19. Aachener Kolloquium Fahrzeug- und Motorentechnik

Transcription

1 19. Aachener Kolloquium Fahrzeug- und Motorentechnik Studie zur Partikelfilterapplikation an einem modernen homogenen turboaufgeladenen 2L DI-Ottomotor A Study about Particle Filter Application on a State-of-the-Art Homogeneous Turbocharged 2L DI Gasoline Engine Dr.-Ing. Ingo Mikulic, Hein Koelman Dow Automotive Systems, Schwalbach im Taunus, Germany Steve Majkowski, Paul Vosejpka Dow Automotive Systems, Auburn Hills, USA Summary The application of short, uncoated DOW AERIFY TM particulate filters solves the particle number emission issue of a state-of-the-art gasoline light passenger car versus the threshold of 6x10 11 #/km, which is in place for diesel light-duty passenger cars. It is a plug-and-play solution using AERIFY gasoline particulate filters (GPF) which are optimized for direct injection (DI) gasoline engine operation conditions. No modification of the original TWC system was necessary. No negative impacts on the gaseous tailpipe emissions or fuel consumption were detected under NEDC test conditions. In on-road testing, no soot accumulation was found over extended standardized low load, city, and mixed driving conditions. A fuel consumption penalty of 2-3% was measured at vehicle speeds of 180 km/h and higher comparing the results for the exhaust system with and without a filter. 1 Introduction The reduction of fuel consumption and CO 2 emissions is an important goal for current and future development in the automotive powertrain industry. In recent years the usage of direct injection, turbo-charging, and downsizing has become a powerful approach to accomplish these goals for state-of-the-art gasoline engines. The fuel consumption of gasoline light passenger cars is significantly reduced, with attractive torque performance and driveability fully delivered [1-2]. The upcoming Euro 6 type approval legislation requires the introduction of a particle number limit (PN) for light-passenger cars in Europe. The European legislators may

2 2 19. Aachener Kolloquium Fahrzeug- und Motorentechnik 2010 set a value of 6x10 11 #/km as an appropriate threshold in the type approval test for gasoline engines (NEDC). It is the limit which has already been defined in Euro 5 and Euro 6 legislation for diesel light passenger vehicles. Recent measurements of tailpipe particle number emissions show that state-of-theart gasoline light passenger engines do not comply with a limitation of 6x10 11 #/km [3-4]. To provide compliance, two different approaches can be taken: Solution via engine-in means Solution via exhaust aftertreatment Important items like cost, application effort, system complexity and controllability will be key arguments for the OEM decision to pursue one solution over the other. In this study, the usage of mullite honeycomb particulate filters was investigated to solve the particle number tailpipe issue of a state-of-the-art gasoline light passenger car. Important performance items like particle number filtration, pressure drop, and fuel penalties due to the addition of a filter have been studied to highlight the pros and cons of a gasoline particulate filter (GPF) application. 2 Exhaust Gas Configurations with GPF There are various alternatives for the integration of a GPF into the exhaust system of a DI gasoline light passenger engine. Figure 1 illustrates some solutions with respect to the close-coupled TWC system. Actual baseline concepts contain a three-way catalyst (TWC) converter with one or two bricks delivering the required TWC functionality for emission compliance. GPF concept 1 describes the introduction of a bare GPF. It can be packaged in its own converter can in the underfloor position where sizing in length is relatively open. The drawback of this position is that there is significant thermal loss between the TWC and the underfloor GPF. It is questionable whether a continuous clean-out of accumulating soot will occur over all reasonable driving conditions due to this exhaust heat loss between the TWC and the underfloor GPF. A close-coupled location of the GPF, ideally in the same can and placed right behind the TWC system, would provide a continuous soot clean-out (e.g. plug-and-play system). Here, the limiting factor is the packaging constraint. The TWC system needs to be downsized to enable the integration of a close-coupled bare GPF. And even in that case, the length of the filter will be limited due to the typical lack of space in this location. A shorter filter is needed for a close-coupled bare GPF compared to underfloor packaging location. GPF concept 2 describes the introduction of a coated filter. It is a more complicated development approach, since it requires a higher level of optimization between

