Identification Allshift

Transcription

1 Identification Allshift S.A.K van Loenhout DCT 2005 Automotive Engineering Science Eindhoven University of Technology December

2 Preface This report describes the result of the assignment performed for my first internship at the department of Mechanical Engineering of the Eindhoven University of Technology. The assignment was performed in cooperation with the company Drivetrain Innovations BV (DTI) located in Eindhoven. I would like to thank dr.ir Alex Serrarens, ir Rob Pullens, ir. Michiel Pesgens and Wictor Wocjik for their help and advice. December 2005, Stefan van Loenhout 2

3 Contents Preface... 2 Contents Introduction Problems AMT AMT as bolt on system Fuel economy and costs Working principles AMT Brake Shift Technology (BST) Testing Test programme Test vehicle Measurements: Acceleration on roller bench Measurements: Downshifts at cornering Simulations Animation Conclusions Literature

4 Chapter 1 1 Introduction In cooperation with Eindhoven University of Technology, Drivetrain Innovations BV (DTI) develops a Brake Shift Transmission-module (BST-module) which can eliminate the driving torque interruptions during gearshifts of an automated manual transmission (AMT). A Mitsubishi Colt with a 6-speed AMT is in use to test this new module. The gearshift programme of this vehicle is called Allshift. In order to compare the performance of the BST in the future with a reference this test vehicle needs to run a test programme. Aspects like driveability, acceleration and fuel economy are very important in such a test programme. The assignment was to make and perform a test programme, which can map the most important driving characteristics of the Mitsubishi. Because of time shortage, fuel economy and emissions could not be measured within this project. 4

5 Chapter 2 2 Problems AMT In the literature many examples of problems concerning AMTs can be found. There are many developers and manufacturers of gearboxes working on solutions to overcome these problems. In this section such problems are described in order to show that there is a great market for solutions that make the AMT more driveable and popular. 2.1 AMT as bolt on system Motorsports and automotive technology consultant Prodrive and Italian transmission manufacturer Graziano in Automotive Engineering International : The problem is that current AMT systems are just as the name suggests automated versions of existing manual gearboxes. The addition of hydraulic or electric actuators merely replicates the action of the driver-operated manual gear lever. It means that this type of AMT system is limited by the mechanical gearbox and clutch hardware and as such exhibits manual gearshift characteristics, such as slow up shifts with long torque interruptions and unpleasant torque reversal effects. The result is that unlike conventional automatics, such systems do not provide smooth shifting with carefree throttle operation, and unsatisfactory AMT clutch control leads to poor creep and park functions. 2.2 Fuel economy and costs Hitachi Group: In order to achieve widespread acceptance as a replacement for conventional transmissions, the next generation transmission must provide the good fuel economy of an MT, the effortless shifting of an AT, and must also be compact and affordable. Unfortunately, the conventional AMT fails to achieve the seamless shifting of an AT and the twin-clutch AMT is difficult to implement compactly and cost effectively. 5

6 Chapter 3 3 Working principles 3.1 AMT In this section the working principle of the AMT is explained. The AMT in our test vehicle -a 1.5 litre Mitsubishi Colt- is a 6-speed gearbox with a shift programme which is called Allshift. The gearbox consists of speed gear sets with a special layout with two individual drums, which control the shift forks of the different gears. The so-called twin drum shift system is the hart of quicker shifting. Each drum moves a set of two forks, which are connected, to the synchromeshes. As can be seen in figure 1 drum one is connected to the shift fork controlling 1 st and 3 rd gear and the shift fork controlling 5 th and 6 th gear and drum 2 is connected to the shift fork controlling 2 nd and 4 th gear and reverse. By controlling both drums we are able to move the synchromesh towards the 2 nd gear at the same time as we release the synchromesh from the 1 st gear when shifting from 1 st to 2 nd gear. Fig. 1: layout gearbox. This principle also holds for other shifts among neighbouring gears. Gearshifts between gears that are controlled by the same synchromesh will take more time. For example a 4-2 downshift. 6

