In recent years, turbine inlet temperatures in gas turbine systems have continued to rise in order to enhance efficiency and performance. However, higher temperatures results in increased levels of NOx emission and this in turn seriously affect the environment. Development of prediction methods for NOx emissions is important, since this allows one to evaluate the environmental contributions of gas turbine systems. The gas turbine combustor geometries are very complex so that the flow pattern, chemical and thermal field inside the combustor is very complex. Numerical methods for predicting the performance of the combustor and NOx formation and emission levels are attractive as they provide inexpensive solutions to the complex geometry problem without physically building the hardware. This paper provides numerical simulations for predicting the NOx emissions from gas turbine systems using: (i) chemical equilibrium and simplified kinetic reaction approaches, and (ii) flamelet and reaction progress variable approaches.
A parallel implementation of vortex methods dealing with unsteady viscous flows on a distributed computing environment through Parallel Virtual Machine (PVM) is reported in this paper. We test the recently developed diffusion schemes of vortex methods. We directly compare the particle strength exchange method with the vorticity distribution method in terms of their accuracy and computational efficiency. Comparisons between both viscous models described are presented for the impulsively started flows past a circular cylinder at Reynolds number 60. We also present the comparisons of both methods in their parallel computation efficiency and speed-up ratio.
It has become increasingly important to calculate magneto-hydrodynamics (MHD) flows under alternating-current (AC) magnetic fields, in order to control the molten metal processing step in the furnace by electromagnetic force. In this paper, three dimensional problems for an arbitrary-shape model are solved by using the hybrid finite element method-boundary element method (FEM-BEM) with the A-φ (magnetic vector potential-electric scalar potential) method in the electromagnetic field and using the arbitrary Lagrangian Eulerian-finite element method (ALE-FEM) in the velocity field with a free surface. The strong point of the ALE method is that the element itself moves along the velocity field of the fluid, and consequently, the boundary between the fluid and air is distinct. Using these methods, we can observe the complicatedly tangled physical parameters of the electromagnetic field in the cold crucible and confirm the unique shape and vibration of the surface of the molten metal.
The influence of cross-sectional configuration of a cylindrical body on the lock-in phenomenon of Kármán vortex shedding was investigated using a mechanical oscillator for cross-flow oscillation of the cylinder. A circular, a semi-circular and a triangular cylinder with an equal height were used as test cylinders to see the effect of the movement of the separation point. The lock-in criteria accounting for the spanwise coherency of the Kármán vortex were discussed based on the experimental data under a fixed Reynolds number of around 3500. The lock-in region on the plane of non-dimensional cylinder frequency and the non-dimensional amplitude was almost the same for all of the cylinders in spite of differences in the range of the separation point movement. The minimum value of non-dimensional threshold amplitude for lock-in was much smaller than the value of 0.05 which was reported so far. Results obtained in this work imply that the movement of the separation point is a result of the lock-in phenomenon, rather than an essential cause.
The flow around a rotating circular cylinder, which is placed near a moving plane wall, is investigated experimentally for high Reynolds number (Re=33000). The cylinder spanned the test section of a wind tunnel and was aligned with its axis parallel to the under wall of the test section and normal to a uniform flow. The wall of the test section under the cylinder was moved at the same speed as the uniform flow. Surface pressure measurements, velocity surveys and flow visualizations suggest the existence of three flow patterns in the near wake behind the cylinder by the change of the gap between the cylinder and the moving wall. The wake from the cylinder is characterized by the behavior of a jet passing through the gap. Furthermore, the behavior of the jet is sensitive to both the centrifugal force generated by the rotating cylinder and the inertial force of the jet.
