Moving Behavior of an Object in Gas-Solid Fluidized Beds

Mar.2005J. China Univ, of Mining Tech (English Edition)Vol 15 No. 1Moving behavior of an Object in Gas-SolidFluidized bedsWEI Lu-bin, WANG Geng-yu, HAO Liang, ZHAO Yue-minSchool of Chemical and Environmental Engineering, China University of Mining Technology, Beijing 100083, ChinaAbstract: The settling behavior of coarse particles in a gas-solid fluidized bed was experimentally studied by usingmagnetic tracer. It is well known that the calculation of terminal velocity is of interest in dense medium separationHowever, this problem has not been completely solved up to now. In this work, the terminal velocity of an object mov-ing in a gas-solid fluidized bed was experimentally measured and theoretically calculated. The experimental results in-dicated that the plastic viscosity and yield stress of the bed increase as the size of fluidized particles increases, but itaries little when some coarser particles are mixed with the fluidized particles. The resistance to a rising object was anorder magnitude greater than that to a settling object. The efficient buoyancy on a flaky object, which lies flatly on thegas distributor, was much less than that calculated by the Archimedes principle. The object does not always rise or settle with minimal projective area owing to radial motion of the fluidized particles. But in the lower part of the bed, thebar-shaped objects were likely with minimal projective area rising or settling.Key words: object motion; fluidized bed; settling velocity; coal dry beneficiationCLC number: TD 94: TD 9421 Introductionsion is much larger than that of fluidized particle. Theinteraction of particles and object has important effectA gas-solid fluidized bed possesses many prop-on fluidized bed separation. In this paper we discusserties similar to that of a liquid. It is natural to try to on calculation of settling velocity of object in theuse fluidized beds for dry beneficiation of mineralsThe first process for coal was described by Fraser and fluidized bed, behavior of rising object and effect ofthe physical characteristics of fluidized particles andYancey, and since then many studies have beencarried out 2-5. The gas-solid fluidized bed processthe shape of object on the behavior of object motionbelongs to dry dense medium concentration. It is wellknown that the settling behavior of separated material2 Apparatus, Method and MaterialPropertiand the calculation of terminal velocity are of interestin dense medium separation. However, this problemThe experiment apparatus is showed in Fig. 1has not been completely solved up to now 12-31The settling velocity of sphere is measured by fiveThere are two kinds of particles in the operation inductance coils winding on the outer wall of the bedof fluidized bed separation. Fine particles are used as with equal intervals of 40 mm. The inner diameter ofseparating medium to form fluidized bed, called as bed中国煤化工 e lowest coil to gasparticle. Meanwhile, coarser particles, called as ob- distCNMH Ger sphere passesject, are separated materiel whose geometry dimen- through an inductance coil, the inductance and qualReceived 20 October 2004: accepted 5 November 2004Projects(50474030,90210035, 50025411)supported by National Natural Science Foundation of ChinaCorrespondingauthor.Tel:+86-10-62331897,E-mailaddress:wlb@cumtb.edu.cnJ. China Univ of Mining& Tech (English Edition)ity factor of the coil will vary, thus leading to the like a Bingham fluid -. a popular way of calculatchange of output voltage. Therefore, the time that the ing drag force on the object in a Bingham fluid is totracer passes through the coil is obtained, then, the modify the Reynolds number of drag formula foraverage velocity of the tracer passing between two Newtonian fluids. If employing the empirical correlacoils can be known. The measurement of settling tion presented by Schiller and Naumann, the corvelocity in this work is more accurate than relation isDaniels(1+0.15Re0687(1)ValveRe=_dompeRotameterwhere Rem is the modified Reynolds number; CD isthe drag coefficient; do is the object diameter, pb isthe density of the fluidized bed; u is the plastic vis-Fig 1 Schematic diagram of experimental apparatuscosity; to is the yield stress; ur, is the relative velocityThe most tracers are spherical. Their diameters between the object and the fluidized particle; u =urange from 5.00 mm to 12.81 mm and their densities "p(where uo and p are the velocities of the object andfrom 1520 kg/m to 7830 kg/m. A few tracers are the particle, respectivelyflaky and bar-shaped. Their equivalent volume di-It is valid for Re<1000 and thus covers the rangemeters range from 8.36 mm to 13. 53 mm and their of the fluidized bed separation iI. When the fallingdensities from 130 kg/m'to 1480 kg/m Fluidized velocity o reaches the terminal settling velocity, theparticles are four grades of silica sands, their propecorrelation based on force balance can be written asties are listed in Tablel. The plastic viscosity and4d(P -P,gyield stress tabulated in Tablel are measured by usingfalling spheres. The mass percent of 0.2-0 1 mm where u, is the terminal slipping velocity of the fal-accounts for 8% in 0. 2-0.076 mm grade of fluidized ling object(terminal relative velocity between theparticles. In the settling experiment, firstly infuse air object and the fluidized particle), po is the objectto make fine particles fluidized, then put one mag- density. Based on Eq (1)and Eq(2),we can obtainnetic tracer onto the bed surface and measure the set-tling velocity of the tracer. In the floating experiment,Remap。