The paper proposes a new class of scheme based on wavelet decomposition of the Navier-Stokes equations and their solutions. The solutions are composed of scaling functions by projecting the nonlinear terms onto the space spanned by the scaling functions. The scheme is first applied to a Riemann problem of the onedimensional inviscid Burgers' equation and the numerical solutions show excellent agreements with the exact one. The scheme is also applied to the Euler equations with the Sod condition, and captures the sharp shock front and the density discontinuity. It is then extended to two-dimension and applied to the cavity flow problems with the Reynolds number up to 20 000. The results, the velocity profile of the centerline at Reynolds number 1 000 in particular, present excellent agreements with the results that are frequently referred to.

View full abstract

A new numerical algorithm for incompressible fluid flow subjected to strong system rotation is proposed. The Coriolis term in the Navier-Stokes equation is diagonalized by introducing a complex velocity and the term is advanced implicitly. Two examples of numerical simulation are presented. The first example is the source-sink flow between rotating disks and it is proved that the computational cost of the present algorithm is ten times cheeper than the previous one. The second example is the source-sink flow between the rotating housing with a spherical wall and the rotating core with combined spherical and cylindrical walls, which is the flow field applied to a particle classification system. The second example indicates the applicability of the present algorithm to complex geometry.

View full abstract

To give appropriate assessment of thermo-fluid phenomena in the space applications under various gravity conditions, the algorithm of CCUP was restructured and expanded for the three-dimensional flows in the curved linear coordinate system. In the modified algorithm, called TCUP, not density but temperature was selected as the independent variable so that the heat transfer even in the incompressible flow field should be simulated adequately. In the present paper, both the supersonic flow field around a right circular cone and the natural convection in a square cavity were simulated with the developed code and the each of numerical results was compared with the theoretical solution or the benchmark result, respectively. Consequently, quite a good agreement was obtained between them.

View full abstract

An implicit preconditioning method based on the so-called preconditioning method proposed by Weiss and Smith for solving very low-Mach-number flows and the higher-order finite-difference method developed by the authors for solving transonic viscous flows is presented. Especially, an implicit flux-splitting form in general curvilinear coordinates is newly introduced in this paper. This form is applicable not only to the flux-difference splitting in Roe's Riemann solver but also to the implicit fluxes in the LU-SGS scheme. Very low-Mach-number flows anound NACA0012 airfoil are calculated to validate the numerical accuracy of the present method and a natural convection around a cylinder is calculated to compare the results with the experiments.

View full abstract

The sounding mechanism in air-reed instruments is investigated based on the computational fluid dynamics. While previous studies have assumed a sound source based on acoustic theory, the authors attempt in the present study to elucidate exactly the source of sound as an essential factor defining the timbre of musical instruments. To deal with the large computational grids required to capture minute changes in pressure, the Hitachi SR 8000 massively parallel supercomputer is employed for computation. The computational results are consistent with Brown's experimental equation, and new frequency component revealed only in this three-dimensional analysis is identified, attributed to the harmonics of the air jet.

View full abstract

The generation and the development of the magnetic flux density fluctuation due to the coherent fine scale eddies are investigated in detail by analyzing DNS data of MHD homogeneous turbulence. With the addition of the magnetic field, two circulations of magnetic flux density fluctuation are generated around the eddy on the plane perpendicular to the axis of coherent fine scale eddy in the early stage, while these circulations disappear with time. The current density shows the maximum and the minimum value on the same plane. The distance between the center of the eddy and the location of the maximum or the minimum is almost same with the radius of the fine scale eddy. If the axis of the fine scale eddy is perpendicular to the direction of the applied magnetic field, the Lorentz force weakens the swirling motion of the eddy and works as shear stress for it. However, the Lorentz force tends to expand the eddy if the axis of the fine scale eddy is parallel to the direction of the applied magnetic field.

View full abstract

Highly loaded nozzle blades of steam turbines are developed. The pitch-chord ratio of these nozzles increases by 14% compared to one of conventional nozzles. And the increase of the stage efficiency is about 0.35% at the design point. This reason is that the strong adverse pressure gradient along blade suction surface is avoided in spite of the increase of blade loading, because the blade loading is equally distributed to the whole blade and the curvature of suction surface are optimized. In the first step of development, the nozzles are designed by the aid of computational fluid dynamics technique. In the next step, the increases of stage efficiency are assured by the air turbine test results.

