Small Gasoline Direct Fuel Injection Two-Stroke Engines for Range Extender Applications

Small Gasoline Direct Fuel Injection Two-Stroke Engines for Range Extender Applications Pierre Duret 1, Stéphane Venturi 2, Prakash Dewangan 1 1: IFP School, Rueil-Malmaison, France 2: IFP Energies Nouvelles, Rueil-Malmaison, France Abstract: The main purpose of this paper will be to discuss various possibilities of using a small gasoline direct injected two-stroke engine as a range extender (REX) for electric vehicles (EV). For such REX applications, the main requested specifications are: low noise and vibrations (NVH), compactness, lightweight, minimum production cost and efficiency. It is considered that meeting the emissions standards will be in any case compulsory. Considering these specifications, this paper explores how the gasoline DI 2-stroke engine, thanks to its advantages resulting from its double combustion cycle frequency, can be a good candidate compared to other possible engine technologies. We will illustrate that through three examples of range extender applications based on DI 2-stroke engines: a small displacement 2-cylinder scooter based engine as a safety device for an EV city car. a modified 2-cylinder snowmobile engine as a REX (series hybrid) for a small multi-usage lightweight electric car, a 3-cylinder marine outboard engine as REX for a multi-usage high performance EV sport car, Various design and simulation studies have been performed with these three configurations and will be presented. The results will show that emissions standards can be met when gasoline direct injection and ultra-low NOx Controlled Auto-Ignition (CAI) are combined together with an oxidation catalyst for aftertreatment. Beside the achievement of the Euro 6 NOx target, remarkably low level of average CO2 emissions can be achieved with impressively increased vehicle range (compared to the pure EV range) with only a few litres of gasoline. Keywords: Two-stroke engines, direct injection, CAI Controlled Auto-Ignition, range extender, electric vehicle 1. Introduction Electric vehicles (EV) can be seen as a way to mitigate the GHG emissions according to the solution used for the production of the required electricity. Nevertheless, it is generally considered that, due to its drawbacks mainly linked to the electric energy storage system (heavy, bulky and expensive battery), the purely electric vehicle will be limited in short-medium term for some specific applications. Even if significant progresses can be expected in the battery technology, during the transition, the solution to increase the chance of acceptance of EV by the public in a large scale could be to keep a limited pure EV range (with therefore minimum battery cost) corresponding to most of the urban usages and to equip the vehicle with a lightweight range extender. This range extender could allow to multiply by several times the pure EV range without sacrifying the global CO2 emissions. It is interesting to remind ourselves that Citröen presented in 1998 at the Paris Auto Show an electric vehicle (based on a Saxo Citröen model) and equipped with a small direct injected gasoline 2- stroke engine as range extender. This innovative vehicle (vehicle mass 1050 kg; max speed 120 km/h) was presented with a pure EV range of 80 km and an extended range up to 340 km. The auxiliary power unit used was a prototype DI 2-stroke engine technology, 2 cylinder opposite 200 cc, delivering a power of 6,5 kw and directly coupled with a starter generator. The auxiliary power unit (thermal engine + starter generator) was remarkably packaged with overall dimensions of vol. 30x30x25 cm & a mass of 20 kg. With such small size, it was possible to implement this auxiliary unit under the rear seat of the Saxo car. A schematic view of the whole powertrain of the car is presented in Figure 1. Figure 1: The Citroën Saxo Dynavolt: an electric vehicle concept presented in 1998 with a small DI 2- stroke engine as range extender [1] But this very interesting project was not further investigated after the 1998 Paris Auto Show for two main reasons: firstly, it was not the right period for electric vehicles (too much in advance) and 2-stroke gasoline DI technology was not yet mature. In parallel, during the last decades the DI 2-stroke technology has been further developed outside automotive and successfully applied in production for marine outboards and 2-3 wheelers engines [2-4]. Page 1/9

