An investigation was made on the method of concentrating oxygen in air under the gauge pressure of 0∼3g/cm2 at a normal temperature by using Itaya-zeolite composed mainly of clinoptilolite. Approximately 50 volume % of oxygen-concentrated air could be obtained by a two-cylinder small-sized pressure swing device. Though Itaya-zeolite has less adsorbing power of nitrogen than the molecular sieve 5A or 4A, it will be useful for oxygen-enriched air in the fields such as waste water treatment and medical attendance.
The quantitative determination method of the chemical composition of a solid matter by using an Electron Probe Micro Analyzer has been widely applied in the analysis of metals and minerals. However, this method has not been used for catalysts, synthetic minerals and composite materials which are porous matters. It comes from several reasons: (1) surface roughness of porous matter, (2) increase in electron beam penetration depth, and (3) electric charge up in an insulator such as oxides and organic substances. In the conventional quantitative analysis of EPMA, nonporous matter is preferred not only as the measuring samples, but also as the standard samples. The main purpose of this study is to find out a method to calculate the true composition of a porous matter referrable to the data of nonporous samples. The present report describes the influences of the pore structure of analyzing samples on the X-ray intensity observed and on the composition corrected by the conventional Bence and Albee calculation method. Three series of samples known as industrial catalysts, i. e., SiO2 gel, SiO2Al2O3 (W) and SiO2MgO (SM) were used, and each series consisted of several samples controlled to various porosities by calcination. The pore structure was measured by a mercury porosimeter. The results obtained were as follows. (1) The X-ray intensity was found to be greatly influenced by the porosity and the pore structure of the samples. (2) The experiments of SM calcined at lower temperatures and of porous metal (Pb) plate show that the data are meaningless unless a proper linear relation between the incident beam current and the absorbed one exists, since the sample can be charged up electrically. (3) The sum of the mass fraction corrected by the Bence and Albee method approached to 1.0 with decreasing porosity of samples W and SM.
An experimental study on the producing mechanism of ultra fine powders in the initial wear region was carried out by a rotating friction mill devised to produce submicron powders by mutual friction grinding of two specimens of marble. The following results were obtained. (1) The frequency distribution of inclination of a tiny surface protuberance showed the shape of a log-normal distribution and the relation of Hmax/tanθ=a·Hmax+b was obtained between the average inclination and surface roughness of specimens. (2) The frequency distribution of diameter of contact protuberance area showed also the shape of a log-normal distribution and the average diameter agreed with the theoretical value. (3) The number of contact point of a tiny surface protuberance increased as the surface roughness decreased. (4) The standard deviation of a roughness distribution curve increased with the surface roughness of specimens. (5) Two theories of cohesive wear and abrasive wear were examined as the producing mechanism of ultra fine powders in the initial wear region by mutual friction grinding. When the surface roughness of a specimen was under about 10μ, the formation of powders was predominated by the cohesive wearing mechanism and over about 10μ powers was produced by the abrasive wearing mechanism. It is concluded that the surface roughness of a specimen is an extremely important factor in mutual friction grinding.
Grinding of quartz was carried out in cetyl alcohol with two kinds of ball mill; the one was made of glass and the other was a rotating micro autoclave made of stainless steel. The results were compared with those of the ones without cetyl alcohol. The grinding conditions were from 2 houres to 64 hours at room temperature or high temperatures. The effect of cetyl alcohol, temperature and grinding hour on the size distribution of quartz particles were investigated by sedimentation and scanning electron microscopy. It was found that the percentage under 15μ diameter of the treated quartz was larger than that of the non-treated quartz and the degree of coagulation of the treated particles was less than that of the non-treated one. The effect of temperature was not clearly seen.
To investigate ball mill grinding from the kinetic point of view, the effects of hardness of samples, feed weight and feed size on the rate constant of weight fraction increase in the ground product have been studied. The samples used were silica glass, borosilicate glass, quartz, feldspar, limestone, marble and talc. The pebble mill consisted of a 13.2cm-long cylinder of 12.8cm diameter and 40 balls of 3cm diameter having the total weight of 1180g. In the case of wet grinding, 400 cc-water was added to the mill. The tests were done at the speed of 88.1% of critical one. The experiments were performed on the samples of 5 different particle sizes from 14 to 65 mesh for dry grinding and 4 different ones from 20 to 65 mesh for wet grinding to obtain the grinding rate constant. The following results were obtained. Dry: Kx=0.024HV-0.47(WS/WB)-1.5·(x/xf)·(xf/xo)0.7 (xo=0.119cm) Wet: Kx=0.05HV-0.47(WS/WB)-1·(x/xf)·(xf/xo)0.5 (xo=0.084cm) where Kx is the zero order increasing rate constant of weight fraction less than size x, HV the dimensionless Vickers hardness, WS the feed weight, WB the ball weight, xf the feed size and xo the optimum feed size.
