INTRODUCTIONThroughout the world, many efforts are being made to improve the thermal efficiency of internal combustion engines. One relatively new approach is gasoline partially-premixed compression ignition (PPCI) that was introduced by Kalghatgi [1] and first tested by Johansson [2]. A high octane fuel was injected late on the compression stroke of a boosted diesel engine operating with high EGR. The injection process was complete prior to the start of combustion enabling partial mixing of the fuel and air prior to heat release. Very low fuel consumption, NOx, and PM emissions were measured. This early work established that gasoline-like fuels with high resistance to autoignition are preferred for PPCI. These fuels have greater ignition delay relative to high cetane fuels, which

gives them more time for mixing. The high volatility of gasoline also helps mix the fuel and air quickly after injection. Kalghatgi taught that PPCI requires some stratification of the fuel-air mixture (“partially premixed but not fully mixed”) to achieve ignition and controlled burn duration.

Since this early work, many researchers have performed additional engine experiments and simulations, which have contributed to a significant body of PPCI test data and experience. Tests have been performed on diesel engines using various gasoline-like fuels at Shell [3, 4, 5, 6], Lund [7, 8, 9, 10], Cambridge [11], UW-Madison [12, 13, 14, 15, 16], Argonne [17, 18, 19, 20], Tsinghua [21], and RWTH [22]. In practically all cases, diesel-like efficiency was measured. Various injector designs and piston designs have also been

Development of a Gasoline Direct Injection Compression Ignition (GDCI) Engine

Mark Sellnau, Matthew Foster, Kevin Hoyer, Wayne Moore, James Sinnamon, and Harry Husted

Delphi Automotive

ABSTRACTIn previous work, Gasoline Direct Injection Compression Ignition (GDCI) has demonstrated good potential for high fuel efficiency, low NOx, and low PM over the speed-load range using RON91 gasoline. In the current work, a four-cylinder, 1.8L engine was designed and built based on extensive simulations and single-cylinder engine tests. The engine features a pent roof combustion chamber, central-mounted injector, 15:1 compression ratio, and zero swirl and squish. A new piston was developed and matched with the injection system. The fuel injection, valvetrain, and boost systems were key technology enablers.

Engine dynamometer tests were conducted at idle, part-load, and full-load operating conditions. For all operating conditions, the engine was operated with partially premixed compression ignition without mode switching or diffusion controlled combustion. At idle and low load, rebreathing of hot exhaust gases provided stable combustion with NOx and PM emissions below targets of 0.2g/kWh and FSN 0.1, respectively. The coefficient of variation of IMEP was less than 3 percent and the exhaust temperature at turbocharger inlet was greater than 250 C. BSFC of 280 g/kWh was measured at 2000 rpm-2bar BMEP. At medium-to-higher loads, rebreathing was not used and cooled EGR provided NOx, PM, and combustion noise below targets. MAP was reduced to minimize boost parasitics. At full load operating conditions, near stoichiometric mixtures were used with up to 45 percent EGR. Maximum BMEP was about 20 bar at 3000 rpm. For all operating conditions, injection quantities and timings were used to control mixture stratificaton and combustion phasing.

Transient co-simulations of the engine system were conducted to develop control strategies for boost, EGR, and intake air temperature control. Preliminary transient tests on a real engine with high rate of load increase demonstrated potential for very good control. Cold start simulations were also conducted using an intake air heating strategy. Preliminary cold start tests on a real engine at room temperature demonstrated potential for very good cold starting. More work is needed to calibrate the engine over the full operating map and to further develop the engine control system.

CITATION: Sellnau, M., Foster, M., Hoyer, K., Moore, W. et al., "Development of a Gasoline Direct Injection Compression Ignition (GDCI) Engine," SAE Int. J. Engines 7(2):2014, doi:10.4271/2014-01-1300.

2014-01-1300Published 04/01/2014

Copyright © 2014 SAE Internationaldoi:10.4271/2014-01-1300

saeeng.saejournals.org

developed with significant improvements. Delphi [23, 24, 25, 26, 27] reported single-cylinder and multi-cylinder engine test results with various injectors using single, double, and triple injection strategies.

Engine tests have also been performed using naphtha fuels on both modified spark-ignited engines [28] and modified diesel engines [29]. Naphtha has significantly lower octane number (RON 70 to 84) and significantly lower processing cost compared to current market gasoline. This work has shown compatibility of the PPCI combustion process with lower octane fuels in a longer term perspective.

PPCI has demonstrated very good potential for very high fuel efficiency with low engine-out NOx and PM emissions using a range of gasoline-like fuels. However, towards a production solution, significant issues remain. Due to the lower exhaust enthalpy of low temperature engines using PPCI, it is difficult to produce intake boost with acceptable boost system parasitics. A practical powertrain system with robust PPCI combustion is needed, including injection, valvetrain, boost, and exhaust subsystems. The engine must also meet vehicle packaging requirements under hood and satisfy cold start and transient response requirements.

As part of a US Department of Energy funded program, Delphi has been developing a multi-cylinder engine concept for PPCI combustion with current US market gasoline (RON91). The engine has four cylinders with a 1.8L displacement and was designed based on extensive simulations and single-cylinder engine tests. A multiple-late-injection (MLI) strategy with GDi-like injection pressures was selected without use of a premixed charge. The absence of classic knock and pre-ignition limits in this process enabled a higher compression ratio of 15. The engine operates “full time” [25] over the entire operating map with partially premixed compression ignition. No combustion mode switching, diffusion controlled combustion, or spark plugs were used. Delphi uses the term Gasoline Direction Injection Compression Ignition (GDCI) in reference to this combustion process.

One program objective was to build a practical GDCI engine that achieves diesel-like fuel efficiency using current market gasoline (RON91) with engine-out NOx and PM emissions below that needed for NOx or PM aftertreatment. Table 1 lists initial targets for engine testing and development. Combustion noise level (CNL) limits are shown in Figure 1. Other program objectives include demonstration of good transient load response and room temperature cold starts.

In the current work, analysis and simulation tools were used to design and fabricate a new multi-cylinder GDCI engine. Design tools were used to package the powertrain in a D-class vehicle. Engine dynamometer tests were conducted over a range of operating conditions and included preliminary calibration mapping. With this data, a competitive assessment of brake specific fuel consumption (BSFC) was performed using published data for gasoline, diesel, and hybrid vehicle engines.

Finally, aggressive transients with high rate of load increase were simulated and then tested on the real engine. Cold starts at various ambient temperatures were also simulated and tested on the real engine.

Table 1. Preliminary Targets for Engine Testing.

Figure 1. Combustion Noise Level (CNL) Limits used for Testing.

