森吉 泰生

<p>It is required to reduce both the friction and lubricant oil consumption (LOC) in order to reduce CO₂ in internal combustion engines. In addition, as the requirement to reduce particulate number (PN) may be realized by reducing LOC, the relationship between LOC and PN should be examined. Elucidation of the mechanism of LOC belongs to a basic and applied field and a common subject, however, only the LOC is measured by the weight method in steady conditions and the mechanism is not clarified so far. Thereby, the relationship between LOC and PN was measured in a wide operation area including transient conditions using a production engine in this study.</p>

In this research, the improvement of mixing and pulsation in exhaust manifold with a design and implementation of bypass adapter at exhaust port were deeply investigated. This in-turn can improve the post-oxidation phenomena and hence emissions and engine performance could be enhanced. This research investigation includes 1-D, 3-D simulations and experimental validation on a 4-cylinder turbocharged spark ignition (SI) engine. Firstly, the 1-D and 3-D simulation models were developed and calibrated with the experimental results. Then, the simulations were used for the detailed investigation of mixing and pulsation in exhaust manifold with and without bypass adapter. Thereafter, experimental test for the post-oxidation were conducted with and without consideration of the bypass adapter and results were compared. From the simulation and experimental results, it was proven that by using bypass adapter at the exhaust port, the mixing of exhaust gas species was observed to be significantly improved to some extent. Also, the unbalance between exhaust port and turbocharger upstream gas species were reduced. This also reduced the exhaust gas pulsation. By the improvement of mixing between scavenged O and unburned gas species, the post-oxidation reaction was also noted to have improved and consequently the emissions and turbo-speed were found to be better that led to an improved IMEP and thermal efficiency of the engine. 2

The stochastic pre-ignition phenomenon plays a vital role to limit the further increasing BMEP for natural gas engines. In this study, the pre-ignition propensities were examined in a highly boosted premixed natural gas engine by various engine loads and air/fuel ratios, as well as different methane number (MN) altered by hydrogen addition. A proper pre-ignition evaluation method was proposed referring to intake temperature. Moreover, the limits of in-cylinder temperature and pressure for the onset of pre-ignition were estimated. The results show that both higher IMEP and richer mixture conditions readily lead to pre-ignition. The significant increases of pre-ignition frequency and heavy-knocking pre-ignition cycle present with lowering MN.

In this research, a novel methodology for the post-oxidation in a turbocharged spark ignition (SI) engine is proposed and investigated that can improve the emissions along with the reduction in turbo-lag. In this research, both simulation and experimental activities are performed. The 1-D simulation model was used for the identification of efficient scavenging. Thereafter, experimental validation tests for modeling and post oxidation were conducted on a 4-cylinder turbocharged SI engine. From the results, it was revealed that efficient scavenging and post-oxidation can be obtained at lower speed and higher load. The enthalpy in exhaust manifold increased due to the post-oxidation reaction which in turn increased the temperature and pressure of the exhaust gases and hence emissions reduced. Also, due to the increased enthalpy at turbine upstream, the turbocharger speed increased and as a consequence, reduction in the turbo-lag was observed. It was also noted that the post-oxidation is limited at higher load and overlap in an inline 4-cylinder engine due to the strong scavenging which increased the cooling effect in in-cylinder and exhaust manifold due to excess air.

Effects of measurement method, coolant temperature and fuel composition on soot emissions were examined by engine experiments. By reducing the pressure fluctuation in the sampling line, the measured soot emissions with better stability and reproducibility could be obtained. With lower coolant temperatures, larger soot emissions were yielded at much advanced fuel injection timings. Compared to gasoline, soot emissions with a blend fuel of normal heptane, isooctane and toluene were significantly decreased, suggesting the amounts of aromatic components (toluene or others) should be increased to obtain a representative fuel for the predictive model of particulate matter in SIDI engines.

