Document byz5m7edXDYoomkpQoXYZJ2po

> g) filtration of asbestos-cement suspension, the output of the sieve cylinders and their influence on the physical-mechanical properties of the end product based on work by Dr. I. 1. Berney -- part VII It is impossible to investigate the output of the sieve cy linders independently of its influence on the physicalmechanical properties of the manufactured product. Any calculation of the output of the sieve cylinders can only be accepted as correct, if ueder the operating conditions, the yielded product is of satisfactory quality. Otherwise, such calculations will not be representative and will ha-ve no practical use. The same applies to the influence of the conditions of filtration on output and quality of product. It is proposed to analyse the data obtained experi mentally, on the above criteria; this method will permit to draw correct conclusions and secure answers to the fundemental problems of the asbestos-cement manufacturing cycle among which are the improve- ment of quality and the increase of .production. The relation existing between the conditions of filtration of the asbestos-cement suspension and the physical mechanical properties of the manufactured end product, have been investigated both in the labo ratory and under actual manufacturing conditions, Experiments. A quantity of asbestos-cement suspension was in vestigated using the special filter (see AC/M No. 7-8, July-Aucust 1968). The characteristics of the sus pension were the following: ratio asbestos to cement: 14,4 : 85,6%; asbestos furnish: all chrysotile, 50% grade 5 and 50% grade 6; Fineness of grinding of cement: 3200 Blaine. Cement type: 400. The sample was made adopting the method described in AC/M No. 7-8, 196S; the sample was composed of a number of individual layers. As the properties of the manufactured product depend to a great extent on the pressure applied and to exclude any variations which could result from pressure condi tions, each layer was pressed at 3 kg'em- and the layers to each other at 20 kg/cm2; temperature of the suspension was also maintained constant at 20 C and only the conditions of filtration were allowed to change during the formation of the ele mentary asbestos-cement layer. The volume of the filtrated material was registered and the average rate of filtration was calculated. Average rate of filtration times concentration cf the suspension gives the delivery of asbestos-cement per unit of surface of the wire which in other w'ords represents the specific productivity of the sieve cylinder (see AC/M No. 1-2/1969). Relative values of the specific productivity, will in all cases, be expressed in percent of the specific productivity, under certain known conditions ad mitted to represent 100 percent. In order to compare the physical mechanical pro perties of samples obtained in the laboiatory with the properties of the product obtained on the factory machines, a special series of experiments were car ried out as follows: Asbestos-cement suspension was drawn in equal quantities fiotn the first and third vats of a ma chine in operation. MAkCM.APRIL I9S9 The AC/M v. 2. CTD001161 c*f The mixer of the special filter used (see AC/M No. 7-8, 1968) was fed with this suspension. A number of samples measuring 10 x 100 and 400 x 400 were then produced in the laboratory. The moisture content of the layer from which these samples were made and the moisture content and the strength of these samples were measured. Measurements were also taken of the thickness of the layer from which the industrial product was manufactured of the machine and also of its moisture content before suction and of the moisture content and strength of the green-end product. On the strength of the results of fifteen experiments, it was established that when using the same raw materials, the properties of the samples made in the laboratory correspond to those of the product manufactured on the factory machine if the respec tive values are multiplied by the following coef ficients: For moisture content of layer before suction Moisture content of fresh product Tensile strength of fresh product Flexural strength of hardened product after 7 days Water absorption Specific weight 0,94 0,95 0,68 0,88 0,85 1,08 Variations, resulting from different technogilocal conditions, of the above coefficients, do not exceed 6-8% and the above information led to the con clusion that the method of forming the samples in the laboratory was quite satisfactory. The influence of the thickness of the elementary asbestos-cement layer, on the physical mechanical properties of the product. The thickness of the elementary asbestos-cement layer, in industrial production, is one of the most varying values. The non constancy of the concen tration of the suspension in the vats seems to be the basic cause of this variability. In the course of the conducted experiments, the concentration of the suspension was made to vary from 6 to 14% and the thickness of the laser from 0,7 to 