3 19. Aachener Kolloquium Fahrzeug- und Motorentechnik coating and substrate technology as well as system control. Actual concepts to implement coated GPF target the replacement of the second TWC brick that typically controls tailpipe NO x emissions by a coated GPF. The coated GPF has to accomplish the same catalytic performance goals as the second TWC brick of the original TWC system. Baseline Close-coupled Concept of Actual Exhaust Systems (No GPF) Engine TWC TWC Close-coupled GPF Concept 1: Bare GPF Engine TWC TWC Close-coupled GPF Underfloor - GPF underfloor - No modification of TWC concept - Continuous soot clean-out? - Plug-and-Play? Engine TWC TWC GPF Close-coupled - TWC system must shrink - GPF length limited in TWC canning - Continuous soot clean out - Plug-and-Play GPF Concept 2: Integration of TWC coating in GPF Engine TWC TWC/GPF Close-coupled - All bricks in same can - Continuous soot clean out - TWC functionality? - Higher complexity Figure 1: Alternative solutions of GPF implementation with respect to the closecoupled TWC system 3 DOW AERIFY Particulate Filters AERIFY particulate filters provide extremely low back pressure for higher engine performance while reducing the fuel consumption and CO 2 penalty in all the shown engine configurations. AERIFY filters feature an advanced, highly porous, chemically treated ceramic honeycomb with a unique microstructure that provides high-filtration performance. The technology is a robust, lightweight solution that enables improved engine performance and total system cost reductions compared to other substrate materials. The key product benefits are: High-efficiency particle number filtration, even without a soot cake Low back pressure Robust strength

4 4 19. Aachener Kolloquium Fahrzeug- und Motorentechnik 2010 Ability to downsize for smaller packaging without compromising performance Reduced fuel consumption penalty Lower carbon dioxide emissions Regulation compliance at lower cost High coatability Superior chemical / ash resistance The AERIFY material type is an advanced ceramic called Acicular Mullite [5] and can be used for diesel and gasoline applications. 4 Vehicle Test Setup and Procedures In this study, the usage of AERIFY GPF with different honeycomb structures and lengths was investigated to solve the tailpipe particle number issue of a state-of-theart gasoline light passenger car. For purposes of testing, an Audi A5 Coupe 2L TFSI light-duty passenger gasoline car was purchased. The vehicle was Euro 5 compliant. Audi A5 Coupe 2L TFSI MY 2007 Combustion system Displacement Exhaust system Turbocharged DI Gasoline engine with lateral injector position L Close-coupled TWC Emission compliance Euro 5 Maximum speed Mileage at test begin 226 km/h Approx.1500 km Table 1: Representative data of the test vehicle For the purpose of GPF testing, a replaceable filter can was used approx. 90 cm downstream of the close-coupled TWC system of the test vehicle. The replaceable canning allows the integration of different 4.66 inch diameter filter samples with a

5 19. Aachener Kolloquium Fahrzeug- und Motorentechnik maximum length of 5 inches. Additional sampling ports have been applied for measurement of exhaust gas composition and temperatures, absolute pressure, A/F ratio upstream of the TWC, pressure drop over GPF, and tailpipe particle mass and number emissions (see Figure 2). Upstream TWC Downstream TWC Upstream GPF Downstream GPF Ex. gas temperature Ex. gas pressure Ex. gas composition Rel. air/fuel ratio Ex. gas temperature Ex. gas composition TWC Ex. gas temperature Ex. gas pressure GPF Ex. gas temperature / pressure Ex. gas composition Particle number Particle mass Delta pressure R R/2 Engine values (OBD scan tool) GPF temperature measurement upstream and downstream Figure 2: Principal measurement set-up The following AERIFY particulate filters have been exposed to vehicle testing. Diameter, inch Length, inch Honeycomb structure DPF DPF DPF GPF GPF GPF GPF Table 2: Test samples

