Bayesian Experimental Design and its Application to Engine Research and Development

<div class="section abstract"><div class="htmlview paragraph">The introduction of the Worldwide Harmonised Light Vehicles Test Procedure (WLTP) and Real Driving Emissions (RDE) test requirements have put increased strain on vehicle and engine performance as well as the development of advanced engine technologies and emissions mitigation strategies. This requirement for increased development is a direct result of the need for new vehicles to comply with present and emerging emissions standards across an extended range of boundary conditions that include ambient temperature, altitude and driving style.</div><div class="htmlview paragraph">To reduce the significant number of on-road test permutations that would ordinarily be required to validate a given vehicle across the defined RDE boundary conditions, a Road to Rig (R2R) development approach known as RDE Plus (RDE+) is being evolved at HORIBA MIRA. This will enable real world driving scenarios to be developed and deployed much further upstream during the development lifecycle of the vehicle and engine using aspects of virtual calibration; the aim being to significantly reduce vehicle and engine development costs and associated timescales.</div><div class="htmlview paragraph">On-road RDE data is presented in this paper from vehicles of different powertrain technologies gathered from several European test locations under conditions representing the boundaries defined by the RDE protocol. Upon gathering a combination of vehicle CANBus, custom instrumentation and emissions (Portable Emissions Measurement System (PEMS)) data, recorded RDE routes were replicated on a chassis dynamometer with a view of developing a standardised practical approach to replicate RDE tests in the laboratory environment.</div><div class="htmlview paragraph">Lab tests were performed at HORIBA MIRA in the Advanced Emissions Test Centre (AETC) with the aim of defining a chassis dynamometer test methodology suitable for different powertrain technologies that results in accurate representations of on-road conditions with respect to vehicle load, ambient temperature, altitude and emissions.</div><div class="htmlview paragraph">Following the road to chassis dynamometer correlation, the vehicle engines will be installed on an engine dynamometer with the objective of developing a suite of “worst case” RDE cycles using an Engine-in-the-Loop (EiL) format complemented with a Design of Experiments (DoE) type mapping approach. These “worst case” cycles, most likely operating near to or at the RDE boundary condition limits, can then be adopted for engine and vehicle development purposes as well as providing evidence of complete RDE compliance. The methodology to develop this EiL strategy in combination with a detailed overview of RDE+ project as a whole is presented in this paper.</div></div>

<div class="htmlview paragraph">As the environment protection is a key issue for the 21<sup>st</sup> century, the world-wide auto manufacturers are now considering the pollutant emission reduction as a major task for the future. Thanks to a combustion occurring in a lean well premixed charge, the HCCI (Homogeneous Charge Compression Ignition) concept offers the possibility to almost reach the near zero-NOx and particulate emission ambitious target. Nevertheless, the customer application of such a system must clearly preserve the well known advantages of Diesel engines such as high specific power and torque and low fuel consumption. For this reason the full load application of such an engine still uses conventional combustion whilst the homogeneous charge compression ignition is used at part load.</div> <div class="htmlview paragraph">This work presents how the pre-optimization of a combustion chamber, dedicated to HCCI Diesel combustion, is carried out at full load thanks to the use of design of experiments and a CFD code. Consequently the part load optimization under homogeneous conditions is not discussed here.</div> <div class="htmlview paragraph">The bowl shape was investigated by the mean of width to depth ratio, the injector nozzle number of holes as well as protrusion and spray angle were optimized as regard to the cylinder head swirl ratio.</div> <div class="htmlview paragraph">Besides its interest in reducing the number of calculation configurations, the design of experiments allows us to point out the most important interactions between the above parameters.</div> <div class="htmlview paragraph">Main results show that the specific power at maximum speed is not sensitive to the bowl shape as far as the other parameters are optimized in order to ensure good mixing conditions and long ignition delays. Early injection together with adapted injection configuration - eg spray angle, number of holes and swirl ratio, are found to be of uppermost importance to reach good performance with low smoke level. Optimization at the specific torque shows that an appropriate bowl width to depth ratio allows the best use of oxygen as well as an efficient soot post-oxidation.</div>

