Caking tests

The specific resistances obtained are independent of applied load, suspension concentration and membrane type, as expected for non-compressible filter cakes. Tests with uniform latex particles have given permeabilities in very good agreement with Equation 2, using a value of 5 for the Carman-Kozeny constant. 

Performance of Drying-Pyrolysis Process Indirect steam drying operation was available to all kinds of dewatered cakes tested and its efficiency (overall heat transfer coefficient U) varied with the kinds of cakes. 

Public Broadcasting Service (PBS). The Layer Cake Test, Available online. URL http //www.pbs.org/wgbh/amex/bomb/ peopleevents/pandeAMEX60.html. 

Gilbert,RichardL.,Jr., andPaulW.Knapp. 1970. Development of an Accelerated Caking Test for Urea," Journal of Agriculture and Food Chemistry, 18(3) 397-400. 

On the last point it is appreciated that formulae exist which should enable one to extrapolate from a thin cake tested under laboratory conditions to a thicker cake expected on a full scale plant. In our opinion these formulae are virtually useless. 

In a 500 ml. bolt-head flask, provided with a mechanical stirrer, place 70 ml. of oleum (20 per cent. SO3) and heat it in an oil bath to 70°. By means of a separatory funnel, supported so that the stem is just above the surface of the acid, introduce 41 g. (34 ml.) of nitrobenzene slowly and at such a rate that the temperature of the well-stirred mixture does not rise above 100-105°. When all the nitrobenzene has been introduced, continue the heating at 110-115° for 30 minutes. Remove a test portion and add it to the excess of water. If the odour of nitrobenzene is still apparent, add a further 10 ml. of fuming sulphuric acid, and heat at 110-115° for 15 minutes the reaction mixture should then be free from nitrobenzene. Allow the mixture to cool and pour it with good mechanical stirring on to 200 g. of finely-crushed ice contained in a beaker. AU the nitrobenzenesulphonic acid passes into solution if a little sulphone is present, remove this by filtration. Stir the solution mechanically and add 70 g. of sodium chloride in small portions the sodium salt of m-nitro-benzenesulphonic acid separates as a pasty mass. Continue the stirring for about 30 minutes, allow to stand overnight, filter and press the cake well. The latter will retain sufficient acid to render unnecessary the addition of acid in the subsequent reduction with iron. Spread upon filter paper to dry partially. 

In a 500 ml. three-necked flask, fitted with a reflux condenser and mechanical stirrer, place 121 g. (126-5 ml.) of dimethylaniline, 45 g. of 40 per cent, formaldehyde solution and 0 -5 g. of sulphanilic acid. Heat the mixture under reflux with vigorous stirring for 8 hours. No visible change in the reaction mixture occurs. After 8 hours, remove a test portion of the pale yellow emulsion with a pipette or dropper and allow it to cool. The oil should solidify completely and upon boiling it should not smell appreciably of dimethylaniline if this is not the case, heat for a longer period. When the reaction is complete, steam distil (Fig. II, 41, i) the mixture until no more formaldehyde and dimethylaniline passes over only a few drops of dimethylaniline should distil. As soon as the distillate is free from dimethylaniline, pour the residue into excess of cold water when the base immediately solidifies. Decant the water and wash the crystalline solid thoroughly with water to remove the residual formaldehyde. Finally melt the solid under water and allow it to solidify. A hard yellowish-white crystalline cake of crude base, m,p. 80-90°, is obtained in almost quantitative yield. RecrystaUise from 250 ml. of alcohol the recovery of pure pp -tetramethyldiaminodiphenylmethane, m.p. 89-90°, is about 90 per cent. 

Materials suitable as filter aids include diatomaceous earth, expanded perilitic rock, asbestos, ceUulose, nonactivated carbon, ashes, ground chalk, or mixtures of those materials. The amount of body feed is subject to optimisa tion, and the criterion for the optimisa tion depends on the purpose of the filtration. Maximum yield of filtrate per unit mass of filter aid is probably most common but longest cycle, fastest flow, or maximum utilisation of cake space are other criteria that requite a different rate of body feed addition. The tests to be carried out for such optimisation normally use laboratory or pilot-scale filters, and must include variation of the filtration parameters such as pressure or cake thickness in the optimisation. 

