— 4 - A - A TEMPERATURE AND AIR FLOW ANALYSRS IN A SPRAY DRYER “nest: for $0 Degree of M. 5. MICHIGAN STATE UNIVERSITY Jhareswar Prasad Pal 1959 ‘féFQQS M ichigan State University TEMPERATURE AND AIR FLOW ANALYSIS IN A SPRAY DRYER by Jhareswar Prasad Pal AN ABSTRACT Submitted to Michigan State University of Agriculture - and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Engineering . 1959 Approved 20% u/ M, J/”/"7 ABSTRACT A fundamental study of engineering aspects involved in milk spray drying is important to improve efficiency. Air temperature, air velocity and time of drying greatly influence the efficiency of drying. Temperature patterns were obtained inside a Spray dryer under actual operating conditions. The air flow pattern inside the dryer was determined. Unfortunately, very few attempts have been made by pre- vious research workers to determine temperature and air flow patterns in a large scale spray dryer during operation. The droplets and dried particles tend to follow the air flow in a spray dryer, which is of considerable value in predicting the air flow patterns along with temperature patterns. A temperature and air flow analysis was made in a spray dryer under operating conditions. Temperature and velocity patterns of various cross-sections of the dryer are presented. Approximately 90 percent of the drying process is accom- plished within the first five feet of the drying chamber. The effective utilization of dryer chamber was limited within ten feet of its length. Possibly a dryer having twelve feet length would be as good as the existing dryer which has a length of seventeen feet. It was also observed that a faulty air flow design and an uneven temperature distribution would be greatly responsi- ble for poor spray dryer performance. TEMPERATURE AND AIR FLOW ANALYSIS .IN A SPRAY DRYER by Jhareswar Prasad Pal A THESIS Submitted to Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Engineering 1959 TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . . . OBJECTIVE. . . . . . . . . . . . . . REVIEW OF LITERATURE . . . . . . . . PROCEDURE . . . . . . . . . . . . . Part I. Temperature Measurement Part II. Air Flow Pattern . . . PRESENTATION AND DISCUSSION OF DATA. CONCLUSIONS. . . . . . . . . . . . . SUGGESTIONS FOR FURTHER STUDY. . . . REFERENCES . . . . . . . . . . . . . APPENDIX . . . . . . . . . . . . . . ii Page 12 12 16 21 55 55 56 57 ACKNOWLEDGEMENTS The author wishes to express his deep gratitude to Dr. Carl W. Hall for his consultation, suggestions and comments in conducting the experiments and the preparation of this thesis. Sincere thanks is expressed to Dr. J. R. Brunner of the Dairy Department for his valuable assistance. The use of the facilities of the Dairy Department was greatly appreciated by the author. Sincere thanks are due to Dr. T. I. Hedrick for his assistance in arranging the facilities in the Dairy De- partment. The author wishes to express his sincere appreciation to Dr. Arthur W. Farrall, Head of the Department of Agricultural Engineering and to Dr. Merle L. Esmay of the Department of Agricultural Engineering for their efforts in arranging the Graduate Assistantship. iii LIST OF FIGURES Figure Page 1 General view of the spray dryer . . . . . . . lO 2 Assembly View of the apparatus. . . . . . . . 11 5 Layout of the apparatus . . . . . . . . . . . 15 h Cross-sectional temperature patterns inlet air at 220° F (1 to 6 feet from frontend)................19 5 Cross-sectional temperature patterns inlet air at 2200 F (7 to 10 feet from frontend)0.0000000000000020 6 Longitudinal temperature patterns inletairatZZOoF..o.........22 7 Cross-sectional temperature patterns inletairat500oFcecccccoccee21+ 8 Longitudinal temperature patterns inletairatioooFeceeeecoecoo25 9 Longitudinal temperature patterns inlet air at 555° F . . . . . . . . . . . . 26 10 Longitudinal temperature patterns inlet air at 555° F . . . . . . . . . . . . 27 ll Cross-sectional temperature patterns inlet air at 555° F . . . . . . . . . . . . 29 12 Velocity profiles of air flow . . . . . . . . 