GDCI CONCEPT AND INJECTION STRATEGYThe GDCI engine concept combines the best of diesel and spark-ignited engine technology. Like diesel engines, the compression ratio is high, there is no intake throttling, and the mixture is lean for improved ratio of specific heats. GDCI uses a new low-temperature combustion process for partially-premixed compression-ignition. Multiple-late-injections of gasoline (RON91) vaporize and mix very quickly at low injection pressure typical of direct injected gasoline engines. Low combustion temperatures combined with low mixture motion and reduced chamber surface area result in reduced heat losses.

A schematic of the GDCI combustion chamber concept is shown in Figure 2. The engine features a shallow pent roof combustion chamber, central-mounted injector, and 15:1 compression ratio. A quiescent, open chamber design was chosen to support injection-controlled mixture stratification. Swirl, tumble, and squish were minimized since excessive mixture motion may destroy mixture stratification created during the injection process. The piston and bowl shape were matched with the injection system and spray characteristics. The bowl is a symmetrical shape and was centered on the cylinder and injector axes.

The GDCI injection strategy is central to the overall GDCI concept and is depicted in the Phi-T (equivalence ratio-temperature) diagram shown in Figure 3. The color contours in Figure 3 show simulated CO emissions concentration. The injection process involves one, two, or three injections during

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

the compression stroke and are shown as quantities Q1, Q2, and Q3. Each injection begins in the upper left of the Phi-T diagram (liquid) and quickly vaporizes and mixes to Phi less than 2 by the start of combustion (SOC). At SOC the fuel-air mixture is stratified to achieve stable ignition and controlled heat release. Wall wetting is minimized and fuel is kept away from cold zones such as the piston topland and cylinder liner that may impede complete oxidation.

Figure 2. GDCI combustion chamber concept.

The ideal “stratification line” is shown in Figure 3 and represents the ideal injection process. To achieve low NOx and low PM emissions simultaneously, combustion must occur “in the box” shown in Figure 3 (avoiding soot and NOx formation regions). To also minimize CO emissions, which can compromise efficiency, combustion must occur in the region 0<Phi<1.2 with 1300<T<2200 degrees K. Therefore at SOC, which is approximately top-dead-center (TDC), all parcels in the combustion chamber should be no richer than Phi of approximately 1.2. This corresponds to the top of the stratification line in Figure 3. The stratification line is inclined due to charge cooling effects with the richest parcels cooling more than the leanest parcels.

Figure 3. GDCI Injection Strategy depicted on Phi-T Diagram with NOx, Soot, and CO Contours.

KIVA simulations were performed with this combustion system and an injector spray model developed using spray chamber test data. A full 360-degree mesh with a fine grid was necessary to capture the injection, mixing, and combustion

phenomena. Results for a double injection strategy at 2000 rpm and 11 bar indicated mean effective pressure (IMEP) are shown in Figure 4a,b,c,d,e at various crank positions.

Figure 4a. Phi-T diagram at −50 CAD atdc (during first injection).

Figure 4b. Phi-T diagram at −34 CAD atdc (start of second injection).

Figure 4c. Phi-T diagram at −10 CAD atdc (prior to start of combustion).

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

Figure 4d. Phi-T diagram at 0 CAD atdc (near start of combustion).

Figure 4e. Phi-T diagram at 21 CAD atdc (near end of combustion).

Figure 4a at −50 crank angle degrees (CAD) after-top-dead-center (atdc) corresponds to the first injection. The fuel spray is entraining air rapidly with very fast atomization and vaporization in the bulk gas. Each data point plotted in Figure 4a corresponds to a different parcel in the combustion chamber with color indicating the fuel mass fraction. Figure 4b at −34 CAD atdc corresponds to the beginning of the second injection. Like the first injection, fuel remains in the bulk gas without contacting cool wall zones. This second and last injection controls the ignition process. Figure 4c at −10 CAD atdc shows the mixture stratification prior to start of low-temperature reactions. As fuel vaporizes and mixes, the equivalence ratio of the richest part of the charge is decreasing. If too lean, the charge will be overmixed and ignition compromised. If too rich, soot and CO emissions will increase. The stratification line is inclined to lower temperatures for the richer elements of the charge due to charge cooling. Figure 4d at TDC, is approximately start of combustion with some temperature rise visible in the figure. Parcels with Phi near stoichiometry are the first to release heat. Figure 4e at 21 CAD atdc corresponds to near end-of-combustion (EOC). CO has burned out across the chamber and small amounts of NOx were produced in the hottest portions of the chamber. The temperature in the Phi-T plot has now shifted to the right due to heat release and temperature rise for each parcel. Richer parcels realize higher temperature rise than the leanest parcels.

This sequence of KIVA simulations represents the intended injection, mixing, and combustion processes for GDCI combustion. Because combustion occurs at relatively low peak gas temperatures, it is referred to as low-temperature combustion. The lower combustion temperatures are a main factor in reducing heat transfer through the cylinder head, liner, and piston and contribute to very low indicated specific fuel consumption and emissions.

ENGINE DESCRIPTIONThe absence of classic combustion knock and preignition for GDCI means that a GDCI engine can be operated on RON91 gasoline at high compression ratio for high efficiency. For this reason, GDCI is also a good candidate for aggressive downsizing, down-speeding, and boosting, which is a proven strategy for high vehicle-level fuel economy. These were two main considerations in base engine design.

Table 2. Base Engine Specifications.

Figure 5. GDCI Multi-cylinder Engine.

Table 2 lists the base engine specifications. The engine is a four-cylinder 1.8L engine with four-valves-per-cylinder. Bore, stroke, and connecting rod length are 82 mm, 85.2 mm, and 144.7 mm, respectively, for B/S of 0.96 and r/l of 0.29. With

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

geometric compression ratio of 15, this engine has geometry similar to modern light duty diesel engines. A photograph of the first assembled GDCI engine is shown in Figure 5.

Cylinder HeadThe cylinder head is an all new design with four-valve-per-cylinder rated at 200 bar peak cylinder pressure (PCP) for development purposes. Extensive analysis work was conducted to develop the cylinder head using conventional aluminum alloys to meet cylinder sealing requirements. A new multilayer steel head gasket was developed for this application. The injector is central-mounted without offset or inclination relative to the cylinder axis. The combustion chamber features a shallow pent roof with valves flush with the pent roof surface. The low pent roof angle was selected to achieve high compression ratio, while maintaining low chamber surface area and reasonable valve sizes. Intake and exhaust ports were developed for good flow characteristics but without the need for swirl, tumble, or squish. This was a significant advantage and produced good intake flow efficiency as a function of valve lift as measured on a steady flow bench. A double overhead camshaft, Type II valvetrain with roller finger followers was selected due to low friction and compact packaging. No spark plug exists in the cylinder head but there is provision for cylinder pressure sensors. In addition to cylinder pressure sensing, Kistler cylinder pressure transducers were mounted vertically at the cylinder periphery. Overall, the cylinder head is very compact and employs conventional aluminum casting methods, while meeting demanding structural requirements.