<div class="section abstract"><div class="htmlview paragraph">In recent years the 2-wheelers engines companies have been focused on increasing the overall engine efficiency, that can be achieved amongst other by engine down-speeding, engine down-sizing and by reducing the frictions; however, to maintain or improve vehicle performance, it is necessary to provide a corresponding increase in specific power.</div><div class="htmlview paragraph">In accordance with these trends, the studied approaches and methodologies have been exploited, during the development of the new Piaggio small scooter engine.</div><div class="htmlview paragraph">In this work a multi-objective analysis has been applied to the valve train system design, in order to optimize the engine performance in terms of friction reduction, power curve and timing system response. Along with this optimization methodology, a robust design has been studied to make the peak cranking compression pressure insensitive to the engine starting device working. These calculation methodologies have been followed using commercial software as GT-SUITE® code and the modeFRONTIER® platform for the multi-objective optimization analysis.</div><div class="htmlview paragraph">The several configurations investigated have allowed to reach the desired trade-off between the appropriate and efficient valve train behavior and the engine performances.</div><div class="htmlview paragraph">The global engine friction reduction has been obtained thanks to low resisting cam torque profiles, a specific valve springs set and a new compression releaser device (during engine starting phase).</div><div class="htmlview paragraph">The results have been further validated by an experimental campaign carried out on the test bench for engine performance and on the specific cold test bench to evaluate friction losses.</div></div>

<div class="section abstract"><div class="htmlview paragraph">High pressure EGR provides NO<sub>x</sub> emission reduction even at low exhaust temperatures. To maintain a safe EGR system operation over a required lifetime, the EGR cooler fouling must not exceed an allowable level, even if the engine is operated under worst-case conditions.</div><div class="htmlview paragraph">A reliable fouling simulation model represents a valuable tool in the engine development process, which validates operating and calibration strategies regarding fouling tendency, helping to avoid fouling issues in a late development phase close to series production.</div><div class="htmlview paragraph">Long-chained hydrocarbons in the exhaust gas essentially impact the fouling layer formation. Therefore, a simulation model requires reliable input data especially regarding mass flow of long-chained hydrocarbons transported into the cooler.</div><div class="htmlview paragraph">There is a huge number of different hydrocarbon species in the exhaust gas, but their individual concentration typically is very low, close to the detection limit of standard in-situ measurement equipment like GC-MS.</div><div class="htmlview paragraph">Therefore, a new measurement and analysis approach has been developed, where the exhaust gas is guided to a metal foam collector, in which HC`s are deposited. The probe is then analyzed in a suited thermogravimetrical system (TGA) in nitrogen atmosphere, temperature range 25°C to 650°C. Analyzing the TGA curve, HC concentration data for 6 different boiling temperature ranges are obtained, provided to an adapted 1-d fouling simulation model.</div><div class="htmlview paragraph">Using these data along with further input parameters like cooler geometry, gas temperature, pressure, flow, particle size distribution and coolant temperature, the simulation model has proven as a suitable tool to predict the fouling and identify engine settings for fouling reduction.</div></div>

