Boosting the Starting Torque of Downsized SI Engines GT-Suite User s Conference 2002

Size: px

Start display at page:

Download "Boosting the Starting Torque of Downsized SI Engines GT-Suite User s Conference 2002"

Alisha Curtis
5 years ago
Views:

1 GT-Suite User s Conference 2002 Hans Rohs Inst. For Combustion Engines (VKA) RWTH Aachen Knut Habermann, Oliver Lang, Martin Rauscher, Christof Schernus FEV Motorentechnik GmbH Acknowledgement: Some of the presented results are taken from a research program funded by the European Community (GRD ) Good morning Ladies and Gentleman. This presentation shows results of a study about boosting the starting torque of downsized SI engines. First, I want to acknowledge the contribution of my fellow co-authors Knut Habermann, Oliver Lang, Martin Rauscher and Christof Schernus and please note some of the presented results are taken from a research program funded by the European Community. Rohs/Habermann/Lang/Rauscher/Schernus 1

2 2 Contents Turbocharging Challenges of SI Engines Additional Boosting Device: e-booster Additional Boosting Device: Roots blower System Comparison Summary Now, let me then outline the structure of this presentation. As an introduction I d like to give a brief overview about the challenges regarding the turbocharging of SI engines especially under downsizing circumstances. Then I m going to show you how these challenges can be met with the implementation of an additional boosting device, either an e-booster or a Roots blower. For both of these I will discuss steady state and transient system performance and a comparison of the two systems will be presented. Finally I will summarize the results. Rohs/Habermann/Lang/Rauscher/Schernus 2

3 rpm (bar) 3 Starting Torque Characteristics of TC and NA Engines Turbos nat. aspirated SI engines rpm (bar) BMEP BMEP TC rpm NA rpm This plot shows the starting torque characteristics of turbocharged and naturally aspirated SI engines. To illustrate how to read this plot two sample full load curves are added. The main plot contains the steady-state full load BMEP values for 1000 rpm and 2000 rpm. Therefore the further to the upper left an engine is located in this plot the steeper the descend of BMEP towards lower engine speeds is. Compared with naturally aspirated engines the turbocharged engines generally have a lower relative BMEP at 1000 rpm. Now, if the turbocharging is applied as part of a downsizing concept this results in less then desired torque at very low engine speeds even if torque and power at higher engine speeds match or even surpass those of a non downsized naturally aspirated engine. And this gets even worse if you look at the transient behaviour of such an engine concept because of the turbo lag. The goal of this study is to show how both, the steady-state and the transient performance of a downsized SI engine, can be significantly improved with the implementation of an additional boosting device. Rohs/Habermann/Lang/Rauscher/Schernus 3

4 4 Simulation model Intercooler Intake manifold TC Exhaust system Exhaust manifold Wastegate BMEP (bar) BMEP at full load Engine Speed (rpm) This GT-Power model of an 1. litre 4 cylinder turbocharged SI engine is used as the base model for these investigations. The highlighted components are the turbocharger, the intercooler, the intake and exhaust manifolds, the wastegate and the exhaust system. The calculated full load BMEP curve of this engine illustrates the lack of starting torque. While at slightly above 2000 rpm the engine has a BMEP of 20 bar this drops to approximately 11 bar at 1000 rpm. To evaluate the transient behaviour of this engine we have performed load step simulations at multiple engine speeds. Rohs/Habermann/Lang/Rauscher/Schernus 4

5 5 Turbo Engine Load Step Response Example (bar) Load Step BMEP=2 bar to Full 1750 rpm = const. transient BMEP steady state BMEP 1000 rpm TC shaft speed (bar) boost pressure PR TC compressor Time after load step (s) Time after load step (s) For instance, this slide shows the load step response at 1750 rpm from 2 bar BMEP to full load. The transient model is set up so that the simulation converges at a base operating point of 1750 rpm and 2 bar BMEP. Then at the start of the load step the throttle is opened and the wastegate is closed (both within 0.2 seconds) and the engine speed is kept constant. This slide shows the engine BMEP in the upper left plot, the boost pressure in the lower left plot, the Turbocharger shaft speed in the upper right plot and the Compressor pressure ratio in the lower right plot (all plotted versus the time after the load step). With the opening of the throttle the boost pressure instantly increases up to the ambient pressure like a naturally aspirated engine would behave, but from that time on the pressure increase continues much slower.the turbocharger takes time to get up to speed from the initial value of below rpm up to the steady state full load value of approximately rpm. Therefore even 3.5 seconds after the load step the engine hast not reached its steady state BMEP value. To generate a map of the transient capability of the engine we have selected discrete time values after the load step and marked the respective reached operating points in the full load BMEP map. Rohs/Habermann/Lang/Rauscher/Schernus 5