PistonA new piston was designed for GDCI based on extensive KIVA simulations. The piston bowl shape and injector spray characteristics were matched for typical GDCI injection timings. This resulted in significantly reduced piston surface area and less propensity for wall wetting relative to typical diesel pistons. The bowl is symmetrical and is located on the cylinder axis.

Table 3. Main Features of GDCI Piston relative to typical LD diesel piston [27].

The main features of the GDCI piston are listed in Table 3 [27]. The piston, pin, and ring pack were rated at 200 bar peak cylinder pressure. Surface area, squish area and blowby area (defined by ring end gap and second land radial clearance) were significantly reduced relative to typical light-duty diesel piston designs (BMW M57 piston used as reference). Back-to-back tests were conducted on a single-cylinder engine using

the GDCI and baseline diesel pistons. Significant fuel consumption improvements were measured at several speed-load points as shown in Figure 6 [27].

Figure 6. Fuel consumption and combustion noise for GDCI piston compared to diesel baseline from Delphi [24].

Injection SystemPrevious work by the authors has demonstrated improved GDCI combustion performance through injection system developments [23, 24, 25, 26, 27]. Simulation tools were used to develop injector spray characteristics that provide very fast vaporization, low spray penetration, while also providing the necessary injection rate (flow rate). Designed experiments on a single-cylinder engine were used to determine best injection timings, quantities, and injection pressures for various injector designs. Injection pressures are in the range used by production gasoline direct injection (GDi) engines. Lower injection pressures are desirable to reduce pump parasitics and fuel system cost.

The fuel pump was mounted at the end of the cylinder head and was driven by the exhaust camshaft. The pump shaft has four lobes and exhibits excellent rail pressure control via modulation of the spill valve. The fuel rail has inlet and outlet orifices for dampening of pressure pulsations in the system. Injector drive waveforms were developed to support the multiple late injection strategies used for GDCI. Importantly, the system delivers accurate fuel quantities for multiple injections in all cylinders.

Valvetrain SystemFull-time GDCI combustion over the operating map is achievable by using exhaust rebreathing (RB) at low loads and cooled EGR at medium to higher loads. Rebreathing (RB) is accomplished using the exhaust valvetrain system, which provides a secondary exhaust lift event during the intake stroke. It is an effective method to recuperate heat from hot exhaust gases in order to raise mixture temperatures. This heat promotes autoignition at low loads when boost pressure is zero or low. Rebreathing also collapses the pumping loop (intake and exhaust valves are open at same time) and raises

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

exhaust temperatures significantly. In this way, a close-coupled oxidation catalyst will remain warm over a wide range of low loads without special maintenance heating.

The valvetrain system is a Type II, continuously-variable mechanical system that provides the rebreath functionality. The valvetrain has good packaging characteristics with low friction. A Delphi electric camshaft phaser is used to actuate the exhaust valvetrain and control the secondary valve lift with very fast response. Figure 7 shows a photograph of the front of the engine with electric camshaft phasers.

Figure 7. Photograph of front of engine with electric phasers.

Boost SystemIntake boosting is difficult for low-temperature engines because the exhaust enthalpy to drive a turbocharger is very low. For medium-and-high-load operation, GDCI requires increased cooled EGR to dilute the charge and maintain proper combustion phasing. This puts greater demand on the boost system. Extensive engine simulations were performed [26] to develop a system that provides the necessary intake boost with low boost parasitics. Several boost architectures with various boost devices were studied including a 1-stage turbocharger (TC), a 2-stage TC, a 1-stage supercharger (SC), and a 2-stage TC-SC system. The 2-stage system with one variable geometry TC, one SC, and two charge coolers was selected for the most efficient operation over the operating range.

Figure 8 shows a schematic of the boost system. No intake throttle is used. The overall system including liquid-cooled charge air coolers was developed and packaged for minimum system volume, good cylinder-to-cylinder flow uniformity, and low flow restriction on the vehicle. Since the supercharger is a positive displacement device driven by the engine, transient response of the boost system is very fast.

The LPL-EGR system is also shown in Figure 8. The exhaust gases exit the turbine and pass through a close-coupled oxidation catalyst to reduce HC and CO emissions. Exhaust gases at this point are already relatively cool. The EGR flow enters an EGR cooler located upstream of the EGR valve. The EGR cooler operates at engine coolant temperature to prevent water condensate at the TC compressor inlet. The EGR system is designed for high flow in a compact layout for fast response on the vehicle.

Figure 8. Schematic of boost system with charge air coolers and LPL-EGR.

Exhaust System and AftertreatmentFigure 9 shows a photograph of the exhaust system used on multicylinder GDCI engines. The manifold has a 4-into-1 configuration with compact, symmetrical design, and tubular construction. The runner lengths are nearly identical for each cylinder for uniform flow and pressure dynamics. This insures that rebreathing of exhaust gases will also be uniform across the cylinders. The exhaust system is insulated to maintain sufficient exhaust temperatures for aftertreatment.

The aftertreatment system is comprised of only an oxidation catalyst. The intent is to satisfy tailpipe NOx and PM emissions targets using clean GDCI combustion technology. The oxidation catalyst is close-coupled to the turbocharger outlet for heat conservation. The catalyst brick is a round, alumina monolith with high loading of platinum and palladium for low temperature light off of CO and HC species. The entire exhaust system including flex pipes and sensors is packaged for the GDCI vehicle application. An exhaust pipe and muffler were used for dynamometer testing to emulate the restriction of the vehicle exhaust system.

Figure 9. Photograph of GDCI exhaust system on a dynamometer engine.

EXPERIMENTAL METHODSExperiments were performed on an engine dynamometer at Delphi. The test fuel was Shell regular unleaded gasoline with approximately 10 percent ethanol and research octane number (RON) of 91, as is representative of commercial gasoline fuel in the United States. Fuel property data are listed in Table 4.

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

Table 4. Properties of Shell Test Fuel.

Engine air flow was measured by a Meriam laminar flow element with a 55 gallon surge tank located upstream of the engine. Combustion air was conditioned for temperature and humidity. Fuel flow was measured with an AVL 735 Fuel Mass Flow Meter in combination with an AVL 753 Fuel Temperature Controller. Fuel flow data for operation on E10 fuel was corrected to E00 based on the lower heating value of the test fuel.

