The hydraulic hardening property as well as other properties of water granulated blast furnace slag powder (hereafter refered to as slag) has been widely utilized. Many efforts have been made to fully utilize these properties through particle control of the slag. The usage of an air classifier as the means of manufacturing very fine slag has been suggested and carried out actually. Since the production of very fine slag has become popular and indispensable, the improvement of classifying efficiency is an important and urgent issue. The classifying efficiency of a classifier is remarkably influenced by the degree of dispersion of fed slag. The aim of the present study was to examine the effectiveness of the addition of classifying aids in the classifying process for slag. As the classifying aids, ethanol, diethylene glycol (DEG), and triethanolamine (TEA) were used. The results obtained are as follows: (1) The efficiency of slag classification was considerably improved by the addition of classifying aids; in particular DEG and TEA were effective under the experimental condition employed. (2) The effectiveness of the addition of classifying aids could be evaluated in terms of the dispersability measured by a powder-falling test which was employed in this study. (3) The optimum dosage of classifying aids was almost same as that of grinding aid. However, the functions of the both aids were slightly different.
Natural Itaya and Nakano zeolites were treated with hydrochloric acid for 1hr at 100°C. The original zeolites and the treated ones were characterized by X-ray diffraction, chemical analysis, thermo-gravimetric analysis and adsorption. The X-ray diffraction patterns of the zeolite treated with 1mol·dm-3 acid and those of the untreated zeolite were almost identical, but the ratio of SiO2 to Al2O3 was different. The zeolite treated with acid rapidly increased its inner adsorption volume. The monolayer capacity of nitrogen gas rapidly increased as the amount of cations in the zeolite decreased. The effect of hydrochloric acid treatment on gas separation properties of natural Itaya and Nakano zeolites was investigated. Hydrogen, oxygen, nitrogen, methane and carbon monoxide were completely separated on the acid treated Itaya zeolite column, but not on the natural Nakano or its acid treated zeolite column.
It was presented in the previous paper that the simulation model of mixing in a two dimensional particle system was not so effective as to estimate the state of actual mixing process in a three dimensional particles system. This is because the area of mixing zone was determined on the basis of the results obtained from the photographic observation of only the upper surface of mixer during the actual mixing process. The purpose of this paper was to investigate how to derive a simulation model applicable to the mixing in a three dimensional particle system. The area of mixing zone in the three dimensional particle system was determined by introducing a representing equation where the area of mixing zone decreases with the depth of a plane to which the mixing zone belongs. Furthermore, the size of three dimensional model system was determined on the basis of the size of particles and mixer used. The electric resistance of a particle was its observed value. The conclusions obtained are as follows; (1) The validity of the simulation model derived was confirmed by comparing the parameters (σ, R) of the actual mixing process with those of the simulation model. (2) The size of mixing zone decreased exponentially with the depth of a plane to which the mixing zone belongs. (3) The area of mixing zone has an upper limit which varied with the change in length of a blade. The correlation coefficient between σc and Rs was greater than 0.8, when the upper limit area of the mixing zone was given; where σc is the standard deviation calculated in simulation experiments, and Rs the total electric resistance calculated by applying the approximation method.
In order to simplify the calculation of the total electric resistance of a binary particle system of two dimensional f·c·c (face centered cubic) structure, we propose a conventional method of transforming a lattice from the f·c·c to the square. This transformation is practiced by the following procedure. At first, an original f·c·c particle system whose total electric resistance should be calculated is rotated in an angle π/4 as shown in Fig. 1(a), (b). Then, with the addition of four triangles to the slant f·c·c system of Fig. 1(b), it is possible to complete the system of Fig. 1(c) whereby the total electric resistance may be easily calculated. The results obtained are summarized as follows; (1) The calculation of the total electric resistance of a f·c·c system can be symplified by using the method of transformation described above. (2) The edge effect of the mixing zone of a f·c·c structure is greater than that of the two dimensional square system. (3) There is a high correlation (correlation coefficient<0.94) between the total electric resistance of the system PQRS and the standard deviation of mixing of the system ABCD (Fig. 1(c)).