6 6 Steady State Performance and Load Step Response Turbocharged SI Engine steady state full load BMEP response after 0.0 s (BMEP=2 bar) 0.25 s 0.5 s 1.0 s 2.0 s 3.0 s BMEP (bar) Load 1750 rpm 3.00 s 2.00 s 1.00 s 0.50 s 0.25 s 0.00 s Engine Speed (rpm) By repeating the load step calculation at different engine speeds, we gain a map of both, steady state performance and load step response. If you have a look at the dark blue line for instance you can read the maximum BMEP 1 second after a load step from 2 bar BMEP which is significantly lower then the full load value. To use this engine in a downsizing concept the low end torque and the transient performance would both need a significant boost. To achieve this, several solutions come to mind. As the low performance is mainly a result of insufficient air supply a modification of the boosting system is necessary. Rohs/Habermann/Lang/Rauscher/Schernus 6

7 7 Required Boost Pressure in Compressor Map Compressor PR desired boost pr Surge Limit rpm η = 70 % steady state full load 75 % 60 % 50 % 40 % corr. flow rate (m 3 /s) If we have a look at the compressor map with the steady state full load curve and the desired boost pressure, we can see that an electrically assisted turbocharger won t be capable of increasing the steady state BMEP at low engine speeds because the operating points of the base engine are already very close to the surge limit. Thus the application of an additional boosting device is more promising. As I said in my introduction we have investigated the use of an additional e- booster as well as an additional roots blower. Rohs/Habermann/Lang/Rauscher/Schernus 7

8 Implementation of e-booster in GT-Power model Bypass E-Booster Electrical load Turbocharger This picture shows the implementation of the e-booster in the GT-Power model. The turbocharger remains the same as in the base model. The e-booster is applied upstream of the turbocharger and consists of the compressor, a free shaft and a torque object. The respective electrical load for 1.0 kw effective compressor power is subtracted from the engine cranktrain for steady state operation. A bypass is provided for operating points where the e-booster is idling, switched off or causes a pressure loss (This can occur in transient operation). For the load step simulations the following assumptions have been made: At 2 bar BMEP the bypass is fully opened and the e-booster is idling. This is realised in the GT-Power model by running the e-booster at as little speed as possible in the provided compressor map. At the start of the load step a constant power of 1.0 kw is provided instantly while the closing of bypass and wastegate as well as the opening of the throttle take 0.2 seconds. Rohs/Habermann/Lang/Rauscher/Schernus

9 9 Additional e-booster: Full load and load step response Turbo + e-booster steady state full load BMEP response after 0.0 s (BMEP=2 bar) 0.25 s 0.5 s 1.0 s 2.0 s 3.0 s BMEP (bar) Engine Speed (rpm) This figure shows the response map of the turbo engine with the e-booster compared to the results of the engine equipped only with the turbocharger, the latter ones being dimmed in the graph. This shows a significant increase in steady state and transient performance. The area between the black and the grey line marks the gain in steady state full load BMEP. At 1000 rpm the system with e-booster has more then 15 bar BMEP which is an increase of approximately 4 bar. The maximum BMEP of 20 bar is reached at 2000 rpm versus 2250 rpm without the e- booster. A comparison of the transient results shows that, while for the first quarter of a second there are no benefits detectable, half a second after the load step the system with e-booster begins to pull ahead. One second after the load step the system with e-booster already has a BMEP of 11 bar at 1000 rpm which equals the maximum steady state value for the turbocharged base engine at this speed. Another second later the system with e-booster has 2 to 4 bar more BMEP for the whole engine speed range below 2000 rpm. And this stays that way further on, illustrated here, three seconds after the load step. Rohs/Habermann/Lang/Rauscher/Schernus 9

10 10 E-booster steady engine speeds rpm Comparison of Compressor Wheel Diameters 37 and compressor wheel diameter: 37 mm surge 70 % 1.6 compressor wheel diameter: 43 mm /min /min 60 % /min 70 % PR /min engine speed 1.2 surge 60 % corr. flow rate (m³/s) corr. flow rate (m³/s) Our simulations have shown that choosing the right compressor size for the e-booster is critical. The plot shows an efficiency map for for a compressor with a wheel diameter of 37 mm and the operating points for steady state full load from 1000 rpm ( the upper left point) to 2000 rpm ( the lower right point). This configuration results in compressor speeds up to rpm which is above the current technical limit of rpm maximum for electric motors used in e-boosters. With an increase of the chosen wheel diameter to 43 mm the required speeds drop to more favourable values but the operating point at 1000 rpm engine speed is getting close to the surge line. Rohs/Habermann/Lang/Rauscher/Schernus 10

