mz 1. HENRY RUSHTON,
MIXING ILLINOIS
INSTITUTE OF TECHNOLOGY, CHICAGO, ILL.
-
URING the past year articles have appeared in the technical journals having t o do with process applications for different kinds of mixing devices, as well as new data and compilations on performance of mixers in liquid agitation systems. A marked increase in interest in the subject of mixing has been evidenced by a number of talks on the technology and principles of mixing given before local sections of the AMERICAN CHEMICAL SOCIETY and other technical and industrial groups. All publications this year emphasized the quantitative aspects of mixing technology; this is most encouraging and augurs well for the development of a better understanding of this important unit operation of chemical engineering. During the year a considerable number of industrial research laboratories have installed laboratory size and pilot plant size equipment for study of mixing variables in relation to specific processes of industrial importance. Such experimentation by industry in cooperation with equipment manufacturers and university laboratories is proving to be valuable to the chemical process industries. Several important large scale processes, involving pilot plant work for a thorough study of the effect of mixing, have recently been placed in operation. Process Applications. A continuous mixer has been described by Bridger, Wilson, and Burt for the manufacture of a concentrated superphosphate (1). This mixer is a funnelshaped device. The cone of the funnel has an angle of 30' and is approximately 27 inches in height, tapering from 18 inches at the top t o 3 inches a t the bottom. There are no moving mechanical parts in the mixer, and mixing is effected by distribution of four streams of acid in such a way as to cause a swirling motion. Crushed phosphate rock is fed to the central top part of the cone just below the acid nozzles. Because of the tangential introduction of acid a t a fairly high velocity, a vortex is created, and the phosphate rock fed to this vortex becomes mixed with the acid while being thrown t o the sides of the container by centrifugal force and the downward motion through the cone. This mixer has operated successfully a t a production rate of 30 t o 35 tons of superphosphate per hour. Holdup time in the mixer is about 2 seconds. This indicates that there is rapid intermingling of the crushed rock and acid. After leaving the continuous mixing device the mixture of rock and acid is spread on a moving belt, and within a short length of traveling and time the chemical reaction is completed. A good grade of relatively dry superphosphate forms and is discharged from the moving belt. This development depends upon the integration of continuous and automatic feed and control devices. The energy required for effecting rapid distribution of rock and acid is obtained through motion induced by liquid flow due t o pressure head rather then by a mixing impeller. It is claimed that excellent mixing together with low operating and maintenance costs can be achieved, and that results are superior to any previously obtained by the use of batch mechanical mixers. The mixer size is small considering the capacity of the unit. An article by Chaddock (2) is concerned primarily with the proper and practical applications of various types of agitating devices to different processes. Considerable practical information is given based on the experience of the author in applying present-day mixing equipment to various chemical processes. Chaddock classifies agitator impellers into four principal types: propellers, paddles, turbines, and anchors. He states that fully 95Tc of all mixing problems can be served by one of these stand49
ard type impellers. He discusses the method by which power input per unit volume is used to specify the degree of agitation, and presents curves for the approximation of degree of agitation as a function of power input. Data are presented in the forms of curves and nomographs whereby impeller sizes and speeds can be selected for any desired amount of power required, and the degree of agitation selected for a given process. These data are not new but are presented here in relation to selection of agitator type and size for various process requirements. Chaddock also calls attention to the fact that the vessel design is of considerable importance, and that it must be considered with relation t o the type and size impeller. He recommends from his experience that impellers be operated approximately 4 to 8 inches above the bottom of the tank, and the illustrations picturing positioning and types of impellers are so located. In general such low positioned impellers (especially those 8 inches and more in diameter) are not normal practice; more installations are made between "4 to 1 impeller diameter off bottom, provided due consideration is given to the velocity of flow necessary t o sweep the tank bottom. It is further recommended in Chaddock's paper that best agitation in deep vessels may sometimes by obtained by the use of two impellers of opposite pitch placed so that the lower impeller discharges downward and the upper impeller discharges upward. Thm arrangement is generally considered not to be best practice. Common practice is to arrange multiple propellers, each discharging downward. Impeller Performance Data. Four articles have appeared during the year on the power requirements for rotating impellers. The first of these by Martin (6) was a review and compilation of existing data and summaries. He compared data on the power consumption of various impellers on the basis proposed by the earlier work of White and of Hixon. Martin concludes that White's equation-namely, that a dimensionless power factor is a function of Reynolds number-can be used to correlate and predict impeller performance, and that constants in the relation are unique for each geometric shape of agitator and tank. This conclusion is thoroughly in accord with all published data and serves a t this time to emphasize that generality which is all too little understood. An article by Olney and Carlson (6) presents a number of new data on the power absorption in mixers using arrowhead turbine type dispersers, spiral turbines with stators, and flat paddles. They also summarize data on some 16 other impeller-tank combinations previously reported in the literature. I n this work data were obtained for performance in eight different liquids over approximately a threefold range of density and a hundredfold range of viscosity. The data are presented in the form of speed-power curves and also as log-log plots of power function vs. Reynolds number. I n some cases immiscible liquid combinations were used, and these two-phase power studies were correlated with single phase ones. The work was performed under conditions approximating those in commercial operation. Baffles were used, or in the case of the spiral turbine a stator ring was used, replacing baffles at the wall. The results correlate well with the few data in the literature for baffled and commercial operation conditions. The authors were able to correlate the effect of vessel diameter, impeller blade width, and depth of liquid with the power requirements for flat paddles. Finally, they present a nomograph for
INDUSTRIAL AND E N G I N E E R I N G CHEMISTRY
50
estimating net power requirements of mixers. The data for the nomograph were obtained from their own and from those previously noted in the literature. Two articles by Mack and Uhl (8, 4 ) have appeared pertaining to performance of agitators. The fist deals with agitator power requirements for gas-liquid contacting equipment, and the second discusses mass and hea,t transfer in agitator systems. Both articles are intended to show how impellers can be selected and operated for a given process application based on existing data in the literature, or from data from plant or pilot plant operation. There are a few new data in these articles, but the articles are of value primarily in calling attention to existing information already in the literature which may have been overlooked by newcomers t o the field. I n the fist article Mack and Uhl present again data on gasliquid contacting by mixers and extend the data for oxygen, carbon dioxide, chlorine, and hydrogen. They also give data for the power characteristics of wire impellers. The wire impellers are simply described as having been made with piano wire, and the configuration and size are not mentioned. An analysis of power function os. Reynolds number is made along the conventional lines established by White and others, and it is emphasized that power absorbed by impellers operating in baffled tanks may be as many as five times greater than power consumption in unbaffled tanks. Several hypothetical illustrations are given to show how to scale up from pilot plant data t o large size in order to illustrate the method of handling the power relations discussed in the article. These examples are exercises in the use of the dimensionless equations correlating impeller performance and are of use for this reason. Illustrative examples on gas-liquid contacting concerning the absorption of carbon dioxide in a slurry of metallic hydroside are also given. This is an extension of previous work of the author with others. Several other hypothetical examples pxtaining to gas-liquid contacting are also worked out.
