Search

A laser 2-focus velocimeter (L2F) has been applied for measurements of velocity and size of droplets in the dense region of diesel spray. Investigated was the fuel spray injected intermittently into the atmosphere from an injector nozzle with the orifice diameter of 0.113 mm. The injection pressure was set at 40 MPa by using a common rail system. The L2F had a micro-scale measuring volume which consisted of two foci. The focal diameter was about 3 μm, and the distance between two foci was 17 μm. The maximum data sampling rate of the L2F system was markedly high as 15 MHz. Measurement was conducted at 25 mm downstream from the nozzle. The local mass flow rate and the number density were calculated at several specified times from the measured velocity, size, and arrival time of droplets. It was clearly shown that the mass flow rate was higher at the spray front than the inside, and was higher at the spray center than the periphery. The accumulated mass flow rate in the spray section measured by the L2F was almost equal to the amount of injected fuel mass within 5%.

Three dimensional flow separation occurring in a vaneless diffuser at small flow rates was analyzed by using the numerical simulation with the low Reynolds-number turbulence model (LRN k-e) as well as the momentum integral method (MI). Comparison between the LRN k-e calculation and the experiment indicates quite a good agreement in the 3-D boundary layer velocity profile. The numerical investigation shows that the LRN k-ecalculation can predict the reverse flow in the vaneless diffuser, while the MI-method behaves relatively poor although it captures the overall flow pattern correctly. It is clarified that a small gradient in the inlet tangential velocity distribution with respect to the diffuser depth is the key factor which dominates the reverse flow behavior in the vaneless diffuser at small flow rates, as demonstrated in the experiment.

The authors have proposed a new technique to suppress successfully the rotating stall in the vaneless diffuser. By positioning the completely rough wall between R=1.2 and 1.6 on the hub side diffuser wall alone, the flow rate of rotating stall inception was decreased by 42% at a small pressure drop less than 1% of the blower pressure rise. This is based on that the local reverse flow occurs firstly at the hub side in most centrifugal blowers with backswept blade impeller mainly depending on the distortion of tangential component of inlet main-flow velocity. The 3-D boundary layer calculation shows that the local reverse flow is decreased remarkably by an increase in wall friction, and it disappears if the friction coefficient is 0.01 about twice of the smooth wall. By analyzing the effect of wall friction and turbulence induced by the rough wall, it is clearly shown that the rough wall positioned in the limited region, where the wall-limiting streamline flow angle shows the negative maximum, is quite effective to suppress the rotating stall at a small drop of the diffuser pressure recovery.

The three-dimensional flow behavior in a low pressure steam turbine under the low load operation was analyzed numerically by applying the Modified FLIC (Fluid In Cell) method to the quasi three-dimensional flow computation. And it was also measured by using the five-hole Pitot probe in the scale-model turbine with the last stage. By comparing the calculated flow angles with the measured ones at the inlet and exit of the rotor, it is shown that the present computation method is a useful skill to predict an occurrence of the reverse flow in the last stage under the low load operation. Both calculated and experimental results show that the reverse flow occurs at the root of the rotor, if the axial exhaust Mach number becomes smaller than 0.3 even though the stator works well.

In order to clarify the characteristics of combustion fluctuation in diesel engines quantitatively, the fluctuations of pressure and flame temperature were measured, and the root mean square value, the autocorrelation coefficient and the power spectrum density of the measured fluctuations were analyzed by comparing the cases with and without pilot injection. The following concluding remarks are obtained. (1) The frequency power spectrum of the in-cylinder pressure history is reduced markedly by pilot injection in two frequency ranges from 0.2 to 2.0 kHz and from 2 to 5 kHz. (2) The fluctuation of combustion pressure is dependent on the maximum rate of pressure rise, which is dependent on the ignition delay. (3) The maximum rms of the flame temperature fluctuation during diffusion combustion has a clear correlation with that of the pressure fluctuation in the initial combustion. (4) The fluctuation of the flame temperature during diffusion combustion has the characteristics of isotropic turbulence.