Comparatively high Reynolds number channel flows up to Reτ=640 are simulated by DNS. Noticeable property of the high Reynolds number flow is large-scale streaks which are larger not only in scale but also in their spanwise separation than those well-known in the buffer layer. It is revealed that the low-speed streaks are the region of high turbulent activity and accordingly of high turbulent shear stress with densely populated elementary vortices. These properties are little influenced by wall condition, such as drag controlled or not. LSE (Linear Stochastic Estimation) technique is applied to extract elementary vortex pairs. Time evolution of these vortices submerged in a laminar flow but having velocity distribution of turbulent one reveals the process of generation of large-scale streaks and accompanying densely populated elementary vortices. The formation of large-scale streaks by the model simulation is also little influenced by the wall condition and this modeled process is expected to mimic that taking place in the near-wall layer of practical channel flow. The existence of the low-speed streaks itself is the key for the formation of the structure, thus suggesting the self-sustenance of turbulence by both large-scale streaks and vortices concentrated area.
In a centrifugal blower with diffuser vanes, the noise level may increase at certain rates of revolution. In particular, the phenomenon of acoustic resonance due to tones arising from aerodynamic interaction between the impeller and the diffuser is serious. In this paper, we develop a physical model to describe this phenomenon. The model assumes that the flow paths in both the impeller and the diffuser behave as quasi-one-dimensional acoustic tubes. The model predicts the critical rates of revolution at which resonance occurs, and the prediction of the model is verified by experiments. We also propose an effective method for suppressing the resonance. This involves the use of diffuser vanes with slits. Each slit is located just behind the end of the overlapping region between neighboring vanes. The physical model determines this location. Experiments showed that the method successfully reduces sharp peaks in the noise spectrum.
The existing lifting-line theory for a supercavitating hydrofoil in two-dimensional shear flow (where velocity varies in both spanwise and vertical directions) between two parallel planes is applied to partial cavitation under the same flow conditions. A lifting-line equation is derived from combining the assumption holding the lift coefficient two-dimensionally at any spanwise position with an up-wash velocity induced along a lifting-line. The hydrofoil for numerical examples is flat in section, rectangular in plan form, and has a 5 degree attack angle. Effects of two kinds of shear parameters, cavitation number, and taper ratio on local lift coefficient, total lift coefficient and induced drag coefficient are clarified through numerical calculations. Cavity length distributions are obtained on the basis of the assumption that the relationship between cavitation number and attack angle is two-dimensional at any spanwise position.
Turbo pumps have weak points, such as the pumping operation becoming unstable in the rising portion of the head characteristics and/or the cavitation occurring under the intolerably low suction head. These points have been individually overcome in general, since such phenomena originate in essentially different causes. To overcome the above weak points simultaneously, however, we propose a unique pumping system with counter-rotating mechanism and discuss its advantages. This system consists of two-stage impellers and a unique motor with double rotors. The front and rear impellers are driven respectively by the inner and outer rotors of the motor, keeping the relative rotational speed constant and counter-balancing the rotational torque. Such driving conditions not only improve smartly the unstable performance at low discharge but also suppress smartly cavitation at high discharge, under the optimum cooperation with the impeller works and the rotor outputs. The model test proved that there is no rising portion of the head characteristics and that the front impeller speed is gradually decreased with increase of the discharge as if taking the place of the inducer.
A numerical investigation for three-dimensional natural convection in concentric and eccentric annuli with open ends is made by a zonal grid approach, which extends the outlet boundary from the open end of the annuli to a far enough outside position that can be reasonably specified with the ambient flow properties. It is found that the eccentric annulus has a poorer natural convection heat dissipation rate, as compared to the concentric annulus. The inner cylinder surface temperatures decrease toward the outlet plane. It is also found that higher temperatures around the inner cylinder occur in the region near its bottom (the contacting point of the inner and outer cylinders) for the eccentric annulus, whereas in the region near its top for the concentric annulus. The variation of the inner cylinder surface temperatures is smaller for the case of concentric annulus, as compared to the eccentric annulus. The maximum inner cylinder surface temperatures occur right at the bottom of the inner cylinder for the eccentric annulus, whereas right at the top of the inner cylinder for the concentric annulus.