View full abstract

Probe characteristics are numerically obtained in a wide range of electron temperatures and densities for the collision-dominated plasma sheath in which the fundamental Langmuir theory of the electrostatic probe fails. The asymptotic theory of Cohen for continuum conservation equations is applied to the ellipsoidal probe, and to the plane probe as the limiting case. It is found that, for the collision-dominated plasma sheath, significant errors in measured electron temperature are expected if the fundamental Langmuir theory is used to evaluate the temperature from the measured probe current and voltage. For quick evaluation of the electron temperature in such plasmas, the bound triple-probe method is proposed, and numerical results are summarized in a graphical form for practical use.

View full abstract

The effects of both generation frequency and bubble diameter on the behaviors of single bubble chains in stagnant water were experimentally investigated. We developed the bubble generator equipment to control precisely both bubble generation frequency and bubble diameter, independently. Bubbles were generated with diameters from 1.0 mm to 2.5 mm, and generation frequencies from 1 Hz to 20 Hz. We investigated bubble chain patters developments in stagnant water, by taking both generation frequency and diameter as parameters. Two types of bubble chain patterns were investigated carefully. The first type was the bubbles rising in zig-zag or helical. In this case the behaviors of bubbles with l/d>100 (l : mutual distance d : diameter) were observed to be identical to those of single bubbles, on the contrary, those of bubbles with l/d<50∼100 were influenced by the interaction by the other bubbles. The second type was the bubbles rising in line. The behavior of bubbles with l/d>200 were indistinguishable from those of single bubble.

View full abstract

For an oscillating single hydrofoil with tip clearance, tip leakage vortex cavitations were observed experimentally. It was found that the oscillation of the cavity size was delayed behind the oscillation of the angle of attack. The maximum value of the cavity size decreased when the frequency of oscillating hydrofoil increased. To simulate the unsteady characteristics of the tip leakage vortex cavitation, a simple calculation of 2-D unsteady flow based on the slender body approximation was conducted with taking into account the effects of cavity growth. The calculation results of the cavity volume fluctuation showed qualitative agreement with the experimental results.

View full abstract

To estimate the total amount of adsorbed surfactants onto the surface of a single rising bubble, the Navier-Stokes equation, the convection-diffusion and the surface-diffusion equations of the surfactants were numerically solved at the steady state and the effects of the Reynolds number (Re) and the bulk concentration of the surfactants on the surface concentration distribution were investigated. The results reveal that the surface concentration of the surfactants at the top of the bubble is lower than that at the bottom and the difference between them increases as the Re increases or the bulk concentration of the surfactants decreases. The critical Re (Re_c) can be estimated for the bubble to behave as a solid particle and the averaged concentration of surfactants coincide with the equilibrium concentration corresponding to a zero net adsorption-desorption flux for Re<Re_c.

View full abstract

In the present paper, shock oscillations around transonic symmetric double circular blade are investigated to clarify the behavior of the unsteady flow fields and the cause of oscillations. First, the flow around isolated blade in a duct is examined experimentally and numerically. Their results show good agreements in terms of shock structures, frequencies and oscillation modes. In addition one-pitch-periodic cascade is examined numerically. New flow models of linear stability analyses are proposed. Based on the models, it was found that Kelvin-Helmholtz instability, which occurs due to the interaction between the main flow and the wake, causes the shock oscillation.

View full abstract

It is possible to control the spread of a plane jet passively by the vortices from a two-dimensional cylinder installed in the nozzle contraction. In this study, the wake characteristics of a cylinder and the coherent structure of a jet by this passive control method were examined by flow visualization and quantitative velocity measurements. In wide range of strouhal number based on vortex shedding frequency and jet exit velocity, the spreads of the jet were found to be suppressed. This suppression is caused by the high frequency vortices from the control cylinder which inhibits growth and pairing of the low frequency vortices. On the other hand, in certain cases of the strouhal number being 0.05-0.10, the spread of the jet was remarkably enhanced associated with distinctive peaks of spectrum at higher order of vortex shedding frequency caused by the deformation of the vortices in flow contraction nozzle. In the case of enhanced mode, vortex from the cylinder moves to the opposite side shear layer with the same vorticity sign as the vortex, and it enhances roll up of the shear layer.

