Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The application of exhaust gas recirculation to a single cylinder compression ignition engine fuelled… McTaggart-Cowan, Gordon 2002

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2002-0180.pdf [ 4.54MB ]
Metadata
JSON: 831-1.0099623.json
JSON-LD: 831-1.0099623-ld.json
RDF/XML (Pretty): 831-1.0099623-rdf.xml
RDF/JSON: 831-1.0099623-rdf.json
Turtle: 831-1.0099623-turtle.txt
N-Triples: 831-1.0099623-rdf-ntriples.txt
Original Record: 831-1.0099623-source.json
Full Text
831-1.0099623-fulltext.txt
Citation
831-1.0099623.ris

Full Text

THE APPLICATION OF E X H A U S T GAS RECIRCULATION TO A SINGLE CYLINDER COMPRESSION IGNITION ENGINE F U E L L E D WITH N A T U R A L GAS By Gordon McTaggart-Cowan B.Eng., University of Victoria, 1999 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF MECHANICAL ENGINEERING We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December 2001 ©Gordon McTaggart-Cowan, 2001 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an a d v a n c e d d e g r e e a t the U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r a g r e e t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d b y t h e h e a d o f my depar tment o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f MztHA/VX/lAL £o^/^/"yiyzgZ5gf^4-The U n i v e r s i t y o f B r i t i s h C o l u m b i a V a n c o u v e r , Canada Date J ft" pGr.e*<iK en. 7 On i Abstract Reducing the emissions of Nitric Oxides from diesel engines is one of the main challenges facing diesel engine designers. Many different methods have been investigated for reducing NO x emissions, including exhaust gas recirculation (EGR) and the high pressure direct injection (HPDI) of natural gas. Combining these two techniques offers the potential to reduce NO x emissions further than either method can individually. To test the effects of EGR on an HPDI engine, the University of British Columbia's Alternate Fuels Research Group recently received a new Cummins ISX heavy-duty truck engine, modified for single-cylinder operation. The new engine was commissioned on HPDI and a series of tests were run to compare its performance and emissions to a six-cylinder HPDI engine. These results showed good agreement for performance, but some significant differences in emissions between the two engines. Although emissions data are not directly transferable to a six-cylinder engine, the trends and general effects identified through testing on the single cylinder engine should be applicable to all HPDI engines. The new engine has also been used to study the combination of EGR and HPDI. While of a preliminary nature, the results indicate that significant NO x reductions can be achieved, with the greatest effects being found at low-speed, moderate-load operating conditions. Reductions in NO x emissions of as much as an order of magnitude were detected, but these extreme reductions came at the price of increased hydrocarbon emissions and reduced engine performance. More moderate reductions in NO x can be achieved with little penalty in either performance or emissions. ii Table of Contents Abstract ii Table of Contents iii List of Tables v List of Figures vi Nomenclature viii Acknowledgements ix Chapter 1 - Introduction 1 1.1 Solving the NO x problem 2 1.2 The SCRE 2 1.3 Project Goals and Objectives 3 1.4 Thesis Overview 4 Chapter 2 - Commissioning a Single Cylinder Research Engine for High Pressure Direct Injection of Natural Gas Testing 5 2.1 Introduction 5 2.2 Background 6 2.3 Experimental Apparatus 7 2.3.1 Control and Operational Systems 8 2.3.2 Instrumentation and theDAQ 10 2.3.3 Operational Methodology and Error Analysis 12 2.4 Results 13 2.4.1 Engine Baseline 14 2.4.2 Repeatability '. 15 2.4.3 Effect of Back-Pressure 16 2.4.4 Comparison to a Six-Cylinder Engine 17 2.5 Conclusions 22 Chapter 3 - NOx Reduction from a Natural Gas Fuelled HPDI Engine using EGR 24 3.1 Introduction 24 3.2 Background 26 3.2.1 NOx Formation Mechanisms 26 3.2.2 Methods for Reducing NOx 27 3.2.3 Exhaust Gas Recirculation 28 3.3 Experimental Apparatus 31 3.3.1 Experimental Methodology 33 3.3.2 Data Analysis and Experimental Uncertainty 35 3.4 Results 36 3.4.1 Effect of EGR on HPDI Operation 36 3.4.2 Influence of Engine Speed on EGR Effects 43 iii 3.4.3 Influence of Engine Load on EGR Effect 45 3.4.4 Effect of Replacement vs. Supplemental EGR 48 3.4.5 Effect of EGR Temperature 51 3.4.6 Effect of Gas Injection Timing 53 3.4.7 A Strategy for Applying EGR to an HPDI Engine 54 3.5 Conclusions 56 Chapter 4 - General Conclusions and Future Work 59 4.1 Conclusions 59 4.2 Recommendations for Future Work 61 References 64 Appendices 66 Appendix 1 - NO Formation Mechanisms 66 Appendix 2 - Relevant Equations 72 Appendix 3 - Idle Timing Sweep Results 75 Appendix 4 - Motoring Torques 77 Appendix 5 - Diesel Consumption Data 78 Appendix 6 - SCRE Instrumentation List 79 Appendix 7 - EGR Testing Procedures 80 Appendix 8 - Baseline and Comparison Test Summary 82 Appendix 9 - EGR Test Summary 85 iv List of Tables Table 2.1 Engine specifications 8 Table 2.2 Injector parameters 10 Table 3.1 SCRE characteristics 31 Table 3.2 Test points for EGR testing 34 Table A. 1 Motoring torque data 77 Table A.2 Approximate diesel consumption at each test point 78 Table A.3 SCRE instrumentation list 79 Table A.4 Baseline and comparison test summary 82 Table A.5 EGR test summary 85 v List of Figures Figure 2.1 Air exchange system schematic 9 Figure 2.2 Engine HPDI operating map, IMEP vs. engine speed 14 Figure 2.3 Repeatability of IMEP and THC emissions at 1200 RPM, 10 Bar IMEP 15 Figure 2.4 Effect of exhaust back-pressure on IMEP and NO x emissions, at 800 RPM and 10 Bar IMEP 17 Figure 2.5 Volumetric efficiency vs. speed at full rated load at each speed 18 Figure 2.6 Volumetric efficiency vs. load at 1200 RPM 18 Figure 2.7 ISFC vs. speed, at full rated load for the given speed 19 Figure 2.8 ISFC vs. load at 1200 RPM 20 Figure 2.9 NO x and THC emissions vs. load at ESC modes 5, 6, and 7 20 Figure 2.IOCO2 emissions vs. load at ESC modes 5, 6, and 7 21 Figure 3.1 SCRE air exchange system 32 Figure 3.2 NO x emissions vs. EGR fraction at 1200 RPM, 10 Bar IMEP 37 Figure 3.3 THC emissions vs. EGR fraction at 1200 RPM, 10 Bar IMEP 37 Figure 3.4 C0 2 emissions vs. EGR fraction at 1200 RPM, 10 Bar IMEP 38 Figure 3.5 ISFC vs. EGR fraction at 1200 RPM, 10 Bar LMEP 39 Figure 3.6 Peak pressure location and timing vs. EGR fraction at 1200 RPM, 10 Bar IMEP 39 Figure 3.7 Cylinder pressure trace at different EGR fractions at 1200 RPM, 10 Bar LMEP 40 Figure 3.8 Gross heat release rate at different EGR fractions at 1200 RPM, 10 Bar IMEP 41 Figure 3.9 Net heat release rate at different EGR fractions at 1200 RPM, 10 Bar LMEP 43 Figure 3.10 Maximum EGR fraction at 10 Bar JJVIEP as a function of speed 44 Figure 3.11 NO x emissions at 10 Bar IMEP as a function of EGR fraction at varying speeds 45 vi Figure 3.12THC emissions at 10 Bar IMEP as a function of speed and EGR fraction 45 Figure 3.13 Maximum EGR fraction at constant speed as a function of load 46 Figure 3.14NOx emissions at constant speed as a function of load and EGR fraction 47 Figure 3.15 THC emissions at constant speed as a function of load and EGR fraction ....48 Figure 3.16 Pressure trace of supplemental and replacement 20% EGR cases at 800 RPM, 10 Bar IMEP 49 Figure 3.17Net heat release of supplemental and replacement 20% EGR cases at 800 RPM, 10 Bar IMEP 50 Figure 3.18Peak cylinder pressure vs. EGR fraction for supplemental and replacement EGR at 800 RPM, 10 Bar IMEP 50 Figure 3.19NOx and THC emissions vs. EGR fraction for supplemental and replacement EGR at 800 RPM, 10 Bar IMEP 51 Figure 3.20Intake manifold oxygen concentration vs. EGR fraction for supplemental and replacement EGR at 800 RPM, 10 Bar IMEP 52 Figure 3.21 Effect of intake manifold temperature on NO x and THC emissions as a function of EGR fraction, at 800 RPM, 10 Bar IMEP 53 Figure 3.22 Effect of injection timing on NO x and THC emissions as a function of EGR fraction, at 1600 RPM, 10 Bar IMEP 54 Figure 3.23 NO x - THC tradeoff at ESC mode 5 for a range of EGR fractions 55 Figure 3.24NOx - ISFC tradeoff at ESC mode 5 for a range of EGR fractions 56 Figure A.l Effect of absolute timing of natural gas injection on NO x emissions at 600 RPM idle 75 Figure A.2 Effect of absolute timing of natural gas injection on THC emissions at 600 RPM idle 76 Figure A.3 Effect of absolute timing of natural gas injection on indicated power at 600 RPM idle 76 Figure A.4 Motoring torque as a function of engine speed and oil temperature 77 vii Nomenclature °CA Crank angle degree ATC After top dead centre B T C Before top dead centre CNG Compressed natural gas CO Carbon Monoxide C0 2 Carbon Dioxide EGR Exhaust gas recirculation F/A Overall fuel-to-air ratio HPDI High pressure direct injection HRco2 High range CO2 sensor reading (vol %) ikWhr Gross indicated power, in k Win-IMEP Indicated mean effective pressure ISFC Indicated specific fuel consumption K NO x correction factor LHV Lower heating value of the specified fuel LRco2 Low range CO2 sensor reading (vol. %) m Mass flow rate of indicated species MAP Manifold air pressure (kPa(g)) MAT Manifold air temperature (°C) MW Molecular weight of the specified species N Engine speed (RPM) nr Number of revolutions per cycle (=2) NGEOI Natural gas end of injection NGSOI Natural gas start of injection NO Nitric Oxide N0 2 Nitrogen Dioxide NO x Nitric Oxides (NO, N0 2 , ...) P Cylinder pressure Pind Indicated power Pinj Injection pressure of natural gas PM Particulate matter PSOI Pilot start of injection Qn Apparent net heat release rate RIT Relative injection timing (time after end of diesel injection) SCRE Single cylinder research engine Tjntake Ambient air temperature THC Total hydrocarbons Y Specific heat ratio CO Specific humidity of ambient air Volumetric efficiency Thermal (fuel conversion) efficiency P = Density of the indicated species viii Acknowledgements On this project, I was privileged to work with an outstanding group of enjoyable, dedicated and experienced personnel. First and foremost, I would like to thank my supervisors, Dr. Kendal Bushe and Dr. Philip Hill, for their invaluable support and guidance throughout this project. From them, I learned a great deal more than could ever be put in a thesis. Special recognition also needs to be given to the Westport staff, especially Mr. Dale Goudie and Mr. Paul Grant for their work on the initial installation of the facility, and to Dr. Sandeep Munshi, for his inestimable assistance through the ups and downs encountered in commissioning the SCRE. I also want to thank the staff at UBC, and particularly Mr. Richard Van Dolder, for their large contributions to the success of this project. Finally, I want to thank my family for their encouragement and advice, and Helen, to whom this report is dedicated and whose unwavering patience and support made the difference during my time at UBC. ix Chapter 1 Introduction The diesel engine is an efficient and reliable source of power, widely used in applications where reliability and efficiency are in high demand. But it is in medium and heavy ground transportation where the diesel engine is most prevalent, propelling most commercial transport vehicles. By providing the motive power for inexpensive, long distance haulage, the diesel engine has been instrumental in fuelling economic growth around the world. The rapid increase in the number of diesel-powered vehicles on the roads has emphasised one of the diesel's main drawbacks - its harmful emissions. These emissions include many different species, but among the most damaging are the oxides of nitrogen (NOx). NO x are key components in the formation of photochemical smog, one of the most prevalent, noxious, and hazardous forms of local air pollution. The effects of photochemical smog include reduced visibility and increased respiratory distress, especially among children and the elderly. Concern over the effects of poor air quality (especially in urban areas) has led to strict restrictions on vehicle emissions. Among the most tightly regulated of these pollutants are NO x, of which diesel engines produce a disproportionate share. To meet current and future emissions regulations, heavy-duty engine designers have been forced to develop novel solutions for reducing emissions, while retaining the reliability and efficiency that have made the diesel engine so popular. 1 1.1 Solving the NOx problem Researchers in the diesel engine field (strictly speaking, compression ignition engines burning diesel fuel) have investigated many different methods for reducing emissions. They have been hindered by the fuel being used - diesel is an innately 'dirty' fuel, and forms a major barrier to emissions reductions. One concept to overcome this problem, developed at UBC and in the process of commercialisation by Westport Innovations Inc., is to replace most of the diesel fuel with natural gas. The High Pressure Direct Injection (HPDI) of natural gas into the combustion chamber has been shown to dramatically reduce harmful emissions without affecting engine performance. However, to meet future emissions regulations, even the relatively clean emissions from current HPDI engines must be reduced. Another NO x reduction concept is the use of exhaust gas recirculation (EGR). EGR is commonly used in spark ignition engines, but its application to diesel engines has proved much more challenging. The primary hurdle is that, while EGR reduces NO x emissions, it increases the amount of fine particulate matter (PM) emitted. By using HPDI, PM emissions are greatly reduced; therefore, it is expected that EGR can be used to significantly reduce NO x, without increasing PM emissions to levels near those of a diesel engine. 1.2 The SCRE The application of EGR to an HPDI engine was the first set of experiments carried out on UBC's new single cylinder research engine (SCRE). The core of the SCRE facility is a state-of-the-art Cummins ISX heavy-duty truck engine, modified so that only 2 one cylinder is firing. This modification provides for research to be carried out on a modern truck engine, but without the costs, time delays, and experimental variabilites that plague multi-cylinder research engines. The SCRE facility includes all the auxiliary systems needed to run the engine and the instrumentation required to monitor and analyse the test results. The facility was designed from the outset for EGR testing, with provision for more control over air flow rates, temperatures, and pressures than is possible on a standard engine. This provides a unique system, ideal for investigating both 'regular' EGR and the many variables that can be modified to maximise EGR performance. 1.3 Project goals and objectives The SCRE project is part of a five year program, jointly funded by NSERC and Westport Innovations. The first challenge was getting the SCRE operational on HPDI and validating the experimental results. To accomplish this, three specific objectives were laid out: 1) To operate the SCRE fuelled by high-pressure natural gas. 2) To establish that the data generated by the SCRE is repeatable. 3) To develop a baseline map of the engine and show that the SCRE results are comparable to the 6-cylinder engine. Of special importance for this work was establishing the comparability between the SCRE and an equivalent six-cylinder engine. By identifying and explaining the key differences in the SCRE's performance and emissions, the scope of application for the SCRE data can be determined. With the engine operating satisfactorily, the actual research into EGR could be carried out. With the many variables that affect EGR's role in NO x formation, carrying 3 out a comprehensive study was beyond the scope of the current work. This investigation focused on: 4) Investigating NOx reduction through the use of EGR on an HPDI engine over a range of speed and load conditions. 5) Studying the effect on NOx emissions of supplementing, rather than replacing, the intake air flow with recirculated exhaust gas. 6) Developing a preliminary EGR + HPDI optimisation strategy. One of the major drawbacks of the application of EGR to diesel engines is increased PM emissions. The study of the effects of EGR on the PM emissions of an HPDI engine is the subject of research currently underway on the SCRE. By meeting the six objectives laid out above, the SCRE will be proven to be an effective research tool, and the basic effects of EGR on HPDI engine performance will be established. From these results, further research can be focused on optimising EGR, in conjunction with the many other variables that influence NO x formation, thereby achieving further significant NO x reductions from HPDI engines. 1.4 Thesis Overview This thesis is composed of four chapters, the first being this introduction. Chapter 2 focuses on the installation and commissioning of the SCRE, as laid out in objectives 1 -3 . Chapter 3 involves the preliminary EGR testing carried out on the SCRE, as identified by objectives 4 - 6 . Chapter 4 presents general conclusions and gives some suggestions for future work. Chapters 2 and 3 were written so that they may be submitted as separate articles for publication in technical journals with little or no modification and as such are intended to be self-supporting documents. 4 Chapter 2 Commissioning a Single Cylinder Research Engine for High Pressure Direct Injection of Natural Gas Testing 2.1 Introduction The diesel engine has a long history of efficient, reliable power production in many different applications. Escalating concern over internal combustion engine emissions has resulted in increasingly restrictive emissions regulations, especially on heavy-duty road use diesels. These restrictions have forced engine designers to reduce emissions of, among other species, nitric oxides (NOx) and particulate matter (PM). Conventional diesel engines are limited by the fuel being used, which is injected as an atomised liquid (high PM) and burns at high temperature (increasing NOx). One way to reduce these emissions is to replace the direct injection of liquid diesel with the high-pressure direct injection (HPDI) of natural gas [1]. PM is reduced by natural gas' lower carbon content as well as HPDI's primarily gaseous-phase injection. NO x emissions are reduced due to natural gas' lower flame temperature. An HPDI-based system is under development at Westport Innovations, in conjunction with the University of British Columbia (UBC). UBC's Alternate Fuels research group has received a new Cummins ISX400 engine, modified for single cylinder operation. With only one cylinder firing, cylinder-to-cylinder variability is removed and only a single prototype injector is needed. Unlike many other single cylinder engines, the UBC SCRE is fundamentally the same as a production ISX engine, maximising the similarities between the SCRE and equivalent six-cylinder engines. This helps to ensure that research carried out on the SCRE is directly applicable to multi-cylinder engines. 