Resistive particulate matter sensor (PMS) is a promising solution for the diagnosis of diesel/gasoline particulate filter (DPF/GPF) functionality. Frequently triggered regeneration of their sensing element, for cleaning the soot dendrites deposited on the surface, leads to experience high temperature and thermal stress and pose high risk of developing cracks in the electrodes or sensing substrate. A semiconductor with a dopant concentration of 100 ppm~10000 ppm is applied as a sensing element for PMS self-diagnosis. Upon cooling at air, the polarization doped-insulating layer in a resistive PMS starts to resume the electrical conductivity in the wake of experiencing high regeneration temperature, through the electron and hole directional mobility. At a temperature slightly lower than carbon equilibrium temperature of 600°C, the charge carriers' migration rate reaches the maximum value and then steps down to near zero with cooling of the sensing element due to the carrier thermal movement and polarization. On the other hand, during percolation/quiescent process, the charged soot dendrites appear to be anchored between the gap of electrodes, not well interconnected but rather standing independent of each other. No conductive path forms between electrodes, but the mobility of charge carrier between carbonaceous particulate matter and sensing element suddenly increases and then is kept constant (saturation drift current occurrence) until the first soot conductive path is formed. This article tries to figure out electron transferring pathways from amorphous carbon to the insulating layer doped by a defined dopant and to shed some light on the mechanism of choking phenomenon.

<p>The "Innovative Combustion Technology" program, a national project is established under the Cabinet Office, Government of Japan as a part of the" Cross-ministerial Strategic Innovation Promotion Program (SIP)". The "Gasoline Combustion Team" is one of teams of the "Innovative Combustion Technology" program.</p><p>The "Gasoline Combustion Team" is comprised of Keio University as a Leader university and 29 universities as a Cluster university. Upon agreement with the Japan Science and Technology Agency (JST), we have been conducting the research on the "Super-Lean Burn for Gasoline Engines" with a support of the Research Association of Automotive Internal Combustion Engines (AICE) under the strong industry, academia and government collaboration.</p><p>This paper introduces both scientific and technological approaches to innovative combustion technologies to realize the "Super- Lean Low Temperature Combustion", targeting a thermal efficiency of 50%.</p>

<p>The prechamber combustion characteristics were studied using a rapid compression and expansion machine (RCEM) to improve the efficiency of cogeneration natural gas engines. The torch flames generated by a prechamber were used to investigate the effect that a prechamber has on the main combustion. In our previous study, we observed the correlation between the torch flame and the main flame (which is a so-called "prechamber combustion") as well as the knocking phenomena for various prechamber configurations. In this study, in order to investigate the effects of engine size on prechamber combustion characteristics, large size RCEM were developed.</p>

<p>The prechamber combustion characteristics were studied using a rapid compression and expansion machine (RCEM) to improve the efficiency of cogeneration natural gas engines. The torch flames generated by a prechamber were used to investigate the effect that a prechamber has on the main combustion. In our previous study, we observed the correlation between the torch flame and the main flame (which is a so-called "prechamber combustion") as well as the knocking phenomena for various prechamber configurations. In this study, we have investigated the effect of prechamber combustion on main chamber combustion characteristics using a constant volume combustion chamber.</p>

<p>Real driving conditions change because of road, traffic environments, weather and driver's characteristics. These changes cause differences between real driving and regulations. To improve the differences, a real driving emission (RDE) regulation has been introduced and implemented in Europe for diesel passenger vehicles. Similar regulation is expected to be introduced in Japan from 2022. Since environmental conditions as mentioned above vary real-driving test results, the reproducibility of experiments is low. In order to figure out the variation, it is necessary to carry out a large number of tests which take a lot of time and cost. On the other hand, a prediction method using simulation model can efficiently evaluate the emissions reflecting various driving patterns and environmental conditions. In the authors' previous study, prediction methods of NOx emission were developed and verified using a diesel passenger car. However, the verification was carried out under equivalent environment conditions. In this study, the accuracy of measured and predicted data was confirmed by considering the environmental factor in the prediction model. In case of considering the environmental factor, the predictability of model was increased by comparing the previous model.</p>

直噴ガソリンエンジンのすす生成モデルの構築のための基礎データとして、エンジン実験におけるすす測定の安定化を試みた。測定位置や圧力変動低減のためチャンバーの有無などを変化させることで、すす排出挙動の異なる条件であってもすすの排出量と粒子数および粒径分布を安定かつ再現性よく測定することが可能となった。

