When reaching the stall point of an axial-flow pump, the pump head characteristic becomes unstable and the pump head suddenly drops. Before this happens however, at even higher flow rates the NPSH3 and the pump body and shaft vibrations increase dramatically. For effectively increasing the available operating range, it is essential to find a solution for all three problems without reducing the pump efficiency at design. The paper describes an experimental investigation on the outlined subject that gives insight into the flow phenomena leading to stall. Based on this knowledge a very simple type of casing treatment was chosen and investigated. It was found to satisfy all mentioned requirements. Subject to the investigations is a highly loaded axial-flow pump having a n q of 150 (SI units). The overall pump performance was investigated measuring pump head, efficiency, NPSH3, and casing as well as shaft vibrations. Further-more, oil flow pictures taken at the pump casing and at the rotor blades, and video captures of the cavitating core of the tip clearance vortex were analyzed for understanding the flow phenomena leading to stall (see also related paper Part I, Schrapp et al. (2004)). From the video captures it was realized that the behavior of the tip clearance vortex which was found to perform so-called spiral-type vortex breakdown is triggering stall inception in this machine.

Marshall Space Flight Center has developed and demonstrated a measurement device for sensing and resolving the hydrodynamic loads on fluid machinery. The device — a derivative of the six-component wind tunnel balance — senses the forces and moments on the rotating device through a weakened shaft section instrumented with a series of strain gauges. This “rotating balance” was designed to directly measure the steady and unsteady hydrodynamic loads on an inducer, thereby defining the amplitude and frequency content associated with operating in various cavitation modes. The rotating balance was calibrated statically using a dead-weight load system in order to generate the 6 × 12 calibration matrix later used to convert measured voltages to engineering units. Structural modeling suggested that the rotating assembly first bending mode would be significantly reduced with the balance’s inclusion. This reduction in structural stiffness was later confirmed experimentally with a hammer-impact test. This effect, coupled with the relatively large damping associated with the rotating balance waterproofing material, limited the device’s bandwidth to approximately 50 Hertz. Other pre-test validations included sensing the test article rotating assembly built-in imbalance for two configurations and directly measuring the assembly mass and buoyancy while submerged under water. Both tests matched predictions and confirmed the device’s sensitivity while stationary and rotating. The rotating balance was then demonstrated in a water test of a full-scale Space Shuttle Main Engine high-pressure liquid oxygen pump inducer. Experimental data was collected a scaled operating conditions at three flow coefficients across a range of cavitation numbers for the single inducer geometry and radial clearance. Two distinct cavitation modes were observed: symmetric tip vortex cavitation and alternate-blade cavitation. Although previous experimental tests on the same inducer demonstrated two additional cavitation modes at lower inlet pressures, these conditions proved unreachable with the rotating balance installed due to the intense dynamic environment. The sensed radial load was less influenced by flow coefficient than by cavitation number or cavitation mode although the flow coefficient range was relatively narrow. Transition from symmetric tip vortex to alternate-blade cavitation corresponded to changes in both radial load magnitude and radial load orientation relative to the inducer. Sensed moments indicated that the effective load center moved downstream during this change in cavitation mode. An occurrence of “higher-order cavitation” was also detected in both the stationary pressures and the rotating balance data although the frequency of the phenomena was well above the reliable bandwidth of the rotating balance. In summary the experimental tests proved both the concept and device’s capability despite the limitations and confirmed that hydrodynamically-induced forces and moments develop in response to the unbalanced pressure field, which is, in turn, a product of the cavitation environment.

Moderate cavitation was used to visualize the tip clearance vortex in a highly-loaded single-stage axial-flow pump. Two different techniques were utilized to produce photographs and videos. From these it was found that for flow rates lower than design an initially well defined clearance vortex undergoes spiral-type vortex breakdown. This result was theoretically confirmed in an approximate manner by solving the Bragg and Hawthorne equation for an isolated vortex with a measured casingwall pressure distribution as a prescribed boundary condition.

The response of a compressor rotor blade tip to disturbances simulating unsteady stator rotor interaction in the casing region of a turbomachine such as an axial propulsion pump or axial compressor has been studied. The measurements were performed using 24 microphones embedded in the surfaces of two blades of a linear cascade with tip gap and single sensor-hot wire anemomemetry within the tip gap at a single tip gap ( t g / c = 0.033). The unsteady velocity measurements were made perpendicular to the blade camber line at the same locations at which the microphones were embedded. Relative motion of the blade tips and casing was simulated with a moving endwall. Unsteady periodic inflow from an upstream stator row was modeled using a set of vortex generators attached to, and moving with the endwall. The unsteady pressure was measured at eight tip gaps ( t g / c = 0.00825, 0.0165, 0.022, 0.033, 0.045, 0.057, 0.079, 0.129) along the chord and the span of the blades for a blade chord Reynolds number of 390,000. Upon initial investigations, the pressure fluctuations appear to be almost a linear variation with tip gap up to t g / c = 0.057. With increasing tip gap past this, the pressure fluctuations begin to decrease. Both the pressure and velocity fluctuations increase significantly near the trailing edge of the cascade blade suggesting that the response involves both viscous and inviscid mechanisms.