Conventional filtration theory has been challenged a two-phase theory has been appHed to filtration and used to explain the deviations from paraboHc behavior in the initial stages of the filtration process (10). This new theory incorporates the medium as an integral part of the process and shows that the interaction of the cake particles with the medium controls filterabiHty. It defines a cake-septum permeabiHty which then appears in the slope of the conventional plots instead of the cake resistance. This theory, which merely represents a new way of interpreting test data rather than a new method of siting or scaling filters, is not yet accepted by the engineering community. 

The KDF filter was first tested in prototype on a coal mine in northern Germany. It was installed in parallel with existing vacuum filters and it produced filter cakes consistendy lower in moisture content by 5 to 7% than the vacuum filters. Two production models have been installed and operated on a coal mine in Belgium. The filter is controlled by a specially developed computer system this consists of two computers, one monitoring the function of the filter and all of the detection devices installed, and the other controlling the filtration process. The system allows optimization of the performance, automatic start-up or shut-down, and can be integrated into the control system of the whole coal washing plant. 

The test results reported show the advantages of pressure filtration quite clearly, ie, the dry cake production capacity obtained with the test soHds (coal suspensions) was raised 60 or 70% and the final moisture content of the cake reduced by as much as 5 to 7% by increasing the pressure drop from 60 to 200 kPa. Further increases in the operating pressure bring about less and less return in terms of capacity and moisture content. 

TDF Drum Filter. This is a fairly conventional dmm filter housed in a vertical pressure vessel. Test data, obtained with the smallest model of only 0.75 m filtration area, is available (18). Larger models have also been announced, ranging up to the filtration area of 46 m and very large vessels. The operating pressures are moderate, up to 25 or 35 kPa, and the dmm speeds fairly conventional from 0.3 to 1.5 rpm. The range of dry cake production quoted is from 250 to 650 kg/m h for fine coal. 

Results of test work with this filter, producing cakes of 1 mm thickness using a 3 mm clearance, have been pubUshed (33,34). The cake formed on the medium was generally stable, giving high filtration rates over long periods of time, and the precoat type cake did not blind with time. There was no evidence of any size selectivity of the process the only exception was conventional filter aids which were preferentially picked up by the rotating fluid. This... 

The scale-up of filtration centrifuges is usually done on an area basis, based on small-scale tests. Buchner funnel-type tests are not of much value here because the driving force for filtration is not only due to the static head but also due to the centrifugal forces on the Hquid in the cake. A test procedure has been described with a specially designed filter beaker to measure the intrinsic permeabiHty of the cake (7). The best test is, of course, with a small-scale model, using the actual suspension. Many manufacturers offer small laboratory models for such tests. The scale-up is most reHable if the basket diameter does not increase by a factor of more than 2.5 from the small scale. 

Vacuum filters are usually simulated with a Buchner funnel test or filter leaf test (54). The measured parameters are cake weight, cake moisture, and filtration rate. Retention aids are usually evaluated using the Britt jar test, also called the Dynamic Drainage Jar, which simulates the shear conditions found on the paper machine and predicts performance (55). 

Free-swelling tests are commonly used to measure a coal s caking characteristics. A sample of coal is packed in a cmcible or tube, without compaction, and heated at a fixed rate to about 800°C. Infusible coals distill without changing appearance or state of agglomeration. The fusible coals soften, fuse, and usually sweU. The profile of the resultant coke is compared to a series of reference profiles so that a swelling index can be assigned. The profiles represent indexes between 0 and 9. The best cokes come from coals having indexes between 4 and 9. 

For each independent variable, form the average value at which it was run in the complex. Draw a line from the coordinates of the worst cake through the average point—called the centroid—and continue on that line a distance that is twice that between these two points. This point will be the next test point. First decide if it is feasible. If so, bake the cake and discover if it leads to a cake that is better than the worst cake from the last set of cakes. If it is not feasible or it is not better, then return half the distance toward the average values from the last... 

Pretreatment Chemicals Even though the suspended solids concentration of the slurry to be tested may be correct, it is frequently necessary to modify the sluriy in order to provide an acceptable filtration rate, washing rate, or final cake moisture content. The most common treatment, and one which may provide improvement in all three of these categories, is the addition of flocculating agents, either inorganic chemicals or natural or synthetic polymers. The main task at this point is to determine which is the most effective chemical and the quantity of chemical which should be used. 

---