Tensile and compressive strengths of fine magnesia, alumina and silica powder beds were measured. Although the direct tensile test has been commonly used for powder beds with sufficiently high tensile strength, the radial compression test was employed in the present study because of its simplicity and easiness. The size effect of test pieces and the distribution of its strength were discussed using the specimens with different diameter (D) and thickness (W). The results are as follows. (1) The tensile strength (St) of the specimens with W/D=1/2 slightly decreased with increasing diameter of powder beds as shown in Fig. 3. (2) For the powder beds having the same diameter but different thickness, the tensile strength had a peak at W/D=1/2, beyond which it decreased gradually and below which it decreased rapidly as shown in Fig. 4. (3) The distribution of strength for powder beds with W/D=1/2 and D=20mm could be approximated by a Gaussian's distribution or Weibull's distribution with the high coefficient of uniformity (m>20) as shown in Fig. 5. The compressive strength (Sc) of the specimens with W/D=2 was measured by the conventional uni-axial compression test. As shown in Fig. 7, Sc Was related to St by the equation, Sc≈7 St. On the other hand, the Griffith's theory with the assumption of random crack distributions gave Sc=8St. From this relationship and the above (3), it is suggested that powder beds are highly uniform brittle materials.
The frit powders coated with methyl methacrylate polymer were prepared by polymerization of monomer. The reaction was carried out under various conditions by adding sulfurous acid solution to the mixed suspensions of methyl methacrylate monomer, powdered frit and water. The relationship between the polymer content in the polymer-coated frit and the reaction condition was studied by thermal analysis. The polymer content was found to be affected by the reaction temperature, the methyl methacrylate monomer to frit ratio, the amount of sulfurous acid and dissolved oxygen in water. The surface potential decay was measured in the corona-charged polymer-coated frit layers. The corona-charging properties of the powders were found to depend on the reaction condition. The polymer-coated frit prepared under the dissolved oxygen-free condition generally showed a high surface potential and its initial surface potential was as high as that of the methyl methacrylate polymer powder. On the other hand, the initial surface potential of the polymer-coated frit prepared in the oxygen-containing water was about half of the value of methyl methacrylate polymer powder. This may be attributed to the difference in corona-charging properties of the coatings of methyl methacrylate polymer.
In order to investigate the effect of the shape of an immersed body on the settling velocity of slurry, the experiments were carried out by using several kinds of immersed bodies and CaCO3- slurry. It was found that the settling velocity increases by immersing a body having many projections in the slurry. The velocity increased with increasing the number of projections on the body, because clear water is produced by settling under the projections. The experimental results were discussed by making some assumptions about settling, and the settling velocity of slurry accelerated by an immersed body was estimated. It was possible to calculate the increase in settling velocity on the assumption that newly produced clear water or liquid under the projections moves upward passing through some slits existing on an immersed body without remixing with slurry.
In the present study the effects of non-ionic surface active agents added to suspending liquid on the wet spherical agglomeration process and the micromeritic properties of agglomerates were investigated. Sulfisomidine particles (100g) having a diameter of 73.3 or 79.3μm were dispersed in chloroform (2l) and were agglomerated with buffer solutions (pH 10.2, 50ml). The system was agitated by a turbine type agitator with 6 blades at 630rpm. Each of the six different surface active agents with HLB values in the range of 1.8 to 15.0 was added to the suspending liquid prior to starting the agglomeration. The length mean diameter of agglomerates and the geometric standard deviation decreased with increasing the concentration of surface active agent and attained approximately to the minimum at the critical micelle concentration. The above phenomena were enhanced as the HLB increased, especially in Tween 80 (HLB, 15). The agglomeration process was described as the first order kinetics. The kinetics parameters such as the rate constant, the population densities of flocculate and equilibrium agglomerate increased with increasing the concentration and the HLB of surface active agents. Finally the micromeritic properties of the agglomerate and the kinematic parameters could be correlated quantitatively with the one parameter, i.e. the interfacial tension between the bridging liquid and the suspending liquid, irrespective of the concentration and the HLB of surface active agents.
