The effects of casting conditions of cellulose acetate membrane (Batch-47 type) on its reverse osmosis properties have been investigated by x-ray diffraction, electron microscopy, water vapor sorption and reverse osmosis performance. It was found that the membrane consisted of a thin dense layer and a porous substructure. The variation of diffusion coefficient of water calculated from the sorption data corresponded to the change of water flux by a reverse osmosis. This indicates that as soon as a casting solvent is lost from the membrane surface, the dense layer is formed. The membrane annealed at about 75°C showed a steep increase in salt rejection but a decrease in water flux. Some changes in the compaction effect and membrane shrinkage were noted in the same temperature range as described above. X-ray diffraction and deuteration methods revealed no distinct increase in crystallinity upon annealing (40-80°C). This may be due to the pore size reduction on the surface accompanied by the shrinkage and also to the increase of degree crystallinity in the dense layer.
Concentration polarization effect has been investigated for a batch type reverse osmosis cell with the membranes prepared in 3% NaCl solution at various annealing temperatures. It was found that this effect became stronger for a higher water flux membrane. Effects of salt-gelation on the reverse osmosis properties of cellulose acetate membrane have been investigated also. The membrane shrinkage increased with salt concentration of gelation medium. Correspondingly the salt rejection increased but the water flux decreased. The salt-gelation effect was in accord with the order in the lyotropic series. No evidence of asymmetric character of the reverse osmosis has been observed for the salt-gelled membrane.
The effects of membrane preparation variables on the reverse osmosis properties of cellulose acetate membrane have been investigated by water vapor sorption. Dependence of isotherm on the annealing temperature was not observed in casting conditions of one minute solvent evaporation and of annealing temperature of 60-80°C. This may indicate that the primary membrane structure is formed in the evaporation step and the dense layer is thin. The membrane prepared from Eastman 398-3 was more porous than one prepared from Wako Pure Chemicals. Reverse osmosis properties of cellulose membrane was discussed, based on the selective water adsorption on cellulose in the amorphous region, and a crevasse-like structure model was proposed. By using this model, a good agreement was obtained between the experimental and the calculated data.
The process of crack propagation was continuously studied by extending thin foils of pure Al with pre-existing cell structures in an electron microscope. The dislocation structures near the fractured part of pure Al sheet specimens of 100μ thickness were also observed. The results are summarized as follows; (1) When a crack propagates through the cell structures, thinning occurs in the cells lying in the direction of crack propagation due to the dislocations generated from the crack tip. However, pre-existing cell walls still remain in the thinned zone and become sharpened, and finally these cell walls attain the shape of low angle boundaries. Dislocations moving in the thinned zone are stopped at both sides of the zone and the cell walls with high dislocation density are formed there. (2) The possible mechanisms of crack propagation in the thinned zone, in which the sharp boundaries and the dense dislocation arrays are remaining, are as follows; (i) shear rupture by the generation of dislocations from the crack tip, (ii) coalescence of the microcracks with the main crack, and (iii) failure along the sharp boundary. (3) Fracture teeth of less than several thousands angstroms thick are formed at the fractured parts of the sheet specimen. Well-defined low angle boundaries remain in the fracture teeth. In the still thicker parts near the root of fracture teeth, dislocation tangles or fine cell structures exist. The processes of formation of these structures can be explained from the direct observations of crack propagation.
In order to obtain the fundamental knowledge in clarifying thermal stress induced in the stator windings, the interaction mechanism between the conductor and insulation of the stator windings under axial load were examined theoretically and experimentally. The stress analysis was made with the aid of the equivalent model in which the stator windings was regarded as a composite structure consisting of the conductor and the insulation bonded by the adhesive. The experiments were carried out on six different kinds of model windings and the strain distributions induced on the insulation surface were measured by strain gauges under axial load. The electrical properties, such as tangent delta and breakdown voltage, were also measured under axial load and the variations in the electrical properties caused by the occurrence of mechanical damages were examined. The results obtained are as follows (1) The interaction mechanism between the conductor and insulation of the stator windings followed the derivation shown in the stress analysis. (2) When the axial load was applied, the tensile stress in the insulation was maximum at the center portion and the shearing stress in the adhesive was maximum at the end portion. Therefore, it was considered that the insulation windings should bear not only the tensile strength of insulation but also the shearing strength of the adhesive or of the insulation in the vicinity of the adhesive. (3) The closed relationship existed between the occurrence of mechanical damages and the variations in the electrical properties. That is, the apparent variations were observed in the tangent delta characteristics when bonding failures took place in the adhesive and the remarkable reductions were observed in breakdown voltages when the insulation was broken.