1,5 mm. Hydrostatic pressure was 40 crus. The prevailing conditions of filtration, the moisture content of the layer and the physical mechanical properties of the freshly formed samples are shcv.'n in table 11; for strength and specific weight of the hardened product, see fig. 29. Variation of the concentration of the suspension and the resulting change in the thickness of the layer have a great overall influence on the technological cycle. The following differences can be noted where one A/C sheet was obtained with a concentration of 8% (see table 11, experiment No. 2) and another Tadlk 11 Conditions of filtration, moisture content of the layer and the physical mechanical properties of the fresh asbestos-cement product. Expe riments Time of filtration- sees. Percent concern (ration of the suspension Thickness of layers mms. Number of layer per item Percent moisture content of layer on the wire Percent moistu re conienl of ln>er on the felt, before suction Moisture content of product Strength of green product kg/cm * R.ite of filtration cm/sec. Percent specie output of sir'-e O linder i3 23 33 43 53 6 0,732 8 1,100 10 1,120 12 1,3-10 14 1,500 6 5 4 3 3 72,3 55,06 30,3 2,74 0,33 100.0 71,7 51,40 28,5 2,09 0,31 125.0 - 69,6 48,20 27,2 2,09 0,27 155.0 69,2 47,50 26,0 1.98 0,24 14-' 67,4 . 45,S0 28,0 US 0,21 149.2 'I6 CTD001162 *1X1 AS U r* with a concentration or 12% (see table 11, experi ment No. 4): ) The layers forming the first sheet were filtrated at a higher rate (0,31 cms/sec) than those forming the second sheet (0,24 cms/sec). As it was already established (see AC/M No. 9-10/ 1968), different rates of filtration influence the struc ture of the layer and the ratio asbestos to cement in the layer; consequently the layers forming the sheets under consideration will be quite different not only in thickness but mainly in structure. ) The first sheet (exp. No.' 2) has been formed with 5 layers, in other words it passed 5 times under the Harvester roll and the second sheet was formed with 3 layers and consequently passed only 3 times under the Harvester roll. This is a considerable difference in pressure and will in fluence the strength characteristics in as much as the first sheet will result to be more compact. The above 2 differences are not necessarily connected and can be met separately; for instance when a machine is running with a felt speed of 36 m min. and concentration of the suspension equal to 10 %. the thickness of the layer will be the same as that delivered by a machine running with a felt speed cf 31 m/min. and concentration of the suspension equal to 8%. Consequently the number of times the sheet under formation is pressed in this case, will be the same but the formation of the layer on the sieve c\ finder will take place under different conditions as the rate of filtration of a faster machine will be higher and forcibly the structure of the layers delivered will diner. This establishes a fact: that with 2 machines manu facturing a product with layers of the same thick ness, the respectively yielded products can ha\e quite different physical mechanical properties. In order however to determine the reasons which cause the difference in the properties of the products Y<* Rg=kg/cm* Rp = kg/cm* ya Rp^kg/cm* Rg=kg/cm* * 1____ 0,06 0,0$ 0,12 ' Gi* Concentration of suspension in gm/cm* Fig. 29. The influence of the thickness of the l.iscr. on the phjsic.ilmech.micjl propcitio of the enj product. with chjnginc rates of filtrjlion Rg = flexur.tl strength Jt 7 d.'^s of ugc; Rp -- ten-ale strength of green product; vi - specific weight in gm.cm* to) tlS ojd tj: Concentration of suspension gm'cni* <2 15 20. Temperature of suspension tl Fig. 30. - The influence of the thickness of the b>cr on the mcchanical properties of the end product, when rale of liUra:;:t constant. 10 tWE AC CTD001163 Tafile 12 The conditions.under which n constant thickness of the layer is secured with different values of the rate of filtration. Percent concentration Time of filtration secs. Rale of filtration cm/sec. Specific output of sieve cylinder percent 10 1.8 0,378 257 9 2,5 0,322 196 -8 3,5 0,286 155 7 5,0 0,210 100 Table 13 The conditions of filtration with changing sallies of hydrostatic pressure. Hydrostatic Percent con* Time pressure, ccniralion of of filtration cms. suspension sees. Rate of filtration cm/secs. Specific ,Output percent 10 10 3 0,162 100 15 9 3 0,204 | 113 20 8 3 0,267 131 35 7 3 0,284 123 manufactured by the 2 machines, it has been neces sary to investigate separately the influence of the two factors i.e. the thickness of the layer and the rate of filtration. In the course of the experiments, the results of which are referred in table 11, both the thickness of the layer and the rate of filtration were changing at the same time and having more than one variable at a lime, it was quite impossible to establish which of the two factors was responsible for the change of the properties of the product. It became therefore necessary to conduct