<div class="htmlview paragraph">This report describes the possibility of quantitative analysis of Schlieren photographs as an internal combustion engine diagnostic.</div> <div class="htmlview paragraph">Using a recently developed photographic analysis system, it was attempted to analyze Schlieren photographs. Results showed simple integration calculations produced significant distortion in the analyzed results. To eliminate the distortion, some correction techniques were developed in this research. Accuracy of the analyzed results were evaluated roughly with uncertainty analysis. The results showed that this analysis technique can be one of the approximate diagnostics for the measurement of fuel vapor and density distribution in internal combustion engine research.</div> <div class="htmlview paragraph"><b>SCHLIEREN</b> photography is one of the most popular visualization techniques in the research of internal combustion engines. Although the photographs have density information in themselves, they are used mainly for the purpose of qualitative visualization. The objective of this research is to develop a quantitative analysis of Schlieren photographs for internal combustion engine research.</div> <div class="htmlview paragraph">One important parameter in engine research, for example, is the air entrainment in fuel jet vapor. Its measurement has previously been possible by gas sampling, laser holography, or laser Raman spectroscopy. The gas sampling method provides only point data, and the laser methods are very sensitive to vibration or optical settings. With this present Schlieren photographic analysis technique, it will be possible to measure instantaneously the whole space of the density field with a very simple optical system which is commonly used in engine research. Lakshminarayan and Dent(<span class="xref">1</span>)<span class="xref">*</span> analyzed interferometric photographs of jet sprays impinging on a wall. A major part of their (analysis method involved the counting of fringes which are sometimes difficult to distinguish in photographs. Nishida et al.(<span class="xref">2</span>)(<span class="xref">3</span>) analyzed Schlieren photographs reconstructed from laser holograms of jet vapor. However, they did not refer to the accuracy of the results.</div> <div class="htmlview paragraph">This report describes the background of the quantitative analysis and shows some analyzed results. The analysis was made for a CO2 jet injected into air. The Schlieren photographs were analyzed by an image processor and a personal computer.</div> <div class="htmlview paragraph">The results showed that the elimination of integrated error due to non-uniform background level was one of the most important parameters to be considered. In this study, a correction method is shown to determine the non-uniform background level and also to improve the accuracy of the results. This report also describes analyzed results for transient free jets and wall-impinging jets, which were compared with experimental results.</div>

<div class="htmlview paragraph">This paper presents the basic test data obtained from tests of a cabin air distribution system in a simulated Space Station U.S. Lab-A module.</div> <div class="htmlview paragraph">The cabin air distribution system controls the flow of air in the open space of a Space Station module. In order to meet crew comfort criteria the local velocities for this cabin air are required to be distributed within a specified range with upper and lower limits. Achieving this desired velocity distribution is dependent upon the: (1.) design of the cabin air supply equipment and cabin air return equipment, (2.) total flowrate of air supplied to and subsequently returned from the cabin, and (3.) interactive effects of any other additional air flow streams which enter and exit the cabin.</div> <div class="htmlview paragraph">The basic Space Station design for the cabin air supply and air return equipment was used in this test program. Only directional adjustments to vanes in supply air diffusers were made during the test. This equipment is included in the Temperature and Humidity Control (THC) subsystem for the Space Station. Two other air streams, in addition to what is defined as THC cabin airflow, which were involved in the test program were:</div> <div class="htmlview paragraph"> <ol class="list nostyle"> <li class="list-item"> <span class="li-label">1.</span> <div class="htmlview paragraph">Standoff airflow – Standoffs are described in the paper. Briefly, they are enclosures within the cabin which formerly included ventilation fans. The consequent standoff airflow interacted with THC cabin airflow. The present baseline THC subsystem does not include the standoff fans.</div></li> <li class="list-item"> <span class="li-label">2.</span> <div class="htmlview paragraph">Intermodule Ventilation (IMV) Airflow – IMV assemblies are used to interchange airflow between Space Station modules. These modules include the U.S. Lab-A module, the nodes, and, later in the Space Station program, the international modules.</div></li></ol></div> <div class="htmlview paragraph">Test results from eight (8) of the test runs which demonstrate the ability for the air distribution system to meet its objectives are presented. These test run results are compared with each other with respect to their standing relative to meeting the test objectives.</div> <div class="htmlview paragraph">The test program investigated the effects of the variables noted above in meeting the test objectives. The basic variable in these test runs was the THC airflow. This airflow varied from 300 cfm to 500 cfm in the 8 test runs presented here. Some changes were made in supply air diffuser vane directional settings. Standoff airflow was used in one test run (number 4) and IMV airflow was used in test runs 5, 6, 7, and 8. Based on the test results, the desired velocity range and an engineering evaluation of the test results, a THC airflow rate for U.S. Lab-A is recommended.</div>

