force

Gravity G i m i g

Particle–?uid interaction force

Viscous drag force f D,i

0:63t

4:8

Re p,i

!2

r

f

9u

i

àv i9eu iàv iTp d2

i eàb

i Pressure gradient force f pg,i V p,i r P

where:n?R i=R i,v ij?v jàv i+o j?R jào i?R i,v n,ij?(v iján)án,v t,ij?(v ij?n)?n,

^o ij?x ij=o ij,Re p,i?ed i r

f

e i9u iàv i9T=em

f

T,

b?3:7à0:65exp?àe1:5àlogRe p,iT2=2 ,e?1àe

P k c

i?1

V iT=eD V cT

K.W.Chu et al./Chemical Engineering Science66(2011)834–847835

particles,so that Newton’s third law of motion is satis?ed(Xu and Yu,1997).

The above CFD–DEM principles have been well established.Our previous CFD–DEM programs are all in-house codes.For compli-cated?ow systems,the code development for the solution of?uid phase could be very time-consuming.On the other hand,com-mercial CFD software packages such as Fluent,CFX and Star-CD are readily available for this purpose.In order to take advantage of this CFD development,we have extended our CFD–DEM code with Fluent as a platform,achieved by incorporating a DEM code into Fluent through its User De?ned Functions(UDF).This approach has been successfully used in our recent study of various complicated ?uid–solid?ow systems(Chu and Yu,2008a,b;Chu et al.,2009a, 2009b),and is used in this work.

3.Simulation conditions

The cyclone considered is a typical Lapple cyclone.Fig.1(a) shows the geometry and notations of the cyclone dimensions and Table2gives their values.Fig.1(b)shows the computational domain,containing47,750CFD cells.The whole computational domain is divided by unstructured hexahedron grids.At the zone near wall and vortex?nder the grids are dense,while at the zone away from walls the grids are re?ned.Three grid domains were tested in our preliminary computation,containing25,900,47,750, 95,350cells,respectively(Wang et al.,2006).The difference is less than5%for all variables examined,suggesting that computed results are independent of the characteristics of the mesh size.

Totally6runs of simulation is carried out.The only parameter changed in each run is the ratio of solids to gas by mass at the cyclone inlet,which is0,0.5,1, 1.5,2and 2.5,respectively (correspondingly volumetric ratios at the inlet are0,0.002, 0.004,0.006,0.008and0.01,respectively).The other simulation conditions are summarized in Table3.Note that in the DEM simulation,the computational effort increases with the decrease of particle size exponentially.Therefore,in order to reduce computa-tional effort and to develop a mechanistic understanding,coarse particles are used in this work although?ne particles are normally encountered in gas cyclones.Moreover,the particles are assumed to be spherical.In principle,CFD–DEM simulations can be per-formed for non-spherical particles as,for example,recently done for ellipsoidal particles(Zhou et al.,2009;Hilton et al.,2010). However,such simulations are computationally very demanding, particularly for complex?ow systems like cyclones.On the other hand,it is noticed that for particles which are not so non-spherical, their?ow behavior can be simulated by adjusting the sliding and rolling friction coef?cients(e.g.see,Zhou et al.,1999;Yu and Xu, 2003;Zhou et al.,2004).The effectiveness of this approximation will be examined in details,in connection with our on-going work on hydro or dense medium cyclones(Wang et al.,2007;Chu et al., 2009b).In the simulation,the gas?ow is?rst solved by use of Fluent to reach its macroscopic steady?ow state(de?ned as the state when the macroscopic pattern and parameters of the?ow do not change much by time).Then particles are introduced at the inlet of the cyclone,with the number of particles added per second calculated according to the pre-set solid mass?owrate.After the ?ow reaches its steady?ow state,the pressure drop is calculated by the average static pressure at the cyclone inlet minus that at the outlet of the cyclone vortex?nder.

Physical experiments have also been conducted to validate the numerical model.The experimental method is the same as that in our previous work(Wang et al.,2006).In such an experiment,air was blown into the inlet of the cyclone,with its?owrate measured by a?owmeter.The inlet gas velocity was set to20m/s.The exit tube was open to the air and the gas pressure at the top of the vortex ?nder was around1atm.A?ve-hole probe consisting of

an

Fig.1.Schematic and grid representation of the cyclone considered,together with the de?nitions of the sections for discussion:(a)3D view of geometry;and(b)3D view of CFD grids;(c)top view of the sections;and(d)front view of the sections.

Table3

Parameters used in the work.

Solid phase a Gas phase

Particle diameter(mm)2Type of gas Air

Particle density(kg/m3)2500Density(kg/m3) 1.225

Particle inlet velocity(m/s)3Gas inlet velocity(m/s)20

Sliding friction coef?cient0.3Viscosity(kg/m/s) 1.8?10à5

Young’s modulus(N/m2)1?107Time step(s)1?10à4

Poisson’s ratio0.3Grid number62,105

Damping coef?cient0.3Grid type Hexahedron

Rolling friction coef?cient6?10à5

a For convenience,the wall is assumed to have the same properties as particles

but in?nitely diameter.

Table2

Geometry of the cyclone considered(D?0.2m).

a/D b/D De/D S/D h/D H/D B/D

0.250.50.50.625 2.0 4.00.25

K.W.Chu et al./Chemical Engineering Science66(2011)834–847

836

adjustable frame and ?ve pressure transducers was used to measure the pressure of the gas ?eld at the cyclone inlet.When the ?ve-hole probe was placed in a ?ow ?eld,voltage signals obtained through the ?ve pressure transducers were transferred to an ampli?er.The magni?ed voltage signals were acquired through a data acquisition system containing a microprocessor and a personal computer.The material used both in simulation and experiment is mono-sized glass beads with diameter of 2mm.

4.Results and discussion 4.1.Model validation

Pressure drop is one of the most important operational para-meters for gas cyclones and is used as the major parameter for the purpose of validation in this work.For pure gas ?ow,Fig.2shows that the simulated pressure drops under different inlet gas velocities agree with the experimental results.For gas–solid ?ow,Fig.3shows that the predicted pressure drop decreases with the increase of solid loading steadily,which is in good agreement with the experimental measurement.It also qualitatively agrees with the trends reported in the literature (Yuu et al.,1978;Hoffmann et al.,1992;Fassani and Goldstein,2000;Bricout and Louge,2004;Cortes and Gil,2007).

The phenomenon that the tangential velocity decreases with the increase of solid loading was found in experiments (Yuu et al.,1978)and can be captured by the current numerical model.From Fig.4it can be seen that the tangential velocity apparently decreases with the increase of solid loading and the trend agrees with the experimental measurements.Note that the comparison is only qualitative since the present simulations were not carried out under the same conditions as the experiment.

The information shown in Figs.2–4suggests that the developed CFD–DEM model can capture the major ?ow characteristics in a gas cyclone.One of the advantages of CFD–DEM model is that it is a ?rst-principle approach and at a particle scale,and can thus provide rich microdynamic information about the gas–solid ?ow.The analysis of this information can lead to better understanding of the ?ow structure and mechanics in a gas cyclone as discussed in the following sections.4.2.Gas–solid ?ow pattern

4.2.1.Particle ?ow

It is known that the macroscopic ?ow pattern of solids in gas cyclone is that most particles congregate at the wall immediately after entering the inlet and then descend in strands or bands (Muschelknautz and Greif,1997;Wang et al.,2006;Li et al.,2009).

This ?ow pattern in strands has not been well predicted in some of the previous LPT models (Wang et al.,2006;Derksen et al.,2008).However,as shown in Fig.5,it can be satisfactorily captured by

the

https://www.wendangku.net/doc/b212635693.html,parison of the simulated and measured pressure drops at different inlet gas

velocities.

https://www.wendangku.net/doc/b212635693.html,parison of simulated and measured pressure drops at different solid loading ratios.

Outer cylinder wall

Inner cylinder wall

Outer cylinder wall Inner cylinder wall

Solid loading ratio (kg solid per kg air)

Fig.4.Distributions of the tangential velocity of gas phase at the cylinder section of the cyclone under different solid loading ratios:(a)simulation results and (b)experimental measurement (Yuu et al.,1978).

K.W.Chu et al./Chemical Engineering Science 66(2011)834–847837

current CFD–DEM model.It can be seen that the ?ow reaches its macroscopically steady state ?ow after 2s.Particles have a higher descending angle and velocity in the cylindrical part than in the conical part,because of the supporting forces acting on particles by the cyclone conic wall (Wang et al.,2006).

It would be interesting to know the differences in the solid ?ow patterns when the reaction forces of solids on gas ?ow are ignored (one-way coupling)and considered (two-way coupling).Fig.6compares the ?ow patterns when solid load ratio is 2.5.It can be seen that the ?ow patterns are similar in the cylinder section of the cyclone except that the descending angle of solids under the one-way coupling is smaller than that under the two-way coupling condition.For the ?ow in the conic section,the ?ow patterns are however quite different.Under the one-way coupling condition,strands ?ow pattern is not clearly observed in the conic section.There are much more particles residing in the conic section than that under the two-way coupling condition.Furthermore,the axial velocities of particles under the one-way coupling condition are generally higher than those under the two-way coupling condition.The differences originate from the fact that solids affect the ?ow of gas signi?cantly,especially in the conic section as discussed in Section 4.2.They also highlight the need for two-way coupling for cyclones.In the following,unless otherwise speci?ed,all the results are based on the two-way coupling approach.

Fig.7shows the comparison of the solid ?ow patterns under different solid loading ratios.Clearly,strands ?ow patterns are observed,irrespective of solid loading ratios.However,the details

are different.It can be seen that the number of turns travelled by solids in the cyclone decrease with the increase of solid loading ratio.The ‘‘strands’’are wider when solid loading ratio is high,which qualitatively agrees with the ?nding by Trefz and Muschelknautz (1993).The overall axial velocity of particles decreases with the increase of solid loading ratio.

Fig.8shows that the mass of solids residing in the cyclone increases more than linearly with the increase of solid loading ratio.When solid loading ratio is 0.5,the mass of solids residing in the cyclone is 0.11kg (Point A).However,if the mass residing in the cyclone is assumed to increase linearly with solid loading ratio,then the mass of solids in the cyclone should be about 0.067kg (Point B),64%less than that in the real case.The reason why solids accumulate in the cyclone easier at lower solid loading ratios is that particles tend to keep rotating in the conic section when the ?ow is dilute and particle–particle interaction force is not big enough to make particles move downward.

4.2.2.Fluid ?ow

Pressure drop is one of the most important parameters in the operation of gas cyclone since it represents the energy consump-tion.It varies with solid loading ratio,as shown in Fig.3.Increasing solid loads decreases the pressure drop,which is consistent with the literature (Yuu et al.,1978;Hoffmann et al.,1992;Fassani and Goldstein,2000).This phenomenon was thought to be contrary to the rationale.For example,in ?uidization or pneumatic conveying,pressure drop usually increases with solid loading ratio.A notice-able ?nding from the present CFD–DEM simulations is that the pressure drop increases slightly at the beginning (t o 0.1s)and then decreases sharply,as shown in Fig.9.Actually,from Fig.5it can be seen that the ?ow at the inlet of the cyclone is similar to the one in pneumatic conveying,which results in the increase of pressure drop when solids are loaded at the beginning (t o 0.1s in the current case).The sharp decrease of pressure drop after t ?0.1s is caused by the gas–solid interaction which will be discussed in Section 4.3.

