Excess Gas Flow Research Articles

BLANKET TUNGSTEN: A NON-ISOTHERMAL APPROACH

The metal in the blanket tungsten process for microelectronic interconnects is deposited in via holes or trenches of sub-micron radius/width and depths of the order 2 μm. This process is usually isothermal, based on chemical vapor deposition. However, the resulting concentration gradients along the depth of the via promote formation of a void (keyhole) in the feature, unless processing is at low temperatures and hence, for longer duration. Using process modeling, the feasibility of the two approaches are investigated that rely on a non-isothermal process, which are equally fast and reduces the volume of the keyhole defect. In the first, application of a thermal gradient is examined, which can mitigate the tendency to form keyholes. However, the process conditions require excessive gas flow rates. In another, continuous cooling of the substrate while deposition occurs simultaneously is examined. The method based on continuous cooling appears more feasible in optimizing between keyhole size and process time.

Computational approach to granular segregation in tumbling blenders

Binary systems of particles of different size but equal density are fluidized in a 30-cm diameter bed with a perforated plate distributor. This work described the extensive experimentation, and relates the mixing/segregation properties to the visible bubble flow rate, the particle size ratio, and other parameters of minor influence. Experimental data are expressed as mixing index, correlated in terms of the decisive parameters. Comparison with previous empirical equations for the mixing index is also included. The excess gas flow rate required to avoid segregation in a fluidized bed of wide size distribution powders can be calculated from the expression for the mixing index (Eq. (15)).

Preparation and characterization of Pb(Nb,Ti)O3 thin films by metalorganic chemical vapor deposition

Lead niobium titanate, Pb(Nb,Ti)O3, thin films were prepared on various substrates by metalorganic chemical vapor deposition from the Pb(C11H19O2)2–Nb(O⋅C2H5)5–Ti(O⋅i-C3H7)4–O2 system. Polycrystalline films with (111) orientation and epitaxially grown films with c-axis orientation were deposited on the (111)Pt/Ti/SiO2(100)Si and PbTiO3/(111)Pt/Ti/SiO2/(100)Si substrates, and (100)SrRuO3//(100)LaAlO3 substrate, respectively, at 620 °C. Nb content in the film can be controlled only by changing the input gas flow rate of Nb(O⋅C2H5)5 under the excess gas flow rate of Pb(C11H19O2)2. The composition of the film was independent of the deposition temperature from 400 to 620 °C. The single phase of Pb(Nb,Ti)O3 was deposited up to the Nb content of about 5 at % and pyrochlore phase was codeposited above this Nb content. Leakage current density decreased with the increasing Nb content in the film up to 6.5 at %. Coercive field (Ec) of the film was about 130 kV/cm and was independent of the Nb content and the kinds of substrates. On the other hand, the remanent polarization (Pr) of the film on the (111)Pt/Ti/SiO2/(100)Si substrate increased from 9 to 29 μC/cm2 with the increasing Nb content from 2.1 to 4.8 at %. Pr of the film on the (100)SrRuO3//(100)LaAlO3 substrate was smaller than those of the (111)Pt/Ti/SiO2/Si and PbTiO3/(111)Pt/Ti/SiO2/Si substrates.

Segregation by size difference in gas fluidized beds

Binary systems of particles of different size but equal density are fluidized in a 30-cm diameter bed with a perforated plate distributor. This work described the extensive experimentation, and relates the mixing/segregation properties to the visible bubble flow rate, the particle size ratio, and other parameters of minor influence. Experimental data are expressed as mixing index, correlated in terms of the decisive parameters. Comparison with previous empirical equations for the mixing index is also included. The excess gas flow rate required to avoid segregation in a fluidized bed of wide size distribution powders can be calculated from the expression for the mixing index (Eq. (15)).