5 The purpose of the work described in this chapter is to establish how well results from the SCRE can be directly applied to the six-cylinder engine. The full range of SCRE operation has to be established and the reliability of the test results needs to be proved. The differences in the experimental results between the SCRE and a six-cylinder engine have to be determined. Finally, methods of how best to mitigate these effects are to be suggested. 2.2 Background The HPDI system involves the injection of natural gas fuel from a new, dual fuel, high pressure injector into an otherwise unmodified engine. HPDI requires a small amount of diesel fuel to ignite the natural gas; the bulk of the power is then delivered from the combustion of the natural gas. The quantity of pilot fuel varies as a fraction of the total heat release over the engine operating range between 3% and 10%. This system has been shown to retain the thermal efficiencies of the diesel engine [2] over most operating conditions, while significantly reducing emissions. In particular, NO x emissions are reduced by 40%. This is attributed primarily to the lower adiabatic flame temperature of the natural gas fuel [3]. A substantial reduction in particulate matter emissions has also been observed, on the order of 50-80%, although those results are of a preliminary nature. C O 2 emissions are also reduced, due to the lower carbon to energy ratio of the fuel, by approximately 25% [2]. While maintaining performance and reducing emissions, the HPDI system does not require significant modifications to the base engine. Aside from replacing the fuel injector, other required modifications include the removal of the engine's diesel fuel pump and the addition of compression systems for the natural gas and diesel fuels. The 6 hydraulic/mechanical control of the diesel injector is replaced by electronic control for the HPDI system, providing improved control and increased optimisation opportunities. The HPDI injection is a two-stage process, where the diesel fuel is initially injected, followed after a short delay by a natural gas injection through separate ports. The absolute and relative timing of the two injections affects the engine's efficiency and emissions. Changes in injection timing affect the peak cylinder temperature[4], which has a large impact on NO x formation due to the high dependence on temperature of the primary NO x formation mechanism (the extended Zeldovich mechanism[5]). Other emissions, including CO, total hydrocarbons (THC) and PM, are affected to a lesser extent. 2.3 Experimental Apparatus The core of the SCRE facility is a modified Cummins ISX400 series heavy-duty diesel engine, originally rated at a power of 300 kW and 1966 Nm of torque. The engine was modified by Cummins Inc. to allow for single cylinder operation (cylinder'6", nearest the flywheel, being the firing cylinder). The alterations included installing dummy injectors, blocking the intake and exhaust ports, and removing the piston rings from the non-firing pistons. These pistons were also drilled through (diameter = 30 mm) to reduce compression in the unused cylinders, with the removed mass being replaced by lead in the wrist pin to maintain the engine's balance. The modifications did not include any other changes to the internal workings of the engine - the fuel rails, internal air intake manifold, cam shafts and timing, firing cylinder piston, etc. are all the same as for a production engine. The actual specifications for the SCRE are given in Table 2.1. 7 Displacement 2.5 L Compression Ratio 19:1 Bore 137 mm Stroke 169 mm Connecting Rod Length 261 mm Rated Speed 1800 R P M Table 2.1 Engine Specifications. 2.3.1 Control and Operational Systems The largest drawback of the modified engine is that, at some operating conditions, the engine's frictional losses are greater than the power generated by the combustion in i the operational cylinder. To overcome this, supplemental torque is provided by a Baldor 40 hp electric motor, controlled by a 'vector'-type torque-control drive. The motor's power is insufficient to overcome the engine's internal friction at high speed, so that certain high speed, low load operating conditions are not achievable. Any excess power is dissipated by a General Electric model TG eddy-current type water-cooled dynamometer capable of absorbing 150 kW. The dynamometer is controlled by a Digalog dynamometer control unit, which uses PID control to hold the engine at the set speed. The dynamometer is directly coupled to the engine through a spider coupling, and the vector drive is connected to the dynamometer via a toothed belt. To avoid excessive warm-up times, an auxiliary heating loop composed of a timer-controlled water-circulating pump and heater, as well as an oil pan heater, is used. For cooling the engine, a shell and tube type coolant-to-water heat exchanger replaces the standard radiator. The engine thermostats regulate the engine temperature to 80°C under most operating conditions, except for some low speed/low load conditions. At these conditions the engine temperature drops slightly, as combustion in the single cylinder is not producing as much waste heat as is being lost to the surroundings. 8 To overcome the unusual conditions and air-line pulsations of single cylinder operation, a custom air-exchange system, shown in Figure 2.1, was installed. The intake side includes a Lysholm 1600 AX twisted lobe screw compressor type supercharger driven by a 40 hp electric motor, a water-cooled charge-air cooler, and a 35 gallon surge tank, located immediately upstream of the intake manifold. The surge tank removes at least 95% of the pulsations from the intake air stream [6]. After exiting the cylinder the exhaust passes through an exhaust surge tank and then through an electrically-controlled butterfly valve. This valve simulates the back-pressure that would be experienced if a turbocharger were present. The back-pressure valve can also be used to maintain a positive pressure gradient between the exhaust and intake lines, to drive the exhaust gas through the EGR loop. The EGR loop comprises another electrically operated flow-control valve and an air-to-water EGR cooler. The operational flexibility of the air system provides versatility not available on a standard engine. Intake surge tank Air filter 1 Engine Emissions Sample Lim Supercharger EGR cooler ' EGR Valve BP^valve Exhaust surge tank Figure 2.1 Air exchange system schematic. For the SCRE to be run on HPDI, diesel and natural gas flows at pressures exceeding 20 MPa are required. The natural gas is compressed by a two-stage system, with the final compression being supplied by a hydraulic intensifier feeding into high pressure storage tanks. The diesel fuel is compressed in one step by a Dynex-Rivett 9 5000 psi (345 bar) hydraulic pump driven by a 5 hp (3.73 kW) electric motor. After being compressed, the diesel is cooled to 40°C. Both high pressure fuels are supplied to the injector through the engine's internal fuelling rails. The design and operation of the dual-fuel injector has been extensively studied [7],[8]. The model used was a J31-019 HPDI injector with the following characteristics: Number of Hole Dia. Jet Angle Holes (mm) (from firedeck) Pilot (diesel) 7 0.12 18° Natural Gas 8 0.72 18° Table 2.2 Injector Parameters The injector was fully characterised and linearised before installation. Tests showed that the injector delivers natural gas at approximately 7 kg/hr per 1 ms pulse, while the diesel injection volume was 4 mm /injection for a similar 1 ms pulse [6]. 2.3.2 Instrumentation and the DAQ The data acquisition system (DAQ) is composed of National Instruments hardware, the core being a SCXI 1001 chassis and a PCI-MIO-16E-1 DAQ card, capable of 1.25 MHz sample speed. The signals from the instruments are collected by four plug-in modules, three for low speed (an SCXI 1102 for the thermocouples and SCXI 1100 and 1102B modules for the other sensors) and one (another SCXI 1100) for the high-speed signal from the in-cylinder pressure transducer. The low speed data is sampled four times per second, while the high-speed data is collected every lA degree of crankshaft revolution, at up to 22 kHz depending on the engine's speed. The low-speed data was collected at one second intervals and averaged over 20 samples; for the high-speed data, 20 sequential cycles were collected and averaged. 10 Data recorded by the SCRE includes air pressures and temperatures at multiple locations through the air system. Temperatures are measured using type 'K' thermocouples and pressures by Energy Kinetics type-209 transducers. The most critical pressure sensor is the in-cylinder transducer, a flush-mounted AVLQC33C which is water cooled to reduce thermal shock effects. The airflow is measured by a turbine-type Superflow meter, calibrated using a Superflow flowbench. The natural gas flow is recorded by a Micromotion coriolis-effect mass flow meter. Crank angle position is recorded at 0.5° crank angle resolution using a BEI XH25D shaft encoder. The torque being exerted by the dynamometer is determined using an Artech Industries load cell. An emissions bench with instruments for NO x, C O 2 (multiple ranges), CO, THC, and O2 was used to collect emissions data. The sample, taken from the exhaust stream near the exhaust port, runs through a sample line, pump, and filters (all heated to 190°C), and is then chilled to remove condensate. The NO x analyzer is a chemiluminescent type manufactured by Advanced Pollution Instrumentation (model 200 AH), with ranges between 5 and 5000 ppm. There are two C O 2 analysers, both non-dispersive infrared, a Beckman 880 running at 0-20% and a California Analytical Instruments model .100, run at either 0-2% or 0-10% (the latter is also used for EGR flowrate calculation). The CO is measured by a Siemens 2IP non-dispersive infrared sensor, the THC by a Ratfisch RS55 flame-ionization detector, and the O2 by a Siemens Oxymat 5E paramagnetic oxygen detector. The most significant problem encountered was the inability of the diesel flow measuring 'pail and scale' system to give an accurate reading. Vibrations and pulsations reduced the precision of the scale being used, while the scale itself (at 30 kg) was too. large to detect, in a reasonable time frame, typical SCRE HPDI diesel flow rates on the 11 order of lOOg/hr. For the test points of greatest interest, the diesel flow rate was determined by running the engine over an extended duration (at least one hour) and recording the change in mass over that time. Unfortunately, this method does not account for changes in diesel flow due to changes in fuel line or combustion chamber pressure, or any other transient effects. The experimental uncertainty of this method is estimated to be on the order of ±30%. However, as the diesel provides (at most) 10% of the total energy being delivered, this error in diesel flow rate will not seriously reduce the precision of the experimental results. 2,3.3 Operational Methodology and Error Analysis Operation of the SCRE is more complicated than a standard diesel engine, due to both the added complexities of single-cylinder operation and the greater number of fuelling variables to be controlled in the HPDI system. Basic engine operation is regulated by the fuel and air flow rates, engine speed, and injection timing. To maintain similarity with the 6-cylinder engine, engine speed and air and fuel mass flow rates were matched for each operating condition. To maximise comparability, the timings of the diesel and natural gas injections should also be matched. This does not account for slight differences between the commanded and actual injection timings, which are influenced by the injector itself as well as the injection control hardware. Due to internal friction losses, the SCRE's brake torque at any given operating condition is not comparable to the brake torque of a six-cylinder engine. All the results will therefore be compared on the basis of gross indicated performance. The indicated performance is itself subject to errors, which were quantified by inducing an error in the 12 raw pressure data and observing the effect on the calculated results. The IMEP was observed to vary by as much as ± 5% under the influence of a 0.5° offset in crankshaft position. The effects of errors in pressure transducer calibration, valve timing, and manifold pressure on IMEP were observed to be much less significant, not more than 2% for induced errors as large as 5%. The errors in the low-speed data were determined by evaluating the variation in repeated readings at a single speed, load, and timing condition, after all the mitigating factors had been removed. This results in an uncertainty of 5% in the NO x reading, 10% in the THC reading, 4% in the airflow reading, and 1.5% in the fuel flow reading. The fuel flow uncertainty leads to a 2% error in the ISFC. Other errors are noted as necessary in the analysis of the results. 2.4 Results A 28-mode test was used to evaluate the similarities and differences between the SCRE and six-cylinder HPDI engines. The 28 modes are a combination of four loads (25, 50, 75, 100%) at each of seven speeds (600, 800, 1000, 1200, 1400, 1600, 1800). The loads are specified as a percentage of the maximum load at that speed. The SCRE's maximum load is defined to be the same IMEP as that of an equivalent six-cylinder engine. As no injection timing data was available for the six-cylinder test set, the SCRE timing was adjusted to match the peak cylinder pressure and location from the six-cylinder testing. The second set of tests involved exactly matching the SCRE timing with that of a six-cylinder engine to study any discrepancies in emissions. The engine speed, air and fuel mass flow rates were also matched. The test points used for these conditions were from the ESC 13 mode test, specifically modes 5, 6, and 7 (1066 RPM, 50, 75, and 25% load). 13 2.4.1 Engine Baseline Matching SCRE fuel and air flow rates over the 28 points prescribed in the standard HPDI 28-point test results in the map shown in Figure 2.2. The low load / high speed area in the lower right hand corner of the plot, delimited by the 'limit' line, is the operating range where the engine is unable to operate due to inadequate supplemental power. In terms of the more common AVL 8-mode and ESC 13-mode test cycles, the unachievable points are AVL modes 5 and 6, and ESC mode 11. These points, and suggested alternatives, are labeled on Figure 2.2. The replacement points were chosen to ensure that the per-cycle air and fuel flow rates are the same. For ESC mode 11 and AVL mode 6, a point of 1450 RPM, 8.6 Bar IMEP is suggested. For AVL mode 5, 1000 RPM, 5.5 Bar IMEP is the best replacement available. While these points provide the best similarity to the test points, the effects of the lower speed on in-cylinder conditions can not be ignored. 20 15 5 25% Load 50% Load 75% Load -ii-100% Load —X— Limit Line K • / T "' \ / , ' 1 < T J ^ S C T A V L 6, E S C 11 i / ^s^^ Replacement 5——!^^^'^ + AVL 6 ^ ^ A V L 5 Replacement ^ + E S C 11 + AVL 5 0 500 1000 1500 2000 Engine Speed (RPM) Figure 2.2 Engine operating map, IMEP vs. engine speed, from 28-mode test. 14 2.4.2 Repeatability To establish the repeatability of the SCRE test data, two test points were selected (1200 RPM, 50% load, and 800 RPM, 25% load on Figure 2.2) and were repeated throughout the duration of the testing. A large number of samples were taken over an extended period of time, providing a better indication of the distribution around the mean for the experimental values. Some of the observed variations in experimental results are secondary effects resulting from variations in operating conditions. For example, the standard deviation of the recorded exhaust temperature is 4.1% of the mean. By including the effect of fuel flow rate (whose standard deviation is 5.5%), the standard deviation of the exhaust temperature is reduced to 2.1%. Similar effects are seen in other performance and emissions data. With these effects taken into consideration, the repeatability of most of the test parameters is acceptable. For example, the repeatability of the calculated IMEP and the measured THC emissions are shown in Figure 2.3. The relatively large amount of scatter in the THC data is not surprising, as the sensor's fixed range is 0-10 000 ppm. 12 Q. S X * X X X* X * * X X x * XX x X * * * x * * X * x x * • IMEP x T H C 160 120 E a 80 3 o X 40 15 Test points 30 Figure 2.3 IMEP and THC emissions vs. test number at 1200 RPM; 10 Bar IMEP; MAT=30°C, MAP=45kPa(g); PSOI=-3ms; RIT=2.5ms; PinJ=19 MPa. 15 2.4.3 Effect of Back-Pressure On the SCRE, the back-pressure normally exerted by a turbocharger is replicated by an electric valve in the exhaust line. Exhaust pressure data is not available for the six-cylinder engine, so the pressure setting was chosen to be the same as the intake manifold pressure. To establish the influence of back-pressure, a set of tests were run at a single operating condition where the back-pressure was varied from atmospheric pressure to 20% above the intake manifold pressure (0 - 60 kPa). The effects on performance (IMEP) and emissions (NOx) of the variations in back-pressure are shown in Figure 2.4. The effect on performance is very small, with the slight reduction in EVIEP (less than 1%) being within experimental uncertainty. The volumetric efficiency was reduced by approximately 1%, again well within experimental uncertainty. The effect on brake performance, while smaller than the experimental uncertainty (which exceeds 20%), does indicate a reduction in brake torque, as would be expected due to the increased pumping work required to expel the burnt gases. For the emissions, a change in 20% around the anticipated back-pressure results in a change of 2% in the NO x reading. At pressures approximating the exhaust manifold pressure (±15% of the intake manifold pressure) the effects of back-pressure on performance and emissions are, in all cases, smaller than the experimental uncertainty of the results. 16 18.00 15.00 12.00 o Z 9.00 ? n OQ ST 6.00 UJ 5 3.00 0.00 I i i 4—t I i i • IMEP X NOx (g/ikWhr) 15 30 45 Exhaust Backpressure (kPa(g)) 60 Figure 2.4 Effect of exhaust back-pressure on IMEP and NOx emissions, at 800 RPM and 10 Bar IMEP. 2.4.4 Comparison to a Six-Cylinder Engine Ideally, when comparing the performance and emissions of the SCRE with those of a six-cylinder ISX HPDI engine, all the performance and emissions data would be compared to one set of tests. Unfortunately no emissions or timing data was available for the 28-mode test used for performance comparison and no in-cylinder pressure data was available for the ESC mode tests used for emissions comparison. Due to the increased pulsations in, and lack of tuning of, the SCRE air exchange system, significant differences in the volumetric efficiency comparison are to be expected. The variation with speed can be clearly seen in Figure 2.5, with very similar results between SCRE and six-cylinder operation at the lowest speeds (600 and 800 RPM), but significant differences (as much as 10%) at higher speeds. This effect is seen to occur at all loads, as shown in Figure 2.6, where the magnitude of the offset can be seen to be 8%. While the differences are somewhat larger than expected, they might be explained by the possibility that the acoustic wave patterns experienced by the SCRE are 17 the SCRE are affecting its ability to expel exhaust and ingest fresh air. The influence of this effect on the SCRE's performance and emissions can be minimised by matching the six-cylinder's air mass flow rate. The resulting slight increase (between 1 and 5 kPa) in the intake manifold pressure should not have a significant impact on the combustion processes. 