6 6 19. Aachener Kolloquium Fahrzeug- und Motorentechnik 2010 AERIFY samples indicated as DPF are designed for use in LDV diesel applications where a significant soot accumulation takes place. AERIFY samples indicated as GPF include new product features to significantly reduce the pressure drop in a clean state while maintaining a high particle number filtration performance. It is expected that there will be minimal soot mass accumulation under DI gasoline engine operational conditions, especially when the filter is packaged close to the engine. For both kinds of filter designs, a study of different filter lengths (e.g. 2-5 inches) will show the performance data trends with respect to particle number filtration and pressure drop. The test series was divided into three parts. On the roller chassis, NEDC type approval tests have been run to investigate the impact of different AERIFY particulate filters on gaseous and particle tailpipe emissions and fuel consumption of the test vehicle. The PN emissions were measured with an Engine Exhaust Condensation Particle Counter (EECPC) from TSI Incorporated, Shoreview, USA. The measurements were done and results reported in line with the referenced procedure [6]. After each NEDC test, an extended, full load acceleration was executed on the roller chassis to measure the peak pressure drop over each individual filter. In the second part of the study, on-road tests were done to investigate the system behavior under real world conditions. First, standardized on-road cycles with low engine loads and exhaust temperatures have been run to try to find driving conditions that will lead to a significant soot build-up in the filter. Secondly, the fuel consumption over a standardized mixed mode driving cycle was measured by accurately monitoring the refilled fuel quantity. Thirdly, the fuel consumption at constant vehicle speeds was assessed on the test track by reading the ECU fuel consumption data. At the end of the above test series, the particle emission behavior of the test vehicle was investigated in repeated cold start scenarios in the cold chamber at approx. -7 C ambient temperature. 5 Test Results Roller Chassis The roller chassis test results display the vehicle behavior in regards to gaseous emissions, particle emissions, fuel consumption, and pressure drop. Table 3 shows the NEDC test results with and without AERIFY particulate filters in the exhaust system. Two separate tests have been done without a filter in the exhaust system. The tailpipe particle number emissions are in a range of approx. 1.8x10 12 #/km which is about three times higher than the PN limit for Diesel light passenger cars (e.g. 6x10 11 #/km). All of the applied AERIFY filters reduce the tailpipe PN emissions clearly below this threshold. All particle mass emissions are clearly below the legislation limit of 4.5 mg/km.

7 19. Aachener Kolloquium Fahrzeug- und Motorentechnik The gaseous HC and CO emission does not show a consistent difference in tests with or without filter. Surprisingly, the NO x emissions are significantly reduced in tests with filters applied (approx %). This goes in hand with a reduction of the CO 2 emission and the fuel consumption (approx. 3%). This phenomenon could be caused by increased internal EGR in the engine map area of the NEDC type approval test, which is caused by the increased backpressure with the filter in the exhaust system. The application of a particulate filter as a solution of the particle number issue of modern DI gasoline light passenger engines certainly cannot be seen as a tool to generally decrease NO x, CO 2, and fuel consumption. With respect to the different filter lengths applied in this study, no consistent impact on gaseous and particulate emissions as well as on fuel consumption could be seen. HC [g/km] CO [g/km] NOx [g/km] CO2 [g/km] Particle Mass [g/km] Particle Number [#/km] Fuel Consumption [l/100 km] w/o GPF E w/o GPF E DPF, 4.33x3 DPF, 4.33x4 DPF, 4.33x5 GPF, 4.33x2 GPF, 4.33x3 GPF, 4.33x4 GPF, 4.33x E E E E E E E Table 3: NEDC emission test results gaseous, particulate, fuel consumption

