<div class="section abstract"><div class="htmlview paragraph">Performances of battery electric vehicles (BEV) are affected by the thermal imbalance in the battery packs under driving cycles. BEV thermal management system (VTMS) should be managed efficiently for optimal energy consumption and cabin comfort. Temperature changes in the brick, module, and pack under the repeated transient cycles must be understood for model-based development. The authors conducted chassis dynamometer experiments on a fully electric small crossover sports utility vehicle (SUV) to address this challenge. A BEV is tested using a hub-type, 4-wheel motor chassis dynamometer with an air blower under the Worldwide Harmonized Light Vehicles Test Cycle (WLTC) and Federal Test Procedures (FTP) with various ambient temperatures. The mid-size BEV with dual-motor featured 80 thermocouples mounted on the 74-kWh battery pack, including the cells, upper tray, side cover, and pack cover. The authors analyzed battery pack temperature distribution behavior by changing the battery’s initial state of charge (SOC) and cell temperatures.</div><div class="htmlview paragraph">Performance metric data such as battery voltage, current, SOC, pack temperature, coolant properties, pressure drop, and flow rate are recorded and analyzed. The results show the temperature variations under diverse driving conditions, with an average difference of 1°C between modules and 2.3°C between bricks in module 1, and a maximum temperature difference of 3.9°C is recorded in the battery pack. The results provide valuable insights into the optimal operational range for maintaining battery pack temperature stability. The measured results can provide a fundamental understanding of the peak temperature location on the brick-to-brick and pack-to-pack variation under transient cycles. These results provide a fundamental understanding of the thermal performance of battery brick, module, and pack, which can be used to develop a VTMS model.</div></div>

<div class="section abstract"><div class="htmlview paragraph">Battery thermal management system (BTMS) significantly improves battery electric vehicle (BEV) performance, especially under cold weather. A tradeoff between battery power consumption and cabin heating performance must be considered in cold driving conditions. This preliminary study aims to develop an integrated battery pack and coolant channel model to predict the thermal behavior of a BEV thermal management system. In this study, we develop and calibrate the physical baseline model with testbench data using finite element and CFD software. First, an electrochemical battery cell 1D model (Pseudo-2D or P2D) is built and calibrated against the cell reference data. An integrated pack model consisting of four modules (each has 23 and 25 bricks and a total of 4416 cells) with coolant flow channels is developed using reference and benchmarked data. Then, the model is calibrated against experimental results from a mass-production, mid-size battery-electric SUV operated under cold driving conditions at a constant vehicle speed of 60 km/h for 5800 sec. The integrated model considers the heat transfer characteristics from cell-to-brick and brick-to-coolant channels. As a result, a thermodynamic model that can predict the average battery temperature rise, temperature changes, and pressure drops of the coolant of the battery pack is constructed. The result shows that the battery pack model can predict the coolant pressure drop of the pack with a mean absolute percentage error (MAPE) of 0.49 % and the coolant temperature rise with a 5.23% MAPE. The calibrated battery pack model can reasonably reproduce the terminal voltage with a MAPE of 0.30%. A 3D-CFD simulation result of the battery brick model is also reported on the cell-to-cell temperature distribution of 46 cells.</div></div>

<div class="section abstract"><div class="htmlview paragraph">Synthetic fuels can significantly improve the combustion and emission characteristics of heavy-duty diesel engines toward decarbonizing heavy-duty propulsion systems. This work analyzes the effects of engine operating conditions and synthetic fuel properties on spray, combustion, and emissions (soot, NOx) using a supercharging single-cylinder engine experiment and KIVA-4 code combined with CHEMKIN-II and in-house phenomenological soot model. The blended fuel ratio is fixed at 80% diesel and 20% n-paraffin by volume (hereafter DP). Diesel, DP1 (diesel with n-pentane C5H12), DP2 (diesel with n-hexane C6H14), and DP3 (diesel with n-heptane C7H16) are used in engine-like-condition constant volume chamber (CVC) and engine experiments.</div><div class="htmlview paragraph">Boosted engine experiments (1080 rpm, common-rail injection pressure 160 MPa, multi-pulse injection) are performed using the same DP fuel groups under various main injection timings, pulse-injection intervals, and EGR = 0-40%. Once the 3D-CFD model is validated with the CVC and experimental engine data, in-cylinder combustion and emissions are analyzed. The CVC experiments show that DP2 and DP3 liquid penetrations are shorter than diesel oil. In engine tests, NOx did not change much for all DP fuels for the same engine operating condition. However, shorter-penetrated DP2 and DP3 reduce soot emissions by more than 60% and CO without worsening brake-specific fuel consumption compared to diesel oil. The 3D-CFD results show that n-hexane shifts the penetration of the high-carbon number to the low-carbon fuel. Vapor penetrations are found to be shortened by blending low-volatility fuels with diesel oil. Visualizations of the in-cylinder confirmed a decrease in the amount of soot formation near the wall for DP2 and DP3 fuels. In addition, equivalence ratio – temperature (phi-T) maps of these fuels indicate that at 40% EGR, soot emissions are reduced at lower equivalence ratios than diesel oil.</div></div>