Exhaust emissions were measured conventionally. Non-dispersive infrared analysers were used for CO2, CO, and intake CO2 species. Intake CO2 data was used to calculate EGR levels. A chemiluminescence analyser was used for NOx species, a heated flame-ionization detector was used for total HC, and a paramagnetic analyser was used for O2. Exhaust smoke was measured with an AVL 415S smoke meter. Exhaust particulate size distribution was measured with a TSI 3090 Engine Exhaust Particle Size Spectrometer [30, 31]. The sampling system for exhaust particulate is described in [24].

Cylinder pressure was measured with a flush-mounted Kistler 6125CU20 pressure transducer. These transducers were inserted from the top of the cylinder head and located flush with the chamber near the cylinder bore. A Kistler 2613B crankshaft encoder provided crank position data and was dynamically aligned with engine TDC using a Kistler 2629B TDC probe. Cylinder pressure data was sampled every 0.5 CAD. Indicated mean effective pressure (IMEP) was reported on a gross basis. Combustion noise level was measured using an AVL FLEXIFEM Noise Meter [32].

Engine ControllerThe engine controller used for GDCI dynamometer engine testing is based on National Instruments hardware and LabView software [33], and was built by National Instruments-San Antonio. The system schematic is shown in Figure 10. The controller features real-time combustion analysis using NI Combustion Analysis System (NI-CAS) and allows rapid prototyping of test cell engine control software using LabView. The controller is comprised of a National Instruments 8110 Real-Time processor, two PXi 7813R 3M gate field

programmable gate arrays (FPGA), two PXi 6123 modules for fast data acquisition, a PXi 8512 for CAN communications, and numerous cRIO modules for interfacing to engine sensors, actuators, and other test cell instrumentation.

Figure 10. Schematic of Test Cell Engine Control System for GDCI Multicylinder Engine Testing.

Engine control software for dynamometer testing was developed by Delphi. The software balances IMEP for each cylinder based on cylinder pressure data. A multiple-injection-control utility [23] was developed to control the fuel injection quantity, Q, for each injection event. The Q for each injection is determined using instantaneous rail pressure, cylinder pressure, and an embedded calibration map for each injector. Fuel rail pressure is tightly controlled using the rail pressure sensor and the GDi fuel pump spill valve. Intake boost is controlled by the supercharger and turbocharger rack position for minimum boost parasitics. Engine coolant temperature and liquid-cooled, charge air coolers are controlled via temperature feedback to maintain prescribed set points. The engine controller provides various other controls functions for the EGR valve, SC clutch, 2-stage oil pump, etc.

A CAN bus was used to communicate among the main controller, actuator smart controllers, and the PUMA test system in the test cell. The CAN bus passes information for site safeties, emissions, fuel flow, as well as pressure transducers and thermocouple values (Figure 10). Data acquisition of all high-speed combustion data and low-speed data is performed by the engine controller and saved in a single, binary file.

PRELIMINARY CALIBRATION TESTSEngine calibration tests were performed over a wide range of steady-state, warmed-up, operating conditions. The engine included all subsystems including all accessories (water pump, air conditioner compressor, alternator, accessory drive), and complete vehicle exhaust system. Typical operating settings for these tests are listed in Table 5.

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

The engine underwent a break-in schedule and motoring friction was measured using the “motoring method”. Motoring friction mean effective pressure (M-FMEP) is defined as brake mean effective pressure (BMEP) minus net mean effective pressure (NMEP), where NMEP is carefully measured using cylinder pressure transducers. M-FMEP is shown in Figure 11 and is reasonably low for the first build of this new engine.

Table 5. Engine temperatures, air humidity, fuel and lubricant used for calibration mapping tests.

Figure 11. Motoring Friction Mean Effective Pressure (M-FMEP) for first multicylinder GDCI engine.

Figures 12a and 12b show GDCI combustion characteristics at 1000 rpm and 3 bar IMEP load. Start of combustion is near TDC with short combustion duration for high expansion of all burned fuel for high thermodynamic efficiency. The heat release profile does not have a diffusion tail, however, is long enough for quiet combustion. The PV diagram in Figure 12b shows very low pumping loop of 3 kPa.

Figure 12a. Cylinder pressure and heat release for GDCI combustion at 1000 rpm and 3 bar IMEP.

Figure 12b. Log PV diagram for GDCI combustion at 1000 rpm and 3 bar IMEP.

Calibration tests were performed in two test sessions. The first session was for low loads (2-4.5 bar BMEP) and low speeds (800-1500 rpm) for which exhaust rebreathing was used. Results for these tests are shown in Figure 13. The second test session was for medium-to-higher loads (8-17 bar BMEP) in the medium speed range (1500-2500 rpm) for which exhaust rebreathing is not used and cooled EGR is used. Results for these tests are shown in Figure 14. Injection strategies vary as a function of operating condition. For all tests, “design of experiments” methods were not used due to the high number of control factors and limited time for testing. Test results are considered preliminary and not yet optimized.

For all operating conditions, targets for NOx, PM, CNL, and coefficient of variation of IMEP (COV IMEP) were fully met. Brake specific fuel consumption at all conditions was very good with indicated specific fuel consumption (ISFC) below 180 g/kWh over very wide operating ranges. Minimum BSFC of 214 g/kWh was observed at 2000 rpm-13.5 bar BMEP. GDCI BSFC data is compared to other competitive engines in the following section.

For low loads, smoke was practically absent, combustion noise was well below targets, and combustion stability was typically less than 3 percent. Exhaust rebreathing was very effective to maintain exhaust temperatures at the turbocharger inlet typically above 250 C for the lowest loads tested. Combustion efficiency ranged from 92 to 96 percent. Further improvements in combustion efficiency are expected with improved calibrations.

For medium-to-high loads, NOx emissions exhibited a unique decreasing trend with increasing load. This is contrary to typical NOx trends for spark-ignited and diesel engines and is attributed to increasing EGR levels with load. This is a very important outcome and will be a key enabler to meet tailpipe NOx targets in upcoming vehicle testing.

For all tests, peak exhaust temperatures at the turbocharger inlet were below about 500 C, which is desirable for durability of exhaust system components. For all tests, peak cylinder pressure was less than 160 bar, inlet manifold pressure was less than 3.5 bar absolute, and EGR was less than 45 percent.

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

Figure 13. Measured fuel consumption, emissions, and combustion analysis results for low-load calibration tests.

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

Figure 14. Measured fuel consumption, emissions, and combustion analysis results for med-high load calibration tests.

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

Figure 15. Competitive engines selected for comparison.

Table 6. Estimated BSFC data (g/kWh) for competitive engines at various test conditions.

