喜歡這套資料就充值下載吧。。。資源目錄里展示的都可在線預(yù)覽哦。。。下載后都有，，請放心下載，，文件全都包含在內(nèi)，，【有疑問咨詢QQ:414951605 或 1304139763】 喜歡這套資料就充值下載吧。。。資源目錄里展示的都可在線預(yù)覽哦。。。下載后都有，，請放心下載，，文件全都包含在內(nèi)，，【有疑問咨詢QQ:414951605 或 1304139763】

Chemical Engineering Journal 78 (2000) 107113 Experimental investigation of the heat and mass transfer in a centrifugal uidized bed dryer M.H. Shi , H. Wang, Y .L. Hao Department of Power Engineering, Southeast University, Nanjing 210096, China Received 9 November 1998; received in revised form 25 June 1999; accepted 29 June 1999 Abstract An experimental study of the heat and mass transfer characteristics of wet material in a drying process in a centrifugal uidized bed (CFB) dryer was carried out. The rotating speed ranged from 300 to 500 rpm. Wet sand, glass beads and sliced food products were used as the testing materials. The gas temperature and the wet bulb temperature at the inlet and outlet, as well as the bed temperature, were measured. The moisture contents were determined instantaneously by the mass balance method in the gas phase. Inuences of the supercial gas velocity, particle diameter and shape, bed thickness, rotating speed of the bed and initial moisture on the drying characteristics were examined. One empirical correlation which can be used to calculate the heat transfer coefcients of the gas particles in the centrifugal uidized dryer were obtained. 2000 Elsevier Science S.A. All rights reserved. Keywords: Drying; Heat and mass transfer; Centrifugal uidized bed 1. Introduction Centrifugal uidized bed (CFB) drying is a new technol ogy in which the wet material undergoes a highly enhanced heat and mass transfer process in a centrifugal force eld by rotating the bed. The bed essentially is a cylindrical basket rotating around its symmetric axis with a porous cylindrical wall. The drying material is introduced into the basket and forced to form an annular layer at the circumference of the basket due to the large centrifugal forces produced by rota tion. The gas is injected inward through the porous cylin drical wall and the bed begins to uidize when the forces exerted on the material by the uidizing medium balance the centrifugal forces. Instead of having a xed gravitational eld as in a vertical bed, the body force in a centrifugal bed becomes an adjustable parameter that can be determined by the rotation speed and the basket radius. Minimum uidiza tion can, in principle, be achieved at any gas ow rate by changing the rotating speed of the bed. By use of a strong centrifugal eld much greater than gravity, the bed is able to withstand a large gas ow rate without the formation of large bubbles. Thus, the gassolid contact at a high gas ow rate is improved and heat and mass transfer can be achieved during the drying process. For this reason, the CFB dryer has received much attention in the drying industry. Corresponding author. Only a few research works dealing with drying in the CFB could be found in the literature. Lazar and Farkas 1,2 and Brown 3 have conducted the drying process in a CFB for sliced fruits and vegetables, while Carlson 4 investigated the drying of rice in the CFB. These research works are very instructive, but they are mainly focused on the possibility of an industrial application for CFB. The ow behaviour and drying characteristics in the CFB are very complicated and still unclear. A knowledge of heat transfer from the gas to the material is desirable in order to estimate the material surface temperature from the measured temperature of the gas. A quantitative knowledge of the heat transfer characteristics of CFBs is therefore necessary for design purposes 5. In this paper, an experimental study of the ow behaviour and gassolid heat and mass transfer characteristics in a CFB dryer was performed and the main factors which inuence the drying process were examined and discussed. 2. Experimental apparatus A schematic diagram of the experimental apparatus is shown in Fig. 1. A cylindrical basket rotated about a hor izontal axis is mounted in a sealed cylindrical casing. The basket is 200 mm in diameter and 80 mm in width. The side surface of the basket contains 3 mm diameter holes which serve as a gas distributor, with an open area of 22.7%. A 13858947/00/$ see front matter 2000 Elsevier Science S.A. All rights reserved. PII: S13858947(00)001480M.H. Shi et al. / Chemical Engineering Journal 78 (2000) 107113 109 Fig. 3. The uidized curve of sand in the CFB (d p D0.245 mm, nD400rpm). Material (up/down): (m/h) sand; (d/s) glass beads. speeds during the drying tests. In the initial uidizing stage, the pressure drop increases linearly with increasing gas ve locity. After reaching the critical point, the pressure drop will be almost constant. However, different results are observed for sliced and block materials. The pressure drop curve