Analytical research was conducted to study the heat transfer from horizontal surfaces to normally impinging circular free-surface jets under arbitrary-heat-flux conditions. General expressions of heat transfer coefficients and recovery factor were obtained in the stagnation region and boundary layer region as well as viscous similarity region by the integral method. Then by using the analytical results, the conjugated heat transfer between a water jet and the impinged wall was numerically solved. Finally, an experimental study was also performed to characterize conjugated heat transfer coefficient on a thick copper target base impinged normally by circular free-surface jet. The present predictions of the mean conjugated heat transfer coefficient were good agreement with the experimental data.
Numerical simulations are employed to investigate the heat transfer and fluid flow in an inhomogeneous porous medium. The permeability heterogeneity, characterized by a correlation length and variance, are seen to have strong effect on both the flow and heat transfer. The velocity fluctuation generated by permeability heterogeneity has the tendency to alter the flow features and orientation predominantly at the scale of the correlation length. For an isotropic permeability distribution, the most pronounced effect of flow alternation and fluctuation is observed at intermediate correlation length, which agrees with previous findings, thus leads to the best performance of convective heat transfer. In a permeability field with anisotropic correlation length, larger correlation length along the stream-wise direction makes the velocity fluctuation oriented more likely parallel to the main averaged velocity, and weakens the transverse transport effects. Therefore, better heat transfer performance is found if larger correlation length is aligned along the transverse direction. The simulations also show better heat transfer effect as the variance of the permeability field or global flow rate increases.
The present paper deals with a new method of defrosting using the frost sublimation phenomenon, which occurs below the triple point of water (273.16K, 610.5Pa). The present experimental study examines the mass transfer of the annular frost layer developed on a cooling pipe exposed to an impinging jet flow. The morphology of the frost layer during sublimation was observed using a digital video recorder. It was understood that the mass flux of the frost layer increased with increasing the jet flow velocity and the difference in the mass concentration of water vapor between the frost surface and the impinging jet flow. The non-dimensional correlation equations of mass transfer of defrosting were derived as functions of various parameters.
The transient response of a 2D pin fin with a time- and space-variation base temperature, is analyzed using the Laplace transformation method and the Duhamel's method. The temperature distributions of the pin fin are identical with published analytical solution for the case using the method of separation of variables, while the latter method can not treat 2D problem presented in this paper. The heat flux at fin base and the actual heat flux transferred from the lateral surface and the tip surface of the pin fin to the surroundings are also obtained. For all cases analyzed in this paper, the temperature distributions and the heat flux of the fin reach a steady periodic response after t=2 of dimensionless time. The results show that the effects of 2D conduction are large, particular at large time. Furthermore, as Bia value is small, the effect of the tip convection on heat transfer is quite significant for short pin fins.
The structure of a multiple-pass heat exchanger composed of pipes and plate fins is similar to that of heat exchangers used for melting snow. In this study, we investigated the characteristics of a multiple-pass heat exchanger having two types, a regular pitch type and an irregular one, focusing on the brine pipe pitch of the multiple-pass heat exchanger. The perfect melting condition and the melting efficiency were related to the dimensionless parameters of the heat exchanger and its operating conditions. The calculated results for the perfect melting condition and the melting efficiency agreed with the results obtained from field tests on melting of falling snow. Applying the irregular pitch extends the critical condition for perfect melting when the flow rate of brine is low or the area for melting of falling snow is large. Also, the melting efficiency of the irregular pitch is higher than that of the regular pitch.
Bubble behaviors of subcooled nucleate pool boiling are investigated both experimentally and theoretically. The analytical study based on the dynamic microlayer model predicts that one cycle of an individual bubble experiences four stages, i. e., the initial growth with a semi-spherical shape, the final growth with a spherical segment geometry, the condensation process and the waiting time. Also, vapor bubbles on the Pt wire are experimentally observed by a high-speed camera at the speed of 10000 frames/sec in the nucleate pool boiling of subcooled water. The predicted four stages of an individual bubble are clarified experimentally. Relatively good agreement is shown between the present predictions of the total periods of individual bubbles and experimental data.
