Search Results

Although whether the cylinder gas oscillation is provoked by end-gas autoignition in a certain cycle or not is a irregular phenomenon, autoignition itself takes place in almost all of the cycles in the knocking condition. Detection of the autoignition makes it possible to realize a knock anticipating strategy. Using the decay rate of the effective heat release rate as the index, delayed autoignition with small auto-ignited mass fraction can be detected. Applying this index for the analysis of the autoignition in the acceleration process, it was clarified that heavy autoignition immediately after the acceleration caused by the selective induction of the low boiling point gasoline components into the cylinder is followed by the period where the low combustion chamber wall temperature reduces the autoignited mass fraction and suppresses the cylinder gas oscillation.

The ignition delay times of n-C7H16, i-C8H18, and a blend of them at different fuel, O2 and N2 concentrations were computed using a detailed chemical kinetic mechanism generated by KUCRS. For each fuel, the dependences of ignition delay time on fuel, O2 and third body concentrations and on the heat capacity of a mixture were distilled to establish a power law equation for ignition delay time. For n-C7H16, ignition delay time τhigh without low-temperature oxidation at a high initial temperature between 1000 K and 1200 K was expressed using the scaling exponents for fuel, O2 and third body concentrations and heat capacity of 0.54, 0.29, 0.08, and - 0.38, respectively. Low-temperature oxidation induction time τ1 at a low initial temperature between 600 K and 700 K was expressed using the scaling exponents for fuel, O2 and third body concentrations and heat capacity of 0.03, 0.18, 0.04, and - 0.17, respectively.

Ignition process chemistry was analyzed using a detailed chemical kinetic model of n-heptane generated by KUCRS (Knowledge-basing Utilities for Complex Reaction Systems), wherein pressure-dependent rate constants of the O₂ addition to alkyl radicals and hydroperoxy alkyl radicals and the thermal decomposition of ketohydroperoxides have been introduced. Then, the effect of the initial pressure and the individual effects of the initial fuel, O₂ and N₂ molar concentrations on a relationship between the initial temperature and the ignition delay were discussed. When the initial temperature increases, the branch of C₇H₁₄OOH removal into the second O₂ addition and the decomposition into C₇H₁₄cyO and OH is more sensitive to the pressure and the O₂ concentration, and thus, the LTO preparation phase is more affected by the pressure and the O₂ concentration. The LTO phase terminates mainly by the OH removal by intermediate species.

Methyl butanoate (MB) and methyl decanoate (MD) are surrogates for biodiesel fuels. According to computational results with their detailed reaction mechanisms, MB and MD indicate shorter ignition delays than long alkanes such as n-heptane and n-dodecane do at an initial temperature over 1000 K. The high ignitability of these methyl esters was computationally analyzed by means of contribution matrices proposed by some of the authors. Due to the high acidity of an α-H atom in a carbonyl compound, hydroperoxy radicals are generated out of the equilibrium between forward and backward reactions of O₂ addition to methyl ester radicals by the internal transfer of an α-H atom in the initial stage of an ignition process. Some of the hydroperoxy methyl ester radicals can generate OH to activate initial reactions. MB has an efficient CH₃O formation path via CH₃ generated by the β-scission of an MB radical which has a radical site on the α-C atom to the carbonyl group.

The ignition delay times and heat release profiles of CH4, C2H6, C3H8, i-C4H10, and n-C4H10 and dual-component CH4-based blends with these alkanes in air were determined using a detailed chemical kinetic model. The apparent activation energy of C2H6 in the relationship between initial temperature and ignition delay time is higher than those of the other alkanes because OH formation is dominated by H2O2(+M)=OH+OH(+M) from the beginning over a wide range of initial temperatures. The heat release rate of C2H6 is higher than those of the other alkanes in the late stage of ignition delay time because H2O2 is accumulated with a higher concentration and promotes the OH formation rate of H2O2(+M)=OH+OH(+M). These ignition characteristics are reflected in those of CH4/C2H6.

Some experimental data indicate that an HCCI process of a highly diluted mixture is characterized with a two-stage profile of heat release after the heat release by low-temperature oxidation, and with slow CO oxidation into CO₂ at a low temperature. In the present paper, these characteristics are discussed using a detailed chemical kinetic model of normal heptane, and based on an authors' idea that an ignition process can be divided into five phases. The H₂O₂ loop reactions mainly contribute to heat release in a low-temperature region of the TI (thermal ignition) preparation phase. However, H+O₂+M=HO₂+M becomes the main contributor to heat release in a high-temperature region of the TI preparation phase. H₂O₂ is accumulated during the LTO (low-temperature oxidation) and NTC (negative temperature oxidation) phases, and drives the H₂O₂ loop reactions to increase the temperature during the TI preparation phase.