50 iv LIST OF TABLES Page 1. Air flow data ------------------ - 18 INTRODUCTION The proper drying and storing of milk is an important operation today. There has been a demand for dry milk powder at low moisture content in recent years in recognition of the inverse relationship between moisture content and storage sta- bility. This has added to the difficulty of dryer operation, particularly in drying highly heat sensitive materials, such as milk. Of all the methods of drying of milk the Spray dry- ing is by far the best method used in industry. One of the major advantages of spray drying is that heat damage may be held at a low level. Product density can be varied within a given range. Particle size of the product may be controlled or varied in a given range by a proper variation in the oper- ating conditions. Spray drying frequently preserves the qual- ity of the product because the wet droplets in the hot drying zone are kept cool by evaporation and the total contact time can be controlled so that the dried product does not remain in the dryer long enough to become overheated. Owing to the lack of adequate data on which to base de- sign, spray dryers generally have been built on an empirical basis. Widely different designs have been evolved. The ef- fect of variation in operating conditions on powder charac- teristics may not be the same in all dryers, and the complex- ities of dryer operation tend to obscure these relationships even in the same dryer. Hence a fundamental study of engineer- ing aspects involved in spray drying is necessary to improve the efficiency of spray drying. Factors which influence heat damage, therefore, such as temperature, air velocity, and time of drying, and those determining the moisture content of the powder are of primary importance. These factors greatly in- fluence the efficiency of the drying operation. No study has been made to measure these factors which may provide use- ful information to improve the design of spray dryers. Hence there is a need to analyze temperature and air velocity inside a spray dryer. Moreover, one important aspect of spray dryer design and operation, besides atomization, is the prediction and control of air flow patterns within the drying chamber. Relatively little experimental work has been done on this phase of spray drying. Because the droplets and dried parti- cles tend to follow the air flow in a spray dryer it is of considerable value to predict these flow patterns along with temperature patterns. OBJECTIVE The chief interest was concerned with the study of tem- perature and air flow patterns inside a horizontal co-current milk spray dryer. The main objective was to determine how effectively the dryer was utilized for the drying process and to find the main drying zone from the study of temperature and air flow patterns. REVIEW OF LITERATURE Dried milk is the most important of the dried products of milk. References to milk powder date back to medieval history. To Marco Polo (125h-152h), the celebrated Venetian traveler, is attributed the reference to a dried milk by the Tartars during the 15th century. Next we hear of dried milk in tablet form made in 1810 by the French scientist, Nicolas, from milk concentrated slowly to a dough-like consistency in a current of dry air. Milk drying processes were perfected for commercial use in the last half of the 19th century. Spray drying has been successfully practiced in special ap- plications since the last quarter of the 19th century (1888). Extensive application of this drying process in the chemical industries did not occur until nearly fifty years later. Spray drying has been used with varying degrees of success in a wide variety of industries. It has been used to dry milk, soaps, detergents, pharmaceuticals, fine chemicals, organic and inorganic chemicals. The annual production of dried skim milk in the United States increased from Al million pounds in 1920 to h81 mil- lion pounds in 19uo and 676 million pounds in l9u7. Milk is spray dried in quantities approaching a billion pounds annu- ally, and the product is widely used in animal and human food. The increase in consumer packaging of nonfat dry milk is in h keeping with the marked per capita civilian use of nonfat dry milk which has risen steadily from 0.2 pounds in 1920 to 1.5, 2.2, 5.6 and u.2 pounds in 1950, 19ko, 1950 and l95h, respectively. Latest figures indicate that the a- mount of nonfat dry milk sold for packaging continues to increase at about 2h percent per year (1955). Spray drying usually is visualized as involving three operations: (a) atomization, (b) evaporation from drops, and (c) flow and mixing of gases and particles. The dis- tinguishing characteristic of spray drying is the large sur- face of product presented to the heated