When a concrete specimen is subjected to repeated loading or sustained loading, its strength does not always decrease but sometimes increases. It seems likely that the increase or decrease of strength of concrete results from its internal microcracking. The main purpose of the present paper is to verify this presumption. The compressive strength of concrete specimens after a repeated loading, and the microcracks developed due to the repeated loading were studied. The major repeated load levels studied were 44, 59, and 74% of the ultimate strength, The numbers of load cycles were 0, 2, 4, and 15. The microcracks developed in the concrete by the repeated loading were divided into four types, viz., the paste, bond, aggregate, and mortar cracks. The paste cracks were tiny microcracks which developed in cement matrix. They were examined with a microscope using 150X. The three microcracks other than the paste cracks were investigated with a stereomicroscope using 10X. The test results obtained are as follows. (1) With increasing number of load cycles, the strength of concrete fluctuates rhythmically, (2) Whether the strength increases or decreases after the repeated loading depends on the microcrackings, and (3) The paste cracks raise the strength of concrete but the other microcracks, i.e., bond, aggregate, and mortar cracks, lower it.
Though some investigations have been done to clarify the corrosion fatigue strength of steels, there is few investigation ever made on their corrosion fatigue damage. In this paper some results of the corrosion fatigue test and the measurement of electropotentials are described to discuss the corrosion fatigue damage of 0.82%C piano wire. From the results, the following conclusions were obtained; (1) Piano wire of 0.82%C is heavily damaged within the first 20% of the total life. (2) The progress of the damage under the corrosion fatigue process is devided into three stages; the first stage is the formation of corrosion pits, the second stage is the progression of some pits into cracks, and the third is a stage in which one of the crack propagates causing the specimen to break. (3) The number of cycles at which the electrode-potential reaches its maximum value, shows the end of the first stage. (4) The relation among the stress amplitude σ, the number of cycles to failure N and the number of cycles at maximum electrode-potential Np is as follows. log10σ=-0.65log10Np+5.2 log10N=2.32log10Np-7.1
In order to investigate the notch effect on creep-fatigue interaction of metallic materials at elevated temperatures, three types of tests of static tensile creep, zero to tension low cycle fatigue and fatigue with the hold time under a constant tensile stress were performed. For the tests, three types of notched cylindrical specimens with the stress concentration factors of 2.0, 4.8 and 6.4 were used. Super alloy steel_A286 was used and the tests were performed at 650°C. From the experiments, the following conclusions were derived. (1) Different notch effects on the rupture time of the material were observed in the three types of tests. In the creep test, the rupture time ratio of the notched specimen to the smooth specimen was greater than unity for all of the elastic stress concentration factor Kt employed and it took the maximum value at Kt of nearly 4. On the other hand, in the fatigue test, the rupture time ratio was smaller than unity for all Kt employed and the decreasing rate became smaller at a larger Kt. In the fatigue test with the hold time, the rupture time ratio was partly greater than unity at Kt of nearly 2 and decreased monotonously to the value of smaller than unity with increasing value of Kt. (2) From the calculation obtained by applying the cummulative damage rule to the rupture life of materials in the fatigue test with the hold time, the combined damage φ of both creep damage φc and fatigue damage φf of the material was apparently dependent on the elastic stress concentration factor Kt and the fatigue damage φf was predominant in the condition of a larger Kt.
Studies on the synthetic thermosetting resins obtained by the reaction of p-sulfoamide-benzamide (SABA) and terephthalamide (TA) with formaldehyde (FA) were carried out. The results were as follows. (1) Colorless and transparent SABA resins were synthesized provided that the reaction temperature was more than 100°C and the molar ratio of FA to SABA was not less than 2.5 at the beginning of the reaction. When a proper quantity of ammonia or sodium hydroxide as a catalyst was added under the above mentioned condition, the similar resins were also obtained. It seemed that the molecular weight of resins tended to be larger in the presence of ammonia, but smaller in the presence of sodium hydroxide than that of resins without any catalysts. (2) Colorless and transparent TA resins were synthesized provided that the reaction temperature was more than 100°C and the molar ratio of FA to TA was not less than 7.5 at the beginning of the reaction, and that a proper quantity of ammonia was added as a catalyst. (3) It was found that when the reaction temperature was higher than 140°C at SABA-FA or 150°C at TA-FA system, a hard, infusible and insoluble resin was formed due to relatively larger rate of the reaction. (4) It was concluded that the relations between molecular weight and softening point of the resins, between molecular weight and reduced viscosity, and between softening point and reduced viscosity were linear.
The use of electronic computers has brought about many structural analysis methods which can be applied to structural analysis problems in highly complicated conditions. The Tensor Code is one of these methods and has often been used for the transient dynamic structural analysis. The principle of the Tensor Code is to apply the equation of motion expressed by Lagrangian co-ordinates to the stress field in the equation by means of difference approximation. In this study, a three dimensional dynamic structural analysis method was developed based on the principle of the Tensor Code, and the outline of the method is described in this paper.