<div class="section abstract"><div class="htmlview paragraph">Pre-chamber (PC) natural gas and hydrogen (CH<sub>4</sub>-H<sub>2</sub>) combustion can improve thermal efficiency and greenhouse gas emissions from decarbonized stationary engines. However, the engine efficiency is worsened by prolonged combustion duration due to PC jet velocity extinction. This work investigates the impact of cylindrical PC internal shapes to increase its jet velocity and shorten combustion duration. A rapid compression and expansion machine (RCEM) is used to investigate the combustion characteristics of premixed CH<sub>4</sub> gas. The combustion images are recorded using a high-speed camera of 10,000 fps. The experiments are conducted using two types of long PC shapes with diameters <i>φ=</i>4 mm (hereafter, long<i>φ</i>4) and 5 mm (hereafter, long <i>φ</i>5), and their combustions are compared against a short PC shape (<i>φ</i>=12 mm). For all designs of the PC shapes, the PC holes are 6 with 2 mm in diameter. Initial recorded results using only CH<sub>4</sub> show that jet extinction does not occur using the short and long 5mm types. The combustion duration of <i>φ</i>=4 mm PC is the shortest compared to the short-type and 5 mm PC. With CH<sub>4</sub>-H<sub>2</sub> blending (0%H<sub>2</sub>, 10%H<sub>2</sub>, 20%H<sub>2</sub>) and 4 mm shape, the combustion durations of 10%H<sub>2</sub> and 20%H<sub>2</sub> can be shortened compared to the CH<sub>4</sub>-only case (0%H<sub>2</sub>). However, the jet extinction probability is not zero.</div><div class="htmlview paragraph">3D-CFD combustion simulations are performed using CONVERGE software, and a 1/6 sector-mesh, GRI-Mech 3.0, G-equation combustion, and law-of-wall models are utilized in conjunction with RNG k-epsilon turbulence model. Simulated in-cylinder pressure and burning rate are validated against the recorded data. PC jet velocity, turbulent kinetic energy (TKE) of the main chamber, PC jet temperature, total heat loss in PC, and PC heat loss rate of 12 mm, 4 mm, and 5 mm are compared. Experimental combustion images and 3D-CFD temperature distribution with and without H<sub>2</sub> blending are also reported. The results show that long PC types can accelerate the jet velocity and shorten combustion duration. The jet extinction can be prevented by designing a small PC diameter area larger than the nozzle’s cross-sectional area. With H<sub>2</sub> blending, laminar burning velocity increases, and the jet ejection timing can be advanced. The result shows that the combustion duration can be shortened by 15 degrees using CH<sub>4</sub>-20%H<sub>2</sub> against CH<sub>4</sub> only.</div></div>

