The construction and performance of a sodium sulfur cell with dynamic sodium and sulfur electrodes are described. The cell was constructed with a sodium feed into a [3"-alumina tube and a sulfur feed into an annular sulfur electrode. Lowresistance graphite felt was tightly packed around the ~"-alumina tube. Sodium pentasulfide was removed from the sulfur electrode. The cell was stably charged in the two-phase region and a high charge acceptance of 95% was obtained. The cell capacity and the discharge voltage increased with the sulfur and sodium feeds. The internal resistance was decreased by thinning the sulfur electrode and using a single zone of low-resistance graphite felt.

The density distributions in the axial and radial cross sections of two typical sodium sulfur cells were obtained by a high-energy x-ray computed tomography (CT) system with a linear accelerator. The tomograms of the low-resistance cell showed a uniform density distribution of sulfur and sodium pentasulfide in the graphite felt of the sulfur electrode, while those of the high-resistance one were indicative of a nonuniform density distribution. The difference in the cell characteristics between the cells could be accounted for by the difference in the density distribution of the active material in the sulfur electrode. The low discharge voltage of the high-resistance cell, calculated using an internal-parallel-cell model, agreed well with experimental results. The tomograms of cells taken by x-ray CT can be applied to evaluate sodium sulfur cell characteristics. ) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 155.33.120.209 Downloaded on 2015-03-17 to IP ) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 155.33.120.209 Downloaded on 2015-03-17 to IPABSTRACT This paper describes a study of mass-transfer process and particle motion in a plating barrel. The mass-transfer coefficient was measured with a diffusion-controlled metal-dissolution reaction, and the motion of metal particles was examined with a video recording system. Three kinds of particle motion were observed in a horizontal barrel: (i) slumping motion at a low barrel rotational speed of 3 rpm; (it) falling motion at the barrel rotational speeds of 6 to 13 rpm; and (iii) cascading motion at 15 to 17 rpm. When the barrel was tilted at an angle of 30 to 60 ~ from the horizontal position, the particles in the lower portion of the barrel rotated like a rigid body. There was no relative particle movement in this regime.The falling motion of particles occurred only near the top surface of the particle load. At a barrel tilt angle of 90 ~ (i.e., a vertical barrel), all the particles rotated like a rigid body, and no falling layer was observed. The mass-transfer rate to the particles increased with increasing barrel rotational speed, and decreased with increasing tilt angle from the horizontal position. When the barrel tilt angle was less than 60 ~ the mass-transfer rate decreased with increasing barrel loading. The effect of barrel loading on mass transfer decreased with increasing tilt angle; at a barrel tilt angle of 90 ~ the barrel loading had a negligible effect on the mass-transfer rate in the barrel. A set of empirical equations was obtained to correlate the Sherwood number to the Reynolds number, Schmidt number, Grashof number, barrel tilt angle, barrel loading, and barrel immersion.) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 155.33.120.209 Downloaded on 2015-03-17 to IP

The azimuthal and axial velocity of the source-sink flow in a rapidly rotating cylinder were measured by using the back scattered mode of the laser-Doppler velocimeter. A rotating optical system was devised for measuring the gas flow with a high angular velocity of 5.2×102 rad/s. The incident laser beams were introduced into the cylinder through the rotating optical system which synchronously, but independently, rotating with the cylinder. The Doppler-shifted signal was observed from the light back-scattered by small paraffin mist particles of about 1 μm in diameter. The accuracies of the measurements were within ±5% for the azimuthal flow-velocity component and ±8% for the axial one. The E1/4 layers formed on the sidewall and the feed radius were clearly observed due to the small Ekman number E=2.2×10−5. The theoretical results (boundary layer theory and numerical calculation) agree quantitatively with the experimental ones, except in the case of the feed radius.