Taking into account these two considerations (results achieved in the 90's with a DI 2-stroke with a non-mature technology & more recent availability of well proven DI 2-stroke technology outside automotive) it seems natural to wonder if for the range extender application, a small DI 2-stroke engine could be a relevant candidate for such application [5,6]. For this purpose, three different case studies of range extender versus vehicle applications undertaken by several teams of IFP School students are examined in this paper and their potential results presented. 2. EV city car with 15 kw range extender The first project presented in this paper is a small displacement 2-cylinder scooter based engine that could be designed and optimized as a safety device for a typical EV city car. 2.1 Requested specifications for the thermal engine For this project, we chose to study a 2-cylinder 250 cc DI 2-stroke that could be based on the use of existing 125 cc Honda Pantheon 2-stroke engine production components (piston, rod, CAI-AR valve, ). The reason of the choice of this engine is that it is already designed to operate in Controlled Auto-Ignition (CAI) also named Activated Radicals (AR) combustion [7-12]. The other following specifications were fixed at the beginning of the project: maximum power of 15 kw at moderate engine speed for noise control Range extender engine operating point in CAI for NOx & PM control Euro 6 emissions standards Minimum packaging in 160 liters: the idea was to try to integrate all the REX components (engine with intake, exhaust, fuel circuit, cooling system, generator, fuel tank, ) into a kind of box that could be considered as a REX kit or module. The choice of a packaging within 160 liters was chosen in reference to a Wankel based benchmark [13] 2.2 Engine architecture for REX application Several possible 2-cylinder designs were considered: In-line, combustion phasing @ 180 deg In-line, combustion phasing @ 0 deg In-line, combustion phasing @ 90 deg V2 90 Opposed cylinders (boxer) Opposed pistons Figure 2: illustration of the opposed piston configuration Each 2-cylinder engine configuration has then been rated according to the most important selection criteria considered for a range extender application [6,14]: NVH / balancing: with balancing shaft for in-line configurations NVH / torque fluctuation: combustion frequency and phasing Packaging / volume: intake & exhaust systems, accessories, Packaging / weight: engine, exhaust system, balancing shaft, Cost: complexity of manufacturing of the engine and components (exhaust system, balancing shaft, belts & accessories, ) Efficiency: scavenging process, exhaust tuning, The results of this rating are summarized in the Table 1. From this analysis, the Opposed pistons configuration seems the most promising followed by the 2-cylinder in-line with combustion phasing every 180 deg. CA. Table 1: comparison of various engine architecture versus REX application criteria 2.3 Energy management optimization The purpose of this sub-section is to study the optimum energy management strategy for the range extender application. The main tasks undertaken are as follows: to determine the optimized REX operating point necessary to undertake the NEDC with maintained battery SOC, taking into account efficiency of the whole chain from the requested power at the wheels to the corresponding power of the REX thermal engine; Page 2/9

to estimate (based on the extensive IFP experimental DI CAI 2-stroke engine data base build during the last 25 years of experience [15-18]) the pollutant emissions level in comparison with Euro 6 legislation and the fuel consumption and corresponding EV range extension. This energy management optimisation has been simulated with 4 examples of vehicle application. The weight and estimated SCx of the 4 selected vehicles are shown in the Table 2. Figure 3: revised NEDC driving cycle with 90 km/h top speed Table 2: example of vehicle applications studied We considered a range extender giving a maximum power output of 15 kw (which is a quite reasonable power for a 250 cc 2-stroke engine that can be achieved at a moderate engine speed of about 4500-5000 rpm for minimum REX noise). Taking into account the various efficiencies from the REX crankshaft to the wheels, we can estimate that it should correspond to a power at the wheels of about 11,33 kw. From this value and taking into account the vehicle characteristics in Table 2, it is possible to estimate the top speed of each vehicle in REX mode and for different road slopes. The results are summarized in the Table 3 here below. Table 3: maximum vehicle top speed in REX mode with 0, 2 and 4% of slope The type of REX considered in this study are more like a safety device allowing to be able to continue to drive only when the battery charge becomes too low. Therefore, since it will be an exceptional mode of operation, it is possible to accept some reduced vehicle performance. During REX operating mode, it was then chosen to limit the vehicle speed to a maximum of 90 km/h for all the 4 vehicles. In such case the NEDC Cycle used for the emissions calculation is also limited to 90 km/h as shown in Figure 3. It is then possible to calculate the average power at the REX crankshaft that is necessary for each vehicle to perform this revised NEDC cycle. This calculation includes the various efficiencies and also the regenerative braking during deceleration. In such case, if the REX engine runs during the whole NEDC at the same average power of the Table 4, it will allow to maintain the same battery state of charge between the beginning and the end of the cycle. Table 4: average power of the REX for maintained battery state of charge during the revised NEDC This required average power can be achieved at different engine speeds. The choice of the engine speed results from a trade-off between NOx emissions which decrease when the engine speed increases (because the engine load decreases) and the engine noise which increases when the speed increases. In fact the optimum operating point corresponds to the minimum engine speed for Euro 6 NOx emissions. 4000 rpm seems in this case a good compromise. At this engine speed, from our DI CAI 2-stroke data base, the REX should be able to meet Euro 6 NOx limit up to a power of 2,97 kw (corresponding to 1,78 bar BMEP) without DeNOx. This power is slightly above the power required for the 4 vehicles. If we now look to the other pollutant emissions (see Table 5), we see that without aftertreatment, the CO level should be just above the Euro 6 limit and the HC level about 6 times above. From previous experience, we can be confident that, with such raw CO and HC emissions levels, Euro 6 limits can be achieved with an appropriate oxidation catalyst design (location, dimensions, cells density, precious metal formulation) and an appropriate fast oxi-cat lighting strategy. Table 5: estimated emissions of the REX (2,97 kw @ 4000 rpm) Page 3/9