The inner ?ow structures of gas phase under different solid loading ratios at a vertical central plane of the cyclone are shown in Figs.10–13.Fig.10shows the pressure distributions with

different

Fig.5.Snapshots showing the dynamic ?ow of particles in the cyclone when solid loading ratio is 2.5(particles are coloured by particle velocity in Z

-direction).

https://www.wendangku.net/doc/b212635693.html,parison of solid ?ow patterns when the reaction of solids on gas ?ow is:(a)not considered (one-way coupling)and (b)considered (two-way coupling).Solid loading ratio was 2.5kg solid per kg air for both cases and snapshots were taken at t ?2s.Particles are coloured by the axial velocity of particles.

K.W.Chu et al./Chemical Engineering Science 66(2011)834–847

838

solid loading ratios.It can be seen that the pressure is high on the cyclone wall and can be negative in the centre of the cyclone.The magnitude of pressure near the cyclone wall decreases steadily with the increase of solid loading ratio.When comparing Fig.10(a)with Fig.10(b)–(d),it can be seen that the negative pressure zone in the apex of the cyclone disappears after loading solids.This suggests that the solids have a higher impact on the gas ?ow in the apex than in the other regions.Another obvious trend shown in Fig.10is that the length of the negative pressure zone decreases as the solid loading ratio increases.

Fig.11shows the tangential velocity distributions with different solid loading ratios.An obvious trend is that the tangential velocity decreases with the increase of solid loading ratio,especially in the apex of the cyclone.It also suggests that the solids have a higher impact on the gas ?ow in the apex than the other regions.Another trend shown in Fig.11is that the distribution of the tangential velocity is not symmetric.This is caused by the fact that there is only one inlet which is not symmetric in geometry.

There are limited experimental and numerical data about the axial and radial velocities in gas cyclones when solids are loaded.However,those velocities can be readily produced by the current

model.From Fig.12(a)and (b)it can be seen that when the solid loading ratio is 0.5,the high axial velocity region moves toward to the centre of the cyclone.This is particularly so in the lower region of the conic part.From Fig.12(b)–(d)it can be seen that the highest axial velocity region moves upward as solid loading ratio increases further.This trend is believed to be caused by the accumulation of particles in the cyclone which prevents gas from ?owing down-ward since the reaction force acting on gas phase by particles generally points upward in the axial direction (see Fig.19).

Fig.13shows the different radial velocity distributions with different solid loading ratios.Fig.13(a)demonstrates that the forced vortex in the centre of the cyclone for pure gas ?ow looks like a helical twisted cylinder.The axis of the forced vortex is not straight but curved.After a small amount of solids are loaded,as shown in Fig.13(b),the axis of the forced vortex at the cylinder part of the cyclone becomes almost straight while the twisted

forced

https://www.wendangku.net/doc/b212635693.html,parison of solid ?ow patterns (at t ?2s)under different solid loading ratios:(a)0.5;(b)1.0;(c)1.5;(d)2.0;and (e)2.5.

A B

Fig.8.The mass of solids residing in the cyclone under different

solid loading ratios.

Fig.9.Pressure drop vs.time under different solid loading ratios.

K.W.Chu et al./Chemical Engineering Science 66(2011)834–847839

vortex becomes more intensive in the conic area of the cyclone.As solid loading ratio increases further,the forced vortex in the conic part is dampened greatly (Fig.13(c)and (d)).

Fig.14shows the spatial distributions of pressure and tangential velocity of the gas phase in more details.It con?rms the signi?cant change of the ?ow after loading solids.It also shows that ?ow is largely symmetric even though the cyclone geometry is not symmetric,i.e.,there is only one inlet in the cyclone.

The change of gas velocities after loading solids can be further illustrated by considering the total kinetic energy of the motion of gas phase,as shown in Fig.15.Here,the kinetic energy is the sum of

the kinetic energy (P cell number cell ?1

e1=2Tm cell 9v cell 92

)in each computa-tional cell.m cell and v cell are the mass and velocity of gas phase in a CFD cell,respectively.The kinetic energy is also shown in different directions,calculated using the tangential,axial or radial compo-nents of v cell .According to this de?nition,the total kinetic energy is the sum of the kinetic energy in the tangential,axial and radial directions.It can be seen that the tangential kinetic energy decreases signi?cantly with the increase of solid loading ratio while the radial and axial kinetic energy are almost constant.As the kinetic energy is directly related to velocity,these ?ndings

con?rm

Fig.10.Pressure (Pa)distributions at the central section vertical to the cyclone inlet under different solid loading ratios:(a)0.0;(b)0.5;(c)1.5;and (d)

2.5.

Fig.11.Tangential velocity (m/s)distributions at the central section vertical to the cyclone inlet under different solid loading ratios:(a)0.0;(b)0.5;(c)1.5;and (d)

2.5.

Fig.12.Axial velocity (m/s)distributions at the central section vertical to the cyclone inlet under different solid loading ratios:(a)0.0;(b)0.5;(c)1.5;and (d)

2.5.

Fig.13.Radial velocity distributions at the central section vertical to the cyclone inlet under different solid loading ratios:(a)0.0;(b)0.5;(c)1.5;and (d)2.5.

K.W.Chu et al./Chemical Engineering Science 66(2011)834–847

840

the signi?cant decrease in the tangential velocity.Moreover,Fig.15shows that the total tangential kinetic energy is much larger than the axial and radial kinetic energies and thus https://www.wendangku.net/doc/b212635693.html,paring Figs.3and 15suggests that the relationship between the pressure drop and solid loading ratio is very similar to that between the tangential velocity of the gas phase and solid loading ratio.Thus,the decrease of the pressure drop is strongly related to the decrease of the tangential velocity.

It should be noted that the pressure drop and tangential velocity may not always decrease with the increase of solid loading ratio in certain extreme cases,especially when particles are extremely ?ne (Bricout and Louge,2004).When a large portion of particles fed at the inlet are too ?ne to be collected in the spigot of the cyclone,the

majority of particles would not ?ow as strands in the region close to the outer wall of the cyclone.Instead,they would ?ow in the inner region of the cyclone and ?nally escape through the vortex ?nder.In this situation,the impact of particles in the inner region on gas ?ow may be different from that in the region close to the outer wall.Therefore,the effect of solids on gas ?ow obtained in this work is only applicable to the situation where particles are large enough to be collected at the spigot through the cyclone outer wall,which is important since about 95%of particles can be collected at under-?ow in most of the operations of gas cyclones.4.3.Forces governing the motion of particles

According to the model framework of the current work,the motion of particles in a cyclone is governed by the particle–?uid,particle–particle and particle–wall interaction forces.In this sec-tion,the three forces are examined to better understand the nature of the ?ow in a cyclone.

4.3.1.Particle–?uid interaction force

Two particle–?uid forces,i.e.,the gas drag force and pressure gradient force (PGF)are considered in this work.Fig.16shows the

total (P N p i ?19f i 9)and averaged (e1=N p TP N p

i ?19f i 9)forces acting by gas phase on particles for different solid loading ratios,where N p is the total number of particles residing in the cyclone.For conve-nience,the particle–particle and particle–wall interaction forces are also shown in this ?gure.The forces on a particle are all normalized by dividing the gravity force of the particle.Fig.16

(a)

Fig.14.Pressure (I)and tangential velocity (II)distributions at different sections (de?ned in Fig.1):(a)pure gas ?ow and (b)gas–solids ?ow when solid loading ratio is 2.5.

Fig.15.Kinetic energy of gas phase at a macroscopically steady state as a function of solid loading ratio.

K.W.Chu et al./Chemical Engineering Science 66(2011)834–847841

shows that both the total drag force and PGF increase with solid loading ratio but Fig.16(b)shows that the average values decrease slightly with the increase of solid loading ratio.The increase of the total forces should be due to the increased mass of solids residing in the cyclone (see Fig.8).The decrease of the average drag force and PGF should correspond to the decrease of the tangential velocity (see Figs.4and 11)and the pressure drop (see Figs.3and 10),respectively.Fig.16also shows the magnitude of the drag force is much larger than that of the PGF,suggesting the drag force is more dominant for this ?ow system.

The spatial distributions of the drag force and PGF acting on individual particles,when the solid loading ratio is 2.5,are shown in Figs.17and 18,respectively.Both ?gures show that the spatial distributions of the two forces are not uniform.They are higher in the inlet region of the cyclone and there are two separating regions in the strands:the regions close to the cyclone wall have lower drag forces and PGFs while the regions away from the wall have higher drag forces and PGFs.The drag force on particles is low because of the no-slip condition between the cyclone wall and gas phase,which may lead to low gas velocity and hence low drag force.The PGF is relatively low close to the cyclone wall because the pressure does not decrease signi?cantly near the wall regions but decreases signi?cantly in the centre regions (from positive to negative),as shown in Fig.10.PGF is proportional to the pressure gradient,as shown in Table 1.

In this work,the ?uid drag force is the only force considered that is related to the gas velocity.Therefore,it is expected that the distribution of the ?uid drag force will follow the trend of the velocity of the gas ?ow.This is indeed con?rmed by Fig.17,which shows that the ?uid drag force directs dominantly in the tangential direction.As shown in Figs.12–15,gas ?ows mainly tangentially since its tangential velocity is much larger than the axial and radial

velocities.On the other hand,Fig.18shows that the direction of the PGF is almost vertical to that of the drag force,pointing from the wall to the centre of the cyclone.This agrees with the radial

pressure

Fig.16.Normalized total (a)and averaged (b)forces in the cyclone under different solid loading ratios:line I,particle–particle interaction force;line II,particle–wall interaction force;line III,drag force;and line IV,pressure gradient

force.

Fig.17.Spatial distribution of the gas drag forces acting on particles in the cyclone when solid loading ratio is

2.5.

Fig.18.Spatial distribution of the pressure gradient forces acting on particles in the cyclone when solid loading ratio is 2.5.

K.W.Chu et al./Chemical Engineering Science 66(2011)834–847

842

distribution where the pressure decreases gradually from the wall to the centre of the cyclone (see Fig.10).

According to the current model framework,the forces acting on particles (f p àf ,i )by gas phase are at a particle scale and the reaction forces acting on gas phase (V cell F f àp )by particles are at a CFD cell scale,while they obey Newton’s third law of motion,i.e.

F p àf ?e1=V cell TP k c

i ?1f p àf ,i .This force is shown in Fig.19.Two phenomena can be observed from the ?gure:The ?rst one is that the reaction force of particles on gas phase increases with solid loading ratio in the upper part but decreases generally in the lower part of the cyclone.This is because the magnitude of the reaction force depends on two factors:one is the magnitude of ?uid force on particles and the other one is the number of particles in a CFD cell or solid concentration.When solid loading ratio is low,the reaction force is small in the upper part of the cyclone since the solid concentration is low.When solid loading ratio is high,the reaction force is small in the lower part of the cyclone since the gas velocity and pressure gradient are relatively low in that part (see Figs.10–13).The other phenomenon is that most of the vectors are shown in red,which means that the reaction force acted by particles on gas phase is generally upward.The presence of solids in the cyclone will produce a resistant force to the downward ?ow of gas.