a further scries of experiments allowing only one variable -- in this case the thickness of the layer -- while the rate of filtration was maintained constant. This is pos sible by increasing the concentration and the tem perature of the suspension as necessary to maintain a constant rate of filtration. This experiment has a practical significance as it is exactly by this method i.e. by increasing the concentration and temperature of the suspension, that the production of the machine is increased in many plants. In comparing these values with those of previously conducted experiments, the resulting differences should not be seen as absolute values because the cement used in this scries differed in composition; however, for the purpose of establishing the influence of the thickness of the layer on the physical-mecha nical properties of the end product, such relative values are sullicient. Graph fig. 30 gives an entirely different picture from that of graph fig. 29. The elimination of any inliuencc from a change in the rate of duration, shows that the strength charactcri-dics of products manufactured with lasers of a thickness ranging from 0.76 to 1,55 mm, arc the same; only ;14: the green product showed a decrease in strength. Having established that within certain limits, the thickness of the layer hardly influences the strength characteristics of the end product when the rate of filtration is kept constant, it is now necessary to establish the influence of the rate of filtration n the thickness of the layer is maintained constant. In order to establish the influence of the rate of filtration on the strength characteristics of the pro duct, a scries of laboratory experiments were con ducted. In the course of these experiments, the thickness of the layer was maintained constant when the concentration was increased., by reducing the time the filter was kept immersed in the suspen sion. Similar conditions can be obtained with the forming machines if the speed of the fell is increased when concentration is increased. This is the method for increasing the production of asbestos-cement manufacturing machines. The conditions under which the elementary asbestos- cement layer was obtained in the course of these experiments, are referred in table 12. Hydrostatic pressure, in the course of the experi ments was 40 cms. The thickness of the Liver was always equal to 0.9-1.0 mm. The sample in all cases was formed with 5 lavers. The moisture content of the lever on the felt was varying between 60.-1 and 52.2% and the moisiure content of the ficsli sample ranged hetween 27.1 and 28,2%. These variations can be considered negligible and consequently their influence The influence that the rate of filtration exerts on me physical mechanical properties of the end product, when the thickness of the layer is maintained con-mint are shown in fig. 31. ' CTD001164 lHt AC >.! Fig. 31. * The influence of the rate of filtration on the physical-mechanical properties of the end product, when the thickness of the layer is constant. Another series of experiments was conducted for the purpose of investigating the influence of hydro static pressure, on the properties of the asbestoscement product. The conditions under which these experiments were conducted, are given in table 13. In the course of the experiments, the thickness of the layer was maintained between 0,6 and 0,7 mm. Moisture content of the layer on the felt between 52,6 and 53,9% and the moisture content of the green end product between 27,3 and 28,9,%. The influence of hydrostatic pressure on the strength and density of the hardened asbestos-cement product can be seen on fig. 32. A series of experiments were also conducted on a sheet forming machine and the method of carrying out these experiments has previously been illustrated (see AC/M No. 7/8-1968). The results of these expe riments. i.c. the influence of the thickness of the layer, of the speed of the felt, of the rate of filtra tion and of the hydrostatic pressure on the physical mechanical properties of the asbestos-cement pro duct, are shown in table 14. |is; The results of the laboratory and factory investiga tions are analysed hereunder. Conditions of operation of the sieve cylinders and their influence on the quality of the product. The best conditions of operation of the sieve cylin ders can be established by analysing the results of the investigations of the influence of the conditions of operation of the sieve cylinder on the quality of the end product. The investigation will first imoKe the dependence of the properties of the product on the thickness of the laser. The experiments carried out both in the laboratory and under actual manufacturing conditions, indicate that when the concentration of the suspension is increased from 6 to 12% and the thickness of the layer is increased from 0,732 to 1.340 mm. the strength of the product both fresh anti hardened decreases and so does the specific weight. Under the above conditions, however, the