View full abstract

An experimental investigation of pulsating flow through a tapered curved tube is presented to study the blood flow in the aorta. The experiments are performed in a U-tube with the curvature radius ratio of 3.5 and including 50 percent reduction in the cross-sectional area between the inlet and the exit of the curved section. Laser Doppler velocity measurements are made for the Womersley number of 10, the mean Dean number of 400 and the flow rate ratio of 1. Flow visualizations are used to investigate qualitatively the nature of the flow, complementing the quantitative LDV measurements. The velocity profiles for the steady and pulsating flows in the tapered U-tube are compared with the corresponding results in a U-tube having a uniform cross-section. The striking effects of the tapering on the flow are exerted on the axial velocity profiles in the section from the latter half of the bend to the neighborhood of the inlet of a downstream tangent. A depression phenomena in the velocity profile appears at the position of smaller bend angle Ω and the degree of the depression against the cross-sectional averaged velocity is reduced. Strong secondary fluid motions occur near the bend exit and that leads a velocity profile to depress weakly in the downstream tangent. The value of β_m which indicates a uniformity in the velocity profiles decreases with an increase of Ω, but it rapidly increases just after the bend exit.

View full abstract

An LDA flow rate meter which is based on the LDA (Laser Doppler anemometer) technique has been developed. The principle of the LDA flow rate meter applied for circular pipe is following. The center line velocity of pulsating pipe flow was measured by the LDA. The velocity distribution profile in the pipe cross section can be numerically estimated, and then the instantaneous flow rate was calculated by the integration of the profile. Firstly, it was applied and evaluated at the pulsating flow made by an artificial heat-lung system. The system has a double roller pump with a vinyl tube of Ω-shape. Secondly, an intermittent flow rate at high fuel pressure injection system was also measured. Swirl injector generated the oscillating pipe flow under high line pressure of 5 MPa at injection frequencies of 20 and 40 Hz. There was a little difficulty at the measurement of injection flow depending on the zero velocity shifting. Finally, a correction method of zero shifting was proposed and the applicability of the flow rate meter was demonstrated.

View full abstract

Cavitation bubbles in a starting water jet discharged from a circular nozzle into still water are investigated by a conventional photography technique in a moderately low range of jet exit velocity. Most of bubbles are initially generated in starting vortices and connected with each other in the form similar to vortex rings. Nearly axisymmetric lumps of disconnected bubbles are also observed frequently. By analyzing photographic data acquired from the side-and end-view pictures of the ring-like bubbles, their average properties, such as trajectory, geometry and size, are evaluated.

View full abstract

The characteristics of cavitation instabilities were investigated through the measurements of pressure and blade stress fluctuations. Under the condition of head break down at lower cavitation number, a higher order rotating cavitation was observed. It can be understood in two ways : 2 cells with forward propagation velocity ratio of 4, or 1 cell with backward propagation speed ratio of 5. From high-speed video pictures, it was found that the cavity in the tip leakage flow fluctuates mainly near the trailing edge. It causes pressure and blade stress fluctuations near the trailing edge. These characteristics are different from those of conventional rotating cavitation at higher cavitation number. Higher order cavitation surge with the frequency ratio of 5.1 was also observed under the same condition.

View full abstract

To investigate the role of large-scale structures in the layer away from the wall, DNSs (Direct Numerical Simulations) of turbulent channel flow having small size of the computational box are conducted ; i. e., the size of spanwise direction is set narrower than the half channel width to prevent the formation of streamwise roll-mode structures : the axis of rollers are directed to the nearly streamwise direction and their diameter is the order of channel half width. It is turned out that the mean streamwise velocity remarkably increases in the region away from the wall, while the nearwall turbulence retains their universal property independent of the modulation of large-scale flow. POD (Proper Orthogonal Decomposition) method is applied to extract the large-scale flow. Although the large-scale structures are drastically altered from streamwise roll mode to spanwise one, the large-scale flow, regardless of the type of rotation mode, plays a crucially important role in the distribution of turbulent shear stresses and the formation of coherent elementary vortices.

View full abstract

Turbulent channel flows rotating about the spanwise, streamwise and wall-normal axes are calculated with a low-Reynolds-number second-moment closure. The stabilizing and destabilizing effects of the spanwise rotation are reproduced by the model due to the exact stress production term related to the rotation. In the case of weak streamwise rotation, a shear stress component induced by the rotaion changes its sign at a position between the wall and the centerline. This feature is captured by the model and the predicted velocity profiles are in reasonable agreement with DNS data. At high rotation rates, however, the sign reversal disappears in DNS and the predicted sign disagrees with that of DNS almost over the whole region, leading to poor predictions of the velocity profiles. The budget of the stress component suggests a need for a redistribution process which works against the exact rotation production. The wall-normal rotation directly affects the mean flow field. This rotation leads to relaminarization of the flow when the rotation rate becomes high. Good predictions are obtained for weak rotation but the model gives relaminarization at a lower rotation number than DNS.

