Chemical Engineering Science62(2007)2–
27
/locate/ces
Bubble formation and dynamics in gas–liquid–solidfluidization—A review
G.Q.Yang,Bing Du,L.S.Fan∗
Department of Chemical and Biomolecular Engineering,The Ohio State University,Columbus,OH43210,USA
Available online25August2006
Abstract
Current worldwide commercial activities in converting natural gas to fuels and chemicals,or gas-to-liquids technology u slurry bubble column reactors with column sizes considerable larger than tho currently in practice.Such commercial activities have prompted further fundamental rearch interest influid and bubble dynamics,transport phenomena and the scale up effects of three-phafluidization systems. The fundamental behavior of particular relevance to the activities is associated with the elevated temperature and pressure conditions.
This review attempts to summarize the salient characteristics of liquid,bubbles,and particles and their interactive behavior and dynamics in the process of bubble formation and bubble rising in gas–liquid–solidfluidization systems.Measurement techniques including both intrusive techniques such as the probes,and non-intrusive techniques such as tomography,that are ud to studyfluid and bubble properties in gas–liquid and gas–liquid–solid systems,are illustrated.Governing mechanisms of bubble–particle collision and bubble breakup are discusd.The state-of-the-art computational techniques,that consider both the discrete and the continuum approaches for movement of the particle and bubble phas along with the discrete simulation results,are prented.Of particular emphasis is the effect of pressure and temperature on thefluid and bubble dynamics in three-phafluidization.
᭧2006Elvier Ltd.All rights rerved.
Keywords:Bubble formation;Bubble dynamics;Measurement techniques;Gas–liquid–solidfluidization;Pressure;Computationalfluid dynamics(CFD); Bubble–particle collision;Bubble breakup
1.Introduction
Gas–liquid–solidfluidization systems have been applied extensively in industry for physical,chemical,petrochemical and biochemical processing(Shah,1979;L’Homme,1979; Ramachandran and Chaudhari,1983;Fan,1989).Current worldwide commercial activities in converting natural gas to fuels and chemicals,or gas-to-liquids technology u slurry bubble column reactors with column sizes considerable larger than tho currently in practice(Sookai et al.,2001).Such com-mercial activities have prompted further fundamental rearch interest influid and bubble dynamics,transport phenomena, and the effects due to scale up of three-phafluidization systems.The fundamental behavior of particular relevance to the activities is associated with the elevated temperature and pressure conditions.
In gas–liquid–solidfluidization systems,bubble dynamics plays a key role in dictating the transport phenomena and
∗Corresponding author.Tel.:+16146883262.
E-mail address:fan@chbmeng.ohio-state.edu(L.S.Fan).
0009-2509/$-e front matter᭧2006Elvier Ltd.All rights rerved. doi:10.s.2006.08.021ultimately affects the overall rates of reactions.It has been rec-ognized
that the bubble wake,when it is prent,is the dominant factor governing the system hydrodynamics(Fan and Tsuchiya, 1990).In general,consideration of theflow associated with the bubble wake near the bubble ba,whether laminar or turbu-lent,is esntial to characterize the complete behavior of the rising bubble,including its motion.Converly,examining the shape,ri velocity,and motion of a bubble can provide an in-direct understanding of the dynamics of the liquid–solidflow around the bubble.
Most of the three-pha process with considerable commercial interest are conducted under high pressure and high temperature,for example,methanol synthesis(at P=5.5MPa and T=260◦C),resid hydrotreating(at P=5.5–21MPa and T=300–425◦C),Fischer–Tropsch syn-thesis(at P=1.5–5.0MPa and T=250◦C),and benzene hydrogenation(at P=5.0MPa and T=180◦C)(Fox,1990; Jager and Espinoza,1995;Saxena,1995;Mills et al.,1996; Peng et al.,1999).Fundamental study of bubble dynamics in the gas–liquid–solidfluidization systems,particularly under high-pressure and high-temperature conditions,is thus crucial.