BSFC COMPETITIVE ASSESSMENTThe primary objective of this program is to achieve diesel-like fuel economy in a practical gasoline multi-cylinder GDCI engine. Toward this goal, a competitive assessment was performed using recent published data for leading engines in the same displacement and power class. Competitive engines selected for comparison included two diesel engines [34, 35], two hybrid vehicle engines [36, 37], and two turbo GDI engines equipped with variable valvetrain systems [38, 39]. Figure 15 shows photos of these engines.

Engines designed for different markets in the world use different fuels and are compliant with different emissions regulations. In Europe, gasoline octane number and diesel cetane number are significantly higher than in the US. As well, in the US, emissions standards are significantly more stringent than in Europe. There may also be differences in accessories used for testing, and this can affect fuel consumption. Therefore, it is difficult to make direct comparisons of fuel consumption among some engines based on published data alone.

Table 6 lists the engines in the assessment with their respective model year, fuel used, emission compliance, and reference. On the right side of Table 6, brake specific fuel consumption (g/kWh) is listed for various operating conditions including the “best efficiency point” on the map. The data reported for the GDCI engine is obtained directly from the calibration mapping results presented in the last section. No corrections for small differences in fuel lower heating value (LHV) were made.

For the “best efficiency point” on the map, the BSFC of the GDCI engine matches the hybrid vehicle engines, is about 5 percent higher than the two diesels, and is significantly better

than the spark ignited engines for which data is published. For the low speed/load range, BSFC for GDCI is better than most of the diesel engine data. BSFC for GDCI is also significantly better than the SI engine data and comparable to the very efficient hybrid vehicle engines. At the 2000 rpm-2 bar BMEP world test point, BSFC for GDCI was measured at 280 g/kWh, which is lower than all gasoline and diesel engines for which data could be found. For the medium-to-high speed/load range, BSFC for GDCI was within 1 to 5 percent of diesel data, and significantly better than the spark-ignited engines. The hybrid vehicle engines did not produce enough torque at these loads for a comparison and are labelled as “off map”.

In summary, the first multi-cylinder GDCI engine produced very good BSFC in preliminary testing. The results are not considered optimized. Further BSFC improvements are expected through near term developments of the combustion system, improved firing friction, and refined engine calibrations.

FULL-LOAD PERFORMANCE CHARACTERISTICSPreliminary engine dynamometer tests were performed to characterize full-load performance of the GDCI multicylinder engine. For full load, the air-fuel ratio was typically at or near stoichiometry and cooled EGR was used up to about 45 percent. Intake air temperature was controlled to 55 C and inlet manifold absolute pressure (MAP) was limited to 3.5 bar.

Results are shown in Figure 16. For engine speeds of 1500, 2000, and 2500 rpm, BMEP was measured up to about 18 bar. For these tests, NOx, smoke, CNL, and COV IMEP met stringent program targets. Even at full load, the engine emitted

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

NOx less than 0.02 g/kWh and smoke less than 0.08 FSN. This is expected to be lower than engine-out emissions of most if not all production gasoline and diesel engines in the world.

Figure 16. Measured full-load BMEP 1500-2500 rpm and simulated (GT Power) full-load BMEP for higher engine speeds.

For engine speeds above 2500 rpm, tests were interrupted due to a component failure external to the engine, thus making this data unavailable for this paper. As an estimate of expected performance, simulations were conducted using GT Power [40]. The model was correlated with measured full load data at 1500, 2000, and 2500 rpm, and is described in a previous publication [26]. The model predicted maximum BMEP of about 20 bar at about 3000 rpm with 45 percent EGR. Engine test data will be obtained to confirm these simulations.

The GDCI engine and boost system are capable of providing greatly increased full-load performance if lower EGR levels are permissible. Figure 16 shows simulation results for lower EGR, with up to 35 bar BMEP predicted with zero EGR. To operate at these lower EGR levels, a late injection strategy for diffusion-controlled combustion is expected to be necessary. While this could be developed, Delphi is pursuing the use of clean GDCI combustion over the entire operating map. Overall, the preliminary full-load results for the first GDCI multicylinder engine meet preliminary expectations.

TRANSIENT CO-SIMULATION AND TESTAccurate combustion control during speed/load transients is an important requirement for GDCI multicylinder engines. Robust auto-ignition with proper combustion phasing and heat release rates must be maintained. The combustion process depends on many interactive operating variables such as intake manifold pressure (MAP), intake air temperature (IAT), exhaust gas recirculation (EGR) and exhaust rebreath quantity (RB); all of which affect the state of the trapped cylinder charge (temperature, pressure, composition). Optimal steady-state values for these parameters are determined in dynamometer testing and used as target values during transients. The overall goal for transient control is to maintain MAP, IAT, and EGR as close as possible to the optimal targets

The combustion process also depends on fuel injection parameters (number of injection pulses, timing and quantity of each pulse), which are also optimized in dynamometer testing. Injection parameters are fast parameters and can be accurately controlled on an individual-cylinder, individual-cycle basis with very fast control capability. However, MAP, IAT, and EGR responses are relatively slow due to transport delays, thermal inertia and actuator slew rate limits, which are intrinsic to the system. For the best possible control, the intake, exhaust, and EGR systems were designed and packaged very compactly with minimal pipe lengths and plenum volumes.

To aid in controls developments, a detailed and comprehensive GT-Suite [40] model of the engine system was created and coupled in co-simulation to a Simulink model [41] of the control algorithms. Figure 17 shows the upper level of the system model. Figure 18 shows the upper level of the engine model consisting of the base GDCI engine, an engine thermal model, an engine cooling system model, and a charge air cooling system model. The base engine model includes the low-pressure EGR system with EGR cooler, turbocharger, the first charge air cooler (CAC-1), supercharger, and the second charge air cooler (CAC-2). The engine cooling system model includes cylinder heat transfer to head, liner and piston, coolant pump, engine radiator, and thermostat. The charge air cooling system model includes the variable-speed coolant pump, low-temperature radiator mounted ahead of the engine radiator, and a blend valve to maintain CAC coolant temperature in a desired range at low ambient temperatures by blending a controlled amount of engine coolant. Combustion was not the focus of this particular simulation and was modelled simply with a wiebe function. Wiebe parameters were fit to typical combustion test data from the engine.

Figure 17. System model, GT-Suite Simulink co-simulation.

Simulation results are shown in Figure 19 for a tip-in tip-out transient starting from a low-speed, light-load cruise condition of 1000 rpm and 3 bar IMEP, to a moderate load condition of 2000 rpm and 10 bar IMEP. This moderate load condition is typical of the heaviest acceleration transients on the FTP test for this vehicle. The rate of IMEP rise during tip-in is 7 bar/sec.