導入が予定されているRDE試験は排出ガス規制である一方、同時に燃費の測定も可能である。その際に得られる燃費は、モード燃費や実燃費と比較してどのような水準となると見込まれるのか、直噴ターボガソリン車を用いてRDE試験法に準拠したコースを走行して比較を行い、とくに実燃費との関係について考察した。

ディーゼル機関の後燃え要因である噴霧先端の過濃混合気塊生成抑制のため，主燃料分割噴射の効果を調査した．分割によりメイン噴射量が低減しキャビティ内の過濃混合気塊生成が抑制されること，及びアフター噴射燃料は高温場で余剰空気と共に急速燃焼することで後燃えが低減し，熱効率向上要因となっていることが判明した．

火花点火機関の高希釈高過給化により，点火環境は厳しくなる一方である．スパークプラグでの放電路挙動と予混合気への点火性能の関係を明らかにすることが重要である．本研究では定容燃焼容器を用い，希釈流動場における放電と着火性の関係を調査した．各放電パラメータの着火性への寄与度は，流動強度により差異が確認された．

<p>Cycle-to-cycle variation (CCV) of in-cylinder flow occurs in internal combustion engines. It is necessary to analyze CCV of flow to separate averaged-flow (as low frequency / low wave number) from turbulence (as high frequency / high wave number), because an averaged flow varies from cycle to cycle. Two averaging methods are used for the extraction of mean component from instantaneous flow. One is temporal-averaging method, the other is spatial-averaging method. In the temporal -averaging method, a fluctuation of flow is captured at fixed point in Eulerian, turbulence is regarded as the high frequency component, and it is removed by a low pass filtering. In the spatial-averaging method, the turbulence in spatial arrangement of flow velocity is directly averaged by using vortex scale as a threshold (e.g. Moving-averaging filter and Gaussian-averaging filter). However, the temporal-averaging and the spatial-averaging have completely different characteristics. Therefore, it is necessary to clarify the difference of filtering characteristics in each averaging filter. In this study, comparisons of averaged flow patterns of temporal-average and spatial-average are carried out. Moreover, variable sized spatial filter which is based on Taylor's frozen-turbulence hypothesis is proposed. As a result, variable sized filtering is found close to the filter characteristic of the time average method.</p>

実路走行時におけるNOx排出量を予測するシミュレーションについて，著者らは既研究で，NOx低減触媒を装着していない車両で実路走行時の予測を可能とした．本研究では， NOx吸蔵還元触媒を装着した車両を対象に，簡易な予測モデルを検討し，実路走行時の排出量と比較・検証を行った.

The purpose of this paper is to find a way to extend the high load limit of homogeneous charge compression ignition (HCCI) combustion. This paper presents the effect of in-cylinder flow and stratified mixture on HCCI combustion by experiments and three-dimensional computer fluid dynamics coupled with a detailed chemical reaction calculation. The first study was conducted using a rapid compression and expansion machine (RCEM) equipped with a flow generation plate to create in-cylinder turbulent flow and with a control unit of in-cylinder wall temperature to create in-cylinder temperature distribution. The study assesses the effect of the turbulent flow and the temperature distribution on HCCI combustion. In the second study, the numerical simulation of HCCI combustion was conducted using large eddy simulation coupled with a detailed chemical reaction calculation. The study analyzes the interaction between in-cylinder turbulent flow and mixture distribution and HCCI combustion. The result shows that turbulent flow effects a change in the distribution of chemical species which is a trigger of combustion initiation.

<p>DI Gasoline engine tends to increase the particulate mass concentration in the case of cold start, warming-up process and high load condition. In this study, PM emission was investigated from cold start to warming-up stage. The engine coolant temperature is set to 8°C, 30°C and 80°C, and start of fuel injection is changed from -320 deg.ATDC and -90 deg.ATDC. The excess air ratio, the load (gross indicated mean effective pressure, Gross IMEP) and the combustion phase (CA 50) was adjusted to 1.0 and 0.7 MPa, ≈ 9 deg. ATDC, respectively. The fuel injection pressure was kept constant at 10 MPa. As a result, when the coolant temperature was changed to 8°C, 30°C and 80°C, it was found that the soot at the coolant temperature of 8°C increased. In addition, it was found that soot was less affected by coolant temperature from -300 deg. ATDC to -120 deg. ATDC. Soot tends to increase at the injection condition -320 deg. ATDC and -90 deg. ATDC where the piston top surface is approaching the injector. The soot increased because fuel adhered to the piston top surface and pool combustion occurred. Also at -90 deg.ATDC, soot formation from the rich region due to mixing failure is considered to be an influencing factor.</p>