otupm. of the machine increases by 44,8%. It has been a common CTD001165 THE AC M (&) u \ 1 error to think that the decrease in strength of a product made, with a suspensiQn of high concen tration is connected with the greater moisture con tent of the thicker layer. This is wrong and it can be seen in table 11 which shows that the moisture content of a thicker layer, on the wire, is dropping continuously. The moisture content of the thicker layer on the felt, before suction, is also considerably lower. It is evident that due to the above, the moisture content of the product, within certain limits (thick ness of layer up to 1,3 mm), is also lower. When a highly concentrated suspension is used, the resulting decrease of the moisture content of the layer on the wire and on the felt allows to produce a product of relatively low moisture content and this in spite of the fact that further dewatering by pressure (exercised by the Harvester rolls or pressure rolls) is considerably reduced because the product under formation is pressed less limes. It is for the same reasons that the specific weight of the product decreases only insignificantly (see graph fig. 29). It is in fact in consideration of the above characte ristics of the process of filtration that a great number of plants throughout the world have been able to operate satisfactorily with an increased production by increasing the thickness of the layer by 5C-$0o in respect to previously recognised principles. I 0 SO 20 'JO *i0 50 00 Mydrosuiic pressure cmi. Fir.. 32. * The influence of h\ilrosuuic pressure on the phy\ica1*mcchniicjl properties of (he cml product \Ou*n thickness of Ihc l.ocr is consl.tnt. ; 20 1HE AC/SI A O 4^ 7 This of course does not mean that the quality cha racteristics of the product did not change at all. In fact the crcen product obtained with a thin layer, is considerably stronger than the product obtained with a thicker layer (see table II). It can also be observed that the products manufactured with 1,5 mm layers have a smaller specific weight and a greater water absorption than the products manu. factured with layers of 1 mm thickness, although the moisture content of both, at the green state is the same. In order to clear the reasons of these phenomena, it is necessary to distinguish clearly between density of the individual layer and density of the product as a w'hole; later, it will be demonstrated that the compactness of the end product depends not only "biTlhe force of pressure of the pressing equipment (Harvester roll or pressure equipment) but also on the number of times the elementary layer undergoes such pressure. In fact when a machine operates with a thin layer, the individual layers which form the product, in spite of the fact that they have a lower specific weight, are pressed more compactly as a result of the increased number of times they pass under the pressure equipments. When the ma chine works with a thicker layer, these layers have a higher specific weight but are not as compactly pressed in the product under formation and conse quently, in between them will exist a great number of pores filled with air. It results that the thicker the layer is, the greater will be the difference of compactness of the individual layers and also the average compactness of the end product. The presence of pores filled with air has no influence on the moisture content of the green product. When on the other hand^the water absorption of the har dened product is determined, the pores become filled with water and it is due to this phenomenon that the ratio between moisture content of the green product and the capacity of the hardened product to absorb water, is disturbed. The premature destruction of some roofing sheets for example is due to the loosening of the bond between layers with a resulting greater porosity and this porosity increases with the increase of the thick ness of the elementary layer during manufacture. The'porosity among the lay ers appears to be the weakest point of these roofing materials. When, however, the porosity among the layers in creases -- this happens when working with a thick layer -- the average compactness of the product I22, changes only insignificantly and is hardly detectable with the currently used laboratory equipment. This explains why products still deteriorate, although when tested they showed negligible changes and in some cases even a stability in strength and in capacity to absorb water. On the basis of the above, the conclusion can be drawn that the increase of porosity is connected with the pressure equipment and not with the filtration equipment of the machines. The fact that with a higher concentration of the suspension a more com pact layer is delivered, has already been demonstrated. Furthermore any change in the moisture content of the layer is due to the effect of hydrostatic pressure and not to the change of the concentration of1 the suspension. This is even more evident when consider ing that the upper part of