The effect of port water-injection on the exhaust NOx reduction was examined experimentally as well as theoretically. In the experiment, water was injected into each suction port of a 4 cylinder turbocharged DI diesel engine using gasoline injectors. The exhaust NOx was reduced significantly by port water-injection, and about 50% reduction in NOx concentration was obtained by injecting an amount of 0.035kg of water/kg of air under various engine operation conditions. By comparing between the experiment and the analysis based on the authors' two-zone model, it is shown that the NOx reduction is mainly caused by a decrease in the temperature of the burned gas resulting from an increase in the specific heat due to humidification.

A newly developed stratified fuel-water-injection system has been tested for NOx reduction, targeting the future strict restrictions to be set for slow-speed cross-head marine diesel engines. In addition to this system, we have developed anothor possible measure of a direct water injection system. We introduce the details of various test results in comparison to stratified fuel-water-injection system, and also discuss the expected effects on cylinder lubrication of the two different water-injection systems.

The effect of water on NO formation was analyzed theoretically from the point of view of changes in the entrained air rate and specific heat of the burned gas in the combustion chamber. The specific heat was calculated accurately based on the chemical equilibrium composition analysis considering the absolute humidity of the inducted air, the amount of water in the emulsified fuel and the residual gas fraction. It is clarified by the two-zone combustion model analysis that about 20% reduction in NOx is attained by an increase of only 0. 01 kg/kg in the absolute humidity, and the NOx reduction due to the water-emulsified fuel is caused by increases in both gas weight and specific heat of the burned gas whereas the entrained air rate is hardly affected by the water in the emulsified fuel.

A semiconductor laser with a wide active stripe was utilized as the power source of a laser 2-focus velocimeter. The focus was expanded to about 0.2mm wide to form the two parallel focal planes. The backscattered light from seeding particles was detected by photodiodes, and the time-of-flight between two foci was counted by means of a digital clock. The flow velocity was measured in a circular jet stream, and the results were compared with those measured simultaneously by Pitot tube. After checking the accuracy of measurements, the blade-to-blade distribution of velocity and flow angle were measured at the exit of a centrifugal impeller to confirm the applicability of the semiconductor-L2F with band focus.

The relationship between Bosch smoke density and the dry soot concentration in a particulate emission is examined experimentally using various test fuels including water-emulsified fuel in the wide operation range of a small-sized, high speed-turbocharged DI diesel engine. A good correlation has been confirmed between them in the combustion tests of automotive diesel oil and marine diesel oil. However, parts of the data have deviated from the above correlation curve in the cases of water-emulsified fuel and heavy fuel. The reasons for the deviation have been clearly shown by electromicroscopic studies of trapped particulates in the filter and by analyzing a relationship between the sulfur content in the fuels and the sulfate content in the dry soot.

In order to clarify the combustion behavior in the combustion chamber of a high-speed DI diesel engine, the cylinder pressure and the radiant heat emission were measured simultaneously by using a small-sized piezo pressure transducer and an optical fiber thermometer (OFT) with the light pipe sensor respectively. The OFT output was compared with the soot formation and burn-up processes estimated by the simplified in-cylinder two-zone model. It was found that the diffusion burning stage usually started at a time close to the first maximum heat release, and a good correlation was shown between the measured exhaust smoke density and the integrated quantity of apparent diffusion combustion.

The authors have developed a method to analyze the turbulence scale accurately from the descrete data, which are usually obtained by LDV measurements, at low sampling rate with non-uniform time intervals. In this method, the turbulence correlation is calculated from the data of many engine cycles, and successively the turbulence spectrum is calculated from the inverse Fourier-transform of the turbulence correlation. The reliability of this method is confirmed by the model data sampled randomly on a sinusoidal process, and by the data measured in the cylinder of a motored diesel engine having a direct injection system. Furthermore, it is clearly shown that the spectrum in the frequency range higher than the Nyquist frequency defined by the mean data rate can be analyzed by this method, and that it is not necessary to change the range of frequency analysis despite a little variation of mean sampling rate.