In order to clarify the turbulent burning velocity of multi-component fuel mixtures, both lean and rich two-component fuel mixtures, in which methane, propane and hydrogen were used as fuels, were prepared while maintaining the laminar burning velocity approximately constant. A distinct difference in the measured turbulent burning velocity at the same turbulence intensity is observed for two-component fuel mixtures having different addition rates of fuel, even the laminar burning velocities are approximately the same. The burning velocities of lean mixtures change almost constantly as the rate of addition changes, whereas the burning velocities of the rich mixtures show no such tendency. This trend can be explained qualitatively based on the mean local burning velocity, which is estimated by taking into account the preferential diffusion effect for each fuel component. In addition, a model of turbulent burning velocity proposed for single-component fuel mixtures may be applied to two-component fuel mixtures by considering the estimated mean local burning velocity of each fuel.
Stratification features in DI gasoline combustion were studied using a constant-volume combustion vessel. Indicated pressure analysis and high-speed combustion observation were carried out by changing injection-sparking interval τint, spray cone angle and swirl ratio. There exists a τint range where the mixture is ignitable and the shortest τint gives the highest maximum pressure rise rate (dP/dt)max and the highest (dP/dt)max reaches a maximum at a certain swirl ratio. The fairly large scatter found in(dP/dt)max data plotted against τint is markedly reduced by plotting them against total ignition delay τtot(=τint+τd), where τd denotes ignition delay. Maximum volumetric burning velocity (Sv)max was proposed as a measure of the stratification degree, based on the thermodynamic analysis which was carried out under the concept that a higher stratification degree increases the stoichiometric range of the mixture. It is noteworthy that as the spray cone angle is increased, (Sv)max increases and becomes less affected by SR.
Friction forces of a piston ring pack for a typical SI engine were measured using a floating liner system, in which the effects of cylinder pressure, oil starvation and piston secondary motion were excluded. Friction patterns of each individual ring, represented by measured friction forces, were classified into five frictional modes with regard to the combination of predominant lubrication regimes (boundary, mixed and hydrodynamic lubrication) and stroke regions (mid-stroke and dead centers). Those modes were identified on a Stribeck diagram of the dimensionless bearing parameter and friction coefficients; the coefficients were evaluated at mid-stroke and at dead centers. Frictional modes were evaluated by varying operation parameters (such as engine speed and cylinder wall temperature). Compression rings operated in the mode in which hydrodynamic lubrication was dominant at mid-stroke, while mixed lubrication was dominant at dead centers in steady conditions. However, oil control rings operated in the mode in which mixed lubrication was dominant throughout the entire stroke.
This paper describes a detailed experimental investigation of heat transfer in a reciprocating spiral tube with particular reference to the piston cooling application. The flow studied is turbulent upon entering the coils but transits into laminar in the further downstream of the coils. A selection of heat transfer measurements with which the physics of pulsating and buoyancy forces interactively affect the heat transfer along the inner and outer edges of the reciprocating coiled tube is illustrated. The pulsating force with buoyancy interaction causes the considerable heat transfer modifications from the static results. Although enhancing the buoyancy level improves heat transfer, the local Nusselt number in the reciprocating coils is initially impaired from the static value with weak reciprocation; but recovered at the higher level of pulsating force. This study focuses on the development of the experimental procedure that could lead to a physically consistent empirical correlation, which assists to evaluate the local heat transfer in the reciprocating coils by permitting the individual and interactive effects of centrifugal force, torsional force, pulsating force and reciprocating buoyancy on the forced convection to be quantified.