100 80 o J 60 o £ ui u 1 40 E 3 o > 20 • 6-Cyl ISX -o -SCRE 500 - 1000 1500 Speed (RPM) 2000 Figure 2.5 Volumetric efficiency vs. speed, at full rated load at each speed, from 28-mode test. 100 80 .£ 60 u E UJ o 5 40 o E 3 O > 20 • 6-Cyl ISX oSCRE 25 50 75 Load (% of max at 1200 RPM) 100 125 Figure 2.6 Volumetric efficiency vs. load (as a%of the maximum rated load), at 1200 RPM, from 28-mode test. 18 Of all the performance variables investigated, the volumetric efficiency was most significantly influenced. Better correlation between the SCRE and six-cylinder engines was found in, for example, the indicated specific fuel consumption. The differences between the two engines in ISFC, as a function of speed, are shown in Figure 2.7. At most speeds, the agreement is well within the experimental uncertainty for ISFC. Only at 600 RPM, where the two values differ by more than 20%, is the difference significant. There is no clear reason why the SCRE should consume more fuel at lower speeds. Similar results are observed as a function of the load, as shown in Figure 2.8. At higher loads, there appears to be good agreement between the SCRE and six-cylinder engines, but at the lowest loads, significant differences are again observed. The combination of these two observations indicates that there may be an effect of the significantly lower fuel flow rates at low load and low speed conditions. Whether this is an instrumentation issue or an operating effect for which no explanation is evident needs to be further investigated. Despite discrepancies such as this, the comparison between the performance of the SCRE and the six-cylinder engine is generally acceptable. 0.25 0.20 0.15 in 0.05 - • - 6 - C y l ISX - c i - SCRE 0.00 0 500 1000 1500 2000 Speed (RPM) Figure 2.7 ISFC vs. speed, at full rated load for the given speed, from 28-mode test. 19 0.05 0.00 -6-Cyl ISX -SCRE 25 50 75 Load (% of max at 1200 RPM) 100 125 Figure 2.8 ISFC vs. load (as % of the maximum rated load) at 1200 RPM from 28-mode test. For the ESC modes used to replicate emissions, the fuel and air mass flow rates were matched, as were the start-of-injection timings. The emissions of greatest interest to HPDI testing are NO x and THC, the comparisons of which are shown in Figure 2.9. The THC readings can be seen to be fairly similar between the two engines, with the largest difference being 20% - outside the experimental uncertainty of the SCRE data, but not including consideration of the uncertainty in the six-cylinder data. Figure 2.9 NOx and THC emissions vs. load at ESC modes 5, 6, and 7. 20 From Figure 2.9, the NO x readings can be seen to be significantly less similar, with differences as large as 50% being observed. One possible reason for the higher NO x emissions can be deduced from the plot of C O 2 emissions, as shown in Figure 2.10. The SCRE C O 2 emissions are about 20% higher at all test points. This indicates increased fuel consumption in the SCRE, which would result in higher temperature combustion (as the air flow rate is the same for the two engines) leading to higher NO x emissions. As no in-cylinder performance data is available for the six-cylinder engine, it is not possible to accurately compare the power produced from the two engines. The only method for ensuring equivalent power delivery is by matching the fuel flow rates. If one of the fuel flow rates contained a significant error, there would be no way to confirm this. Further support for this theory is provided by a carbon-balance analysis, which showed a significant discrepancy between inlet and exhaust carbon in the six-cylinder engine. This would also explain the smaller difference in the THC readings, as THC emissions are less sensitive to cylinder temperature than NO x. Further investigation into the accuracy of the fuel flow rate measurements from both the SCRE and the six-cylinder engine are needed. I IMEP (Bars) • Figure 2.10 C02 emissions vs. load at ESC modes 5, 6, and 7. 21 2.5 Conclusions The SCRE provides a unique facility for cheaper, quicker, and more precise research into HPDI engine operation. The equipment, including the modified engine and all the auxiliary systems required, has been shown to operate over most of the same conditions as an equivalent six-cylinder engine. Some high speed, low load test points were not achievable due to excessive internal friction, but substitute points have been identified which should provide adequate in-cylinder similarity. As well, the results from the SCRE have been shown to be repeatable and to give reasonable ranges of experimental uncertainty. Continued refinement of the apparatus and experimental techniques should further reduce these uncertainties. In comparison to an identical, unmodified ISX HPDI engine, the greatest performance difference can be seen to be in the engine's air exchange system. A significant reduction in volumetric efficiency at moderate and high speeds is observed, but can be compensated for by ensuring that air mass flow rate (rather than intake air pressure) is closely matched. This effect needs to be further examined, preferably through the use of an engine simulation program. Most of the other aspects of performance compare fairly well over most of the engine's operating range. The majority of emissions from the SCRE are similar to those from the six-cylinder engine. Significant differences were detected in the NO x emissions. Some discrepancies can be explained (and should be expected) due to differences in experimental techniques and equipment. Emissions (especially NOx) are so sensitive to changes in equipment and operating conditions that even a minor change can have a dramatic impact on NO x readings. This makes exactly reproducing the emissions from 22 the six-cylinder engine nearly impossible. However, the magnitudes of the NO x discrepancies are so large that further investigation into their causes are required. Overall, the SCRE can be used to represent the performance of an equivalent six-cylinder engine. It is not advisable to directly compare numerical values from the SCRE with data from a similar six-cylinder engine. However, trends and general effects that are identified on the SCRE are directly applicable to an equivalent six-cylinder engine. The SCRE should also be able to provide further insight into more fundamental questions involving HPDI combustion and operation. 23 Chapter 3 NOx Reduction from a Natural Gas Fuelled HPDI Engine using EGR. 3.1 Introduction The diesel engine is widely used in many applications that demand high efficiency, good reliability and low life-cycle costs. Common applications include marine propulsion and local power generation, but it is in heavy-duty ground transportation that diesel engines are most prevalent. The widespread use of diesels in heavy automotive applications has exacerbated concern over the diesel engine's main drawback, its harmful emissions. Among these emissions are nitric oxides (NOx), key participants in the formation of photochemical smog - one of the world's most severe local air pollution problems. The permitted emissions of smog-forming species, including NO x, are being reduced. Diesel engines, which produce a disproportionate share of NO x emissions, are being required to achieve dramatic reductions in a relatively short timeframe. This is forcing diesel engine designers to investigate aggressive methods for controlling NO x. One method for reducing harmful emissions while retaining the performance of a diesel engine is replacing the diesel fuel with high pressure direct injection (HPDI) of natural gas. This method, initially conceived at the University of British Columbia (UBC) and under development at Westport Innovations Inc., involves replacing the standard diesel injector with a more complicated dual-fuel injector. A small amount of diesel fuel pilot is used to initiate the combustion (as natural gas will not auto-ignite under the conditions found in a typical diesel engine), with the bulk of the energy 24 delivered from the combustion of the natural gas. This method has been shown to significantly reduce emissions of NO x, particulate matter (PM), and carbon dioxide (CO2). Another NO x reduction method, widely used in automotive spark ignition engines, is exhaust gas recirculation (EGR). EGR has been extensively studied for diesel engines, but has the drawback of increasing PM emissions at the same time as it reduces NO x. Not only are PM emissions harmful, but significant concerns have been raised regarding fouling of the engine's intake components by the recirculated PM, degrading the engine's performance and longevity. The use of HPDI has been shown to drastically reduce the PM emissions - so by combining EGR and HPDI, it may be possible to reduce NO x emissions, without an unacceptable increase in PM emissions. To investigate this, a new research facility was established in UBC's Alternate Fuels research group. This core of this facility is a Cummins ISX engine modified for single cylinder operation. Using this engine, a series of EGR tests were carried out to: 1) Establish that EGR and HPDI can be used in combination to reduce NOx without serious performance and emissions penalties. 2) Investigate the effect of EGR on the combustion processes in an HPDI engine. 3) Investigate the effects of intake air replacement and recirculation temperature on the emissions and performance of an HPDI engine. 4) Develop an initial strategy for optimising an engine using EGR and HPDI. Specific areas of study included the engine operating conditions where EGR can be most effective and what fractions of EGR are required to achieve the best results. Simultaneously, it was hoped that a better understanding of the effects of EGR on the HPDI combustion and NO x formation processes could be achieved. 25 3.2 Background Of the many pollutants emitted by internal combustion engines, nitric oxides (NOx) are among the most harmful. NO x (primarily NO and N O 2 ) are not directly hazardous at the concentrations found in the atmosphere. The damage is done when NO x reacts with hydrocarbons, oxygen, and sunlight to form photochemical smog. The health effects of the inhalation of smog include lung damage and cardio-pulminory distress, sometimes leading to premature death. In attempts to eliminate smog, the diesel engine has become a prime target. For heavy-duty road use diesels, the two most important current emissions standards are the European Euro V and American EPA's proposed 2007 standard. The Euro standards require NO x emissions to be reduced to 2 g/kWhr by 2008 [9], while the EPA has proposed limits of 0.2 g/bhphr (0.15 g/kWhr), to be fully implemented by 2010 [ 10]. 3.2.1 NOx Formation Mechanisms The formation of NO in internal combustion engines has been widely studied, with the hope of finding the key to reducing NO x emissions. Virtually all the NO x formed in an IC engine is formed as NO, some of which then reacts with excess oxygen to form N O 2 . There are four mechanisms that contribute to NO formation in an internal combustion engine [5]. The mechanisms, discussed in more detail in Appendix 1, are: 1) Thermal (extended Zeldovich)[5] 2) Prompt (Fenimore)[ll] 3) Nitrous Oxide[l2] 4) Fuel[l3] The most significant mechanism is the highly temperature-dependant thermal mechanism. Even small reductions in peak combustion temperature can result in a large reduction in 26 the rate of NO formation through this mechanism. The other mechanisms, while much less significant under diesel engine conditions, are not as temperature dependant. The prompt route depends primarily on the presence of CH radicals (limited to within the flame zone), while the nitrous oxide mechanism requires very high pressures to enhance its rate-limiting third body reaction. For diesel, the fuel route is largely irrelevant, as diesel fuel contains only minute quantities of nitrogen. Unrefined natural gas can contain up to 25% N2 [14], but most of this is gaseous and will participate in the reactions in the same way as atmospheric nitrogen, so is of little relevance. Most in-cylinder NO x reduction strategies focus on controlling the formation of thermal NO by reducing peak combustion temperature - but the roles of the other mechanisms may limit the NO x savings achievable in this manner. 3.2.2 Methods for reducing NOx Many methods have been investigated for reducing NO x emissions from compression-ignition (CI) engines. Three-way catalytic converters, widely and effectively used in spark-ignition engines, are not used in CI engines. Fouling concerns from particulate matter, the need for an overall near-stoichiometric air-fuel ratio, and the need for high temperatures in the exhaust to activate the catalysts are all problems yet to be overcome. A more advanced version, selective catalytic reduction, is under development. Problems with the durability of such a complicated system when exposed to diesel engine exhaust have thus far prevented its implementation [15]. Another more widely used method to reduce NO x is to simply retard the injection timing. This reduces NO x by delaying the combustion, which lowers the burning temperature (and hence 27 reduces thermal N0X) [4]; however, it also results in a thermal efficiency penalty and increased particulate emissions [16]. The preceding methods for reducing emissions from diesel engines are limited by the qualities of the fuel being used. By replacing most of the diesel fuel with the direct injection of high pressure natural gas, substantial emissions reductions can be achieved. This has been shown by research carried out at UBC and Westport Innovations Inc. working on the patented high pressure direct injection (HPDI) system ([1],[2],[3]). Of particular interest are the dramatic reductions in NO x (30-50%) and particulate matter (PM, 50-80%) which are achieved while maintaining the performance at the same level as that of a standard diesel fuelled engine. The NO x savings from HPDI are a direct result of the reduction in adiabatic flame temperature in switching from diesel to NG fuel. Unfortunately, not even current HPDI engines meet the more stringent future emissions targets. One method for further reducing NO x, widely researched for diesel engines, is the use of exhaust gas recirculation (EGR) [13]. The combination of EGR and HPDI offers the potential of further enormous NO x savings without impairing engine performance. 3.2.3 Exhaust Gas Recirculation EGR involves removing some of the gases from the exhaust stream and adding them to the engine's intake air stream. Commonly used for NO x control in spark ignition engines, EGR has been shown to work experimentally on direct injection compression ignition engines ([13],[17],[18], for example). Reductions between 50 and 100 K in combustion temperature (and corresponding reductions in thermal-generated NOx) at 28 moderate EGR fractions have been observed [19]. Increased PM emissions and fouling concerns have thus far prevented its widespread introduction to diesel engines [15], although these are being overcome. EGR reduces NO x by lowering the combustion temperature by diluting the oxidizing agent (the intake air) with exhaust gas. The presence of (primarily) inert species in the combustion zone results in reduced flame temperatures and longer mixing times. That this effect is the most significant influence of EGR has been shown by many researchers, for example [16], [18], [20]. Other effects that have been shown to be less significant include changes in specific heat, increased ignition delay, and the involvement of exhaust species in the combustion process [18]. If uncooled, EGR also acts by changing the intake manifold density by increasing the intake air temperature. Lower density means less mass in the combustion chamber to absorb the heat released during combustion, and hence higher in-cylinder temperatures [21]. By cooling the EGR, the in-cylinder temperature is decreased, which should reduce thermal NO x formation. This has not been conclusively shown, with some researchers (e.g. Durnholtz [22]) reporting no change in NO x, while others (e.g. Ladommatos [23]) report significant savings by cooling the recirculated exhaust gas. From an operational point of view, EGR can cause significant complications. By removing the exhaust gas upstream of the turbine and adding it downstream of the compressor, turbocharger fouling concerns can be alleviated. The resulting reduction in air mass flow rates will significantly affect turbocharger performance, influencing both the intake air boost pressure and the air mass flow rate into the engine. The exact influence depends on the engine and turbocharger operating characteristics. The two 29 extremes are that either the exhaust gas replaces an equal volumetric amount of fresh air, or the exhaust gas is completely supplementary to the existing intake airflow. Non-EGR and the two extreme conditions are shown in the following diagram, where the fresh airflow is white and the exhaust gas is grey: Non-EGR Condition Jj^ll Replacement EGR Condition Supplemental EGR Condition The effect of replacement vs. supplement on NO x emissions is still under debate, with some researchers indicating worsened emissions from supplemental (Uchida et al.[24]) and others showing improved emissions (Ladommatos et al., [23]). Knowing which extreme provides better performance is vital to maximising the effectiveness of the EGR system. A real engine will operate somewhere between these two extremes, depending on the operating characteristics of the engine and turbocharger. That EGR can be used to reduce NO x from diesel engines has been clearly shown. Ladommatos et al. [23] reported a 3-fold reduction in NO x emissions, while Durnholtz et al. [22] show similar results at low load, but reducing effectiveness as load is increased. Comparable results are reported by other researchers (for example Zelenka [15], Uchida [24], Lapuerta [25]). EGR does have other impacts on diesel engines, including increased PM emissions (by 57% at 16.8% EGR, Kreso et al. [26]). Increased fuel consumption (as much as 4% at 25% EGR) has also been observed, although this is removed by cooling the EGR [27] (similar results were reported by Durnholtz et al. [22]). AH of these studies were carried out on diesel-fuelled engines; the effects of EGR on compression-ignition, direct-injection, natural gas fuelled engines have not been studied. 