There is an urgent need to establish a mold for dental precision casting of pure titanium which has an excellent biocompatibility. The present authors have already developed a useful mold composed of refractory ceramic particles as a material for the aggregate and metal particles as a material for controlling expansion due to oxidation during heating. In order to design a mold having much more excellent dimensional precision and favorable other characteristics for quality control of the casting, it is necessary to analyze fundamentally the relationship between the amount of expansion of mold and other factors, such as the void fraction, content of expansive particles, radius ratio of metal and aggregate particles, and binding strength between particles. The present investigation was undertaken to develop two kinds of personal computer programs for simulation of (i) a packing of different sized circles under a gravitational falling condition and (ii) an expansion due to specific expanding circles in such a packed bed as (i). For the simulation (ii) process, the two dimensional finite element method was used. Since there is a good correlation between the results from the present trial simulation and the experimental data, it will be possible in the future to analyze the expansion mechanism of the mold in detail, by applying these developed programs.
Fracture toughness testing was carried out on solid wood, heat-treated wood and polyethylene glycol (PEG) impregnated wood. The main results are summarized as follows: (1) The load-displacement curves had very different patterns for solid wood, heat-treated wood and PEG impregnated wood. (2) The fracture toughness value of the heat-treated wood was equal or larger than that of the solid wood. (3) The fracture toughness value of the PEG impregnated wood was larger than those of the solid wood and the heat-treated wood. (4) The Young's modulus of the heat-treated wood was larger than that of the solid wood. (5) The Young's modulus of the PEG impregnated wood was smaller than that of the heat-treated wood. (6) The cumulative acoustic emission energy of PEG impregnated wood was very much smaller than those of the solid wood and heat-treated wood
The relation between the macroscopic plastic response and the corresponding grain scale plastic events for a polycrystalline aggregate was studied by using the finite element method. The numerical model used was based on the constitutive relation combining a few simple and fundamental semi-microscopic plastic mechanisms which can describe a large variety of plastic behaviors observed in lots of experiments. The numerically evaluated plastic accommodation parameter, or the principle and representative scalar function, which relates the microscopic mechanisms on the grain scale with the macroscopic behavior, was compared with those suggested by other theories accepted in polycrystalline plasticity.
This paper deals with the stress concentration analysis of weld reinforcements under tensile load, and the optimum weld condition for decreasing the stress concentration factor (SCF). SCF was systematically calculated using a numerical analysis with various geometrical parameters. The stress field induced by a distributed force on a segment in an infinite plate was used as a fundamental solution to solove this problem. In the calculated range of the geometrical parameters, SCF and the optimum weld condition were found to be as follows. (1) SCF is almost constant when a flank angle θ is over 60°, when h/t is over 0.2, or when w/t is over 1.0; where h is the height of the reinforcement, w is the width of the reinforcement, t is the thickness of the plate. (2) SCF is almost proportional to the square root of t/ρ, where ρ is the radius of the notch. (3) In pulse arc welding, the best welding condition for decreasing SCF is the one to controll θ except ρ.
The distribution of fatigue crack propagation life with retardation was investigated by the use of a Markov approximation method. By supposing the situation where some overloads are superimposed on cyclic loading process, the probability distribution of total delay time was first derived according to the Arone's phenomenological method. Next, it was applied to revise the residual life distribution function in order to take into account the effect of overloads, under the condition that both stress amplitude and crack propagating resistance were random. Finally, with the aid of numerical calculations, it was clarified that the effect of retardation due to overloads was insignificant in the range of high reliability.
