A simulation of powder compaction on two dimensional model introducing cohesion probability P' has been reported previously2). In the present work the coordination number of a disc was taken into consideration in the process of shifting the disc to the lower layer, and another probability P" was introduced in the step of insertion of a disc into the hole created by shifting of the former disc. The data of porosity and tapping number followed Kawakita's equation when P' was relatively large and P" small. On the other hand, Kuno's equation was applicable when P' was small and P" large. Furthermore, comparisons between two-dimensional porosity εA and three-dimensional porosity εV were made using both the regular packing model and the random packing model.
It has been considered that a variation of bulk density value, which is one of the most important characteristics of powdered materials, depends on the physical properties of the particles and the ways of packing. The important problems for the determination of bulk density of powder are how to get (1) loose packing, (2) high precision and good reproducibility and (3) simple and speedy measurement. The authors have tried to solve these problems and obtained that a loose packing can be achieved by employing horizontal vibration without accompanying an initial large shrinkage of volume, which is common in tapping or vertical vibration. The actual measurement of bulk density indicates that the date lie within 1% coefficient of variation in the case that the horizontal vibration is applied during the flow-in packing process and the leveling-off process simultaneously.
It is well-known that the measured value of the angle of repose varies dependig upon the method of measurement. The present author indicates that, although the angle of repose is placed in one category by its definition, it should be divided into two different groups. They are named the maximum angle of repose αRM and the stable angle of repose αRS. On the analogy of the tilting box method, the maximum angle of repose αRM corresponds to the angle of slope at the moment that a box begins to slip, whereas the stable angle of repose αRS corresponds to the angle of the residual slope after slipping (Fig. 1). The relation between these two angles of repose and the measuring method has been investigated based on the previous results in the literatures. The sliding method (the tilting box test or the revolving cylinder test) gives the maximum angle of repose, while the draining method gives the stable angle of repose. The pouring method (the free pile method or the restricted pile method) yields the values scattered between the maximum angle of repose and the stable angle of repose, which varies also with the technique of supplying powder. These facts are clearly seen in the reports of Takahashi2) (Fig. 3) and of Train12) (Fig. 4). In order to evaluate the flowability of powder, the maximum angle of repose should be chosen as the angle of repose. The tilting box method is the best for measuring the maximum angle of repose, because of its reliability of measured values, its clearness of forming (or tilting) speed, its possibility of measurement at any bulk density, and its accurate inspectation of two dimensional scale effect.
The relation between the surface kinetic angle of powdered particles and their shapes was studied by using two kinds of measuring apparatus: The one in which a base plate was fixed horizontally and the other in which a barrel turned slowly around a horizontal axis. The experimental results obtained from both methods indicated that the geometrical piling configulations of the powdered particles were controlled by the law of sliding friction until the pile angle with a horizontal plane reached to 30°. Based on the Helz's theory of elasticity and the laws of solid friction, the surface kinetic angle was expressed theoretically by the following equation: μr=tanφr=K(π/6·d3Ψ)-1/3α.r2/3 in which a new concept of a shape factor of powdered particles named‘degree of ratio to sphere’ was introduced. The calculated values from the above equation coincided fairly with the experimental values.
The angle of internal friction is one of the most important properties of powder and is often used to characterize powder in industry. In general, the angle of internal friction is determined from the slope of yield locus. However, it has been reported in many previous works that the shape of yield locus was affected by the preconsolidation or compaction pressure. Thus the significance of this property has been confused considerably. In this paper, the shape of yield locus was studied to establish the rheological significance of angle of internal friction through observations mainly from the view point of microscopic structure of powder beds. For lactose and calcium carbonate powders compacted under various pressures, the yield loci were determined. The results obtained show that the curves of shear index determined from the slope of yield locus can be divided into two characteristic parts at the specific compaction pressure, and that the porosity and strength of compacted powder bed change discontinuously also at this critical point. It can be concluded that the contact mechanisms of particles have been affected by compaction pressure, and also that the shape of yield locus is determined by the contact mechanisms.