To execute the computational fluid dynamics coupling with fuel chemistry in internal combustion engines, simplified chemical kinetic models which capture the low-temperature oxidation kinetics would be required. A steady-state analysis was applied to see the complicated reaction mechanism of alkylperoxy radicals by assuming the steady state for hydroperoxyalkyl (QOOH) and hydroperoxyalkylperoxy (OOQOOH) radicals. This analysis clearly shows the systematic trend of the reaction rate for the chain-branching and non-branching process of alkylperoxy (ROO) radicals as a function of the chain length and the carbon class. These trends make it possible to classify alkylperoxy radicals by their chemical structures, and suggest a reduced low-temperature oxidation chemistry.

A novel leanburn concept, ‘Barrel-Stratification’ is proposed. Fuel is introduced into the cylinder through one of the intake ports of a dual-intake-valve engine of which the tumbling air motion is intensified by the sophisticated intake port design. Because the velocity component in the direction parallel to the axis of tumble is small, charge stratification realized during the intake stroke is maintained until the end of the compression stroke. By the effects of charge stratification and the turbulence enhancement by tumble, stable combustion is realized even at extremely lean conditions. The concept was verified by flow field analysis applying a multi-color laser sheet technique and the flame structure analysis employing the blue-end image intensification realized by the interference mirror and the short delay phosphor.

An extensive effort has been made, at Mitsubishi Motors, in the technology field of new catalysts and of the catalyst reaction control for the purpose of further improvement of the emission control with the GDI engines [1-2]. A new NOx-trap catalyst has been developed to satisfy the required higher catalyst performance under high-temperature condition. The new catalyst contains potassium (K) of excellent NOx-storage capacity under high-temperature region in the catalytic atmosphere, and to retain K stability zeolite is mixed in the catalyst layer as well as the substrate is coated with silica (SiO2). This new catalyst has been proven of the improved NOx conversion efficiency, and solved the long-pending problems particularly those experienced under high-temperature operation.

Gasoline includes various kinds of chemical species. Thus, the reaction model of gasoline components that includes the low-temperature oxidation and ignition reaction is necessary to investigate the method to control the combustion process of the gasoline engine. In this study, a gasoline combustion reaction model including n-paraffin, iso-paraffin, olefin, naphthene, alcohol, ether, and aromatic compound was developed. KUCRS (Knowledge-basing Utilities for Complex Reaction Systems) [1] was modified to produce paraffin, olefin, naphthene, alcohol automatically. Also, the toluene reactions of gasoline surrogate model developed by Sakai et al. [2] including toluene, PRF (Primary Reference Fuel), ethanol, and ETBE (Ethyl-tert-butyl-ether) were modified. The universal rule of the reaction mechanisms and rate constants were clarified by using quantum chemical calculation.

NOx adsorber has already been used for the after-treatment system of series production vehicle installed with a lean burn or direct injection engine [1,2,3]. In order to improve NOx adsorbability at high temperatures, many researchers have recently been trying an addition of potassium (K) as well as other conventional NOx adsorbents. Potassium, however, reacts easily with the cordierite honeycomb substrate at high temperatures, and not only causes a loss in NOx adsorbability but also damages the substrate. Three new technologies have been proposed in consideration of the above circumstances. First, a new concept of K-capture is applied in washcoat design, mixed with zeolite, to improve thermal stability of K and to keep high NOx conversion efficiency, under high temperatures, of NOx adsorber catalyst. Second, another new technology, pre-coating silica over the boundary of a substrate and washcoat, is proposed to prevent the reaction between potassium and cordierite.

Three zone mixture preparation model, assuming that fuel and air are distributed in three separate zones, fuel air and mixture zone, was proposed. Air Utilization Efficiency derived from the model was used to evaluate the mixing nonuniformity. Effect of the large scale nonisotropic turbulence downstream of the dimple or edge in the intake port of MPI engine on the convective mass transfer from fuel film was clarified by the proposed nondimensional index, Local Sherwood Number. It was found that when the fuel is injected toward the wall where large scale turbulence exists, almost all of the fuel is seeded in the air passing the region at the beginning of the intake process, resulting in the time-resolved nonuniformity of the mixture strength at the intake valve. Using the Air Utilization Efficiency, it was elucidated that time-resolved mixing nonuniformity at intake valves induces spatially nonuniform fuel/air distribution in the cylinder.

The relationships between the octane number and the carbon atom number and the molecular structure of alkanes were comprehensively analyzed by using the detailed kinetic model generated by there automatic reaction scheme generation tool, KUCRS [1, 2]. The octane number is an index showing the ignition delay in the engine temperature regime, that is, the engine ignition temperature range. The high octane number is observed in the following two cases; 1 The ignition delay of the low temperature region is large. 2 The ignition delay of the low temperature region is the same, but the transition temperature for NTC (Negative Temperature Coefficient) region is low.