<div class="section abstract"><div class="htmlview paragraph">A previous paper on this topic presented the use of design of experiments (DOE) to evaluate the sensitivity of vehicle dynamics simulation of the postimpact motion of a vehicle that included high initial rotational rates. That investigation involved only one software package and thus was confined to one simulation model for the purposes of developing and refining the analysis method rather than including a variety of simulation models for broader application. This paper expands the application of the method to investigate the comparative behavior and sensitivity of several other vehicle dynamic simulation models commonly used in the field of crash reconstruction. The software packages included in the studies presented in this paper are HVE (SIMON and EDSMAC4), PC-Crash and VCRware. This paper will present the results of the study, conducted using DOE, involving these models. The eleven factors selected for the study presented here were chosen based largely on the results of the prior study. The experimental design was expanded from 16 trials to 32 trials to provide additional insight into the interactions between the factors. Three response variables were used: the <i>x</i>-coordinate, <i>y</i>-coordinate and total rotational displacement of the vehicle at rest.</div><div class="htmlview paragraph">The results of the analysis show the sensitivity of all four programs to the eleven factors as well as the interactions between the factors. The results show that, consistent with prior results, and the expectation of the authors, the factor that all four programs showed as the most sensitive is the tire-roadway drag. Additionally, consistent with a reconstructionist’s intuition, is that the programs are sensitive to the initial (postimpact) velocities. However, the level of the sensitivities of the four programs to a given change in the same factor are different in many cases.</div><div class="htmlview paragraph">As compelling as the results are that show the largest sensitivities, the results show factors that some, or all, of the programs show little sensitivity. For example, for response using the <i>y</i>-coordinate and rotational displacement of the vehicle at rest, all four of the programs were relatively insensitive to changes in the lateral and longitudinal stiffnesses of the tires over the range selected.</div></div>

<div class="section abstract"><div class="htmlview paragraph">The engine research and development has a significant contribution to meet the stringent emission norms and the changing global market demands. Leveraging the available virtual engineering methods to improve performance, velocity, quality and diminish the lead time is the key for any global brand to stay in the competition. It is the key element to reduce the research and development costs substantially by virtually developing the idea as it is conceived.</div><div class="htmlview paragraph">Engine development test cells consist of expensive test and measurement systems which demand skilled labor and advanced equipment. Effective utilization of the test cells is essential to meet the scheduled project deadlines and cost targets. Engine Design process and tools when used effectively can increase the efficiency and lower the test cell operation costs substantially. This paper discusses the examples for this application in the area of engine installation, sensitive instrumentation/assembly.</div></div>

<div class="htmlview paragraph">The results of a comprehensive experimental investigation into the exhaust emission performance and combustion properties of neat and blended Gas-To-Liquids (GTL) diesel fuel are presented. A sulphur-free European diesel fuel was used as the reference fuel, and two blends of the GTL diesel fuel with the reference fuel, containing 20% and 50% GTL diesel fuel respectively, were investigated. The study was based on a Mercedes Benz 2.2 liter passenger car diesel engine and presents emission data for both the standard engine calibration settings, as well as settings which were optimized to match the characteristics of each fuel.</div> <div class="htmlview paragraph">Vehicle emission tests showed that the GTL diesel fuel results in reductions in HC and CO emissions of greater than 90%, while PM is reduced by 30%, and NOx remains approximately unchanged. Engine bench dynamometer tests showed reductions in soot of between 30% and 60%, and NOx reductions of up to 10% with the GTL diesel fuel, depending on the operating point. Results with the GTL diesel blends showed a pronounced non-linear response, with reductions in emissions generally being larger than indicated by the blending ratio, to the extent that the emission performance of the 50% blend was similar to that of the neat GTL diesel fuel.</div> <div class="htmlview paragraph">Optical engine tests were carried out to investigate the combustion properties of the neat GTL diesel fuel in more detail. In addition to reflecting the better ignition quality of the GTL diesel, these tests also showed that it has better vaporization characteristics and a more uniformly distributed flame structure.</div> <div class="htmlview paragraph">A design of experiments (DOE) approach was used to optimize the injection settings and EGR rates for each fuel in order to obtain the most favourable NOx-soot trade-off. By this means it was estimated that simultaneous reductions of 35% in both NOx and PM could be obtained in the vehicle emission test with GTL diesel fuel, with an optimized engine calibration. The corresponding estimated reductions with the 50% and 20% blends of GTL diesel are 30% and 15% respectively, again reflecting the non-linear exhaust emission response of the blended fuels.</div>