Research was conducted to develop an advanced three-way catalyst with better low-temperature activity and durability than those currently available. Honeycomb monolithic catalysts were designed with different Rh/Pd, Rh/Pd/Pt loading ratios in a double-layer washcoat. A single-cylinder gasoline engine was used to estimate the conversion efficiency of HC and CO. BET, SEM (scanning electron microscopy), and TEM (transmission electron microscope) analysis were performed to confirm the thermal aging process macroscopically. The characteristics of the catalytic reaction were also analyzed by TPR (temperature-programmed reduction) and TPD (temperature-programmed desorption). An Rh/Pd catalyst with a 1/18 loading ratio showed the best conversion efficiency and thermal stability in our study. The light-off temperature for 50% HC conversion of this catalyst was 260°C. The double-layer washcoat enhanced the thermal stability by inhibiting the negative interaction between Pd and Rh during the thermal aging process.
Advanced steam turbine systems which employ Ultra Super Critical pressure steam conditions provide a cost effective and efficient replacement for existing oil-fired power plant units. By application of this advanced steam turbine technology, CO2 emissions can be reduced by 10 percent and generating costs reduced by up to 20 percent. This report describes the optimization of these turbine systems and provides an economic comparison with conventional super critical pressure units.
A numerical experiment is conducted to study the bypass transition of compressible plane Poiseuille flow. The space and time development of the disturbances into streaks is captured by employing a spatial DNS, which although requires more computational resources than the temporal one, allows having a full picture of the process. The generation and algebraic growth of the streaks, which are lifted away from the walls and oscillate in the spanwise direction, characterize the transition. The numerical results give some answers related to the transition of compressible plane channel flows, initiated by large amplitude random disturbances, and point to the universality of the breakdown mechanism in wall bounded flows (incompressible and subsonic) and to the important role of the near-wall streaks.
Large eddy simulation (LES) of compression ramp flow with a shock wave was successfully performed at a free-stream Mach number of 2.9 and a Reynolds number of 106 based on the boundary layer thickness. A compressible Smagorinsky-type eddy viscosity model was employed. It was shown that the present LES predicts not only the time-averaged quantities but also fluctuations such as pressure deviation and shock motion near the compression corner. Also, the present simulation predicts the three-dimensional structure of shock/turbulent boundary interaction, which is difficult to simulate by using Reynolds averaged Navier-Stokes equations (RANS).
A detonation-driven shock tube firstly designed by H. R. Yu is considered to be a useful apparatus for producing high-enthalpy flow. In this apparatus, a strong shock wave is generated by detonating an oxygen-hydrogen mixture (oxy-hydrogen) and the driver gas temperature and pressure are extremely high compared with those of a conventional shock tube. However, the structure of the detonation wave is not uniform, e. g., the detonation wave has three-dimensional cellular structures and multiple transverse waves. Furthermore, the detonation wave is followed by a Taylor expansion fan and the performance of detonation-driven shock tube is not well understood. In this preliminary study, a detonation-driven shock tube is constructed and its performance is experimentally investigated by measuring pressure histories and the profile of the ionization current behind the detonation wave. As a result, (i)the pressure history of the detonation wave is clarified and shows reasonable agreement with the result obtained by the KASIMIR shock tube simulation code. (ii)The propagation velocity of the detonation wave coincides well with the theoretical prediction assuming a Chapman-Jouguet detonation wave. (iii)The equivalence ratio of the oxy-hydrogen mixture to produce the highest Mach number of the shock wave is evaluated to be φ approximately equal to 1.7.
The superconductor thermal stability is investigated under the effect of the hyperbolic heat conduction model. Two types of superconductors are considered which are types I and II. The superconductor thermal stability under the effect of different design, geometrical and operating conditions is studied. The effects of the time rate of change of the disturbance initial temperature and the disturbance duration on the superconductor stability is investigated. In general, it is found that the wave model predicts a wider stability region as compared to the prediction of the diffusion model.