The effect of pre-adsorbed water on profiles of adsorption isotherms of water vapor on NaCl powder was investigated. Measurements were also made on volume shrinkage of the powder bed and surface area change of the particles during the water adsorption. And an interrelationship among the above three properties was discussed. Each of the adsorption isotherms obtained showed a stepped increase in the amount of water adsorbed at the vapor pressure where the formation of two molecular layers completed. When the samples were heat-treated at above 200°C or ground in a dry atmosphere, the ratio of the step-width to monolayer capacity in the isotherms decreased. But when the samples were exposed to the repeated cycles of adsorbing ca. two layers of water followed by outgassing at room temperature, the above ratio was restored to the value for the untreated salt. When the amount of water adsorbed exceeded three layers, the powder particles began to aggregate and the surface area decreased.
In order to examine the effects of several manufacturing conditions on the surface physical properties of magnesium silicates, five kinds of magnesium silicates were produced by four different methods. SiO2/MgO mole ratio, apparent specific volume, specific surface area, acid strength and acidity at pKa=+1.5 and pKa=+3.3 were measured. It was found that these physical properties were related to the adsorptive power for dyes (M·B, M·G, S, B·M·G) and to pH condition during manufacturing of magnesium silicates. Magnesium silicate with the strongest adsorptive power had the following properties: (1) larger SiO2/MgO ratio, (2) smaller apparent specific volume and (3) smaller specific surface area. Acid strength and acidity of the surface of magnesium silicates had no relation to the adsorptive power for dyes.
Mechanochemical behaviors of several hydrated crystals having such three types of crystallization water as H2O, OH radical and hydrogen bond were investigated. Mechanochemical dehydration progressed in accordance with the order of thermal stability of the hydrates. Typical structural changes of the hydrates during grinding process were as follows; (1) CaSO4·2H2O→CaSO4·1/2H2O→IICaSO4, (2) Mg(OH)2→brucite gel (in the air), (3) Mg(OH)2→MgO(in vacuum), (4) Al(OH)3 gibbsite→κ-Al2O3→α-Al2O3, (5) CaHPO4·2H2O→CaHPO4→Ca2P2O7 (almost amorphous phase), (6) TiO2·nH2O gel→anatase→rutile. It was found that most of the mechanochemical dehydrates except (5) existed as aggregation of fine crystals (1/100μ order) and did not transform into amorphous phase contrary to the phenomena observed in general products of grinding. Surface properties of the dehydrates, distorted by shear stress, were discussed in terms of their acidity and heat of wetting.
In order to perform the evaluation of airbone dust in facilities for the atomic energy and to study the inhalation of radioactive aerosols, proper attention must be paid on the particle size distribution of the aerosols. Particularly, the monodispersed aerosols of proper size are required for the evaluation of internal exposure and the calibration of particle counters. This paper describes the study on the mechanism of generation of aerosols from a spinning disc aerosol generator and the characteristics of the generated particles. Monodispersed aerosols of the size from 2 to 12μm with the geometric standard deviation of less than 1.2 could be generated from the solution of methylene blue by a 2cm-diameter glass spinning disc spun by an electric motor at 20000 to 40000rpm. The particle size or mass median diameter of the generated aerosols indicated a linear relationship with cube root of the concentration of feeding solution. Aerosols were collected by thermopositor and their sizes were measured by taking their microscopic photographs. As to the relation between the surface tension of feeding solution and projecting distance of main droplet, the following results were obtained: The projecting distance of main droplet was in inverse proportion to the angular velocity of disc but was in proportion to the surface tension of feeding solution and to the square of K value of the equation (2). The experimental values agreed approximately with the calculated values.
The relation between contact number and porosity in randomly packed beds of mixtures of various-sized spheres was investigated. The materials used were polystyrene spheres of 1-6mm in diameter. The spheres were separated by sieving into five size fractions. Three samples having various size distributions were prepared by remixing the appropriate spheres of different size fractions. The packed beds were prepared by various methods of packing. The porosity values were ca. 0.64 for loose packing and ca. 0.37 for dense packing. Red ink was poured into the packed bed, filled it up and then drained. The ink was held by capillary force at each contact point after draining. After drying, the marked spheres were separated by sieving into original size fractions, and the distribution of contact numbers was obtained by counting the number of ink marks on each sphere in the same fraction. The contact number in each fraction showed normal distribution, and it increased with increasing particle size and with decreasing porosity. When the porosity became above 0.45, the relation between the mean contact number and porosity coincided with the result obtained by Ridgway and Tarbuck.