has a maximum value that corresponds with the critical uidiza tion point as shown in Fig. 4. In the initial uidizing stage, the pressure drop increases slowly with increasing gas ve locity. After reaching the critical point, the pressure drop will decrease with increasing gas velocity. This is because the selflock phenomenon of the sliced material under a cen trifugal force eld will be weakened and because the bed becomes uniform. This causes a decrease in the ow resis tance. Decreasing the bed rotating speed would decrease the bed pressure drop and the critical gas velocity remarkably, as also shown in Fig. 4. This is because decrease in the bed rotating speed would weaken the centrifugal force eld and cause the ow resistance to decrease. It can be seen from Fig. 4 that the critical uidized velocity for pieces of potato is somewhat smaller than that of blocks of potato owing to the Fig. 5. Intermittent drying curve in the CFB (sand, d p D0.411 mm, MD2.48 kg, !D41.9 rad s 1 , U 0 D1.71 m s 1 , H in D0.016 kg kg 1 ): (1) T g;in ; (2) T g;out ; (3) T b ; (4) R; (5) x. Fig. 4. The uidized curves for materials with different shapes: (4) pieces of potato 10 mm 10 mm 1.5 mm, nD300 rpm; (h) blocks of potato 5mm 5mm 5 mm, nD300 rpm; (s) block of potato 5 mm 5mm 5 mm, nD250 rpm. larger upwind surface area for pieces of material. Further more, pressure drop of the piece material bed is also smaller than that of the block material bed because the pieces of material show better uidization character in the CFB. The initial uidizing relationships obtained from the theoretical model for granular material 6 do not t the sliced mate rial. The initial uidizing conditions for the sliced material with different shapes should be determined experimentally and individually. 3.2. Drying curves Typical gas temperature and bed temperature curves as well as the drying curve of wet sand in the intermittent dry ing process are shown in Fig. 5. This shows that the drying110 M.H. Shi et al. / Chemical Engineering Journal 78 (2000) 107113 Fig. 6. Variations in the moisture content (Curve 6) and drying rate (Curve 7) for sliced potato. characteristics of materials like sand in the CFB, in which the moisture content is mainly surface water, are the same as in an ordinary dryer, i.e. the whole drying process can be divided into three stages. At a short initial stage, the material is preheated and the drying rate increases rapidly; the bed temperature is increased to a stable value. The second stage is a constant drying rate stage in which the heat transferred from gas to material is expended totally for evaporation of the surface water of the material. The material temperature remains constant and the drying rate is also constant. The last stage is called the falling rate stage in which the ma terial temperature increases gradually and the drying rate decreases until the end of drying. The drying behaviour for sliced food products in the CFB is somewhat different from sand as shown in Fig. 6. It is obvious that sliced potato has a drying character in the CFB that is basically similar to that in the conventional drying process. In the beginning, there is a short initial period. In this period, the bed material is preheated; the bed temper ature approaches a stable value quickly and the drying rate increases very rapidly. This initial period is followed by a period of a constant rate of drying. In the constant rate pe riod, the surface of the test material is covered with a thin water lm. The heat transferred from the gas ow to the ma terial is used completely to evaporate the moisture, so that the temperature of the sliced material remains at an equi librium temperature and the drying rate is at the maximum value. As the main moisture content in potato is cell water, the constant rate period is very short. The most important drying process is completed in the falling rate period. In the falling rate period, the dry layer appears and gradually becomes thicker near the surface owing to the larger trans port resistance of the inner moisture outward. This causes the heat transfer resistance to increase and the drying rate to decrease rapidly in the rst stage. After the dried layers temperature has increased to a certain value, a slow decrease in the drying rate occurs. This indicates that the falling rate period for the sliced potato in the CFB dryer can be divided into two different stages. This is signicant for engineering design and operation. The experimental results show that the pieces of potato in the drying process have a larger drying rate and a shorter drying time than blocks of potato in the CFB. This is be cause the transport distance of moisture from the inner cell to the outer evaporating surface in the pieces of ma terial is much shorter than in the blocks of material; in particular, the second stage of the falling rate period is shorter for the pieces of material during the drying process. In general, because the sliced material could be uidized and mixed very well in the CFB, the drying time is ex tremely short. For example, the drying time is 15 times shorter in the CFB for sliced potato than in the tunnel dryer and ve times shorter than in the conventional uidized dryer. 