Document Qg4qNkYyE9rNadx9Mxy4VMrnv

Formation of the elementary asbestos-cement layer during the process of filtration based on work by Dr. I. I. Berney -- part IV In many processes of manufacture, filtration is only a means of dewatering the suspension and the solids retained by the filter are usually removed by scrap ping or some other method, forming a shapeless mass; the interest is limited to the moisture content of the obtained product. Asbestos-cement manufacture differs in as much as the end product is formed from a number of indi vidual layers and their compactness depends not only on the moisture content but also on the structure of the delivered layer and the distribution in it of cement and fibres. The asbestos-fibre/cement ratio in the layer depends on the properties of the asbestos-cement suspension and on the method of filtration and this relation will be examined in detail later; in the meantime the results of experimental work referred in table 4 should be examined as they are indicative of the influence of the average rate of filtration on the weight of the layer. These experiments were carried out on the special filter described in No. 7/8 - 1968 AC/M, using the same suspension but pressures of 30 and 240 g'cm-. Jt can be noted that when the rate filtration of is higher than 1 cm/sec,, the delivered layers, with angles of inclination of the filter of 90" and 135, have the same weight and an almost equivalent mean specific resistance; as the rate of filtration decreases, how ever, the weight ratio between layers obtained with angles of 135" and 60 respectively, drops while mean specific resistance of the layer obtained with angle of inclination of the filter of 135" increases, in comparison to that of the layer obtained at an angle of 60. The results of these experiments are particularly si gnificant as they reveal an important fact i.c. the in crease of the rate of filtration reduces considerably SEPTFMBER OCTOBER I9B* the structural instability of the layer and makes possible a better collection of the particles of the solid phase, during the process of filtration of asbestos cement suspension. Examination of the data referred in tables 2. 3 and on graph fig. 15 (see AC/M Nos.' 7/8-1968) and table 4 enables to determine the causes for which the structure of the layer is influenced by the method of'filtration. The influence on the process offiltration of the relation: rate offiltrationjratc of sedimentation of the solidpar ticles of the suspension. Let us imagine that fig. 16 is a non rotating sieve cylinder immersed in the suspension for 3/4 of its diameter and connected to a vacuum pump; that agitation of the suspension is intense enough to prevent sedimentation of the solid phase in the bottom of the vat and to maintain uniformity of the concentration in all parts but not intense enough to wash off the sediment deposited on the surface of the sieve cylinder; that .filtration is taking place at constant pressure for a determined time. An imaginary line passing through the centre of the transversal section of the sieve cylinder, will divide the suspension into two parts, upper and lower (zones I & 2). In zone 2, the suspension is kept in agitation by means of special agitators. The suspension thu> agitated, halier', against the vvnlN and bottom of the sat and against the surface of the sieve cylinder moving disorderly in different directions. When considering the ((washing olT of a solid body by a turbulent (low. the existence of a boundary layer having a laminar flow, is admitted. Fluid velocity through this layer decreases as it comes in proximity CTD001136 |15 X Table 4. The influence of filtration of asbestos-cement suspension, on the specific resistance and weight of the layer. Filtration at pressure of 240 g/cm* Filtration at pressure of 30 g/cm' Time of filtration t/sec. 1,2 2,15 3,11 4,08 5,i2 1,05 2,07 3,03 4,03 5,01 Average rate of fil tration cm/sec. 1,02 0,78 0,62 0,50 0,43 0,58 0,40 0,32 0,26 0,24 Average specific res istance of layer cm/g. (V = 60) 373 760 1110 1090 195 260 312 330 Weight ratio of layer obtained at angles of 135? and 60 ' Average specific res istance of layer cm/g. (5 = 135) 1,01 0,92 0,86 ' 0,82 0,79 390 850 1320 1350 0,73 0,67 0,65 0,63 0,62 280 394 490 534 N.B. Composition of the solid phase of the suspension: Chrysotile asbestos grades 5 & 6. S0' each: cement type 400. fineness of grinding: residue on sieve 4900 - 7,J %. Solid phase composed by 13.5 % asbestos and 86,5 % cement (by ^eight). Concentration: 0,08 g/cni*. to the solid body until it is equal to zero on its surface. If the surface of the moving in the liquid, solid body is a filtrating surface (as is the surface of a sieve cylinder) then the velocity of the fluid through the boundary layer, will be the geometric sum of the two velocities: flow velocity washing off the surface