Fig.20shows the vectors of the reaction forces of particles on gas phase in a horizontal plane when solid loading ratio is 2.5.From Fig.20(a)it can be seen that the reaction force points upward,which means particles will prevent gas from ?owing downward.From Fig.20(b)it can be seen that the reaction force mainly points tangentially,because the ?uid drag force is mainly in the tangential direction and the PGF is much smaller than the drag force even though PGF mainly points radially.According to Newton’s third law of motion,this tangential force is in the reverse direction of the tangential velocity of gas phase.Thus,according to Newton’s second low,this force will decelerate the tangential ?ow of gas phase,which may explain why the tangential velocity of gas phase decreases signi?cantly after loading solids (see Figs.3and 10).When the tangential velocity of the gas phase decreases,the pressure drop will decrease correspondingly.

Yuu et al.(1978)suggested that the decrease of the tangential velocity is caused by the increase of friction due to the movement of particles attaching to the outer cylinder wall.They coated a layer of particles on the cylinder wall,making it rough and again measured the tangential velocity,and found that the decrease of the tangential velocity under rough outer cylinder wall condition was not as great as that under gas–solid ?ow condition.In our work,the decrease of the tangential velocity of gas phase after loading solids is captured,even though the effect of the presence of solids on the friction between gas and the wall is not considered.Thus,it is thought that the decrease of the tangential velocity of gas phase after loading solids is mainly due to the reaction force of particles on gas phase in the current conditions.Trefz and Muschelknautz (1993)has described this phenomenon as that ‘‘the exchange of tangential momentum between the slowly slipping wall layer of solids and the gas decelerates the vortex ?ow’’while they did not provide any

evidence.

Fig.19.Spatial distributions of the reaction force of particles on gas phase under different solid loading ratios:(a)0.5;(b)1.5;and (c)2.5.The vectors are coloured by the axial particle–?uid volumetric

force.

Fig.20.Vectors of the reaction force of particles on gas phase on a horizontal plane when solid loading ratio is 2.5:(a)front view and (b)top view.The force is àF fp term in Eq.(4).

K.W.Chu et al./Chemical Engineering Science 66(2011)834–847843

4.3.2.Particle–particle interaction force

Particle–particle interaction relates to particle breakage or attrition,which may be important when separating solid products from gas.The information regarding the particle–particle interac-tion at different solid loading ratios is shown in Figs.16and 21.Fig.16clearly shows that both the total and averaged particle–particle interaction forces are much larger than the other three forces,suggesting that the inclusion of particle–particle interaction is important in the modelling of gas–solid ?ow in gas cyclones.Moreover,this ?gure shows that the magnitudes of both the total and averaged particle–particle interaction forces increase with solid loading ratio.This is because the solid concentration in the cyclone increases with solid loading ratio.When solid concentra-tion is high,the chance for particle–particle interaction is high.

Fig.21shows the spatial distribution of particle–particle inter-actions.It can be seen that there are strong particle–particle interactions inside the strands of the solids.Outside the strands,there are very few particles in red,which means the particle–particle interaction force is quite small there.Fig.21(a)also demonstrates that the particle–particle interaction in the upper part of the cyclone is not strong when solid loading ratio is low,because a strand is not clearly formed there.Fig.21(d)shows the transient variation of the total particle–particle interaction force with time.It can be seen that the total particle–particle interaction forces increase from zero to a maximum value and then ?uctuates around a constant.The trends are different when the solid loading ratio is different.As solid loading ratio increases,the total particle–particle interaction force reaches its stable state quicker and the magnitude of both the ?uctuation and the absolute value become larger.Those trends look similar to that of the pressure drop as shown in Fig.8.

4.3.3.Particle–wall interaction force

Particle–wall interaction force in cyclones relates to the wearing of cyclone wall which could be a serious problem for cyclones operating at high solid loading ratios.For example,it is observed in practice that the apex of a cyclone in the preheating system in cement industry can be seriously worn out.The information of

particle–wall interaction is shown in Figs.16and 22.Fig.16shows that the magnitude of the particle–wall interaction is slightly lower than the particle–particle interaction,but higher than the particle–?uid forces.This suggests that particle–wall interaction is an important factor in the modelling of gas cyclones.Moreover,Fig.16(a)shows the total particle–wall interaction force increases with solid loading ratio but Fig.16(b)shows that the averaged particle–wall interaction force decreases with the increase of solid loading ratio.The increase of the total particle–wall interaction should be due to the increase of the solids residing in the cyclone,as shown in Fig.8.The decrease of the averaged particle–wall interaction force may be caused by the ‘‘shielding’’effect of particles.Not all particles can collide with the wall since some particles only collide with the particles bouncing back after colliding with the wall.

In this work,the spatial distribution of the particle–wall interaction is quanti?ed by use of the so-called Time Averaged Collision Intensity (TACI)which is related to wear (Finnie,1972),de?ned by TACI ?

P t ?T t ?0

P k m

i ?1

9f cn ,i tf dn ,i tf ct ,i tf dt ,i 9e5T

where A is the surface area of a sample wall,T is the simulation time or sampling time,k m is the number of particles making contact with the sample wall at a given time.Physically,it can be understood as the particle–wall interaction forces per unit (wall surface)area per unit time.

Fig.22shows the spatial distribution of TACI between particles and wall for different solid loading ratios.Generally speaking,the high TACI regions mainly establish on the wall opposite to the inlet of the cyclone and the conic and apex wall.The TACI on the wall opposite to the inlet is high because particles enter the cyclone from the inlet and then heavily collide with the cyclone wall where they change their ?ow directions signi?cantly.The conic wall gets high TACI because it constricts the diameter of particle orbits by particle–wall contacts.Fig.22also shows that the distribution of TACI varies with solid loading ratio,especially on the conic part

of

Fig.21.Spatial distributions of particle–particle interaction force (normalized by dividing particle weight)under different solid loading ratios:(a)0.5;(b)1.5;and (c)2.5,and (d)transient variation of the total particle–particle interaction forces with time.

K.W.Chu et al./Chemical Engineering Science 66(2011)834–847

844

the cyclone.As solid loading ratio increases,the area of the cyclone wall experiencing particle–wall interaction decreases while the intensity of the particle–wall interaction increases.

Fig.22(d)shows the transient variation of the total particle–wall interaction force with time.It can be seen that the total particle–wall interaction forces increase from zero to a maximum value and then ?uctuates around a constant.The trends are different for different solid loading ratios.As solid loading ratio increases,the total particle–wall interaction force reaches its stable state quicker and the magnitude of both the ?uctuation and the absolute value become larger.Those trends are similar to that of the pressure drop (Fig.8)and the total particle interaction force (Fig.21(d)).None-theless,an obvious difference is that the magnitude of the total particle–wall interaction force is not so sensitive to solid loading ratio.It suggests that the wear may be improved under high solid loading conditions for a given amount of solid to handle.

5.Conclusions

A CFD–DEM model has been developed to describe the gas–solid ?ow in a gas cyclone.Its validity is demonstrated by its successful capturing the key ?ow features in a cyclone separator such as the ?ow pattern of particles along the cylinder wall in strands,and the decrease of pressure drop and tangential velocity after loading solids.The effect of solid loading ratio has been analysed on this basis,leading to the following conclusions:

As the solid loading ratio increases,there are less turns travelled

by solids,especially in the apex region of a cyclone,and the width of the strands increases.The total mass of solids residing in the cyclone does not linearly increase with solid loading ratio.More solids tend to accumulate in the cyclone when the solid loading ratio is low.

The gas pressure drop ?rstly increases and then decreases to reach a stable value when solids are loaded.When the ?ow reaches a macroscopically steady ?ow state,the pressure

drop and the tangential velocity of gas phase decrease steadily with the increase of solid loading ratio.At the same time,the high axial velocity region moves upwards and the radial ?ow of gas phase is signi?cantly dampened,especially in the apex region.

In general,the magnitudes of the particle–particle and particle–wall interaction forces are much larger than that of the particle–?uid force.On the other hand,the magnitude of the drag force is larger than the PGF.The drag force is mainly in the tangential direction but the PGF is mainly in the radial direction.The reaction force of particles on gas ?ow is mainly in the tangential direction and directs mainly upward in the axial direction.The reaction force in the tangential direction will decelerate gas phase and the upward axial force will prevent gas phase from ?owing downward in the near wall region.Both the total and averaged particle–particle interaction forces increase with the increase of solid loading ratio.The most intensive particle–particle interaction regions locate within the particle strands.The total particle–wall interaction increases but the averaged particle–wall interaction decreases with solid loading ratio.The intensive particle–wall collision regions mainly locate in the wall opposite to the cyclone inlet and the cone wall.When solid loading ratio increases,the area of the cyclone wall experiencing particle–wall interaction decreases but the intensity of the particle–wall interaction increases.

The ?ndings should be useful to the development of a compre-hensive understanding of the gas–solid ?ow in cyclones.However,it should be pointed out that the present study is carried out for large,cohesionless and spherical particles because of the current computer capability.Thus,the present study is largely qualitative and for fundamental understanding,while showing the capability of the CFD–DEM approach applied to gas cyclone.Therefore,more detailed,systematic studies are necessary in order to understand the effects of variables related to operational conditions,particle and material properties,and cyclone geometry,and hence generate results useful to engineering

application.

Fig.22.Spatial distributions of particle–wall Time Averaged Collision Intensity under different solid loading ratios:(a)0.5;(b)1.5;and (c)2.5,and (d)transient variation of the total particle–wall interaction forces with time.

K.W.Chu et al./Chemical Engineering Science 66(2011)834–847845

Nomenclature

A area,m2

c damping coef?cient,dimensionless

d particl

e diameter,m

E Young’s modulus,Pa

f c contact force,N

f d dampin

g force,N

f pàf particle–?uid interaction force,N

F volumetric force,N/m3

g gravity acceleration vector,9.81m/s2

G gravity vector,N

I moment of inertia of a particle,kg m

k cell number of particles in a computational cell, dimensionless

k i number of particles in contact with particle i, dimensionless

k m number of contacts in a sample,dimensionless

m mass of a particle,kg

M rolling friction torque,N m

n unit vector in the normal direction of two contact spheres, dimensionless

N p particle number,dimensionless

P pressure,Pa

D P pressure drop,Pa

R radius vector(from particle centre to a contact point),m R magnitude of R,m

Re Reynolds number,dimensionless

t time,s

T total simulation time,s

T driving friction torque,N m

u mean?uid velocity vector,m/s

u0?uctuating?uid velocity vector,m/s

v particle velocity,m/s

V volume,m3

V velocity vector,m/s

Greek letters

b empirical coef?cient de?ned in Table2,dimensionless d vector of the particle–particle or particle–wall overlap,m

d magnitud

e o

f d,m

e porosity,dimensionless

m f?uid viscosity,kg/m/s

m r coef?cient of rolling friction,m

m s coef?cient of sliding friction,dimensionless

n Poisson’s ratio,dimensionless

r density,kg/m3

s viscous stress tensor,N/m3

x angular velocity,rad/s

o magnitude of angular velocity,rad/s

^x unit angular velocity

Subscripts

c contact

cell computational cell

d damping

D drag

f?uid phase

ij between particle i and j

i(j)corresponding to i(j)th particle

max maximum

n in normal direction p particle phase

pg pressure gradient

pàf between particle and?uid phases

t in tangential direction

Acknowledgement

The authors are grateful to the Australia Research Council(ARC) for the?nancial support of this work.

