The previous measures used to evaluate quantitatively fluid mixing states have defined mixedness with only a few parameters. However, these measures can not to express the characteristics of a mixing pattern that shows complicated changes throughout the mixing process. In this study, statistical thermodynamic formalism based on escort probability measure is applied to analyze fluid mixing patterns, and the mixing indices are defined as spectra. These spectra enable us to describe comprehensively the characteristics of mixing patterns from the viewpoints of information, theory multifractal theory and thermodynamics. Many of the indices previously proposed to express the degree of mixedness are included as special cases of these spectra. The spectra proposed in this study represent an improvement of the traditional concepts of degree of mixedness.
Mixing and reaction processes of two reactive liquids in glycerin contained in a cylindrical vessel with an eccentric cylindrical rod were experimentally investigated. The eccentricity, defined as the distance between the centers of the vessel and the rod normalized by the difference between the radii of the vessel and the rod was varied between 0.1 and 0.7. The cylindrical vessel and the cylindrical rod were rotated alternately was keeping the Stokes flow condition. Ferric nitrate (Fe(NO3)3) and potassium ferrocyanide (K4[Fe(CN)6]) were used as reactants, resulting in the following reaction : Fe3+ + K4[Fe(CN)6] → KFe[Fe(CN)6] + 3K+ Since the product (KFe[Fe(CN)6]) is dark blue in color, the region where it is present can be identified. This reaction is so fast compared to the flow motion that it occurs as soon as two reactants mix in a molecular scale. The heat of reaction is small enough to treat the chemical reaction as passive with regard to the flow motion. The chemical reaction is significantly enhanced in the chaotic region and the reaction region gradually spreads over the chaotic region, regardless of the initial location of reactants. The chemical reaction is much enhanced when reactants are placed in the region where the stretching rate of elements is much larger.
The velocity field and flow patterns in a fully baffled cylindrical flat-bottomed tank with a Rushton turbine impeller were investigated by the spectral analysis of the velocity time-series measured with two-dimensional laser doppler velocimetry (LDV) and by laser sheet flow visualization. The dynamical behavior in the range near the turbine revolution to the range related to trailing vortices and Kolomogorov wave numbers was shown to be induced by nonlinear coupling between at least three frequencies, the blade passage frequency fp, the turbine revolution frequency fp/6 and the macro-instabilities frequency fu(fMI) due to the change in the eddies or vortices (movement of the center and the size) near the turbine impeller. The dynamic analysis using the Fourier spectrum under different conditions of off-bottom clearance supported the finding that the macro-instabilities frequency fu(fMI) is a component of velocity field dynamics.
The rise characteristics of bubbles, such as rise velocity and chord length, in a liquid are measured using fiber-optic probes. The optical probe consists of three plastic optical fibers of identical diameter, 250-µ m OD, of which one is used as the probe tip and the other two are used as transmitting lines of incident and reflected lights. A pair of probes with different vertical tip locations (by ~1 mm) is utilized to register the time difference resulting from the bubble rise as well as a pair of bubble-piercing times, or loosely, the bubble residence time. These two types of times and the inter-probe distance are, in turn, used to estimate the bubble rise characteristics. Combined with high-speed imaging, it is found that the chord length or local "bubble height" can be estimated with reasonable accuracy when using the average bubble rise velocity measured based on the image data. When only the probe data are utilized, however, the local bubble rise velocity is overestimated— under certain bubble-piercing conditions— by as much as fivefold. The main reason for this discrepancy is found to be bubble deformation, especially near the bubble edge, triggered by the bubble-probe incidence and prolonged during the piercing of the rising bubble by the probe tip.