air. Because of large surface area, the drying rate is rapid and the drying time is very short. Drying usually occurs in two more or less distinct stages. The first is a period of surface evaporation known as constant rate period and the second, a falling rate period, during which the rate of evaporation continually decreases. The evaporation of pure liquid drops would involve only the first or constant rate period. The drying of droplets from atomizers in a spray dryer is a si- multaneous heat and mass transfer operation in which heat for evaporation is transferred by conduction and convection from hot gases to the drop surface and vapor is transferred by diffusion and convection back into the gas stream. The overall rate of drying is a function of the temperature, hu- midity, and transport properties of the gas, the diameter, temperature, and relative velocity between the drop and its surroundings, and the nature of solid material dissolved or suspended in the liquid. The principle of spray drying is that the milk is atom- ized to a fog-like mist in the presence of currents of heated air. The countless number of milk particles and the exceed- ingly minute particle size so produced, expose a vast area of milk surface to the hot dry air currents. The milk particles surrender the moisture practically instantaneously and drop to the bottom of the drying chamber in the form of small grains or flakes of dried milk. Several types of atomizers are inuse, namely, the compressed air atomizer, the pres- sure spray atomizer, and the disk centrifugal atomizer. In a compressed air atomizer a Jet of high temperature compressed air passes at high velocity through a stream or through mul- tiple streams of preheated milk. The milk streams issue from a battery of simple nozzles. The high velocity jet of hot air at 500° F strikes the milk streams at a right angle, atom- izing the milk intensity. This method of spray drying is used by some of the European milk drying firms, but has been largely superseded by the pressure spray. Pressure spray nozzles are extensively used in American milk drying factories. The milk is forced under high pres— sure (usually about 2500 to 5000 pounds) through a small or- ifice. Uniform fineness of spray and maximum efficiency des- iccation demand uniform pressure of the milk. A centrifugal Spray atomizer consists of a disk or slotted basket that revolves at a velocity (about 5000 to 20,000 r.p.m.) depending upon the diameter of the rotating device. The centrifugal force of the rapidly rotating disk causes the milk to spread in a film of uniform thickness toward the periphery where it is thrown off in the form of a thin veil resembling a fog. The centrifugal atomizer has the advantage of absence of small orifices that are subject to clogging. It thus permits the spray drying of highly concentrated milk (containing as high as 50 percent solids) and no pump pressure is required. Atomization with centri- fugal or swirl-type pressure nozzles is produced by liquid turbulence and by the attenuation effect of the tangential velocity component of the liquid. Tate and Marshall (1955) and others have shown that the tangential velocity usually has a greater effect than the axial component in causing liquid breakup. The exact mechanism of breakup in the swirl- type nozzles is not completely understood for all pressures. At low pressures, the cone formed by the nozzle is readily observed to be a film or sheet of liquid which expands out- ward from the orifice until the sheet thickness is no longer sufficient to resist the disruptive forces created by the waves and undulations in the sheet. A simple classification of spray dryers on the basis of air-flow direction in the chamber relative to the spray can be made. Three principal flows may be termed: (1) co- current, (2) counter-current, and (5) mixed flow; that is, combined co-current and counter-current flow with reference in each case to the relative directions of the air and spray. Each of the three classes indicated above may be further sub- divided into the categories of (a) straightline flow, or (b) vertical or spiral flow. It has been found from studies that counter-current flow does not permit so much flexibility of inlet air arrangements as does co-current flow. Mixed flow spray dryers may also take a variety of forms. Since these dryers, as the name implies, combine both co-current and counter-current