A fuel design concept for an HCCI engine based on chemical kinetics to optimize the heat release profile and achieve robust ignition was proposed, and applied to the design of the optimal methane-based blend. Ignition process chemistry of each single-component of natural gas, methane, ethane, propane, n-butane and isobutane, was analyzed using detailed chemical kinetic computations. Ethane exhibits low ignitability, close to that of methane, when the initial temperature is below 800 K, but higher ignitability, close to those of propane, n-butane and isobutane, when the initial temperature is above 1100 K. Furthermore, ethane shows a higher heat release rate during the late stage of the ignition process. If the early stage of an ignition process takes place during the compression stroke, this kind of heat release profile is desirable in an HCCI engine to reduce cycle-to-cycle variation during the expansion stroke.

One of the authors has proposed to use the decay rate of EHRR, the effective heat release rate, d2Q/dθ2 as an index for the rapid local combustion [1]. In this study, EHRR profiles and the cylinder gas pressure oscillations of the low and high speed knock are analyzed by using this index. A delayed rapid local combustion, such as an autoignition with small burned mass fraction can be detected. In the cases of the low speed knock, it has been agreed that a rapid local combustion is an autoignition. Although whether the cylinder gas oscillation is provoked by an auto ignition in a certain cycle or not is an irregular phenomenon, the auto ignition takes place in almost all of the cycles in the knocking condition. Mixture mass fraction burned by an auto ignition is large. A small auto ignition may induce a secondary auto ignition, in many cases, mass burned by the secondary auto ignition is extremely large.

A systematic chemical lumping method has been proposed, based on the detailed kinetic analysis of hydrocarbon fuel ignitions. The model constructed by using this method contains two reaction sets, RO2 and fragment reaction package. The ignition characteristics of each fuel can be reflected by only adjusting several rate parameters in RO2 reaction package. From the comparison with detailed model, it was confirmed that this simplified model well reproduces the results of detailed one without missing the kinetics of hydrocarbon ignitions. We concluded that this new lumping approach has the possibility to be applicable to every hydrocarbon fuels.

Spray motion visualization, mixture strength measurement, flame spectral analyses and flame behavior observation were performed in order to elucidate the mixture preparation and the combustion processes in Mitsubishi GDI engine. The effects of in-cylinder flow called reverse tumble on the charge stratification were clarified. It preserves the mixture inside the spherical piston cavity, and extends the optimum injection timing range. Mixture strength at the spark plug and at the spark timing can be controlled by changing the injection timing. It was concluded that reverse tumble plays a significant role for extending the freedom of mixing. The characteristics of the stratified charge combustion were clarified through the flame radiation analyses. A first flame front with UV luminescence propagates rapidly and covers all over the combustion chamber at the early stage of combustion.

A two-stage combustion is one of the Mitsubishi GDI™ technologies for a quick catalyst warm-up on a cold-start. However, when the combustion is continued for a long time, an increase in the fuel consumption is a considerable problem. To solve the problem, a stratified slight-lean combustion is newly introduced for utilization of catalysis. The stratified mixture with slightly lean overall air-fuel ratio is prepared by the late stage injection during the compression stroke. By optimizing an interval between the injection and the spark timing, the combustion simultaneously supplies substantial CO and surplus O2 to a catalyst while avoiding the soot generation and the fouling of a spark plug. The CO oxidation on the catalyst is utilized to reduce the cold-start emissions. Immediately after the cold-start, the catalyst is preheated for the minimum time to start the CO oxidation by using the two-stage combustion. Following that, the stratified slight-lean combustion is performed.

The interaction between spark discharge and low-temperature oxidation (LTO) was investigated using an optical compression and expansion machine fueled with n-C7H16 or i-C8H18 for an equivalence ratio of 0.33. Charge pressure was adjusted so that the compression stoke could induce LTO for n-C7H16, but could not lead to high-temperature reactions. A spark was discharged in the field before, during, or after the LTO for n-C7H16 or in the field without LTO for i-C8H18. Reaction zones were induced in the field after the LTO, whereas no reaction zones were induced in the fields before the LTO and without LTO. Local ignitions were induced in the areas surrounding the propagating reaction zones. The reaction zone propagation with the low equivalence ratio must be a different phenomenon from conventional flame propagation. The reaction zones can compress or heat the surrounding areas containing H2O2 and CH2O, and accelerate an H2O2 regeneration loop in the pre-reaction zones.

Detailed reaction path analyses of DME (dimethyl ether, CH3OCH3) and n-heptane (n-C7H16) were performed computationally with the “contribution matrix” showing the contribution ratios of important elementary reactions to formation or removal of every species or heat release at transient temperatures. It was found that the “H2O2 reaction loop” defined by the authors plays an important role in the initiation of thermal ignition. This is a reaction loop composed of four reactions, H2O2 + M → 2OH + M, OH + CH2O → HCO + H2O, HCO + O2 → HO2 + CO and 2HO2 → H2O2 + O2. The overall reaction is 2CH2O + O2 → 2H2O + 2CO + 473 kJ. This loop begins to be active, when the OH formation by H2O2 + M → 2OH + M becomes dominant against those by cool-flame reactions with NTC's (negative temperature coefficient) at about 950 K. The loop releases a significant amount of heat without consuming H2O2.