A technique to apply a fluidized bed of electrically conductive beads is proposed to improve the uniformity of the electric field intensity in microwave applicators, which is required for uniform heating of media. The principle of this effect lies in the dynamic control of random reflections of microwave due to the motion of the conductive beads in the bed. The electrically conductive beads were prepared by wrapping aluminum foil around styrene foam balls and cellophane tape above it. The diameter and density of the beads were 13 mm and 123 kg/m3 respectively. The minimum fluidization velocity of the beads agreed with that predicted by to Wen-Yu equation when the distributor was a porous plate. However, fluidization of the beads took place at lower gas rate than the minimum fluidization velocity due to a jet flow from the small pores for the distributor of a perforate plate. The intensity in the applicator of a commercial microwave oven became the most uniform when the beads were fluidized with a uniform hold-up profile along the height of the bed placed in front of the applicator walls. The dynamic effect making the field uniform by fluidization was improved with increases in the area of apparent reflection by the fluidized bed and in the hold-up of beads, and was superior to a conventional stirring blade.
The effect of adsorption of water onto NaMOR (Na+-type mordenite) and HMORs (H+-type mordenites) with various SiO2/Al2O3 ratios on microwave heating was investigated. The adsorption of water markedly influenced the microwave heating of NaMOR with steady feeding of water vapor. For example, in the presence of water vapor, NaMOR was heated more rapidly and to higher temperature (432 K at highest) than under a dry condition. On the other hand, the adsorption showed little effect of water on the microwave heating of HMORs. A TPD experiment indicated that water was more strongly adsorbed on NaMOR than on HMORs. These results show that the strong interaction of water with NaMOR improves the heating rate and the temperature reached in the microwave heating process.
A theoretical model of phase inversion using structural instability of population balance in agitated liquid-liquid dispersion is proposed. The model describes the hysteresis of phase inversion and the effects of stirrer speed and physical properties on phase inversion. Phase inversion experiments were performed on liquid-liquid dispersion systems composed of water and hydrocarbon mixtures. The physical properties of the organic component were varied by mixing the hydrocarbon with carbon tetrachloride. The volume fractions of dispersed phases of various densities, viscosities and interfacial tensions were measured and found to show good agreement with those estimated by the model.
A modified compartmentalized reactor where the inflow of substrate and the outflow of product can be separately controlled was devised for studying self-sustained oscillations of an enzyme reaction. This reaction-diffusion system consists of a well-stirred reactor inserted between two membranes and reservoirs. By changing the characteristics of the two membranes, such as their thickness and permeability, the reactor inflow and outflow can be modulated. The reaction involved was a hydrolysis of N-α -benzoyl-L-arginine ethyl ester by papain, a proteolytic enzyme. The regions in the parameter space where self-sustained oscillations can be induced were determined via numerical simulation. The mechanism of the oscillations was also discussed. As a result, we found that with the appropriate parameters, the sizes of the regions for the modified compartmentalized reactor were larger than those for the former type. This shows the significance of the separate control of the reactor inflow and outflow. It is also suggested that this reactor system could be used as a model one for studying the dynamics of biological systems and for developing biomimetic devices.
Emulsion polymerization of vinyl acetate was investigated in batch and continuous flow operations performed at the same ratio of initiator, emulsifier and monomer concentrations, and at an emulsifier concentration lower than CMC. The batch operation showed a time-independent particle size distribution (PSD) of primary particles smaller than 0.3 µm, which indicated that agglomeration of colloidal polymer particles to produce secondary particles did not occur. The continuous flow operation with separate feeds of monomer and aqueous solution of initiator and emulsifier showed in the first stage (the first 90 min) the same steady-state PSDs as in batch operation, and in the following stage (after 120 min), gave time-dependent PSDs, the median of which shifted gradually from 0.1–0.3 to approximately 1.0 µm over about 3 hours. The oligomer and fine particles continuously generated by homogeneous nucleation were precipitated selectively onto large secondary particles in the second stage of continuous operation. After all secondary particles that had finished growing by selective precipitation had been discharged from the reactor, the same growth process of secondary particles was repeated.
Global reaction enhancement by forced composition cycling was investigated by calculation. In this study, a catalytic reaction mechanism proposed for NO reduction by hydrocarbons over alumina was employed. Langmuir-Hinshelwood kinetics was assumed in the mechanism. The characteristics of the mechanism were as follows: (1) two successive intermediates, (2) competitive consumption of the active second intermediate. An important finding was that global enhancement was attained by the forced cycling. Additionally, optimization of the periodic operation indicated that the maximum enhancement factor was about 1.7. The reason for the global reaction enhancement was discussed. In this system, the number of system condition functions, coverage on the catalyst surface by reactive species, is greater than that of the operating variables, concentrations of reactant gases. This makes the trajectory of the condition functions under the forced cycling frequently move apart from, or into the region where it is restricted at the steady state. This seems essential for the enhancement under unsteady conditions in the model.