3.3. Inuences of the operational parameters 3.3.1. Supercial gas velocity It is obvious that an increase in the supercial velocity would increase the degree of uidization, and thus, the heat and mass transfer between the gas and the solid phase would be greatly enhanced. This causes the drying rate to be larger and the drying time to be shorter, as shown in Fig. 7. The critical moisture content would be increased with increas ing gas velocity, indicated by the broken line in Fig. 7. For food material, the experimental results show that the dry ing rate in the constant rate period and the rst stage of the falling rate period would increase with increasing gas velocity in the low gas velocity range. Thus, the total dry ing time would be decreased. However, when the gas veloc ity is increased to a certain value, the constant rate period would disappear, the rst stage of the falling rate period Fig. 7. The inuence of supercial velocity on the moisture con tent (d p D0.411 mm, MD2.50 kg, !D41.9 rad s 1 , H in D0.016 kg kg 1 ): (1) U 0 D1.66 m s 1 ; (2) U 0 D2.17 m s 1 .M.H. Shi et al. / Chemical Engineering Journal 78 (2000) 107113 111 would decrease and the second stage would increase. The total drying time would remain unchanged; this is because the main water content in potato is the inner cell water and the main drying process is in the second stage of the falling rate period. With an increase in the inlet gas temperature, the drying rates in all drying periods increase and the total drying time will decrease. However, the increase in gas tem perature would be limited by the quality of the dried food products. In our test, the best inlet gas temperature is about 100110 C. The experimental results also show that pieces of radish with given dimensions show a larger drying rate than pieces of potato under the same operating conditions. This is be cause the microstructures of the test examples indicate that radish has a larger cell structure with a more regular ar rangement than potato, and furthermore, the liquid in the radish cell is less viscous; these structural characteristics make radish easy to dry. 3.3.2. Rotating speed At the same gas velocity, a decrease in the bed rotating speed will reduce the centrifugal force acting on the material and increase the uidized degree of the material; this causes the heat and mass transfer between the gas and the solid phase to increase. Thus, when decreasing the bed rotating speed, the drying rate will be larger, as shown in Fig. 8, and the drying process will be much more uniform over the whole bed. This means that, for a given material drying in the CFB, the bed rotating speed should be as low as possible until the uidization state cannot be maintained. When it is desired that the drying process be enhanced by increasing the gas velocity in the CFB dryer, the bed rotating speed must be increased simultaneously to avoid the drying material from blowing out of the bed. Theoretically, the bed can be operated in the optimum uidized condition at any gas velocity by regulating the bed rotating speed in the CFB. Fig. 8. The inuence of rotating speed (d p D0.411 mm, MD2.41 kg, U 0 D 1.43 m s 1 , H in D0.0123 kg kg 1 ): (1) !D52.4 rad s 1 ; (2) !D41.9 rad s 1 . Fig. 9. The inuence of particle diameter (MD2.4 kg, !D41.9 rad s 1 , U 0 D 1.43 m s 1 , H in D0.0123 kg kg 1 ): (1) d p D0.245 mm; (2) d p D0.411 mm. 3.3.3. Partial diameter Fig. 9 shows the inuence of particle diameter on the drying behaviour in the CFB. It is clear that, owing to the larger slip velocity between gas and solid particles for parti cles with larger diameters, the heat and mass transfer in the drying process would be enhanced; thus, the drying rate in the CFB would increase with increasing particle diameter as shown in Fig. 9. However, with increasing material dimen sions, the internal heat and mass transfer resistance would be increased; thus, for a given material to be dried, it is im portant to determine the optimum material dimensions in the drying process under certain given operating conditions. 