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

Figure 18. Engine system model using GT-Suite.

Figure 19. Simulated transient response for the system.

The supercharger and turbocharger are used to control MAP during the transient. Boost control is very good with maximum deviation from target of about 4 kPa. The boost system is capable of significantly faster load transients than simulated in this case. The simulated MAP lags the desired by approximately 50 milliseconds (1st order time constant).

IAT is controlled by modulating low-temperature loop coolant flow rate. IAT is generally well controlled with a small positive excursion on tip-in and a small negative excursion on tip-out. When high rates of boost are applied, the air in the manifold between the CAC-2 outlet and the intake valves is compression heated during tip-in (or expanded during tip-out). There is no opportunity to cool (or heat) this air. In this simulation, the IAT maximum deviation from target is +/− 3 deg C. The system is able to maintain IAT to within 1.0 C of the desired value in 1.3 seconds.

Figure 20. Preliminary transient response on a real GDCI multi-cylinder engine.

EGR control produces a smooth EGR response with a lag of about 0.5 second between the EGR valve, where it is controlled, and the intake valves. The lag is due partially to gas transport and partially to mixing in the heat exchanger inlet and outlet headers. This is a relatively low lag and results from compact design of the EGR and intake systems. The EGR flow path includes an oxidation catalyst located directly at the turbocharger outlet, an EGR cooler, and a short pipe to the EGR valve. EGR is introduced close to the turbo-compressor inlet, where EGR gases are mixed with incoming fresh air. The flow must also pass through CAC-1, which is sized for good high load thermal control at elevated ambient temperatures. The EGR response lag will cause engine EGR to significantly deviate from targets during transients. Since GDCI combustion

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

depends on EGR, compensation using other operating variables during transients is used to maintain robust combustion.

Preliminary transient tests were performed on a real GDCI multi-cylinder engine and results are shown in Figure 20. This test demonstrates excellent boost system response for a very fast load transient (∼15 bar/sec). The pressure ratios across the supercharger and turbocharger are plotted separately, which shows that fast SC boost control can compensate for the slower TC response. This preliminary work demonstrates very fast response capability for precise boost control.

COLD START SIMULATION AND TESTA robust cold start with low emissions is another important requirement for GDCI multi-cylinder engines. The goal is to obtain robust starts without a spark plug to avoid the additional cost, system complexity and the controls difficulties associated with combustion mode switching. The GDCI cold start system includes intake air heating and leverages the pressure and heat produced by the supercharger.

Co-simulation of the engine system was performed to understand the effect of intake air heating and intake boost on in-cylinder gas pressure and temperature. Cold cranking was simulated for starter motor acceleration up to 200 rpm in one second. Initial temperatures of all engine components and gases were set equal to ambient temperature. The heater power delivered to the intake air was varied as a parameter, but held constant during each cranking run.

Figure 21 shows results for ambient temperatures of 25 C and 0 C. Intake air heat increases MAP significantly due to heat addition downstream of the charge air cooler. However, at cranking speeds, intake heating causes a one to two bar decrease in peak cylinder pressure due to increased in-cylinder heat transfer. Note that without heating, the intake air remains at ambient temperature, even though the supercharger delivers significant boost and temperature rise. This is because the charge air cooler absorbs nearly all the heat generated by the supercharger. However, as shown in Figure 21, heat addition of 100 and 200 watts per cylinder is sufficient to significantly increase both intake air temperature and peak cylinder gas temperature.

Figure 22 summarizes the effect of heat addition on peak cylinder temperature for the full range of simulated ambient temperatures (25 to −25 deg C) and heating rates (0 to 300 W). It is estimated from KIVA simulations that a peak gas temperature of 700 to 800 K is sufficient to achieve robust auto-ignition. These results suggest that room temperature (25 C) starts may require a small amount of heater power (∼50W). These results also suggest that intake air heating may enable cold starts at a low ambient temperature of −25 C if 175 W per cylinder can be delivered to the air.

Figure 21. Simulation results for cold cranking.

Figure 22. Effect of intake air heat on peak cylinder temperature.

Figure 23 shows measured engine speed and cylinder pressure for each cylinder during a preliminary room temperature cold start (∼ 20 C) on a real multi-cylinder GDCI

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

engine. Cranking speed of 300 rpm was quickly achieved and one motoring engine cycle occurs prior to the first fuel injection. A small amount of intake air heating enabled combustion on the first cycle, which is consistent with the simulations described above. Stable combustion is quickly achieved at the target idle speed of 800 rpm. This preliminary test demonstrates a good cold start at 20 C with good potential for robust cold starting at lower temperatures.

Figure 23. Preliminary cold start test at room temperature on a multicylinder GDCI engine.

SUMMARYBuilding on previous developments and extensive simulation work, a 1.8L four-cylinder GDCI engine was developed and tested using market gasoline (RON91) at Delphi. The engine uses a new low-temperature combustion process for gasoline partially-premixed compression ignition. Central to these advancements was a fuel injection system and injection strategy combined with a new piston design. Using multiple late injections and GDi-like fuel pressure, the fuel-air mixture could be stratified but sufficiently mixed. This produced very clean and efficient combustion as simulated using KIVA code.

Preliminary steady-state, calibration tests were performed. Test results showed that targets for NOx (0.2 g/kWh), smoke (0.1 FSN), combustion noise (load dependent), and stability (3% COV IMEP) were met. Injection parameters could be used at all operating conditions to control combustion phasing.

BSFC was measured and compared to competitive gasoline and diesel engines over a wide range of operating conditions. BSFC for GDCI was significantly better than advanced production SI engines, and comparable to the very efficient hybrid vehicle engines at their best efficiency conditions (214 g/kWh). Compared to new diesel engines, BSFC for GDCI at light loads was comparable or better, and at high loads was about 5 percent higher. At all loads, GDCI was remarkably clean with the potential for no aftertreatment for NOx and particulate emissions. Further BSFC improvements are expected through planned development work.

Preliminary transient simulations and tests were performed for warmed-up operating conditions. Very good transient response was observed in preliminary tests. The high boost rates of the supercharger could offset the slower response of the

turbocharger. Preliminary cold start simulations and tests were also performed using intake air heating. This work showed the potential for practical GDCI cold starts. More work is needed to optimize the engine calibration over the operating map and develop engine controls for the vehicle.

REFERENCES1. Kalghatgi, G., “Auto-Ignition Quality of Practical Fuels and

Implications for Fuel Requirements of Future SI and HCCI Engines,” SAE Technical Paper 2005-01-0239, 2005, doi:10.4271/2005-01-0239.