features, the resulting air flow patterns are complex, and there is a high degree of turbulence and mixing in the drying chamber. Edeling (1950) reported some experimental and theoreti- cal considerations of air and particle motion in spray dry- ing. He reported the concept of spiral motion of air as it passes through a drying tower with reSpect to the spray from a pneumatic atomizer. Thordarson (1952) made some observations of the air flow patterns in a transparent, plastic spray dryer model and it was possible to photograph air flow patterns for various air- inlet arrangements. When the air entered through four inlets at an angle to the radius, the diameter of the spiral appeared to become smaller the more closely the inlets coincided with the radius. When the air was caused to enter the chamber in spiral motion through a central opening in a false ceiling, the pattern developed is shown to be spiral flow of air out of a central opening. The study on the effects of drying air temperature on bulk density by Duffie and Marshall (1955) showed that bulk density decreases with increasing drying air temperature. This effect is attributed to an increase in particle size with air-temperature increase, resulting in a decrease in particle wall thickness and hence in particle density. There have been very few attempts by previous researchers to study the temperature pattern inside a spray dryer. 10 .nfipaaamne on» co .3? Seasons .m .mE PROCEDURE Part I. Temperature Measurement The testing apparatus used consists of basically a frame, a potentiometer, a selective switch system, a ther- moflask for reference junction, copper-constantan thermo- couples and a thermometer. A frame was built which would fit inside the Rogers stainless steel dryer chamber. This frame, Fig. 2, was built into two sectors that could be folded together at the middle to facilitate easy handling. Thirty-eight thermocouples were fastened at different grid points one foot apart as shown in Fig. 5. These thermo- couples were connected to a selective switch systan. A portable type potentiometer (Leeds and Northrup Company, serial number 1501h75) was used to measure the millivolt readings and corresponding temperatures were recorded from a thermocouple table. The reference junctions of these thermocouples were maintained at 52° F. Practical diffi- culties were encountered because of the unfavorable condi- tion for stabilizing the frame in position inside the dryer. The milk powder dust, humidity, and high.temperature are most unfavorable conditions to get accurate readings. How- ever, much care and precaution was taken to get accurate readings under the existing conditions. 12 z”? / ' 22 2 3 23 24 25 4 5 26 27 2—7—3 9 2e —2‘9J' Io II I2 so my DRYER T3 I4 I5'_'3'2_ v \ I6 l7 IS 34 3 . \Ie Eiffel-‘7 FRAME 3 lm.__4, REF-JUNCTION SELECTIVE SWITCH FIG-3° 3 Erin] POTENTIOME TE R LAYOUT OF THE APPARATUS Concentrated skim milk at ho percent solids and at 155° F was used for drying. The inlet air'temperature was maintained at 500° F and the outlet air from the dryer was maintained at 200°F. The thermocouple readings were recorded at different cross-sections at every two feet interval along the dryer. The experiments were repeated to obtain constant readings. During this period of experiment the feed rate was maintained at 20.6 gallons per hour nozzle. Two atomizers were opera- ting at a pressure of 2500 psi. No. 69 orifice size (0.0292 inch diameter) was used for the nozzles. This apparatus could not be used for first four feet of the dryer where approximately 90 percent of the drying takes place as milk droplets start accumulating on the thermocou- ples. A separate apparatus was built to mount thermocouples. It consists of a 2 inch diameter rod 12 feet long. The appa- ratus with ten thermocouples fastened at one foot intervals was held horizontally inside the dryer to measure temperatures. This apparatus worked satisfactorily in the main drying zone. Although there was a little deposition of milk powder on the thermocouples, satisfactory readings could be taken under the existing condition. To check any effect on thermocouple read- ings due to different types of insulation, a copper-constantan thermocouple (20 gage B.S. cotton braid over enamel) was com- pared with a second'set of thermocouples. The second set of thermocouples was made of duplex 2h gage cOpper-constantan wire having an insulation of glass fiber