of the sieve cylinder and rate of filtration. On the surface of the cylinder, when fluid velocity is equal to zero, the rate of movement of the liquid -- velocity and direction -- is equal to the rate of filtration. It can consequently be stated that the a/m concepts will be correct both with rotating and non-rotating sieve cylinders. I.el us now examine how the process of filtration is influenced by the movement of the particles of the solid phase of the suspension under the effect ,of gravity. At the moment a particle reaches the filtrating sur face, the speed of its movement will be equal to the rate of sedimentation (IJoc) under the effect of gravity plus rate of filtration (UVq. Speed of sedimentation can be figured as a function of the SfPlTMflER.OCrOBFR T96fl diameter of the particle. Uoc = f (dm) (2-22) Particles of irregular form are subject to the same rules provided the size of the irregular particle is equal to the size of an equivalent spherical particle. It is only through the very thin first layer deposited on the surface of the sieve cylinder that flow of the liquid and filtration are equal regarding rate and direction. The motion of particles of the solid phase of the suspension will be in some way similar to the motion of particles in a vertical stream. Its course can be imagined in the following way: As a consequence of the intense agitation of the suspension, particles of different sizes arc pu-hed by the blades of the agitators towards the Mirluee of the filter, but not all of those particles remain on this surface. The filter retains only those panicle the rate of sedimentation of which is lower than the rate of filtration; the larger particles, sedimenting at a rate higher than that of filtration, will not be retained and will again pass in the suspension. A CTD001137 17 r* jv V Fig. 16. Diagram of a sieve cylinder. 3/4 or its diameter immersed in a suspension, showing direction and rate of filtration and direction and rate of sedimentation of the solid particles. similar process can be observed when the sieve cylinder rotates. Rates of sedimentation have been plotted against percent corresponding weights of particles contained in the suspension the rate of sedimentation of which is equal or less than Uoc, to construct graph figure 17. Referring to zone 2, fig. 16 (lower zone) sedimentation and filtration are directed in inverse sense. When assuming that at a given movement the rate of fil tration is equal to U'd, then on the surface of the layer will be retained only part of the particles of the solid phase, those for which Uoc < U'4 (2-23) The quantity of such particles in the suspension 1181 (see fig. 17) constitutes n,%. Consequently when determining the volume unit of the filtrated materiai. c >> which expresses the weight of the layer deli vered on the filter, this will be less than a (weigh; of concentration) and will be n, % of a the reason being that the layer will not contain the panicles whose rate of sedimentation is higher than the rate of filtration. At the moment the rale of filtration will be equal to IJ`<^ the largest particles delivered on the layer will have a speed of sedimentation Uoc -- LA'. As the process of filtration progresses at constant pressure its rate gradually decreases and conse quently the quantity of maximum and minimum size particles, passing from the suspension to the layer, will also decrease. CTD001138 1H AC, 'J 3 ut a * ;'t corresponding percent weight of particles, the rate ol sedimentation of which is equal or less than lloc. Fig. 17. - Graph of distribution of the solid phase of a suspension. e /o When UY will be less than Uoc max. the weight of the layer delivered per volume unit of filtrated material V and average diameter of particles, will have changing values. Movement of particles at point 2 (fig. 16) has been examined. For the points of the surface of the cylinder in the lower zone (where the directions of filtration and sedimentation constitute the angle y - see fig. 16), the inequality (2-23) must be expressed in the following form: Uoc-cos (ISC0--y) < U-^ or U'd -- Uoc . cos . (ISO0 -- y) < 0 (2-24) During the process of filtration, the following cases arc possible in the lower zone. (fig. 16). 1. The rate of filtration on all points of the lower zone, is higher than the projection, at maximum |20| speed against the filtration flow, of the sedi menting particles of the solid phase of the su spension. This would mean that the rate of filtration is so high that inequality (2-24) would be correct for all the particles of the suspension, including the largest. In this case, the laser would be formed from particles with the same unchanged correla tion, as this exists in the suspension. Panicle size distribution, through the whole thickness of the layer, would be the same and c\ery \olume -- unit of filtrated suspension w ould deii\ er on the filter in the course of the entire cycle. a seement of equal weigh! in proportion lo the concern an of the suspension. The value c " wouk. be constant during the whole cycle and the average specific resistance of the layer will also have a constant value. CTD001139 THE AC/.'