G.Q.Yang et al./Chemical Engineering Science62(2007)2–273
This review describes theflow behavior of liquid,bubbles, and particles in gas–liquid–solidfluidization systems.The mechanisms of the bubble formation,bubble instability,and bubble ri dynamics along with pertinent forces governing such mechanisms are illustrated.The review also surveys mea-surement techniques that are ud to quantify theflow and bubble properties in gas–liquid and gas–liquid–solid systems. The techniques include both the intrusive ones such as the probes,and non-intrusive ones such as tomography.Salient bubbling phenomena,related to bubble–particle collision and the hydrodynamic similarity rules,are also discusd.The state-of-the-art computational techniques that consider both the discrete and the continuum approaches for the particle and bubble phas as well as some discrete simulation results are prented.Of particular emphasis in this review are the pres-sure and temperature effects on thefluid and bubble dynamic properties in three-phafluidization.
2.Measurement techniques
The quantification of bubble characteristics in gas–liquid–solidfluidized beds is normally made through direct visualiza-tion or by employing instruments.To visually obrve bubble behavior in thre
e-phafluidization systems,a two-dimensional (2D)fluidized bed is commonly ,Chen et al.,1989, 1994;Fan and Tsuchiya,1990;Kim and Kim,1990;Fan et al., 1992;Tzeng et al.,1993;Kluytmans et al.,2001,2003;Bech, 2005;Vandu et al.,2005;Zaruba et al.,2005).By analyzing the data obtained by the photography or video images,the dynamic behavior of bubbles,including bubble shape,bubble wake, bubble size and bubble ri velocity,is quantified.The bubble flow behavior in a2Dfluidized bed is treated as a vertical slice of the three-dimensional(3D)system.Chen et al.(1994)found that there were some similarities between theflow structures of2D and3D beds.However,the direct visualization only provides limited information regarding bubble dynamics in3D systems.
On the other hand,a large number of measurement tech-niques,including intrusive and non-intrusive methods,have been developed to investigate the bubbleflow behavior in the3Dfluidized bed systems.A comprehensive review of measurement techniques in gas–liquid and gas–liquid–solid reactors can be found in Boyer et al.(2002).In the following, some measurement techniques that quantify bubble character-istics,along with recent advances in measurement techniques for gas–liquid and gas–liquid–solidfluidization rearch are discusd.
2.1.Intrusive techniques
专注的近义词
花园作文Considerable intrusive techniques have been developed to study bubble behavior in gas–liquid and gas–liquid–solidflu-idized systems.The intrusive techniques include impedance (conductivity or resistivity)probes,opticalfiber probes,ultra-sound probes,endoscopic probes and hotfilm anemometry.
A brief summary of the techniques is given below.
The impedance probe has been applied to measure the bub-ble volume fraction,bubble length and bubble ri velocity in three-phafluidized beds with relatively high liquid conductiv-ity.The method utilizes the difference in conductivity between the liquid and the gas pha.For the three-phafluidized beds with low liquid conductivity,the addition of some salts into the system is required(Boyer et al.,2002).Hills and Darton(1976) investigated the bubble rising velocity in a bubble column by an impedance probe.Matsuura and Fan(1984)studied the bub-ble size and bubble ri velocity in three-phafluidized beds under three differentflow regimes by using a dual electrical re-sistivity probe.Tang and Fan(1989)applied a dual-resistivity probe to study the bubble size distribution and the axial distri-bution of gas volume fraction.Liu(1993)ud a dual-nsor resistivity probe to measure the bubble size,bubble ri velocity and bubble frequency in the bubble column.Chen et al.(1998) applied a dual-resistivity probe to measure the axial and radial distributions of bubble diameter,bubble ri velocity,bubble frequency and gas volume fraction in a th
ree-phafluidized bed.Zenit et al.(2001)applied a dual impedance probe to study the gas volume fraction,bubble velocity and bubble collision in a vertical channel.To obtain accurate bubble volume frac-tion using the impedance probe technique,the interaction be-tween bubbles and the probe must be considered(Zenit et al., 2003).Bad on the statistical,fractal,chaos and wavelet anal-ys,the conductivity bubble probe signal can be analyzed to discern the localflow structure of the three-phafluidized bed (Briens and Ellis,2005).保肝养肝的食物