braid, impregnated ‘ over double-wrap glass fiber impregnated over enamel. The second insulation is a good resistant to high temperature and humid conditions. They gave the same readings for the same measuring junctions. The experiments were repeated and the same procedure was followed to record temperature patterns inside the dryer under different operating conditions. During the second period the inlet air temperature was maintained at 220° F and the outlet air temperature was maintained at 190° F. The central atomizer alone was used in this case at 1000 psi. The feed rate of concentrated.milk having no percent solids was 15 gallons per hour. Under this operating condition a set of readings was taken for the temperature pattern at dif- ferent cross-sections inside the dryer. To collect more information on temperature pattern, the same procedure was followed to record temperatures at dif- ferent cross-sections when no milk was dried. During this experiment inlet air temperature was maintained at 555° F and the outlet air temperature was at 510° F. Readings were taken three times for a particular point to get consistent readings. Difficulties were encountered in conducting the experi- ments on a full scale commercial dryer. In the first place, the dryer is not accessible when in operation. Secondly, it takes considerable length of time before the systen comes to an equilibrium state and becomes ready for conducting ex- periments. Since it is a commercial dryer and was not opera- ted frequently, it took a long time to conduct a limited num- ber of experiments. Part II. Air Flow Pattern The second part of the exnriment*was to find the air flow pattern inside a horizontal co-current spray dryer under operating conditions. To make a preliminary investigation on air flow a conventional type of pitot tube was built with.a % inch copper tubing. A copper lead tube having the same dia- meter was used to withstand high temperature. One end of the copper tubing was attached to the pitot tube by means of a coupling and the other end was connected to a U-tube mano- meter outside the chamber. This apparatus could not be used because of collection of milk powder on the pitot tube open-' ing and secondly, there was not much.pressure drop due to low velocity of air. No other suitable instrument was available that could be used to measure air velocity inside the inac- cessible closed stainless steel drying chamber under opera- ting conditions. Considering the difficulty of measurement, it was decided to study air flow patterns inside the dryer when no milk was being dried but other conditions were the same. A vane anemometer (1% inch diameter) was used to mea- sure the air velocity at different positions inside the dry- er. Vane anemometer readings were noted at different heights along the dryer. 17 The dryer is provided with a turbo-blower from the North American Manufacturing Company which delivers 500 cubic feet per minute at 5&20 rpm at 15.8 inches water column. At the exhaust and of the dryer there is an ex- haust fan from the American Blower Corporation, size 15, which operates at 1790 rpm and delivers ALDO cubic feet per minute at 11.5 inches water column. The dryer works under a partial suction draft. TABLE 1 AIR FLOW DATA Velocity,AI§et]sec. Distance From Vertical Distance Front End, From Bottom, Forward Reverse feet feet Flow Flow 2 l 5.01 g l 5.05 1 5.06 8 1 5.16 10 1 5.18 2 2 5.8M % 2 5.87 2 5.87 8 2 5.87 10 2 5.92 2 5 3-89 2+ 5 5.9 5 5-9 8 5 3-9 10 5 5.92 2 6 1.855 g 6 1.86 6 2.155 8 6 2.17 10 6 2.19 2 8 k.o I 2 5. 5 8 8 5.80 10 8 5.uo mu 1 MI 2 so”! OUI L! TAII I90”! I' FROM ‘FQNTEND .NLETAIR 220" )UTLET All"! IBOT J'bs'm HEIDI-”END .NLETAIR 22°F OUTLE T Alli ISDF a; __/ ‘I‘I‘JI 'R"Nl (NF 19 IbLt'AII . 7 ‘1'. "JD! I! l AIR .101 7'FI9I F‘I‘hT 6‘“: mi: 1 Am 2 20 I out LE 1 AIR Ieo'r |N_[YA 522C" OT'LIYM‘ '50‘ -. "ARIN? [ND Fig. h. Cross-sectional temperature patterns (1 to 8 inlet air at 220° F feet from front) INLET AIR 220‘F OUTLE TAIR I90'F 7' FROM FRONT END INLETAIR2 20’F OUTLETAIR I90‘F A Is lest-1901= ,, \\\C 9 FROM FRONT END 20 INLETAIR 220'F OUTLET AIR I90‘F B'FROM FRONT ENI' INLET AIR 220‘s OUTLETAIR I90'F Lax I 70'F IBO'F I90'F I95‘F (If)? IO'FROM FRONT END Fig. 5. Cross-sectional temperature patterns inlet air at 220° F (7 to 10 feet from front) PRESENTATION AND DISCUSSION OF DATA Three sets of data collected under the three operating conditions were