Vol. 40, No. 1
The second article dealing with mass and heat transfer makes use of dimensionless correlations of Hixon and Baum and of the heat transfer data of Chilton, Drew, and Jebens. Several examples for hypothetical cases are worked out t o illustrate the use of these data. A table for film coefficients of heat transfer for a large number of liquids is given. These coefficients are either those reported by Chilton, Drew, and Jebens or have been cornputed by the authors based on the dimensionless correlation of Chilton, Drew, and Jebens. There are no new experimental data t o warrant the assumptions and conclusions refcrred t o in the article. Several assumptions and inferences are made regarding heat transfer with relation to power input that are open to serious question, and it is not likely that the data of Chilton, Drew, and Jebens can be extended with any degree of accuracy beyond the condition specifically stated in their work. Mack and Uhl make extensions of the data to baffled conditions which were not covered by those reporting the data, and the assumption upon which the calculations are made cannot as yet be substantiated from experimental evidence. Review. A review article covering information on mixing during the preceding year was presented by Rushton ( 7 ) . This, . ~ N DENGINEERIKG together with previous reviews in INDUSTRIAL CHEMISTRY, includes a bibliography which will bring up to date all information in the technical literature pertaining to mixing. LITERATURE CITED
(I) Bridger, G. L., Wilson, R. A., and Burt, R. B., IXD. ENG.CHEDI., 39, 1265 (1947). (2) Chaddock, R. E., Chem. Eng., 53, 151 (Nov. 1946). (3) Mack, D. E., and Uhl, V. W., Ibid., 54, No. 9, 119 (1947). (4) Ibid., 54, No. 10, 115 (1947). ( 5 ) Martin, J. J., Trans. Am. Inst. Chem. Engrs., 42, 777-81 (1946). (6) Olney, R. B., and Carlson, G. J., Chem. Eng. Progress, 43,No. 9, 467 (1947). (7) Rushton, J. H., IND. ENG.CHEY.,39, 30, 43 (1947).
RECEIVED November 10, 1947.
SEDIMENTATION AND HYDRAULIC CLASSIFICATION mg
ANTHONY ANABLE, THE
DORR
C O M P A N Y , NEW YORK 22, N. Y.
EDIMENTATION and hydraulic classification are very old unit operations, the theory of which is well understood. Therefore, new developments of a basic character seldom occur in the short space of time of a year or two. On the other hand, improvement in the design of mechanical equipment is a continuous process which leads steadily towards increased efficiency and a general over-all bettering of operation. Sedimentation. The Trebler clarifier, which was mentioned in this journal's annual review of January 1946, has been rechristened the Dorr Duo-Clarifier (Figure l ) , and a dcscription of this unit in its prcsent form may be of interest. It is basically a modification of the standard Dorr clarifier and consists of such a unit divided diametrically into two compartments by a vertical dividing wall extending from above the liquid level down to the sludge level. One compartment is used for primary scttling and the other for secondary settling. Tlic fccd well and the overflow collection trough are bisected by the dividing wall with separate feed and overflow lines. Recirculation take-offs are also separated and function independently.
Figure 2 shows this unit used in connection with a Dorrco DuoFilter for treating a biological waste a t a milk plant. Flow goes fist to the primary filter and then, in sequence, to the primary clarifier, secondary filter, and final clarifier. Because of lower installation costs of a Duo-Clarifier compared with two standard clarifiers of equal combined area, plus lower maintenance costs and simpler operation, the unit is proving attractive for small communitics, institutions, and trade waste-treatment plants. From two milk waste-treatment plants the following data are available. Results based on daytime composites over a 16-day Deriod. Flow, Gal./Min. 31 21 Flow
Gal / i U i n . 31 21
Recirculation Ratio B.O.D., P.P.M. Second5% RePrimary ary Raw Final iiioved 1.26 3.52
2 23 4.52
533 648
95.8
20 33
95.
Suspended Solids. P.P.M. % ReRaw Final moved 330 31 90.3
Total Solidi, P.P.II1 % ReRaw Final moved 1182 712 40.
363
1211
30
90.1
461
61.5