<div>Regulations limiting greenhouse gas (GHG) emissions in the transport sector have become more restrictive in recent years, drawing interest to synthetic fuels such as e-fuels and biofuels that could “decarbonize” existing vehicles. This study focuses on the potential to increase the thermal efficiency of spark-ignition (SI) engines using ethanol as a renewable fuel, which requires a deep understanding of the effects of ethanol on combustion behavior with high compression ratios (CRs). An important phenomenon in this condition is pre-spark heat release (PSHR), which occurs in engines with high CRs in boosted conditions and changes the fuel reactivity, leading to changes in the burning velocity. Fuel blends containing ethanol display high octane sensitivity (OS) and limited low-temperature heat release (LTHR). Consequently, their burning velocities with PSHR may differ from that of gasoline. This study therefore aimed to clarify the effects of ethanol on SI combustion behavior under PSHR conditions. Combustion behavior was studied by performing single-cylinder engine experiments and chemical kinetics simulations. The experimental measurements were performed to characterize the relationship between the occurrence of PSHR and the main combustion duration. Analysis of this relationship showed that the ethanol-blended fuel has a lesser PSHR and a longer combustion duration than the non-ethanol fuel by approximately 5% in high engine load conditions. Simulations using input data from the experiments revealed that the ethanol-blended fuel has a lower laminar burning velocity due to the lower temperature in the unburned mixture caused by its PSHR. Additional simulations examining the chemical effect of partially oxidized reactants caused by PSHR on the laminar burning velocity showed that partially oxidized reactants increase the laminar burning velocity of the ethanol-blended fuel but decrease that of a reference fuel without ethanol. A large number of fuel radicals and oxides of the ethanol-blended fuel enhances chain-branching reactions in the pre-flame zone and possibly increases its laminar burning velocity. However, the thermodynamic effect of PSHR on laminar burning velocity exceeds the chemical effect, and thus the ethanol-blended fuel has a lower turbulent burning velocity in the PSHR conditions.</div>

Abstract

Using thermoelectric generators (TEG) to reduce exhaust heat loss from internal combustion engines can improve emissions and the fuel economy of conventional and electrified vehicles. However, TEG potentials have not been investigated in hybridized, compressed natural gas (CNG), twin-turbocharged, spark-ignited (SI) engines. This work demonstrates TEG's effectiveness in boosting a hybridized 3.0 L CNG engine using model-based development. TEG experiments are performed to measure thermal performances under different inlet gas conditions for model validations. Simplified user-defined functions of flow friction and heat transfer coefficients are used to calibrate the model. A fast-calibration model can reproduce measured heat transfer, pressure drop, and thermal performances. The engine performances are validated against measured 35 steady-state conditions from the production engine used in light-duty CNG trucks under the JE05 drive cycle. Next, the model is connected to the turbocharging system downstream of the well-calibrated 4-cylinder SI engine model. Under the peak performance condition (peak brake thermal efficiency BTE at 2400 RPM and 102 kW load), the results show that the engine BTE is improved by 0.56% using a 7×9 TEG module arrangement (3-sheet TEG with 1.5× A4 size). A 9×10 arrangement can enhance the BTE to 0.8%. Effective electrical power is generated up to 1.168 kW from the TEG, depending on the JE05 operating regions, without significant brake power loss.

Abstract

This work demonstrates the enhancement of brake thermal efficiency (BTE) of an advanced, turbocharged, production-intent 2.2 L diesel engine with a thermoelectric waste heat recovery system (TEG-WHR). The integrated engine model with the TEG is developed using 0D/1D software. Experimental data from the corrugated fin TEG under fin pitch = 1.0–2.0 mm, inlet gas temperatures (200–300 °C), and mass flow rates (5.0–15.0 g/s) are used for validating the model. The TEG model can reproduce measured pressure drop, heat transfer, and thermal performance characteristics. A 1-cylinder engine model parented from the advanced turbocharged diesel engine is developed. Under motoring and firing conditions, measured exhaust pressure, temperature, velocity, mass flow rate, and enthalpy are validated under various valve timings. Finally, the 3-layer TEG model is connected to the 4-cylinder engine to maximize its performance under a highly efficient (peak BTE) operating condition at 2250 RPM. Optimal size and thermoelectric module arrangement of the TEG system in the engine system considering a tradeoff between the TEG generated electrical power and engine pumping losses are suggested. The effective power of 1.1 kW and 1.1 % BTE improvement are obtained from the 3-sheet TEG system. As a result, a 49.9 % engine BTE is demonstrated without brake power loss.