the layer formed on the wire of the sieve cylinder, being nearer to the liquid phase of the suspension, will have a lower moisture content when the suspension is heavily concentrated and consequently also the average moisture content of the whole layer will necessarily be lower. Often noted reductions in strength of products obtained with highly concentrated suspensions, have led to the error of believing that the deterioration of the- quality of the product has been caused by the increase of the thickness of the layer. The experi mental work carried out has proved beyond doubt that such an assumption is incorrect. The assumption that any increase of the output of the machines, inevitably leads to a deterioration of the quality of the product, could have seemed correct at the time when the technological process of manu facture was not properly investigated and more specifically when the influence of the rate of filtration and of other factors, on the properties of the pro duct, was not known. The increase of the concentration of the suspension and the increase of the thickness of the layer which follows, are not the only means in the hands of the technicians for increasing the output of a plant. The output can in fact be increased without increasing the concentration, by increasing the speed of the felt in which case, sometimes, the thickness of the layer is even reduced (see AC/M No. 1 2-1969). Output of a plant can of course be further increased, by increasing both the concentration of the suspen sion .and the speed of the felt to such values ihst the thickness of the layer will remain constant. Such a case was created in the course of the experiments for the investigation of the influence of the t.ate CTD001167 THE AC. M Table 14 The influence of-the working conditions of the sieve cylinder, on the physical mechanical properties of the ashestos-cement end product. Data based on the experimental work carried out on a factory sheet forming machine. For the characteristics of the raw materials see table 8, AC/M No. 1/2 - 1969 Thickness of layer cms. Hydrostatic pressure , Speed of felt m/min. 0,030 48 17,7 0,028 48 23,6 0,040 ! i 48 23,6 0,027 48 . - 28,9 0,037 48 28,9 0,030 0,031 i 48 30 33,0 23,6 0,024 15 23,6 Rate of filtration cm/sec. Flexural stregth kg/cm1 Percent water absorption Specific weight Specific output gm/cm* percent 0,204 270 27,5 1,44 0,250 301 20,7 1,67 0,225 290 20,92 1,61 0,266 318 19,5 ' 1,70 0,253 312 18,9 1,72 0,253 314 19,2 1,73 0,240 320 18,4 1,74 0,200 380 15,8 1,80 100 124 177 146 200 187 137 107 of filtration on the physical mechanical properties of the product (see table 12 and fig. 31). When con ditions such as those shown in table 12 are met in a machine equipped with sieve cylinders of S50 mm 0, ,5 seconds of filtration will correspond to a felt speed of 18 m/min. and 2,5 seconds will correspond to a felt speed of 36 m/min. The increase of the felt speed from IS to 36 m/min. and of the concentration from 7 to 9% will correspond theoretically to an increase of the output by 96% although the thick ness of the layer remained constant. At the same time the strength of the product will increase by 11 % (see fig. 31). This was also established in the course of the expe riments carried out on a sheet forming machine (sec table 14). The increase of the felt speed from 17,7 to 33 m/min. resulted in an increase of the strength characteristics of the sheet by 15%. The thickness of the layer in this case was practically unchanged. As it has been possible to demonstrate that the strength of the product can be increased while the thickness of the layer is constant, it results that quality depends not only on the thickness of the layer but on other factors too and specifically on the rate of filtration. When comparing the two methods described hereabove, for increasing the output of a plant, i.c. by -.decreasing only the concentration and secondly by increasing both the concentration of the suspension and the speed of the felt, a considerable difference will be observed in the conditions of filtration of the asbestos-cement suspension. In the first case (see table 11) the rate of filtration decreases as the concentration of the suspension increases while in the second case the rate of filtra tion increases (see table 12) together with the increase of the concentration. It has already been established that an increase of the rate of filtration causes the asbestos-cement layer to become less heterogeneous and improws the ratio of asbestos to cement in the layer (see AC/'M No. 9. 