今後乗用車に導入される可能性があるパワートレイン関連の新技術を導入した場合のWLTC走行時の燃費改善効果を予測するために，燃費シミュレーションツールに新たなモデルを導入し，その予測精度について実機を用いて検証を行った．次いで，このモード走行燃費シミュレーションコードを使って，各新技術の燃費改善効果を評価した．

火炎伝播の過程を観察する手法としてシュリーレン法が広く知られている．しかし火花放電による着火の場合，プラズマ形成や予熱による密度変化の影響で初期火炎の形状や大きさが判別できない．そこで本稿では可視域高速度カメラと赤外高速度カメラを用い，放電路と初期火炎の形成の様子を捉えることを試みた．

<p>To increase thermal efficiency of internal combustion engine, lean burn and EGR (Exhaust Gas Recirculation) system have been developed with spark ignition coils generating larger discharge current and discharge energy than current mass production coils. Several researches clarified that larger discharge current increases discharge channel extension and decreases possibility of discharge channel blow-off and possibility of misfire. However, these investigations don't mentioned effect of larger discharge current and energy on air-fuel ratio and combustion period. Then purpose of this research is to investigate relation among air-fuel ratio, combustion period and coil specification in order to clarify control factor of air-fuel ratio of lean burn. In this study, five coils having different current profiles were evaluated under 2000 rpm and 0.6 MPa NMEP (Net indicated Mean Effective Pressure) at lean mixture condition by combustion test and in-cylinder optical measurement test with research single cylinder engine. The combustion test results showed a correlation between lean limit air-fuel ratio and initial combustion period. Moreover, optical measurement test showed that initial combustion period has a correlation with discharge energy before 1st restrike and discharge channel extension rate and variation of initial combustion period under stable control condition doesn't depend on discharge current.</p>

<p>Cycle-to-cycle variation (CCV) of combustion is an important issue because it affects emissions and drivability. Improvement of CCV of combustion has been carried out using electronic controls (e.g. ignition timing, fuel injection and variable valve timing) in motor vehicle's engines. However, electronic devices are hardly used for motorcycle's engines because of limited space and cost. Therefore, the engine performance itself must be improved to reduce CCV of combustion in motorcycle. Though CCV of combustion is caused by CCV of in-cylinder flow pattern, fuel distribution, temperature and residual gas, and ignition energy, it is difficult to measure and analyze these factors. In this study, the simultaneous measurement of high-speed PIV and direct photographing of flame propagation was carried out. CCV of in-cylinder flow was evaluated as temporally-averaged flow that was obtained by instantaneous flow using low-pass filtering and cut-off frequency. As a result, in-cylinder temporally-averaged flow pattern fluctuated between individual cycles. Especially, the flow pattern on the surface of piston at BDC was different between the highest and the lowest cycle in IMEP. This difference is considered to be due to the location offset of tumble flow. Also the fluctuation of turbulence kinetic energy (TKE) is caused by tumble flow offset. TKE distribution near the spark plug at ignition timing affected the direction and speed of flame propagation.</p>

大気環境のさらなる改善のためにディーゼルエンジン排出ガスからの大幅なNOx低減が求められる中、弊社は平成28年排出ガス規制に適応したNOxの還元剤として軽油を用いる尿素フリーの後処理システム（DPR-II）を実用化した。本システムの排出ガス低減技術を発表する。

ディーゼル機関の後燃え低減による熱効率向上を目指し，市販のディーゼル機関の燃焼を紫外自発光撮影手法により可視化することで，後燃えの熱発生領域を調査した．三次元数値計算結果と共に考察した結果，後燃え期間中の熱発生は，噴霧火炎先端に形成された過濃混合気が滞留し，その外周部で発生している．