recorded. Temperature contours were pre- pared for different sections. A study of the temperature patterns in Fig. h and 5 shows that there was uniform dis- tribution of temperature inside the dryer at a particular cross-section. The temperature gradients varied uniformly with few exceptions. It was observed that there were some local low temperature zones. A close look at these tempera- ture patterns also showed that the cold spots were either within or below the main drying front. It could be explained in terms of the physical phonomenon that the hot air came in contact with atomized milk spray, and it transferred much of its heat to the relatively cold atomized milk Spray. There was a tendency for this cold air to settle down due to higher density. The second reason for the formation of low tempera- ture centers was due to a low velocity air mass formed at a separation zone of air velocity. In Fig. A and 5 it was ob- served that temperature gradients did not follow any definite trend. At some cross-sections it had the tendency to move upward whereas at other sections, it had the reverse trend. The wide variation of this trend might be due to non-uniform distribution of atomized milk spray in the dryer. Secondly, uneven distribution of air and eddy current effects inside the dryer probably accelerated this abnormality. 21 22 m oomN ad has poaaa anaconda oaapohonaou amnaugumnoq .w .mfim 0.... 0.4 20:.Qum JQZlQDFSVJOd ‘ . panacea so: tall... , xxxmbmm ¢.<»o: hflbb 20_h:mm 4<2_u3».0204 \IJfi ab~_. IIIIIIIIIIIyomWhHHHHHHHIIIIIIIIILma_ J UV. Pam. roam \ NDNL zo.kuwm 4(2.0:».0204 vbtb rn_bcut 4....2 gs. ‘ . J flask WI 1.0m. 90m. \\\Moe_ so a Fonooaa .\\l6mLHHH///// Tl, 024 «.4 For 111 .00. wanhdmwazm» m_(»w4»30 n.0NN wcahdmwaimh. 1.4 hm..2_ xm>mu morghuww mmuuuq 25 Fron the,study of the temperature patterns along the longitudinal sections of the dryer on Fig. 6 it can be ob- served that there was a temperature drOp around the main dry- ing zone. Further, it is observed that there were disconti- nuities in temperature patterns. This might be due to the presence of turbulence in the moving air inside the dryer. Temperature patterns for cross-sections at 12, 15, 1k and 15 feet of the dryer are presented in Fig. 7. They correspond to the inlet air temperature of 298° F and ex- haust air temperature of 200° F. Note that the right side of the dryer section had higher temperature patterns than the left side. .The reason for this concentration of tem- perature gradients was that the right end spray atomizer was not used; as a result, hot air entering the right side of the dryer section did not lose as much heat as the left side of the dryer. However, it would be observed that the temperature difference between the left and the right sides of the dryer gradually decreased as air moved further toward the rear end. This pattern could be well visualized in Fig. 7 at the 15 feet distance from the front. Fig. 8 shows the longitudinal temperature patterns for the last four feet of the dryer.‘ An analysis of these patterns revealed a low tem- perature zone surrounded by a high temperature zone. This could be explained in recognition of the fact that most of the drying process took place at the central core of the dryer and air in this region became colder compared to the \\ \\4 '20 , [I / 1' 2/15'6 22017 [23 2'? .' { /// J 222.4 ./ f“. H J INLETAIRSO‘F ‘ f INLET/Al/AZ 4‘s OUTLETAnazoo OUTLE/AVN2IO'F I2' FROM FRONT END I3' FRON FRONT END 208 2007 204°F208°F ZIZ'F I INLRM zso‘F INR AIR soo’r OUTLET AIRISS - OUTLE T IRZO |5° FROM FRONT END l4'FROH FRONT END Fig. 7. Cross-sectional temperature patterns inlet air at 500° F ‘ I322). // 4/ \/ 7.4 LONGITUDINAL SECTION L’ll-I ‘— L-ZI:2,LOVGITUDINAL SECTION L LONGITUDINAL SECTIONL-3L'3 ( I f/ 2:29,; LON GITUDINAL SECTION L‘TL—T ‘ W 29\ 22A'F/{2\0'F L ONGITUDINAL SECTION b6 I:6 zE/ . (4/ 220 I J1 LONGITU DINAL SECTION [:3 U3 .———_~ ————-+ »————+ HOT AIR AND PRODUCT FLOW INLET AIR TEHPERATURE 298 F OUTLET AIR TEMPERATURE 20°F LONGITUDINAI SECTION t4L'4 Fig. 8. Longitudinal temperature patterns, inlet air at 500° F 25 26 a ommm as has pods“ 0933.3 092390980» Hdnuunuamcoq .m .wam 0.40... zo_»ouw sz.o:k_ozo.. TOMB/f? 4 Nb NH. 20.». owm J¢o no 20:.Uum mmomo < 27 a ommm as he and: 25033 asuma0980u Heapspdwnoq .OH .wam 0.0 202.0um J¢h80N_¢Oz I'M. IOTPuua 4482—3450204 h_004w>m_( m t n. N . O . N n QM O #20"...