2. Johansson, B., “High-Load Partially Premixed Combustion in a Heavy-Duty Diesel Engine, Diesel Engine Emissions Reduction (DEER) Conference Presentations, 2005, Chicago, IL.

3. Risberg, P., Kalghatgi, G., Ångstrom, H., and Wåhlin, F., “Auto-Ignition Quality of Diesel-Like Fuels in HCCI Engines,” SAE Technical Paper 2005-01-2127, 2005, doi:10.4271/2005-01-2127.

4. Kalghatgi, G., Risberg, P., and Ångström, H., “Advantages of Fuels with High Resistance to Auto-ignition in Late-injection, Low-temperature, Compression Ignition Combustion,” SAE Technical Paper 2006-01-3385, 2006, doi:10.4271/2006-01-3385.

5. Kalghatgi, G., Risberg, P., and Ångström, H., “Partially Pre-Mixed Auto-Ignition of Gasoline to Attain Low Smoke and Low NOx at High Load in a Compression Ignition Engine and Comparison with a Diesel Fuel,” SAE Technical Paper 2007-01-0006, 2007, doi:10.4271/2007-01-0006.

6. Kalghatgi, G., “Low NOx and Low Smoke Operation of a Diesel Engine using Gasoline-Like Fuels”, ASME 2009 International Combustion Engine Division Spring Technical Conference, ICES2009-76034, 2009.

7. Manente, V., Johansson, B., and Tunestal, P., “Partially Premixed Combustion at High Load using Gasoline and Ethanol, a Comparison with Diesel,” SAE Technical Paper 2009-01-0944, 2009, doi:10.4271/2009-01-0944.

8. Manente, V., Johansson, B., and Tunestal, P., “Half Load Partially Premixed Combustion, PPC, with High Octane Number Fuels. Gasoline and Ethanol Compared with Diesel”, SIAT 2009 295, 2009.

9. Manente, V., Johansson, B., and Tunestal, P., “Characterization of Partially Premixed Combustion with Ethanol: EGR Sweeps, Low and Maximum Loads”, ASME International Combustion Engine Division 2009 Technical Conference, ICES2009-76165, 2009.

10. Manente, V., Tunestal, P., Johansson, B., and Cannella, W., “Effects of Ethanol and Different Type of Gasoline Fuels on Partially Premixed Combustion from Low to High Load,” SAE Technical Paper 2010-01-0871, 2010, doi:10.4271/2010-01-0871.

11. Weall, A. and Collings, N., “Gasoline Fuelled Partially Premixed Compression Ignition in a Light Duty Multi Cylinder Engine: A Study of Low Load and Low Speed Operation,” SAE Int. J. Engines 2(1):1574-1586, 2009, doi:10.4271/2009-01-1791.

12. Ra, Y., Yun, J.E., and Reitz, R., “Numerical Simulation of Diesel and Gasoline-fueled Compression Ignition Combustion with High-Pressure Late Direct Injection,” Int. J. Vehicle Design, Vol. 50, Nos. 1,2,3,4. pp.3-34, 2009.

13. Ra, Y., Yun, J.E., and Reitz, R., “Numerical Parametric Study of Diesel Engine Operation with Gasoline,” Comb. Sci. and Tech., 181:350-378, 2009.

14. Dempsey, A. and Reitz, R., “Computational Optimization of a Heavy-Duty Compression Ignition Engine Fueled with Conventional Gasoline,” SAE Int. J. Engines 4(1):338-359, 2011, doi:10.4271/2011-01-0356.

15. Ra, Y., Loeper, P., Reitz, R., Andrie, M. et al., “Study of High Speed Gasoline Direct Injection Compression Ignition (GDICI) Engine Operation in the LTC Regime,” SAE Int. J. Engines 4(1):1412-1430, 2011, doi:10.4271/2011-01-1182.

16. Ra, Y., Loeper, P., Andrie, M., Krieger, R. et al., “Gasoline DICI Engine Operation in the LTC Regime Using Triple- Pulse Injection,” SAE Int. J. Engines 5(3):1109-1132, 2012, doi:10.4271/2012-01-1131.

17. Ciatti, S., and Subramanian, S., “An Experimental Investigation of Low Octane Gasoline in Diesel Engines,” ASME International Combustion Engine Division 2010 Technical Conference, ICEF2010-35056, 2010.

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

18. Das Adhikary, B., Ra, Y., Reitz, R., and Ciatti, S., “Numerical Optimization of a Light-Duty Compression Ignition Engine Fuelled With Low-Octane Gasoline,” SAE Technical Paper 2012-01-1336, 2012, doi:10.4271/2012-01-1336.

19. Ciatti, S., Johnson, M., Das Adhikary, B., Reitz, R. et al., “Efficiency and Emissions Performance of Multizone Stratified Compression Ignition Using Different Octane Fuels,” SAE Technical Paper 2013-01-0263, 2013, doi:10.4271/2013-01-0263.

20. Das Adhikary, B., Reitz, R., and Ciatti, S., “Study of In-Cylinder Combustion and Multi-Cylinder Light Duty Compression Ignition Engine Performance Using Different RON Fuels at Light Load Conditions,” SAE Technical Paper 2013-01-0900, 2013, doi:10.4271/2013-01-0900.

21. Yang, H., Shuai, S., Wang, Z., and Wang, J., “High Efficiency and Low Pollutants Combustion: Gasoline Multiple Premixed Compression Ignition (MPCI),” SAE Technical Paper 2012-01-0382, 2012, doi:10.4271/2012-01-0382.

22. Won, H., Peters, N., Pitsch, H., Tait, N. et al., “Partially Premixed Combustion of Gasoline Type Fuels Using Larger Size Nozzle and Higher Compression Ratio in a Diesel Engine,” SAE Technical Paper 2013-01-2539, 2013, doi:10.4271/2013-01-2539.

23. Sellnau, M., Sinnamon, J., Hoyer, K., and Husted, H., “Gasoline Direct Injection Compression Ignition (GDCI) - Diesel-like Efficiency with Low CO2 Emissions,” SAE Int. J. Engines 4(1):2010-2022, 2011, doi:10.4271/2011-01-1386.

24. Sellnau, M., Sinnamon, J., Hoyer, K., and Husted, H., “Development of Full-Time Gasoline Direct-Injection Compression Ignition (GDCI) for High Efficiency and Low CO2, NOx and PM,” Aachen Colloquium 2011, 2011.

25. Sellnau, M., Sinnamon, J., Hoyer, K., and Husted, H., “Full-Time Gasoline Direct-Injection Compression Ignition (GDCI) for High Efficiency and Low NOx and PM,” SAE Int. J. Engines 5(2):300-314, 2012, doi:10.4271/2012-01-0384.