Mathematical modelling of a bubbling fluidised-bed coal gasifier and the significance of ‘net flow’

An isothermal model, incorporating the two-phase theory, has been developed to evaluate the performance of a bubbling fluidised-bed coal gasifier. A distinctive feature of this model is the consideration of a net flow term from the emulsion phase to the bubble phase in the conservation equations. Simulations with consideration of the net flow term indicate that the overall results compare favourably with available experimental data from an industrial fluidised-bed gasifier reported in the literature. The net flow is significant, in the range 71–87% relative to the feed gas rate, strongly depending on the coal rank, heterogeneous reaction rates and volatile matter released in the bed. The higher the coal rank, the lower the net flow and total excess gas flow. The large volume of net flow generated can significantly change the fluidisation conditions in the bed and thus alter the reaction rates and mass transfer properties. Simulations without the net flow deviate significantly from the experimental results.

Properties of thin Ta–N films reactively sputtered on Cu/SiO 2/Si substrates

This work studied the thin film properties of 200 Å Ta and Ta–N films reactively sputtered on the Cu/SiO 2/Si substrates. The nitrogen atomic concentration in the Ta–N film was found to saturate at about 30%. Excessive sputtering gas flow, especially the N 2 gas, caused surface damage to the sputter deposited films. Thermal annealing in N 2 ambient at temperatures up to 700°C did not change the atomic concentrations and the chemical states of Ta and N in the Ta and Ta–N films. The thermal annealing resulted in the grain growth and the healing of sputter damage, but it also induced new defects in the Ta–N films due to thermal stress. This presents a reliability problem in process application involving high temperature thermal cycle.

Analysis of Throughflow Velocity in Two-Dimensional Fluidized Bed Bubbles

The formation of gas bubbles is one of the most characteristic phenomena of fluidized beds. Many unique properties of fluidized beds can be related directly to the presence of bubbles and are dominated by their behavior. Therefore, accurate prediction of parameters such as bubble shape and size, voidage variation and throughflow are practically important. In the present analysis, an approximate model, based on a strongly idealized picture of the bubble formation has been presented. The bubbling gas fluidized bed has regions of low solids density comprised of gas pockets or voids. The observed voids exhibited a variety of shapes (Halow and Nicoletti, 1992), depending upon the material and fluidization velocity. In the low-velocity experiments with the finer materials, rounded voids are observed. However, with coarser materials, voids were typically large and bluntnosed. In the image analyses work, reported by Gautam (1989), in a bed operating slightly above the incipient fluidization, elongated bubbles (a > b, as shown in Figure 1) were observed for glass beads (sp. gravity = 2.5) of mean diameter 500 μm and flattened bubbles (a < b) were seen for mean particle diameter of 350 μm. Also, he noticed the dependence of throughflow velocity on the elongation of the bubble as it traverses up the bed. Additionally, throughflow velocity was found to be independent of the excess gas flow rate through the bed. The digitized image of a typical bubble (refer Gautam et al., 1994) which shows that the bubble were elongated in the vertical direction and were more elliptical than circular. Therefore, description of a bubble on the basis of just one diameter, either the horizontal or the vertical or an equivalent diameter, as has been done by many researchers in the past, is rather incomplete. It is inferred from the present work that the bubble aspect ratio plays an important role in predicting an accurate gas flow through the bubble.

Prediction of the particle flow conditions in the freeboard of a freely bubbling fluidized bed