30 3.3 Experimental Apparatus Studying the general effects of EGR applied to an HPDI engine was one of the main goals of UBC's new single cylinder research engine (SCRE) facility. The SCRE itself is a Cummins ISX 400 truck engine, modified so that only one cylinder fires. The modifications, carried out by Cummins Inc., result in a research engine that behaves like the standard engine, while removing cylinder-to-cylinder variability and requiring only a * single prototype injector. Table 3.1 gives the important characteristics of the SCRE: Displacement 2.5 L Compression Ratio 19:1 Bore 137 mm Stroke 169 mm Connecting Rod Length 261 mm Rated Speed 1800 RPM Table 3.1: SCRE Characteristics The modifications to the SCRE itself focused on reducing parasitic losses from the non-firing cylinders while maintaining engine balance and operability. More detailed information on the research apparatus, control systems, and instrumentation suite can be found in Chapter 2, along with a validation of the experimental systems. One of the most important components of the research facility is the system that controls the fuelling of the dual-fuel injector. The injector itself is a fully characterised J31-019 HPDI injector. Fuel is supplied to the injector at high pressures (around 20 MPa) from high-pressure storage tanks, and the timing of the diesel pilot and natural gas injections are controlled by a fuel delivery computer. The natural gas flow rate is measured by a Micromotion coriolis-force mass flow meter. Other key instruments include an AVL water-cooled in-cylinder pressure transducer, a BEI 0.5° crank angle resolution shaft encoder and a Superflow fan-type intake air flow meter. The instruments for measuring emissions include a chemiluminescent analyzer for NO x, non-31 dispersive infrared for CO and C O 2 , a flame-ionization detector for THC, and a paramagnetic sensor for O2. The SCRE's air exchange system, shown in Figure 3.1, was specifically designed to provide the flexibility of operation needed for EGR testing. The fresh air is compressed using a Lysholm supercharger powered by an electric motor, allowing better flexibility in intake air mass flows and pressures than could be provided by a standard turbocharger. After exiting the supercharger, the air is cooled in an air-to-water heat exchanger. Pulsations resulting from the engine's single cylinder operation are damped out using 35-gallon surge tanks on both the inlet and exhaust streams. The back-pressure normally induced by a turbocharger is simulated using an electrically controlled butterfly valve. Maintaining the exhaust pressure higher than the intake air pressure allows exhaust gas to pass through the EGR loop. While this results in a higher back-pressure than on a standard engine, the effects of this change were investigated and determined to be minimal. The EGR loop is comprised of an electrical control valve and another cooler. The quantity of EGR flow is controlled by adjusting the back-pressure, the opening of the EGR valve, and the speed of the supercharger (thereby adjusting the fresh air flow rate and intake manifold pressure). The EGR flow is determined by measuring the C O 2 concentration in the intake and exhaust streams. Intake surge tank Air cooler ^4 K Z h f r g ) Air filter Engine Exhaust 7u Ajj Supercharger EGR cooler EGR Valve =00= sample line Exhaust surge tank BP valve Figure 3.1 SCRE air exchange system. 32 3.3.1 Experimental Methodology To meet the stated objectives, a series of EGR tests at varying speeds and loads were carried out. The first set of tests involved running the engine over a range of speed-load setpoints, and carrying out tests at random EGR fractions between 0 and 40%. The EGR fraction (%) is defined by Heywood [13]: Tfl %EGR = x 100 (3. 1) m freshair + m EGR To ensure that the results at all EGR fractions were comparable, the injection timing was held constant at each speed-load point, but the fuel flow rate (controlled by the natural gas pulse width) was increased to maintain constant IMEP. The intake air temperature was held constant at 30°C to remove any influence of changes in intake air temperature. To carry out the testing, the back-pressure was set approximately 10 kPa higher than the intake pressure, and both intake and exhaust pressures were held constant for all EGR fractions by adjusting the back pressure valve opening and supercharger speed. The first set of testing was replacement EGR, using the test points detailed in Table 3.2. The table includes the manifold conditions and the timing of the diesel and natural gas injections. The diesel pilot start of injection (PSOI) is advanced with increased speed to compensate for the constant time delay between pilot and natural gas injections (relative injection time, RIT). The pilot start of injection is measured as the time before TDC at which the pilot injection starts (when the pilot solenoid closes). The RIT is the delay between the closing of the pilot solenoid valve and the start of the natural gas injection. 33 Test Speed Load IMEP Manifold Manifold PSOI RIT Point (RPM) (%r Pressure Temperature (Bar) (kPa(g)) CQ (ms/°ATC) (ms) 1 600 Idle 2 -1 30 -2.0 / -7.2 1 s 1.8 2 800 25 6 20 30 -2.0 / -9.6 1.8 3 800 60 10 40 30 -2.0 / -9.6 1.8 4 800 100 16 85 30 -2.1 /-10.0 1.8 5 1200 25 8 30 30 -2.2/-15.8 1.8 6 1200 50 10 50 30 -2.2/-15.8 1.8 7 1600 50 10 96 30 -2.25 / -22.0 1.8 Table 3.2: Test Point specifications for EGR testing During EGR testing, the engine manifold air temperature and pressure were held constant for all EGR fractions. This held the net mass airflow into the engine approximately constant. Changes in the intake charge density due to the presence of exhaust gases were fairly small (at most 3%), and their effect on the airflow is smaller than the experimental uncertainty. All the test points for this set were repeated twice, and any points that did not show good agreement were re-tested to ensure accuracy. The second set of tests involved the use of 'supplemental' rather than 'replacement' EGR. For these tests, the intake air mass flow rate was held constant, and the EGR fraction was increased by increasing the exhaust back-pressure, thereby forcing more exhaust through the EGR loop and raising the intake air pressure. The injection timing, IMEP, and intake air temperature were held constant. Test points 2, 3, and 5 from the preceding table were used, with EGR fractions up to 20% being studied. The final test investigated the effects of EGR temperature. Due to limitations of the experimental apparatus, carrying out true 'hot' EGR testing (where the EGR is mixed with cooled air just upstream of the intake manifold) was not possible. Instead, it was necessary to control the output temperature from the intercooler to try and simulate the intake manifold temperature of a 'hot' EGR system. Due to time and facility limitations, * % Load is defined as the % of the maximum indicated torque at the given speed. 34 it was only possible to carry out testing at 800 RPM, intermediate load (point 3) and with an intake temperature of 45°C. This should be sufficient to provide a preliminary indication of the effect of EGR temperature, but will not indicate the effect of truly 'hot' EGR at high EGR fractions. In carrying out this test, the manifold air pressure, fuel timing, and IMEP were held constant. 3.3.2 Data Analysis and Experimental Uncertainty The brake power developed by the SCRE is substantially different from a six-cylinder engine, and is often negative. As a result, most of the SCRE data is presented in indicated-specific terms, based on the indicated power calculated from the in-cylinder pressure trace. The pressure trace is also used to estimate the heat release rate, using[13]: f S ^ J L A - U * , (3.2) dt l - y dt y - l dt Where Q„ is the apparent net heat release rate, P is the cylinder pressure, Vis the cylinder volume, and y is the estimated specific heat ratio. To increase the precision in the NO x readings, the effect of day-to-day variations in ambient humidity had to be accounted for using an empirical correction factor (K), as suggested in [28]: 0.044x 0.0038 A •(G>-10.7)+1.8- -0.116x-+0.0053-(r^-29.4), (3.3) \ A K = \+l-where F/A is the overall fuel/air ratio and co is the specific humidity of the intake air. For the NO x readings, as well as the other recorded values, the experimental uncertainty was estimated by taking the standard deviation of the readings from a single operating condition, replicated over 30 times throughout the testing period. For 35 calculated values based on these readings, the results of the mean value, and the results of the mean offset by one full standard deviation, were used to determine the experimental uncertainty. One of the most important calculated results is the IMEP, which has an uncertainty of ± 5% due principally to a half degree uncertainty in the crank angle sensor. The ISFC has an uncertainty of ±1.5%, due to the errors in the IMEP and the fuel flow rate. The NO x reading has a standard deviation of ±5%, while the THC and CO readings vary by as much as 10%. The high uncertainties in the THC and CO readings are caused by experimental values which are a small fraction of the sensors' ranges (2% and 4%, respectively). Where possible, the uncertainty ranges are indicated on the corresponding graphs. Best-fit lines, with the order indicated in the caption, have also been used on some of the graphs to indicate general trends. 3.4 Results Both EGR and HPDI have significant influences on the combustion processes within the engine. To study these effects, a moderate speed, moderate load operating point (1200 RPM, 10 Bar IMEP) was selected. While the engine's operating condition will have an influence on the magnitude of EGR's effects, the results from this detailed analysis are likely to be indicative of the general influences of EGR. 3.4.1 Effect of EGR on HPDI Combustion Very significant reductions in NO x emissions can be achieved through the use of EGR on an HPDI engine, as shown in Figure 3.2. The NO x emissions drop almost linearly with EGR fraction, and at the highest achieved EGR fraction (32%), the NO x emissions are reduced by almost an order of magnitude. These reductions do come at a 36 price, as shown by THC emissions depicted in Figure 3.3. At low EGR fractions, increases in THC are slight (less than 10%, at 14% EGR), at which point a reduction of 45% in NO x has already been achieved. As the EGR fraction is increased further, THC emissions increase more rapidly, until at 32% EGR the THC emissions have increased by a factor of 6. Increased emissions of THC from the engine are indicative of poor combustion, with increasing quantities of fuel not fully reacting to reach final products. EGR Fraction (%) Figure 3.2 NOx emissions vs. EGR fraction at 1200 RPM; 10 Bar IMEP; MAT=30°C; MAP=45 kPa(g); PSOI=-2.2ms; RIT= 1.8ms; Pini=19 MPa. 10 20 30 40 EGR Fraction (%) Figure 3.3 THC emissions vs. EGR fraction at 1200 RPM, 10 Bar IMEP; MAT=30°C; MAP=45 kPa(g); PSOI=-2.2ms; RIT=1.8ms; Pinj=19MPa. Emissions of C O 2 , which can be used as an indication of how much of the fuel is being fully consumed, are shown in Figure 3.4. At moderate EGR fractions, C O 2 is not 37 affected, but at 27% EGR, a reduction of 10% in C O 2 emissions is observed. By carrying out a carbon-balance analysis, 80% of this reduction can be explained by the increased THC and CO emissions; the remainder have probably been emitted as increased particulate matter emissions (an estimated increase of 10 g(carbon)/kg of fuel). While this rather crude approximation needs to be experimentally verified (PM is the subject of a study currently under way on the SCRE), it does indicate that PM emissions may be significantly increased at high EGR fractions. Similar analyses at lower EGR fractions did not indicate significant PM increases, suggesting that moderate EGR operation may not have significant PM penalties. Figure 3.4 C02 emissions vs. EGR fraction at 1200 RPM, 10 Bar IMEP-MAT=30°C; MAP=45 kPa(g); PSOI=-2.2ms; RIT=1.8ms; ' Pinj=19MPa. EGR has also been reported to increase fuel consumption at high EGR fractions. Figure 3.5 shows the effect of EGR on indicated specific fuel consumption (ISFC). The ISFC is not significantly increased at moderate EGR fractions, but at higher rates, an experimentally significant increase of 2% in ISFC is observed. The explanation for the reduction in fuel efficiency can be seen from the peak cylinder pressure and location, as 38 shown in Figure 3.6. Moderate EGR fractions do not have a significant impact, but at high fractions, the peak pressure is delayed and reduced in magnitude. This lower pressure combustion will have a lower thermal efficiency, increasing the ISFC. 140 !§ 1 2 0 o u_ w 100 20 EGR Fraction (%) Figure 3.5 ISFC vs. EGR fraction at 1200 RPM; 10 Bar IMEP; MAT=30"C; MAP=45 kPa(g); PSOI=-2.2ms; RIT=1.8ms; Pinj=19 MPa. T 120 ra CQ I 8 80 P n 0. 40 -Peak Pressure (Bars) - Peak Pressure Location -fl O I-20 < 10 20 40 EGR Fraction (%) Figure 3.6 Peak pressure location and timing vs. EGR fraction at 1200 RPM, 10 Bar IMEP; MAT=30°C; MAP=45 kPa(g); PSOI=-2.2ms; RFT= 1.8ms; Pinj=19 MPa.. The effect of EGR on the in-cylinder conditions is clearly shown by the pressure trace, Figure 3.7. An EGR fraction of 14% appears to have no significant effect on the pressure trace, while a higher fraction (27%) significantly retards (5°CA) and reduces the 39 magnitude (10% lower) of the peak pressure. Later and lower pressure combustion contributes to the significantly lower NO x readings (lower pressure indicating less heat release, lower temperature combustion and hence less NO x formation through the thermal mechanism) at high EGR fractions. This is not the primary source of NO x reductions, as at moderate EGR fractions the peak cylinder pressure is not significantly different from the non-EGR case. The primary source of NO x reductions is the lower local flame temperatures induced by dilution of the charge oxygen mass fraction, as was experimentally verified in diesel engines by Ladommatos et al. [23] and Ropke et al. [29] - but this can not be experimentally confirmed with the current research equipment. The same reasoning can be applied to the THC emissions, where at moderate EGR fractions the change in flame temperature will not have a large impact on emissions. At higher EGR fractions, a lack of readily available oxygen, combined with later combustion, results in less of the fuel reaching its final products with more being emitted as partially burned hydrocarbons. 42& I , 9-1 , , 1 -40 -20 0 20 40 60 Crank Angle (°ATDC) Figure 3.7 Cylinder pressure trace for different EGR fractions at 1200 RPM, 10 Bar IMEP; MAT=30°C; MAP=45 kPa(g); PSOI=-2.2ms; RIT= 1.8ms; Pinj=19MPa. 40 Another effect of EGR is the increased variability of the combustion process at higher EGR fractions. With no EGR, the cycle-to-cycle standard deviation of the peak pressure is 0.5%, but this increases to 3.5% at 34% EGR. This increase is not linear, but rather is similar to the THC emissions; the standard deviation increases gradually at moderate EGR fractions (to 0.8% at 15% EGR), but then increases dramatically at high EGR fractions. These increased variations in the cycle-to-cycle conditions are indicative of poor combustion induced by excessive EGR fractions. The effect of EGR on the gross heat release profile is shown in Figure 3.8. This plot clearly shows that, as the EGR fraction is increased, the peak heat release occurs later in the combustion process and has a larger magnitude. This indicates that combustion is occurring later in the cycle, and that the fuel is being consumed faster when it does burn. As the injection timing is constant, this suggests an increased mixing delay caused by the presence of exhaust gases, followed by more pre-mixed combustion. The large amplitude fluctuations of these heat release curves are believed to be a result of noise on the pressure signal. 200 n : 1 i 120 .2 2 g> 80 n> O) S 4 0 cn CD j= 0 0 4 8 12 16 20 Crank Anlge (°ATC) Figure 3.8 Heat release rate for different EGR fractions at 1200 RPM; 10 Bar IMEP; MAT=30°C; MAP=45 kPa(g); PSOI=-2.2ms; RIT=1.8ms; Pinj=19 MPa. 41 The net heat release (Figure 3.9) is also interesting, especially in the relation between the rate of fuel consumption and the timings of the diesel and natural gas injections. The pilot start of injection can be seen to precede the first observable ignition event by approximately 7°CA. About 5°CA after the ignition of the pilot, the natural gas injection begins. Immediately after the commanded end of the natural gas injection, the net heat release curve begins to increase rapidly, indicating ignition of the natural gas. In the case of 27% EGR, the delay between natural gas injection and the increase in combustion rate is increased by 4°CA. As the slopes of the curves preceding the natural gas injection are not significantly different, it can be postulated that the diesel pilot has ignited at approximately the same point. The delay in the natural gas combustion indicates that the diesel pilot is not igniting the natural gas as quickly at high EGR fractions as at lower fractions. This could result from a lack of oxygen in the immediate vicinity of the injector, where the diesel pilot may have consumed much of the available oxygen. The natural gas has to penetrate further into the chamber to reach a near-stoichiometric fuel/air ratio, delaying the bulk of the natural gas combustion. This would also explain the onset of this condition - at moderate EGR fractions, there is still excess oxygen in the vicinity of the injector after pilot combustion. The same can be said for the THC emissions, which will be increased by insufficient available oxygen to complete the oxidation of the fuel. The corresponding loss of efficiency due to the increased hydrocarbon emissions can be seen in the lower net heat release. 42 100% PSOI L / N G E O I g 80% o 0% * 27% E G R • No E G R • 14% E G R -20 -10 0 10 20 30 40 Crank Angle (°ATC) Figure 3.9 Net Heat Release at different EGR fractions at 1200 RPM; 10 Bar IMEP; MAT=30°C; MAP=45 kPa(g); PSOI=-2.2ms; RIT=1.8ms; Pinj=19 MPa. 3.4.2 Influence of Engine Speed on EGR Effects The primary effect of engine speed on the application of EGR is the maximum EGR fraction that can be sustained before the combustion variability becomes excessive. To develop a meaningful comparison of the possibilities for NO x reduction at various operating conditions, a definition of the maximum EGR fraction is necessary. This could be defined as the maximum amount of EGR at which the engine will continue to operate, but this is operationally impractical due to excessive THC emissions and impaired performance. A more reasonable (albeit arbitrary) definition of maximum EGR is the EGR fraction at which the peak cylinder pressure falls to 90% of its maximum value. At most test points, this value coincides approximately with a doubling of the standard deviation of the peak pressure, as well as being the point at which THC and CO emissions begin to increase dramatically. To determine the effect of engine speed on the application of EGR, tests were carried out over a range of speeds while the load was held constant at 10 Bar IMEP. The 43 results, using the definition of maximum EGR fraction outlined above, are shown in Figure 3.10. The maximum EGR fraction decreases linearly by about 2% for every 100 RPM increase in speed. This can be attributed to the reduced overall air/fuel ratio, where the quantity of excess air is reduced as the engine speed increases. As well, more residual exhaust gases are being retained (due to the engine's lower volumetric efficiency at higher speeds), resulting in a larger burned gas fraction in the combustion chamber. 