3.3.4. Bed thickness Fig. 10 shows the effect of initial bed thickness on the drying process. It can be seen that, with increasing bed thick ness, the drying rate would be decreased; this is because the heat and mass transfer driving force between the gas and the solid phase is larger in the shallow bed situation. Fig. 10. The inuence of bed thickness (d p D0.411 mm, !D41.9 rad s 1 , U 0 D1.43 m s 1 , H in D0.0123 kg kg 1 ): (1) L 0 D30 mm; (2) L 0 D20 mm.112 M.H. Shi et al. / Chemical Engineering Journal 78 (2000) 107113 Fig. 11. The initial moisture content (d p D0.411 mm, MD2.48 kg, !D41.9 rad s 1 , U 0 D1.71 m s 1 , H in D0.016 kg kg 1 ): (1) x 0 D0.221 kg kg 1 ; (2) x 0 D0.0574 kg kg 1 . 3.3.5. The effect of initial moisture content It is obvious that a material with a large initial moisture content has a much longer drying time (Fig. 11), but the drying characteristics are the same. The only difference is in the duration of the constant rate stage. 3.4. The heat transfer correlation Sixtyve experimental runs of wet sand and glass beads were carried out under the conditions of a static bed thick ness range from 10 to 30 mm, Reynolds number from 5.47 to 35.3 and centrifugal force from 10.08 to 28 multiples of gravity. The heat transfer coefcients were converted into Nusselt numbers using the mean diameter and the thermal conductivity of air at the average temperature. The dimensionless correlation of heat transfer between gas and particles in the CFB during drying is obtained by use of a regression procedure. The exponent of the diffusivity ratio (Prandtl number) has been assumed to be 1/3; thus, Fig. 12. Comparison of experimental and calculated results. Nu D 5:33 10 5 Pr 1=3 Re 1:59 F 0:48 c L 0 d p 0:21 s g 0:79 (7) The suitable parameter ranges for the above two correla tions are ReD5.042.0, F c D10.028.0. In Eq. (7), the Nus selt number is dened as NuDhd p / ; the Reynolds number is ReD g U 0 d p / ; the Prandtl number is PrDc pg / ; and then, dimensionless centrifugal force is dened as F c Dr 0 ! 2 /g. Comparison of the experimental heat transfer data with the values calculated by Eq. (7) is shown in Fig. 12. The deviation for all test data obtained in this work is within 25%. 4. Conclusions 1. The CFB may be operated in the packed bed, incipient uidization or uidized bed states at a given gas velocity. Steady uidized states can be maintained at large gas ow rates by using a strong centrifugal force eld. 2. There is no evident active region near the distributor of the CFB. The gassolid heat transfer comes under the inuence of the supercial gas velocity, particle diameter, particle shape factor, particle density, bed thickness and rotational speed of the bed. 3. The drying process can be divided into three stages in the CFB dryer and the drying rate increases with increas ing supercial gas velocity and particle diameter and de creasing bed rotating speed and initial bed thickness. 4. Sliced food products can be uidized and mixed very well in the CFB. The pressure drop curve has a max imum value and the critical uidized parameters vary with the shape and dimensions of the drying prod ucts and the material itself, as well as the operating conditions. 5. Sliced food products can be dried very well and ef ciently. The main process of drying is within the falling rate period; the drying rate depends on the shape, dimen sions and material of the drying products, as well as the operating conditions. 5. Nomenclature a particle surface per unit volume (m 2 m 3 ) c pg , c ps specic heat of gas or solid (J kg 1 C 1 ) d p mean particle diameter (m) D AB molecule diffusivity (m 2 s 1 ) F c dimensionless centrifugal force, r 0 ! 2 /g G mass ow rate of gas (kg s 1 ) h heat transfer coefcient (W m 2 C 1 ) H width of bed (m); wettability of gas (kg kg 1 ) L 0 xed bed thickness (m) M weight of dried material (kg)M.H. Shi et al. / Chemical Engineering Journal 78 (2000) 107113 113 n rotating speed of the bed (rpm) Nu Nusselt number, hd p / 1P pressure drop (kPa) Pr Prandtl number, c pg / R drying rate (kg m 2 s 1 ) Re Reynolds number, U 0 d p / T temperature ( C) U 0 supercial gas velocity (m s 1 ) x moisture content (kg kg 1 ) Greek letters porosity heat conductivity (W m 1 C 1 ) viscosity of gas (kg m 1 s 1 ) kinematic viscosity of gas (m 2 s 1 ) g , s density of gas or solid (kg m 3 ) s sphericity ! angular velocity (rad s 1 ) Acknowledgements This project was supported by the National Natural Sci ence Foundation of China. References 1 M.E. Lazar, D.F. Farkas, The centrifugal uidized bed. 2. Drying studies on piece form foods, J. Food Sci. 36 (1971) 315319. 2 M.E. Lazar, D.F. Farkas, J. Food Sci. 44 (1979) 242246. 3 G.E. Brown, D.F. Farkas, Centrifugal uidized bed, Food Technol. 26 (12) (1972) 2330. 4 R.A. Carlson, R.L. Roberts, D.F. Farkas, Preparation of quick cooking rice products using a centrifugal uidized bed, J. Food Sci. 41 (1976) 11771179. 5 D.F. Hanni, D.F. Farkas, G.E. Brown, Design and operating parameters for a continuous centrifugal uidized bed drier, J. Food Sci. 41 (1976) 11721176. 6 C.I. Metcalfe, J.R. Howard, Fluidization, Cambridge University Press, Cambridge, 1978, pp. 276327.