26. Hoyer, K., Sellnau, M., Sinnamon, J., and Husted, H., “Boost System Development for Gasoline Direct-Injection Compression-Ignition (GDCI),” SAE Int. J. Engines 6(2):815-826, 2013, doi:10.4271/2013-01-0928.

27. Sellnau, M., Sinnamon, J., Hoyer, K., Kim, J. et al., “Part-Load Operation of Gasoline Direct-Injection Compression Ignition (GDCI) Engine,” SAE Technical Paper 2013-01-0272, 2013, doi:10.4271/2013-01-0272.

28. Chang, J., Kalghatgi, G., Amer, A., Adomeit, P. et al., “Vehicle Demonstration of Naphtha Fuel Achieving Both High Efficiency and Drivability with EURO6 Engine-Out NOx Emission,” SAE Int. J. Engines 6(1):101-119, 2013, doi:10.4271/2013-01-0267.

29. Chang, J., Viollet, Y., Amer, A., and Kalghatgi, G., “Fuel Economy Potential of Partially Premixed Compression Ignition (PPCI) Combustion with Naphtha Fuel,” SAE Technical Paper 2013-01-2701, 2013, doi:10.4271/2013-01-2701.

30. MODEL 379020 ROTATING DISK THERMODILUTER, Operation Manual, TSI Incorporated, June, 2009.

31. MODEL 3090 ENGINE EXHAUST PARTICLE SIZE SPECTROMETER, Operation Manual, TSI Incorporated, October, 2010.

32. AVL List, “Product Description - FlexIFEM Indi,” http://www.avl.com/c/document_library, Mar. 2013.

33. Labview RT Software, Release 2009, National Instruments, Austin TX, 2009.

34. Hadler, J., Rudolph, F., Dorenkamp, R., Stehr, H., et al., “Volkswagen's New 2.0I TDI Engine Fulfills the Most Stringent Emissions Standards,” Vienna Symposium 2008.

35. Murata, Y., Tajiri, K., Sasaki, Y., Fukushima, H., et al., “Development of the New 2.2I Fuel-Efficient Diesel Engine for the Honda Civic,” Aachen Colloquium 2012.

36. Yonekawa, A., Ueno, M., Watanabe, O., and Ishikawa, N., “Development of New Gasoline Engine for ACCORD Plug-in Hybrid,” SAE Technical Paper 2013-01-1738, 2013, doi:10.4271/2013-01-1738.

37. Adachi, S., and Hagihara, H., “The Renewed 4-Cylinder Engine Series for Toyota Hybrid System,” Vienna Symposium 2012.

38. Schopp, J., Kiesgen, G., Kilias, H., Lechner, M., et al., “The New 1.6-Litre Turbocharged Engines with Direct Injection and Fully Variable Valve Gear for the New BMW 1 Series Car,” Aachen Colloquium 2011.

39. Vent, G., Enderle, C., Merdes, N., Kreitmann, F., and Weller, R., “The New 2.0I Turbo Engine from the Mercedes-Benz 4-Cylinder Engine Family,” Aachen 2012.

40. GT Power Software, Version 7, Gamma Technologies, Inc., Westmont, IL, 2010.

41. Matlab Simulink Software, Mathworks, Inc., Natick, MA, 2013.

CONTACT INFORMATIONMark SellnauEngineering ManagerDelphi Advanced Powertrain3000 University DriveAuburn Hills, MI 48326mark.sellnau@delphi.com

ACKNOWLEDGMENTSThis work was supported by the US Dept of Energy, Office of Vehicle Technology, and administered by Gurpreet Singh under contract DOE DE-EE0003258.

The authors gratefully acknowledge collaboration with and support from Hyundai Motor Company. The authors also appreciate contributions to this work from engineers and technicians at Delphi Powertrain, Delphi Diesel Systems; Delphi Thermal Systems, Delphi Electronics and Safety, and Delphi Korean Automotive Components. At Wisconsin Engine Research Consultants (WERC), simulation work by Professor Rolf Reitz, Professor Chris Rutland, Harmit Juneja, Chris Wright, and Mike Bergin is gratefully acknowledged. At the Engine Research Center (ERC) at University of Wisconsin-Madison, contributions by Professor Jaal Ghandhi and Dr. Jason Oakley are gratefully acknowledged.

DEFINITIONS/ABBREVIATIONSatdc - After Top Dead Center

B - Bore

BDC - Bottom Dead Center

BMEP - Brake Mean Effective Pressure

BSFC - Brake Specific Fuel Consumption

C - Centigrade

CAC - Charge Air Cooler

CAN - Controller Area Network

CE - Combustion efficiency

CFD - Computational Fluid Dynamics

CNL - Combustion Noise Level

CO - Carbon monoxide emissions

COVIMEP - Coefficient of Variation of IMEP

DoE - Design of Experiments

E00 - Gasoline without Ethanol

E10 - Gasoline with 10% Ethanol

EEPS - Eng. Exh Particle Size Spectra

EGR - Exhaust gas recirculation

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)

FIE - Fuel injection equipment

FSN - Filtered Smoke Number

GCR - Geometric Compression Ratio

GDi - Gasoline Direct Injection

GDCI - Gasoline Direct Inj. Comp. Ignition

HC - Hydrocarbon emissions

HCCI - Homo. Charge Comp. Ignition

HRR - Heat Release Rate

IAT - Intake Air Temperature

IMEP - Indicated Mean Effective Pressure

ISCO - Indicated specific carbon monoxide

ISFC - Indicated specific fuel consumption

ISHC - Indicated specific hydrocarbon

ISNOx - Indicated specific nitrous oxide

KIVA - Comput. Fluid Dyn. Code developed at Los Alamos National Labs

I - Connecting Rod LengthLTC - Low Temperature Combustion

LHV - Lower Heating Value

MAP - Manifold Absolute Pressure

M-FMEP - Motoring Friction Mean Eff. Pres.

MLI - Multiple Late Injection

NMEP - Net Mean Effective Pressure

NOx - Oxides of Nitrogen

PCP - Peak Cylinder Pressure

PHI - Equivalence Ratio

PID - Proport., Integ., Deriv. Controller

Pinj - Injection Pressure

PM - Particulate Matter

PPCI - Partially Premixed Comp. Ignition

Q - Quantity injected

r - Crank throw length

RB - Rebreath

RON - Research Octane Number

S - Stroke

SGDI - Stratified Gasoline Direct Injection

SOC - Start of Combustion

SOI - Start of Injection

SC - Supercharging

TC - Turbocharging

TDC - Top dead center

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE International.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE International. The author is solely responsible for the content of the paper.

Sellnau et al / SAE Int. J. Engines / Volume 7, Issue 2 (July 2014)