Performance of a light olefins production process under both periodic and recycling operations was investigated via numerical simulation. In the periodic operation, the reactor wall temperature was set to vary sinusoidally along its axis. The by-products, butylene and pentene, were recycled to the inlet or the middle position of the reactor. As a result, we found that the yield of ethylene could be increased to 0.116 mol-ethylene/moI-methanol and the product ratio of ethylene and propylene could be raised to 0.685, by applying the appropriate combination of the recycling operation and the wall temperature modulation. The mechanism of the enhancement was also discussed from the viewpoint of a synergy effect between the two operations.
Periodic operation of reaction temperature was examined as a means to control latex particle size distribution in the continuous emulsion polymerization of vinyl acetate. The experiment was conducted in a CSTR under conditions below the critical micelle concentration of the emulsifier. The monomer conversion attained in a time-independent steady-state operation showed a dramatic rise from 0 to over 60% at a critical reaction temperature of Tc=53°C. The dynamic operation was performed by switching the reaction temperature between two values, one of which gives a very low monomer conversion and the other a high monomer conversion in the usual steady-state operation. The periodicswitching induced oscillations in monomer conversion, and the particle size distribution (PSD) varied synchronously with the switching operation.
A hybrid dynamic simulation method based on both the compartment model and the CFD model was proposed with the dual purpose of lightening the load of the process designer and of simulating reactive transport phenomena more accurately. To examine the applicability of this method, we carried out dynamic simulation of a process, in which the dynamic behavior inside the reactor was influenced remarkably by the spatial heterogeneity of state variables caused by the flow of feed. By using the hybrid simulation method, we showed that the influence of local heterogeneity in the state variables on the dynamic behavior of whole process varied with the operational conditions of mixing. Moreover, we found that it is necessary for efficient process simulation of an unsteady operation, to determine appropriately the regions to which the compartment model and CFD model could each be applied, and to develop various functions of an interface module for exchanging process data dynamically between these models.
Marangoni convection accompanying mass inflow from a deformable free surface, as is observed in the condenser section of a two-component heat pipe, was analyzed numerically by using a boundaryfitted coordinate system. It was shown that the condensed liquid is stable even in a gravity-free state. When the condensation rate and the Marangoni force throughout the condenser section are constant, condensate thickness distribution converges to a prescribed value regardless of the initial value, but it depends on various factors such as Marangoni number. In the case of a large Bond number, reflux of the condensed liquid by the Marangoni effect may be impossible if the gravitational force due to the difference in the heights of the liquid surfaces exceeds the Marangoni force, even if the heat pipe is horizontal. Calculated condensate thickness and velocity distribution were found to differ from the results by Nusselt's theory, but approximated to the values calculated by a simple parallel flow model as long as the Marangoni number is not very small. It was shown that the absence of Marangoni force in a part of the condenser section does not necessarily make it impossible for the working liquid to flow back, but that if the Marangoni force is absent at the edge of the condenser section, the value of the contact angle between the liquid and the inner wall has a significant effect on the condensation thickness distribution.
The flow through a simplified honeycomb model of eight channels was simulated with the lattice gas automata method. Comparison of automata results shows good agreement with those of the finite difference method, confirming the suitability of the lattice gas automata method. The program was designed to locate any stagnant flow regions that formed, and to automatically convert these regions into obstacles, using this obstacle process. Then the ratio between the maximum and minimum flow rates in the channels decreased by 27%. Also, parallel processing was experimented with in conjunction with the lattice gas automata method, and its adaptability was evaluated. The rate of increase in calculation speed was found to be largely proportional to the number of processors, and that parallel processing was useful and could calculate rapidly.