40 0 , 1 600 1200 1800 Speed (RPM) Figure 3.10 Maximum EGR fraction for stable combustion at 10 Bar IMEP as a function of speed. EGR's effect on NO x emissions as a function of speed is shown in Figure 3.11. The most effective NO x reduction can be seen to occur at lower speeds. This is a result of the greater quantity of exhaust gas that can be recirculated without seriously degrading the combustion. Similar reasoning can be used to explain the effects on THC emissions, as shown in Figure 3.12. These results show that as the speed is increased, both the absolute amount of THC emitted, as well as the percentage increase caused by the maximum EGR fraction, are increased. In terms of performance, there does not appear to be any significant relationship between fuel consumption and speed at the maximum EGR fractions, with variations in ISFC between 0.25 and 1.5% that are within the 44 experimental uncertainty of the ISFC. Overall, the effects of EGR are greatest at low speeds, decreasing as the speed is increased. This is due primarily to the ability to operate at higher EGR fractions, allowing for greater NO x reductions, as a result of higher air/fuel ratios at lower speeds. 40 £ 30 A • 800 RPM • 1200 RPM A 1600 RPM *j: 7*" 0 10 20 30 40 50 EGR Fraction (%) Figure 3.11 NOx emissions at 10 Bar IMEP as a function of EGR fraction at varying speeds. The curve-fits are linear. 70 0 J , , , , 1 0 10 20 30 40 50 EGR Fraction (%) Figure 3.12 THC emissions at 10 Bar IMEP as a function of speed and EGR fraction. The curve-fits are cubic. 3.4.3 Influence of Engine Load on EGR Effects Using the same definition as in the previous section, the maximum EGR fraction as a function of load (at 800 RPM) is shown in Figure 3.13. The maximum EGR fraction 45 at high load is 30% lower than that at low loads. This is to be expected, as at higher loads, more fuel is being injected, and the overall air/fuel ratio will be lower with less excess oxygen. This leaves less capacity in the combustion chamber to replace oxygen with exhaust species before the combustion process starts to suffer from a lack of readily available oxygen. Surprisingly, the maximum EGR fraction at 6 Bar IMEP is lower than that at 10 Bar. There is no clear reason for this, although one possible explanation is that the very short duration of injection (around 0.36 ms) of natural gas at 6 Bar is more subject to ignition problems caused by a lack of fuel than injections at higher loads. Figure 3.13 Maximum EGR fraction for stable combustion at 800 RPM as a function of load In engine performance there does not appear to be a significant relationship between the maximum EGR fraction and the ISFC. All the ISFC increases between the non-EGR and maximum EGR cases were small (less than 2%) and well within the experimental uncertainty, although the general increase in ISFC at high EGR fractions is confirmed. The effect of EGR on emissions as a function of load is also interesting. Figure 3.14 shows the effect of EGR on NO x emissions at the three load conditions studied. The intermediate load condition, which had the highest maximum EGR fraction, 46 also achieves the greatest NO x reductions, of 84% between the maximum EGR and non-EGR conditions. The capacity for NO x reduction at full load (16 Bar), was almost as good, with an 80% reduction achieved at the maximum EGR condition. The least capacity for NO x savings were found at 6 Bar, which saw only a 70% reduction. This coincides with the lower than expected maximum EGR fraction at this load condition. The THC emissions also follow this trend, as shown in Figure 3.15. The greatest increase in THC is observed at 6 Bar, where the THC emissions are increased by a factor of 2.4. The hydrocarbon emission increases at 10 and 16 Bar are almost equal at twice the non-EGR conditions. It is interesting to note that similar limitations at low load were found at 1200 RPM, where both the maximum EGR fraction and the resulting NO x reductions were smaller than at intermediate load, although this effect was not of the same magnitude as was found at 800 RPM. Another interesting observation is that, at high loads, the NO x reductions were found to be non-linear at high EGR fractions. This might indicate that a different NO x formation mechanism (other than the thermal mechanism) is becoming more relevant. A similar effect is reported by Plee et al. [20] at high EGR fractions on a diesel engine. Further research is needed to validate this supposition. 100 0 -I , , , , , . -—I 0 5 10 15 20 25 30 35 40 EGR Fraction (%) Figure 3.14 NOx emissions at constant speed as a function of load and EGR fraction at 800 RPM. 6 Bar curve fit is linear, others are cubic. 47 120 • 6 Bar • 10 Bar A 16 Bar -——*——-— £ t l . u l 11 0 5 10 15 20 25 30 35 40 EGR Fraction (%) Figure 3.15 THC emissions at constant speed as a function of load and EGR fraction at 800 RPM. Curve fits at 10 and 16 Bar are linear, at 6 Bar is cubic. Despite these unexpected findings, it is apparent that EGR can be successfully employed over a wide range of speed and load conditions to reduce NO x emissions from an HPDI engine. The greatest potential EGR savings can be accomplished at low-to-moderate speeds, as would be expected and has been found by other researchers studying diesel-fuelled engines, and at moderate loads. 3.4.4 Effect of Replacement vs. Supplemental EGR Whether the exhaust gas should replace, or be added supplementary to, the intake fresh air flow is directly relevant to the design of the EGR system. To optimise the EGR system, it is important to identify what effects the two extreme cases (fresh air replaced by an equal mass of exhaust, or fresh air mass flow retained constant) have. For this study, a low-speed, moderate load (800 RPM, TO Bar IMEP) test point was used. Experimental limitations restricted the EGR fraction to 20% for the supplemental test. The effect of supplemental vs. replacement EGR on the in-cylinder conditions are 48 clearly shown in Figures 3.16 and 3.17, the pressure trace and net heat release analyses for 20% EGR. The pressure trace shows the effect of the greater charge mass in the supplemental case, resulting in a peak pressure 20 bars higher than for the replacement EGR condition. The supplemental EGR peak pressure also occurs one crank angle degree after the replacement case's. This can be explained by the slower combustion, as shown by the lower slope of the combustion period on the net heat release plot. This plot also shows how the natural gas' ignition delay is shortened by 1°CA by the supplemental EGR. The effect of supplemental EGR on cylinder peak pressure at all EGR fractions is shown in Figure 3.18. The steady increase, which is expected because of increased charge mass, indicates that supplemental EGR will probably be limited at high loads by the peak cylinder pressure. 160 0 ^ , , , , — | -40 -20 0 20 40 60 Crank Angle (°ATC) Figure 3.16 Pressure trace of supplemental and replacement 20% EGR cases at 800 RPM; 10 Bar IMEP; PSOI=-2ms; MAT=30°C;Pinj=19MPa. 49 Crank Angle (°ATC) Figure 3.17 Net heat release of supplemental and replacement 20% EGR cases at 800 RPM; 10 Bar IMEP; PSOI=-2ms; MAT=30°C;Pinj=19MPa 160 40 -Q ' " I • i —• • Replacement EGR • Supplemental EGR 0 5 10 15 20 25 EGR Fraction (%) Figure 3.18 Peak cylinder pressure vs. EGR fraction for supplemental and replacement EGR at 800 RPM; 10 Bar IMEP; PSOI=-2ms; MAT=30°C;Pi„j=19MPa. Curve-fits are linear. The effect of supplemental vs. replacement EGR on emissions is shown in Figure 3.19. While the observed 3% increase in NO x emissions at 20% EGR is less than the experimental uncertainty, a slight increase is indicated by the 10% lower slope of the linear fit line for supplemental EGR NO x emissions. There was no discernible effect of supplemental EGR on the THC emissions. The effect on performance is not significant, the slight observed reduction in ISFC (V2V0 at 20% EGR) due to supplemental EGR being 50 within experimental uncertainty. Supplemental EGR also increases the volumetric concentration of oxygen in the charge air, as shown in Figure 3.20. The concentration of oxygen in the intake air increases by 13% over the range of EGR fractions studied, while the volumetric oxygen concentration for the replacement EGR case decreased 10% at the same fraction. This seems to contradict the importance of the volumetric oxygen concentration in NO x formation through the use of EGR, as suggested by, among others, Li et al. [16]. The increased oxygen concentration does suggest that supplemental EGR should not be as susceptible to the combustion degradation that limits the application of replacement EGR. This could lead to a significant increase in the maximum EGR fraction, and a corresponding further reduction in NO x emissions. The mass fraction of oxygen in the intake charge is does not differ substantially between supplemental and replacement EGR. This indicates that it is the reduced oxygen mass fraction, rather than the volumetric concentration, which causes the observed reduction in combustion temperatures and the corresponding decline in NO x production. 100 • NOx (supplement) A THC (supplement) • NOx (replacement) in 20 1 0 * 0 5 10 15 20 25 EGR Fraction (%) Figure 3.19 NOx and THC emissions vs. EGR fraction for supplemental and replacement EGR at 800 RPM, 10 Bar IMEP. The NOx curve-fits are linear, while the THC is quadratic. 51 0.016 0.012 " E O.C a. 0.008 3" 0.006 0.004 • replacement EGR o supplemental EGR 10 15 EGR Fraction (%) 25 Figure 3.20 Intake manifold oxygen concentration vs. EGR rate for supplemental and replacement EGR at 800 RPM, 10 Bar IMEP. Curve-fits are linear. 3.4.5 Effect of EGR Temperature The effect on engine performance of increasing the manifold temperature while holding the intake manifold pressure constant was negligible. The effect on emissions, while small, is significant as shown in Figure 3.21. The NO x emissions were increased by 5 g/kg of fuel for all EGR rates by increasing the intake manifold temperature. A slight (although non-significant) reduction in THC emissions was also observed. This indicates that the EGR temperature will have a significant impact on emissions. Further testing is needed into the effect of 'hot' EGR at all EGR fractions - the 45°C intake temperature relates to hot EGR at an EGR fraction of only 10%. The effects of varying the temperature should be investigated at different speed and load conditions. Likewise, the effect of increasing the intake air temperature, while maintaining a constant mass flow rate (leading to higher intake air pressures) should be studied. 52 100 80 ° 60 40 E uj • NOx - cold O THC-cold • NOx - hot • THC-hot • " \ 20 30 EGR Fraction (%) Figure 3.21 Effect of intake manifold temperature on NOx and THC emissions as a function of EGR fraction, at 800 RPM, 10 Bar IMEP. Cold MAT is 30°C and Hot MAT is 45°C. NOx curve-fits are cubic, THC quadratic. 3.4.6 Effect of Gas Injection Timing To carry out a preliminary investigation into the interactions between EGR and injection timing, the 1600 RPM, 10 Bar IMEP test point was repeated using two different timings. The diesel pilot injection time (PSOI) was held constant, and the relative natural gas timing (RIT) was shifted from 1.8 to 2.5 ms (equivalent to 5 crankshaft degrees), resulting in absolute natural gas injection timings of 1.3° before TDC and 3.3° after TDC. By advancing the timing, the peak pressure is increased by 12%, (due to combustion occurring closer to TDC), resulting in a higher temperature flame, increasing NO x emissions by 40%. This, along with the effect of introducing EGR, is shown in Figure 3.22. As can be seen, significant NO x reductions are achieved for both timings. The NO x emissions from the more advanced timing remain 40% higher at all EGR fractions. The maximum achievable EGR fraction is increased by advancing the timing, from 28 to 34%, but this only results in a further NO x reduction of 5% over the minimum 53 N0X at the more retarded timing. Advancing the timing had no significant influence on the THC emissions, which increased as a function of EGR fraction for both timings. This is offset by an increase in fuel efficiency of approximately 1.0% resulting from the more advanced timing. At the more advanced timing the THC emissions at the maximum EGR fraction are 20% higher than those for the retarded timing. Offsetting this against the small reduction in NO x indicates that simply achieving a higher EGR fraction will not necessarily improve the tradeoff between NO x and THC. These results indicate that the best strategy for using EGR on an HPDI engine should be to adjust the timings for low NO x, and then use a moderate EGR fraction to further reduce NO x emissions. 80 _ 60 3 40 A NOx RIT=2.5ms • THC RIT=2.5ms AN0xRIT=1.8ms OTHC RIT=1.8ms ' ^ > < r ^ ' 0 10 20 30 40 EGR Fraction (%) Figure 3.22 Effect of gas injection timing on NOx and THC emissions as a function of EGR fraction, at 1600 RPM, 10 Bar IMEP. PSOI=-3ms; Pinj=19MPa. Curve-fits are quadratic. 3.4.7 A Strategy for Applying EGR to an HPDI Engine The goal of EGR optimisation is to minimise NO x while maintaining performance and without significantly increasing other harmful emissions. The first step in applying EGR should be to reduce NO x emissions by retarding the timing, without incurring large penalties either in performance or CO and THC emissions. From there, further NO x 54 reductions can be achieved through the use of moderate quantities of EGR. There will be a trade-off between the achievable NO x reductions and increased THC emissions, as shown in Figure 3.23. As shown, there is potential for further NO x savings, but it comes at the cost of increased hydrocarbon emissions. A similar plot was prepared for the NO x -ISFC tradeoff, as shown in Figure 3.24. Moderate EGR fractions do not impose a significant efficiency penalty. It is only at the highest EGR fraction that a significant increase in fuel consumption is observed. From inspection of these two plots, a reasonable EGR fraction in the vicinity of 10% EGR can be suggested for these operating conditions. This results in a reduction in NO x of 40% over the baseline case, with no significant effect on efficiency. The cost is in the THC emissions, which are correspondingly increased by 50%. In absolute terms, this refers to a reduction in a brake specific NO x (for an equivalent six-cylinder engine's brake power) emission of 2.1 g/kWhr. The minimum achievable emissions rate is 1.18 g/kWhr. Similar procedures could be repeated for each operating point until a comprehensive engine map has been developed. This is just a preliminary technique - further research is required to reduce the THC emissions while maintaining engine performance and maximising NO x reductions. 10.0 T 1 0.0 1.0 2.0 3.0 4.0 NOx (g/ikWhr) Figure 3.23 NOx - THC tradeoff at ESC mode 5 for a range of EGR fractions. N=1067RPM; IMEP=10Bar; PSOI=-0.54ms;RIT=1.8ms; MAP=55 kPa; MAT=30°C; Pinj=19 MPa. 55 200 180 160 {£ 140 w 120 100 20% EGR 16% EGR | ^ /10% EGR No EGR 3% EGR 0.0 1.0 2.0 3.0 NO„ (g/ikWhr) 4.0 Figure 3.24 NOx - ISFC tradeoff at ESC mode 5 for a range of EGR fractions. 3.5 Conclusions The application of EGR to an HPDI engine has been shown to have the potential to significantly reduce NO x emissions. Moderate amounts of EGR do not seriously degrade performance or increase other harmful emissions, while more than halving NO x emissions at most operating conditions. When large EGR fractions are used, significant increases in THC emissions and in ISFC are observed. It is these factors, which are a result of increased combustion variability, that limit the quantity of exhaust gas which can be recirculated. The maximum EGR fraction at which the combustion degradation becomes significant varies with engine speed, load and injection timing. The maximum EGR fraction is lower at high speeds and high loads, due to the reduced overall air/fuel ratio. Contradictory to other research, the maximum EGR fraction at low loads is also lower than at part loads. Further research is needed to fully explain this phenomenon, which appears to be an effect of the HPDI system's two-stage injection process. One aspect of EGR that has featured prominently in diesel engine EGR research 56 temperature has very little effect on the engine's performance, but that cooling the EGR will reduce NO x emissions. More testing is needed, over wider ranges of intake temperatures and engine operating conditions, to establish the significance of this result. Another area of debate in EGR studies is the effect of 'supplemental' vs. 'replacement' EGR. Supplemental EGR does not offer significant improvements in performance at moderate EGR fractions, while the NO x emissions are slightly increased. One area where supplemental EGR has been shown to have a significant influence, but which was not studied, is in increasing the maximum EGR fraction. This could be of importance at high speed, high load conditions, where the maximum replacement EGR fraction will be limited by the overall air/fuel ratio. It could also have a significant influence on reducing the harmful effects of EGR at conditions where the timing has already been retarded to optimise NO x reductions. A strategy for maximising the applicability of EGR to an HPDI engine has been suggested. EGR is best applied to an engine whose timing has already been optimised to reduce NO x for non-EGR operation, without large sacrifices in fuel efficiency or other emissions. Moderate fractions of EGR should then be added (the decision on how much depends on the tradeoff between THC and NO x emissions), so that significant NO x savings can be achieved without seriously increasing THC emissions or the ISFC. Further research is required to realise the full potential for NO x reductions provided by EGR. Areas to be studied more carefully, other than those already mentioned, include the influence of injection timing on EGR application at varying speeds and loads. A thorough study of the effects of EGR on PM emissions needs to be carried out. More test sets are needed to optimise all the aspects of the combustion 57 process to get the maximum possible NO x reductions. However, the potential for significant NO x reductions through the use of EGR on an HPDI engine has been clearly shown. 