References

Anderson,T.B.,Jackson,R.,1967.A?uid mechanical description of?uidized beds.

Industrial&Engineering Chemistry Fundamentals6,527–539.

Boysan,F.,Ayers,W.H.,Swithenbank,J.,1982.Fundamental mathematical-model-ling approach to cyclone design.Transactions of the Institution of Chemical Engineers V60,222–230.

Bricout,V.,Louge,M.Y.,2004.Measurements of cyclone performance under conditions analogous to pressurized circulating?uidization.Chemical Engi-neering Science59,3059–3070.

Chu,K.W.,Wang,B.,Vince,A.,Yu,A.B.,Barnett,G.D.,Barnett,P.J.,2009a.CFD–DEM study of the effect of particle density distribution on the multiphase?ow and performance of dense medium cyclone.Minerals Engineering22,893–909. Chu,K.W.,Wang,B.,Yu,A.B.,Vince,A.,2009b.CFD–DEM modelling of multiphase ?ow in dense medium cyclones.Powder Technology193,235–247.

Chu,K.W.,Yu,A.B.,2008a.Numerical simulation of complex particle–?uid?ows.

Powder Technology179,104–114.

Chu,K.W.,Yu,A.B.,2008b.Numerical simulation of the gas–solid?ow in three-dimensional pneumatic conveying bends.Industrial&Engineering Chemistry Research47,7058–7071.

Cortes,C.,Gil,A.,2007.Modeling the gas and particle?ow inside cyclone separators.

Progress in Energy and Combustion Science33,409–452.

Crowe,C.T.,Pratt,D.T.,1974.Analysis of the?ow?eld in cyclone separators.

Computers&Fluids2,249–260.

Cundall,P.A.,Strack,O.D.L.,1979.Discrete numerical-model for granular assem-blies.Geotechnique29,47–65.

Derksen,J.J.,Sundaresan,S.,van den Akker,H.E.A.,2006.Simulation of mass-loading effects in gas–solid cyclone separators.Powder Technology163,59–68. Derksen,J.J.,van den Akker,H.E.A.,Sundaresan,S.,2008.Two-way coupled large-eddy simulations of the gas–solid?ow in cyclone separators.AIChE Journal54, 872–885.

Di Renzo,A.,Di Maio,F.P.,2007.Homogeneous and bubbling?uidization regimes in DEM-CFD simulations:hydrodynamic stability of gas and liquid?uidized beds.

Chemical Engineering Science62,116–130.

Enwald,H.,Peirano,E.,Almstedt,A.E.,1996.Eulerian two-phase?ow theory applied to?uidization.International Journal of Multiphase Flow22,21–66. Fassani,F.L.,Goldstein,L.,2000.A study of the effect of high inlet solids loading on a cyclone separator pressure drop and collection ef?ciency.Powder Technology 107,60–65.

Finnie,I.,1972.Some observations on erosion of ductile metals.Wear19,81–91. Gidaspow,D.,1994.Multiphase Flow and Fluidization.Academic Press.

Gui,N.,Fan,J.,Cen,K.,2009.Effect of local disturbance on the particle-tube collision in bubbling?uidized bed.Chemical Engineering Science64,3486–3497. Hilton,J.E.,Mason,L.R.,Cleary,P.W.,2010.Dynamics of gas–solid?uidised beds with non-spherical particle geometry.Chemical Engineering Science65,1584–1596. Hoekstra,A.J.,Derksen,J.J.,Van Den Akker,H.E.A.,1999.An experimental and numerical study of turbulent swirling?ow in gas cyclones.Chemical Engineer-ing Science54,2055–2065.

Hoffmann,A.C.,Vansanten,A.,Allen,R.W.K.,Clift,R.,1992.Effects of geometry and solid loading on the performance of gas cyclones.Powder Technology70,83–91. Hu,L.Y.,Zhou,L.X.,Zhang,J.,Shi,M.X.,2005.Studies on strongly swirling?ows in the full space of a volute cyclone separator.AIChE Journal51,740–749.

Ibsen,C.H.,Helland, E.,Hjertager,B.H.,Solberg,T.,Tadrist,L.,Occelli,R.,2004.

Comparison of multi?uid and discrete particle modelling in numerical predic-tions of gas particle?ow in circulating?uidised beds.Powder Technology149,29. Kafui,K.D.,Thornton, C.,Adams,M.J.,2002.Discrete particle-continuum?uid modelling of gas–solid?uidised beds.Chemical Engineering Science57, 2395–2410.

Kawaguchi,T.,Tanaka,T.,Tsuji,Y.,1998.Numerical simulation of two-dimensional ?uidized beds using the discrete element method(comparison between the two-and three-dimensional models).Powder Technology96,129–138. Kuang,S.B.,Chu,K.W.,Yu, A.B.,Zou,Z.S.,Feng,Y.Q.,https://www.wendangku.net/doc/b212635693.html,putational investigation of horizontal slug?ow in pneumatic conveying.Industrial& Engineering Chemistry Research47,470–480.

Li,S.H.,Yang,H.R.,Zhang,H.,Yang,S.,Lu,J.F.,Yue,G.X.,2009.Measurements of solid concentration and particle velocity distributions near the wall of a cyclone.

Chemical Engineering Journal150,168–173.

Li,Y.,Zhang,J.P.,Fan,L.S.,1999.Numerical simulation of gas–liquid–solid?uidiza-tion systems using a combined CFD–VOF–DPM method:bubble wake behavior.

Chemical Engineering Science54,5101–5107.

K.W.Chu et al./Chemical Engineering Science66(2011)834–847 846

Limtrakul,S.,Boonsrirat,A.,Vatanatham,T.,2004.DEM modeling and simulation of

a catalytic gas–solid?uidized bed reactor:a spouted bed as a case study.

Chemical Engineering Science59,5225–5231.

Malone,K.F.,Xu,B.H.,2008.Particle-scale simulation of heat transfer in liquid-?uidised beds.Powder Technology184,189–204.

Meier,H.F.,Mori,M.,1998.Gas–solid?ow in cyclones:the Eulerian–Eulerian https://www.wendangku.net/doc/b212635693.html,puters&Chemical Engineering22,S641–S644. Muschelknautz,E.,Greif,V.,1997.Circulating Fluidized Beds.Blackie,London. Pant,K.,Crowe,C.T.,Irving,P.,2002.On the design of miniature cyclones for the collection of bioaerosols.Powder Technology125,260–265.

Qian,F.,Huang,Z.,Chen,G.,Zhang,M.,2007.Numerical study of the separation characteristics in a cyclone of different inlet particle https://www.wendangku.net/doc/b212635693.html,puters &Chemical Engineering31,1111–1122.

Rhodes,M.J.,Wang,X.S.,Nguyen,M.,Stewart,P.,Liffman,K.,https://www.wendangku.net/doc/b212635693.html,e of discrete element method simulation in studying?uidization characteristics:in?uence of interparticle force.Chemical Engineering Science56,69–76.

Rong,D.G.,Horio,M.,2001.Behavior of particles and bubbles around immersed tubes?uidized bed at high temperature and pressure:a DEM simulation.

International Journal of Multiphase Flow27,89–105.

Slack,M.D.,Prasad,R.O.,Bakker,A.,Boysan,F.,2000.Advances in cyclone modelling using unstructured grids.Chemical Engineering Research&Design78, 1098–1104.

Trefz,M.,Muschelknautz,E.,1993.Extended cyclone theory for gas?ows with high solids concentrations.Chemical Engineering&Technology16,153–160. Tsuji,Y.,2007.Multi-scale modeling of dense phase gas–particle?ow.Chemical Engineering Science62,3410–3418.

Tsuji,Y.,Tanaka,T.,Ishida,T.,https://www.wendangku.net/doc/b212635693.html,grangian numerical-simulation of plug?ow of cohesionless particles in a horizontal pipe.Powder Technology71,239–250. Wan,G.J.,Sun,G.G.,Xue,X.H.,Shi,M.X.,2008.Solids concentration simulation of different size particles in a cyclone separator.Powder Technology183,94–104. Wang,B.,Chu,K.W.,Yu,A.B.,2007.Numerical study of particle–?uid?ow in a hydrocyclone.Industrial&Engineering Chemistry Research46,4695–4705.Wang,B.,Xu,D.L.,Chu,K.W.,Yu,A.B.,2006.Numerical study of gas–solid?ow in a cyclone separator.Applied Mathematical Modelling30,1326–1342.

Xu,B.H.,Yu,A.B.,1997.Numerical simulation of the gas–solid?ow in a?uidized bed by combining discrete particle method with computational?uid dynamics.

Chemical Engineering Science52,2785–2809.

Xu,B.H.,Yu,A.B.,Chew,S.J.,Zulli,P.,2000.Numerical simulation of the gas–solid ?ow in a bed with lateral gas blasting.Powder Technology109,13–26. Yoshida,H.,1996.Three-dimensional simulation of air cyclone and particle separation by a revised-type cyclone.Colloids and Surfaces A—Physico-chemical and Engineering Aspects109,1–12.

Yu,A.B.,Xu,B.H.,2003.Particle-scale modelling of gas–solid?ow in?uidisation.

Journal of Chemical Technology and Biotechnology78,111–121.

Yuu,S.,Jotaki,T.,Tomita,Y.,Yoshida,K.,1978.Reduction of pressure-drop due to dust loading in a conventional cyclone.Chemical Engineering Science33, 1573–1580.

Zhao,Y.Z.,Jiang,M.Q.,Liu,Y.L.,Zheng,J.Y.,2009.Particle-scale simulation of the?ow and heat transfer behaviors in?uidized bed with immersed tube.AIChE Journal 55,3109–3124.

Zhou,Y.C.,Wright,B.D.,Yang,R.Y.,Xu,B.H.,Yu,A.B.,1999.Rolling friction in the dynamic simulation of sandpile formation.Physica A—Statistical Mechanics and its Applications269,536–553.

Zhou,Y.C.,Yu,A.B.,Stewart,R.L.,Bridgwater,J.,2004.Microdynamic analysis of the particle?ow in a cylindrical bladed mixer.Chemical Engineering Science59, 1343–1364.