The opticalfiber probe utilizes the principle that the light transmits in liquid medium and is reflected by the gas medium or bubbles.The optical probe is not effective,however,when the difference in the refraction index between the gas and liq-uid phas is small.Lee et al.(1986)and Lee and De Lasa (1987)measured the local gas volume fraction and bubble fre-quency in a three-phafluidized bed using the U shape optical fiber probe.Yu and Kim(1988)applied the U shape optical fiber probe to study the radial distributions of the bubble size, bubble ri velocity and bubble volume fraction in three-pha fluidized beds.Frijlink(1987)developed a four-point probe to improve the detection of the direction of the movement and the shape of the bubble.Chabot and de Lasa(1993)measured the axial and radial distributions of the bubble chord length,bubble ri velocity and gas volume fraction in a bubble column at high temperature by using the refractive optical probe.Xue et al. (2003)applied
the four-point opticalfiber probe to investigate the bubble size and bubble ri velocity in gas–liquid systems. They found that a preci calibration of the probe by a CCD camera was needed to obtain the accurate measurement on the bubble size and bubble ri velocity.Shoukri et al.(2003)mea-sured the gas volume fraction,bubble size,bubble ri velocity, bubble frequency and interfacial area in a large scale bubble column using a dual optical probe.One of the advantages for the opticfiber probe technique is that it can also be effectively applied to high-pressure and high-temperature conditions for the bubble property measurement(Luo et al.,1997,1998b). Stolojanu and Prakash(1997)obtained the solids concentra-tion and bubble volume fraction in a three-phafluidized bed
4G.Q.Yang et al./Chemical Engineering Science62(2007)2–27
using the ultrasonic technique.Al-Masry et al.(2005)studied the bubble frequency and bubble size distribution in bubble columns using the statistical analysis of acoustic signals.The ultrasonic techniques are not suitable for u under high gas holdup conditions becau of the significant acoustic attenua-tion due to reflection on gas bubbles(Broering et al.,1991). Peters et al.(1983)studied the particle ejections by the bubble eruption at the surface of a bubbling gas–solidfluidized bed using an image system carrying afiber optic probe or an en-doscopic probe.Such a
probe was also ud by Kumar et al. (1992)in the study of the solid concentration effects on the heat transfer in bubbly liquid–solid systems.Wang and Ching(2001) measured the multiple bubble velocities in the gas–liquidflow using a dual-probe hot-film anemometry.However,the hot-film probe is so fragile that it can only be ud for low solids concentration conditions.Furthermore,hot-film anemometry measurement requires uniform temperature distribution in the measured volume.
The probe measures the point properties.One of the disad-vantages for intrusive probe techniques is the probe interfer-ence with theflowfield and hence the bubble dynamics.The bubble could be disintegrated,accelerated or elongated by the immerd probe(Rowe and Masson,1981;Chabot et al.,1992; Kiambi et al.,2003;Zenit et al.,2003;Julia et al.,2005).Al-though reducing the probe size could reduce the interfering effect,the probe could also easily be damaged.The accurate conversion of the chord length distribution to the bubble size distribution is another challenging area as the bubble shape and the size distribution vary with time and the location.
2.2.Non-intrusive techniques
The non-intrusive technique has the advantage of no mea-surement interference with theflowfield.The information provided by the non-intrusive techniques varies from the cross-ctional
bed density profile to the particle trajectory map (Chaouki et al.,1997;Chen et al.,1999;Seeger et al.,2003; Hubers et al.,2005;Warsito and Fan,2005).The non-intrusive techniques that were ud to measure the three-phafluidized bed properties include pressure transducer,visualization tech-nique,particle image velocimetry(PIV),X-ray, -ray,positron emission tomography(PET),radioactive particle tracking (RPT),ultrasonic tomography,nuclear magnetic resonance imaging(NMR),lar techniques and electrical tomography. Some measurement examples in three-phafluidized beds are described below.