<title>Abstract</title>
This work analyzed measured data from a single-cylinder engine operated under gasoline direction injection homogenous charge compression ignition (GDI-HCCI) mode. The experiments were conducted at a 0.95 equivalence ratio (φ) under 0.5 MPa indicated mean effective pressure and 1500RPM. A side-mounted injector delivered primary reference fuel (octane number 90) into the combustion chamber during negative valve overlap (NVO). Advanced combustion phase CA50 were observed as a function of the start of injection (SOI) timings. Under φ=0.95, peak NVO in-cylinder pressures were lower than motoring for single and split injections, emphasizing that NVO reactions were endothermic. Zero-dimensional kinetics calculations showed classical reformate species (C3H6, C2H4, CH4) from the NVO rich mixture increased almost linearly due to SOI timings, while H2 and CO were typically low. These kinetically reformed species shortened predicted ignition delays. This work also analyzed the effects of intake pressure and single versus double pulses injections on CA50, burn duration, peak cylinder pressure, combustion noise, thermal efficiency, and emissions. Advanced SOI (single-injection) generated excessive combustion noise metrics over constraint limits, but the double-pulse injection could significantly reduce the metrics (Ringing Intensity ≤ 5 MW/m2, Maximum Pressure Rise Rate = 0.6 MPa/CA) and NOx emission. The engine's net indicated thermal efficiency reached 41% under GDI-HCCI mode against 36% under SI mode for the same operating conditions. Under GDI-HCCI mode and without spark-ignition, late fuel injection in the intake stroke could reduce NOx to a single digit.

<title>Abstract</title>
Gasoline direct injection (GDI) and negative valve overlap (NVO) are standard strategies to control combustion characteristics and exhaust gas emissions in homogenous charge compression ignition (HCCI) engines. In this work, experimental data from a single-cylinder engine operated under the GDI-HCCI mode were analyzed. The experiments were performed at an equivalence ratio of 0.95 under a mid-load condition. A side-mounted injector delivered primary reference fuel with octane number 90 directly into the combustion chamber during the NVO. The measured results showed advanced combustion phase CA50 under the early start of injection (SOI) timings. Peak recompression pressures were lower than the motoring, emphasizing that the NVO reactions were net endothermic. Zero-dimensional kinetics calculations showed that classical reformate species increase almost linearly as a function of SOI timings.


This work also presents the effects of intake boosting pressure and single versus double pulses injections on CA50, burn duration CA10-90, peak cylinder pressure, combustion noise metrics, thermal efficiency, and emissions. The combustion noise metrics were over the engine constraint limit under advanced SOI timings, but a double-pulse injection could reduce the combustion noise metrics and NOx emission. Late fuel injection in the intake stroke could reduce NOx to a single digit.

<title>Abstract</title>
Gasoline direct injection (GDI) is a promising solution to increase engine thermal efficiency and reduce exhaust gas emissions. The GDI operation requires an understanding of fuel penetration and droplet size, which can be investigated numerically. In the numerical simulation, primary and secondary breakup phenomena are studied by the Kelvin-Helmholtz/Rayleigh-Taylor (KH-RT) wave breakup models. The models were initially developed for diesel fuel injection, and in the present work, the models are extended to the GDI application combined using large-eddy simulation (LES). The simulation is conducted using the KIVA4 code.


Measured data of experimental spray penetration and Mie-scattering image comparisons are carried out under non-reactive conditions at an ambient temperature of 613K and a density of 4.84 kg/m3. The spray penetration and structures using LES are compared with traditional Reynolds-Averaged Navier-Stokes (RANS). Grid size effects in the simulation using LES and RANS models are also investigated to find a reasonable cell size for future reactive gasoline spray/combustion studies. The fuel spray penetration and droplet size are dependent on specific parameters. Parametric studies on the effects of adjustable constants of the KH-RT models, such as time constants, size constants, and breakup length constant, are discussed. Liquid penetrations from the RANS turbulence model are similar to that of the LES turbulence model’s prediction. However, the RANS model is not able to capture the spray structure well.

<title>Abstract</title>
The oxygen-depleted environment in the recompression stroke can convert gasoline fuel into light hydrocarbons due to thermal cracking, partial oxidation, and water-gas shift reactions. These reformate species can influence the combustion characteristics of gasoline direct injection homogeneous charge compression ignition (GDI-HCCI) engines. In this work, the combustion phenomena are investigated using a single-cylinder research engine under a medium load. The main combustion phases are experimentally advanced by direct fuel injection into the negative valve overlap (NVO) compared with that of intake stroke under single/double-pulse injections. NVO peak in-cylinder pressures are lower than that of motoring due to the limited O2 concentration, emphasizing that endothermic reactions occur during the overlap. This phenomenon limits the oxidation reactions, and the thermal effect is not pronounced. The zero-dimensional chemical kinetics results present the same increasing tendencies of classical reformed species of rich mixture such as C3H6, C2H4, CH4, CO, and H2 as functions of injection timings. Predicted ignition delays are shortened due to the additions of these reformed species. The influences of the reformates on the main combustion are confirmed by three-dimensional computational fluid dynamics (CFD) calculations, and the results show that OH radicals are advanced under NVO injections relative to intake stroke injections. Consequently, earlier heat release and cylinder pressure are noticeable. Parametric studies on the effects of injection pressure, double-pulse injection, and equivalence ratio on the combustion and emissions are also discussed experimentally.