Zhou,Z.Y.,Kuang,S.B.,Chu,K.W.,Yu,A.B.,2010.Assessments of CFD–DEM models in particle–?uid?ow modelling.Journal of Fluid Mechanics661,482–510. Zhou,Z.Y.,Pinson, D.,Zou,R.P.,Yu, A.B.,2009.CFD–DEM simulation of gas ?uidization of ellipsoidal particles.In:Proceedings of the Seventh International Conference on CFD in the Minerals and Process Industries CSIRO,Melbourne, Australia,9–11December2009,Paper No.201Zho(CD Rom Edition).

Zhu,H.P.,Zhou,Z.Y.,Yang,R.Y.,Yu,A.B.,2007.Discrete particle simulation of particulate systems:theoretical developments.Chemical Engineering Science 62,3378–3396.

K.W.Chu et al./Chemical Engineering Science66(2011)834–847847

第七 章 CFD仿真模拟

第七章CFD仿真模拟 一．初识CFD CFD是英文Computational Fluid Dynamics（计算流体动力学）的简称。它是伴随着计算机技术、数值计算技术的发展而发展的。简单地说，CFD相当于"虚拟"地在计算机做实验，用以模拟仿真实际的流体流动情况。而其基本原理则是数值求解控制流体流动的微分方程，得出流体流动的流场在连续区域上的离散分布，从而近似模拟流体流动情况。可以认为CFD是现代模拟仿真技术的一种。 1933年，英国人Thom首次用手摇计算机数值求解了二维粘性流体偏微分方程，CFD由此而生。1974年，丹麦的Nielsen首次将CFD用于暖通空调工程领域，对通风房间内的空气流动进行模拟。之后短短的20多年内，CFD技术在暖通空调工程中的研究和应用进行得如火如荼。如今，CFD技术逐渐成为广大空调工程师和建筑师解决分析工程问题的有力工具。 二．为什么用CFD CFD是一种模拟仿真技术，在暖通空调工程中的应用主要在于模拟预测室内外或设备内的空气或其他工质流体的流动情况。以预测室内空气分布为例，目前在暖通空调工程中采用的方法主要有四种：射流公式，Zonal model，CFD以及模型实验。 由于建筑空间越来越向复杂化、多样化和大型化发展，实际空调通风房间的气流组织形式变化多样，而传统的射流理论分析方法采用的是基于某些标准或理想条件理论分析或试验得到的射流公式对空调送风口射流的轴心速度和温度、射流轨迹等进行预测，势必会带来较大的误差。并且，射流分析方法只能给出室内的一些集总参数性的信息，不能给出设计人员所需的详细资料，无法满足设计者详细了解室内空气分布情况的要求； Zonal model是将房间划分为一些有限的宏观区域，认为区域内的相关参数如温度、浓度相等，而区域间存在热质交换，通过建立质量和能量守恒方程并充分考虑了区域间压差和流动的关系来研究房间内的温度分布以及流动情况，因此模拟得到的实际上还只是一种相对"精确"的集总结果，且在机械通风中的应用还存在较多问题； 模型实验虽然能够得到设计人员所需要的各种数据，但需要较长的实验周期和昂贵的实验费用，搭建实验模型耗资很大，有文献指出单个实验通常耗资3000～20000美元，而对于不同的条件，可能还需要多个实验，耗资更多，周期也长达数月以上，难于在工程设计中广泛采用。 另一方面，CFD具有成本低、速度快、资料完备且可模拟各种不同的工况等独特的优点，故其逐渐受到人们的青睐。由表1给出的四种室内空气分布预测方法的对比可见，就目前的三种理论预测室内空气分布的方法而言，CFD方法确实具有不可比拟的优点，且由于当前计算机技术的发展，CFD方法的计算周期和成本完全可以为工程应用所接受。尽管CFD方法还存在可靠性和对实际问题的可算性等问题，但这些问题已经逐步得到发展和解决。因此，CFD方法可应用于对室内空气分布情况进行模拟和预测，从而得到房间内速度、温度、湿度以及有害物浓度等物理量的详细分布情况。 进一步而言，对于室外空气流动以及其它设备内的流体流动的模拟预测，一般只有模型实验或CFD方法适用。表1的比较同样表明了CFD方法比模型实验的优越性。故此，CFD方法可作为解决暖通空调工程的流动和传热传质问题的强有力工具而推广应用。 表1四种暖通空调房间空气分布的预测方法比较 比较项目 1射流公式 2 ZONAL MODEL 3CFD 4模型实验 房间形状复杂程度简单较复杂基本不限基本不限 ?对经验参数的依赖性几乎完全很依赖一些不依赖

医药工业洁净厂房设计要求规范

医药工业洁净厂房设计规范 第一章总则 第1.0.1条为了贯彻执行国家《药品生产质量管理规范》（以下简称ＧＭＰ），提出符合ＧＭＰ要求的生产厂房、设施及设备的设计要求，特制订本规范。 第1.0.2条本规范适用于新建、改建和扩建的医药制剂、原料药和药用辅料的精制、干燥、包装工序，直接接触药品的药用包装材料、无菌医疗器械等医药工业洁净厂房的设计。 第1.0.3条医药工业洁净厂房诉设计必须贯彻国家有关方针、政策。做到技术先进、确保质量、安全实用、经济合理，符合节约能源和保护环境的要求。 第1.0.4条医药工业洁净厂房的设计，既要满足当前产品生产的工艺要求，也应适当考虑今后生产发展和工艺改进的需要。 第1.0.5条在利用原有建筑和设施进行洁净技术改造时，可根据生产工艺要求，从实际出发，充分利用现有的技术设施，符合因地制宜的原则。 第1.0.6条医药工业洁净厂房的设计应为施工安装、维护、管理、检修、测试和安全运行创造必要的条件。 第1.0.7条医药工业洁净厂房的设计除应执行本规范外，还应符合现行的国家标准、规范和规定的有关要求。 第二章生产区域的环境参数 第一节一般规定 第2.1.1条为了保证医药产品生产质量，防止生产环境对产品的污染，生产区域必须满足规定的环境参数标准。

注2：空气洁净度的测试以静态条件为依据，测试方法应符合国家医药管理工业洁净室和洁净区悬浮粒子的测试方法》中有关规定； 注3：对于空气洁净度为100级的洁净室，室内大于等于5μm尘粒的计数，应进行多次采样，当其多次出现时，方可认为该测试数值是可靠的。 第2．2．2条药品生产有关工序和环境区域的空气洁净度等级按国家CMP等有关规定确定。 第2．2．3条洁净室内的温度和湿度应符合下列规定： 一、生产工艺对温度和湿度无特殊要求时，以穿着洁净工作服不产生不舒服感为宜。空气洁净度100级、10000级区域一般控制温度为20～24℃，相对湿度为45～60%。100000级区域一般控制温度为18～28℃，相对湿度为50～65%。 二、生产工艺对温度和湿度有特殊要求时，应根据工艺要求确定。 第2．2．4条洁净室内应保持一定的新鲜空气量，其数值应取下列风量中的最大值： 一、非单向流洁净室总送风量的10～30%，单向流洁净室总送风量的2～4%； 二、补偿室内排风和保持室内正压值所需的新鲜空气量； 三、保证室内每人每小时的新鲜空气量不小于40m3。 第2．2．5条洁净室必须维持一定的正压。不同空气洁净度的洁净区之间以及洁净区与非洁净区之间的静压差不应小于5Pa,洁净区与室外的静压差不应小于10Pa。 青霉素等特殊药物生产洁净区，固体口服制剂配料、制粒、压片等工序洁净区的气压控制，应符合第8.5.1条要求。 第2．2．6条洁净室和洁净区应根据生产要求提供足够的照度。主要工作室一般照明的照度值不宜低于300LX；辅助工作室、走廊、气闸室、人员净化和物料净化用室可低于300LX，但不宜低于150LX。对照度要求高的部位可增加局部照明。

一维CFD模拟仿真设计

CFD simulation in Laval nozzle SIAE 090441313 Abstract We aim to simulate the quasi one dimension flow in the Laval nozzle based on CFD computation in this paper .We consider the change of the temperature ,the pressure ,the density and the speed of the flow to study the flow.The analytic solution of the flow in the Laval nozzle is provided when the input velocity is supersonic.We use the Mac-Cormack Explicit Difference Scheme to slove the question. Key words :Laval nozzle ,CFD,throat narrow. Contents Abstract .................................................. . (1) Introduction .............................................. .. (2) Simulation of one-dimensional steady flow (3)

Basis equations ................................................. (3) Dimensionless .......................................... . (10) Mac -Cormack Explicit Difference Scheme (11) Boundary conditions ................................................ (13) Reference .............................................. (13) Annex .................................................. .. (14) Introduction Laval nozzle is the most commonly used components of rocket engines and aero-engine, constituted by two tapered tube, one shrink tube, another expansion tube. Laval nozzle is an important part of the thrust chamber. The first half of the nozzle from large to small contraction to a narrow throat to the middle. Narrow throat and then expand

洁净厂房设计规范2019

1 总则 1.0.1为在医药工业洁浄厂房设计中贯彻执行国家有关方针政策,做到技术先进、安全可靠、确保质量、节能环保,制定本标准 1.0.2本标准适用于新建、扩建和改建的医药工业洁净厂房设计。生物制品、毒性药品、精神药品、麻醉药品以及放射性药品的生产和质量检验设施除应执行本标准外,尚应符合国家有关的监管规定。 1.0.3医药工业洁净厂房的设计应为施工安装、系统设施验证、维护管理、检修测试和安全运行创造必要的条件 1.0.4医药工业洁净厂房的设计除应符合本标准外,尚应符合国家现行有关标准规范的规定。 2术语 2.0.1医药洁净室pharmaceutical clean room 空气悬浮粒子和微生物浓度,以及温度、湿度、压力等参数受控的医药生产房间或限定的空间。

2.0.2医药工业洁净厂房pharmaceutical industry clean room 包含医药洁净室的用于药品生产及质量控制的建筑物 2.0.3人员净化用室room forcleaning personne 人员在进人医药洁净室之前按一定程序进行净化的房间 2.0.4物料净化用室room forcleaning materia 物料在进入医药洁净室之前按一定程序进行净化的房间 2.0.5受控环境controlled environment 以规定方法对污染源进行控制的特定区域 2.0.6悬浮粒子airborne particles 用于空气洁净度分级的空气悬浮粒子尺寸范围在0.1um~1000um 的固体和液体粒子。 2.0.7微生物microorganIsms 能够复制或传递基因物质的细菌或非细菌的微小生物实体 2.0.8含尘浓度particle concentration 单位体积空气中悬浮粒子的数量