<div class="section abstract"><div class="htmlview paragraph">State-of-the-art motorcycle engines consist of numerous variable components and require a powerful motor management to meet the growing customer expectations and the legislative requirements (e.g. exhaust and noise emissions, fuel consumption) at the same time. These demands are often competing and raise the level of complexity in calibration. In the racing domain, the optimization requirements are usually higher and test efficiency is crucial. Whilst the number of variables to control is growing, the time to perform an engine optimization remains the same or is even shortened. Therefore, simulation is becoming an essential part of the engine calibration optimization. Considering the special circumstances in racing, involving valuable hardware, as well as extremely short development and calibration iteration loops, only transient testing is possible.</div><div class="htmlview paragraph">By utilizing model-based testing and optimization, Ducati Corse, the racing team division of the well-known motorcycle manufacturer Ducati, improved the ability to optimize their race engines efficiently. By using an engine model it is possible to make extremely quick calibration adaptations. All parameters can easily be optimized with respect to potential constraints without running the engine on a testbed. Moreover, the re-use of the engine model for co-simulations is applicable and sharing it with other departments in the company is possible to increase efficiency even more.</div><div class="htmlview paragraph">AVL CAMEO™ - the intelligent automated calibration environment - supports all engine optimization requirements with a consistent workflow from the task definition to the verification. For this specific racing use case, the software solution was implemented for the test planning using DoE (Design of Experiment), the data plausibility check and the empirical engine modeling. In addition, AVL CAMEO™ was the tool for realizing the model-based optimization and map creation.</div><div class="htmlview paragraph">With the implementation of the model-based approach, the motorcycle manufacturer has successfully improved the engine performance optimization. An exact model of the engine is now available which supports a deep understanding of the engine behavior. Through realizing this calibration approach, quick office and race track adaptations are possible and alternative optimizations for different tracks or conditions are easy to execute.</div></div>

<div class="htmlview paragraph">The 1975 federal CVS test procedure and emission standards have placed a heavy emphasis on minimizing the time required after engine start up to begin converting exhaust emissions. Four major improvements in catalytic HC/CO converters have been made in the past year to reduce catalyst ignition time:</div> <div class="htmlview paragraph"> <ol class="list nostyle"> <li class="list-item"> <span class="li-label">1.</span> <div class="htmlview paragraph">Use of low thermal capacity monolithic catalyst supports.</div> </li> <li class="list-item"> <span class="li-label">2.</span> <div class="htmlview paragraph">Locating converter after thermal reactor and/or closer to the exhaust manifold.</div> </li> <li class="list-item"> <span class="li-label">3.</span> <div class="htmlview paragraph">Development of catalysts with superior stability to long exposure at high temperatures.</div> </li> <li class="list-item"> <span class="li-label">4.</span> <div class="htmlview paragraph">Development of catalysts with improved low temperature activity.</div> </li> </ol> </div> <div class="htmlview paragraph">Small laboratory samples of potential HC/CO oxidation catalysts were screened for activity in a “laboratory simulated cold-start” test with exhaust from a 2-cyl engine. Several improved catalyst candidates were evaluated in a full-size converter “cold-start” test with exhaust from a 351 in<sup>3</sup> engine.</div> <div class="htmlview paragraph">Locating the converters closer to the exhaust manifold results in higher catalyst operating temperatures which require greater stability at high temperatures in the presence of trace contaminants. Stability, when exposed to wide temperature variations and changes from oxidation to reduction environments, has been evaluated in a thermal-cycling test for several improved catalysts. Catalyst stability to long-term exposure has also been evaluated in an engine dynamometer mileage accumulation test operated under simulated road conditions.</div> <div class="htmlview paragraph">Unfortunately, the rapid thermal-chemical response of the improved catalyst, operated closer to the exhaust manifold, increases the occurrences of excessive catalyst temperatures. Mathematical model studies of the effect of the improved catalyst properties on thermal response indicate a need to prevent unexpected high CO or HC concentration peaks in the engine exhaust, or to improve properties of existing catalysts. System changes, such as moving the low temperature catalyst farther from the exhaust manifold, may be required to obtain some protection from catalyst overtemperatures. These changes increase emission levels and require a breakthrough in engine control of raw emissions or improved catalysts.</div>