11 11 Required Kin. Energy for Transient E-booster Operation Load Step BMEP=2 bar to Full 1750 rpm = const s Compressor PR s /min 1.0 s s /min d = 37 mm d = 43 mm Energy (J) Choosing a wrong compressor size can also have negative impacts on the transient behaviour. This plot shows the required kinetic energy for the e-booster during a load step from 2 bar BMEP to full load at 1750 rpm. Marked are the time since the load step with the green lines and compressor speeds with the blue lines. The black line represents a compressor with a wheel diameter of 37 mm and the dashed red line with 43 mm. Both are operated with a starting speed of rpm which equals a kinetic energy of about 200 J. The slight differences in the kinetic energy are due to the different moments of inertia. The bigger Compressor reaches a pressure ratio of 1.05 after an increase in the kinetic energy of approximately 200 J. By contrast the smaller compressor requires more then double that amount to reach the same pressure ratio due to the much higher speed needed. On top of that the pressure ratio of the smaller compressor even drops to values below 1.0 for a significant amount of time because the compressor speed is not high enough for the required air mass flow. Rohs/Habermann/Lang/Rauscher/Schernus 11

12 Additional Roots blower Large gear ratio compact booster Long full load operation Breathing characteristics similar to engine: high boost @ low revs booster speed (rpm) 12000 000 4000 0 Gear Ratio

12 12 Additional Roots blower Large gear ratio compact booster Long full load operation Breathing characteristics similar to engine: high low revs booster speed (rpm) Gear Ratio Engine/Booster = 1:4 Engagement Simple: purely speed controlled Disengagement engine speed (rpm) Booster engaged only at low revs: ½No delay at start ½No turbo lag ½Boost pressure gradient only depends on throttle ½Small power consumption Small blower operating speed range high average efficiency Secondary air injection capability Low backpressure sensitivity Let me list some of the properties of this system. The Roots blower, which has the benefit of similar breathing characteristics to the engine, is connected to the belt drive with a clutch which can be engaged and disengaged automatically at selected speeds. That way a high gear ratio can be chosen to support the turbocharger at low engine speeds and the roots blower can be disengaged at higher speeds where the turbocharger is in full effect. This results in no delay at the start of the blower and thus no turbo lag. The boost pressure gradient only depends on the throttle. The low speeds mean low power consumption. Additionally the roots blower has secondary air injection capability and low bachpressure sensitivity. As the blower is operated only in a small speed range it can be optimized for a high average efficiency. As the required power is provided by the engine itself there is no problem with long full load operation. The large gear ratio results in a compact booster. Rohs/Habermann/Lang/Rauscher/Schernus 12

13 13 Implementation of roots blower in GT-Power model Bypass Roots blower Mechanical connection Turbocharger The implementation of the roots blower into the GT-Power model is handled the same way as with the e-booster. The Roots blower is applied upstream of the turbocharger and a bypass is provided. The compressor object is connected to the engine crankcase with a gear connection and an additional torque object is implemented to simulate the mechanical friction of the Roots blower. Again for 2 bar BMEP the bypass is fully opened and the Roots blower is disengaged. In the GT-Power model the gear ratio of the gear connection is set as low as possible with the available compressor map data. The Torque object to simulate the friction is disconnected entirely. Again the switching of throttle, bypass and wastegate takes 0.2 seconds and in the same amount of time the clutch is engaged, meaning that the gear ratio for both the Roots blower and the torque object are set to the chosen ratio of 4.2 : 1. Rohs/Habermann/Lang/Rauscher/Schernus 13

14 14 Additional Roots blower: Full load & load step response Turbo + Roots steady state full load BMEP response after 0.0 s (BMEP=2 bar) 0.25 s 0.5 s 1.0 s 2.0 s 3.0 s BMEP (bar) engine speed (rpm) This plot shows the simulation results for the system with roots blower. An overlay of the results of the base engine reveals a similar performance increase as with the e-booster. The steady state BMEP at 1000 rpm is increased to 15 bar and the maximum BMEP of 20 bar is reached at 2000 rpm. We can see, that in transient operation it is desirable to keep the Roots blower engaged for speeds up to 3000 rpm. The load step simulation results are even more impressive then with the e-booster. After half a second the system with the Roots blower has gained up to 3 bar BMEP versus the base engine.. Another half a second later it has already surpassed the steady state full load curve of the base engine for all speeds below 2000 rpm. Two seconds after the load step the BMEP is approaching its steady state full load values for engine speeds above 1500 rpm. Now, what is the cause for this impressive behaviour? Rohs/Habermann/Lang/Rauscher/Schernus 14