The density and flux of particles in the freeboard of a freely bubbling fluidized bed is a function of the size, aspect ratio and frequency of bubbles erupting at the surface of the dense bed. To predict the velocity and amount of material thrown into the freeboard, a previous model for the eruption of a single bubble (Yule and Glicksman, 1989, Proceedings Engineering Foundation International Fluidization Conference, Canada) was reformulated to deal with the eruption of a multiple bubbles undergoing coalescence at the bed surface. The solution includes the explicit mechanics of particle elements in the nose of the leading bubble as well as the particle elements in the wakes of bubbles trapped between coalescing bubbles. The motive force for the particle acceleration is the excess gas flow through the cavity at the bed surface. The gas flow, in turn is found from a numerical solution spanning the bed from the upper surface to the distributor. The solutions for a single erupting bubble, two vertically aligned bubbles, and three vertically aligned bubbles were obtained numerically. The predicted vertical mass flux of particles decreases exponentially with height. The smooth fall off of the predicted flux with height above the bed is due to the consideration of separate particle elements which have different initial conditions at different points around the nose of the bubble. The solution for the erupting single sphere agrees with the measurements of George and Grace (1978, A.I.Ch.E.Symp. Ser. 74). For a freely bubbling bed, the frequency of bubble eruptions and the probability of coalescence at the bed surface can be obtained from a statistical model of the bubbling process which has been used before to predict accurately the expansion of bubbling beds (Glicksman et al., 1991, Chem. Engng Sci. 46, 1561–1571). Combining the statistical model with the predictions for single eruptions results in a prediction of the flux and density distributions in the freeboard. Measured density distributions above a freely bubbling bed, containing horizontal tubes, are in general agreement with the model predictions. The distributions are found to be primarily controlled by the bubble size at the bed surface, the volume fraction of bubbles in the bed and the excess gas velocity U U mf . The predicted exponential coefficient for density and mass flux vs height is consistent with the correlation of experimental results given by Kunii and Levenspiel (1990, Powder Technol. 61, 193).

A system for monitoring the delivery of ventilating gas to the oxygenator during cardiopulmonary bypass

Present methods for monitoring delivery of ventilating gas to the oxygenator during cardiopulmonary by pass have not been critically examined despite the occurrence of catastrophic gas flow disturbances. Currently, a flowmeter regulates the delivery of gas into the proximal portion of the gas circuit, and at these institutions, an oxygen (O 2) analyzer positioned downstream of the flowmeter monitors oxygen concentration of the flowing gas. However, these devices are incapable of accurately analyzing the flow of gas into the oxygenator housing. In response to this deficiency, the authors developed and evaluated, in a laboratory model, an alternative system for monitoring ventilating gas flow. The system measured quantity of gas flow by pneumotachography, and oxygen concentration of the ventilating gas with a polarographic oxygen analyzer at the gas inlet port of the oxygenator. Five types of disturbances in gas delivery, which have been documented in the medical literature, were created in a laboratory model: (1) O 2 concentration variation, (2 and 3) oxygenator gas line disconnection at each of two locations in the gas circuit, (4) a decrease in gas flow to the oxygenator secondary to a leak in the gas circuit, and (5) excess gas flow to the oxygenator. The capability of currently used monitoring devices (flowmeter and “passive” O 2 analyzer) with the proposed monitoring system (pneumotachograph and “aspirating” O 2 analyzer) was compared for detecting these abnormalities in gas delivery. The currently used devices were only able to satisfactorily detect a change in oxygen concentration. In contrast, the proposed system was able to rapidly detect all types of gas delivery disturbance. Based on these findings, the authors recommend further development and clinical testing of this proposed system.

Predictable nitrous oxide uptake enables simple oxygen uptake monitoring during low flow anaesthesia.

The uptake rate of oxygen and nitrous oxide were studied during low flow anaesthesia with enflurane or isoflurane in nitrous oxide with either spontaneous or controlled ventilation. The excess gas flow and composition were analysed. The nitrous oxide uptake rate was in agreement with Severinghaus' formula VN20 1000.t-0.5. The composition of excess gas was predictable and the following formula for oxygen uptake could be derived: VO2 = VfgO2-0.45 (VfgN2(0)-(kg: 70.1000.t-0.5)) where oxygen uptake rate (VO2, ml.min-1) equals oxygen fresh gas flow (VfgO2) minus 0.45 times the difference between the fresh gas flow of nitrous oxide (VfgN2O), ml.min-1 and estimated uptake of nitrous oxide. The equation assumes constant inspired gas concentrations of 30% oxygen and 65-70% nitrous oxide. The oxygen uptake rates calculated from this formula were in good agreement with measured uptake rates. Thus, continuous monitoring of oxygen uptake rates is possible by using only reliable flowmeters and analysis of inspired oxygen concentration.