-pbgfirstly lay the lighter object, whose density is less180+015Rcm063)pbthan bed density, in the bottom of the bed, then adjust(3)the gas velocity upper to the given value as soonpossible and measure its floating velocity.u,to can be measured by falling spheres, if pb, PoTable I Physical properties of silica sands and experi- do is determined, the corresponding u, can be obmental conditionstained by means of a numerical solution of Eq (3)0.3-0.20.2-0.10.2-0.0760.1-00768.564.351.40The object settling velocity can be recorded as uro, so115011001085HPa·s)06750.2630.0880092coarse obiect. uto>> up, therefore8.115.964.01Gas superficial velocity/(cm s )10.29 5.17232.17中国煤化工 plying E,(3)areCNMH Gd values. Fig.2compares calculated values with the experimental3 Calculation of Settling Velocityvalues of settling velocities for 0. 1-0.076 mm gradeThe studies show that the fluidized bed behaves of fluidized particlesWEI Lu-bin et alMoving Behavior of an Obiect in Gas-Solid Fluidized BedsMinus means that the velocity of particles is upward.0605The velocity of particles in the upper part of bed islarger than that of particles in the lower part of bed,which is agreeable with Lin's conclusion"!.Thecalculated velocity values approach to modified ex-Fig 2 Comparison of calculated settling velocity with ex- perimental values, however, the error between calcu-perimental data for denser and coarser objectslated values and modified experimental values tend tobe magnified with the decrease of the density andWith the decrease of object density and size, ve- size. Thus, a tendency can be derived that the settlingocity measuring is affected by the moving of fluid- area of light and small object transit from the ascendized particles. In this situation, ur equals uo up. It is stream of bed center to the descend stream of the bedimpossible for object to settle in a constant speed dueto the motion of fluidized particles. Comparatively,the settling velocity between the first coil and the Table 2 Calculated and measured results of settling ve-second coil is approximate to that of the second andlocity for lighter and smaller objectsthe third, recorded mu as object velocity in the upper No. l(kg m)mmu," eyalpart of the bed; the settling velocity between the third6.731.8350.33250.67and the forth is approximate to that of the forth and9.442.751.4-1.3520-0.758.185.821-30838-1.38the fifth, recorded u, as object velocity in the lower415308945.26203.2637-1.56part of the bed. Table 2 lists the calculated and ex515209.345.63253.134.0-1.63perimental values of settling velocities615808896342.1-4.244.5-184938143.04.146.3-0.84From Table 2 it appears that there are some dif- 8 1550 10.08 7.52 3.9_3.62 64 -1.12erence between u, calculated with Eq (3)and um, un 9 1670 9.11 8.67 3.5-5.17 6.2 -2.47measured through experiments. It indicates that the 10 1640 9.67 9.02 4.0-5.027.2-1.8200810214.65617.7-251motion of fluidized particles have a predominant ef-2175010.18124376-4.83110-143fect on settling behavior of object. Though the total 13 1700 10.97 12.77 7.7-5.079.9-2.871425100816311.05.3114.0-23average value of up is zero, the local motion of fluid1523409.0420.2915.05.29180-229ized particles still exists. The particle circulation pat904204416.0444190-144term is induced that particles ascend at the center anddescend near the wall of the fluidized bed. In spite of 4 Effect of Size on the Settling Behaviorthe velocity of fluidized particles up being not measured directly, we get the estimative values of particleThe size of fluidized particles has a predominantffect on the settling velocity. Fig 3 shows that thelocity up from un, un, and u, by applying Eq (4)settling velocity varies with different grade of silicaFrom Table 2, it is easily found that the values of unu, or uIr-u, is approximately equal for different desands as fluidized pasity and size, which shows that objects may settleFig 3 shows that the settling velocity of spherewithin the same region. In experiments of No I anddecreases with the increase of the size of the fluidizedNo. 2, the size or the density of object is so small thatparticles. With the increase of the size of the particles,its settling velocity measured has too much rarthe type of fluidized particles translates from Geldartdomicity because of the motion of fluidized particles.. Group A powder to Geldart Group B powderS;theSo they are not included when we calculate the aver-中国煤化工 plastic viscosity andage value of up. By averaging Wn u, and ur-ur of No. 3 yCNMH Ge. The velocity ofto No 16, the average velocity of fluidized particles in sphere in the grade of 0.2-0.076 silica sands is almostthe upper part of the bed upu is-4. 44 cm/s, thethe same as that of the grade of 0. 1-0.076, whichage velocity of lower part of the bed upl is-1. 