The sodium sulfur battery has a high theoretical specific energy, 760 Wh kg Ϫ1 , with a low rate of self-discharge. Hence, considerable effort has been devoted to its development for large-scale energystorage applications, such as in load leveling systems for electric power plants. [1][2][3] The number of sodium sulfur cells needed to store the required energy, however, becomes very large in such systems. These cells are arranged in series and parallel chains, and each cell must have the same charge and discharge characteristics, high efficiency, and good reliability. Therefore, much work has focused on making the cell characteristics uniform and improving the energy efficiency by optimizing the cell structure, especially regarding the sulfur electrode. [4][5][6] Typical sodium sulfur cells under development consist of sodium and sulfur electrodes separated by a solid electrolyte, ␤Љ-alumina. Liquid sodium and mixtures of liquid sulfur, sodium polysulfide, and graphite felt fill and the respective electrode containers. During repeated charging and discharging, sodium ions pass through the wall of the ␤Љ-alumina, react with sulfur or sodium polysulfide at the surface of the graphite felt in the sulfur electrode, and then form different ionic species. Therefore the charge and discharge characteristics of the cell depend strongly on the distribution of the active materials in the sulfur electrode.Cell characteristics can be effectively evaluated by measuring the active material distribution. However these cell characteristics are generally evaluated from cell voltage and current during charge-discharge operations. 4-6 Because the active materials, sodium and sodium polysulfide, react readily with moisture and oxygen in air at room temperature, the distribution of the active material cannot be measured accurately by destructive methods. 7 It should instead be obtained by a nondestructive method. The behavior of active material has been observed using X-ray radiography. 8 But this method cannot clarify the distribution of active material on a cross section of the cell.To solve these problems, in our previous report, we have observed a cross section of the cell at room temperature using X-ray computed tomography (CT). 9 In the present report, an in situ X-ray CT system with a high-energy linear accelerator was applied to image cross sections of sodium sulfur cells during charge and discharge at 350ЊC. The cell charge-discharge characteristics were evaluated using the tomograms. Sodium Sulfur CellSodium sulfur cell reaction.-The discharge reaction of the sodium sulfur cell is written as follows 2Na ϩ xS r Na 2 S x (x ϭ 5, 4, 3)[1]The cell is discharged from the two-phase region composed of sulfur and sodium pentasulfide to form sodium trisulfide. A reaction model of the sodium sulfur cell is shown in Fig. 1. During discharge, the sodium ions penetrate the ␤Љ-alumina, reacting with the sulfur or sodium polysulfide in the graphite felt of the sulfur electrode. During the charging operation, the sodium polysulfide is decom...

An improved sodium sulfur cell design is proposed to enhance capacity and to decrease internal resistance. The cell is constructed with the sodium negative electrode inside a ~"-alumina tube and an annular sulfur positive electrode that utilizes unique and separate charge and discharge electrodes, outside the tube. The active material in the positive electrode is circulated, using the density difference between sulfur and sodium pentasulfide. An experimental cell was stably charged and discharged in the two-phase region composed of sulfur and sodium pentasulfide. Compared to a conventional cell design, the effective cell capacity was increased without increasing the surface area of a ~"-alumina tube while the internal resistance was decreased. The resistance decrease was due to a thinner positive electrode and the unique charge and discharge electrodes. ABSTRACTThis paper presents artificial neural networks (ANNs) to model thermally based microelectronic manufacturing processes. The specific processes chosen are the chemically vapor deposited (CVD) epitaxial deposition of silicon in a horizontal reactor and "pool boiling" as applied to vapor-phase soldering. In the CVD processes, an analytic model is used to generate data under simulated production conditions. Part of the data sets are used to train the neural network models. These models, referred to hereafter as physiconeural models, are then used to predict the output as a function oi input parameters for the other part of the data sets. For pool boiling, an empirical correlation is used to train the ANN model. A comparison of these predictions with the physical model's computational results for the CVD process, and the experimental data for the pool boiling shows good agreement. These results show the effectiveness of the artificial-neural-network technique for modeling complex processes. Further work is in progress to exploit fully the potential of neural models, singly or in conjunction with physical models for run-to-run and real-time process control.