In order to investigate the effect of humidity in the atmosphere on the cohesion of powder particles, the adsorption weight of water and the cohesion forces between particles were measured in various humidities. Spherical glass powders with three kinds of particle sizes (17-100μ) and hydrophobic glass powder made by silicon treatment of the former powders, were used as samples. Cohesive forces were calculated from the tensile strengths of powder beds measured by Du Noüy surface tension balance. In the adsorption isotherms at 25°C, the adsorption weights were nearly constant until the relative humidity reached about 60%. Beyond this point, the weights increased with increasing humidity. When the adsorption isotherms were represented in the form of adsorption weight per unit surface area of particles, the isotherm curves for the powders of different particle size roughly coincided in the constant range, but they were different in the untreated samples and the hydrophobic samples. These constant values corresponded to 13 water molecular layers and 6 layers, respectively. On the other hand, the tensile strengths were low and roughly constant in this constant adsorption range, but they increased sharply with an increase in humidity, and then decreased. On the basis of the above results, it was considered that the water vapor was uniformly adsorbed on the surface of particles first, and then condensed at the contact points between particles, forming the water bridges. The liquid bridge forces were calculated and compared with the theoretical results by Rumpf on the assumption that the tensile strengths were given by the product of the number of contact points and the cohesive force of a liquid bridge. The experimental results corresponded to the theoretical curves with the relative distance of about 10-2 and a contact angle of 0°for the untreated samples and 40-60°for the hydrophobic samples. The relative distance is the ratio of the distance between the particle surfaces at a contact point to the particle diameter. The bridge forces, however, had maximum values which were about one order of magnitude lower than the theoretical values and decreased sharply. It might be considered that irregular distribution of adsorbed water or the formation of secondary structure of particles in the powder bed occured.
Though photo-extinction sedimentometry is a very sensitive measuring method of size distribution, it has been recognized that this method brings some measuring error which is attributed to the instability of suspension. Kaye et al. has pointed out already that this instability is caused by the dropping of temperature at the liquid surface by evaporation. In this report, this instability phenomenon is analyzed quantitatively by using two types of photo-extinction sedimentometers. Four kinds of liquids were selected as dispersing media, each of which had different thermal properties. The starting time of instability was obtained experimentally, and the corresponding Rayleigh number (λ=Pr·Gr) was calculated with the aid of surface temperatures, which were obtained theoretically from the evaporation rate determined with the same cell. It can be concluded that the critical Rayleigh number is in the range of 3000∼5000, beyond which most probably instability in liquid suspension will occur.
The surface of silica gel was treated in an autoclave by the solution of alcohol or phenol with some functional groups in hexane or acetone, and their surface properties were investigated by determining their preferential dispersion into the media consisting of two immiscible components such as n-hexane and water. Furthermore, the identification of the surface groups and their thermal and chemical natures were examined through the infra-red spectra, differential thermal analysis (DTA), thermogravimetry (TGA) and qualitative analysis. The conclusions obtained are as follows: (1) The silica gels treated by diethylene glycol, 3-chloropropanol, phenol, or salicylic acid were hydrophilic whereas the ones treated by o-, m-, or p-cresol and hydroxybenzol were hydrophobic (lipophilic). (2) The characteristic absorptions due to alcoxy or phenoxy groups were observed in the infra-red spectra of the surface-treated silica gels. Since the intensity of the absorption at 950cm-1 due to bending vibration of Si-OH was almost the same for both the untreated and phenol-treated silica gels, it is still questionable whether silanol combines chemically with phenol or not. (3) DTA showed the exothermic due to dissociation and oxidation of the surface groups at 250°C, which is higher than the normal boiling point, and the remarkable exothermic due to oxidation of deposited carbon. Therefore, the surface groups are considered to combine with silanol chemically, except phenol. (4) These surface groups are confirmed to be alcoxy or phenoxy, since they were dissolved into boiling water or boiling dil. H2SO4 and, furthermore, from the qualitative analysis of dissolved solution, they were confirmed being alcohol or phenol used in this study.
The surface properties of silica gels treated with n-pentanol were investigated by the adsorption of argon at 77°K, water vapour at 288°K and n-heptane at 273°K and by the heat of immersion of water and n-heptane at 298°K on the samples having various numbers of surface groups. The following results were confirmed: (1) For the silica gel samples which were untreated or had a small number of surface groups, the shape of isotherms of water vapour adsorption was the type II in BET classification, whereas it was the type I for the samples with a large number of surface groups. (2) The shape of isotherms of n-heptane adsorption was the type II for both the untreated and surface-treated samples. The existence of many capillary pores less than 20Å in diameter was suggested, since the volume of pores larger than 20Å in diameter appeared to be exactly the same for both the untreated and surface-treated silica gels. (3) The surface areas obtained from the absorption of argon, water and n-heptane showed a linear dependence on the number of surface groups. From their slopes, it was concluded that each of the surface groups covered 1.5 unreacted silanol and these surface groups would presumably be distributed on the surface of silica gels uniformly. (4) The heat of immersion of water showed a linear relation with the monolayer adsorption value obtained from water vapour adsorption.