The effects of fluid transients, or water hammer, in closed loop systems are somewhat different than those observed in open ended systems. The open loop system has received much attention in the literature, not so for the closed system. The generally accepted method of characteristics (MOC) technique was applied herein to investigate closed loop systems. The magnitudes of the pressures during fluid transients were investigated for examples of rapid valve closures, and the operations of parallel pumps. The effects of trapped air in the system were also considered for these examples. To reduce the pressures caused by the transients, the installation of slow closing valves were evaluated for different conditions.

The flow-fields around airfoils in a uniform flow under the generation of noise were numerically studied and compared with experimental data. The numerical simulation was carried out by a large-eddy simulation that employs a deductive dynamic model as a subgrid-scale model. The result for a symmetrical airfoil at small angle of attack α = 3°–6° indicates that the discrete or narrow-banded frequency noise is generated when the separated laminar flow reattaches near the trailing edge of the pressure side and a strong instability thereafter affects positive vortices shed near the trailing edge. This type of forced transition or late transition instabilities near the trailing edge of the pressure side, interacting with convected vortices in an attached T.B.L. on the suction side, can be found in many practical airfoils of impellers rotating at moderate speeds under design conditions. The sound spectra derived from the aero-acoustic computations of airfoils indicate a dipole nature of sound having a narrow-banded or discrete peak by laminar instability and turbulent vortex shedding from their trailing edges of finite thickness at a Strouhal frequency, a quadrupole sound by turbulent broadband boundary-layer noise, or a mixed mode depending on flow conditions near the T.E.

A novel fuel cell design uses porous bed electrodes. Liquid electrolyte, with fuel and oxidizer entrained, is circulated through the electrodes. An energy balance has been done for the cell and auxiliaries. The design calls for the CO-CO 2 mixture leaving the anode to be burned to supply an O 2 -N 2 -CO 2 mixture to the cathode. Two heat exchangers are employed: an air preheater cooling the exhaust and a waste heat exchanger cooling the gases entering the cathode. Results show that CO/CO 2 ratio is the most important parameter for determining performance. For other conditions ideal, plant efficiency varies from ∼90% for pure CO 2 to ∼20% for pure CO. If decoupled from CO/CO 2 ratio, the main effect of increasing cell temperature is to increase heat transfer rate in the air preheater. This is also the main effect of increasing excess air. Decreasing internal cell efficiency lowers power output and increases waste heat rejection.

In this paper, we present a method of calculation by partial or total condensation of the water vapor contained in the humid air, over the smooth or finned tubes-heat recupurators. This study presents an implantation of the film method in a computer code developed here. The mathematical model used is validated by our experimental approach, using tubes bundles in staggered and aligned arrangements. The heat transfer coefficient by convection around the fin is supposed too be constant. The computer code predicts the heat flux exchanged in a range of 20% and 5%, in wet and dry regime, respectively. The apparent heat transfer coefficient by condensation can exceed 10 times the value of the heat transfer coefficient. The mathematical model used is validated with the experimental data obtained in this study.

The significance of mass transfer enhancement in polymer electrolyte fuel cells (PEFC) is presented and studied in this work based on experimental investigation. A novel structure of reactant gas distributors in PEFC is proposed for mass transfer enhancement purpose. For the PEFC with novel gas distributors, it is found that the large drop of the cell voltage, generally caused by a weak mass diffusion, is postponed to occur at relatively higher current density even though the same or less amount of air is fed when comparing to a PEFC with gas distributors in conventional structure. As a result, the maximum obtained electrical power in a PEFC and a PEFC stack both are dramatically improved under both free convective and forced convective airflow conditions.

A mathematical model has been developed as a design optimization tool for the prediction of the aerodynamic behavior of a web in an industrial dryer. In particular the key modeling objective is to examine the propagation of stress waves within the web to determine its lateral motion and hence its stability. Numerical solutions were obtained through a coupling of computational fluid dynamics (CFD) calculations and structural dynamic simulations (SDS). The viscoelastic behavior of the web and the presence of turbulence were found to significantly affect the stability of the web.