2.4 In-vehicle integration The following Table 6 summarizes the main performance results achieved with this case study of a 15 kw range extender for EV city car. Lightweight 3-seat mid-size vehicle Plug-in hybrid 25 km full electric range Acceleration: 0-100 km/h < 10s Top speed: 185 km/h Average fuel consumption target: 2 l/100 km [19] Table 6: summary of the simulated main performance results of the 15 kw REX Finally, about the initial 160 litres packaging target, a CATIA predesign study has been undertaken based on the choice of the opposite pistons engine configuration. Figure 5 : the Segula Hagora vehicle project 3.2 Thermal engine specifications In this study, the thermal engine will develop about 30-35 kw at moderate engine speed. Its main specifications are as follows: based on a modified 2-cylinder snowmobile production ROTAX 600 HO engine high power / weight ratio longest stroke available engine already equipped with direct injection (E-Tec) modified for high torque at low speed and limited maximum speed Euro 6 emissions capability (exhaust valve for ultra low NOx CAI) Figure 4: CATIA pre-design study of the REX packaging in a 160 liters volume The Figure 4 shows that all the following components of the complete range extender kit can be incorporated in such a volume: DI/CAI 250cc 2-stroke with intake & exhaust generator and its coupling with the engine cooling system with radiator and fan DC/DC convertor fuel tank (10 liters) Figure 6: the production ROTAX 600 HO engine The following Table 7 show the new targets for reduced power & speed of the modified engine in order to achieve Euro 6 emissions limits. 3. Lightweight urban sport plug-in hybrid 3.1 Vehicle specifications (Segula Hagora project): In this second case study of range extender application, the selected vehicle is a urban sport plug-in hybrid vehicle concept developed by Segula and named the Hagora project. Its main specifications are: Table 7: new performance targets of the modified engine Page 4/9

3.3 GT Power modeling of modified configuration versus performance targets During this project, GT Power has been first calibrated on the production engine. In a second step, it has been used to select the best transfer and exhaust ports timings for the new engine performance targets. The results are summarized in the Table 8. Figure 8: selected hybrid architecture Table 8: compared port timings & exhaust tuning speed between production versus modified engine The motivations for such engine modifications were: to reduce the maximum engine speed to increase the low speed torque to decrease the raw emissions (HC) to improve the ability to CAI combustion in order to finally reach the efficiency & emissions of IFP experimental DI CAI 2-stroke data base (see Figure 7). 3.5 Simulink energy management optimization Hybrid strategy An online Equivalent Consumption Minimization Strategy (ECMS) in Simulink for energy optimization has been firstly studied. The following Figures 9 & 10 show the selected engine operating points resulting from this minimum NOx optimisation strategy on NEDC cycle under charge sustaining constraints. Figure 9: DI 2-stroke engine operating points along the iso-bsfc map in hybrid mode Figure 7: Iso-BSFC map used for DI CAI 2-stroke engine 3.4 Selected Plug-in Hybrid architecture In this project, a specific innovative plug-in hybrid architecture has been chosen as illustrated in Figure 8. This particular architecture allows a lot of possibility in terms of the optimization of the energy management of the whole powertrain. In a first step of this study, the possibility to use the DI 2-stroke engine in the hybrid mode has been investigated. Figure 10: DI 2-stroke engine operating points along the iso-nox emissions map in hybrid mode Page 5/9

