Search Results

In this study, the authors analyze the concentration distribution of an evaporative spray mixture with LIEF (Laser induced exciplex fluorescence) method, which is a type of optical measurement. LIEF method is one of the optical measurements for obtaining the spray concentration distribution for separating vapor/liquid phases based on the fluorescence characteristics. In this paper, a quantitative concentration distribution analysis method for wall impingement spray in heterogeneous temperature field has been proposed. Then, a series of experiments were performed in varying injection pressure and ambient density. As a result, a two-dimensional concentration distribution was obtained for the free spray and wall impingement spray.

Three-dimensional large eddy simulation (LES) has been conducted for a diesel spray flame using KIVALES which is LES version of KIVA code. Modified TAB model, velocity interpolation model and rigid sphere model are used to improve the prediction of the fuel-mixture process in the diesel spray. Combustion is simulated using the Eddy-Dissipation model. CIP method was incorporated into the KIVALES in order to suppress the numerical instability on the combustible flow. The formation of soot and NO was simulated using Hiroyasu model and KIVA original model. Three different grid resolutions were used to examine the grid dependency. The result shows that the LES approach with 0.5 mm grid size is able to resolve the instantaneous spray with the intermittency in the spray periphery, the axi-symmetric shape and meandering flow after the end of injection as shown in the experimental results.

Large Eddy Simulation (LES) is applied to non-evaporative and evaporative diesel spray simulations. KIVALES, which is LES version of KIVA code, is used as the LES computational code. Modified TAB model is used as breakup model, and interpolated donor cell differencing scheme is employed to calculate convective terms. To validity LES simulation, LES results using KIVALES are compared with experimental results and simulated results with conventional RANS approach using KIVA3V res.2. The results show that the LES simulation of non-evaporative spray depends on the grid size in comparison with RANS simulation, and good agreement is obtained between experimental results and the LES results with fine grid (720,000 cells). Furthermore, asymmetric non-evaporative spray which has intermittency at the outer edge of sprays is simulated, since instantaneous turbulent flow field can be predicted directly in LES case.

It is significant for understanding the phenomena in a stratified charge engine and an SI engine with direct injection system to carry out the fundamental research. The experiments were conducted in a constant volume chamber with atmospheric condition. The pre-mixed charge composed of ethylene and air was charged with various equivalence ratio, the second charge with the same composition was injected into the chamber, thereafter, the combustion started by a spark plug. The phenomena were analyzed by use of the experimental results of shadowgraph, [OH] natural emission, pressure history and NOx and UHC in the exhaust gas.

Diesel sprays injected into a combustion chamber of a small sized high-speed CI engine impinge surely on a piston surface and a cylinder wall. As a consequence, their vaporization, mixture formation and combustion processes are affected by impingement phenomena. And the other important factors affecting on the processes is the injection pressure. Then, the distribution of the vapor concentration in a single diesel spray impinging on a flat and hot wall was experimented by the exciplex fluorescence method, as a simple case. The injection pressure was varied in the range from 55 MPa to 120 MPa. It is found that the distribution of the vapor concentration in this case is much leaner than that in the case of the low injection pressure of 17.8MPa.

This paper deals with the particle distribution in Diesel spray under the non-evaporating condition from the analytical aspect based on our experimental results. In the analysis, TAB method of KIVA II code and the k-ε turbulent model were used, and the mono-disperse distribution of the initial parcel's diameter, whose size equals to the nozzle hole diameter, was utilized in conjunction with the breakup model. The size distribution of atomized droplets (i.e. the χ-squared distribution function) is justified with the degree of freedom. It is shown that the ambient gas, which is initially quiescent, is induced and led to a turbulent gas jet. The turbulent gas jet which has a equivalent momentum with the Diesel spray was also examined by Discrete Vortex method. The quantitative jet growth was shown to be possible for the estimation and determination in its initial boundary values at the nozzle.

The combustion of a diesel spray includes very complex processes, that is, atomization, evaporation, diffusion, turbulent mixing and burning. On the other hand, there are no phenomena of atomization and evaporation in the combustion of a transient gas jet. However, the latter jet can be treated as a fundamental of the former spray. From the standpoint mentioned above, acetylene gas was injected into the ambient during short duration as a transient gas jet and its flow characteristics were investigated by means of photography with a sheet of laser light and LDV to detect the turbulent vortex generated in the boundary layer between it and surroundings, in the experiments presented here. And the experimental results show that the jet itself is divided into four peculiar regions and the modelling of each region is carried out by use of the results to understand the mixture formation process owing to the turbulent diffusive mixing.

The purposes of this study are to clarify the atomization mechanism, the change over time in droplet size distribution, and the change in spray characteristics dependent on back pressure on diesel fuel spray. Diesel spray injected into a quiescent gaseous environment under high pressure is observed by taking direct microscopic photographs varying the moment of exposure, the back pressure, and the ambient density. The results show that the mechanism of spray atomization is divided into 4 processes, and spatial distribution of breakup droplets and a droplet volume rate are assessed for the whole spray region. Total and local distributions of droplet size are expressed by empirical equations as a function of time elapsed from the moment of injection. It is confirmed that the uniformity of the distribution, Sauter mean diameter of droplets, and droplet production rate change with time. Mean droplet diameter is further described in relation to the pressure drop and the ambient density.