Sampled gas need not be returned during low-flow anesthesia

The purpose of this investigation was to study the N2 flux between the patient and the breathing circuit, and the excess gas during N2O anesthesia with the low, fresh gas flow technique. Forty patients were studied. After a 6-minute high, fresh gas flow denitrogenation period, the O2 fresh gas flow was set at about 4 ml/kg/min and the N2O fresh gas flow was set to maintain an inspired O2 fraction of 0.30. The excess gas flow and N2 excretion were measured by a variant of the Douglas bag method. The mean inspired N2 concentration reached a peak of 5.9% at 40 minutes. The estimated mean N2 excretion was 39 ml/min at 10 minutes, declining to 18 ml/min at 60 minutes. A calculation of N2 homeostasis during closed-circuit anesthesia based on the results of the patient study indicated that sampling for gas analysis actually reduces the gas costs if the sampled gas is scavenged instead of returned to the circle system, since intermittent flushing with high, fresh gas flow for denitrogenation is unnecessary in the former situation. Regardless of the fresh gas flow used, sampled gas need not be returned during N2O anesthesia.

Slugging behaviour of fluidized beds of large particles

An experimental study has been carried out to investigate the slugging behaviour of fluidized beds of large particles (Group D according to the Geldart classification). The influence of the bed diameter on the transition from bubbling to slugging flow conditions has been quantified by means of a simple relationship, which allows faithful predictions. Slug length and frequency, and average bed expansion have been measured as functions of the excess gas flow rate, u— u mf, with beds of different diameters and overall heights. An attempt is presented to correlate the experimental data in terms of the main physical and operating variables.

Bed expansion and the visible bubble flow rate in gas fluidized beds

The paper describes the experimental and theoretical investigations on (1) the variation of fluidized bed expansion with gas flow rate, particle size, temperature and the presence of immersed surfaces, and (2) the visible bubble flow rate and its relation to the excess gas flow rate. Both experiments conducted by the authors and literature data are analysed.

充填層内の通気性及び伝熱特性からみた新塊成鉱プロセスの評価

A new iron ore agglomeration process was experimentally evaluated in comparison with the sinter process on the basis of permeability and heat transfer characteristics in the bed. Desirable operating conditions for the new process were also investigated. The main results obtained are as follows : (1) The productivity of the new process increased linearly in proportion to the gas volume passing through bed, while the maximum productivity of the sinter process was estimated at 1.6 t/m2 h due to the limit of the permeability in the bed by the capacity of the suction blower. (2) According to the sintering test results for the two processes using a pot made of a transparent silica glass tube, the new process's melting zone temperature seemed to be higher than that of the sinter process, owing to be the efficient combustion of fine coke coated on the green ball. (3) An evaluation of gas flow characteristics in the bed by a mathematical model, the productive yield for the new process seemed to be superior to that of the siner process, in that it controlled excessive gas flow near the side plate. (4) It was proved by both mathematical and experimental evaluations that the size of the green balls should be mainly controlled between 5 and 10 mm in order to achieve sufficient heat conduction within the pellet.

Control of the quality of fluidization in a tall bed using the variance of pressure fluctuations

A quantitative relation was established experimentally between the excess gas flow above minimum fluidization and the variance of the differential pressure fluctuations in a bed. The measured variance was used as part of a computer control scheme to maintain a tall fluidized bed very close to minimum fluidization throughout its length. This was achieved by bleeding gas from the fluidized bed at several different levels. The control strategy has been successfully demonstrated using a 200-mm dia., 3.2-m tall fluidized bed.