車両燃費シミュレーションではトランスミッション等のサブシステムをモデル化する必要がある．本研究は，実車での計測により車両燃費シミュレーションに必要なトランスミッション等のモデルを簡易に構築する手法を検討した．さらに，本シミュレーションを活用してタイヤ性能の違いが燃費に及ぼす影響を定量的に評価した．

<p>It is required to examine the characteristics of turbochargers for automobiles using one-dimensional simulation from the viewpoint of estimating total engine performance. In this study, a mathematical model to predict mechanical loss generated in a turbocharger is proposed. Friction works generated in a journal bearing and a thrust bearing are modeled, separately. As for the calculation of the friction work with the thrust bearing, the thrust force is calculated from the fluid force which is formulated analytically and calculated numerically based on one-dimensional flow taking account of relevant boundary conditions. According to the developed model, the friction work generated in a journal bearing is larger than that in a thrust bearing. Difference of thrust force and flow rate of oil has less impact on the friction works in a turbocharger. Finally, calculated total friction work based on the model proposed in the present study is compared with that obtained from the oil temperature method.</p>

NOx emissions from diesel passenger vehicles affect the atmospheric environment. It is difficult to evaluate the NOx emissions influenced by environmental conditions such as humidity and temperature, traffic conditions, driving patterns, etc. In the authors' previous study, real-driving experiments were performed on city and highway routes using a diesel passenger car with only an exhaust gas recirculation system. A statistical prediction model of NOx emissions was considered for simple estimations in the real world using instantaneous vehicle data measured by the portable emissions measurement system and global positioning system. The prediction model consisted of explanatory variables, such as velocity, acceleration, road gradient, and position of transmission gear. Using the explanatory variables, NOx emissions on the city and highway routes was well predicted using a diesel vehicle without NOx reduction devices. However, the prediction model had some limitations owing to the effects of NOx reduction devices. In this study, among various NOx reduction systems, a diesel vehicle with NOx storage catalyst (NSC) was chosen to predict the NOx emissions under a catalytic system. To improve the accuracy of the NOx emissions under the NSC, a catalyst model was added to the prediction model and used to predict the catalyst properties. By adding the catalyst model, the accuracy of NOx emissions of the prediction model was improved compared to the previous prediction model with individual explanatory variables. The NOx emissions were well predicted compared to the measured data.

To drastically improve thermal efficiency of a gasoline spark-ignited engine, super-lean burn is a promising solution. Although, studies of lean burn have been made by so many researchers, the realization is blocked by a cycle-to-cycle combustion variation. In this study, based on the causes of cycle-to-cycle variation clarified by the authors' previous study, a unique method to reduce the cycle-to-cycle variation is proposed and evaluated. That is, a bulk quench at early expansion stroke could be reduced by making slight fuel stratification inside the cylinder using the twin-tumble of intake flows. As a result, the lean limit was extended with keeping low NOx and moderate THC emissions, leading to higher thermal efficiency.

Among the emerging technologies in order to meet ever stringent emission and fuel consumption regulations, Exhaust Gas Recirculation (EGR) system is becoming one of the prerequisites particularly for diesel engines. Although EGR cooler is considered to be an effective measure for further performance enhancement, exhaust gas soot deposition may cause degradation of the cooling. To address this issue, the authors studied the visualization of the soot deposition and removal phenomena to understand its behavior. Based on thermophoresis theory, which indicates that the effect of thermophoresis depends on the temperature difference between the gas and the wall surface exposed to the gas, a visualization method using a heated glass window was developed. By using glass with the transparent conductive oxide: tin-doped indium oxide, temperature of the heated glass surface is raised. This new method was applied to an EGR cooler, which successfully enabled real-time visualization without soot deposition on the window surface. Consequently it was visually confirmed that the soot adheres to the heat exchanging surface. This observation also confirms that basic mechanism of the deposition is due to the thermophoretic phenomenon. Furthermore, there were notable findings observed in the experiment on running a diesel engine. First, position of water condensation occurrence depends on the local thickness distribution within the soot deposit. In addition, the size of the soot peeling area can be effectively increased by inducing high gas velocity with Vortex Generator fin (VG fin) when condensed water exists. The VG fin has heat transfer promoting effect by generated vortex along the gas flow direction on the heat transfer surface. In this paper, these observations are discussed in detail as well as the development of the visualization method.