— OZWhZOEB‘OCK 1N H a 3 .0: 3:3... m m? H >k_00.._w>44_x‘ 1| 1 #0 m J O.N_.mv_u 3 RR .0 3 . I. .‘Iulllll .71 (III! 10— use .3 33093 333.5 .mH .ummxh wwuz.>toooqw> 1.4 n t N . _ N n men. ‘74 1 ozw 9.20m“. 10:“. .0. ha. wiuomn. » »_OOJU> ...<_x< \ \. .\ I \.nE, ‘IIIIE I hzomm ll' .oumxhumu z. rkaoqu) ¢_( o e n ~ . e _ u n e n pzocu ozupzocazoma .0 2. 3:0,... :53“; .232 méel o o. 1333 NI .LH9I3H 133i NI LHOIIH uwm\»uuaz.>tood>¢i o n ~ . o _ N n o o d 1 J J 1 i 11 1 A“ J_ pzomu _ ozuhzomazoma i .o: 3.4.2... >tooeu>45xq . w\..\ o-m_ o: .uwm\buum z. >h.004w> 1.4 :2. I: ‘ 02w block 10¢... .N #4 uiuoca r2004m> J<_x< ll ' Dbl , Ill bloc... (.N. 4:... o. 133$ NI LHOI3H $1 upper half of the dryer, the air flow direction was for- ward whereas at the lower half of the dryer the air flow was in the reverse direction. There was vortex pattern formed by the air inside the chamber. Air mass formed a vortex where the air rotates as a rigid body before it gees out of the chamber. The zone of separation has been drawn in Fig. l2-F; it was the region of very low velocity and at some place between there was no air velocity at all. Although the air flow patterns so presented may be slightly different when the milk is sprayed, the general trend is expected to re- main the same with more turbulent effects. The separation of the product in the drying chamber should be better with the reversal of air flow direction, because the dried milk particles have more time to settle to the bottom of the dryer by the force of gravity. The exhaust fan under the operating condition has the discharge rate of h,h00 c.f.m., and the average cross-section of the dryer is 57.1 square feet. If there was unidirection- al flow of air in the dryer, the theoretical average velocity would be 2 feet per second. But in actual condition the upper cross-section of the dryer has the forward flow, where- as the air flow in the lower cross-section is in the opposite direction. The measured average air velocity for the upper half of the dryer was found to be 5.71 feet per second and that of the lower half was 5.62 feet per second in the oppo- site direction. The cross-sectional area occupied by the iforward air flow was approximately 22.1 square feet and the area occupied by the reverse flow at the lower half was approximately 15 square feet. The ratio of distribution of area was approximately 1%:1. The forward flow of air had an average velocity of 1.86 times that of the calculated value. CONCLUSIONS 1. Ninety percent of the drying process is accomplished within the first five feet Of the dryer chamber. The effec- tive utilization of the dryer chamber was limited to the first ten feet of its length. Possibly a dryer with a length of twelve feet would be as good as the existing one of seventeen feet in-iength if operated under similar conditions. 2. An extreme temperature drop occurred in the vicin- ity of the atomizer. The temperature drop was more predOmi- nant in the zone of separation, where air velocity was very low. The above statement is supported by comparing the tem- perature and the air velocity patterns with respect to their positions of formation. 5. Air velocity patterns under actual operating condi- tions could not be found as the experimental problem was ex- ceedingly difficult because of formation of dried product on the pitot tube. However, faulty and unsound air flow design would cause poor spray dryer performance as Observed at points of air stagnation. h. Analytic investigation Of temperature and velocity patterns showed that no two sections Of the dryer have the same temperature and air velocity. Moreover, the temperature pattern at any cross-section, when using symmetrical nozzle patterns, is not symmetrical about the central vertical plane. 35 3h- The nOn-symmetrical trend in temperature patterns is possibly due to the uneven distribution of air flow, eddy current ef- fect, and a higher localized rate of evaporation of spray drops due to variability of position of spray nozzles. SUGGESTIONS FOR FURTHER STUDY Through the experience gained from.this work, further studies are needed at other inlet air temperatures to re- establish the general temperature patterns inside the dryer to improve utilization Of dryer space. Further study should be devoted to air flow patterns under actual operating conditions. The problem Of air flow measurement would be simple if the upper half Of the dryer chamber could be made of some transparent material like plas- tic. This would give better visibility Of the inside of the chamber. The effect of inlet air humidity at different air veloc- ities on temperature patterns in the dryer should be deter- mined. Further effort should be encouraged on the investiga- tion of efficiency of spray drying of milk considering all factors involved in the operation. 