.' Filtration with such characteristics delivers a homogeneous layer and equation (2-18) which is used for designing vacuum filters, applies to such filtration. 2. The rate of filtration on all points of the lower zone, is lower than the projection at maximum speed against the filtration flow, of the sedi menting particles of the solid phase of the sus pension. In this case, particle size distribution would be dif ferent in the different sections of the layer. As we come nearer to point 2 -- thus increasing cos. (180 -- y), the inequality (2-24) will be correct for the particles with smaller value of Uoc. This means that the size of the particles forming the layer, will continuously decrease. As filtration progresses, and the rate of filtration decreases, the average diameter of the particles delivered in the layer, with constant cos. will decrease (the change of the value of cos. 180 -- is only one of the causes .for the delivery of a homogeneous layer). Particle size distribution in the layer will not correspond to particle size distribution in the suspension and the quantity or small diameter particles, delivered in the layer, will be progressively increasing causing, as a consequence, the increase of the average specific resistance because the size of the pores in the layer will be smaller not only on account of pressure but mamly on account of the decrease of the average diameter of the particles which form the layer. Accordingly, the weight c of the layer delivered on the filter in the lower zone, will also be decreasing because as the rate of filtration progressively decreases, the filter will retain only those particles for which the inequality (2-24) is correct. Thus, diameter of particles, weight of layer, average specific resistance and other factors which characterize the process in the lower zone, depend not only on the rate of filtration but also on the point of the circumference of the cylinder on which the picking up takes place. Under these conditions the delivered layer will have unequal particle size distribution in its different sections and is a heterogeneous laser. In the course of experimental work carried out, the obtained data referred in tables 2 and 4 (increase of the average specific resistance during the process of filtration, the influence of the angle of inclination y and of the rate of filtration on SEPTEMBER -OCTOBER 1968 the specific resistance and weight of the layer, the decrease of c during the process) indicate that filtration of asbestos-cement suspension is a process of filtration with formation of a hete rogeneous layer. 3. The rate of filtration on all points of the lower zone is higher (U'<4 max.) at the beginning and lower (U'^ min.) at the end of the cycle, than the projection at maximum speed against the filtration flow of the sedimenting panicles of the solid phase of the suspension. At the be ginning of the cycle, a homogeneous layer will be delivered (as in case 1) until the rate of fil-. tration will be corresponding to the equation U'^ = Uoc max. cos. (180 -- -y). After this moment and until the end of the cycle, the delivered layer will be heterogeneous (case 2). 4. The rate of filtration, is higher at the begin ning and lower at the end of the cycle, than the projection, at minimum speed of the sediment ing particles of the solid phase of the suspension, against the filtration flow. Under these condi tions, filtration takes place only up to the mo ment when the rate of filtration and speed of projection against the filtration flow of the seimenting smallest particles, become equal. At this moment the rate of filtration will be insufficient to retain even the smallest particles of the sus pension, on the layer. The filtrated material . will pass through the layer with a constant mini mum speed (U*rf min.) but the particles will not remain in the layer. 