Generally, for glass fiber reinforced plastics (GFRP) in acid environments, the relationship between the crack propagation rate da/dt and the stress intensity factor at a crack tip K is expressed experimentally in the form of da/dt=AKm where A and m are constants. In this paper, discussion was focused on the theoretical examination of the constant m. The model of crack propagation due to stress corrosion cracking in GFRP proposed by Hull et al. was modified by taking account of the existence of microcracks on the surface of glass fibers. On the basis of this modified model, a micromechanics study was made. It was possible to obtain the value of m which is appropriate to the experimental value.
Based on the maximum tensile stress criterion for the cleavage fracture of steel, correlation between fracture toughness and yield stress, or cleavage fracture stress has been investigated on 13 different types of steel. By incorporating micromechanistic criteria of failure to analytical or numerical crack tip stress and strain solutions, known as the Neuber-RKR (Ritchie, Knott and Rice) model, a certain relation can be derived between fracture toughness and flow/fracture properties of materials. In the previous work, a good experimental correlation according to analytical prediction was obtained. The generality and limitation of the correlation, and the specimen size effect on fracture toughness are discussed in the present work. From the results of the previous and present works, analytical description of fracture toughness expressed in terms of tensile properties is confirmed to be valid and a possibility of the prediction of fracture toughness from the conventional round bar tensile test is shown.
Fracture toughness tests of high speed steels were carried out to develop a method for evaluating fracture toughness. A precrack introduced by the BC (Bridge Compression) method, at the front of which the damage region with microcracks was formed, could not be assumed as an ideal crack. On the other hand, a precrack introduced by the FP (Fatigue Precracking) method, where the precrack was introduced under a sufficiently low level of Kfmax, could be assumed as an ideal crack. The KIC-values evaluated by the BC method were lower than those by the FP method. Therefore, it is necessary to introduce a fatigue precrack, which can be assumed an ideal crack, by the FP method for evaluating fracture toughness. The size and volume fraction of carbides in the sintered high speed steel were smaller and higher than those in the melted high speed steel. The high volume fraction of carbides gave low fracture toughness, while the small size of carbides improved fracture toughness. Since in the present sintered high speed steel these effects by the size and volume fraction of carbides compensated each other, the fracture toughness of the sintered high speed steel was almost equal to that of the melted high speed steel.
The measurements of fatigue crack growth characteristics at room and liquid nitrogen temperatures as well as the observations of microstructure and fracture surface topography of metastable 304L were carried out. The results obtained were compared with those of stable austenitic stainless steels (310 and high Mn stainless steel). The effective stress range ratio of 304L steel at liquid nitrogen temperature was larger than that at room temperature or that of 310 steel. The blocks of α' martensite transformed from austenite via ε martensite were observed in the close neighborhood of cracks in 304L steel, whereas no α' was found in the high Mn stainless steel. The fracture surface of 304L steel at liquid nitrogen temperature was smooth and flat. Based on these observations, the fatigue crack growth mechanisms in stainless steels at cryogenic temperatures were briefly discussed with emphasis on the crack closure.
In order to clarify the effect of vibratory stress of high stress ratio on the crack propagation of sintered silicon nitride, the crack propagation rate was measured on its compact tension specimen. The main results obtained are as follows. (1) Under the cyclic load with a stress ratio smaller than 0.92, the crack propagation rate was accelerated by the cyclic load compared to the crack propagation rate under the static load. The cyclic load with a stress ratio larger than 0.92 produced the same crack propagation rate as that under static load. (2) In case of KImax smaller than KISCC, the crack propagated mainly cycle-dependently. Therefore, care must be taken in case of the fatigue design of ceramics under the vibrational stress. On the other hand, the crack propagated mainly time-dependently when KImax was larger than KISCC. (3) Debris induced crack-closure was observed under the cyclic load even with a large stress ratio of 0.8. (4) In contrast to the two different behaviors of metal under cyclic SCC and under dynamic SCC, the crack propagation behavior of sintered silicon nitride under a cyclic load of high stress ratio and high frequency was same as that under a cyclic load of low stress ratio and low frequency.