The pressure drop measurement together with the statisti-cal analysis techniques has been ud to study the bubbleflow behavior in the bubble column and three-phafluidized beds (Drahos et al.,1991,1992;Johnsson et al.,2000;Kluytmans et al.,2001;Briens and Ellis,2005;Chilekar et al.,2005).The pressure transducer is usually positioned on the wall of the bed. The pressurefluctuation signal is a reflection of the overall hy-drodynamic behavior in the column.That is,the signals are contributed from such sources as bubbleflow,bubble coales-cence and bubble breakup,bubble burst at the top surface,and bubble formation at the distributor.
Direct visualization is also uful for property measurement
in the systems with relatively low gas holdup and solids load-
ing.Jiang et al.(1995),Luo et al.(1998a)and Yang et al.(2000)
utilized the visualization technique to study the bubble charac-
teristics and bubble formation behavior in a largely transparent
apparatus operated at high-pressures(up to20MPa)and high-
中搜temperatures(up to220◦C).They also developed visualization
techniques for in situ measurement of the physical properties
of reacting or non-reactingfluids,such as dynamic surface ten-
sion,andfluid density and viscosity at high pressure and high
temperature(Lin and Fan,1997).
心若兰兮终不移The X-ray technique has been widely ud to investigate the
bubbleflow behavior including bubble shape,bubble size,bub-
ble ri velocity,bubble growth and bubble breakage in the
gas–solid or gas–liquid–solidfluidized beds.The X-ray tech-
nique consists of the X-ray source to generate the X-ray beam
to pass through thefluidized bed,an image intensifier to pro-
duce an image,a CCD video camera to capture the image,and
the image analysis system.Fournier and Jeandey(1993)mea-
sured the void fraction in the gas–liquid two-phaflow using
the X-ray attenuation technique.The X-ray computer assisted
tomography(CAT)was developed to investigate thefluidization
characteristics(Kumar et al.,1997).The X-ray CAT technique
is able to provide rather high spatial resolution(1%),while its
temporal resolution is low.Seeger et al.(2003)measured the
local solids velocity and local solids holdup in a three-pha
fluidized bed by using the X-ray bad particle tracking ve-
locimetry(XPTV).Hubers et al.(2005)applied the X-ray CT
technique to measure the pha holdups in the three-phaflu-
idized bed.
The -ray density gauge technique has been applied to study
保留金the bubble size,bubble frequency and bubble coalescence in a
fluidized bed.The voidage between the radiation source and
detector in the bed is obtained by relating the ionization of
gas to the amount of radiation received by the detector.Seville
et al.(1986)studied the jet and bubble behavior above the
distributor of a gas–solidfluidized bed by the -ray tomogra-
phy technique.The system included a -ray source and a NaI
detector,which rotated along the axis of thefluidized bed.The
total scan time was up to7.5h.The same measurement tech-
nique can be extended to the liquid system.Kumar et al.(1995)
measured the voidage distribution in the bubble column using
the -ray tomographic scanner.Veera and Joshi(2000)ud the -ray tomography to investigate the radial distributions of the gas holdup in a bubble column.Jin et al.(2005)studied the
pha holdups in a pressurized bubble column using the -ray
densitometry.Due to long scanning time,the -ray tomography
technique is only suitable for studying the time-averagedflow
properties.It is not suitable for the measurement of bubble
formation and bubble dynamics in the bed.
Non-intrusive lar techniques are also widely ud to study
bubble behavior,including the PIV,lar Doppler anemometry
(LDA),pha Doppler anemometry(PDA)and lar Doppler
velocimetry(LDV).Chen and Fan(1992)and Ree et al.
(1995,1996)investigated the bubble characteristics in the slurry
bubble column using the2D and3D PIV techniques.Lee et al.