Thermosensitive microgel particles were prepared by emulsion polymerization using a polymerizable anionic surfactant, SE-10 N. N-isopropylacrylamide and N,N'-methylenebisacrylamide were used as primary monomer and cross-linker, respectively. The gel particles reversibly swelled and shrank in response to temperature change. The effects of synthesis conditions on the particle diameter and the swelling degree were examined. Polymerizable surfactant did not affect the diameter of gel particles, if it was present at above a certain concentration. The gel diameter was dependent on the concentrations of the initiator and accelerator. The swelling degree of gel particles was strongly affected by the concentration of the cross-linker. From these results, the mechanism of particle formation was discussed.
MFI zeolite membranes having different Si/Al ratios were prepared by secondary growth of crystalline colloidal seeds under hydrothermal conditions. An MFI membrane of Si/Al=∞ showed a n-C4H10 permeance of 1.5×10−5 m3 (STP)m−2 s−1 kPa−1 and a n/i-C4H10 selectivity of 15 at 150° C. By addition of Al into MFI membranes, the zeolite layers appeared to be randomly oriented, and n/i-C4H10 selectivities were decreased. Methylation of toluene using MFI catalytic membrane reactors was carried out at 400–500° C. Toluene conversions increased with reaction temperature and a maximal p-xylene selectivity of 80% was achieved. suggesting the effectiveness of MFI catalytic membrane reactors.
Adsorption of chromium, nickel and cobalt on porous glass was examined at 298 K. These metals were adsorbed by ion-exchange from exchange solutions that did not contain any ammine complex ions. It was considered that adsorbate ions were aqua ions of these metals and, in the cause of chromium, also Cr(OH)2+ ion. Much more chromium was adsorbed from the non-equilibrium exchange solution soon after preparation than from the exchange solution at chemical equilibrium, because chromium underwent hydroxylation for six days. Chromium was adsorbed in the pH 2.7 to 3.3 region and nickel and cobalt were adsorbed in the pH 4.0 to 6.4 region. Thus, the pH region of adsorption differed with the charge carried by the metal.
The particle size distribution of silica powders in diluted dispersed suspension was evaluated by ultrasonic attenuation spectroscopy and compared to that analyzed by X-ray sedimentation. With increasing diameter of the silica particle, the ultrasonic attenuation was dominantly influenced by viscous losses as well as scattering losses. It was shown that the calculated particle distribution based on measurement of these two losses was in good agreement with that obtained by X-ray sedimentation.
In a compression test of kneaded wet powder in a cylinder to evaluate the result of kneading, the ratio of pressure pL at the bottom of the cylinder to the pressure pU under the pressing stamp changes with the height of compressed powder. By assuming the dependence of friction μK between powder and cylinder wall upon pressure, the relation between the ratio pL/pU and the height of powder was analysed for both compression under constant pressing velocity and compression under constant pressure, and the analysis was confirmed by experimental data.
The culture conditions of human keratinocytes were examined in serum-free medium with trypsin and/or trypsin inhibitor to achieve subculture operation in the successive passages of the cells. Although the cells inoculated in the trypsin-containing medium suffered serious suppression in the potentials of adhesion to the culture surface and subsequent growth in the early culture phase, the coexistence of trypsin inhibitor with trypsin in the medium resulted in a potential of cellular adhesion similar to that in the conventional medium without the treatment agents. Concerning the delay in cell growth, the cells treated with trypsin and its inhibitor showed a doubling time of 23 h, a comparable value to that of the untreated cells, in the exponential growth phase after exchanging the medium with a conventional one. This indicated that the potential of cell growth was restored by the medium exchange to remove trypsin and its inhibitor. In the cultures with trypsin and its inhibitor, it was found that adherent cell concentration at 96 h was maximized when time of first medium exchange (tm) was in the range of 24 to 36 h, and that prolonging tm caused an increase in lag time until cell division in a linear manner. The profile of cell growth was expressed by employing a cell placement model in the successive culture comprising the first and second passages, and the final cell concentration in the first passage and tm value in the second passage were determined so as to give an enhanced cell production rate under the examined conditions. Based on the calculated results, the successive culture was performed through subculture operation with the trypsin and trypsin inhibitor treatment, and the cell production rate obtained was at the same level as the culture associated with centrifugation.