To understand the mechanism of the combustion by torch flame jet in a gas engine with pre-chamber and also to obtain the strategy of improving thermal efficiency by optimizing the structure of pre-chamber including the diameter and number of orifices, the combustion process was investigated by three dimensional numerical simulations and experiments of a single cylinder natural gas engine. As a result, the configuration of orifices was found to affect the combustion performance strongly. With the same orifice diameter of 1.5mm, thermal efficiency with 7 orifices in pre-chamber was higher than that with 4 orifices in pre-chamber, mainly due to the reduction of heat loss by decreasing the impingement of torch flame on the cylinder linear. Better thermal efficiency was achieved in this case because the flame propagated area increases rapidly while the flame jets do not impinge on the cylinder wall intensively. This means that the optimization of orifice diameter and its number is quite important to enhance the flame propagation without increasing heat loss.

<p>The effects of fuel injection timing and coolant temperature on soot and particulate emissions at stoichiometric mixture condition under heating mode were examined in a single cylinder spark ignition direct injection gasoline engine. Single injection was employed and the start of fuel injection timing was adjusted in the intake stroke. The gross indicated mean effective pressure and the 50% heat release angle were set to 0.7 MPa and 9 deg.ATDC respectively by adjusting the intake air pressure, injected fuel mass and ignition timing. Regardless of coolant temperatures, with advancing the injection timing close to the top dead center, the exhausted soot mass concentration and particulate number are increased, mainly due to more liquid fuel impinging the combustion chamber walls and resulting in wider and more rich fuel-air mixture around the walls. Regardless of fuel injection timings, with decreasing coolant temperature, the exhausted soot mass concentration and particulate number are increased. To maintain the same load and combustion phasing, more amount of liquid fuel should be injected, due to the lower in-cylinder temperature and wall temperture with lower coolant temperature, resulting in more unevaperated fuel and wider zone for rich fuel-air mixture near the combustion chamber walls. Compared to coolant temperature, the injection timing should be well designed to decrease the particulate matter emission.</p>

<p>This article shows a research on the effect of in-cylinder flow on thermal stratified homogeneous charge compression ignition (HCCI) combustion to extend a high load limit of HCCI combustion. The study was conducted by three dimensional computational fluid dynamics coupled with simulation of chemical reactions. The in-cylinder temperature distribution was set as an initial condition. The in-cylinder flow was added before low temperature oxidation (LTO) and during high temperature oxidation (HTO), respectively. The results show that the in-cylinder flow before LTO leads to slow combustion due to heat loss to combustion chamber walls. The results also show that the in-cylinder flow during HTO is effective in fast combustion. This is mainly because in-cylinder flow expands hot burned mixture to unburned mixture, which accelerates the combustion.</p>

Reduction in the cycle-to-cycle variation (CCV) of combustion in internal combustion engines is required to reduce fuel consumption, exhaust emissions, and improve drivability. CCV increases at low load operations and lean/dilute burn conditions. Specifically, the factors that cause CCV of combustion are the cyclic variations of in-cylinder flow, in-cylinder distributions of fuel concentration, temperature and residual gas, and ignition energy. However, it is difficult to measure and analyze these factors in a production engine. This study used an optically accessible single-cylinder engine in which combustion and optical measurements were performed for 45 consecutive cycles. CCVs of the combustion and in-cylinder phenomena were investigated for the same cycle. Using this optically accessible engine, the volume inside the combustion chamber, including the pent-roof region can be observed through a quartz cylinder. CCV of in-cylinder flow for 45 continuous firing cycles were measured by Time-Resolved Particle Image Velocimetry (TR-PIV) technique. The in-cylinder flow was measured at intervals of 2 crank angle degrees from the intake to compression strokes using a dual-cavity, high-frequency Nd:YLF laser. In order to analyze the CCV, the measured instantaneous flow was converted to a time-averaged flow by low-pass filtering to remove the high-frequency component. Moreover, CCVs of fuel distribution at intake valve closing (IVC) and just before ignition timing were obtained by Planar Laser Induced Fluorescence (PLIF) technique. The fourth harmonic generation of a dual-cavity Nd:YAG laser was used as the excitation light source. 3-Pentanone, which was mixed with iso-octane and injected to the intake port, was used as a PLIF tracer. These two visualization techniques were applied simultaneously during the continuous firing cycles. As a result, it was confirmed that the CCVs of in-cylinder flow and fuel distribution significantly affect the CCV of combustion at low-load conditions. In particular, flow in a direction opposite to the tumble flow was observed in the lowest load cycle.