8 8 19. Aachener Kolloquium Fahrzeug- und Motorentechnik 2010 Figure 3 shows the tailpipe particle number emissions of Table 3 as a bar chart. The non-compliance of the exhaust system without a filter with respect to the PN limit for diesel light passenger cars is obvious. All filter solutions provide excellent filtration performance. DPF honeycombs show efficiencies clearly higher than 90% while GPF honeycombs efficiencies of about 90%. The filtration performance has been slightly relaxed by changing the honeycomb structure from DPF to GPF. In addition, there is no consistent impact of length on filtration performance. 2.0E E E E E+12 Particle number / -/km 1.4E E E E E+11 Diesel Euro 6 limit 4.0E E E E+10 w/o GPF w/o GPF DPF; 4.66" x 3" 1.44E+11 DPF; 4.66" x 4" 9.94E+10 DPF; 4.66" x 5" 2.17E+11 GPF; 4.66" x 2" 1.48E+11 GPF; 4.66" x 3" 2.38E+11 GPF; 4.66" x 4" 1.40E+11 GPF; 4.66" x 5" Figure 3: Particle number tailpipe emissions in NEDC Figure 4 shows the online PN counts for three selected cases the test without the filter, for the DPF with 3 inch length, and the GPF with 3 inch length. The excellent filtration performance of both filters can be seen over the NEDC test schedule. Significant counts are created in the cold start phase when the catalyst heating calibration mode is active. During the 2 nd 4 th ECE, the PN emissions are very low while the vehicle is operated with speeds between 0 km/h and approximately 50 km/h. Significant PN emission are created during accelerations in the EUDC part of the test cycle. The acceleration phases of the EUDC test phase are known as relatively moderate compared to real world accelerations. Hence, the cold start phase with catalyst heating and frequent sharp accelerations may be identified as major PN emission sources of DI gasoline light passenger cars.

9 19. Aachener Kolloquium Fahrzeug- und Motorentechnik W/o GPF DPF; 4.66" x 3" GPF; 4.66" x 3" Particle flow / #/s Veh. speed /km/h Time / s Figure 4: Online particle number counts in NEDC After each NEDC type approval test on the roller chassis, the test vehicle was exposed to an extended, full load acceleration over approx. 90 sec. The peak backpressure measured upstream of the TWC was reported after the exhaust temperature was stabilized. Figure 5 shows the test results. Without a filter in the exhaust system, an absolute pressure of approx mbar was measured upstream of the TWC. This is the reference line. The usage of AERIFY particulate filters DPF or GPF certainly provide an extra backpressure (see p values in the bar chart, Figure 5). However, the pressure drop penalty measured for DPF honeycombs is cut in half by introduction of the GPF honeycombs. A pressure drop minimum seems to be given at about a 4 inch length for both honeycomb types. For the GPF, a reduction to 3 inch and shorter lengths seems to lift pressure drop more sensitively than for the DPF.

10 Aachener Kolloquium Fahrzeug- und Motorentechnik Exhaust gas pressure upstream TWC at full load Exhaust gas pressure uptream TWC / mbar p = 178 mbar p = 171 mbar p = 188 mbar p = 127 mbar p = 88 mbar p = 95 mbar 1200 w/o GPF DPF; 4.66" x 3" DPF; 4.66" x 4" DPF; 4.66" x 5" GPF; 4.66" x 3" GPF; 4.66" x 4" GPF; 4.66" x 5" Figure 5: Peak pressure drops in extended full load accelerations on the roller chassis 6 Test Results On the Road On-road tests have been done to detect critical driving conditions which may lead to a soot build-up. A second test goal was the measurement of the fuel consumption. This was done in two ways, by monitoring the refilled fuel quantity and through reading the ECU fuel consumption data. All on-road tests with GPF were exclusively done with the 3 inch long AERIFY GPF. Figures 6-8 show the exhaust temperatures at the GPF and vehicle speeds over three drive cycles: one standardized low load, one standardized city, and one standardized mixed driving cycle respectively. A quick, thermal soot oxidation can be expected at elevated exhaust temperature of approximately 600 C and higher. Apart from the mixed driving cycle (see Figure 8), the other two cycles do not provide appropriate thermal conditions for a quick soot clean-out. Hence it may be expected that soot build-up appears in the low load and city cycle leading to increased pressure drops over the 3 inch long AERIFY GPF. Table 4 shows the maximum and average speeds of the three standardized test cycles as well as the mileage over which the AERIFY GPF was operated in each of