For the purpose of adsorptive removal of penta- and tri-valent arsenic existing as arsenate and arsenite anions, respectively, in water, a novel adsorption gel was prepared by crosslinking pectic acid with formaldehyde. In order to obtain fundamental information on the adsorption behavior of this gel, before the adsorption of arsenic, the order of selectivity among metals which exist as cationic species in water was investigated and found to be as follows: Sn(IV) ≫ Sn(II) ≫ Pb(II) ~ Fe(III) ≫ Cu(II) > Al(III) ≫ Zn(II) ~ Ni(II) ~ Ca(II) ≧ CO(II) > Mn(II). The maximum loading capacity for ferric ion was 0. 57 mol/kg-dry gel. Because the ferric iron is known to have a high affinity to arsenate and arsenite anions, iron(III)-loaded crosslinked pectic acid gel was prepared to investigate its adsorption behavior for these arsenic species in water by means of batch and column tests. Arsenate was predominantly adsorbed in the weakly acidic region, while arsenite was predominanthly adsorbed in the weakly alkaline region. Arsenate was found to be selectively adsorbed on the gel in the presence of excess chloride and sulfate ions in both of batch and column adsorption tests. In the adsorption tests using a packed column, arsenate exhibited a more favorable breakthrough profile than arsenite. Arsenate and arsenite were eluted from the column with 1 M hydrochloric acid, together with the loaded iron. Arsenate was also eluted with 1 M sodium hydroxide solution together with a small amount of the loaded iron.
Adsorption removal of organic chlorine and bromine compounds from water was carried out using a temperature-sensitive polymer hydrogel synthesized from polyvinylalcohol (PVA). We previously reported that the volume and surface properties of temperature-sensitive polymer hydrogel changed with temperature in the range from 275 K to 323 K, and that the amount of adsorbed 1,2-dichloroethane increased remarkably at temperatures above about 310 K. In this study, we examined the adsorption characteristics of toxic organic compounds in water using temperature sensitive hydrophilic hydrogel. Ten organic compounds, namely, chloroform, bromoform, trichloroethylene, phenol, p-chlorophenol, o-chlorophenol, m-chlorophenol, 2,4-dichlorophenol, 2,4,6-trichlorophenol and benzoic acid, were used for adsorption removal test. It was found that the amount of adsorbed organic compound also increased at temperatures above about 310 K, and that the driving force for adsorption of an organic compound onto PVA polymer hydrogel with temperature swing is hydrophobic interaction.
The sources of tap water have recently become polluted with a number of chemicals discharged from various industries. Among them, pesticides from golf links and agricultural areas can be considered as priority chemicals for water quality management. In water treatment processes, activated carbon is generally used to remove pollutants. However, humic substances also contaminate environmental water, such as river water. In this study, we examined adsorption isotherms and rates on activated carbon at various temperatures, in order to examine the basic characteristics of adsorption using activated carbon for removal of 2,4-dichlorophenoxy acetic acid (2,4-D), humic substances, and their combination. The adsorption isotherms were correlated with Freundlich-type equation in the concentration ranges tested. Surface diffusion was considered to be dominant. For the selective adsorption treatment for 2,4-D in solotion with humic substances, activated carbon with small pores was more effective than that with large pores.
To immobilize lead compounds in fly ash from a melting process of MSW incineration fly ash, fly ash was mixed with waste plastic, sulfur and calcium hydroxide in a bi-axial kneader at 523 K. The waste plastic samples used were: polyethylene (PE), polypropylene (PP) and polystyrene (PS). The physical immobility of treated fly ash with plastics and the applicability of plastics to sulfurating medium were examined by using a dissolution test. As a result, lead concentration in the eluant decreased with an increase in the amount of plastic added. However, Japan's lead emission standard of 0.3 mg/l was not achieved by this treatment. On the other hand, there was no detectable lead in the eluant from fly ash mixed with plastic (2.5 g/10 g-fly ash), sulfur (1.2–1.4 g/10 g-fly ash) and calcium hydroxide (1.0–1.4 g/10 g-fly ash), confirming the complete insolubilization of lead compounds in fly ash.