58 Chapter 4 General Conclusions and Future Work The SCRE has great potential to provide a flexible and reliable research platform for investigations into HPDI operation and optimisation. As would be expected from such a unique test facility, the results from the initial tests are of almost as much use in detecting problems with the research equipment as they are in finding new facts about the area of study. Even so, the initial testing was successful in breaking new ground by achieving substantial NO x reductions from a natural gas fueled, HPDI engine. 4.1 Conclusions The first goal of this project was to get the SCRE operational on HPDI. This was successfully accomplished, complete with the auxiliary systems required to operate the engine and to record the experimental results. The results have been shown to be repeatable, although some of the experimental uncertainties are larger than is desirable. Most test points of interest are accessible by the SCRE. A few high speed, low load points are not achievable, where the engine's frictional losses are too great to be overcome. When operational parameters are matched, the SCRE's performance closely resembles that of a six-cylinder HPDI engine in most aspects. There are some discrepancies, including volumetric efficiency, but the effects of these can be mitigated with careful experimental techniques. The emissions from the SCRE do not compare as favourably, with some very significant differences being observed. To a certain extent 59 this is to be expected, as even small differences in operating parameters can have dramatic influences on emissions of some species. These discrepancies do mean that absolute values from the SCRE can not be used interchangeably with six-cylinder data. While direct comparison of SCRE and six-cylinder engine results is not advisable, the SCRE can be used to carry out research that is directly applicable to similar six cylinder engines. Trends and relative differences in SCRE emissions closely resemble those of the six-cylinder engine. The SCRE is capable of testing and optimising almost every aspect of HPDI engine operation, and doing it more economically than could be achieved by a six-cylinder engine. For final validation, the optimum conditions indicated by the SCRE can be applied to a six-cylinder test engine. The main experimental work, and the first to be carried out on the SCRE, involved an initial investigation into the application of EGR to an HPDI system. EGR has the ability to significantly reduce NO x emissions from an HPDI engine. The limiting factor for EGR operation is that, as the EGR fraction is increased beyond a certain level, the cycle-to-cycle fluctuations in the combustion process become too severe. This leads to much higher CO and THC emissions, although NO x emissions can reach exceedingly low levels (as low as 60 ppm). The EGR fraction at which this variability becomes a limiting factor depends on the engine operating conditions and the injection timing. Initial testing indicates that EGR fractions between 15% and 25% may be appropriate, but that this is highly dependent on operating conditions and timing. High load and speed conditions, which were not tested, are expected to have significantly lower maximum EGR fractions. One method widely studied for increasing the effectiveness of EGR is to supplement (rather than replace) the fresh intake air with recirculated exhaust gases. At 60 low loads, this has been shown to slightly increase NO x emissions at moderate EGR fractions, with no benefits in performance or other emissions. At higher loads, or at more retarded timings, this method might permit higher maximum EGR rates due to the increased oxygen concentration in the intake mixture. This could result in a significant net reduction in NO x emissions. This possibility needs to be studied further. The effect of the temperature of the recirculated exhaust gas also affects the combustion process. While a slight increase in intake air temperature did not affect the engine's performance, the emissions of NO x were increased. These preliminary results indicate that cooling the EGR is worthwhile, despite the increased operational complexity. Testing at more operating conditions and higher temperatures is required to confirm these results. Maximising the performance of an HPDI engine using EGR will require further study. By reducing emissions through retarding the timing, and then introducing moderate EGR rates, it should be possible to significantly reduce NO x emissions without inducing serious performance or emissions penalties. Further testing on the SCRE is needed to optimise the significant savings in NO x emissions that have been achieved through the combination of EGR and HPDI. 4.2 Recommendations for Future Work The commissioning and initial use of any new research facility will always highlight areas of improvement in the experimental apparatus. The SCRE is no exception, with the primary need being to reduce the experimental uncertainties by improving the precision of the data acquisition system. As well, further detailed testing 61 of the comparisons between the SCRE and the six-cylinder engine need to be carried out, including a detailed look at the effect of injection timing. To this end, carrying out a timing sweep on the SCRE and comparing the results with an identical sweep on a six-cylinder engine would be very useful. More analysis of the identified differences in emissions between single and six-cylinder engines is also needed. The postulated reason for the discrepancy in volumetric efficiency should be confirmed, preferably through the use of an engine simulation program. The development and validation of a workable particulate matter sampling system will greatly enhance the scope of experimental possibilities for the SCRE facility. To maximise the long-term usefulness of the SCRE, it is important to carry out HPDI timing sweeps at every test point of interest. With a full database of timing sweeps, it will be possible to decide on timings for future experiments based on fixed criteria of emissions and performance. This would reduce the SCRE's dependence on the use of Westport injection timings, thereby improving the flexibility of future tests. There is still an extensive amount of EGR testing to be carried out. This includes a more detailed investigation into the effects of intake air temperature and of supplemental EGR at higher load conditions. Another area of interest is to carefully study the effect of EGR on particulate matter emissions. The determination of the relationship between THC, PM, and methane emissions, and how these are affected by EGR operation, is vital for future steps to reduce these harmful emissions. The SCRE provides unique opportunities in many areas of HPDI research. With a fully optimised and well-characterised engine, tests can be carried out on injector and combustion chamber design. These tests benefit the most from the need to only modify a 62 single cylinder. The flexibility of the air exchange and auxiliary systems also provides a wide scope for future research into EGR and related applications. With some more work, the full potential of this unique and valuable research facility can be realised in reducing emissions and improving the performance of HPDI engines. 63 References [I] Douville, B., P. Ouellette, A. Touchette, and B. Ursu. Performance and Emissions of a two-stroke engine fueled using high-pressure direct injection of natural gas. SAE paper 981400. [2] Hodgins, K.B., P.G. Hill, P. Ouellette, and P. Hung. Directly Injected Natural Gas Fueling of Diesel Engines. SAE paper 941692 [3] Ouellette, P., B. Douville, P.G. Hill, and B. Ursu. NOx Reduction in a directly injected natural gas engine. Proceedings of the 1998 Fall Technical Conference of the ASME, ICE Engine Divisions, ICE Vol 31-3, Clymer, NY, Sept 1998. [4] Yu, R.C. and S.M. Shahed. Effects of Injection Timing and Exhaust Gas Recirculation on Emissions from a DI Diesel Engine. SAE Paper 811234 [5] Warnatz, J. U. Maas, and R.W. Dibble. Combustion - Physical and Chemical Fundamentals, modeling and simulation^ xperiments, pollutant formation. 2nd Edition. Springer, Berlin, 1999. [6] Munshi, S. Personal Communications. Nov 2001. [7] Touchette, A. et al. Gaseous and Liquid fuel injector. US Patent no. 6 073 862, June 13,2000. [8] Ouellette, P. et al. Hydraulically actuated gaseous or dual fuel injector. US Patent no. 5 996 558, Dec. 7, 1999. [9] www.dieselnet.com/standards/eu/hd.html [10] http://www.epa.gov/otaq/regs/fuels/diesel/factshet.pdf [II] Fennimore, CP. Thirteenth Symposium (International) on Combustion, The Combustion Institute, Pittsburgh, 1971. [12] Hewson, J.C., and M. Bollig. Reduced Mechanisms for NOx emissions from hydrocarbon diffusion flames. 26th Intl. Symposium on Combustion. The Combustion Institute, 1996. pp 2171-2179. [13] Heywood, J.B. Internal Combustion Engine Fundamentals. McGraw Hill, New York, 1988 [14] Ingersoll, J.G. Natural Gas Vehicles. Fairmont Press, Upper Saddle River, NJ. 1996. p 44 [15] Zelenka, P. et al. Cooled EGR - A Key Technology for Future Efficient HD Diesels. SAE Paper 980190. 1998 [16] Li, Jianwen et. al Effect of Intake Composition on Combustion and Emission Characteristics of DI Diesel Engine at High Intake Pressure. SAE Paper 970322 [17] Acroumanis, C. et. al Effect of EGR on Combustion Development in a 1.9LDI Diesel Optical Engine. SAE Paper 950850. [18] Ladommatos, N., S.M. Abdelhalim, H. Zhao, and Z. Hu. The effects of carbon dioxide in exhaust gas recirculation on diesel engine emissions. Proceedings of the Institute of Mechanical Engineering, Vol 212, Part D., 1998. p 27. [19] Plee, S.L., Ahmad, T. and Myers, J.P. Flame temperature correlation for the effects of exhaust gas recirculation on diesel particulate and NOx emissions. SAE Paper 811195 64 [20] Plee, S.L. et al. Diesel NOx emmissions - a simple correlation technique for intake air effects. 19th Intl. Symposium on Combustion, The Combustion Institute, 1982. pp 1495-1502. [21] Ladommatos, N., S.M. Abdelhalim and H. Zhao. Effects of exhaust gas recirculation temperature on diesel engine combustion and emissions. Proceedings of the Institution of Mechanical Engineering, Vol 212 Part D. 1998 p. 479 [22] Durnholtz, M., G. Eifler, and H Endres. Exhaust-gas Recirculation -A measure to reduce exhaust emissions of DI diesel engines. SAE Paper 920725, 1992 [23] Ladommatos, N., S. Abdelhalim and H. Zhao. The effects of exhaust gas recirculation on diesel combustion and emissions. International Journal of Engine Research, Vol 1, No. 1. 2000. pp. 107-126. [24] Uchida, N., Y. Daisho, T. Saito, and H. Sugano. Combined effects of EGR and Supercharging on Diesel Combustion and Emissions. SAE Paper 930601, 1993. p. 104-108. [25] Lapuerta, M, J.M. Salavert and C. Domenech. Modeling and Experimental Study about the Effect of Exhaust Gas Recirculation on Diesel Engine Combustion and Emissions.. SAE Paper 950216, 1995 [26] Kreso, A.M et al. A Study of the Effects of Exhaust Gas Recirculaton on Heavy-Duty Diesel Engine Emissions. SAE Paper 981422. 1998. [27] Ladommatos, N., R. Balian, R. Horrocks, and L. Cooper. The Effect of Exhaust Gas Recirculation on Soot Formation in a High-Speed Direct-Injection Diesel Engine. SAE Paper 960841, 1996. [28] SAE J177, Rev.l Jun95, section 8. [29] Ropke, S. G.W. Schweimer, T.S. Strauss. NOx formation in diesel engines for various fuels and intake gases. SAE Paper 950213, 1995. 65 Appendix 1 - NO Formation Mechanisms Currently, there are thought to be four significant mechanisms for NO formation in an internal combustion engine [i]. These are: 1) Thermal (extended Zeldovich) NO 2) Prompt (Fenimore) NO 3) Nitrous Oxide NO 4) Fuel NO The relative importance of these different sources is highly dependant on local temperature, concentration gradients, and many other factors, all of which change during the combustion cycle. A.1.1 Thermal NO The thermal source is the largest source of NO in conventional IC engines. It was also the first mechanism to be widely accepted as being a good indicator of the production of NO. The extended mechanism, the basis of which was first postulated by Y.B. Zeldovich in 1946, is given below [i]: 1 : O + N 2 <^> NO + N 2 : N + 02 <=> NO + O 3 : N+OH <=> NO + H The rate coefficients (in cm /mol*s) for the three reactions are: kj = 1.8*1014 exp(-318/(R*T)) k2 = 9.0* 109 exp(-27/(R*T)) k3 = 2.8*1013 The rate limiting step in this reaction is the production of nitrogen radicals by the first reaction, especially at low and medium temperatures (at 2000 K, the rate coefficient of reaction (1) is five orders of magnitude lower than that of reaction (2)). As the temperature increases, the rates of the first and second reactions increase. Similarly, the rates of dissociation of both molecular oxygen and nitrogen to their corresponding radicals also increase with temperature. This means that there are more radicals available to react. 66 Not only will each reaction that occurs proceed faster, but there are also more reactants available, so more reactions will occur. This is why this mechanism is referred to as 'thermal' - because it is highly dependant on the local temperature. The high temperature is needed to break the initial triple bonds in the N2 molecule and thereby generate sufficient radicals for the production of NO. Concentrations of most of the species involved in these reactions are controlled by other reactions, which are so much faster than the thermal mechanism that the species remain in their partial equilibrium conditions. For example, the equilibrium state of the O2 <-> 20 dissociation reaction (which is a function of temperature) controls the concentrations of O and O2 in the flame. Only the quantities of N and NO are significantly changed by the Zeldovich reactions, as the equilibrium reaction of the dissociation of nitrogen molecules is negligible under common combustion conditions (at 2500 K , Kgq = 8.51E-14 for N 2 <-> 2N compared with Keq=2.07E-4 for 02<-^20). The bulk of the nitrogen free radicals available are produced by the first step of the Zeldovich mechanism. This also explains the initially very slow rate of these reactions. It takes time for the first reaction to occur (due to its high activation energy), and as it occurs, the N is used up by (2) and (3) almost as quickly as it is generated. This, however, generates further O radicals, increasing the rate of generation of N and leading to a gradual increase in production of NO. If left undisturbed at a sufficiently high temperature, these steps eventually come to equilibrium. The time taken to reach equilibrium is a function of the fuel/air ratio and the temperature of the system. A simplified rate expression for the thermal mechanism has been suggested by Heywood [ii] (p 575). The derivation of this expression uses an equilibrium approximation for the relationship between the oxygen atom and molecule concentrations, 67 and a steady-state approximation for the nitrogen radical concentration. The strong dependence of NO formation on temperature for this mechanism is evident: d[NO] 6xl0 1 6 ( " 6 9 0 9 0 W w . T J dt T In this equation, [J represents concentration, T represents temperature, and the subscript e denotes equilibrium conditions. A J.2 Prompt NO Due to the slow initial rate of the thermal mechanism, virtually no NO will be generated at the flame. Experiments showed, however, that a discernable amount of NO was being generated immediately within the flame. This was referred to as prompt, or 'Fenimore', NO. The reaction is much more rapid than that for the formation of thermal NO. It involves the CH radical (prevalent in the reaction of hydrocarbon fuels) and an N2 molecule, as shown below [i]: CH + N 2 -> HCN + N -> ... -> NO or N 2 The intermediate reaction which uses the nitrogen radical can lead to either NO (through, for example, recombination with an 0 2 atom as from reaction (1) in the thermal mechanism) or back to an N 2 molecule. The first step in the prompt mechanism is the limiting step. The activation energy of the first reaction is much lower than that of the thermal mechanism (75 kJ/mol vs. 318 kJ/mol). Therefore, the prompt mechanism is not nearly as temperature dependant as the thermal mechanism. The rate-limiting step in the prompt mechanism requires CH radicals. These radicals are highly reactive and only exist within the flame. They are produced by oxidation of CH4 to CH. The richer the flame, the 68 more CH radicals will be present, and hence the more NO will be produced through the prompt mechanism. A.l .3 Nitrous Oxide NO Another mechanism by which NO is generated within a flame is the nitrous oxide (N2O) mechanism. Experimentally, distinguishing between the prompt and the N 2 O mechanisms is virtually impossible. Using numerical simulations, the relative importance of these two rapid mechanisms becomes evident. The first part of the mechanism for the generation of NO from nitrous oxide is similar to the first reaction in the Zeldovich mechanism, but with a third body (M) involved: N 2 + O + M -> N 2 0 + M The intermediate produced in this step is nitrous oxide. The N 2 0 is then oxidized by another free oxygen radical to form NO: N 2 0 + 0 ^ 2NO Higher pressures favor this reaction, due to the involvement of the third body in the initial reaction. The higher the pressure, the more frequent the three-body collisions will be, and the more N 2 O (and eventually, NO) will be generated. At high temperatures, the first reaction of the Zeldovich mechanism (N2 + O —> NO + N) is much more likely to occur. Low temperatures, which limit the thermal mechanism (due to the very high activation energy of the first step), do not affect the formation of N 2 0 as adversely. The third body reaction has a relatively low activation energy, so the primary restriction (from temperature) on this formation mechanism is the quantity of oxygen radicals available. This mechanism is only relevant at temperatures between 1000 and 2000K, where the 69 thermal mechanism is not significant, but where there the equilibrium concentration of O is not negligible. A.lAFuelNO The composition of most fuels varies on a basis of where it is from and how it is refined. This is especially true for Natural Gas, where the composition varies quite significantly between different providers. It is quite common for natural gas to contain some nitrogen (one accepted chemical formula for natural gas is C„Hj.8„No..,n, which gives the nitrogen in the fuel a mass percentage of 8.1%)m. However, as this nitrogen is free and is not 'bound' to the fuel (as is the case for a solid fuel, like coal, or a liquid fuel, like diesel) it will behave similarly to the molecular nitrogen in the air. A.1.5 NOx Decomposition Not all the NO produced during the combustion process will be emitted, as some will decompose later in the combustion cycle. The most relevant route for decomposition is still under debate, with some (including Nishioka et al. [iv]) claiming that the HCN recycle route: HCCO + NO -> HCN + C 0 2 , is the most significant. As this reaction is independent of temperature (activation energy of 0) the concentrations of HCCO and NO control its rate. The concentration of HCCO is highest within the flame, and in a non-premixed flame, some of the already formed NO will be passing through the combustion process later in the cycle. This results in a small but significant quantity of the NO being 'recycled', and is especially important in a non-premixed flame. 