11 19. Aachener Kolloquium Fahrzeug- und Motorentechnik the given test cycles. For example, we conclude from the first row that the test vehicle was operated over 532 km without exceeding a maximum speed of 33 km/h. A potential pressure drop drift caused by accumulating soot was checked on the road by running the vehicle in a standard driving condition on a straight road. No pressure drop increase was detected by this approach in any of these three on-road tests. After accumulating mileages of 532 km, 146 km, and 806 km, the GPF canning was opened to check the inlet face for soot depositions. In all three cases, the inlet face was white to light-grey, which indicates that no significant soot accumulation took place in the three driving cycles: the nonstop cycling with low load (e.g. 532 km), the in-city (e.g. 146 km), and the mixed driving (e.g. 806 km). The vehicle speed profile of the low load and the city cycle reflect driving conditions seen between 200 sec and 800 sec in the NEDC test (compare Figure 4 with Figure 5-6). The particle number emission is low under this driving condition with very moderate vehicle speeds and accelerations. This can be seen as an explanation for the lack of soot accumulation and any pressure drop increase in the low load and city cycling. Figure 4 shows that even moderate accelerations in the EUDC part produce plenty of PN emission. Hence we may assume that the mixed driving cycle is a source of a significant PN emission (see Figure 8). However, the exhaust gas temperature enters frequently a range between 600 C and 700 C where quick soot oxidation is possible. This probably explains the lack of soot accumulation and pressure drop increase in the mixed driving cycle Driving Cycle Max Speed Average Speed Mileage w/o p Increase Low Load 33 km/h 22 km/h 532 km City 53 km/h 29 km/h 146 km Mixed 131 km/h 67 km/h 806 km Table 4: Max speed, average speed, and mileage accumulation in standardized onroad cycles

12 Aachener Kolloquium Fahrzeug- und Motorentechnik Exhaust gas temperature / C Vehicle speed / km/h Upstream GPF Downstream GPF Time / s Figure 6: Exhaust temperatures and vehicle speed over standardized low load cycle 800 Exhaust gas temperature / C Vehicle speed / km/h Upstream GPF Downstream GPF Time / s Figure 7: Exhaust temperatures and vehicle speed over standardized city cycle

13 19. Aachener Kolloquium Fahrzeug- und Motorentechnik Exhaust gas temperature / C Vehicle speed / km/h Upstream GPF Downstream GPF Time / s Figure 8: Exhaust temperatures and vehicle speed over standardized mixed driving cycle The on-road fuel consumption performance was assessed by two different approaches in this test series. In the first approach, the fuel consumption was measured by measuring the refilled fuel quantity after the standardized mixed driving cycle. This was done for the vehicle with the baseline exhaust system without a GPF and the exhaust system with the 3 inch long AERIFY GPF. The standardized mixed driving cycle was run once per day with the same driver at the same time day for consistency. Three tests were run per exhaust gas system configuration on three consecutive test days. Variations in fuel consumption obviously appear due to day-today variation in traffic (see Figure 9). However, the fuel consumption averaged over three cycles differs less than 1%. This result is in line with the fuel consumption measurements during the NEDC tests. In the mixed driving cycle (see Table 4) as well as in the NEDC test (see Figure 4), the maximum vehicle speeds barely reach levels of km/h. The additional backpressure caused by the 3 inch long AERIFY GPF is not high enough to provide a measureable fuel penalty under those test conditions.

14 Aachener Kolloquium Fahrzeug- und Motorentechnik GPF; 4.66" x 3" W/o GPF Fuel consumption / l/100km #1 #2 #3 Mean value Figure 9: Fuel consumption measured by refilling over three cycles of standardized on-road fuel consumption cycling As a second approach to assess on-road fuel consumption data, the ECU values were read during constant cruising on a test track. Figure 10 reports the results for constant cruising from km/h as well as for maximum vehicle speeds in the gears D and S. A fuel consumption penalty of 2-3% is indicated by the ECU data for vehicle speeds of km/h, while there is practically no fuel penalty at lower vehicle speeds for the exhaust system with the GPF in comparison to the baseline exhaust system without a GPF. The maximum vehicle speed and the corresponding exhaust mass flow were read from the ECU during full load operation on the test track. There is an approximately 4% reduction of air mass flow for the exhaust system with the 3 inch long AERIFY GPF over the exhaust configuration without the GPF (see Figure 11). However, the maximum vehicle speeds are essentially the same.

