1、PDF外文：http:/ B：英文原文及翻译 A study on the two-phase flow in a stirred tank reactoragitated by a gas-inducing turbine abstract Experimental and numerical studies of a gasliquid two-phase flow have been applied to a non-baffled laboratory-scale stirred tank reactor, mechanically agitated by a gas-in
2、ducing turbine. The dispersion of air as gas phase into isopropanol as liquid phase at room temperature under different stirrer speeds was investigated. The X-ray cone beam computedtomography (CBCT) measurements have been taken at five different stirrer speeds starting from 1000 rpm at which the gas
3、 inducement occurs for the given operating conditions.The considerable difficulties in acquiring the phase distribution due to beam hardening and radiation scattering effects have been overcome by developing a suitable measurement setup as well as by calibration and software correction methods to ac
4、hieve high accuracy. The computational fluid dynamics analyses of the stirred tank reactor have been performed in 3D with CFX 10.0 numerical software. The simplified numerical setup of mono-dispersed bubbles, constant drag coefficient and the ke turbulence model was able to capture both the bubble i
5、nduction and dispersion and the free surface vortex formation. Despite the assumed simplifications, the numerical predictions exhibit a good agreement with the experimental data. Keywords:Stirred tank reactor；Gas-inducing impeller；CFD；X-ray computed tomographyMixing. 1. Introduction Gasliquid
6、mixing in stirred tank reactors is a commonprocess in the industry. It is regarded as one of the mostdifficult to tackle because of its complexities in terms of flowregimes and multiphase operations. Traditionally, the gasliquid stirred tankreactor is equipped with an impellerresponsible for dispers
7、ing the gas phase, which is usuallysupplied via a single pipe or a ring sparger mounted beneaththe impeller. The gas-inducing impellers provide an alternativegas injection, in which case the gas is sucked via ahollow shaft and fed directly into the stirrer region (Evanset al., 1990). More gas bubble
8、s can be broken-up into smallones when such configuration is applied, which consequentlycould provide higher mass transfer (Rigby and Evans, 1998).Among the long lasting efforts to establish precise butpractical measurementtechniques for the analysis of multiphasefluid dynamic processes in chemical
9、reactors (Boyeret al., 2002), methods based on ionising radiation are mostpromising since they are applicable at higher gas fractions,and they give linear measurements regardless of the structurecomplexity inside the vessel. An advanced tomographictechnique is cone beam X-ray computed tomography (CB
10、CT).With CBCT, a volume densitydistribution is reconstructedfrom a set of two-dimensional radiographs obtained from anobject at different projection angles. This technique isespecially suitable for time-integrated gas fraction measurements.The use of X-ray CT for gas hold-upmeasurements hasbeen desc
11、ribed by Pike et al. (1965), and recently by Hervieuet al. (2002), with application to two-phase flow in a pipe, byKantzas and Kalogerakis(1996), who monitored the fluidisationcharacteristics of a fluidised bed reactor, by Reinicke et al.(1998), and Toye et al. (1998), who used it in packed catalyst
12、beds, and by Vinegar and Wellington(1987),who measuredfluid transport in porous media. All the above-mentionedtechniques yield time-averaged rather than instantaneousphase distribution images. Rotationally symmetric materialdistributions such as thephase distribution in an un-baffledreactor enables
13、even a rather fast tomography, since oneradiographic projection is sufficient to compute a completeaxial and radial gashold up profile. Though CBCT is widelyused today in medical imaging and material research, it hastwo inherent problems that must be addressed for quantitativegas fraction measuremen
14、ts in a stirred chemical reactor. Itis inevitably necessary, in particular, to devise specialcorrection steps to account for beam hardening andscatteredradiation. With such corrections, the calculated gas fractionvalues will fit within the error limits of less than four percent(Boden et al., 2005).
15、Modelling a stirred tank using computational fluid dynamics(CFD) requiresconsideration of many aspects of the process(Marden Marshall and Bakker, 2002). The geometry of the stirredtank, even in the case of a complex one especially when theimpeller is explicitly modelled, needs to be embedded in acom
16、putational grid. Special care has to be taken to account forthe impeller motion especially at low rotating speed when theturbulence and the corresponding turbulence damping wallfunctions have significant effect on the flow (Ranade, 2002).Lane et al. (2005), along the lines proposed by Brucato et al.