70 A differing opinion regarding NO decomposition is postulated by Easley and Mellor [v]. They found that the main routes for NO decomposition are the reverse of the Zeldovich and Fenimore mechanisms. This means that those conditions which aggravate NO x formation, such as high temperatures for the thermal mechanism, also increase the NO decomposition rate. Similar results are reported by Ahmad and Plee [vi]. By whichever mechanism, however, significant quantities of the NO formed in the combustion process will be consumed later in the cycle, reducing the net NO emissions. A.l. 6 References for this Appendix: i Warnatz, J., U. Mass, R.W. Dibble. Combustion — Physical and Chemical Fundamentals, Modeling and Simulation, Experiments, Pollutant Formation. 2nd Ed. Springer, Berlin, 1999. ii Heywood, J.B. Internal Combustion Engine Fundamentals. McGraw-Hill, New York, 1988. iii ibid, p 915. iv Nishioka, M., Y. Kondoh, T. Takeno. Behavior of Key Reactions on NO Formation in Methane-Air Flames. 26th International Symposium on Combustion. The combustion institute, 1996. pp. 2139-2145. v Easley, W.L., Mellor, A.M. NO Decomposition in Diesel Engines. SAE Paper 1999-01-3546. 1999. vi Ahmad, T. and Plee, S.L. Application of flame temperature correlations to emissions from a direct-injection Diesel Engine. SAE Paper 831734. 1983. 71 Appendix 2 - Relevant Equations Volumetric Efficiency: The volumetric efficiency used in this thesis is based on the intake manifold pressure and temperature. The equation used is (Heywood, p. 54): N • P manifold ' ^cyl Where nR (the number of revolutions of the crank shaft per cycle) is 2 for a 4-stroke engine, N is the engine speed and Vcyi is the cylinder displacement. Indicated Power: The indicated power is calculated from the in-cylinder pressure trace. The basic equation used is (Heywood, p. 48): NJp-dV p 1ind nR Where p is the instantaneous in-cylinder pressure. For this calculation, a numerical integration is performed using the mean pressure between two successive encoder pulses {Vi° CA). Indicated Mean Effective Pressure: The IMEP is calculated from the indicated power by (Heywood, p. 50): IMEP = ind "-R Vcyl-N 72 Thermal Efficiency: The thermal efficiency is the ratio between the power developed and the heat energy available from combustion (Heywood, p. 52): p. V, = mCNG -LHVCNG +mdiesel ' LHVdiesel Where Pirui is the indicated power and LHV is the lower heating value for the corresponding fuel. Diesel Equivalent CNG flow: When studying the total fuel flow rate, it is useful to convert the natural gas flow rate to the mass flow rate of diesel which would carry an equivalent amount of chemical potential energy: ^CNG (diesel _ equivalent) ^CNG ^ LHV CNG yLHVdiesel Where p is the density of the indicated fuel. With this, the total flow rate is simply the sum of the diesel equivalent natural gas flow rate and the diesel pilot flow rate. Indicated Specific Fuel Consumption: The ISFC is calculated from the observed fuel flow rates (natural gas corrected to the equivalent diesel flow rate) and the indicated power using (Heywood, p. 52): ISFC — °<-diesel -equivalent) diesel P rind 73 EGR Rate: The EGR rate was calculated using the intake and exhaust CO2 concentrations, using the following equation: % K H ! _ _ 3 = x l O O = ^ - f t 0 3 > x ' W - ^ x l O O m, fresnair+Kr (HRco2-0.03) MW In this equation, LRQOI is the CO2 concentration (volume %) in the intake manifold, HRcoi is the C 0 2 concentration in the exhaust stream, and the MW's refer to the molecular weight of the exhaust gas and the intake manifold air mixture (including the exhaust gas). The molecular weight of the mixture in the intake manifold is calculated from: MW - MW x l r l v r manifoldair 1 Y 1 v v exhaust ^ '' LRC02 -0.03^ ffi?CO2-0.03 + MWfreshair >< fx LRC02 -0.03^ HRC02 - 0.03 y y LUi J Where the terms are the same as in the preceding equation. The exhaust molecular weight is determined from the mass fractions of each of the major species in the exhaust stream, as measured by the emissions analysers. The H 2 0 concentration is estimated based on the intake air humidity and the hydrogen content of the two fuels. 74 Appendix 3 - Idle Timing Sweep Results The effects of injection timing are critical to the operation of the SCRE. Timing sweeps need to be carried out at every test condition to determine the trade-offs to be made between NO x, THC, and CO emissions, as well as performance. A timing sweep was carried out at idle (600 RPM, 2 Bar IMEP), where the pilot start of injection varied from 1 to 4 ms before top centre. Relative injection times (RIT, the delay between the end of diesel injection and the start of natural gas injection) between 1.3 ms and 3.8 ms were tested. The diesel pulse width was constant at 0.6 ms, and the natural gas pulse width was held constant at 0.37 ms. The primary influence on performance and emissions appears to be the absolute timing of the natural gas injection. Relative injection timings that were too low (below 1.3 ms) led to large fluctuations in the combustion process, and were not tested. The effect of the natural gas injection timing is shown in the following figures: 120 -i 1 90 Z 30 -0 "I 1 , I 1 1 -10 -5 0 5 10 15 Absolute start of natural gas injection (°ATC) Figure A.l Effect of absolute timing of natural gas injection on NOx emissions at idle operating conditions. 75 Figure A.2 Effect of absolute timing of natural gas injection on THC emissions at idle operating conditions. Figure A3 Effect of absolute timing of natural gas injection on indicated power at idle operating conditions. As these figures show, there is no optimal timing that will minimise NO x, THC, and performance simultaneously. It is also apparent that the more retarded the timing, the greater the NO x reduction. But this comes at the cost of reduced performance and increased THC emissions. The 'ideal' operating point will be a trade-off between these different effects. 76 Appendix 4 - Motoring Torques One of the main problems with the SCRE is the very high internal friction. The 40-hp motor provided to overcome this is not sufficient at all speed and load conditions, resulting in some (high speed, low load) being unreachable. The engine was motored on a number of occasions, the results of which are tabulated below and shown in Figure A.4. Figure A.4 Motoring Torque as a function of Engine Speed. Crosses indicate oil temperature of 78°C, solid diamonds an oil temperature of 100"C. Speed (RPM) Load (Nm) E O T (oC) 600 90 101 800 135 100 1000 170 100 1200 210 98 200 38 78 300 49 77 430 75 75 540 101 76 605 115 76 755 170 81 890 190 80 1045 211 79 1120 230 78 Table A.l - Motoring Torque Data. As the figure shows, the motoring torque increases linearly with speed. It is also interesting to see the effect of oil temperature, which reduces the frictional torque by 25 Nm. This indicates how important it is for the engine to be operating at full temperature. 77 Appendix 5 - Diesel Consumption Data The diesel mass flow rate sensor (a pail-and-scale system) does not give an accurate or reliable reading for the mass flow rate of pilot diesel. The best way to determine this rate was to run the engine under fixed conditions over an hour (approximately), and record the total change in mass over that time. While this is not an ideal situation, it provides the best available estimate of diesel flow rate. The results for the test points studied is given in Table A.2: Speed Load diesel flow (RPM) (% of max) (9/hr) 600 idle 110 800 25 101 800 60 122 800 100 100 1200 25 126 1200 50 133 1600 50 164 1067 ESC5 104 1067 ESC7 107 Table A.2 Approximate diesel consumption at each test point These values are valid only for the timings used - significant changes in timing will affect the fuel flow rate. 78 Appendix 6 - Instrumentation List for the SCRE The list on the following page is the up-to-date instrumentation list for those instruments installed and operational on the SCRE for the duration of the commissioning and EGR testing. ID | PHYSICAL QUANTITY LOCATION UNITS RANGE MANUFACTURER T E M P E R A T U R E 1A EGR temperature after EGR cooler Controller/DAQ Deqrees C 0-200 Omeqa (K-type) 1B Post intercooler temperature (air) Controller/DAQ Deqrees C 0-200 Omeqa (K-type) 1C exhaust surqe tank (downstream) temp DAQ/CP Deqrees C 0-600 Omeqa (K-type) 1D Dyno case temperature (alarm) CPx2 Deqrees C 0 - 2 0 0 Omeqa (K-type) 1F enqine coolant temperature out alarm CPx2 Deqrees C 0-200 Omeqa (K-type) G EGR coolant outlet temperature C P Deqrees C 0-70 Omeqa (K-type) 1H Intercooler coolant outlet temperature C P Deqrees C 0-70 Omeqa (K-tvoe) 11 dynamometer coolant outlet temperature C P Deqrees C 0 - 2 0 0 Omeqa (K-type) 1J engine oil temperature in enqine C P Deqrees C 0 - 2 0 0 Omeqa (K-type) 1K coolinq tower temperature C P Deqrees C 0 - 2 0 0 Omeqa (K-type) 1L cell ambient temperature C P Deqrees C 0 t o 5 5 Omeqa (K-type) 1M supercharqer intake temperature DAQ Deqrees C 0-200 Omeqa (K-type) 1N supercharqer exhaust temperature DAQ Deqrees C 0-300 Omeqa (K-type) 10 pre-intercooler temperature of air DAQ Deqrees C 0-300 Omeqa (K-type) 1P exhaust manifold temperature DAQ Deqrees C 0-600 Omeqa (K-type) 1Q intake manifold temperature DAQ Deqrees C 0-150 Omeqa (K-type) 1S Supercharqer Body Temperature C P Deqrees C 0-200 Omeqa (K-type) 1T SO Oil Inlet Temperature C P Deqrees C 0-100 Omeqa (K-type) 1U SC Oil Outlet Temperature C P Deqrees C 0-600 Omeqa (K-type) P R E S S U R E 2A Exhaust system pressure DAQ psi 0-50 Enerqv Kinetics 2B enqine oil pressure (alarm) C P psi 0-200 Enerqy Kinetics 2C Pre-intercooler air pressure DAQ/CP psi (q) 0-50 Enerqv Kinetics 2D Post-intercooler air pressure DAQ psi (q) 0-50 Enerqy Kinetics 2G inlet air pressure after flow sensor DAQ "H20 (abs) 0-10 from 6V-92 M I S C E L L A N Y 3A Enqine Speed Diqiloq/DAQ rpm Maqnetic Pick-up 3B Enqine Torque Diqiloq/DAQ Nm 0-300 Artech Load Cell 3C Crankshaft position DAQ 1/2 deqrees C A . 0-360 3E Exhaust Backpressure valve position C P % open 0-100 from electric valve 5B EGR valve position C P % 0-100 Bernard Controls 3F Vector Drive Torque CP /DAQ Nm 0-230 From vector drive 8G Cell CH4 Sniffer C P %LEL 0-100 Gastech Can. Prof2000 GUEl .STATUS ilB^^MlllilHtllllMill 4A diesel fuel temperature DAQ Deqrees C 0 - 2 0 0 Thermocouple K 4B CNG fuel temperature DAQ Deqrees C 0 - 2 0 0 Thermocouple K 6A dieselfuel pressure DAQ psi 0 - 5000 Enerqy Kinetics 6B CNG fuel pressure DAQ psi 0 - 5000 Enerqy Kinetics 6C diesel fuel level in tank (hiqh) C P on/off float switch 6D diesel fuel level in tank (low) C P on/off float switch C Y L I N D E R P R 7A cylinder pressure DAQ bar 0 - 2 0 0 AVL QC33C F L O W 8A diesel fuel mass DAQ kq 0-30 Rice Lake Scale 8B CNG fuel flow DAQ kq/hr 0-12 Micromotion / Brooks 8C airflow DAQ/CP cfm 0-300 6" Superflow 8D Main coolant intake (alarm) C P flow on/off McDonell Flow switch 8E Main coolant discharge (alarm) C P flow on/off McDonell Flow switch E M I S S I O N S B E N C H 10A/H Low-ranqe C 0 2 DAQ % 0-2,0-10 California Analytical 10B Hiqh-ranqe C 0 2 DAQ % 0 t o 2 0 Beckmann 10C 0 2 DAQ % 0to21 Siemens 10D CO DAQ PPM 0-3000 Siemens 10E NOx DAQ PPM 0-3000 API 10G THC DAQ PPM 0-10000 Ratfisch Table A3 SCRE Instrumentation, effective Nov J 2001. 79 Appendix 7 - EGR Testing Procedures The following list outlines the procedures used to achieve stable, repeatable operation of the SCRE with EGR over a range of operating conditions and EGR fractions. The following list refers to replacement EGR operation at fixed intake manifold temperature. Testing of other parameters will require some changes to these operating instructions. 1) Let the engine reach full temperature by running at moderate speed and load (at least 95°C). 2) Set up the low-range C O 2 sensor for EGR operation. Turn the low-range selector valve (on the North outside wall of the test cell) to 'EGR'. Turn both handles on the C 0 2 selector valves (inside the emissions bench, Cabinet 1, LHS) to 'low range'. Ensure that sample flow rate is 1 1pm (flow meter on RHS at back of cabinet) and that the drier flow rate (flow meter attached to back door of cabinet) is 2 1pm. 3) Determine the desired air and fuel flow rates and injection timing for the test point. 4) Adjust the supercharger speed, engine speed, engine load, and fuelling so that the engine is running under the desired conditions. 5) Adjust the back-pressure valve (using the control knob on the CP) until the BP reads 5-10 kPa above the intake manifold pressure. 6) Let the engine run for 5-10 minutes or until all the temperatures stop changing. Carefully monitor the intake air temperature. 7) Open the manual EGR valve (black handle, above the supercharger) so that EGR can flow to either the supercharger exhaust (handle to the left) or to the supercharger intake (handle to the right). 8) Open the remote EGR valve slowly. Watch the low-range C 0 2 analyser output to detect when EGR flow starts. Adjust the EGR valve, supercharger speed and back-pressure valve to hold the manifold air pressure constant while adjusting the EGR fraction (displayed on the DAQ screen). Adjustments need to be made slowly, as the intake C 0 2 sensor has a time delay of approximately 30 seconds. 9) During all adjustments, monitor the pressure trace and THC emissions - a rapid increase in THC or a significant retardation of the peak cylinder pressure indicates the 80 onset of high combustion variability, which can lead to misfires and dangerous quantities of CH4 in the exhaust line. 10) Fine-tuning of the EGR flow rate is best achieved through stepwise adjustments of the supercharger (50 RPM) and the BP valve. Unfortunately, the BP valve does not always turn even when the %open display changes, so care has to be taken to ensure that an excessive increase (or decrease) in back-pressure is not caused. 11) It often requires more than one attempt to reach the desired EGR fraction. Care and patience are required to reach the desired operating condition. 12) Once at the operating condition, let the engine run for about 5 minutes to allow all the readings to stabilise. Once low and high speed readings are taken, repeat steps 9-11 to get to the next desired operating condition. 13) To shut off the EGR system, reduce the opening of the EGR valve to 15-20%. Simultaneously (in stages) increase the supercharger speed and open the back-pressure valve so that the intake and exhaust manifold pressures are held approximately constant. Close the EGR valve the rest of the way, increase supercharger speed to reach the desired intake pressure, and adjust the back-pressure valve for the desired pressure. 14) Follow the standard procedures for adjusting the engine speed and load to reach the next desired operating condition. It is strongly recommended that the engine be run at a fixed (repeatability) point for at least 10 minutes between EGR test runs to purge all exhaust gases from the intake system and to ensure that the engine is operating in the expected manner. 81 Appendix 8 - Baseline and Comparison Test Summary The following table contains a summary of all the testing carried out for this thesis. Detailed data is available upon request from the Alternate Fuels Group at UBC. HPDI 28 Point Baseline ID Time Speed (RPM) Load (approx %, PSOl (ms) PPW (ms) RIT (ms) GPW (ms) LS File .CSV HS File .CSV Test points 2 / /09 /01 BL1 12:40 1200 50 -3 0.6 2.5 0.54 Slow-BL1 Fast-BL1 BL2 12:54 1200 25 -3 0.6 2.3 0.36 SIOW-BL2 Fast -BL2 B L 3 13:14 1200 75 -3 0.6 2.8 0.93 S low-BL3 Fast -BL3 BL4 13:40 1200 100 -3 0.6 2.8 1.53' S low-BL4 Fast -BL4 BL5 14:08 800 100 -3 0.6 2.5 1.65 S low-BL5 Fast -BL5 BL6 14:40 800 25 -3 0.6 3 0.345 SIOW-BL6 Fast -BL6 BL7 15:30 800 50 -3 0.6 2.1 0.47 SIOW-BL7 Fast -BL7 BL8 15:45 800 75 -3 0.6 1.45 0.92 SIOW-BL8 Fast -BL8 BL9 16:05 1400 50 -3 0.6 2.2 0.49 S low-BL9 Fast -BL9 BL10 16:20 1400 75 -3 0.6 2.1 0.94 S low-BLIO Fast -BL10 BL11 16:35 1400 - 3 5 -3 0.6 1.9 0.35 Slow-BL11 Fast-BL11 BL12 16:45 1400 100 -3 0.6 2.6 1.8- S low-BL12 Fast -BL12 Test points 01/10/01 •.. • . < > . . . . -BL13 10:20 1600 50 -3 0.6 2.6 0.45 Fast -BL13 BL14 10:30 1600 75 -3 0.6 2.6 1.35 S low-BL14 Fast -BL14 BL15 10:42 1600 100 -3 0.6 2.2 1.75 S low-BLI 5 Fast -BL15 BL16 10:55 1600 - 3 8 -3 0.6 1.9 0.41 S low-BL16 Fast -BL16 BL17 11:25 1000 75 -3 0.6 1.9 0.86 S low-BL17 Fast -BL17 BL18 11:46 1000 100 -3 0.6 2.4 1.92 SIOW-BL18 Fast -BL18 BL19 11:56 1000 50 -3 0.6 2.9 0.42 S low-BL19 Fast -BL19 BL20 12:15 1000 25 -3 0.6 2.6 0.365 SIOW-BL20 Fast -BL20 BL21 12:55 1750 - 5 5 -3 0.6 2.1 0.43 SIOW-BL21 Fast-BL21 BL22 13:00 1750 75 -3 0.6 1.9 0.81 SIOW-BL22 Fast -BL22 BL23 13:10 1750 100 -3 0.6 2.2 1.81 SIOW-BL23 Fast -BL23 BL24 13:51 600 75 -3 0.6 2.3 0.425 SIOW-BL24 Fast -BL24 BL25 14:00 600 100 -3 0.6 2.3 0.7 SIOW-BL25 Fast -BL25 BL26 14:18 600 50 -3 0.6 2.3 0.399 SIOW-BL26 Fast -BL26 BL27 14:21 600 25 -3 0.6 2.3 0.378 SIOW-BL27 Fast -BL27 Tost points 02/10/01 .