<title>Abstract</title>
The turbulent combustion in gasoline engines is highly dependent on laminar flame speed SL. A major issue of the quasi-dimensional (QD) combustion model is an accurate prediction of the SL, which is unstable under low engine speeds and ultra-lean mixture. This work investigates the applicability of the combustion model with a refined SL correlation for evaluating the combustion characteristics of a high-tumble port gasoline engine operated under ultra-lean mixtures. The SL correlation is modified and validated for a five-component gasoline surrogate. Predicted SL values from the conventional and refined functions are compared with measurements taken from a constant-volume chamber under micro-gravity conditions. The SL data are measured at reference and elevated conditions. The results show that the conventional SL overpredicts the flame speeds under all conditions. Moreover, the conventional model predicts negative SL at equivalence ratio ϕ ≤ 0.3 and ϕ ≥ 1.9, while the revised SL is well validated against the measurements. The improved SL correlation is incorporated into the QD combustion model by a user-defined function. The engine data are measured at 1000–2000 rpm under engine load net indicated mean effective pressure (IMEPn) = 0.4–0.8 MPa and ϕ = 0.5. The predicted engine performances and combustions are well validated with the measured data, and the model sensitivity analysis also shows a good agreement with the engine experiments under cycle-by-cycle variations.

<title>Abstract</title>
Injected gasoline into the O2-depleted environment in the recompression stroke can be converted into light hydrocarbons due to thermal cracking, partial oxidation, and water-gas shift reaction. These reformate species influence the combustion phenomena of gasoline direct injection homogeneous charge compression ignition (GDI-HCCI) engines. In this work, a production-based single-cylinder research engine was boosted to reach IMEPn = 0.55 MPa in which its indicated efficiency peaks at 40–41%. Experimentally, the main combustion phases are advanced under single-pulse direct fuel injection into the negative valve overlap (NVO) compared with that of the intake stroke. NVO peak in-cylinder pressures are lower than that of motoring, which emphasizes that endothermic reaction occurs during the interval. Low O2 concentration could play a role in this evaporative charge cooling effect. This phenomenon limits the oxidation reaction, and the thermal effect is not pronounced. For understanding the recompression reaction phenomena, 0D simulation with three different chemical reaction mechanisms is studied to clarify that influences of direct injection timing in NVO on combustion advancements are kinetically limited by reforming. The 0D results show the same increasing tendencies of classical reformed species of rich-mixture such as C3H6, C2H4, CH4, CO, and H2 as functions of injection timings. By combining these reformed species into the main fuel-air mixture, predicted ignition delays are shortened.


The effects of the reformed species on the main combustion are confirmed by 3D-CFD calculation, and the results show that OH radical generation is advanced under NVO fuel injection compared with that of intake stroke conditions thus earlier heat release and cylinder pressure are noticeable. Also, parametric studies on injection pressure and double-pulse injections on engine combustion are performed experimentally.

<title>Abstract</title>
A quasi-dimensional (QD) simulation model is a preferred method to predict combustion in the gasoline engines with reliable results and shorter calculation time compared with multi-dimensional simulation. The combustion phenomena in spark ignition (SI) engines are highly turbulent, and at initial stage of the combustion process, turbulent flame speed highly depends on laminar burning velocity SL. A major parameter of the QD combustion model is an accurate prediction of the SL, which is unstable under low engine speed and ultra-lean mixture.