Figure 11: optimized hybrid mode strategy for best fuel consumption versus NOx trade-off In the Figures 11 here above, it is possible to see as a function of the time corresponding to the NEDC cycle: the vehicle speed in km/h along the NEDC the instantaneous DI 2-stroke engine torque the instantaneous electric motor torque the evolution of the battery SOC during the cycle understandable when looking to the Figure 10 which shows several operating points above the low load ultra-low NOX CAI region CO & HC emissions remain compatible to be decreased by an oxidation catalyst. 3.5 Simulink energy management optimization Range extender strategy Therefore after this first step showing that a DI 2- stroke engine is not adapted for an hybrid strategy without DeNOx, in a second step the range extender strategy is now investigated. Here are the main characteristics of this range extender strategy: the CVT is tuned for lowest NOx emissions the DI 2-stroke supplies constant power all the time o with an average power allowing to cover o the NEDC cycle with ICE only with a low load working point in CAI for ultra-low NOx (Figure 13) even if BSFC is not the best (Figure 12) the battery is used as a buffer the E-motor is also used as a generator when the ICE power is superior to the vehicle demand the system is globally charge sustaining o o the batteries slowly charged during urban part of NEDC (Figures 14) there is a battery depletion during the extra-urban part of NEDC (Figures 14) It can be observed from Figure 10 that the strategy succeeds to place most of the operating points at low NOx regions however being on-line optimization it suffers some full load operations as can be seen in the red circle on Figures 10 and 11. These high loadhigh NOx points could be avoided in an offline optimization which would be less relevant for real application. Table 9: fuel consumption and emissions results for the hybrid strategy The global efficiency and emissions results presented in the Table 9 show that with this optimized hybrid mode strategy: very good fuel consumption can be achieved unfortunately NOx emissions are too high, about 4 times the Euro 6 level, which then cannot be achievable without DeNOx. This can be easily Figure 12: DI 2-stroke operating points along the iso- BSFC map in range extender mode Page 6/9

The results achieved with this range extender strategy are very interesting. They show again that a DI 2-stroke engine when it is used as a range extender has the potential to reach the Euro 6 NOx emissions limit without aftertreatment, and with significant margin in this example. The fuel consumption is obviously slightly increased compared to the efficiency optimized hybrid strategy but the average fuel consumption still remains below the 2 l/100 km target. In addition, CO and HC remain at a level compatible for their conversion by an oxidation catalyst. 4. Multi-usage high performance luxury electric sport car Figure13: DI 2-stroke operating points along the iso- NOx emissions map in range extender mode 4.1 Electric vehicle specifications (Exagon Furtive) In the third and last case study of range extender application in this paper, we will describe briefly another project of using a modified marine outboard DI 2-stroke engine as range extender of a high performance multi usage luxury electric sport car. The vehicle chosen for this study is the Exagon Furtive (Figure 15). Figure 15 : the Exagon Furtive luxury EV sport car Figure 14: optimized range extender mode strategy for minimum NOx emissions Table 10: fuel consumption and emissions results for the range extender strategy Its main specifications are : Weight: 1950 kg SCx: 0,81 m2 Electric motors: 250 kw Electric range: o 300 km on NEDC o 197 km @ 130 km/h 4.2 Enhanced specifications with Range extender The targets of the project are to study a range extender application that could enhance the vehicle specifications as follows: Maximum vehicle speed in REX mode: 140 km/h + 300 km range on NEDC + 247 km range @ 130 km/h with 30 l fuel tank Page 7/9

4.3 Choice of the engine architecture adapted for REX application One of the major target of this project was to use an existing on the shelf engine as a range extender to minimize the development cost of a new engine. In this study, we investigated the possibility of using a DI 2-stroke engine based on a production marine outboard engine. Its main specifications are in the Table 11 and Figure 16 here after. Table 12: selected REX operating point and NOx emissions compliance The results in Table 12 show that here again, the Euro 6 NOx emissions limit can be met without DeNOx after treatment and with regulated vehicle CO2 emissions of 18 g/km. Table 11: main specifications of the marine outboard DI 2-stroke engine investigated as range extender Figure 16: the 3-cilinder Selva marine outboard DI 2-stroke [3] The required engine modifications and adaptations for this range extender applications are limited to what is strictly necessary: no change of the thermodynamics/scavenging characteristics use of existing production DI system (IAPAC compressed air assisted fuel injection) implementation of exhaust throttling valves for CAI combustion new exhaust system to be designed including oxidation catalyst. 4.4 Energy management optimization for minimum NOx in range extender mode Similarly to the two previous case studies, we will keep the main energy management optimization with: a single operating point in CAI combustion for ultra-low NOx the same battery SOC at the beginning and at the end of the NEDC cycle 5. Conclusions From the results presented in this paper, we have been able to draw the following conclusions: To be able to propose electric vehicles equipped with a range extender will help to develop the EV customer acceptance Small engine technologies are well adapted for such application and will therefore bring their contribution for future Sustainable Mobility Among those technologies, the DI 2-stroke engine represents a relevant candidate thanks to its advantages of lightweight, compactness, double cycle frequency and NVH, cost and efficiency One of the major challenge of the use of DI 2- stroke engines for automotive range extender application is the NOx & PM emissions issue Energy management optimization of 3 examples of DI 2-stroke REX / EV combination shows that, when combined with part load Controlled Auto Ignition, Euro 6 NOx emissions can be achieved without DeNOx aftertreatment. 6. Acknowledgement The authors would like to particularly thank the following contributors to this work: IFP School students project teams: Alexandre BORIE, Thomas BRICHARD, Thomas CREMILLEUX, Samuel QUESADA Félix GALLIENNE, Sébastien LEMOINE, Adrian MIGUEL SANCHEZ Ivo LANIAR, Thomas LE BIHAN, Quentin PIRAUD Estelle GRILLIERES, Vijaykannan MOHAN, Vincent ROVERE, Michel SANCHO The support from industry Lorenzo SERRAO DANA Benoit BAGUR Exagon Motors Pascal GIRARD Segula Technologies Page 8/9