(1999)studied the bubble size distribution in the bubble column
G.Q.Yang et al./Chemical Engineering Science62(2007)2–275
and slurry bubble column using the gas dingagement tech-nique together with PIV technique.Vial
et al.(2001)studied the liquid velocity and turbulence in the bubble columns with different distributors by using the LDA technique.Kulkarni et al.(2004)applied the LDA technique to study the bubble size distribution in the bubble columns.Braeske et al.(1998)mea-sured the size,velocity and holdup of bubble and solid phas in the three-phafluidized beds using the PDA technique.Brenn et al.(2002)applied the PDA technique to measure the veloc-ities of liquid and bubbles in the bubble column.Cui and Fan (2004,2005)investigated the turbulence energy distribution in bubble columns and three-phafluidized beds by measuring the liquid velocity using the LDV technique.For all the lar techniques,the lar beam needs to penetrate theflow system. Thus,the lar techniques limited only to the low gas holdup conditions.
Some other non-intrusive techniques are ud for tracking the particle movement,and/or mapping the instantaneous or time-averaged,local or cross-ctional averaged,pha holdups and pha velocities.They include ,Bemro et al., 1988;Stein et al.,2000;Dechsiri et al.,2005;Hoffmann et al., 2005),and ,Cassanello et al.,1995;Larachi et al., 1996,1997;Chaouki et al.,1997;Chen et al.,1999;Kiared et al.,1999;Nedeltchev et al.,2003),ultrasonic tomography (e.g.,Wolf,1988;Xu et al.,1997;Warsito et al.,1999;Utomo et al.,2001),nuclear magnetic resonance imaging(NMR or MRI)(e.g.,Gladden,1994,2003;Chaouki et al.,1997;Leblond et al.,1998;L
e Gall et al.,2001;Lim et al.,2004;Sederman and Gladden,2005;Gladden et al.,2005),electrical impedance tomography(EIT)(George et al.,2001;West et al.,2001;Kim et al.,2005)and electrical capacitance tomography(ECT) (Warsito and Fan,2001,2003).The details of each of the techniques and the specific hydrodynamic parameters they measure can be found in the corresponding references.
The MRI technique has been widely ud in medical appli-cations.This technique,however,has also been ud for the measurement of multiphaflow systems such as thefixed beds and trickle beds(Gladden,1994,2003;Chaouki et al.,1997; Leblond et al.,1998;Le Gall et al.,2001;Lim et al.,2004; Sederman and Gladden,2005;Gladden et al.,2005).Lim et al.(2004)applied the ultra-fast MRI technique to investi-gate the hydrodynamics in the trickle bed reactors.The2D images of the trickle bed can have a higher spatial resolution of351 m×351 m with a slow acquisition time of6.4s, or a low spatial resolution of1.4mm×2.8mm with a rel-atively fast acquisition time of20ms.The MRI can also be ud to quantify theflowfield in bubble column systems.The relatively high cost of the technique and certainfluid prop-erty requirements,however,may hamper widespread usage of the MRI as a process tomography technique.The ECT is developed to image the multipha media with dielectric properties.It can be ud to quantify the dynamic bubble flow behavior in the gas–liquid an
d gas–liquid–solid three-phafluidized beds.Compared to the CT,the ECT technique has a relatively low spatial resolution but a relatively high temporal resolution.The ECT is suitable for process tomog-raphy applications for various multiphaflow systems.The recent development of the3D ECT or electrical capacitance volume tomography(ECVT)with geometrically configured nsor design and the neural network image reconstruction technique has further advanced the imaging technique and al-lows the dynamic3D multiphaflow behavior to be captured (Du et al.,2005;Warsito and Fan,2005).The ECVT with a spatial resolution of5×5×5mm3and a temporal resolu-tion of80Hz has revealed the bubble formation process and bubble dynamics in three-phafluidized beds(Warsito and Fan,2005).Fig.1shows the3D bubble swarms obtained from the ECVT along with corresponding video images.The images in thefigure reprent one cycle of the circular motion of the spiral rising bubble plume,showing the concutive motions of the central bubble plume on the image plane.The direction of the circular motion is not constant,and is mixed with a high frequency back and forth dancing motion of bubble swarm. 3.Single bubble behavior
In the following,the phenomena related to single bubble behavior are discusd,which include bubble formation from a single orifice,bubble shape,single bubble ri velocity and bubble induced liquidflow.Experimental studies of the single bubble behavior have been extensively reported,and the
军检标准
the-oretical account and CFD simulation have provided detailed information on single bubble properties.