CFD仿真验证及有效性指南

CFD仿真验证及有效性指南 摘要 本文提出评估CFD建模和仿真可信性的指导方法。评估可信度的两个主要原则是：验证和有效。验证，即确定计算模拟是否准确表现概念模型的过程，但不要求仿真和现实世界相关联。有效，即确定计算模拟是否表现真实世界的过程。本文定义一些重要术语，讨论基本概念，并指定进行CFD仿真验证和有效的一般程序。本文目的在于提供验证和有效的重要问题和概念的基础，因为一些尚未解决的重要问题，本文不建议作为该领域的标准。希望该指南通过建立验证和有效的共同术语和方法，以助于CFD仿真的研究、发展和使用。这些术语和方法也可用于其他工程和科学学科。 前言 现在，使用计算机模拟流体的流动过程，用于设计，研究和工程系统的运行，并确定这些系统在不同工况下的性能。CFD模拟也用于提高对流体物理和化学性质的理解，如湍流和燃烧，有助于天气预报和海洋。虽然CFD模拟广泛用于工业、政府和学术界，但目前评估其可信度的方法还很少。这些指导原则基于以下概念，没有适用于所有CFD模拟的固定的可信度和精确度。模拟所需的精确度取决于模拟的目的。 建立可信度的两个主要原则是验证和有效（V&V）。这里定义，验证即确定模型能准确表现设计者概念模型的描述和模型解决方案的过程，有效即确定预期模型对现实世界表现的准确度的过程。该定义表明，V&V的定义还在变动，还没有一个明确的最终定义。通常完成或充分由实际问题决定，如预算限制和模型的预期用途。复合建模和计算模拟没有任何包括准确性的证明，如在数学分析方面的发展。V&V的定义也强调准确度的评价，一般在验证过程中，准确度以对简化模型问题的基准解决方法符合性确定；有效性时，准确度以对实验数据即现实的符合性确定。 通常，不确定性和误差可视为与建模和仿真准确度相关的正常损失。不确定性，即在任一建模过程中由于缺乏知识导致的潜在缺陷。知识缺乏通常是由对物理特性或参数的不完全了解造成的，如对涡轮叶片表面粗糙度分布的不充分描述。知识缺乏的另一个原因是物理过程的复杂性，如湍流燃烧。误差即在建模和

车流量仿真分析-Flotran CFD

2006年用户年会论文 基于ANSYS流体动力学的车流量仿真分析1 [刘长虹，郑杰,朱晓华，张海波，黄虎，陈力华] [上海工程技术大学汽车工程学院，上海，201600] [ 摘要 ] 将交通流比拟为管道流体模型并且利用有限元分析软件ANSYS中的FLOTRAN CFD流体分析模块对隧道口交通流进行比拟及仿真，得出相应交通流量模型和车辆流动模拟图。并对不同车速下 交叉道口的通行能力进行模拟，确定出最佳车速比。且对不同入口形状进行车流通畅度的 ANSYA软件比较模拟，通过模拟直观的展示出不同道路入口形状对车流和道路的影响。最后对 高峰路段路口设计提出有关建议。 [ 关键词]交通流，交通流模型，ANSYS，模拟 Simulating to Traffic Flux By the ANSYS Fluid Dynamic Analysis [Liu Changhong, Zheng Jie, Zhu Xiaohua, Zhang Haibo, Huang Hu, Chen Lihua] [Automobile College Shanghai University of Engineering Science, Shanghai 201600] [Abstract ] Firstly, based on the fluid dynamic mechanics of channel, a traffic flow model is built. Secondly, the traffic flow model on cross road is simulated with the finite element method software (ANSYS). Then according to the calculating results, the simulating traffic ability at the entrance of the roadl in different speed and the different entrance figures are calculated directly. Finally, some suggestions of designing the heavy road are given. [ Keyword ] traffic flow, traffic flow simulation, ANSYS, Simulation. 1.前言 当前，社会经济的迅速发展与交通建设的相对滞后，已经构成非常突出的世界性矛盾，在发展中国家尤其突出。在我国许多大城市中，交通堵塞，事故频繁，成了众所周知的“都市顽症”。以上海市为例，上世纪九十年代的资料表明，在交通高峰期，市中心机动车平均车速不到15km/h，最低的车速仅仅为4km/h，即低于正常的步行速度。解决这个矛盾的一个重要办法是大力进行市政交通建设，实现交通的立体化，现代化。同时还要保证建设道路的合理性。交通流理论是解决这类方法的一种理论方法[1,2]，其中有根据流体动力学理 1上海市教委基金项目(041NE31)和上海市科委基金项目(04QMX1452)资助

医药工业洁净厂房设计规范

医药工业洁净厂房设计规 范 Jenny was compiled in January 2021

医药工业洁净厂房设计规范 第一章总则 第条为了贯彻执行国家《药品生产质量管理规范》（以下简称ＧＭＰ），提出符合ＧＭＰ要求的生产厂房、设施及设备的设计要求，特制订本规范。 第条本规范适用于新建、改建和扩建的医药制剂、原料药和药用辅料的精制、干燥、包装工序，直接接触药品的药用包装材料、无菌医疗器械等医药工业洁净厂房的设计。 第条医药工业洁净厂房诉设计必须贯彻国家有关方针、政策。做到技术先进、确保质量、安全实用、经济合理，符合节约能源和保护环境的要求。 第条医药工业洁净厂房的设计，既要满足当前产品生产的工艺要求，也应适当考虑今后生产发展和工艺改进的需要。 第条在利用原有建筑和设施进行洁净技术改造时，可根据生产工艺要求，从实际出发，充分利用现有的技术设施，符合因地制宜的原则。 第条医药工业洁净厂房的设计应为施工安装、维护、管理、检修、测试和安全运行创造必要的条件。 第条医药工业洁净厂房的设计除应执行本规范外，还应符合现行的国家标准、规范和规定的有关要求。 第二章生产区域的环境参数 第一节一般规定 第条为了保证医药产品生产质量，防止生产环境对产品的污染，生产区

注3：对于空气洁净度为100级的洁净室，室内大于等于5μm尘粒的计数，应进行多次采样，当其多次出现时，方可认为该测试数值是可靠的。 第2．2．2条药品生产有关工序和环境区域的空气洁净度等级按国家CMP等有关规定确定。 第2．2．3条洁净室内的温度和湿度应符合下列规定： 一、生产工艺对温度和湿度无特殊要求时，以穿着洁净工作服不产生不舒服感为宜。空气洁净度100级、10000级区域一般控制温度为20～24℃，相对湿度为45～60%。100000级区域一般控制温度为18～28℃，相对湿度为50～65%。 二、生产工艺对温度和湿度有特殊要求时，应根据工艺要求确定。 第2．2．4条洁净室内应保持一定的新鲜空气量，其数值应取下列风量中的最大值： 一、非单向流洁净室总送风量的10～30%，单向流洁净室总送风量的2～4%； 二、补偿室内排风和保持室内正压值所需的新鲜空气量； 三、保证室内每人每小时的新鲜空气量不小于40m3。 第2．2．5条洁净室必须维持一定的正压。不同空气洁净度的洁净区之间以及洁净区与非洁净区之间的静压差不应小于5Pa,洁净区与室外的静压差不应小于10Pa。 青霉素等特殊药物生产洁净区，固体口服制剂配料、制粒、压片等工序洁净区的气压控制，应符合第条要求。

医药工业洁净厂房设计规范GMP-97

医药工业洁净厂房设计规范 GMP—97 目录 医药工业洁净厂房设计规范 (1) 第一章总则 (2) 第二章生产区域的环境参数 (2) 第一节一般规定 (2) 第二节环境参数的设计要求 (2) 第三章厂址选择和总平面布置 (3) 第一节厂址选择 (3) 第二节总平面布置 (4) 第四章工艺设计 (4) 第一节工艺布局 (4) 第二节人员净化 (5) 第三节物料净化 (6) 第五章设备 (6) 第六章工艺管道 (8) 第一节一般规定 (8) 第二节管道材料、阀门和附件 (8) 第三节管道的安装、保温 (8) 第四节安全 (9) 第七章建筑 (9) 第一节一般规定 (9) 第二节防火和疏散 (9) 第三节室内装修 (10) 第八章空气净化 (11) 第一节一般规定 (11) 第二节净化空气调节系统 (11) 第三节气流组织 (12) 第四节风管和附件 (13) 第五节青霉素等药物生产洁净室的特殊要求 (13) 第九章给水排水 (14) 第一节一般规定 (14) 第二节给水 (14) 第三节排水 (14) 第四节工艺用水 (14) 第五节消防设施 (15) 第十章电气 (16) 第一节配电 (16) 第二节照明 (16) 第三节其它电气 (17) 附录一名词解释 (18) 附录二本规范用词说明 (19)

第一章总则 第1.0.1条为了贯彻执行国家《药品生产质量管理规范》（以下简称ＧＭＰ），提出符合ＧＭＰ要求的生产厂房、设施及设备的设计要求，特制订本规范。 第1.0.2条本规范适用于新建、改建和扩建的医药制剂、原料药和药用辅料的精制、干燥、包装工序，直接接触药品的药用包装材料、无菌医疗器械等医药工业洁净厂房的设计。 第1.0.3条医药工业洁净厂房诉设计必须贯彻国家有关方针、政策。做到技术先进、确保质量、安全实用、经济合理，符合节约能源和保护环境的要求。 第1.0.4条医药工业洁净厂房的设计，既要满足当前产品生产的工艺要求，也应适当考虑今后生产发展和工艺改进的需要。 第1.0.5条在利用原有建筑和设施进行洁净技术改造时，可根据生产工艺要求，从实际出发，充分利用现有的技术设施，符合因地制宜的原则。 第1.0.6条医药工业洁净厂房的设计应为施工安装、维护、管理、检修、测试和安全运行创造必要的条件。 第1.0.7条医药工业洁净厂房的设计除应执行本规范外，还应符合现行的国家标准、规范和规定的有关要求。 第二章生产区域的环境参数 第一节一般规定 第2.1.1条为了保证医药产品生产质量，防止生产环境对产品的污染，生产区域必须满足规定的环境参数标准。 第2.1.2条医药工业洁净室和洁净区应以微粒和微生物为主要控制对象，同时还应对其环境的温度、湿度、新鲜空气量、状差、照度、噪场院等参数作出必要的规定。 第2.1.3条环境空气中不应有不愉快气味以及有碍药品质量和人体健康的气体。 第二节环境参数的设计要求 第2.2.1条医药工业洁净厂房空气洁净度按表2.2.1规定分为三个等级。 医药工业洁净厂房空气洁净度等级表2.2.1