55 2. 3. 9. 10. REFERENCES Coulter, S. T. (19h7). Evaporation of water from milk by spray drying. J. Dairy Sci., 50: 995. Duffie, J. A. and W. R. Marshall. (1955). Factors in- fluencing the pro erties of spray dried materials. Chem. Eng. Pros. 9: hl7-h180 Friedman, S. J., F. A. Gluckert, and W. R. Marshall, Jr. fié952é. Centrifugal disk atomization. Chem. Eng. Prog. :11. Giffen, E. and A. Murastew. (1955). The Atomization 22 Liquid Fuels. John Wiley and Sons, Inc., N.Y. Kitzes, A. S. (19h7). Factors influencing the design and Operation of a spray dryer. Ph.D. thesis, university of Minnesota, St. Paul. Marshall, W. R., Jr. (195k). Atomization and spray dry- ing. Chem. Eng. Prog. Man. Series NO. 2, Vol. 50. Rant, W. E. and W. R. Marshall Jr. (1952). Evaporation from drops. Chem. Eng. Prog. hé: lhl, 175. Seltzer, E. and J. T. Steelemeyer. Advances lg Food Research. Vol. 11, pp. 599-520. Academic Press Inc., Tate, R. W., and W. R. Marshall, Jr. (1955). Atomization by centrifugal pressure nozzles. Chem. Eng. Prog. L9: 167, 226. The American Dry Milk Institute. (1955). l95h census of dry milk distribution and production trends. American ‘Dry Milk Institute, Chicago. Bul. 1000, p. h. 56 APPENDIX An estimate of the air and the heat requirements for the dryer. Concentrated milk, initial temperature, 155° F. Solid contents, ho percent on wet basis. Final moisture content of dry milk, h percent. Feed rate, 15 gallons per hr 2 15 x 9.5 lb/gal = 125.5 lb/hr Product flow rate in terms Of solids : 125.5 x 0.h = h9.h lb/hr Initial moisture content, X1 - - 60 - X1 - IUggEU - 357- 1.5 lb water/lb of dry matter Final moisture content Xe in pounds X2 : 16%:E : 0.0hl7 1b water/1b dry matter Pounds of dry solids per hour = h9.h(1-0.0h17) 3 h7.h lb Rate of drying Pounds of dry solids (XI-x2) u7.u(1.5-o.ou17) 69.2 lb water evaporated/hr At 75° F, 50 percent humidity, the absolute humidity of the available air is 0.0095 1b water per lb dry air and this is equal to Y2. Atmospheric air temperature, t8 = 75° F Inlet air temperature, t1 = 220° F Outlet air temperature, t2 3 1900 F 57 Assuming the dryer to be circular in cross-section, the dia- meter of the dryer 7.0 feet and 17 feet long Total surface area =1TDL+2 E (D)2 "(7.0)(17)+g (7.0)2 57h+ 77-0 h51.0 sq ft Taking meannat between dryer and surroundings as (ti - tugfia - ta) ; igZO-ZS)+(;90-75) 7? Ast 2 i-ll = 130° F The estimated combined natural convection and radiation heat transfer coefficient from the dryer to the surroundings is 2 Btu/hr fta °F. (McAdams, w. H., £323 Transmission, 5rd ed., McGraw-Hill Book 00., Inc., New York, l95h.) Estimated heat loss h A A t 2(h51)(150) 117200 Btu/hr (1 Relative enthalpy, Btu for mixture, H1 = (0.2L, +0.Li5 Y1) (t - 52)+1,075.2 Y1 For inlet air, enthalpy, 302 (air at 220° F) [0.2u+0.LLS (0.0095) (220-52)] 4- 1,075.2 (0.0095) (.2hh28) (188)1-10.22 u5.8+1o.22 56.02 Btu/1b dry air O 58 For exit air at 190° F, enthalpy HGl (0.2u+0.u5 Y1) (l90-52)+1,075.2 Y1 (0.2hrfoeh5 Y1) (l58)+-l,075.2 Y1 57.94-11h6.2'Y1 ' Heat capacity of dry milk is 08 3 0.5 and that of water, 0" = 1.0 Btu/lb 0F. Taking to = 520 F, so that enthalpies of gas and solids are consistant, and since t = 155° F and t82 3 190° F 3 The solid enthalpies, Btt/lb dry solids a. = c. (t. Ref 0.5 (155-52H(o.ou17)(1)(155—52) = 0.5 (105)4-0.0u17(105) = 51.5+u.28 3 55.78 Btu/lb H82: 0-5(190-52)4-0-0h17 (190-52) = 0.5 (158)+-o.0u17 (158) - to)+-x cw (ts - to) = 85.56 Btu/1b Moisture balance: LSX1-+ GsYZ = LSXZ + GsYl Ls ixl-xz) = 03(Y1-Y2) #7.} (1.5-0.0017) = 0s (Y1-0.0095) 68 1-03 (Y1-0.0095) Equation 1. Enthalpy balance: ‘L8H81-+ GSHGZ = LSHSZ + GSHG1+ in 5(55.78)+c-3(56 02)= 1.7 5(85 Q56)+Gs(57-9+111+6-2Y1)+Q 59 26hOi’56.02 G3 3 hOSOi—Gs(§7.94'1186.2Y1)+'117200 18.12 GB ' 11h6.2 GsYl + 118610 Equation 2. Solving equations 1 and 2 simultaneously, 18.12 G3 = 11h6.2 (68 +-0.0095 Gs) + 118610 7.22 Gs 11u6.2 x 68 + 118610 G8 = 1 6 10 = 27250 1b of dry air/hr The enthalpy of the fresh air, H85 = 21.5 Btu/1b dry air. Heat load on heater = Gs (H,2 - H53) 27250 (56.02 - 21.5) 27250 x 5h.52 9h0,000 Btu/hr