3.1.Bubble formation
The fundamental study of the bubble formation behavior from orifices is important for understanding the bubble size variation in the system,particularly for the ca of a low gas velocity or a single orifice gas injector.In the situations,the behavior of bubble formation from the distributor mainly deter-mines bubble characteristics.There are two typical mechanical arrangements for bubble formation from a single orifice;that is with the orifice connected or not connected to a gas cham-ber.For bubble formation from a single orifice without a gas chamber,the gasflow rate through the orifice is always con-stant,which is referred to as constantflow conditions.The phe-nomenon of bubble formation from a single orifice connected to a gas chamber varies with gas injection conditions,which are characterized by the dimensionless capacitance number N c de-fined as4V c g l/ D2o P(Kumar and Kuloor,1970;Tsuge and Hibino,1983).When N c is smaller than1,the gasflow rate through the orifice is almost constant during the bubble forma-tion process,similar to thefirst mechanical arrangement.When N c is larger than1,the gasflow rate through the orifice is not constant,and it is dependent on the pressure difference between the gas chamber and the bubble.S
uch bubble formation con-ditions are characterized as variableflow conditions by Yang et al.(2000)or as constant pressure and intermediate conditions by Tsuge and Hibino(1983).
Numerous experimental and modeling studies have been con-ducted over the past decades on bubble formation from a sin-gle orifice or nozzle submerged in liquids,mostly under am-bient conditions(Kupferberg and Jameson,1969;Kumar and
6G.Q.Yang et al./Chemical Engineering Science 62(2007)2–
27
Fig.1.Snapshots of the 3D volume images of bubble plumes compared with photographs:U g =0.02m /s (from Warsito and Fan,2005).
Kuloor,1970;Azbel,1981;Lin et al.,1994;Ruzicka et al.,1997;Kulkarni and Joshi,2005;Zhang et al.,2005;Xiao and Tan,2006).A few studies were conducted at elevated pres-sures (La Nauze and Harris,1974;Idogawa et al.,1987;Tsuge et al.,1992;Wilkinson and van Dierendonck,1994).The studies indicated that an increa in gas density reduces the size of bubbles formed from the orifice.
Bubble formation in liquids with the prence of particles,as in slurry bubble columns and three-pha fluidized bed systems,is different from that in pure liquids.The experimental data of Massimilla et al.(1961)in an air–water–glass beads three-pha fluidized bed revealed that the bubbles formed from a single nozzle in the fluidized bed are larger in size than tho in water,and the initial bubble size increas with the solids con-centration.Yoo et al.(1997)investigated bubble formation in pressurized liquid–solid suspensions.They ud aqueous glyc-erol solution and 0.1-mm polystyrene beads as the liquid and solid phas,respectively.The densities of the liquid and the particles were identical,and thus,the particles were neutrally buoyant in the liquid.The results indicated that initial bubble size decreas inverly with pressure under otherwi constant conditions,that is,gas flow rate,temperature,solids concentra-tion,orifice diameter,and gas chamber volume.Their results also s
howed that the particle effect on the initial bubble size is insignificant.The difference in the finding regarding the parti-cle effect on the initial bubble size between Massimilla et al.(1961)and Yoo et al.(1997)is possibly due to the difference in particle density.
Bubble formation in a hydrocarbon liquid and liquid–solid suspension with significant density difference between the liq-uid and solid phas was investigated by Luo et al.(1998a)and Yang et al.(2000)under various gas injection conditions.A mechanistic model was developed to predict the initial bub-ble size in liquid–solid suspensions at high-pressure conditions.The model considers various forces induced by the particles,and is an extension of the two-stage spherical bubble formation model developed by Ramakrishnan et al.(1969)for liquids.In the two-stage spherical bubble formation model,bubbles are as-
>内裤的英文
