Effects of environmental conditions on the stress corrosion cracking (SCC) of austenitic stainless steels SUS 304 and SUS 316 were studied. Because of unusual cracking due to pitting corrosion, SUS 316 has far higher SCC susceptibility than does SUS 304 in boiling 42 and 38% MgCl2 solution, whereas it has relatively lower SCC susceptibility at lower temperature and concentration of test solution. Threshold stresses for SCC, obtained by the constant-total strain test method, increased proportionally with the temperature difference between boiling and test temperature of each MgCl2 corrodent within the range from 30 to 42 wt %. No cracking was observed below the test temperature of 60°C for SUS 316 and 40°C for SUS 304. Master crack curve employing either Larson-Miller parameter or shift factor is proposed for the estimation of crack initiation time at service temperature of process equipment. Effects of environmental conditions on the crack pattern and morphology such as macrobranching and cracking with pitting corrosion are also discussed.
The flow patterns due to confluence in a two dimensional right-angled T-shaped flow section were obtained experimentally. The confluence loss was evaluated from momentum theory and the effect of downstream flow on energy loss was investigated. The results were as follows : The loss coefficients calculated by momentum theory were quite small compared with experimental results, and the calculated values of η13 (=loss coefficient for main stream) did not change very much, even if m (=sectional area ratio) was changed. The incomplete point of momentum theory was improved by considering the effect of downstream flow.
The effect of bubble turbulence on the performance of a bubble turbulence-type electrodialyzer is studied by assuming two-dimensional flow. As a result, it is clarified that limiting current density per unit conductivity, ic/λw [V/cm], is directly proportional to loss of pressure per unit length of channel, H [G/cm2/cm], without regard to the structure of the dialysis cell, distance between membranes or state of flow. Furthermore, it is clarified that the bubble turbulence effect is expressed as an amount of increasing head loss, ΔH, caused by form drag of bubbles.
The unsteady characteristics of heat transfer in an optically thick semitransparent liquid layer is deduced theoretically and demonstrated experimentally for the case in which the liquid layer is heated stepwise from the upper side by a radiant heat source. The basic differential equation controlling the temperature field in the liquid layer is introduced and solved by using two models : Multiband model and Planck mean model. The theoretical results calculated from both the Multiband model (calc. I) and those from the Planck mean model (calc. II) for the time-change of liquid temperature distribution within the layer are in fairly good agreement with experimental data obtained by using an infrared lamp bundle as the radiant heat source and diethylphthalate as the liquid sample. Comparing “calc. I” with “calc. II”, however, “calc. I” shows a better coincidence with the experimental data than does “calc. II”.
Scale formation on the heat transfer tubes in a multi-stage flash evaporation-type desalination plant was examined by using a 10-stage flash evaporator with a capacity of 100 m3/d. In a 100-hour continuous operation test, sulfate scale did not precipitate, but a large quantity of alkaline scales formed on the surface of the tubes in the brine heater and the condensers of the evaporator. The major component of the alkaline scales was calcium carbonate at the low-temperature section and magnesium hydroxide at the high-temperature section. Formation of alkaline scales is related to the dissociation equilibrium of carbonic acid in the brine. In submerged tube-type evaporators, the brine evaporates on the surface of the tubes. On the other hand, in flash evaporators the brine does not evaporate in the tubes. Therefore the behavior of carbonic acid in the brine in the flash evaporator differs from that in the submerged tube-type evaporator. As the results of calculation of the mass balance of carbonic acid in the flash evaporator using scale deposition rate data crystallization rate coefficients, kR, for calcium carbonate and magnesium hydroxide were obtained as a function of brine temperature.
Experiments on vaporization of binary solutions into a turbulent air stream at atmospheric pressure and room temperature were carried out in a wetted-wall column. When the concentration of diffusive components is dilute, the diagonalization procedure proposed earlier was found not always to be usable under some conditions. The applicable conditions of the above procedure were confirmed experimentally. Approximate equations of mass transfer rates were proposed, utilizing a linearized method. The validity of the equation and the conditions under which it can be applied were then ascertained experimentally. The conditions under which the mass transfer rates of two diffusive components in ternary gas systems can be dealt with as the diffusion of one gas in binary systems were ascertained, using experimental data.