7. References [1] Stan, C. and Personnaz, J.: "Car Hybrid Propulsion Strategy Using an Ultra-Light GDI Two-Stroke Engine, SAE paper 1999-01-2940 [2] Bell G., Brewster S. and Ahern S., 'Beyond 3 Star Emission Capability for Outboard Engines', SAE Paper 2007-32-0052, 2007 [3] Venturi S., Lavy J., Duret P., Selva M. and Tonta G., 'From Development to Industrialization of an IAPAC Marine Outboard DI 2-Stroke Engine', SETC Conference, Pisa Italy, 2001 [4] Strauss S., Zeng Y. and Montgomery D., Optimization of the E-TEC Combustion System for Direct-Injected Two-Stroke Engines Toward 3-Star Emissions, SAE paper 2003-32-0007 [5] Duret P., Range extender for electric vehicles and powertrain for ultra-low cost vehicles: a chance for small gasoline DI 2-stroke engines?, SIA Spark Ignition Conference, Strasbourg, France 2009 [6] Duret P., The Small Gasoline DI 2-Stroke Engine: an Adapted Range Extender for Electric Vehicles?, Keynote paper, SIAT 2011 Symposium, Pune India [7] Ishibashi Y and Tsushima Y, 'A Trial for Stabilizing Combustion in Two-Stroke Engines at Part Throttle Operation, in Duret P, A New Generation of Two- Stroke Engines for the Future?, IFP International Seminar, Rueil-Malmaison, Editions Technip, 1993 [8] Onishi S, Souk Hong Jo, Shoda K, Pan Do Jo and Kato S, "Active Thermo-Atmosphere Combustion (ATAC) A New Combustion Process for Internal Combustion Engines", SAE Paper 790501 [9] Duret P., "Two-stroke CAI engines" in Zhao H., "HCCI and CAI engines for the Automotive Industry", Woodhead Publishing Limited, 2007 [10] Ishibashi Y, Nishida K and Asai M, 'Activated Radical Combustion in High Speed High Power Pneumatic Direct Injection Two Stroke Engine', in Duret P, A New Generation of Engine Combustion Processes for the Future?, IFP International Seminar, Rueil-Malmaison, France, Editions Technip, 2001 [11] Tsuchiya K, Hirano S, Okamura M, and Gotoh T, Emission Control of Two-Stroke Motorcycle Engines by the Butterfly Exhaust Valve, SAE paper 800973 [12] Duret P. and Venturi S., "Automotive Calibration of the IAPAC Fluid Dynamically Controlled Two- Stroke Combustion Process" SAE Paper 960363, [13] Beste F., Fraidl G., Benda V. and Atzwanger M., Range Extender EV, Encyclopedia of Automotive Engineering, 2014 [14] Trattner A., Pertl P. Schmidt St.P. and Sato T., Novel range extender concepts for 2025 with regard to small engine technologies, SETC 2011, SAE paper 2011-32-0596 [15] Duret P., Ecomard A., Audinet M., "A New Two- Stroke Engine with Compressed Air Assisted Fuel Injection for High Efficiency Low Emissions Applications", SAE Paper 880176, 1988 [16] Duret P., Moreau J-F., "Reduction of Pollutant Emissions of the IAPAC Two-Stroke Engine with Compressed Air Assisted Fuel Injection", SAE Paper 900801 [17] Duret P., Venturi S. and Carey C., "The IAPAC Fluid Dynamically Controlled Automotive Two- Stroke Combustion Process" in Duret P, A New Generation of Two-Stroke Engines for the Future? Rueil-Malmaison, France, Editions Technip 1993 [18] Duret P. et al. "The Air Assisted Direct Injection ELEVATE Automotive Engine Combustion System", SAE Paper 2000-01-1899 [19] Journal officiel de l'union européenne Règlement N 83 8. Contact Pierre DURET: pierre.duret@ifpen.fr Page 9/9