医药工业洁净厂房设计规范

国家医药管理局医药工业洁净厂房设计规范 默认分类 2008-07-13 12:25 阅读1075 评论3 字号: 大中小 主编部门:国家医药管理局上海医药设计院 批准部门:国家医药管理局 施行日期:１９９７年１月１日 编制说明 为在我国医药行业深入实施ＧＭＰ,适应医药工业洁净厂房建设的需要,国家医药局推行ＧＭＰ、ＧＳＰ委员会设计规范专业组决定组织编写《医药工业洁净厂房设计规范》。 本规范由上海医药设计院主编,缪德华同志执笔,武汉医药设计院、重庆医药设计院等单位参与。在编写过程中广泛眚注医药行业内有关方面意见,并先后数次组织医药设计单位与大中型骨干企业的专家进行了认真的讲座修改,由局推行ＧＭＰ、ＧＳＰ委员会寓言并原则通过。之后,局综合经济司(原局计划)就规范主要内容向卫生部药政局、药品监督办公室的领导、专家征询了意见并得到了她们的理解与支持。在此基础上,局ＧＭＰ设计规范专业组对本规范作了修改定稿。 本规范编制工作结合国内外ＧＭＰ的进展与医药工业洁净厂房建设、使用的实践经验,提出了我国医药工业洁净厂房设计的基本要求。各单位在新建、改建与扩建的工程设计中遵照执行。并认真总结经验,提出修改意见,以使本规范日臻完善。 国家医药管理局 第一章总则 第1、0、1条为了贯彻执行国家《药品生产质量管理规范》(以下简称ＧＭＰ),提出符合ＧＭＰ要求的生产厂房、设施及设备的设计要求,特制订本规范。 第1、0、2条本规范适用于新建、改建与扩建的医药制剂、原料药与药用辅料的精制、干燥、包装工序,直接接触药品的药用包装材料、无菌医疗器械等医药工业洁净厂房的设计。 第1、0、3条医药工业洁净厂房诉设计必须贯彻国家有关方针、政策。做到技术先进、确保质量、安全实用、经济合理,符合节约能源与保护环境的要求。 第1、0、4条医药工业洁净厂房的设计,既要满足当前产品生产的工艺要求,也应适当考虑今后生产发展与工艺改进的需要。 第1、0、5条在利用原有建筑与设施进行洁净技术改造时,可根据生产工艺要求,从实际出

医药行业洁净车间设计规范

医药工业洁净厂房设计规范 GB/T14294-1993 主编部门：国家医药管理局上海医药设计院 批准部门：国家医药管理局 施行日期：１９９７年１月１日

编制说明 为在我国医药行业深入实施ＧＭＰ，适应医药工业洁净厂房建设的需要，国家医药局推行ＧＭＰ.ＧＳＰ委员会设计规范专业组决定组织编写《医药工业洁净厂房设计规范》。 本规范由上海医药设计院主编，缪德华同志执笔，武汉医药设计院、重庆医药设计院等单位参与。在编写过程中广泛眚注医药行业内有关方面意见，并先后数次组织医药设计单位和大中型骨干企业的专家进行了认真的讲座修改，由局推行ＧＭＰ.ＧＳＰ委员会寓言并原则通过。之后，局综合经济司（原局计划）就规范主要内容向卫生部药政局、药品监督办公室的领导、专家征询了意见并得到了他们的理解和支持。在此基础上，局ＧＭＰ设计规范专业组对本规范作了修改定稿。 本规范编制工作结合国内外ＧＭＰ的进展和医药工业洁净厂房建设、使用的实践经验，提出了我国医药工业洁净厂房设计的基本要求。各单位在新建、改建和扩建的工程设计中遵照执行。并认真总结经验，提出修改意见，以使本规范日臻完善。 国家医药管理局 第一章总则 第1.0.1条为了贯彻执行国家《药品生产质量管理规范》（以下简称ＧＭＰ），提出符合ＧＭＰ要求的生产厂房、设施及设备的设计要求，特制订本规范。 第1.0.2条本规范适用于新建、改建和扩建的医药制剂、原料药和药用辅料的精制、干燥、包装工序，直接接触药品的药用包装材料、无菌医疗器械等医药工业洁净厂房的设计。 第1.0.3条医药工业洁净厂房诉设计必须贯彻国家有关方针、政策。做到技术先进、确保质量、安全实用、经济合理，符合节约能源和保护环境的要求。 第1.0.4条医药工业洁净厂房的设计，既要满足当前产品生产的工艺要求，也应适当考虑今后生产发展和工艺改进的需要。 第1.0.5条在利用原有建筑和设施进行洁净技术改造时，可根据生产工艺要求，从实际出发，充分利用现有的技术设施，符合因地制宜的原则。 第1.0.6条医药工业洁净厂房的设计应为施工安装、维护、管理、检修、测试和安全运行创造必要的条件。 第1.0.7条医药工业洁净厂房的设计除应执行本规范外，还应符合现行的国家标准、规范和规定的有关要求。 第二章生产区域的环境参数 第一节一般规定 第2.1.1条为了保证医药产品生产质量，防止生产环境对产品的污染，生产区域必须满足规定的环境参数标准。 第2.1.2条医药工业洁净室和洁净区应以微粒和微生物为主要控制对象，同时还应对其环境的温度、湿度、新鲜空气量、状差、照度、噪场院等参数作出必要的规定。 第2.1.3条环境空气中不应有不愉快气味以及有碍药品质量和人体健康的气体。

医药厂房建筑要求

药厂的建筑要求 主编部门：国家医药管理局上海医药设计院 批准部门：国家医药管理局 施行日期：１９９７年１月１日 编制说明 为在我国医药行业深入实施ＧＭＰ，适应医药工业洁净厂房建设的需要，国家医药局推行ＧＭＰ.ＧＳＰ委员会设计规范专业组决定组织编写《医药工业洁净厂房设计规范》。 本规范由上海医药设计院主编，缪德华同志执笔，武汉医药设计院、重庆医药设计院等单位参与。在编写过程中广泛眚注医药行业内有关方面意见，并先后数次组织医药设计单位和大中型骨干企业的专家进行了认真的讲座修改，由局推行ＧＭＰ.ＧＳＰ委员会寓言并原则通过。之后，局综合经济司（原局计划）就规范主要内容向卫生部药政局、药品监督办公室的领导、专家征询了意见并得到了他们的理解和支持。在此基础上，局ＧＭＰ设计规范专业组对本规范作了修改定稿。 本规范编制工作结合国内外ＧＭＰ的进展和医药工业洁净厂房建设、使用的实践经验，提出了我国医药工业洁净厂房设计的基本要求。各单位在新建、改建和扩建的工程设计中遵照执行。并认真总结经验，提出修改意见，以使本规范日臻完善。 ?国家医药管理局 ?１９９６北京 目录 第一章总则 第二章生产区域的环境参数 第一节一般规定 第二节环境参数的设计要求 第三章厂址选择和总平面布置 第一节厂址选择 第二节总平面布置 第四章工艺设计 第一节工艺布局 第二节人员净化 第三节物料净化 第五章设备 第六章工艺管道 第一节一般规定 第二节管道材料、阀门和附件 第三节管道的安装、保温 第四节安全 第六章建筑 第一节一般规定 第二节防火和疏散 第三节室内装修 第七章建筑 1. 一般规定 2. 防火和疏散 第三节室内装修 第八章空气净化 第一节一般规定 第二节净化空气调节系统 第三节气流组织 第四节风管和附件 第五节青霉素等药物生产洁净室的特殊要求 第九章给水排水 第一节一般规定 第二节给水 第三节排水 第四节工艺用水 第五节消防设施 第十章电气 第一节配电 第二节照明 第三节其它电气 附录一名词解释 附录二本规范用词说明 第一章总则 第1.0.1条为了贯彻执行国家《药品生产质量管理规范》（以下简称ＧＭＰ），提出符合ＧＭＰ要求的生产厂房、设施及设备的设计要

CFD案例5-发动机仿真

ANSYS对航空工业解决方案（三）航空发动机仿真方案_2 发表时间：2008-10-23 作者: 安世亚太来源: 安世亚太 关键字: 航空航天 CAE 仿真解决方案 ANSYS 安世亚太 第三章航空发动机仿真方案航空发动机行业概况航空发动机研制中的典型CAE问题航空发动机结构力学计算需求及ANSYS实现航空发动机流体力学和温度场的计算需求及ANSYS实现航空发动机电磁场计算需求及ANSYS实现航空发动机耦合场计算需求及ANSYS实现航空发动机关键零部件的设计分析流程简要说明 4航空发动机流体力学和温度场的计算需求及ANSYS实现 航空燃气涡轮发动机内的流场很复杂，不仅动静流场同时存在，同时还伴有多相流、传热、燃烧等现象，即使从物理上进行很大的简化，模型最后仍然是三维、有粘、非定常的可压流动。航空发动机流场数值计算的发展经历了S2流面法、基于一元管道的流线曲率法、有限差分方法求解非正交曲线坐标系中的S1、S2流面基本方程、有限差分、有限体积和有限差分与流线曲率混合的方法对S1流面跨音速流场的计算，而现在由S1与S2流面相互迭代形成的准三元和全三元计算也发展起来了。现在的采用有限体积法求解NS方程全三维流场计算已经广泛采用，航空发动机的流场数值计算已趋于成熟，可以充分考虑旋转流动、转静干涉问题、多相流、燃烧、亚超跨音速等复杂现象。而且现在求解的规模也不断扩大，利用并行等成熟的CFD技术可以计算达几千万甚至上亿的计算网格。因此结果也更为真实有效。 ANSYSCFX凭借TASCFLOW在叶轮机旋转流动的传统优势，结合更为先进的网格处理技术和高效的求解器，更适合航空发动机流动的复杂性，求解问题的规模和计算精度大大提高，一直处于航空发动机流动模拟的最前沿。

洁净厂房洁净区设计规范

洁净厂房洁净区设计规 范 Document serial number【KKGB-LBS98YT-BS8CB-BSUT-BST108】

国家医药管理局医药工业洁净厂房设计规范主编部门：国家医药管理局上海医药设计院 批准部门：国家医药管理局 施行日期：１９９７年１月１日 编制说明： 为在我国医药行业深入实施ＧＭＰ，适应医药工业洁净厂房建设的需要，国家医药局推行ＧＭＰ.ＧＳＰ委员会设计规范专业组决定组织编写《医药工业洁净厂房设计规范》。 本规范由上海医药设计院主编，缪德华同志执笔，武汉医药设计院、重庆医药设计院等单位参与。在编写过程中广泛眚注医药行业内有关方面意见，并先后数次组织医药设计单位和大中型骨干企业的专家进行了认真的讲座修改，由局推行ＧＭＰ.ＧＳＰ委员会寓言并原则通过。之后，局综合经济司（原局计划）就规范主要内容向卫生部药政局、药品监督办公室的领导、专家征询了意见并得到了他们的理解和支持。在此基础上，局ＧＭＰ设计规范专业组对本规范作了修改定稿。 本规范编制工作结合国内外ＧＭＰ的进展和医药工业洁净厂房建设、使用的实践经验，提出了我国医药工业洁净厂房设计的基本要求。各单位在新建、改建和扩建的工程设计中遵照执行。并认真总结经验，提出修改意见，以使本规范日臻完善。 国家医药管理局 第一章总则 为了贯彻执行国家《药品生产质量管理规范》（以下简称ＧＭＰ），提出符合ＧＭＰ要求的生产厂房、设施及设备的设计要求，特制订本规范。 本规范适用于新建、改建和扩建的医药制剂、原料药和药用辅料的精制、干燥、包装工序，直接接触药品的药用包装材料、无菌医疗器械等医药工业洁净厂房的设计。 医药工业洁净厂房诉设计必须贯彻国家有关方针、政策。做到技术先进、确保质量、安全实用、经济合理，符合节约能源和保护环境的要求。