The prechamber combustion characteristics were studied using a rapid compression and expansion machine (RCEM) to improve the efficiency of cogeneration natural gas engines. The torch flames generated by a prechamber were used to investigate the effect that a prechamber has on the main combustion. In this study, we focused on observing the correlation between the torch flame and the main flame (which is a so-called “prechamber combustion”) as well as the knocking phenomena for various prechamber configurations.

In recent years, improvements in the fuel economy and exhaust emission performance of internal combustion engines have been increasingly required by regulatory agencies. One of the salient concerns regarding general purpose engines is the larger amount of CO emissions with which they are associated, compared with CO emissions from automobile engines. To reduce CO and other exhaust emissions while maintaining high fuel efficiency, the optimization of total engine system, including various design parameters, is essential. In the engine system optimization process, cycle simulation using 0-D and 1-D engine models are highly useful. To define an optimum design, the model used for the cycle simulation must be capable of predicting the effects of various parameters on the engine performance. In this study, a model for predicting the performance of a general purpose SI (Spark Ignited) engine is developed based on the commercially available engine simulation software, GT-POWER. The developed 1-D engine model was validated using a 3-D CFD (Computed Fluid Dynamics) simulation and experimental results. The effects of engine speed and load, air-fuel ratio (A/F), spark ignition timing on combustion characteristics were simulated. The simulation results were compared with experimental results to determine the accuracy of the developed model.

An improvement of thermal efficiency of internal combustion engines is strongly required. Meanwhile, from the viewpoint of refinery, CO2 emissions and gasoline price decrease when lower octane gasoline can be used for vehicles. If lower octane gasoline is used for current vehicles, fuel consumption rate would increase due to abnormal combustion. However, if a Homogeneous Charge Compression Ignition (HCCI) engine were to be used, the effect of octane number on engine performance would be relatively small and it has been revealed that the thermal efficiency is almost unchanged. In this study, the engine performance estimation of HCCI combustion using lower octane gasoline as a vision of the future engine was achieved. To quantitatively investigate the fuel consumption performance of a gasoline HCCI engine using lower octane fuel, the estimation of fuel consumption under different driving test cycles with different transmissions is carried out using 1D engine simulation code. As a result, combining high compression ratio and Continuously Variable Transmission (CVT) can improve the fuel consumption in HCCI/SI combustion even using lower octane gasoline.

PFI (Port Fuel Injection) gasoline engines for motorcycles have some problems such as slow transient response because of wall wet of fuel caused by the injector's layout. Hence, it is important to understand the characteristics of fuel sprays such as droplet size and distribution of fuel concentration. Considering the spray formation in a port, there are three kinds of the essential elements: breakup, evaporation and wall impingement. However, it is difficult to observe three of them at the same time. Therefore, the authors have made research step by step. In the authors' previous study, the authors focused on the wall collision, droplet sizes, droplet speeds and the space distribution of the droplets. In this study, the authors focused on evaporation. A direct sampling method using FID (Flame Ionization Detector) for evaporating fuel was established and the concentration distribution of evaporating fuel in the port was measured and analyzed. As a result, it was found that higher velocity in the port increases fuel concentration with enhanced atomization and that evaporating fuel is easier to be affected by the flow and fuel distillation characteristics.