This work investigates the applicability of the combustion model for evaluating the combustion characteristics of a high-tumble port gasoline engine operated under ultra-lean mixture (equivalence ratio up to ϕ = 0.5) which is out of the range of currently available SL functions initially developed for a single component fuel. In this study, the SL correlation is improved for a gasoline surrogate fuel (5 components). Predicted SL data from the conventional and improved functions are compared with experimental SL data taken from a constant-volume chamber under micro-gravity condition. The SL measurements are done at reference conditions at temperature of 300K, pressure of 0.1MPaa, and at elevated conditions whose temperature = 360K, pressure = 0.1, 0.3, and 0.5 MPaa. Results show that the conventional SL model over-predicts flame speeds under all conditions. Moreover, the model predicts negative SL at very lean (ϕ ≤ 0.3) and rich (ϕ ≥ 1.9) mixture while the revised SL is well validated with the measured data. The improved SL formula is then incorporated into the QD combustion model by a user-defined function in GT-Power simulation. The engine experimental data are taken at 1000 RPM and 2000 RPM under engine load IMEPn = 0.4–0.8 MPa (with 0.1 increment) and ϕ ranges are up to 0.5. The results shows that the simulated engine performances and combustion characteristics are well validated with the experiments within 6% accuracy by using the QD combustion model coupled with the improved SL. A sensitivity analysis of the model is also in good agreement with the experiments under cyclic variation (averaged cycle, high IMEP or stable cycle, and low IMEP or unstable cycle).

Gasoline Direct Injection Homogeneous Charge Compression (GDI-HCCI) combustion is achieved by closing early the exhaust valves for trapping hot residual gases combined with direct fuel injection. The combustion is chemically controlled by multi-point auto-ignition which its main combustion phase can be controlled by direct injection timing of fuel. This work investigates the effect of single pulse injection timing on a supercharged GDI-HCCI combustion engine by using a four-stroke single cylinder engine with a side-mounted direct fuel injector.


Injection of primary reference fuel PRF90 under the near-stoichiometric-boosted condition is studied. The fuel is injected during negative valve overlap (NVO) or recompression period for fuel reformation under low oxygen concentration and the injection is retarded to intake stroke for the homogeneous mixture. It is found that the early fuel injection in NVO period advances the combustion phasing compared with the retarded injection in the intake stroke. Noticeable slower combustion rate from intake stroke fuel injection is obtained compared with the NVO injection due to charge cooling effect. Zero-dimensional combustion simulations with multiple chemical reaction mechanisms are simulated to provide chemical understanding from the effect of fuel injection timing on intermediate species generations. The species such as C2H4, C3H6, CH4, and H2 are found to be formed during the NVO injection period from the calculations. The effects of single pulse injection timings on combustion characteristics such pressure rise rate, combustion stability, and emissions are also discussed in this study.

&lt;div class="section abstract"&gt;&lt;div class="htmlview paragraph"&gt;Primary work is to investigate premixed laminar flame propagation in a constant volume chamber of iso-octane/air combustion. Experimental and numerical results are investigated by comparing flame front displacements under lean to rich conditions. As the laminar flame depends on equivalence ratio, temperature, and pressure conditions, it is a main property for chemical reaction mechanism validation. Firstly, one-dimensional laminar flame burning velocities are predicted in order to validate a reduced chemical reaction mechanism. A set of laminar burning velocities with pressure, temperature, and mixture equivalence ratio dependences are combined into a 3D-CFD calculation to compare the predicted flame front displacements with that of experiments. It is found that the reaction mechanism is well validated under the coupled 1D-3D combustion calculations. Next, lean experiments are operated in a SI engine by boosting intake pressure to maintain high efficiency without output power penalty. The peak indicated thermal efficiency are finally achieved under λ=1.3 with intake manifold absolute pressure 150 kPa in experiment. Data of in-cylinder pressure and rate of heat release from the 3D-CFD simulations combined with the validated chemical reaction mechanism are reproduced. NOx emissions from experiment and simulation are also in good agreements under the lean-boost combustion. Further thermal efficiency improvements of the lean-boost SI engine are investigated numerically by using dilution rate, high induced in-cylinder flow, and high knock resistant fuel. The peak indicated thermal efficiency and load of the SI engine is achieved. In addition, methods to prevent knock for high efficiency spark ignition engine are also discussed.&lt;/div&gt;&lt;/div&gt;

This paper aims to improve thermal efficiency of spark ignition engine by numerical calculation with detailed chemistry. Experimental results from a four-stroke-single-cylinder engine are compared with that of simulations. It is experimentally found that peak efficiency is achieved at lean-limit combustion under excess air ratio λ=1.6. Due to engine output power loss, further investigations are conducted under lean-boost operations. The best condition of the lean-boost mode is at λ=1.3 and 150 kPa boosted pressure (abs). To further improve the efficiency without power loss, simulations are conducted under lean-boost combustion with dilution rate, high engine swirl, and high knock resistant fuel.