101968). ll was then assumed that these factors must have a positive effect on the quality of asbestoscement. The experiments ha\e now confirmed tfie correctness of the supposition. It is in fact the im provement of the structure of the laser that explains the improscmenl of the strength and density of the product (see fig. 31). Having established that the strength characteristics of the product depend oil the thickness of the laser and on the rale of filtration, the investigation must be CTD001168 1HE. AC/M carried further in order to determine to what extent each of these factors exerts an influence on the strength of the product. This can be done by analysing graphs figs. 30 and 31. The analysis shows that the increase of the rate of filtration from 0.250 to 0.3S0 cm/sec. (52%) causes an increase in strength of the green product from 1,80 to 2,60 kg/cm- (45%) and that of the hardened product from 155 to ISO kg/cm2 (16%), see fig. 31. It results that excry one percent of increase of the rate of filtration leads to an increase in strength of the green product by 0,S6% and to that of the hardened product by 0,31%. The decrease of the rate of filtration leads to a reduction in strength of the product in the same proportions. On the other hand, the increase of the thickness of the layer, while the rate of filtration is constant, from 0,75 to 1,5 mm, or by 100% (see fig. 30) reduces the strength of the green product front 2,40 to 1,95 kg/cm or by 19%. The strength of the hardened product can be considered as having undergone no change as the 4% registered is within the limits of possible errors. The increase of the thickness of the layer from 0,73 to 1,50 mm or by 105% (see fig. 29) and the si multaneous decrease of the rate of filtration from 0,38 to 0,21 cm/sec. or by 36% (see table 11) reduces the strength of the green product by 4S% and that of the hardened product by 18%. In this case, if only the rate of filtration is decreased at the same proportion without increasing the thickness of the layer the reduction in strength of the green product is 31% and that of the hardened 11%. It results that cserv one percent of increase of. the thickness of the layer, causes a reduction of the strength characteristics of the green product equal to 16% and of the hardened product equal to 0,07% It can consequently be concluded that every 1 % of increase of the rate of filtration exerts a consi derably greater influence (approximately 5 times) on the strength of the product than the equivalent change in the thickness of the laser and that the change of the thickness of the la>er affects only the strength characteristics of the green product. This investigation has explained why a factory could still produce a good quality product even though the thickness of the laser, in respect to presious methods of operation, ssa.s increased in some cases by 1,5 times. This happens because the thickness of the laycr has practically no influence on the strength characte ristics of the end product; strength is afTcctcd by the rate of filtration mostly. It was obserxed that in this case, simultaneously to the increase of the thickness of the layer, the temperature of the suspen sion and the speed of the felt were increased. The above is correct as far as the hardened product is concerned but things are quite different when the green product is involved. Machines working with a thicker layer, yield a green product which has a considerably reduced strength. The result of this investigation can therefore be .summarised as follows: any increase of the thickness of the layer will influence only marginally the strength ! characteristics of the hardened product but will reduce considerably the strength of the product in the green state, and this is confirmed both in the lab_boratory and under actual manufacturing conditions. The assumption that the strength of the end product was influenced by the thickness of the layer, was based oh the fact that the more easily measurable factor -- thickness of the layer -- was only investi gated while the real cause -- the rate of filtration -- was neglected. Considering the influence that the change of the rate of filtration and of the thickness of the laser, have on the properties of the end product, it max be assumed that these arc solely connected with the working conditions of the forming machines; in practice, hoxxex'er, and in the course of the investi gation, cases were met where the rate of filtration and the thickness of the layer were changing while the speed of the fell and the concentration of the suspension were maintained constant. This happened when the filtration properties and the temperature of the suspension were altered. In fact the composition of the rnxv materials, has a great influence on the rate of filtration. Optimum rate is obtained when the fibres are treated and opened to a correct degree. It is in a following article that the influence of the properties of the raw materials, on the filtrating properties of the suspension, will be discussed. At this point, attention should he drawn to the fact that had filtration resulting from dilTerent factors or causes, leads not only io a reduction of the output of machines but also to n decrease of the strength of the product due to the deterioration of the structure oT the asbestos-cement lajer xxhich influences the properties of both the green and hardened end pro duct. Any dctetioratioii of the quality