In order to realize the high compression ratio and high dilution combustion toward improvement in thermal efficiency, the improvement in stability of ignition and initial phase of combustion under the high gas flow field is the major challenge. In terms of the shift on the higher power side of the operating point by downsizing and improvement of real world fuel consumption, the improvement of ignitability is increasingly expected in the wide operating range also including high load and high engine speed region. In this study, the effects of the gas pressure, gas flow velocity near the spark gap at ignition timing, and discharge current characteristics on spark channel formation were analyzed, focusing on restrike event and spark channel stretching in the spark channel formation process. And the relationship between the average discharge current until 1 ms and the EGR combustion limit was considered.

The effect of properties of lubricating oil on low speed pre-ignition (LSPI) was investigated. Three different factors of oil properties such as cetane number, distillation characteristics and Calcium (Ca) additive (with and without) are prepared and examined. Then actual engine test of LSPI was carried out to evaluate the effect and to clarify the mechanism and role of lubricating oil. Finally it is clarified that the oil cetane number and/or Ca additive strongly affect LSPI phenomena.

In this study, the effect of the low octane number fuel on HCCI engine performance was experimentally investigated using a slightly modified commercial four-cylinder gasoline engine. To operate the engine in HCCI strategy with wide operational range, the blowdwon supercharging (BDSC) system proposed by the authors was applied in the test engine. Research octane number (RON) of test fuels was varied from 90 to 78.5 as an experimental parameter. Experimental results showed that in the range of the present study, HCCI operational range, brake thermal efficiency and exhaust emissions during HCCI operation were little affected by the RON of the test fuels. In contrast, during SI operation, thermal efficiency was deteriorated with lower RON fuel because of knocking. In terms of "Well-to-Wheel" energy consumption, it is favorable to use gasoline without any additional treatment for increasing fuel octane number, because additional energy is necessary for increasing octane number of commercial gasoline. It is expected that the "Well-to-Wheel" CO emission can be reduced by application of high efficiency HCCI combustion into production engine operation with low octane number fuel. 2

The authors investigated the reasons of how a preignition occurs in a highly boosted gasoline engine. Based on the authors' experimental results, theoretical investigations on the processes of how a particle of oil or solid comes out into the cylinder and how a preignition occurs from the particle. As a result, many factors, such as the in-cylinder temperature, the pressure, the equivalence ratio and the component of additives in the lubricating oil were found to affect the processes. Especially, CaCO included in an oil as an additive may be changed to CaO by heating during the expansion and exhaust strokes. Thereafter, CaO will be converted into CaCO again by absorbing CO during the intake and compression strokes. As this change is an exothermic reaction, the temperature of CaCO particle increases over 1000K of the chemical equilibrium temperature determined by the CO partial pressure. The possibility of a preignition due to particles including CaCO particles is numerically simulated comparing with the experimental results. 3 3 2 3 2 3

In this study, in order to clarify the mechanism of preignition occurrence in highly boosted SI engine at low speed and high load operating conditions, directphotography of preignition events and light induced fluorescence imaging of lubricant oil droplets during preignition cycles were applied. An endoscope was attached to the cylinder head of the modified production engine. Preigntion events were captured using high-speed video camera through the endoscope. As a result, several types of preignition sources could be found. Preignition caused by glowing particles and deposit fragments could be observed by directphotography. Luminous flame was observed around the piston crevice area during the exhaust stroke of preignition cycles. This implies that the lubricant oil or mixture of oil and liquid fuel which is accumulated in the piston crevice area burns under low oxygen condition, and that the glowing particles which induce preignition would be produced in the oil combustion during the expansion stroke. Preignition induced by a lubricant oil droplet ignition was probably captured using fluorescence technique. Through the investigations, preignition which occurs at the piston crevice area was often observed. It was also found that preignition event often terminated in a single cycle in the present study and hardly occurred in the sequential manner including normal SI combustion as has been frequently reported in the previous studies. This is probably because a scavenging efficiency in the engine operating condition in this experiment is high. In the conventional boosted gasoline engine, an exhaust pressure increases, and some of the particles generated during first preignition cycle are not scavenged and remain inside the cylinder. The residual particles are heated during the subsequent combustion cycle and induce the preignition again. Preignition event continues until the residual particles are consumed. This would be one of the possible mechanism of the sequential preignition occurrence.