半開放形遠心羽根車のシュラウドすきまに基づく効率減少を入力の減少と圧力損失の両面から検討を加え, 種々の実験データから入力の減少は出口すきま比および羽根出口角の余接に比例すること, またすきま損失はのど部すきま比の0.6乗および相対速度の減速に基づく圧力上昇に比例することを明らかにし, 実験式を提案した. しかもこれらの関係を用いればすきまに基づく効率減少がかなりの精度で予測できることを示した.

The velocity distribution is measured at the exit of two different types of unshrouded centrifugal impellers under four different tip clearance conditions each, one with twenty radial blades and inducers and the other with sixteen backward-leaning blades. The secondary flow induced by the tip leakage is similar in both impellers, and the underturning near the blade tip due to the stationary shroud is reduced significantly with the increase of tip clearance. The local change of angular momentum due to the tip clearance is related to the flow rate ratio of leakage-flow and through-flow, and an almost proportional relationship between them is found. Despite a similar secondary flow in both impellers, the input power is hardly changed in the radial blades impeller and is decreased in the backward-leaning blades impeller by the increase of tip clearance.

Swirl, turbulence and the cyclic fluctuation have been examined using LDV in a motored direct injection diesel engine with a spiral port, a troidal piston cavity and a bore of 102 mm. Measurements were made at a low engine speed of 600 rpm at nine points along a cylinder diameter. The authors have developed a new method to analyze turbulence accurately even under the low data rate condition by introducing a concept of confidence interval. It has been clearly shown that there exists a longitudinal vortex in the piston cavity, the tangential velocity distribution shows a forced vortex pattern, and the swirl ratio increases rapidly in the piston cavity. And turbulence in a compression stroke was not always uniform in the cylinder, but in-cylinder averaged turbulence and the cyclic fluctuation were almost the same between the tangential and axial components.

The airflow was measured by a LDV in a motored four-stroke direct injection diesel engine with a troidal cavity combustion chamber, under three different swirl intensity condition. The turbulence scale was calculated by a random sample data analysis utilizing multicycle velocity data. It has been clearly demonstrated that even though the swirl intensity was varied, the integral time scale increased once at the beginning of the compression stroke and decreased toward compression top dead center. On the other hand, the micro time scale hardly changed in suction and compression strokes. Furthermore, it has been revealed that the axial micro time scale was almost uniform in the cylinder under the high swirl condition, but the integral time scale remained large near the piston cavity center.

Cylinder pressure is measured simultaneously at several points in the combustion chamber of a high speed diesel engine with a toroidal piston cavity by means of small-sized piezo pressure transducers and a multichannel combustion analyzer with an accuracy of one fourth degree of crank angle. It has been clearly shown on the cylinder pressure behavior that the farther the distance of measuring point from the cavity center, the latter the rapid pressure rise and the larger the pressure fluctuation amplitude. The time difference of the rapid pressure rise between two measuring points is identical to the propagation time of a pressure wave originated in the piston cavity. The peak frequencies in the cylinder pressure spectrum, which correspond to the two kinds of wavelengths of cylinder bore and two times the root mean square diameter, are due to standing waves in the combustion chamber.

羽根車の翼端すきまに基づく圧力損失は、すきまからの漏れに基づくものと、翼端の環状すきま内の流体を圧力こう配に抗して支えるための損失とからなるとして圧力損失の式を導き、予測される圧力損失を2種類の遠心羽根車の実験データと比較した。更にその特別な場合として軸流羽根車のすきま損失の式を導き、文献の七つの場合について実験と比較した。いずれの場合にもほぼ満足できる結果が得られている。