View full abstract

Two-dimensional numerical computation has been performed for unsteady laminar flow. Spatial periodic boundary conditions were adopted in the streamwise direction and adequacy of this treatment was particularly studied by changing the computational domain size. The following results were obtained. The Strouhal number and the lift coefficient amplitude may change with the domain size in the same flow conditions. Therefore, the simulation for one unit periodic flow field, a "unit cell", is valid for limited flow condition which depends on the Reynolds number, blockage ratio of the rod and its pitch. Uniqueness of the numerical results, which means calculated results remain unchanged with the domain size, is judged by change of the phase difference between neighboring cells. When the numerical results are unique, the Strouhal number, drag coefficient and the lift coefficient amplitude remarkably increase with the blockage ratio. The lift coefficient amplitude decreases with the increase of the rod pitch.

View full abstract

Numerical analyses were performed for the oscillation effect of impingement surface on the impingement heat transfer and flow with a confined wall. Two-dimensional governing equations were solved for the Reynolds number Re=200 and 500, Prandtl number Pr=0.71, the dimensionless space between nozzle and impingement surface H=1.0 and the Strouhal number Sf=0∼1.0, as a moving boundary problem. Oscillation induces the enhancement and depression of the local heat transfer. The local heat transfer was improved at comparatively low frequency due to the flow fluctuation. On the other hand, at high frequency it was depressed due to the flow for upper direction near the impingement surface. The oscillation effect spatially appeared at the downflow side after impingement.

View full abstract

Radiative heat transfer through a cover gas between upper wall and lower liquid sodium surface, filled with polydispersed sodium mist particles, is investigated numerically by Monte Carlo Method. Radiative properties of sodium mist such as extinction coefficient and scattering phase function are calculated using particle distributions for sodium mist by Mie theory. Using these properties the radiative heat flux is computed for the cover gas space of gray isotropic and nongray anisotropic scattering particles, respectively. The results obtained show that, the heat flux decreases with increasing concentration of sodium mist, however, the effect of sodium mist on total heat flux is relatively low.

View full abstract

This study is a theoretical analysis of the frequency response of general-purpose temperature sensors (thermistors) and the compensation for the delay of the response. A "two-thermocouple probe technique" developed previously (Tagawa & Ohta, 1997) was applied to the estimation of the thermal time-constant values of thermistors whose frequency response can be modeled by a firstorder system. Two thermistors of unequal time-constant values were combined in a probe, and the schemes for the time-constant identification and the response compensation were experimentally tested in air and water flows. The results show that the proposed estimation scheme can provide reliable predictions of the thermistor response, and that the compensation technique works successfully.

View full abstract

The characteristics of both flow and heat transfer were investigated experimentally and analytically for ice slurry flow in a return bend with rectangular cross section. The ice slurry, which is a mixture of fine ice particles and ethylene glycol aqueous binary solution, was utilized as a testing material. In this study, heat transfer coefficient at the heated wall was experimentally determined and ice packing factor distribution in the flow channel was observed in detail. Furthermore, a numerical simulation has been carried out to investigate both velocity distribution and solid fraction distribution in the flow channel. The solutions were obtained numerically by adopting the Algebraic Slip Mixture (ASM) model of a CFD code for number of flow path dimensions and flow parameters. The simulation results indicated that the flow behavior in the curved duct was strongly affected by both buoyancy and centrifugal force, therefore, secondary flow appeared within the parameter range covered in the present study.

View full abstract

The heat-transport characteristics of a bubble-driven non-looped heat-transport device were experimentally measured. The device consists of a non-looped meandering copper tube containing water. The inner diameter of the tube is sufficiently small to allow the formation of vapor and liquid plugs in the tube. At steady state, most of the measured wall temperatures on the tube surface fluctuated periodically. These fluctuations had specific peak frequencies, which increased from 1.3 to 2.0 Hz when the heat-transport rate increased from 150 to 250 W. Time-averaged temperature gradients of the device were almost constant and similar to the case of the heat-conduction mode of a metallic rod. The effective thermal conductivity of the device is proportional to the heat-transport rate but does not depend on the gravity effect. These results indicate that the device is governed by oscillation-induced heat transport.