15 19. Aachener Kolloquium Fahrzeug- und Motorentechnik GPF; 4.66" x 3" Fuel consumption / l/100km W/o GPF Vmax (D) Vehicle speed /km/h Vmax (S) Figure 10: ECU fuel consumption data on the race track in constant speed w/o GPF GPF; 4.66" x 3" Ex. mass flow / g/s Vmax / km/h Figure 11: Exhaust mass flow and maximum vehicle speed on the test track

16 Aachener Kolloquium Fahrzeug- und Motorentechnik Test Results In the Cold Chamber Since it was not possible up to this point in the test series to find operational or driving conditions which clearly blacken or build up soot on the 3 inch long AERIFY GPF sample in the test vehicle, a series of cold starts in the cold chamber was done to investigate the soot production potential of this operational condition. After soaking the vehicle until it was stabilized to approximately -7 C, the engine was started and operated for approximately 10 min. Then the engine was stalled to let the vehicle completely cool down for the next cold start. After eight cold starts a clear blackening of the applied 3 inch long AERIFY GPF was detected (see Picture 1). The extended - 7 C cold start and catalyst heating phase creates more particle mass than under normal ambient temperature of the NEDC. In combination with extended driving in low loads or in the city, soot accumulation might appear for DI gasoline light passenger cars under cold thermal ambient. The location of the GPF will determine its ability for a passive clean-out in these conditions. This cold start case would need to be taken into consideration for any future serial GPF application. Inlet Outlet Picture 1: GPF inlet and outlet face after eight cold starts in the cold chamber 8 Conclusions and Path Forward The usage of uncoated 3 to 5 inch long AERIFY GPF samples solves the particle number emission issue of the Audi A5 2L TFSI gasoline light passenger car versus the threshold of 6x10 11 #/km in NEDC. No modification of the original TWC system was done. No negative impacts on the tailpipe emission behavior or fuel consumption could be detected under NEDC test conditions. It is a plug-and-play solution which did not run into soot accumulation issues during three standardized on-road driving

17 19. Aachener Kolloquium Fahrzeug- und Motorentechnik cycles. A fuel consumption penalty of 2-3% was measured at vehicle speeds of 180 km/h and higher comparing the results of the exhaust system without a filter and with filters. For serial application purpose, extreme driving cycles with low loads over short mileages in cold ambient need to be investigated with respect to the self clean-out capability of the GPF system. Here, the location of the filter ( close-coupled versus underfloor ) will play a significant role due to thermal losses along the exhaust pipe. The last outstanding performance requirement that will influence the sizing of a serial bare GPF solution is the ash storage capability. Since the ash storage under typical DI gasoline operational condition is a complete unknown, some well defined mileage accumulation tests need to be run in a GPF packaging configuration which shall be close to the target configuration of the future serial production car. 9 Definitions, Acronyms, Abbreviations TWC Three-way Catalyst NEDC New European Driving Cycle PN Particle Number DPF Diesel Particulate Filter GPF Gasoline Particulate Filter DI Direct Injection LDV Light-duty Vehicle EUDC Extra Urban Driving Cycle ECE European City cycle OEM Original Equipment Manufacturer EGR Exhaust Gas Recirculation ECU Electronic Control Unit

18 Aachener Kolloquium Fahrzeug- und Motorentechnik Literature [1] Wurms, R.; Budack, R.; Böhme, J.; Dornhöfer, R.; Eiser, A.; Hatz, W. The New 2.0L TFSI with the Audi Valvelift System for the Audi A4 The Next Generation of the Audi TFSI Technology 17. Aachener Kolloquium Fahrzeug- und Motorentechnik 2008 [2] Korte, V.; Lumsden, G.; Fraser, N.; Hall, J. 30% Higher Efficiency with 50% Less Displacement MTZ Motorentechnische Zeitschrift, 03/2010 [3] Braisher, M.; Stone, R.; Price, P. Particle Number Emission from a Range of European Vehicles SAE [4] Ericson, P.; Samson, A. Characterization of Particulate Emissions Propagating in the Exhaust Line for Spark Ignited Engines SAE [5] Li, C.; Mao, F.; Swartzmiller, S.; Wallin, S.; Ziebarth, R. Properties and Performance of Diesel Particulate Filters of an Advanced Ceramic Material SAE [6] United Nations E/ECE/324, E/ECE/TRANS/505 Rev.1/Add.82/Rev.3/Amend.2