82shows that the plastic viscosity and yield stress of theJ. China Univ. of Mining Tech(English Editbed vary little when a little coarser particles are when it ismixed with the fluidized particlesIn addition, some experimentdp=0.1-0.076btained from experiment show that the flaky objectis hard to float from the bottom of bed when gas ve"d=0.3-0.2mlocity is relatively small. It is probized particles have been a certain extent classified sothe fluidized particles in the bed bottom have notadequate activity if the gathough the gas velocity is large, the flaky object isFig 3 Variation of settling velocity with size of objects and also hard to float if it is laid flat on the gas distributorof fluidized particlesBut it can rise at once if the object is laid at a distancefrom the gas distributor. The floating velocity of flaky5 Effect of Object Shape on the settlingobject whose equivalent volume diameter is 10.73Behaviormm in Table 3 is measured with its initial position is25 mm away from the gas distributor. The experiIn the experiment, the author measured themental results suggest the initial settling velocity ofstudied the effect of object shape on the behavior of Separated material should approach zero when it isobject motion. Partial exptal results are given and low density particles, which should float onin Table 3. d, is the diameter of the sphere with surface of fluidized bed, its momentum is so lagerequivalent volume of the object, um for floating ve- when it is entering into the bed that it perhaps settleslocity between the first coil and the third coil, up for a long distance and is hard to rise. To avoidfloating velocity between the third and the fifth coil. low-density object sinking onto the bottom of fluidFluidized particles are 0. 1-0.076 mm grade of silica ized bed, we should make the initial settling velocitysands. Minus means that the velocity of objects is approach zero. The effect of object shape on floatingdownwardvelocity for the bar-shaped is larger than that of flakyTable 3 Experimental results of the effect of object shapeobject, which implied the rising or settling of objectis not always with minimal projective area. The pro-Size/mmdmm Po(kg·m)wmsjective area is affected by the radial motion of the8.10X8.50X84948×960×10.001202130or settle with minimal projective area in the lower1020×1158×109813.53part of the bed, probably because the radial motion is280×14.56X15.8610.735.46×6.10×17.6210.391504.673feebleness in the bed area7.20X840×860480×545×20.106 ConclusionsBecause the rising track and projective area of1) The relative settling velocity can be calcuobject has some randomicity, the result of floating lated by the equation belowexperiment is not exactly disciplinary. But we canlearn from the data of Table 3 that the floating veloc-Re,dp。-pBity is an order magnitude smaller than that calculated中国煤化工0by Eq. (3), which reveal the resistance to a rising ob-CNMHGject is an order magnitude greater than that to a set- where:Retling object with equivalent velocity. It may be thatthe dead area" is formed on the top of the object2) The size of fluidized particles has a predomiWEILu-bin et alMoving Behavior of an Obiect in Gas-Solid Fluidized Bedsnant effect on the settling velocity. with the increase4)The flaky object is hard to float from the bot-of fluidized particles size, the plastic viscosity and tom of bed when gas velocity is relatively small.Theyield stress of the bed increase and the settling veloc- flaky object is also hard to float if it is laid flat on theity will decrease. The plastic viscosity and yielegas distributor even though the gas velocity is largestress of the bed vary little when a little amount of5)The rising or settling of an object is not al-coarser particles is mixed with the fluidized particles3)The resistance to a rising object is one order ways with minimal projective area owing to radialmagnitude greater than that of a settling object. The motion of the fluidized particles, but in the lower partinitial settling velocity of separated material should of the bed, the bar-shaped objects tend to rise or settleapproach zero when it is infused into the separator.with minimal projective areaReferences[1] Fraser T. Yancey H F. Artificial storm of air-sand floats coal on its upper surface, leaving refuse to sink, Coal Age, 1926, 3325-327[2] Douglas E, Joy AS, Walsh T, et al. Development of equipment for the dry concentration of minerals. Filtration Separation,19739:532-538.[3] Chan E W, Beeckmans J M. Pneumatic Beneficiation of coal fines using the counter-current fluidized cascade. Int, J. MineralProcessing, 1982,9: 157-16[4] Wei L B, Chen Q R, Luo Z F. Fine coal and three product dry beneficiation with vibration and double-density fluidized bedsJournal of Central South University of Technology, 1998, 5: 104-107Chen Q, Wei L, Luo Z. Theory and practice of coal dry beneficiation with air-dense medium fluidized bed. ( Guo Y GGolosinskiT S Proc. 96 Int Symp. Mining Sci. Technology).Rotterdam: Balkema, 1996: 767-770[6] Daniels T C. Measurement of the drag on spheres moving through gaseous fluidized beds. J. Mech. Eng. Sci., 1962, 4103-110Wei L B, Chen Q R, Zhao Y M. Measuring the viscosity of fluidized suspension using falling sphere. Engineering Chemistrymetallurgy, 2000, 21: 187-190.(In Chinese[8] Anjaneyulu P, Khakhar D V. Rheology of a gas-fluidized bed. Powder Technol, 1995, 62: 261-267.[9] Langenberg-Schenk G Rietema K. The rheology of homogeneously gas-fluidized solids, studied in a vertical standpipe.Powder Technol, 1984, 38: 23-32[10] Wei L B, Bian B X Dynamics in Mineral Separating Processes. Xuzhou: China University of Mining Technology Press,2002.87-89.( n chinese)[11] Wei L B, Bian B X, Chen R Force on an object moving in dense fluidized bed. Jourmal of China University of Mining[12] Lin J S, Chen MM, Chao B T. A novel radioactive particle tracking facility for measurement of solids motion in gas fluid-ed beds. AIChE J, 1985, 31: 465473[13] Geldart D. Types of fluidizatio. Powder Technol, 1973, 7: 285-290中国煤化工CNMHG