Delayed Coking is a principal thermal cracking process utilized by petroleum refineries to decrease residual output. Fired heater performance is critical to coker unit profitability. Adversely affecting heater performance, coke deposits inside heater tubes must be removed periodically (decoked). There are three main methods of decoking fired heaters: online spalling, steam-air decoke, and pigging. For each decoking technique, the most widely-used procedures are described. Also, the merits and drawbacks of each method are discussed. A survey of industry decoking practice has been attempted as part of this study. Fire-side tube metal temperature (TMT) and heater inlet pressure (P IN ) are the critical parameters used to analyze heater tube fouling. However, changes in flow rate, density, process temperature, and steam injection can cause a pass to have higher TMT or P IN than one with greater fouling. This study suggests approximation of coke deposit thickness to obtain a parameter normalized against changes in process variables. When used in conjunction with TMT and P IN , such a parameter can help in evaluating decoke effectiveness and identify the pass with least economically efficient operation. Economic analysis of fired-heater fouling considers direct costs and opportunity costs. Direct costs, such as plant utilities, fuel, in-house manpower, and outside contractor expense are assessed for each decoking method. This study also evaluates opportunity costs associated with reduction of throughput during decoking procedures. An example economic case is presented to illustrate the potential benefit of using estimated coke deposit thickness to schedule decokes.

This experimental study investigates the heat transfer in a narrow, rib-roughened, rectangular channel with the width-to-height ratio of 10 for its cross-sectional shape. Two flow entrances with equal areas are respectively opened at the upper and side ends of this channel that allows for increasing the coolant-flow fed into the enhanced channel without increasing the height of cooling device. At a fixed total coolant mass flow rate, three test scenarios, namely the single-blow from the side or upper entrance and the twin-blow with the coolant mass flow ratio of 2:1 between the side and upper blows are performed. A selection of detailed heat transfer measurements, evaluated from infrared thermal images over the rib-roughened surface, illustrates the manner by which the flow entry direction and the twin-blow condition affect the local and spatially averaged heat transfers. Relative to the heat transfer results acquired from the single-blow tests, the present twin-blow entry condition could augment the spatially averaged heat transfers to the levels of 150–180% of the single-blow reference levels. A regression-type analysis is subsequently performed to develop the correlation of spatially-averaged Nusselt numbers over rib-roughened surface, which permits the effect of Reynolds number, evaluated from the total amount of coolant consumption, on heat transfer to be determined for three test scenarios.

A ceiling fan is the predominating comfort provider in tropical regions worldwide. It consists of an assembly of an electric motor with 3–4 blades suspended from the ceiling of a room. Despite its simplicity and widespread use, the flow induced by a ceiling fan in a closed room has not been investigated, and sub-optimal designs are in wide use. There is vast potential for energy conservation and improved comfort by developing optimized fan designs. This work develops a fundamental understanding of the flow characteristics of a ceiling operating inside a closed room. Using smoke from thick incense sticks, the flow field created by the ceiling fan is visualized. In most regions, the flow is periodic and three-dimensional. Vortices are seen to be attached to the blade tip and hub, which reduces downward flow and increases energy consumption. Only the middle 75% of blade actually pushes the air downwards, and the comfort region is limited to a cylinder directly under the blades; velocities in this region were measured with a vane anemometer. Winglets and spikes attached to the blade tip disrupted the tip vortex, and increased downflow by about 13% without any increase in power consumption.

The evaporation of refrigerant R134a in three different sizes of Brazed Plate Heat Exchangers (BPHE) is analyzed in this article. The BPHEs are a type of compact plate heat exchanger with parallel corrugated plates which are brazed together in series. In this study, these heat exchangers are utilized as the evaporator of an automotive refrigeration system, and their heat transfer coefficients and pressure drops are obtained experimentally. The refrigerant flows on one side and a 50% glycol-water mixture on the other side of the BPHEs, in a counter flow configuration. The heat transfer coefficient for the single-phase glycol-water mixture flow through the BPHEs is obtained using the Wilson plot technique. The single-phase Fanning friction factor correlation is also obtained using experimental results from the hydrodynamic parameters of glycol-water mixture. The results from the single-phase flow analysis are then extended to the two-phase flow analysis to correlate the evaporation heat transfer and pressure drop for the refrigerant R134a in the BPHEs. In summary, the thermo-hydrodynamic behavior of the evaporation of R143a in brazed plate heat exchangers is analyzed and compared with previous studies.

Effective cooling of modern low NOx combustor liners is achieved through combinations of impingement and other heat transfer enhancement methods. In the present study, a combination of impingement and trip strips is studied to determine the optimum location of trip strips with respect to impingement jet arrays. Heat transfer with pure impingement has degradation downstream due to increased cross-flow effects. To counter the cross-flow induced heat transfer degradation, a combination technique wherein impingement is combined with ribs placed in between impingement rows or downstream of the impingement array is studied. Three configurations with increased rib placements and reduced impingement holes are studied and compared with pure impingement cases for the same jet Reynolds number. Three jet Reynolds numbers are studied for Re j = 10000, 20000, and 30000. Detailed heat transfer distributions are obtained using the transient liquid crystal technique. Results show that the presence of ribs increases jet impingement heat transfer on the surface with lower mass flows. The effectiveness of the combination ribs and impingement can provide higher heat transfer with reduced cooling air requirements.