In order to clarify the effects of particle size and density on the segregation of the particles, some experiments on the permeation of the particles having different size and density from those of the bed were carried out in a moving bed by using glass beads and spherical particles of alumina with various sizes and densities. The results indicate that a simple flowing pattern of the particles exists in the moving bed and the degree of permeation can be represented by the following empirical equation. xT/xB=0.8(ρT/ρB)+1.2(DB/DT)n-1 where n=1 at DT≤DB or n=2atDT>DB and xT/xB is the quantity defined as an index of the permeation degree of the particles in a moving bed, which is independent of the moving distance of the bed. It is suggested that the segregation characteristic of the solid mixtures of particles having two different sizes and densities in a horizontal rotating conical vessel is closely related to the value of xT/xB obtained in the moving bed.
An experimental study on the effects of operational conditions on the formation of ultra fine powders was carried out by a reciprocating friction mill with marble as the specimen. The effects of the working speed, the working pressure and the contact area on the production rate of powder were investigated and the following results were obtained. (1) The effects of the working speed and the working pressure were not significant for the particle size of the powders produced by this mutual friction and abrasion method. The mean particle size was almost constant around the size of 0.2∼0.3μ. (2) The production rate of powder was greater when the contact area of specimens was smaller. (3) It seems that the mechanism of abrasion of marble has analogy with the wearing mechanism of metals in consideration of the relation between the mass of produced powder and the working speed. The details have to be experimentally studied from the mechano-chemical point of view. (4) The mass of the produced powder varied with the working speed and the working pressure when the contact area was constant, and the maximum mass was obtained at the speed of about 6cm/sec and the pressure of 1.5Kg/cm2 in this experiment. (5) Between the production rate of powder Q and the working pressure P, the relation of Q=cPn was obtained. From the correlation of the index n and the working speed, it seems possible to conjecture the dependence of abrasion mechanisms on the working speed.
A series of grinding tests were made to investigate the effects of additive amount of grinding aid and feeding rate of cement on the performance of an open circuit vibrating mill during fine grinding of cement. The determination of grinding characteristics were performed, as shown in Table I and II, by varying the additive amount of grinding aid (Ca=0∼0.08%) and the feeding rate of cement (F=10∼44kg/h). At the same time, the mean residence time in the mill, ta, mixing characteristics, σ and E, and hold-up of powder, H, were determined as shown in Table III and IV, by using a small quantity of tracer (barium sulfate powder). The delta response of tracer concentration was found to be represented well by a logarithmic-normal distribution even in the case with grinding aid, as shown in Fig. 2. It was also found that the addition of grinding aid improves grinding efficiencies, enables to make extremely fine cements and improves size distribution curves of products as shown in Table I and II. With the increase of the additive amount of grinding aid Ca or the feeding rate of cement F, the mean staying time ta becomes shorter and the mixing coefficient E becomes larger. The hold-up H decreases with increasing the addition of grinding aid Ca and with decreasing the feeding rate of cement F.
In this report, the effect of gas atmosphere on the efficiency of size reduction of limestone powder is investigated experimentally by using a laboratory-scale vibration ball mill with ceramic balls (18mmφ). Air in the mill was evacuated and then replaced by other gases such as vapour phase of water, ethylether, methyl alcohol and acetone. The grinding experiment with a fine solid additive (aid) and that in a vacuum mill were also tried for comparison. The experimental results, shown in Figs. 3 and 4, reveal that the gas atmosphere of the above material increases the efficiency of limestone grinding distinctly, as same as the grinding with a solid state additive. These results can be represented by the following equations, which were proposed originally in the previous paper by one of the authors and have been modified here; SW>SWR; dSW/dE=ηr0 or (SW-SW0)=ηr0(E-Ei) SW>SWR; dSW/dE=ηr0-a(SW-SWR)n}(2) The improvement of grinding efficiency is found to be due to the increase of incipient efficiency of grinding, ηr0, the decrease of hindrance effect of grinding by fine powder produced in the mill, n, and the increase of the upper limit of Rittinger's range (SWR). The effect of these factors are summarized in Table I. From these, it can be estimated that the mechanism of the aiding effect of gas atmosphere is different from that of fine solid additive. And also there is no clear evidence that the dipole moment of gaseous material is an important factor of this aiding effect.