17、(1998), have shown that in order to acquire the correct phasedistribution in the vessel, inclusion of the non-drag forces andmodification of the drag laws were necessary. Micale et al. (2000)have reached similar conclusions when studying solid suspension in agitated vessels. However, Torre et al. (2
18、007) managed toobtain good agreement between the numerical simulations andthe experimental observation of the free surface vortex plottedfor a gas void of 0.50.9 with simplifying assumptions of aconstant bubble size, a constant drag coefficient and the use ofthe ke turbulence model. Using advanced X
19、-ray computedtomography, the current study demonstrates the CFD abilitiesof predicting the two-phase flow in the stirred vessel when gasinducing impeller is employed. 2. Experimental investigation A detailed description of the measurement setup, thecorrection and reconstructioncalculations and
20、 their evaluationis given elsewhere (Boden et al., 2005), and is only brieflyreviewed here. 2.1. Tomography measurement setup The cone beam tomography setup consists essentially of arotating anode X-ray source(DI-104H-22/60-150, COMET AG, Switzerland) fed by a medical type high voltage generator(MP6
21、01, Ro ntgenwerk Bochum, Germany) and a two dimensionaldigital X-ray flat panel imager (RID1640AL1, Perkin-Elmer Optoelectronics GmbH & Co. KG, Germany) arrangedopposite to each other as shown in Fig. 1. The X-raysource may beoperated at voltages up to 125 kV andelectron currents of maximum 800
22、mA in single exposuremode. The detector provides 1024 by 1024 pixels, each0.4 mm by 0.4 mm in size. The X-ray setup was assembledinside a shielded box. The stirred reactor was placed between the source and the detector at the distances givenin Fig. 1. To ensure the quantitative radiographic image qu
23、ality foreach single exposure, a special data acquisition proceduredescribed below was adopted. The detectors dark currentsignal was determined before each exposure and wassubsequently subtracted from acquiredimage data. QuantitativeX-ray intensity measurements synchronized to the Xraygeneration may
24、 be degraded due to the detectors delayingbehaviour (also known as image lag). Therefore, the detectorssignal integration time was set to persist well beyond theshutoff of each individual X-ray exposure. The duration of anX-ray exposure was 6.3 s which was the longest permissibletime given by the X-
25、ray generator. For the particular reactorgeometry under investigation,this integration time periodwas found to besufficient to achieve the required signal tonoise ratio which means that dynamic processes with time constants of several seconds, such as changing of the stirrerspeed, feeding, extractio
26、n, and mixing, can be observed in realtime. Fig. 1 Schematic view of the CBCT setup. The model fluid used was isopropanol at normal pressureand room temperature. Thecritical stirrer speed at which gasdispersion at the stirrer blades occurs was estimated by opticalobservation to be of 1020 rpm.
27、 Then, the stirrer wassuccessively driven withspeeds in the range of 10001200 rpmat 50 rpm intervals. For each operation point, a CBCT scan wasperformed. 2.3. Scattering correction In an additional measurement, a moving slit technique (Jaffeand Webster, 1975) was applied to synthesise an almostscatt
28、ering free radiographic image of the reactor at moderatestirrer speed well below its critical value. The differencebetween such an image and an un-collimated cone beam Xrayradiography of the same arrangement gives a measuredvalue for the amount of scattered radiation intensitydistribution in the det
29、ector plane. Subsequently acquiredX-ray intensity distributions have been reduced by thatamount to eliminate the contribution of scattered radiationwhich otherwise wouldultimately lead to quantitative errorsin the reconstruction process. This approach isreasonable,since by dispersing little amount o
30、f gas into the fluid, theoverall mass of the object is conserved and there are onlyslight changes in the material distribution and thus onlyslight changes in the amount of scattered radiation areexpected. 2.4. Beam hardening correction Polyenergetic X-ray radiation would be hardened whenpenetrating
31、thick materials, as the effective attenuationcoefficient becomes smaller with increasing penetrationdepth. If uncorrected, this leads to systematic errors inquantitative X-ray measurements. The adopted method forbeam hardening correction can be illustrated with theexperimental setup presented in Fig
32、. 1. Two radiograms, oneof the reactors completely filled with the fluid and anotherfor the same arrangement plus an additional acrylic plate ofthickness d = 0.01 m between reactor vessel and the detector were taken. Both images were synthesised from a series ofslit images according to the method de
33、scribed above, andthus are assumed to be almost free of scattered radiation.From both images, the calibration extinction radiogramE c (r S , r D ) can be computed according to thefollowingequation DSp la te DSe ffDSDSp la teDSc drrl rrlrrE,00c o s),( ),(ln),(
34、; (1) where I denotes the measured intensities, plateDSeff , is theeffective attenuationcoefficient of the acrylic plate accordingto a certain ray path between source and detectorposition r S , r D , and DS,is the angle between the ray andthe detector normal. The indexes S and D stand for sourceand detector respectively. After that, the plate is removed.Now anyimage taken from the reactor with another fluidgasdistribution inside is processed to the extinction radiogram