The refrigeration Brayton cycle, which has been used extensively in various industries, has an excellent potential for use in automotive air conditioning applications. However, the air-cycle system has a couple of drawbacks including fog generation and low cycle efficiency. In this research project, an air-to-air heat exchanger called a ‘mixer’ is designed and used at the outlet of a refrigeration Brayton cycle. The primary function of the mixer is to remove moisture from the secondary warm airflow into the system. Successful moisture removal from the secondary airflow results in achieving the second function of fog dissipation from the primary cold airflow. In order for the system to perform appropriately, the moisture removal rate must be kept at the highest possible rate. The experimental results from this research project reveal that to enhance moisture removal rate, one may either increase the primary cold airflow rate, decrease the secondary warm airflow rate, or the combination of the above airflow adjustments. Furthermore, based on experimental results, one may speculate that there is an optimum point in decreasing the secondary airflow rate. However, in increasing the primary airflow rate, one must be aware of the pressure drop through the cold side of the mixer as the higher pressure drop results in higher power consumption for the Brayton cycle. It is important to point out that appropriate levels of the primary and secondary airflows impacts the mixer effectiveness, and that for a constant cold airflow rate, decreasing the warm airflow rate below the cold airflow rate results in higher effectiveness.

Based on film theory model, an approach is developed to analyze the condensation of a binary mixture, while one is non-condensable component, on a baffled shell and one path tube condenser. Using of this approach leads to have the condensation rate and both mixture and coolant temperatures at the end of each baffle spacing, which is considered as a control volume in this research. This method is checked for a set of steam-air mixture and water as coolant on a counter and a parallel current condenser. The results show that the condensation rate and temperature profiles in each condenser have consistence trend. Also comparisons of output values of both types of the condenser for different air mass fraction indicate an overall superior operation of counter type than the parallel type.

A computational fluid dynamics (CFD) model is developed to simulate isothermal flow in a large quench tank used for heat treating of steel. As a surrogate for a full simulation of quenching, the isothermal model enables a computational economical comparison of many different design configurations of the quench tank. The model includes most of the geometric complexity of the tank including the skip, the heavy beams used to support it, deflector baffles used to control the flow, and the inlet ducting. Partial factorial screening studies are conducted to identify the most important design variables. The second phase of the project involves examining four new conceptual designs for the quench tank. The outcome of the second phase was the identification of a promising new design that could be realized by the addition of ducting to the existing design. The contributions of this research include a better understanding of the parameters affecting large scale batch quench processes and the development of new directions for quench tank design.

Experiments were performed in which the effective emittance of three types of radiator coatings was measured as Martian dust simulant was added to the radiator surface. The apparatus consisted of multiple radiator coupons on which Carbondale Red Clay dust was deposited. The coupons were powered by electric heaters, using a guard-heating configuration to achieve the accuracy required for acceptable emittance calculations. The apparatus was contained in a vacuum chamber that featured a liquid-nitrogen cooled shroud that simulated the Martian sky temperature. Radiator temperatures ranged from 250 to 350 K with sky temperatures from 185 to 248 K. Results show that as dust was added to the radiator surfaces, the effective emittance of the high-emittance coatings decreased from near 0.9 to a value of about 0.5. A low-emittance control surface, polished aluminum, demonstrated a rise in effective emittance for thin dust layers, and then a decline as the dust layer thickened. This behavior is attributed to the conductive resistance caused by the dust layer.

Experiments were performed to investigate the performance enhancement of single-phase flow and boiling heat transfer in the 3D inner finned helical tubes. The tests for single-phase flow and heat transfer were carried out in the helical tubes with a curvature of 0.0663 and a length of 1.15m, the range of the Reynolds number examined varies from 1000 to 8500. In comparison to the smooth helical tube, the experimental results of two finned helical tubes with different inner fin geometry showed that the heat transfer and flow resistance in the 3D inner finned helical tube gains greater augmentation. Within the measured range of Reynolds number, the average augmentation ratio of heat transfer of the two finned tubes are 71% and 103%, compared with the smooth helical tube, and 90% and 140% for flow resistance, respectively. The tests for flow boiling heat transfer was carried out in the 3D inner finned helical tube with a curvature of 0.0605 and a length of 0.668m. Compared with that in the smooth helical tube, the boiling heat transfer coefficient in the 3D inner finned helical tube is increased by 40%∼120% under varied mass flow rate and wall heat flux conditions, meanwhile, the flow resistance coefficient increased by 18%∼119%.