<div class="section abstract"><div class="htmlview paragraph">A gasoline compression-ignition combustion system is being developed for full-time operation over the speed-load map. Low-temperature combustion was achieved using multiple late injection (MLI), intake boost, and moderate EGR for high efficiency, low NOx, and low particulate emissions. The relatively long ignition delay and high volatility of RON 91 pump gasoline combined with an advanced injection system and variable valve actuation provided controlled mixture stratification for low combustion noise.</div><div class="htmlview paragraph">Tests were conducted on a single-cylinder research engine. Design of Experiments and response surface models were used to evaluate injection strategies, injector designs, and various valve lift profiles across the speed-load operating range. At light loads, an exhaust rebreathing strategy was used to promote autoignition and maintain exhaust temperatures. At medium loads, a triple injection strategy produced the best results with high thermal efficiency. Detailed heat release analysis indicated that heat losses were significantly reduced. At higher loads, a late-intake-valve-closing strategy was used to reduce the effective compression ratio. For all tests, intake air temperature was 50 C.</div><div class="htmlview paragraph">3D CFD simulations of fuel injection, mixing, and combustion were important to understand the emissions formation processes. With multiple late injections and low-to-moderate fuel pressure, spray penetration was low, mixing was fast, and wall wetting could be avoided. Fuel sprays were characterized in a spray chamber. Injection rate was measured using a rate tube.</div><div class="htmlview paragraph">Results showed that ISFC was very low. Minimum ISFC of 181 g/kWh was measured at 2000 rpm-11 bar IMEP. For IMEP from 2 to 18 bar, engine-out NOx and PM emissions were below targets of 0.2 g/kWh and 0.1 FSN, respectively, indicating that aftertreatment for these species may be reduced or eliminated. It was found that combustion noise levels, characterized by several noise metrics, could be effectively controlled by the injection process. Measurements of exhaust particulate size distribution indicated very low particle count, especially for a preferred injector with low levels of in-cylinder swirl.</div><div class="htmlview paragraph">Collectively, these results demonstrate the potential feasibility of full-time GDCI using RON 91 gasoline at low-to-moderate injection pressures with high fuel efficiency. While more development work is needed, there is good potential for a practical GDCI powertrain system based on these concepts.</div></div>

<div class="section abstract"><div class="htmlview paragraph">Fundamental engine research is primarily conducted under steady-state conditions, in order to better describe boundary conditions which influence the studied phenomena. However, light-duty automobiles are operated, and tested, under heavily transient conditions. This mismatch between studied conditions and in-use conditions is deemed acceptable due to the fundamental knowledge gained from steady-state experiments. Nonetheless, it is useful to characterize the conditions encountered during transient operation and determine if the governing phenomena are unduly influenced by the differences between steady-state and transient operation, and further, whether transient behavior can be reasonably extrapolated from steady-state behavior.</div><div class="htmlview paragraph">The transient operation mode used in this study consists of 20 fired cycles followed by 80 motored cycles, operating on a continuous basis. The intention of the cycle is to provide a significant transient condition, namely the change from motored to knock-limited fired operation, while also maintaining a repeatable cycle that allows for the collection of statistics during quasi- steady-state operation.</div><div class="htmlview paragraph">This study investigates the effect of transient operation on Knock-Limited Combustion Phasing (KL-CA50) compared to steady-state operation. Three compositionally dissimilar matched Research Octane Number (RON) = 98 fuels are used in this study, allowing for the assessment of fuel-specific effects on differences between steady-state and transient operation.</div><div class="htmlview paragraph">This study first characterizes the 20/80 firing cycle described above, before comparing the transient KL-CA50 measurements to the steady-state KL-CA50 measurements. Analysis of both the steady-state and transient results are used to gain insights into the effects of transient operation on end-gas autoignition, relative to steady-state operation and as a function of fuel composition.</div><div class="htmlview paragraph">The results of this study indicate the significant effect that transient operation has on KL-CA50 behavior of a fuel. This is both universal, in that all fuels show responses to the differences in compression temperatures of the charge, as well as fuel specific, in that the fuel response varies based on the fuel’s sensitivity to temperature, [O2], and trace species. All fuels showed a significant load extension under transient operation, based on tolerance of higher intake pressures. However, transient operation moved operating conditions to “beyond RON” (Octane Index K < 0) conditions, which favored higher-sensitivity fuels. Based on the analysis of system time constants (e.g. cylinder head temperature dynamic response, exhaust gas temperature dynamic response), it is expected that transient operation, and the benefits for knock-limited operation, are highly influential on drive-cycle performance.</div></div>