of the pioduct. due to low temperature and reduced filtration pro- MARCH - APRIL 1959 CTD001169 25 perties of the suspension, compels the machine operators to work with a low concentration of the suspension in the vats and thus reduce the output of the plant. It is also to be noted that increase of the rate of filtration is not unlimited. Graph fig. 31 shows that the increase of the rate of filtration is useful only to a certain extent, up to 380 cm/scc. Going above this rate, will cause a decrease of the density and strength of the product. The limit beyond which the rate of filtration cannot ,be increased,, seems to have strong foundations if the figures referring to the ratio asbestos-cement in the elementary layer (see AC/M No. 7/8 - 1968) are remembered and which show that a too thin layer is never well saturated with cement. In this case, with a rate of filtration higher than the mentioned limits, cement is not only carried away with the filtrated material, but a disturbance occurs which causes not only the free cement to be carried away by the filtrated material, but also part of the cement which had in the meantime adhered to the asbestos fibre. Such a thin layer, formed under the above conditions can neither be strong nor compact. In the asbestos-cement industrial plants, machines are equipped with sieve cylinders of different dia meters; the hydrostatic pressure differs according to the diameter of the sieve cylinder. The data shown on table 13 and fig. 32 (laboratory experiments) and table 14 (experiments on a factory machine) indicate that with the reduction of the hydrostatic pressure, the strength and compactness of the pro duct increases though the increase can be consi dered insignificant when pressure reduction oscil lates between 60 and 20 cms. Consequently, if can be established that the dif ference in hydrostatic pressure resulting from the use of sieve cylinders ranging in diameter between 1200 mm and 500 mm is unimportant and has no influence on the quality of the product. However, the diameter of the sieve cylinders is also connected with the rate of filtration and this dependence will be discussed later. The experimental work carried out has shown that when hydrostatic pressuie is reduced below 20 cms, the strength of the product shows an inciensc of 20-25% (sec fig. 32 and table 14). Operating with a hydrostatic pressuie of 15 cms, it lias been possible to produce fiat asbestos-cement sheets which at the age of 28 days had an average I26! flexural strength (both directions) of 3S0 kg'em1 and water absorption of 15,8%. This is much more than what is currently obtained in practically every asbestos-cement plant and testifies to the fact that the possibility exists to improve the standard quality of asbestos-cement products by improving the tech nological cycle of manufacture. Analysing the above results it became obvious that this increase in strength could only be explained by the improvement in the structure of the layer, as all the other factors were unchanged. When in the work published in AC/M No. 9 101968, the formation and structure of the layer during the process of filtration was investigated, it was pointed out that the part, or the layer which was in direct contact with the wire, is poor in cement content as a result of the fact that cement was car ried away by the filtrated material. It was observed that during the formation of the layer, with high hydrostatic pressure, the particles of cement are pre vented from depositing on the layer because of it.great compactness; however, at the very moment this compactness is reduced to the extent that satur of the layer with cement becomes possible, then ,r.e layer becomes more homogeneous and consequent:;.' the strength of the product improves. Hydrostatic pressure corresponding to such a moment, as the experiments have shown, is between 10 and IT cats. In order to obtain a hydrostatic pressure within the above indicated values, sieve cylinders of 300 to 400 mm 0 should be used. Alternatively hydrostatic pressure can also be reducewhere cylinders of a larger diameter are used, increasing the level of the water inside the siccylinder, thus reducing the cliTerence between the level in the vat and the level inside the sieve cylinder, to the required values. Table 14 contains data on products obtained on a machine, operating with different thicknesses of layer and Kites of filtration. If the data is used for the calculation of the change in strength, it will be found that the strength of (he product should in crease theoretically by 7.2% when the rate of filtra tion is increased from 0,204 to 0.250 cm sec. lire cxpetimenls showed that the increase was I I The compuiison of the results of the with the results of the experimental work, m.me.r.e that the established relation in the labotatory tween the strength characteristics of the prouuet and the conditions of operation of the sieve cy imerr. is also valid under actual manufacturing conditions. CTD001170 m ac -o