Design of Vacuum Systems for MFTF-B Neutral Beamlines

MFTF-B is the largest fusion mirror device currently under construction. When completed in 1988 the mirror will be fueled and heated by 11 neutral beamlines whose combined power output will be over 40 MW. These beamlines are being designed and built by TRW Systems. Associated with each beamline is a vacuum system whose performance plays a crucial role in the operation of neutral beams. Good vacuum is needed in the injectors and dump tanks to limit the beam loss due to reionization and to avoid excessive gas flow into the plasma chamber. This paper will describe the design of these vacuum systems and explain the considerations and tradeoffs made in the process.

Anaesthetic pollution in the dental out‐patient surgery

Five factors influencing pollution in the dental surgery are considered. Despite leakage from round the mask and from the mouth, scavenging can remove the majority of expired anaesthetic. The effectiveness of scavenging can usefully be monitored with a Wright respirometer. Spillage of anaesthetic vapour at induction and at the end of anaesthesia must be avoided. In addition to scavenging, pollution levels are minimised if a Boyle's type machine is used and excessive fresh gas flow avoided. Room ventilation, e.g. extract fans in windows, is vital in reducing pollution levels in the surgery. Room size and non-uniformity of ventilation must also be considered.

A comparison of four methods for measuring respiration in organic material

Four methods for measurement of respiration (CO 2 evolution) were compared on uniformly prepared, birch forest-floor organic matter; the methods included infrared gas analysis (i.r.g.a.), gas chromatography (g.c.), KOH absorption of CO 2 (b.a.) and the Gilson respirometer. Respiration estimates were made at a single temperature and also at variable temperatures at constant moisture contents. The i.r.g.a. and b.a. methods gave maximum estimates for respiration; the g.c. gave lowest, with the Gilson intermediate. Lower estimates for the g.c. are attributed to the closed system employed, which included no means of increasing gas diffusion (such as a KOH-CO 2 sink or continuous gas flow) out of the sample into the unoccupied space of the respiration chamber. At 25°C, the minimum sensitivity of the g.c. and Gilson were 3.8 and 3.6 μg CO 2, respectively. The minimum sensitivity of the b.a. method (44 μg CO 2) could have been improved with use of a micro-liter burette for titrations. The least minimum sensitivity, 0.31 μg CO 2, was attained with the i.r.g.a. Conditions under which any of the techniques may be used to estimate respiration should be carefully worked out. The relationship between sample size, gas flow rate (open system), chamber atmosphere composition (closed system), and chamber design should be established with sufficient confidence so that stimulation of respiration by excessive gas flow rate, or exhaustion of O 2 or CO 2 inhibition through extended chamber closure does not adversely affect respiration estimates.

Prediction of bubble size in a gas fluidised bed

It is shown that existing data on bubble size (when this is not restricted by the vessel dimensions) can be described by ▪ where ( U — U mf ) is the excess gas flow, h is the height above the distributor and h 0, a measure of the initial bubble size, characterises the distributor. h 0 is effectively zero for a porous plate but may be more than a meter for large tuyeres.

Dispersion of tracer gas supplied at the distributor of freely bubbling fluidised beds

The dispersion of tracer gas, continuously injected beneath an isolated area of the distributor plate, has been determined from time averaged concentration measurements in freely bubbling three dimensional air fluidised beds. The concentration profiles at different levels and for a series of flow rates have been plotted dimensionlessly showing them to be substantially alike in shape, this plot has been well fitted with a simple expression [1]. The tracer dispersion at points within the bed has been successfully predicted using a single phase model, depending mainly on the radial dispersion coefficient. The radial dispersion coefficient has been found to be roughly constant at all levels in the bed (except near the bottom where it is high), but increases rapidly, from a value near molecular diffusion, with an increase in the excess gas flow ( U — U m.f.). Within the experimental range considered, it appears that the Peclet number is inversely proportional to the average bubble diameter.