CFD仿真技术在航空发动机中的应用

CFD仿真技术在航空发动机中的应用 摘要：随着科学技术的发展，航空航天和空间技术有了飞跃的发展，在这些飞 跃的发展技术中主要的技术就是CAE技术。航空工业可以说是CAE技术发展的摇篮，各种CAE技术正是在以航空工业为主的实际工业应用的推动下在不到半个世 纪时间里迅猛发展起来的。以ANSYS、LS-DYNA、Nastran、CFX、Fluent等为代表 的高端CAE软件早已活跃在全球航空工业中。 关键词：CFD仿真技术；航空发动机；应用 1 引言 目前国际知名企业的航空发动机研制周期从过去的10～15年缩短到6～8年 甚至4～5年，试验机也从过去的40～50台减少到10台左右。在发达国家的航 空企业里CAE已经作为产品研发设计与制造流程中不可逾越的一种强制性的工艺 规范加以实施，在生产实践作为必备工具普遍应用。 2、CFD技术国内外使用状况简介 CFD作为CAE技术的一种，已经越来越多的被国内外航空企业广泛的得以应用。第一个商用CFD软件包FLUENT，由与美国空军合作的流体技术服务公司Creare公司于1983年推出的。商业CFD软件的开发及应用，加速了航空工业的 发展，使得基于虚拟样机仿真的现代设计方法成为了可能。以波音公司航空研发 发展历史为例，不难发现，波音公司先后采用了经典的实验测试方法、半经验的 方法、空气动力学的计算、政府内部及企业的CFD代码及广泛的采用CFD商业代码。在波音公司2005年的软件应用报告中明确指明，在1998至2005年内，其 公司每年数值仿真成果的增加量都接近84%左右，采用CAE/CFD的速度超过了工 业的成长速度，CFD技术已经成为其设计的主要手段之一。另外从美国软件公司ANSYS公司的销售业绩报告上显示，航空工业上的应用产值是其公司的主要收益 来源之一。 CFD软件正以其强大的优势在研发中发挥的巨大的作用，例如在NISA的报告 中提到，原本需要7年完成的维吉尼亚级潜水艇的设计，通过CFD技术的应用， 5年就顺利完成；而预计需要11年完成的B-2轰炸机的飞行测试，则在短短的4 年内就通过了测试。 国内在CFD技术上的应用一般，特别是在航空发动方面的使用上，起步与国 外相比较晚，力度上也相差较多。 3、CFD技术的应用 目前在航空发动机的实际应用中是最广泛的一款CFD商业软件是ANSYS旗下 的商业软件FLUENT，其不仅容易使用，而且其准确性及行业的广泛性都是其它商业软件所不能比拟的。CFD软件的使用已经遍及了航空发动机的各个部分的研究，接下来本文通过对其它文献的分析逐一介绍CFD在航空发动机中的使用。 3.1 CFD技术在压缩机、涡轮方面的应用 气动稳定性的设计是当代航空发动机发展研制过程中的重要技术问题之一。 在航空发动机中，对气流最敏感的部件是风扇、压气机和涡轮。在以上3个部件中，CFD的主要应用集中在对压气机和涡轮效率分析上，多级压气机/涡轮最主要 的气动问题就是各级流动是否匹配，总的效率是否达到设计要求。在涡轮方面，CFD不仅可以计算涡轮效率，而且对涡轮叶片的冷却效果分析有着重要的应用。

医药工业洁净厂房的消防设计

仅供参考[整理] 安全管理文书 医药工业洁净厂房的消防设计 日期：__________________ 单位：__________________ 第1 页共6 页

医药工业洁净厂房的消防设计 摘要：介绍了医药工业洁净厂房的特点，分析了其火灾危险性及存在的问题，提出了在消防设计上应注意的事项，探讨 关键词：医药洁净厂房；火灾危险；消防设计 目前，我国正在推行药品GMP(药品生产管理规范)认证，此项认证对厂房洁净度的要求应达标。而医药工业生产工艺复杂，生产过程中使用多种化学危险物品，火灾危险性较大，有的药品价格昂贵，一旦发生火灾，损失严重。医药洁净厂房的设计依据是GB50073—2001《洁净厂房设计规范》，并参照执行国家医药管理局发布的《医药工业洁净厂房设计规范》，笔者就此类厂房的火灾危险性、消防设计及审核应注意的问题做一探讨。医药工业洁净厂房的火灾危险性大 (1)医药工业洁净厂房经常使用各种易燃易爆化学危险物品，且本身化学反应也有燃烧爆炸危险，如使用环氧乙烷、酒精、煤气活性镍、钯炭等极易引起燃烧的物品，制剂工艺灭菌工序中的火焰法和环氧乙烷法的火灾危险性都较大。 (2)建筑或装修材料采用易燃材料。由于洁净厂房需要特殊，加之很多新型复合材料的使用，而建设单位缺乏消防知识，有些选用了有机复合材料，或彩钢板加泡沫，增加了火灾荷载，这些材料燃烧后，又会产生大量的有毒气体。 (3)医药洁净厂房由于洁净无菌的工艺要求，密封性好，但易造成发生火灾后，有毒高温烟气不易及时扩散，甚至会使轰燃提前发生。 (4)人员人口不能满足疏散的要求。医药洁净厂房必须经过清洗、男(女)更衣、消毒等隔离间方可进入，且由于洁净无菌的要求，该疏散门不能采用向疏散方向开启的门，火灾时从人口疏散极为复杂和困难。 第 2 页共 6 页

洁净厂房设计规范(GB50073-2013)

洁净厂房设计规范（GB50073-2013） 4. 1 洁净厂房位置选择和总平面布置 4.1. 1 洁净厂房位置选择应符合下列规定，并经技术经济方案比较后确定：1）应在大气含尘和有害气体浓度较低、自然环境较好的 区域。 2)应远离铁路、码头、飞机场、交通要道以及散发大量粉尘和有害气体的工厂、贮仓、堆场等有严重空气污染、振动或噪声干扰的区域。当不能远离严重空气污染源时，应位于最大频率风向上风侧，或全年最小频率风向下风侧。 3)应布置在厂区内环境清洁，人流、物流不穿越或少穿越的 地段。 4.1.2对于兼有微振控制要求的洁净厂房的位置选择，应实际测定周围现有振源的振动影响，并应与精密设备、精密仪器仪表容许振动值分析比较后确定。 4.1.3洁净厂房新风口与交通干道边沿的最近距离宜大于 50m。 4.1.4洁净厂房周围宜设置环形消防车道，也可沿厂房的两个长边设置消防车道。 4.1.5洁净厂房周围的道路面层应选用整体性能好、发尘少的材 料。 4.1.6洁净厂房周围应进行绿化。可铺植草坪，不应种植对生产有害的植物，并不得妨碍消防作业。 4.2 工艺平面布置和设计综合协调 4.2.1工艺平面布置应符合下列规定： 1)工艺平面布置应合理、紧凑。洁净室或洁净区内应只布置必要的工艺设备，以及有空气洁净度等级要求的工序和工作室。 2) 在满足生产工艺和噪声要求的前提下，对空气洁净度要求 严格的洁净室或洁净区宜靠近空气调节机房，空气洁净度等级相同的工序和工作室宜集中布置。 3) 洁净室内对空气洁净度要求严格的工序应布置在上风侧， 易产生污染的工艺设备应布置在靠近回风口位置。 4) 应考虑大型设备安装和维修的运输路线，并预留设备安装口和检修口。 5) 不同空气洁净度等级房间之间联系频繁时，宜设有防止污 染的措施，如气闸室、传递窗等。 6) 应设置单独的物料入口，物料传递路线应最短，物料进入洁净室（区）之前应进行清洁处理。 4.2.2洁净厂房的平面和空间设计应满足生产工艺和空气洁净度等级要求。洁净区、人员净化、物料净化和其他辅助用房应分区布置,并应与生产操作、工艺设备安装和维修、管线布置、气流流型以及净化空调系统等各种技术设施进行综合协调。 4.2.3洁净厂房内应少设隔间，但在下列情况下应进行分隔： 1)按生产的火灾危险性分类，甲、乙类与非甲、乙类相邻的生产区段之间，或有防火分隔要求者。 2) 按产品生产工艺需要有分憬要求时。

传热模拟CFD 总结

CFD总结一 CFD是英文computational Fluid Dynamics（计算流体力学）的简称。它是伴随着计算机技术和数值计算技术的发展而发展的。简单地说，CFD相当于虚拟的在计算机内做实验，用它模拟仿真实际流体的流动情况。而其基本的原理是数值求解控制流体的微分方程，得出流体流动的流场在连续区域上的离散分布，从而近似模拟流体流动的情况。即CFD=流体力学+热学+数值分析+计算机科学。 流体力学就不用多说了，很多专业都要用到，主要的概念有层流和湍流，牛顿流体和非牛顿流体等等。热学包括热力学和传热学。数值分析就是如何用计算机解人工很难完成的计算，如何处理无解析解得方程。计算机科学主要是计算机语言，如c、fortran）还包括一些图形处理技术，如在后处理，为了使用户对结论有一个很直观的认识，就需要若干图表。以下就对经常在CFD使用的软件做简单的介绍。 一、CFD的结构： 1、提出问题——流动性质（内流、外流；层流、湍流；单相流、多项流；可压、不可压……），流体属性（牛顿流体：液体、单组分气体、多组分气体、化学反应气体；非牛顿流体） 2、分析问题——建模——N-S方程（连续性假设），Boltzmann方程（稀薄气体流动），各类本构方程与封闭模型。 3、解决问题——差分格式的构造/选择，程序的具体编写/软件的选用，后处理的完成。 4、成果说明——形成文字，提交报告，赚取应得的回报。 二、CFD实现过程： （一）建模——物理空间到计算空间的映射。 主要软件： 二维： AutoCAD： 大家不要小看它，非常有用。一般的网格生成软件建模都是它这个思路，很少有参数化建模的。相比之下 AutoCAD的优点在于精度高，草图处理灵活。可以这样说，任何一个网格生成软件自带的建模工具都是非参数化的，而对于非参数化建模来说，AutoCAD应该说是最好的，毕竟它发展了很多很多年！ 三维： 1、CATIA： 航空航天界CAD的老大，法国人的东西，NB，实体建模厉害，曲面建模独步武林。本身可以生成有限元网格，前几天又发布了支持ICEM-CFD的插件ICEM-CFD CAA V5。有了它和ICEM-CFD，可以做任何建模与网格划分！ 2、UG： 软件本身不错，大公司用得也多，可是就这么打市场，早晚是走下坡路。按CAD建模的功能来说它排不上第一，也不能屈居第二，尤其是加上了I－DEAS更是如虎添翼。现在关键是看市场了。 3、Solidworks： Solidworks讲的是实用主义，中端CAD软件它功能最强，Solidedge功能是不比它差，但是Solidworks的合作伙伴可能是SE的十几倍，接口也比SE多很多，相比之下Solidworks是最佳选择。Autodesk Inventor也只能算是中端软件，目前说来，我是处于观望态度，看发展再决定。总之，Solidworks目前的发展如日中天，合作伙伴多如牛毛。用起来极其顺手。这里极力向大家推荐的是ICEM-CFD DCI FOR Solidworks！有了这个东西画个全机网格也很容易了。