• BL13b I 10:15| 1600| 50 | - 3 | 0.6| 2.4| 0 .45 |S low-BL13b |Fast -BL13b Test points 03/10/01 BL14b 11:12 1600 75 -3 0.6 2.7 1.34 S low-BL14b Fast -BL14b BL20b 11:26 1000 25 -3 0.6 2.6 0.38 S low-BL20b Fast -BL20b BL15b 11:35 1600 100 - 3 0.6 2.2 1.93 S low-BL15b Fast -BL15b Idle Timing Sweeps ID T ime Speed (RPM) Load (%) P S O l ms P P W ms RIT ms G P W ms LS File .CSV HS File .CSV Testipoirits124/1i0/0i1» Idle - 3 1.8 14:40 600 0 - 3 0.6 1.8 0.365 Slow-01-10-24-14.40.42 Fast-01 -10-24-14.41.52 Idle -2 1.8 14:45 600 0 -2 0.6 1.8 0.365 Slow-01-10-24-14.45.49 Fast-01-10-24-14.45.39 Idle -1 1.8 14:50 600 0 -1 0.6 1.8 0.365 Slow-01-10-24-14.49.53 Fast-01-10-24-14.50.33 Idle -1 1.0 14:55 600 0 -1 0.6 1 0.365 Slow-01-10-24-14.55.02 Fast-01-10-24-14.54.48 Idle -1 2.6 15:00 600 0 -1 0.6 2.6 0.365 Slow-01 -10-24-14.59.26 Fast-01-10-24-14.59.33 Idle -1 3.2 15:03 600 0 -1 0.6 3.2 0.365 Slow-01-10-24-15.03.15 Fast-01 -10-24-15.03.44 Idle -2 1.0 15:06 600 0 -2 0.6 1 0.365 Slow-01-10-24-15.06.44 Fast-01-10-24-15.06.35 Idle -2 2.6 15:10 600 0 -2 0.6 2.6 0.365 Slow-01-10-24-15.10.22 Fast-01-10-24-15.10.51 Idle-2 3.2 15:15 600 0 -2 0.6 3.2 0.365 Slow-01-10-24-15.15.30 Fast-01-10-24-15.16.23 Idle-2 3.8 15:20 600 0 -2 0.6 3.8 0.365 Slow-01-10-24-15.20.53 Fast-01-10-24-15.26.42 Idle - 3 1.3 15:28 600 0 - 3 0.6 1.3 0.365 Slow-01-10-24-15.28.09 Fast-01-10-24-15.28.17 Idle - 3 1.8t 15:30 600 0 - 3 0.6 1.8 0.365 Slow-01-10-24-15.30.40 Fast-01-10-24-15.32.35 Idle-3 2.6 15:35 600 0 - 3 0.6 2.6 0.365 Slow-01-10-24-15.35.08 Fast-01-10-24-15.34.51 Idle-3 3.2 15:40 600 0 - 3 0.6 3.2 0.365 Slow-01-10-24-15.39.18 Fast-01-10-24-15.40.03 Idle-4 3.2 15:42 600 0 -4 0.6 3.2 0.365 Slow-01-10-24-15.42.41 Fast-01-10-24-15.42.41 Idle-4 2.6 15:45 600 0 -4 0.6 2.6 0.365 Slow-01-10-24-15.45.57 Fast-01-10-24-15.45.51 Idle-4 1.8 15:47 600 0 -4 0.6 1.8 0.365 Idle-4 1.2 15:50 600 0 -4 0.6 1.2 0.365 Slow-01 -10-24-15.49.56 Fast-01-10-24-15.49.48 Idle - 3 1.8c 15:57 600 0 - 3 0.6 1.8 0.365 Slow-01 -10-24-15.56.39 Fast-01-10-24-15.57.23 82 B a c k - p r e s s u r e s e n s i t i v i t y t e s t ID Time Speed (RPM) Load (approx %, PSOl (ms) PPW (ms) RIT (ms) GPW (ms) LS File .CSV HS File .CSV Test DOints 9/11/01 > - • is EBPO 9:40 800 10 -2 0.6 1.8 0.475 Slow-01-11-09-09.42.45 Fast-01-11-09-09.41.51 EBP10 9:55 800 10 -2 0.6 1.8 0.475 Slow-01-11-09-09.57.22 Fast-01-11-09-09.54.16 EBP20 10:02 800 10 -2 0.6 1.8 0.475 Slow-01-11-09-10.03.16 Fast-01-11-09-09.59.11 EBP30 10:08 800 10 -2 0.6 1.8 0.475 Slow-01-11-09-10.07.23 Fast-01-11-09-10.04.58 EBP37 10:15 800 10 -2 0.6 1.8 0.54 Slow-01-11-09-10.14.45 Fast-01-11-09-10.13.38 EBP46 10:18 800 10 -2 0.6 1.8 0.54 Slow-01-11-09-10.18.30 Fast-01-11-09-10.17.25 EBP53 10:22 800 10 -2 0.6 1.8 0.54 Slow-01-11-09-10.21.48 Fast-01-11-09-10.20.54 1 2 0 0 R P M 5 0 % L o a d R e p l i c a t i o n s ID Time Speed (RPM) Load approx. % PSOl ms PPW ms RIT ms GPW ms LS File HS File .CSV .CSV Test DOints 27/09/01 ' * RepH 13:54 1200 50 -3 0.6 2.5 0.54 Slow-01-09-27-13.56.17 Fast-01-09-27-13.55.04 Repl2 15:55 1200 50 -3 0.6 2.5 0.54 Slow-01-09-27-15.56.48 Fast-01-09-27-15.55.30 Repl3 17:00 1200 50 -3 0.6 2.5 0.54 Slow-01-09-27-17.00.49 Fast-01-09-27-16.59.29 Test points 01/10/01 ' V"'' " Repl4 10:10 1200 50 -3 0.6 2.5 0.54 Slow-01-10-01-10.11.14 Fast-01-10-01-10.08.58 Repl5 11:05 1200 50 -3 0.6 2.5 0.54 Slow-01-10-01-11.04.57 Fast-01-10-01-11.05.28 Repl6 12:25 1200 50 -3 0.6 2.5 0.54 Slow-01-10-01-12.30.28 Fast-01-10-01-12.29.32 Repl6b 12:35 1200 50 -3 0.6 2.5 0.46 Slow-01-10-01-12.37.41 Fast-01-10-01-12.36.14 Repl7 13:30 1200 50 -3 0.6 2.5 0.54 Slow-01-10-01-13.32.19 Fast-01-10-01-13.32.53 Repl7b 13:35 1200 50 -3 0.6 2.5 0.46 Slow-01-10-01-13.36.16 Repl8 14:30 1200 50 -3 0.6 2.5 0.54 Slow-01-10-01-14.33.28 Fast-01-10-01-14.32.17 Repl8b 14:40 1200 50 -3 0.6 2.5 0.46 Slow-01-10-01-14.40.43 Fast-01-10-01-14.35.05 Repl8c 14:42 1200 50 -3 0.6 2.5 0.53 ? ? Test DOints 02/10/01 Repl9b 10:05 1200 50 -3 0.6 2.5 0.445 Slow-01-10-02-10.07.18 Fast-01-10-02-10.05.24 RepH0b 11:25 1200 50 -3 0.6 2.5 0.45 Slow-01-10-02-11.24.34 Fast-01-10-02-11.23.51 Test points 03/10/01 --• RepH 1 11:00 1200 50 -3 0.6 2.5 0.45 Slow-01-10-03-10.59.35 Fast-01-10-03-10.59.29 RepH 2 11:40 1200 50 -3 0.6 2.5 0.465 Slow-01-10-03-11.39.46 Fast-01-10-03-11.38.46 Tost points 05/10/01 NOTE: At this point, the cal. Procedure on the NOx analyser was changed to spanning for exactly 10 miniros. RepH 3 14:25 1200 50 -3 0.6 2.5 0.45 Slow-01-10-05-14.16.22 Fast-01-10-05-14.13.42 RepH 4 14:59 1200 50 -3 0.6 2.5 0.45 Slow-01-10-05-14.57.47 Fast-01-10-05-14.58.21 Test1pointsl09/10/01 "-RepH 5 10:45 1200 50 -3 0.6 2.5 0.45 Slow-01-10-09-10.46.24 Fast-01-10-09-10.45.53 RepH 6 11:30 1200 50 -3 0.6 2.5 0.45 Slow-01-10-09-11.29.32 Fast-01-10-09-11.29.10 RepH 7 11:50 1200 50 -3 0.6 2.5 0.45 Slow-01-10-09-11.49.14 Fast-01-10-09-11.49.43 Testlpoints 10/10/01 Repl18 | 10:35| 1200| 50| -3| 0.6| 2.5] 0.455|Slow-01-10-10-10.35.37 |Fast-01-10-10-10.29.50 Testponts 11/10/01 . - - - -RepH 9 11:05 1200 50 -3 0.6 2.5 0.45 Slow-01-10-11-11.05.34 Fast-01-10-11-11.05.01 Repl20 13:30 1200 50 -3 0.6 2.5 0.45 Slow-01-10-11-13.22.57 Fast-01-10-11-13.22.21 Repl21 14:30 1200 50 -3 0.6 2.5 0.45 Slow-01-10-11-14.43.41 Fast-01-10-11-14.25.27 Repl22 16:55 1200 50 -3 0.6 2.5 0.45 Slow-01-10-11-16.56.23 Fast-01-10-11-16.56.04 TestlpointsX1/1.1/01 , r •• r-Repl29 | 12:03| 1200| 50%| -3| 0.6| 2.5| 0.45|Slow-01-11-01-09.32.28 |Fast-01-11-01-12.03.01 Test points 2/11/01 • ". •-.•>>• Repl30 13:45 1200 50% -3 0.6 2.5 0.45 Repl31 14:03 1200 50% -3 0.6 2.5 0.45 Test points 5/11/01 . . . . Repl32 11:50 1200 50% -3 0.6 2.5 0.45 Slow-01-11-05-11.51.47 Fast-01-11-05-11.44.25 Repl33 12:46 1200 50% -3 0.6 2.5 0.45 Slow-01-11-05-12.45.55 Fast-01-11-05-12.44.28 Repl34 15:28 1200 50% -3 0.6 2.5 0.45 Slow-01-11-05-15.27.04 Fast-01-11-05-15.26.17 Repl35 16:05 1200 50% -2.2 0.6 1.8 0.455 Slow-01-11-05-16.05.54 Fast-01-11-05-16.04.25 Test points 6/11/01 Repl36 10:57 1200 50% -3 0.6 2.5 0.465 Slow-01-11-06-10.57.13 Fast-01-11-06-10.55.13 Repl37 11:31 1200 50% -3 0.6 2.5 0.465 Slow-01-11-06-11.31.26 Fast-01-11-06-11.31.56 83 8 0 0 R P M 2 5 % L o a d R e p l i c a t i o n s ID Time Speed (RPM) Target IME (bar) PSOl ms PPW ms RIT ms GPW ms LS File .CSV HS File .CSV mestipoints 1/11/01 Repl2-1 12:15 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-12.15.08 Fast-01-11-01-12.15.39 Repl2-2 13:34 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-13.34.53 Fast-01-11-01-13.33.00 Test points 2/11/01 ; i-- X-r ' ' ; Repl2-3 10:46 800 6 -2 0.6 1.8 0.395 Repl2-3b 10:50 800 6 -2 0.6 1.8 0.395 Repl2-4 13:30 800 6 -2 0.6 1.8 0.395 Test points b/11/01 Repl2-5 9:55 800 6 -2 0.6 1.8 0.395 Slow-01-11-05-09.54.10 Fast-01-11-05-09.57.57 Repl2-6 10:58 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-10.58.14 Fast-01-11-05-10.56.33 Repl2-7 11:30 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-11.30.40 Fast-01-11-05-11.29.01 Repl2-7b 11:32 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-11.33.56 Fast-01-11-05-11.33.16 Repl2-8 12:56 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-12.57.06 Fast-01-11-05-12.56.19 Repl2-9 13:17 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-13.22.35 Fast-01-11-05-13.22.03 Repl2-10 15:25 800 6 -2.2 0.6 1.8 0.39 Slow-01-11-05-15.22.45 Fast-01-11-05-15.22.17 Repl2-11 16:10 800 6 -2.2 0.6 1.8 0.39 Slow-01-11-05-16.09.29 Fast-01-11-05-16.08.48 Test points 6/11/01 Repl2-12 9:20 800 6 -2 0.6 1.8 0.39 Slow-01-11-06-09.22.03 Fast-01-11-06-09.20.23 Repl2-13 10:15 800 6 -2 0.6 1.8 0.39 Slow-01-11-0 6-10.14.17 Fast-01-11-06-10.13.37 Repl2-14 14:40 800 6 -2 0.6 1.8 0.39 Slow-01-11-06-14.45.16 Fast-01-11-06-14.45.49 Test points 9/11/01 . . . . . . . Repl2-15 10:27 800 6 -2 0.6 1.8 0.39 Slow-01-11-09-10.28.40 Fast-01-11-09-10.24.54 Repl2-16 11:04 800 6 -2 0.6 1.8 0.39 Slow-01-11-09-11.04.27 Fast-01-11-09-11.01.01 Repl2-17 13:03 800 6 -2 0.6 1.8 0.39 Slow-01-11-09-13.04.55 Fast-01-11-09-13.02.54 ESC Modes for Emissions Comparison ID Time Speed Load PSOl PPW RIT GPW Slow Fast (RPM) % ms ms ms ms .CSV .CSV 14/11-2001 \ E S C Mode 5 - 1 0 6 7 RPM, 983 Nm BT ~ I ESC5-1 I 14:20) 1067| 50%| -0.542| 0.6| 1.8| 0.7|Slow-01-11-14-14.21.45|Fast-01-11-14-14.18.51 E S C Mode 7 - 106/ RPM, 492 Nm B l : ESC7-1 I 15:161 1067| 25%| -0.5111 0.6| 1.8| 0.4|Slow-01-11-14-15.16.011 Fast-01-11-14-15.16.49 E S C Mode 6 - 1067, RPM, 1475 Nm E ESC6-1 | 15:55| 1067| 75%| -0.844| 0.6| 1.8| 1.3|Slow-01-11-14-15.52.13|Fast-01-11-14-15.51.22 Table A.4 Baseline and comparison test summary. 84 Appendix 9 - EGR Test Summary The following table contains a summary of all the testing carried out for Chapter 3. Detailed data is available upon request from the Alternate Fuels Group at UBC. E G R S e t 1 - R e p l a c e m e n t ID Time Speed (RPM) Target IMEP (bar) PSOl (ms) PPW Cms; RIT (ms) GPW (ms) LS File .CSV HS File .CSV Test points 09/10/01 EGRBL1-01 11:17 1200 8 -3 0.6 2.5 0.397 Slow-01-10-09-11.17.37 Fast-01-10-09-11.16.50 EGRBL1-02 11:45 800 14.5 -3 0.6 2.5 1.02 Slow-01-10-09-11.44.12 Fast-01-10-09-11.42.28 Test points 10/10/01 . " , - ' EGR3-15 I 12:00| 1200| 8| - 3 | 0.6| 2.5| 0.397| I Test points 11/10/01 Test Point 3 - 1 6 0 0 R P M 11 Bar. lMEP (HSLU EGR1-3-0 11:25 1600 11 -3 0.6 2.3 0.45 Slow-01-10-11-11.27.36 Fast-01-10-11-11.25.09 EGR1-3-15 11:44 1600 11 -3 0.6 2.3 0.457 Slow-01-10-11-11.49.06 Fast-01-10-11-11.48.06 EGR1-3-30 12:30 1600 11 -3 0.6 2.3 0.469 Slow-01-10-11-12.32.24 Fast-01-10-11-12.31.32 EGR1-3-36 ? 1600 11 -3 0.6 2.3 0.469 EGR1-3-30-18 12:40 1600 11 -3 0.6 1.8 0.465 Slow-01-10-11-12.40.00 Fast-01-10-11-12.39.05 EGR1-3-45+-18 13:05 1600 11 -3 0.6 1.8 0.465 Slow-01-10-11-13.04.40 Fast-01-10-11-13.02.46 EGR1-3-15-18 13:10 1600 11 -3 0.6 1.8 0.465 Slow-01-10-11-13.11.23 Fast-01-10-11-13.10.14 EGR1-3-0-18 13:15 1600 11 -3 0.6 1.8 0.465 Slow-01-10-11-13.16.46 Fast-01-10-11-13.16.08 Test Point 4 - 800 RRMI6lBa«IMEP • ' EGR1-4-0 15:00 800 6 -3 0.6 2.5 0.385 Slow-01-10-11-14.58.07 Fast-01-10-11-14.54.47 EGR1-4-15 15:25 800 6 -3 0.6 2.5 0.387 Slow-01-10-11-15.25.27 Fast-01-10-11-15.26.24 EGR 1-4-30 16:00 800 6 -3 0.6 2.5 0.387 Slow-01-10-11-15.59.57 Fast-01-10-11-15.59.05 EGR1-4-45 16:30 800 6 -3 0.6 2.5 0.389 Slow-01-10-11-16.27.37 Fast-01-10-11-16.26.31 EGR1-4-60 16:40 800 6 -3 0.6 2.5 0.389 Slow-01-10-11-16.41.12 Fast-01-10-11-16.40.44 EGR1-4-60+ 16:55 800 6 -3 0.6 2.5 0.389 Slow-01-10-11-16.46.31 Fast-01-10-11-16.46.19 29/10/01 Testinq . . ~ Test Point 5 - Idle '•• ! • EGR 1-5-0 14:43 600 2.5 -2 0.6 1.8 0.365 Slow-01-10-29-14.42.39 Fast-01-10-29-14.43.08 EGR1-5-0b 14:47 600 2.5 -2 0.6 1.8 0.365 Slow-01-10-29-14.47.43 Fast-01-10-29-14.50.02 EGR1-5-30 14:59 600 2.5 -2 0.6 1.8 0.365 Slow-01-10-29-14.58.51 Fast-01-10-29-14.59.23 EGR1-5-15 15:04 600 2.5 -2 0.6 1.8 0.365 Slow-01-10-29-15.04.24 Fast-01-10-29-15.03.34 EGR1-5-25 15:15 600 2.5 -2 0.6 1.8 0.365 Slow-01-10-29-15.15.34 Fast-01-10-29-15.15.01 EGR1-5-40 15:25 600 2.5 -2 0.6 1.8 0.365 Slow-01-10-29-15.26.20 Fast-01-10-29-15.25.44 EGR1-5-35 15:32 600 2.5 -2 0.6 1.8 0.365 Slow-01-10-29-15.31.52 Fast-01-10-29-15.30.40 EGR1-5-20 15:43 600 2.5 -2 0.6 1.8 0.365 Slow-01-10-29-15.43.10 Fast-01-10-29-15.42.45 EGR1-5-0C 15:45 600 2.5 -2 0.6 1.8 0.365 Slow-01-10-29-15.45.17 Fast-01-10-29-15.44.40 1/11/01 Testlna ; .-oo--- . :5*i v Test Point 6 - 800 R P M ; 6:B'ar IMEP.(repeated) .— EGR1-6-0 9:50 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-09.50.39 Fast-01-11-01-09.49.31 EGR1-6-20 10:00 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-10.01:47 Fast-01-11-01-10.01.00 EGR1-6-20b 10:15 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-10.14.15 Fast-01-11-01-10.15.05 EGR1-6-30 10:23 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-10.23.25 Fast-01-11-01-10.23.19 EGR1-6-10 10:41 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-10.41.15 Fast-01-11-01-10.38.14 EGR1-6-35 10:45 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-10.49.20 Fast-01-11-01-10.44.34 EGR1-6-25 11:00 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-11.00.25 Fast-01-11-01-10.59.33 EGR1-6-45 11:04 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-11.03.56 Fast-01-11-01-11.04.02 EGR1-6-0b 11:10 800 6 -2 0.6 1.8 0.395 Slow-01-11-01-11.08.59 Fast-01-11-01-11.08.01 Test Point 7 - 1200 RPM. 8 Bar IMEP •' -EGR1-7-0 11:15 1200 8 -2.2 0.6 1.8 0.41 Slow-01-11-01-11.16.09 Fast-01-11-01-11.15.18 EGR1-7-15 11:21 1200 8 -2.2 0.6 1.8 0.41 Slow-01-11-01-11.21.48 Fast-01-11-01-11.20.54 EGR1-7-30 11:26 1200 8 -2.2 0.6 1.8 0.41 Slow-01-11-01-11.26.40 Fast-01-11-01-11.25.30 EGR1-7-10 11:32 1200 8 -2.2 0.6 1.8 0.41 Slow-01-11-01-11.32.07 Fast-01-11-01-11.31.28 EGR1-7-20 11:42 1200 8 -2.2 0.6 1.8 0.41 Slow-01-11-01-11.42.07 Fast-01-11-01-11.41.26 EGR1-7-25 11:54 1200 8 -2.2 0.6 1.8 0.41 Slow-01-11-01-11.54.47 Fast-01-11-01-11.54.05 EGR1-7-0b 11:58 1200 8 -2.2 0.6 1.8 0.41 Slow-01-11-01-11.59.31 Fast-01-11-01-11.58.39 1 est Point 8 - 800 R P M . 10 Bar IMEP ' - • EGR1-8-0 12:36 800 10 -2 0.6 1.8 0.53 Slow-01-11-01-12.36.24 Fast-01-11-01-12.34.32 EGR1-8-10 12:38 800 10 -2 0.6 1.8 0.53 Slow-01-11-01-12.42.52 Fast-01-11-01-12.42.22 EGR1-8-15 12:52 800 . 10 -2 0.6 1.8 0.53 Slow-01-11-01-12.52.52 Fast-01-11-01-12.52.53 EGR1-8-20 13:01 800 10 -2 0.6 1.8 0.53 Slow-01-11-01-13.02.03 Fast-01-11-01-13.01.25 EGR1-8-35 13:10 800 10 -2 0.6 1.8 0.53 Slow-01-11-01-13.10.28 Fast-01-11-01-13.10.24 EGR1-8-30 13:20 800 10 -2 0.6 1.8 0.53 Slow-01-11-01-13.21.31 Fast-01-11-01-13.20.25 EGR1-8-12 13:28 800 10 -2 0.6 1.8 0.53 Slow-01-11-01-13.27.01 Fast-01-11-01-13.26.34 EGR1-8-0b 13:30 800 10 -2 0.6 1.8 0.53 Slow-01-11-01-13.30.25 Fast-01-11-01-13.29.46 85 E G R S e t 1 - R e p l a c e m e n t ( c o n t i n u e d ) ID Time Speed (RPM) Target IMEP (bar) PSOl (ms) PPW (ms) RIT (ms) G P W (ms) LS File .csv HS File .CSV 5/11/01 Testinq Test Point 9 - 800 RPM, 6 Bar IMEP (repeated) • ' EGR1-9-0 10:07 800 6 -2 0.6 1.8 0.385 Slow-01-11-05-10.06.59 Fast-01-11-05-10.07.00 EGR1-9-7 10:10 800 6 -2 0.6 1.8 0.385 Slow-01-11-05-10.11.09 Fast-01-11-05-10.12.23 EGR1-9-20 10:16 800 6 -2 0.6 1.8 0.385 Slow-01-11-05-10.16.11 Fast-01-11-05-10.15.35 EGR1-9-27 10:22 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-10.22.45 Fast-01-11-05-10.22.09 Test Point 10 - '1200 RP.M.YIO'Bar IMEP . V-EGR1-10-0 12:00 1200 10 -2.2 0.6 1.8 0.46 Slow-01-11-05-12.02.49 Fast-01-11-05-12.00.34 EGR1-10-5 12:10 1200 10 -2.2 0.6 1.8 0.46 Slow-01-11-05-12.10.01 Fast-01-11-05-12.09.18 EGR1-10-20 12:15 1200 10 -2.2 0.6 1.8 0.46 Slow-01-114)5-12.15.58 Fast-01-11-05-12.15.29 EGR1-10-30 12:25 1200 10 -2.2 0.6 1.8 0.46 Slow-01-11-05-12.24.23 Fast-01-11-05-12.23.48 EGR1-10-25 12:30 1200 10 -2.2 0.6 1.8 0.46 Slow-01-11-05-12.28.24 Fast-01-11-05-12.30.59 EGR1-10-0b 12:40 1200 10 -2.2 0.6 1.8 0.46 Slow-01-11-05-12.40.53 Test Point 11 - 8 0 0 RPM. 10 Bar IMEP (repeated) «... EGR1-11-0 13:30 800 10 -2.2 0.6 1.8 0.52 Slow-01-11-05-13.31.32 Fast-01-11-05-13.30.49 EGR1-11-10 13:35 800 10 -2.2 0.6 1.8 0.52 Slow-01-11-05-13.36.23 Fast-01-11-05-13.35.12 EGR1-11-15 1 3 : 4 0 - 1 3 800 10 -2.2 0.6 1.8 0.52 Slow-01-11-05-13.43.06 Fast-01-11-05-13.41.02 EGR1-11-35 13:50 800 10 -2.2 0.6 1.8 0.52 Slow-01-11-05-13.51.28 Fast-01-11-05-13.51.24 EGR1-11-25 14:02 800 10 -2.2 0.6 1.8 0.52 Slow-01-11-05-14.02.32 Fast-01-11-05-14.01.47 EGR1-11-0b 14:07 800 10 -2.2 0.6 1.8 0.52 Slow-01-11-05-14.07.30 Fast-01-11-05-14.06.50 Tes(!R6ihtH2ii!800!=RBMS16iBariilMEP . EGR1-12-0 14:40 800 16 -2.2 0.6 1.8 1.2 Slow-01-11-05-14.41.10 Fast-01-11-05-14.40.19 EGR1-12-20 14:46 800 16 -2.2 0.6 1.8 1.2 Slow-01-11-05-14.47.29 Fast-01-11-05-14.46.44 EGR1-12-5 14:55 800 16 -2.2 0.6 1.8 1.2 Fast-01-11-05-14.52.08 EGR1-12-10 15:02 800 16 -2.2 0.6 1.8 1.2 Slow-01-11-05-15.02.34 Fast-01-11-05-15.01.36 EGR1-12-30+ 15:07 800 16 -2.2 0.6 1.8 1.2 Slow-01-11-05-15.05.49 Fast-01-11-05-15.05.51 TestlR6inta3PSrt200IRBMI8!BarllMEPJ(repeatia) • ' -EGR1-13-0 15:38 1200 8 -2.2 0.6 1.8 0.39 Slow-01-11-05-15.37.43 Fast-01-11-05-15.34.24 EGR1-13-5 15:40 1200 8 -2.2 0.6 1.8 0.39 Slow-01-11-05-15.40.44 Fast-01-11-05-15.40.10 EGR1-13-10 15:45 1200 8 -2.2 0.6 1.8 0.39 Slow-01-11-05-15.44.24 Fast-01-11-05-15.42.43 EGR1-13-20 15:50 1200 8 -2.2 0.6 1.8 0.39 Slow-01-11-05-15.49.01 Fast-01-11-05-15.47.47 EGR1-13-35 15:53 1200 8 -2.2 0.6 1.8 0.39 Slow-01-11-05-15.52.36 Fast-01-11-05-15.51.23 EGR1-13-30 15:56 1200 8 -2.2 0.6 1.8 0.39 Slow-01-11-05-15.56.22 Fast-01-11-05-15.55.39 EGR1-13-25 15:59 1200 8 -2.2 0.6 1.8 0.39 Slow-01-11-05-15.59.22 Fast-01-11-05-15.58.38 6/11/01 Testinq • - • -v.- , Test Point 1 4 - 1 6 0 0 RPM. 11 Bar (repeated) * EGR1-14-0 10:36 1600 11 -2.25 0.6 1.8 0.5 Slow-01-11-06-10.35.58 Fast-01-11-06-10.35.34 EGR1-14-5 10:41 1600 11 -2.4 0.6 1.8 0.5 Slow-01-11-06-10.41.51 Fast-01-11-06-10.40.48 EGR1-14-15 10:45 1600 11 -2.4 0.6 1.8 0.5 Slow-01-11-06-10.44.52 Fast-01-11-06-10.44.11 Test Point 15 - 1200 RPM. 11 Bar (repeated) •"" . " ' EGR1-15-15 11:08 1200 11 -2.2 0.6 1.8 0.465 Slow-01-11-06-11.08.50 Fast-01-11-06-11.06.25 EGR1-15-25 11:15 1200 11 -2.2 0.6 1.8 0.475 Slow-01-11-06-11.16.25 Fast-01-11-06-11.15.39 EGR1-15-0 11:27 1200 11 -2.2 0.6 1.8 0.475 Slow-01-11-06-11.27.37 Fast-01-11-06-11.27.03 9/i11'01 Testinq Test Point 16 - 800 RPM, 10 Bar - IMEP held constant throughout tost CIMEP1-0 10:35 800 10 -2 0.6 1.8 0.52 Slow-01-11-09-10.35.56 Fast-01-11-09-10.34.40 CIMEP1-22 10:43 800 10 -2 0.6 1.8 0.52 Slow-01-11-09-10.43.41 Fast-01-11-09-10.42.58 CIMEP1-30+ 10:57 800 10 -2 0.6 1.8 0.7 Slow-01-11-09-10.56.36 Fast-01-11-09-10.55.24 E G R S e t 2 - S u p p l e m e n t a l ID Time Speed (RPM) Target IMEP (bar) PSOl (ms) PPW (ms) RIT (ms) G P W (ms) LS File .CSV HS File .CSV 5/11/01 Testinq • ' Test Point 1 - 800 R P M , 6 Bar IMEP EGR2-1-0 10:31 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-10.30.49 Fast-01-11-05-10.30.50 EGR2-1-7 10:34 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-10.34.17 Fast-01-11-05-10.33.41 EGR2-1-20 10:40 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-10.40.03 Fast-01-11-05-10.39.33 EGR2-1-37 10:50 800 6 -2 0.6 1.8 0.39 Slow-01-11-05-10.50.03 Fast-01-11-05-10.48.55 6/11/011'Testinq ~ . Test Point 2 - 800 RPM. 11"Bar IMEP - •- : . -EGR2-2-0 9:28 800 11 -2 0.6 1.8 0.52 Slow-01-11-06-09.29.59 Fast-01-11-06-09.28.31 EGR2-2-15 9:45 800 11 -2 0.6 1.8 0.52 Slow-01-11-06-09.46.24 Fast-01-11-O6-09.45.06 EGR2-2-20 9:50 800 11 -2 0.6 1.8 0.52 Slow-01-11-06-09.50.54 Fast-01-11-06-09.50.22 EGR2-2-2 9:56 800 11 -2 0.6 1.8 0.52 Slow-01-11-06-09.57.37 Fast-01-11-06-09.56.59 EGR2-2-12 10:03 800 11 -2 0.6 1.8 0.52 Slow-01-11-06-10.01.54 Fast-01-11-06-10.01.12 EGR2-2-0b 10:08 800 11 -2 0.6 1.8 0.52 Slow-01-11-06-10.09.05 Fast-01-11-06-10.08.24 Test Point 3 - 1 2 0 0 RPM. 11 Bar IMEP . SSKS? EGR2-3-0 9:28 1200 8 -2.2 0.6 1.8 0.385 Slow-01-11-06-11.37.37 Fast-01-11-06-11.35.39 EGR2-3-7 9:45 1200 8 -2.2 0.6 1.8 0.385 Slow-01-11-06-11.40.57 Fast-01-11-06-11.40.26 EGR2-3-20 9:50 1200 8 -2.2 0.6 1.8 0.37 Slow-01-11-06-11.45.32 Fast-01-11-06-11.44.54 EGR2-3-30 9:45 1200 8 -2.2 0.6 1.8 0.37 Slow-01-11-06-11.49.06 Fast-01-11-06-11.48.41 EGR2-3-0b 9:50 1200 8 -2.2 0.6 1.8 0.385 Slow-01-11-06-11.52.35 Fast-01-11-06-11.51.38 86 EGR Set 3 - Temperature ID Time Speed (RPM) Target IMEP (bar) PSOl (ms) PPW (ms) RIT (ms) GPW (ms) LS File .CSV HS File .CSV 09/11/2001 Test Point 1 - 8 0 0 RPM. 10 Bar IMEP -TT1-0 12:38 800 10 -2 0.6 1.8 0.52 Slow-01-11-09-12.39.10 Fast-01-11-09-12.38.36 TT1-15 12:43 800 10 -2 0.6 1.8 0.52 Slow-01-11-09-12.43.48 Fast-01-11-09-12.43.02 TT1-30 12:47 800 10 -2 0.6 1.8 0.52 Slow-01-11-09-12.48.17 Fast-01-11-09-12.47.16 TT1-30+ 12:50 800 10 -2 0.6 1.8 0.52 Slow-01-11-09-12.51.09 Fast-01-11-O9-12.50.09 TT1-10 12:58 800 10 -2 0.6 1.8 0.52 Slow-01-11-09-12.58.01 Fast-01-11-09-12.57.13 EGR Set 4 - ESC Mode Tests ID Time Speed (RPM) Load % PSOl ms PPW ms RIT ms GPW ms Slow .CSV Fast .CSV 14/11/2001 ESC Mode'5 - '1067 R P M . 983 Nm BT ESC5-1 14:20 1067 50% -0.542 0.6 1.8 0.7 Slow-01-11-14-14.21.45 Fast-01-11-14-14.18.51 ESC5-1 -3% 14:30 1067 50% -0.542 0.6 1.8 0.7 Slow-01-11-14-14.32.28 Fast-01-11-14-14.32.24 ESC5-1-15% 14:40 1067 50% -0.542 0.6 1.8 0.7 Slow-01-11-14-14.37.54 Fast-01-11-14-14.37.04 ESC5-1-10% 14:50 1067 50% -0.542 0.6 1.8 0.7 Slow-01-11-14-14.46.57 Fast-01-11-14-14.46.13 ESC5-1-20% 14:55 1067 50% -0.542 0.6 1.8 0.7 Slow-01-11-14-14.54.57 Fast-01-11-14-14.54.01 ESC5-1-0% 15:00 1067 50% -0.542 0.6 1.8 0.7 Slow-01-11-14-15.01.54 Fast-01-11-14-15.00.52 E S C Modn 7 - 1067 RPM. 492 Nrh B l ESC7-1 15:16 1067 25% -0.511 0.6 1.8 0.4 Slow-01-11-14-15.16.01 Fast-01-11-14-15.16.49 ESC7-1 -5% 15:28 1067 25% -0.511 0.6 1.8 0.4 Slow-01-11-14-15.28.46 Fast-01-11-14-15.28.46 ESC7-1-10% 15:34 1067 25% -0.511 0.6 1.8 0.4 Slow-01-11-14-15.34.56 Fast-01-11-14-15.34.11 ESC7-1-20% 15:38 1067 25% -0.511 0.6 1.8 0.4 Slow-01-11-14-15.39.29 Fast-01-11-14-15.38.27 ESC7-1-10%b 15:44 1067 25% -0.511 0.6 1.8 0.4 Slow-01-11-14-15.43.43 Fast-01-11-14-15.43.30 ESC7-1-0% 15:50 1067 25% -0.511 0.6 1.8 0.4 Slow-01-11-14-15.47.40 Fast-01-11-14-15.45.22 F S C Mode 6 - 1067 RPM. 1475 Nm BT -:"C • ESC6-1 | 15:55| 1067| 75%| -0.844| 0.6| 1.8| 1.3|Slow-01-11-14-15.52.13|Fast-01-11-14-15.51.22 Table A.5 EGR testing summary. 87 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0099623/manifest

Comment

Related Items