View full abstract

Evaporation of water droplet can be enhanced by plasma irradiation that increases the surface wettability. Relation between plasma irradiation time and contact angle was examined first for three metals and then the lifetime of water drop on hot surface was measured changing the surface wettability by plasma irradiation. The lifetime of water drop decreased and the wetting limit temperature increased with the increasing irradiation time of plasma. Hydrophilicity by plasma irradiation is not a permanent effect but it will be useful for enhancement of cooling of hot metal.

View full abstract

Reduction of pressure loss of a rectangular block on the wall of parallel channel was investigated by setting a rod on the wall upstream or downstream of the block. The height of the channel B was 30 mm. The side length, C, of the rectangular block was 30 and 60 mm and the height of the block, H, was from 8 to 20 mm. The mean velocity of the channel ranged from 6 to 10 m/s. The effects of the rod diameter and its position on the pressure loss coefficient ζ were measured. The reduction of pressure loss achieved 30% and 15% for a suitable upstream and downstream rod, respectively. The pressure drop on the sudden contraction reduced by the upstream rod, and the static pressure rise on the sudden expansion increased by the downstream rod. The total reduction of pressure loss was obtained 44% by using simultaneously the upstream and downstream rods.

View full abstract

In order to predict accurately the NO_x emission of turbulent diffusion flame, we proposed the combination method based on the laminar flamelet model. In this method, the detailed kinetics calculation is separated from the turbulent flow calculation, and then they are recombined in terms of recombination parameters to yield the NO_x emission of the whole flame. The recombination parameters have to be defined in common for both calculations. With the aim to find out the recombination parameters and to prove the validity of them, we carried out the numerical calculation for the unsteady combustion characteristics of counterflow diffusion flame, in which the spout velocities of counterflow varied in a cyclic. The obtained results are as follows : The production/consumption rates of major stable species are mostly determined by the scalar dissipation rate at flame surface SDR_q, and the production rates of intermediate products and nitric oxides are mostly detemnined by SDR_q and the maximum flame temperature T_max. Consequently, it is verified that SDR_q and T_max are appropriate as the recombination parameters of the combination method.

View full abstract

In this study, the initial process of the solidification of flowing supercooled water was investigated analytically. In the process, thin ice propagates on the heat transfer surface toward upstream after the solidification nuclei appeared in the flowing supercooled water, and the dendrite ice grew and melted after growth of thin ice. The calculation model was proposed to reproduce the transient process in the flowing supercooled water, and the growth and the melting of the dendrite ice were calculated. In calculation, the inlet temperature of the water, the average flow rate and the surface temperature of the heating plate were varied as parameters. As a result, it was found that the flow disturbed by the dendrite ice at the upstream region melts the dendrite ice at the downstream region and the calculation model in this study can reproduce the solidification process. Moreover, the effects on the solidification process was examined when the parameters were varied.

View full abstract

A thermal-hydraulic analysis program that can be applied to the complex flow network consisting of tubes and headers in the furnace water wall of fired boilers has been developed. This program can evaluate the flow distribution in the tubes at both steady and transient states. Heat flux distribution in the furnace, which determines each tube's heat absorption, is given by the threedimensional combustion analysis. By using this program, detailed information about the water flowing in the tube, such as temperature, quality, mass flux, etc. is obtained and the temperature of the furnace water wall is calculated. This program is applicable to most operating conditions including slagging, soot blowing and water filling at boiler start up. Both steady and transient states analyses show satisfactory agreement with measured values. This program has already been used to design fired boilers, and good performance has been obtained.

View full abstract

An experimental study has been conducted in order to elucidate influences of additions of hydrogen and methane on the combustion synthesis of diamond films. Differences between the flat flame and the conical flame are first examined, and it is reconfirmed that the velocity gradient is one of the dominant parameters in the combustion synthesis of diamond films. Additions of hydrogen and/or methane are examined on the growth rates of diamond films, crystal sizes, and morphology. These results are also confirmed by conducting the similar experiments with a welding torch, instead of the flat flame burner. It is found that an addition of hydrogen reduces the growth rate and crystal size, while that of methane enhances those although homogeneity of the diamond films is reduced.