The thermal conductivity of the sluge generated at a waste water treatment plant was measured by means of a hot wire method to get information needed for designing effective multi-hearth furnaces. The results are summarized as follows; (1) The relation between the thermal conductivity of the sludge λ (kcal/m.h.°C) and water content Φ (wt%) was expressed by the experimental formula λ=(2.32Φ+60.4)×10-3 (0≤Φ≤20) λ=0.00045Φ1.75 (20<Φ≤60) (2) The effect of the solid volume fraction (φ) on the thermal conductivity of the dried sludge (λ) in the atmosphere of air, nitrogen or carbon monoxide was given by the experimental formula λ=0.25φ+0.015 (kcal/m.h.°C) irrespective of the atmosphere. (3) The thermal conductivity (λ) of the ash bed obtained by burning the sludge increased linearly with increasing temperature T(°C) and was given by the experimental formula λ=(1.37T+955)×10-4 (kcal/m.h.°C) This temperature dependency is due mainly to that of air in the pores of the ash bed. (4) The thermal conductivity of the ash particle was estimated to be about 1kcal/m.h.°C by using de Vries' equation.
The group has carried out a four-year cooperative study on the evaluation of damages at roll surfaces by using the X-ray diffraction technique. The results of the group's activity are described in the present resent report. The half value breadth of X-ray diffraction profiles was found to be a useful parameter for evaluating the surface damage of rolls. Quantitative control of dressing is practically possible by using this parameter.
The ultrasonic wave reflection method has been used to determine the elastic constants of various steels at the temperature range from room temperature to -196°C. Liquid nitrogen was used as a cooling agent. The data of the temperature dependence of elastic moduli for commercial alloys and steels were given in tables with their chemical compositions. Moreover, the data were represented by the empirical formulae in the form of exponential function of temperature. From the results, it can be concluded that both moduli (E, G) tend to increase with decreasing temperature.
According to Coleman's thermodynamics, the elastic and the plastic constitutive equations have been deduced theoretically from the Clausius-Duhem inequality. The modified thermodynamic potential and a scalar internal state variable are introduced and a yield restriction is assumed. The dependence of the internal state variable on the yield function is considered to express the isotropic work-hardening. In the elastic state the rates of stress and temperature are chosen arbitrarily, so that it is possible to obtain naturally the elastic constitutive equations. In the yield state they are restricted by the yield condition, and thus the undefinite plastic parts of strain and the entropy, which are subjected to some restriction, can be obtained. When the material is independent of the internal state variable, it reduces to a perfect plastic material. A special simple case is also discussed.
Description on the kinetics of austenite-pearlite and austenite-martensite transformations in connection with temperature variation and also of the elastic-plastic stresses in the course of quenching of steel was made. The volume fractions of the structures in addition to the plastic strain and the hardening parameter were regarded as internal variables, and an emphasis was placed on the metallurgical considerations on the nucleation and the growth of the phases to obtain the evolutional equations of the variables. Finite element formulations based on the weighted residual method and the principle of virtual work were presented for the analyses of the heat conduction with phase transformations and the stress and strain, respectively. As an example, the theory was applied on a solid cylinder of carbon steel cooled from above its eutectoid temperature. Several kinds of experiments were carried out on temperature change and the variation in the structural distribution was observed by microscopy and the residual stresses were measured by the Sachs' method and by X-ray diffraction technique to verify the present theory.
Fatigue crack propagation by periodic overstressing, i. e. by intermittent application of a very small number of cycling of overstress during a very large number of cycling of understress below Kth, was studied with plate specimens under the in-plane reversed bending stress condition. This condition is a more sophisticated test condition from the standpoint of fracture mechanics than the rotating bending test previously reported. Remarkable acceleration of crack propagation amounting to more than one hundred times was observed in the present study as in the previous ones, indicating the validity of the previous results. Crack closure levels were not affected appreciably by the periodic overstressing in most of the tests. An extra-ordinarily well developed dislocation cell structure was observed in thin foil specimens containing a fatigue crack by means of ultra-high voltage electron microscopy. The acceleration could be attributed not to mechanical causes such as crack closure but to microscopic processes such as the change of dislocation structure observed in the present study or the change of crack propagation mechanism revealed by fractography in the previous studies.
Non-propagating microcracks are observed on the surfaces of smooth unbroken specimens loaded more than 107 cycles at or slightly below their fatigue limits under rotating bending fatigue tests of a prestrained 0.17%C steel and under torsional fatigue tests of an annealed 0.54%C steel. In this study, the crack opening displacement at various distances from the tip of a fatigue microcrack was measured accurately on the specimen surface by using electron microscopy, and the opening-and closing-behaviors of the non-propagating microcracks were compared with those of propagating microcracks, in order to investigate the mechanism of non-propagation of fatigue microcracks. The main results obtained are as follows; (1) The opening-and closing-behaviors of non-propagating microcracks observed under rotating bending fatigue is almost the same as the one observed under torsional fatigue. (2) The shape of the rim near the tip of the non-propagating microcrack on the surface of the specimen is very different from that of the propagating microcrack, at the maximum stress in one cycle of the reversed stress under which the respective cracks are formed. The shape of the non-propagating microcrack shows the “cusp” type, while that of the propagating microcrack shows the “blunt” type near the crack tip. The crack opening displacement near the tip of the non-propagating microcracks is much smaller than that of the propagating microcracks formed under the reversed stress whose amplitude is slightly higher than that of fatigue limit. Moreover, the non-propagating microcracks open rarely at or near their tips even under the maximum stress in one cycle of the reversed stress under which the cracks are formed.