Fatigue fracture behavior of the smooth specimens of polycarbonate was investigated in the range of low cyclic stress. In such fatigue tests, specimens fail during the process of crack initiation and propagation like metal. An experimental program was set up to examine the process of initiation and growth of a surface crack. This was accomplished by careful observations by using the transmitted light photographs of a surface crack on the smooth specimen of polycarbonate during the rotating bending fatigue test at the speed of 2Hz at room temperature. It was shown that the initiation of the surface crack appeared from a very small inclusion near the surface of specimen at 10-30% of the fatigue life. The experiment showed that the surface crack propagation was hindered by the shear bands formed at the both ends of a crack. The final length of surface crack was discussed in terms of a combination of the stress amplitude and the stress intensity factor at the end of crack. From these results, it became evident that the characteristic behavior of the crack initiation and growth in the smooth specimen of polycarbonate was affected significantly by the mechanical properties and yield behavior of the material.
The computer image processing system incorporated into the specially designed servo-hydraulic fatigue loading facilities operating in a scanning electron microscope (SEM) has been developed. The image processing system developed is composed of a micro-computer (NEC PC9801, memory capacity is 4 Mbytes) and a high-speed direct memory access type A/D converter (ELMEC EC-2390). The secondary electron signal of SEM was directly transferred to the memory of micro-computer through the converter at a digitizing rate of 5 to 50μsec per point. By using this system, the sampling of successive five flames of SEM image data during fatigue loading of 0.1Hz become possible. Direct, real time observations of fatigue crack growth behavior under both constant amplitude and repeated two-step loadings were made. Both the crack profile and the contours of fine magnesium oxide particles, which were scattered on the specimen surface as location markers, were extracted from the original SEM image data by using the image processing technique, and this processed image data was used in quantitative analyses of crack opening displacement, deformation behavior near the crack tip and crack growth increment.
The mechanical properties of yttria-doped tetragonal zirconia polycrystals (Y-TZP) are known to be degraded by aging at 100∼300°C in air due to tetragonal to monoclinic phase transformation. In the present study, hot-water aging tests were done for several Y-TZP specimens having different grain size which were made from raw powders with different specific surface area. The tetragonal to monoclinic phase transformation zone depth from the surface after the hot-water aging was measured by a thin film X-ray diffraction method, by which the small amount of transformed monoclinic phase existing very close to the surface can be detected. Y-TZP with the small grain size of 0.15μm, made by firing at 1300°C for 3 hours using zirconia raw powder with specific surface area of 30m2/g, showed only a small amount of transformed monoclinic phase even by 200°C hot-water aging for 100 hours. Furthermore, its phase transformation depth was only 2.3μm and the depth did not increase even after a more prolonged aging time.
Austenitic stainless steels are extensively used for high temperature components, such as heater tubes and boiler tubes in chemical and electric power generation plants. It is well known that, for these materials, the metallurgical degradation occurs at an elevated temperature which is caused by the nucleation and growth of carbides and intermetallic phases. In this study, such metallurgical characteristics of the degraded SUS321 steel and SUS316 steel for a long term in service were investigated, and the non-destructive evaluation of the material deterioration was performed by means of an electro-chemical method. The results obtained are summarized as follows: (1) The microstructural change after a long term exposure in a high temperature environment for SUS321 steel is the nucleation and growth or the decrease of TiC, M23C6 type carbide and σ phase. These precipitates except TiC cause the material degradation. (2) These precipitates have active dissolution peaks at a particular potential independent each other in the anodic polarization measurement in 1N·KOH solution. (3) The metallurgical damage by the precipitation can be estimated in terms of the active dissolution current density (IP value), since the amount of the precipitate has a good relation with IP value at the specific potential obtained by the anodic polarization measurement. (4) The material degradation occuring on SUS316 stainless steel in a high temperature exposure can also be evaluated.