The effect of grinding on the transformation from the cubic to the tetragonal structure of Cu-ferrite caused by subsequent heat-treatment was investigated by X-ray diffraction analysis. The sample of cubic Cu-ferrite, which is metastable below 360°C, was prepared from powders of CuO and Fe2O3 by heating at 930°C for 3hr and subsequent quenching into water., The molar composition was Cu1.001 Fe1.999 O3.950. The sample was ground in a ball mill at room temperature for various periods of time up to 80hr. With increasing grinding time, the strain in the lattice increased and the crystallite size decreased. Each of the ground samples was heat-treated at 280∼340°C for 1hr. It was clear that the short-time ground samples showed more obvious progress of transformation from the cubic to the tetragonal structure than the original one, but the long-time ground samples did not transform into tetragonal structure. In the long-time ground samples, the small crystallite size and a number of crystal imperfections may suppress the occurence of the cooperative Jahn-Teller effect, thus blocking the phase transformation to the stable tetragonal structure. The effect of chemical stoichiometry of Cu-ferrite on its phase transformation was also investigated. The samples with different oxygen contents were prepared by firing in the atmosphere having different partial pressure of oxygen. The result of the heat treatment of the samples at 350°C for 1hr revealed that the transformation from the cubic to the tetragonal structure becomes depressed with increasing oxygen deficiency.
It has been well-known that zinc oxide takes easily a nonstoichiometric form. Consequently, it is considered that the lattice distortion and defects appear easily in zinc oxide after mechanical treatments such as compression or grinding. Such mechanochemical effects of oxides have been hitherto studied mainly by the X-ray diffraction measurements. However, the measurements of the physical properties, such as optical absorption spectra and electrical conductivity, seem to be more effective ways of studying the defects of zinc oxide produced after such mechanical treatments. For these reasons, the defect structure of the ground zinc oxide was studied by measuring its absorption spectra, bulk current and photocurrent. Those results were compared with the result obtained by the X-ray diffraction analysis. The results obtained are as follows: Zinc oxide has a fundamental absorption band at the wavelength of 374mμ. As the grinding time of zinc oxide becomes longer, this absorption band becomes broader and shifts toward longer wave-lengths. Moreover, another broad absorption band in the visual range becomes distinctive with the grinding time. Such spectral changes may be due to the increase of interstitial zinc produced by grinding. Photocurrent of zinc oxide decreases remarkably when ground for 50 hours, after that it decreases gradually with grinding time up to 200 hours. Such decrease of photocurrent may be due to the oxygen vacancies in the surface of crystallites, produced by grinding.
In chemical industry and mining, slurry is being used in many ways. For example, it is not only used during thickening and separation processes by a liquid cyclone or centrifuge and wet attrition-grinding, but also applied as a filler for paper manufacturing or a coater, etc. What one should be cautious about in dealing with slurry is the abrasion of various materials by slurry, which results in the deterioration of quality of slurry, the lowering of efficiency of the apparatus, the decline of productivity caused by stoppage of operation when parts are being exchanged, the increase of maintenance cost, and so on. To avoid these demerits, we devised an experimental slurry abrasion tester and investigated slurry abrasion phenomena of various materials through several experiments. The slurry abrasion of various materials can be summarized as follows: (1) Slurry abrasion progresses almost linearly with time. (2) Slurry abrasion progresses in proportion to the concentration of slurry. (3) Slurry abrasion progresses in proportion to about the 2.6th power of the rotating number of the rotor. (4) Slurry abrasion increases with increasing the particles size in slurry. (5) Slurry abrasion decreases with increasing the viscosity of dispersion medium. From these experiments it was found that as metallic materials stainless steel and nitriding steel are resistant to abrasion, and that as organic materials vinyl chloride, rubber, nylon, and polyplopylene have an abrasion-resisting power. And it was also found that the surface of the materials which had suffered from abrasion shows marks of deformation and scratch. The present experimental method for slurry abrasion, though inconvenient because of the batch method, makes it possible for us to know in a relatively short time the progressing state of abrasion of various materials by using the actual slurry and materials.