5. The maximum rate of filtration, at the begin ning of the cycle, is lower than the projection at minimum speed of the sedimenting particles of the solid phase of the suspension. Under these conditions, no particles will be retained by the surface of the filter. In the light of the indicated principles, the processes of filtration can be classified as follows: a) Processes with rate of filtration higher than the speed of sedimentation of the largest particles of the solid phase of the suspension; the deli vered layer is homogeneous. b) Processes with rate of filtration lower than the speed of sedimentation of the largest particles of the solid phase of the suspension; the de livered layer is heterogeneous. CTD001140 i21'. Sufficient data Is available on classification a. Equa tions (2-14 and 2-18) are specifically related to this classification. However, these equations cannot be applied for calculations referring to classification b to which the filtration of asbestos-cement suspension belongs. The investigation of the influence on the process of filtration of the relation rate-of-filtration-fiow/speed of sedimentation of the particles of the solid phase of the suspension, allows to figure more clearly the process of formation and structure of the elementary asbestos-cement layer. The data referred in tables 2 and 4, obtained from experimental investigation, indicate that the increase of the average specific resistance of the layer, during the process of filtration, takes places in all cases and does not depend on the position of the filtrating screen or on the rate of filtration. Considering the presence in the suspension of asbes tos-cement particles and of free grains of cement, so widely differing in size, experiments have been carried out to explain the phenomenon: It has been established that the asbestos-cement particles, delivered on the surface of the wire during the time period 0,9-1.2 seconds from the beginning of the process, have such large pores that the free cement grains pass through them together with the water. The quantity of solid particles carried away during the first 2 seconds, attains 14.6 % of the weight of the layer (see table 2). After 0,9-1,2 seconds (time can differ with suspen sions differing in composition), no solid particles are observed in the filtrated material although filtra tion and delivery of panicles on the filter still goes on. During this period, the layer will retain all the parti cles, which until this moment were carried away by the filtrated material. Consequently after the end of the transition period, delivery of the smaller particles increases and because of this, specific resistance .also increases. It is also important to establish in which part or the layer and how evenly, free cement is deposited. The depth of penetration of free cement depends on the compactness of the layer. The outer pari of the layer being still not sufficiently compact and in the stage of formation, filtration will take place under conditions similar to those which are observed at the beginning of the process. The free cement particles are not retained in this part of the layer but penetrate more deeply and deposit near the surface of the most compact part of the |22| layer whicl^ is in contact with the surface of the filter. However, as the process of filtration progresses, deposition of cement gradually approaches the ouier part of the layer. It results that the layer is not uni formly saturated with cement (see table 3). Saturation of the layer with cement increases as theduration of the process increases but near the surface of the filter, an intermediate layer with insufficient cement content, will still be formed. The saturation with cement of this intermediate layer is impossibu due to its great compactness. The degree of compactness of this intermediate Iay e depends on hydrostatic pressure (compactness creases with the increase of hydrostatic pres- . When filtration takes places with angle of incltna:. < 90 particle size distribution in the layer w be even less uniform. When the rate of filtration increases the solid phase in the layer will have a composition almost identical to that of the suspension. Under these conditions, layers obtained at angles i = 60 and y = 135 "ill differ less both in weight and specific resist e. The correctness of these conclusions has been con firmed by the experimental work the results of which are referred in table 4. The rate of filtration can be increased in different ways. In the course of the above referred experiments, rate of filtration was increased by increasing the pressure but this method should notie recommended as it leads to excessive compactness of the layers in contact with the solid surface of the filter and pre vents their saturation with cement. A recommended method to increase the rate of filtration is by making the cycle shorter. It results from the above thar uniform distribution of cement in an asbestos-ce ment layer depends on the conditions of filtration. Be:-: uniformity of distribution of cement can be obtained by filtrating asbestos-cement suspensions under the following conditions: 1) with high rates of filtration; 2) with short cycle. 3) w'ilh low hydrostatic pressure. Uniformity of distribution of cement in the layer or otherwise degree of homogeneity appears to be ;t basic factor influencing the quality of a>be "cement. In order to establish the effects on quality, a series of experiments lias been carried out and the re.-clis will be examined in another paper. CTD001141 7Hc AC M