This paper is concerned with a new development in extraction. The multiple-solvent extraction proposed here has a new contact style of solvent. In an apparatus, the solute in a diluent is separated in series with more than two immiscible liquids, and thus in this operation extraction and back-extraction occur simultaneously. In this paper, the concept of single-stage, multiple-solvent extraction is explained and the calculation method is shown. The dual-solvent extraction, which is the simplest system of multiple-solvent extraction, is compared with existing extraction methods. Experiments are also carried out, in which acetone in carbon tetrachloride is transferred to water phase and then to benzene phase in series. It can be said that this extraction method enables us to decrease greatly the solute concentration in the raffinate phase and to increase it greatly in the final extract phase. It can be concluded that this method is more efficient than previous extraction methods.
Weight distributions of particle beds on the bottom of a container are obtained from a computer simulation. Three typical types of beds of spherical particles are dealt with : beds formed by particle feeding from the center of the cylinder (a) without and (b) with the wall effect, and beds (c) formed by particle feeding from the periphery. This paper deals with vertical forces only; the transverse forces acting on each sphere are neglected except for the spheres in contact with the wall. In each of these beds the particle number density is uniform in the radial direction of the container. Microscopically, however, the packing structure is found to be formed such that the weight of each particle spreads outwards from the feeding position. That is, the anisotropy in the packing structure is found to play an important role in the weight distributions of the particle beds on the bottom of the container.
This paper deals with the optimum conditions in plug-type pneumatic conveying of granular materials through a pipe. First, taking a force balance in the pipe, a theoretical equation for air pressure required to convey a plug is developed. Second, the values calculated by the equation are confirmed to be in good agreement with the data obtained in experiments performed with horizontal and vertical pipes of diameters 66, 78 and 99 mm using calcium carbonate powder as granular materials. In conclusion, the optimum operating conditions of plug-type pneumatic conveying of granular materials are briefly discussed.
The density, distribution and maximum diameter of the floc particles of cellulose coagulated in a stirred tank were obtained. It was found that (1) there existed a hysteresis phenomenon in floc density, size distribution and maximum diameter with respect to agitation rate, and that (2) the maximum diameter increased with time, reached a maximum value at a point, and then settled down at a smaller constant value. The flocculation processes were discussed by the following four models : (1) coalescence model, (2) dispersion model, (3) breakup model and (4) mechanical syneresis model.
When colloidal solutions containing such as protein are separated and thickened by ultrafiltration, colloidal particles thickened on the filtration membrane become gelatinous and show very large filtration resistance. Electro-ultrafiltration is effective in decreasing gel formation and in increasing filtration flux in the separation of particles from colloidal solutions. But few studies of the characteristics of electro-ultrafiltration and design of electro-ultrafiltration equipment have been carried out, and therefore electro-ultrafiltration has scarcely been used in industry. Electro-ultrafiltration was analysed considering three periods : period of growing concentration polarization, period of forming gel layer and period of constant filtration flux. The results of analyses were experimentally examined by using colloidal solution of bovine serum albumin. Furthermore, the relation between filtration flux in the steady period and electric power consumption was experimentally investigated. The results show that electro-ultrafiltration has excellent characteristics compared with ultrafiltration in the case of gel formation on the filtration membrane.
Fault Tree Analysis (FTA) can be evaluated either qualitatively or quantitatively depending on the scope, extensiveness, and use of the analysis. The purpose of this study is to ascertain the necessity of the quantitative hazard control method by the system using FTA, by actually computing the system failure probability and probabilistic importance of the basic event of the FT model for all operation time. To generalize, exponential distribution and Weibull distribution models are treated in this study, and the robustness of the failure law as a measure of their importance is examined. Use of sound failure data was necessary for such study. Not only in the design stage but also in the operation stage of the system, the system failure probability and the probabilistic importance of the basic event for all time are actually calculated, and safety management using FTA in the operation stage of the system is proposed. This proposed method of safety management is calculated with an average computer time on the order of a second for a fault tree of about ten logic gates.
The pyrolysis of surplus activated sludge was investigated experimentally by means of a batch reaction furnace. The pyrolysis temperatures were varied from 400 to 1, 200°C. The effect of pyrolysis temperatures on the yields of produced gas, liquor, tar and char as well as the their constituents was studied. As a result, the following conclusions were drawn for the range of experimental conditions employed : 1) Pyrolysis of the activated sludge starts to occur at around 400°C. 2) Total gas yield increases approximately proportionally to the square of pyrolysis temperature, and main constituents of gas are H2, CO, CH4 and CO2. The heating value of gas produced at temperatures above 1, 000°C is about 4, 000 kcal/ Nm3. 3) The maximum yield of tar is attained at between 600 and 800°C. 4) At above 600°C, the char yield is scarcely affected by pyrolysis temperature, and is kept at a constant value of about 48 wt% (dry sludge basis). The heating value of char is between 1, 650 and 2, 200 kcal/kg-char.