<div class="htmlview paragraph">In recent years, small diesel engines for industrial use are required to achieve still higher power output per engine size (high power density) and to have the performance to comply with stringent exhaust emissions standards as well. To meet these social demands and market requirements, there are various engineering tasks yet to accomplish. In this paper, two important technical issues have been focused, those are, combustion system improvement to enable lower exhaust emissions and higher power output than conventional models and new cooling system as a solution for increased heat load issue accompanying the progress of high power density.</div> <div class="htmlview paragraph">Firstly, for combustion system improvement, the swirl adjustment technique has been developed that will optimize the swirl ratio for each engine application with different load and speed condition. Intake airflow has been analyzed using Computational Fluid Dynamics (CFD) approach to enable adjustment of swirl ratio that plays an important role in creating air fuel mixture. The swirl was successfully controlled by providing the originally designed configuration on the wall, from cylinder block side, which introduces the intake air from valve throat area into cylinder. As a result, it is now possible to set optimum swirl ratio for each engine application while securing high volumetric efficiency.</div> <div class="htmlview paragraph">Next, a new cooling system has been introduced to cool down the neighborhood of top deck between cylinders where temperature becomes highest in cylinder block. The new coolant channels with built-in multiple water passages was incorporated by casting between cylinders. As a result, the temperatures between cylinders and in piston ring groove could be lowered remarkably.</div> <div class="htmlview paragraph">After overcoming these engineering challenges, the advanced high power density has successfully been materialized in small direct injection (DI) diesel engine for industrial use.</div>

<div class="htmlview paragraph">To examine the effects of exhaust gas recirculation (EGR) on the exhaust emissions and on the cycle-by-cycle variation of combustion process in a methanol engine, the authors have tried to measure the emissions of total hydrocarbon, carbon monoxides, nitric oxides and formaldehyde and to record the cylinder pressure development histories of 1000 cycles duration. Moreover, the standard deviation of peak pressure Pmax, the heat release rate and the averaged pressure history are analyzed.</div> <div class="htmlview paragraph">The results are compared with those of gasoline fueled operation.</div> <div class="htmlview paragraph">The results are summarized as follows:</div> <div class="htmlview paragraph"> <ol class="list nostyle"> <li class="list-item"> <span class="li-label">(1)</span><div class="htmlview paragraph">In the case of a methanol-fueled engine operated under exhaust gas recirculation (EGR), similar to that already carried out in a gasoline-fueled engine, the concentration of NOx, in the exhaust gas can be decreased but those of formaldehyde (HCHO) and unburned methanol are increased.</div> </li> <li class="list-item"> <span class="li-label">(2)</span> <div class="htmlview paragraph">The effect of EGR, such as 20% of it, on Nox reduction is the same as those of hydrous methanol of 50% water content.</div> </li> <li class="list-item"> <span class="li-label">(3)</span> <div class="htmlview paragraph">In the case of EGR applied to an EFI engine, its output and thermal efficiency are decreased within the lean mixture region (λ>1.0) but these recover and the concentration of hydrocarbon is decreased by raising the fuel temperature.</div> </li> <li class="list-item"> <span class="li-label">(4)</span> <div class="htmlview paragraph">When the optimum conditions of the EGR and fuel temperature are selected, it is possible to improve the exhaust gas characteristics concerning NOx, THC, etc., while the decreased in engine output and thermal efficiency is reduced to the minimum.</div> </li> </ol> </div>