15 15 Steady state and transient operation in Roots map speed rootsblower (rpm) % % 55 % s max transient PR PR steady state s s 0.2 s 1.0 full load, engine speed (rpm) speed gain TC clutch engagement volumetric flow rate (m 3 /s) pressure increase rootsblower effective power rootsblower (kw) Lets have a look at the compressor map of the Roots blower. The black squares connected with the black line are the operating points at steady state full load. The red lines mark the effective power of the Roots blower which is about 1.0 kw at 1000 and 2000 rpm Engine speed and up to 1.3 kw in-between. Now, what happens during a load step, for example at 1000 rpm engine speed? Within the first 0.2 seconds the clutch is engaged and the Roots blower is accelerated to the respective speed of 4200 rpm with the gear ratio of 4.2. So far the pressure ratio stays at 1.0 as the bypass is just closing. Now, with the throttle fully opened and wastegate and bypass closed, the Roots blower quickly builds up pressure. 1.4 seconds after the load step the pressure ratio is above the steady state value at this speed. Further on as the turbocharger gets up to speed it slowly drops back to the steady state value. If we repeat this procedure for multiple engine speeds and connect the operating points with maximum pressure ratio we can see that the transient operating range of the Roots blower significantly surpasses its steady state operating range. Especially at 2000 rpm engine speed the maximum effective power of the Roots blower is more then double the steady state value. Rohs/Habermann/Lang/Rauscher/Schernus 15

16 16 Steady state performance w/o and w/ additional booster BMEP (bar) Turbo only Turbo + e-booster Turbo + Roots engine speed (rpm) A comparison of the steady state full load curves shows that at 1000 and 2000 rpm engine speed both systems deliver about the same BMEP.But inbetween the Roots blower benefits from his bigger power supply. As we learned on the previous slide the Roots blower consumes up to 1.3 kw in these operating points while the e-booster has to stay at 1.0 kw maximum. But both system significantly increase the performance of the engine. For the transient comparison the load step from 2 bar BMEP to full load at 1750 rpm is chosen. The steady state full load BMEP at 1750 rpm of all three systems given in this plot Rohs/Habermann/Lang/Rauscher/Schernus 16

17 17 Comparison of Load 1750 rpm Turbo Turbo + Roots Turbo + E-Booster (bar) transient BMEP steady state BMEP 1000 rpm TC shaft speed e-booster speed (bar) boost pressure PR TC compressor Roots e-booster Time after load step (s) Time after load step (s) show up in this plot as dashed lines in the upper left diagram. For both systems it is obvious, that the additional boosters have two major benefits for the transient behaviour of the system. They directly increase the boost pressure after the closing of the throttle and they increase the air massflow and the available exhaust energy resulting in a faster acceleration of the turbocharger during the load step. However, while the full load values for the two systems with additional boosting devices are not to far apart at 1750 rpm, the transient performance shows another picture. Half a second after the load step the system with the Roots blower has approximately 1.5 bar more BMEP then the system with the e-booster. Another half a second later the lead increases to 2.5 bar. The cause for this is visible in the lower right plot. As the roots blower is forced to its full load speed immediately with the engagement of the clutch its pressure ratio rises above its steady state value, while the e-booster at the same time is still accelerating and only slowly gets up to its steady state pressure ratio. This results in a significantly higher boost pressure with the Roots blower. Three seconds after the load step the system with the e-booster catches up to the system with the roots blower. Rohs/Habermann/Lang/Rauscher/Schernus 17

18 1 Summary ½ Supplementary booster supports TC compressor and enhances the available exhaust energy faster TC acceleration ½ E-booster developed to good maturity, easy integration but demanding to power supply and battery ½ Additional mechanical booster provides high boost pressure at low engine speeds (transient & steady state). High gear ratios enable compact boosters to be integrated into belt drive. ½ Generally, additional costs for the additional booster have to be small to achieve good acceptance in the market Let me briefly summarize the results of our study..... Thank your very much for your attention, I m looking forward to your questions. Rohs/Habermann/Lang/Rauscher/Schernus 1

GT-Suite European User Conference

GT-Suite European User Conference E-Charging on a High Performance Diesel engine D. Peci, C. Venezia EMEA Region - Powertrain Engineering Powertrain Research&Technology Frankfurt, Germany October 26th,