View full abstract

The potential of using spatially-and temporally-resolved chemiluminescence measurements to explore the local flame structure of turbulent premixed flames was investigated. This study examined relations between local behaviors of reaction zone and OH^*, CH^*, and C^*₂ chemilumisescence intensities. Chemiluminescence intensity ratio of OH^*/CH^*, C^*₂/CH^*, and C^*₂/OH^* had encouraging linear correlations with equivalence ratio (φ=0.9∼1.4) and temperature (1 860∼2 160 K) in atmospheric laminar CH₄/air and C₃H₈/air premixed flames. Notice that these ratios have little dependence on movement and curvatures of unstretched flame fronts, which should be occured in actual turbulent premixed flames. Thus, the chemiluminescence intensity ratio can be a good indicator for changing of the local flame structure in turbulent premixed flame fronts. Time evolution of OH^*, CH^*, and C^*₂ chemiluminescence intensities measured simultaneously with ion currents in turbulent C₃H₈/air premixed flames showed capabilities of this optical technique to illustrate the structure of local flame fronts. Furthermore, flamelet passing frequency measured by chemiluminescences agreed well with that by ion currents.

View full abstract

Ignition process and burning characteristics of a polymethylmethacrylate (PMMA) rod in highly preheated convective environment are investigated experimentally. The main parameters are O₂ concentration, flow velocity, preheated air temperature and oxygen dilution gas. The convective ignition can be classified into three steps : blue flame generation, attached flame formation and envelope flame development. Convective ignition of PMMA suddenly exposed to high temperature air is initiated in downstream and dim blue premixed flame is formed there. When the pyrolysis of solid fuel increases enough for flame propagation, the blue flame propagates to upstream to form an attached flame or luminous envelope flame. The ignition delay time is almost independent on oxygen concentration because the initial blue flame formation process is mainly controlled by pyrolysis rate. However, the following attached flame and envelope flame formation time increases with decrease in the oxygen concentration since these processes are chiefly dominated by burning velocity.

View full abstract

The turbine inlet temperature of gas turbine engines becomes higher and higher to grow up the power in recent years. The high temperature properties of the ceramic materials offer the potential of higher turbine operating temperatures with no cooling techniques and therefore the ceramic components with higher temperature limits are becoming more attractive. In this study, a ceramic turbine disc is applied to an aircraft gas turbine engine. A tight fit configuration is generally taken for the joint structure between a ceramic turbine and a metal shaft. However, the constrain force is decreased due to the difference of the thermal expansion coefficient between ceramic and metal at the high temperature condition. This paper describes about the design method and the results of the mechanical tests for the joint structure. Furthermore, the ceramic turbine engine tests were demonstrated under full operating condition.

View full abstract

Characteristics of spray injected from the five types of swirl nozzle, which were the same design however differed in the static flow rates as 600, 700, 800, 900 and 1 000 cc/min respectively, have been evaluated by a phase Doppler anemometer (PDA) and a flow rate meter based on laser Doppler anemometer (LDA flow rate meter). The experiments were mainly made under the injection pressure of 7.0 MPa, injection frequency of 16.7 Hz and fuel amount of 8.6 and 24.4 mg/cycle. The lateral position of peak in the radial distribution of mean droplets velocity was increasing with the nozzle static flow rate. The modified time dividing analysis for intermittent spray flow was proposed and applied to the results. It was well explained by the both measurements of PDA and LDA flow rate meter that the instantaneous fuel flow rate at the initial stage of the valve opening duration affected the spray characteristics of droplet diameter and velocity.

View full abstract

The ignition phenomena of diesel fuel spray injected into a constant volume vessel with medium injection pressure were investigated. Results show that ignition occurs at a certain point near the surface of spray liquid column before breakup where activation reaction is advanced one step ahead of others and the impingement plate assists surface ignition. From these results was deduced the following concept that fuel vaporized from the liquid surface begins to be activated when the velocity gradient at the liquid column surface decreases to a certain level, leading to ignition. The blowing-off limited ignition model was proposed from this concept. Based on the functional relation between various factors and the ignition delay derived from this model, the ignition delay date were well arranged into an experimental formula. Next, ignition investigation was carried out using a bottom-viewed single cylinder engine, revealing that the effects of many factors including EGR rate related to O₂ concentration can be analyzed comprehensively based on the above model.

View full abstract

Equations to describe heat transfer from burning gas to the combustion chamber wall have been empirically derived from hydrocarbon combustion engines. Previous research by the authors analyzed the applicability of them to hydrogen combustion, and showed that the equations calculate lower cooling loss than experimental values. It was therefore concluded that a new heat transfer equation is necessary to describe cooling loss in hydrogen combustion engines. This research focused on the gas velocity term in the heat transfer equation and investigated new terms to replace it for better fit to hydrogen combustion. A new equation with a gas velocity term including rate of heat release was newly proposed and showed better application to hydrogen combustion compared to the widely used Woschni's equation.

View full abstract