It is well known that fatigue failure of metallic materials takes place generally through two processes, crack initiation and its propagation. Therefore, their fatigue life can be divided to the crack initiation life and the crack propagation life. In particular, the crack propagation life takes a critical role for the final fatigue failure of structural members having defects such as sharp notches or fine cracks. In this paper, a statistical study was made on distribution characteristics of the fatigue crack propagation life of metallic materials, and, the results derived analytically were compared with the distribution characteristics of the experimental results. The results obtained are summarized as follows. (1) The parameter of crack propagation rule, C, has a logarithmic normal distribution, so that the crack propagation rate under any crack length is governed by the logarithmic normal distribution. (2) On the assumption that the parameter, m, is constant and C is random variable, the distribution function of fatigue crack propagation life can be analytically derived for any propagation length on the basis of the experimental results of the distribution of the parameter C.
Corrosion fatigue behaviors have been studied on low carbon steels and high tensile steels in 3%NaCl. The corrosion fatigue strength against the number of cycles was estimated by the following formula. σac=σa/(1+B·f-p·Nfp+q) where σa is the fatigue strength tested in air, f is the cyclic speed, Nf is the fatigue life to failure, and B, p, q are the material constants. It was obtained experimentally that the material constants were given as follows; (1) For a plain specimen of SM41B steel and a notched specimen of HT 80 steel, B=2.14×10-4, p=0.39, q=0.36 under plane bending corrosion fatigue tests. (2) For a plain specimens of HT 80, HY 110, and HY 130 steels, B=4.78×10-4, p=0.39, q=0.36 under plane bending corrosion fatigue tests. (3) For a plain specimen of SM41B steel, B=1.08×10-4, p=0.39, q=0.36 under tention and compression corrosion fatigue tests. (4) The material constant B is given by the following function. B=8.7×10-4logσw-1.0×10-3 where σw is the fatigue limit tested in air.
Resistors having different degrees of crosslinking were made by curing carbon black-graft polymers (grafted with n-butyl acrylate and acrylic acid) with an epoxy resin on porcelain plates, in order to obtain some information concerned with the effect of crosslinking on the large positive temperature coefficient of electrical resistivity. For the resistors prepared with the ratio 1/1 of carboxyl group/epoxy group, the gel fraction increased with increasing curing temperature or curing time, and therefore, the crosslinks increased. The volume resistivity decreased with increasing crosslinking and was in the range of 1-8Ωcm. The resistance temperature curves showed a distinct positive anomaly at about 80-100°C attributable to volume expansion of polymer. The variation of resistance at 140°C decreased with increasing crosslinking except for the samples which showed bad thermal stability against heating cycle between 40-140°C. On the other hand, all the resistors made of polymer compositions having different ratios of carboxyl group/epoxy group and cured at 160°C for 2hr exhibited good thermal stability against heating cycle between 40-140°C, though their degrees of crosslinking were different. The least crosslinked resistor showed no anomaly of resistivity. It is considered that rearrangement of the carbon black network may occur toward lowering the resistance by static electric force in less crosslinked resistors.
The stoichiometry of the reactions between polycations (protonated polyethyleneimine, ionene, poly (vinylbenzyltrimethylammonium chloride)) and polyanions (sodium polyacrylate, potassium polystyrenesulfonate) was investigated by conductivity measurements and elemental analyses. It was found that they reacted almost in stoichiometry to give a polyelectrolyte complex. The water vapor adsorption and dielectric properties of the complex were investigated. The complex showed the sigmoid type adsorption behavior similar to the adsorption behavior of a hydrophilic material. The complex exhibited a marked dielectric dispersion over a wide frequency range of 102∼106Hz.
The chemical resistance of unplasticized polyvinyl chloride (HPVC) was investigated in several organic solvents at room temperature by using the static and dynamic test methods, and the results were compared with the data in service. The results showed that the estimation of the chemical resistance of HPVC was rather difficult by the measurements of weight, hardness and tensile properties even after testing for 7∼30 days. On the other hand, the creep and stress relaxation test results gave good agreement with the data in service. From these results, a way for evaluating chemical resistance of HPVC by the stress relaxation test method was established, which gives the result in a relatively short time.