7 In the figure below one can see that this system with the unique layout of speed gear sets enables quicker shifting between consecutive gears. Fig. 2: shift time for 1 st gear -> 2 nd gear. This car can be driven in both manual or auto mode. In auto mode the cars ECU chooses the gears by measuring quantities like vehicle and engine speed, accelerator pedal position and throttle valve angle. In manual mode the driver is able to choose the gears by moving the gear lever forwards and backwards. Fig. 3: diagram of AMT shift lever 3.2 Brake Shift Technology (BST) The main design objectives of the BST are to eliminate the driving torque interruption during gearshifts and design the BST module as a plug-in module for the AMT. Therefore, a planetary gear and brake are added to the AMT. In Figure 4 the schematic layout of a BST power train is depicted. The annulus gear is connected to the brake, the carrier gear is connected via a reduction to the secondary transmission shaft and the sun 7

8 gear is connected to the engine flywheel. In this configuration, the BST system acts as a parallel path over the AMT through which torque transmittal to the wheels can be continued when the clutch is open. The geometric parameter z and the reduction are chosen such that the sun gear s speed is zero between 3 rd and 4 th gear. For that reason the BST can only eliminate drive torque interruption in the first four gears. Fig. 4: schematic layout of the BST powertrain. 8

9 Chapter 4 4 Testing 4.1 Test programme In order to compare the performance of the BST in the future with a reference, the test vehicle needs to run a test programme. Table 1 is a representation of the tests performed with the vehicle. One can see that there is a distinction between 3 test-categories. The tests are performed in order to map the driveability and the acceleration performance of the car. It was not possible to do a fuel economy test in this stage of development. All these tests where firstly done on a roller bench. The acceleration-tests from zero to the maximum speed are done with different acceleration pedal percentages. This is done in order to check if there is a difference in time for opening the clutch and synchronising the gears and to measure the difference in torque at the driveshaft. The 100% throttle measurement is also important to compare the acceleration from 0 to 100 km/h, which is an important value when comparing car performances. These tests are done twice, with cold and hot engine, so the engine heat wouldn t influence different measurements. Measurements Why? Driveability Acc pref. Speed (v) Acc ped(%) Gear clutch t synchro t Torque T/t Speed (v) acceleration hot/cold 0-max 20 1->6 X X X hot/cold 0-max 40 1->6 X X X hot/cold 0-max 60 1->6 X X X hot/cold 0-max 80 1->6 X X X hot/cold 0-max >6 X X X X acceleration 40-> X X X X kickdown 3->2 X X X X 4->3 60-> >3 X X X X 4->2 kickdown 5->2 X X X X 80-> >4 X X X X kickdown 6->3 X X X X specials: cornering >2 X X X cornering >1 X X X Table 1: test programme. 9

10 Furthermore acceleration-tests have been performed at starting speed of 40, 60 and 80 km/h. The downshift situation occurs during a so-called kick-down when for a relatively short period extra toque is needed at the wheels, for example, to overtake another vehicle in critical situations. Because our test vehicle has apart from the 100% throttle position a special kick-down button under the acceleration pedal we did this test for both situations. At 60 km/h the tests were done starting both in 4 th and 5 th gear. The test vehicle in auto mode shifts to 5 th gear at this speed. This is probably done in order to save fuel. To inhibit the 4 th gear at 60 km/h we drive in manual mode, which is overruled when pushing a kick-down. A practical experience with the AMT gearbox is the reason for doing some special cornering tests. The AMT will shift back when one is driving at a certain low speed in 2 nd or 3 rd gear and one kicks down the accelerator pedal. This is very annoying because the gearshift will take a large amount of time just when the driver regrets for extra torque to the wheels. This can even lead to dangerous situations in daily traffic. Therefore we want to know what BST can improve on this aspect. This will become clearer in the simulations and animations of these events. (Chapter 8) 4.2 Test vehicle The vehicle is equipped with sensors to measure all the different values that are important to study the car s Allshift programme. In appendix A. all measured quantities and sensors are listed. Fig. 5: Mitsubishi Colt. 10

11 Chapter 5 5 Measurements: Acceleration on roller bench. In this section the results of the measurements on the roller bench are discussed. Not all the tests are figured here. In appendix B all the test results are summarized in a table. Looking at the figures you can distinguish in clockwise direction: Driveshaft torque, engine speed +AMT primary speed + AMT secondary speed, position of Shift actuator A + B + clutch, position of clutch actuator + throttle valve angle (deg) + accelerator pedal (%), vehicle speed. It was not possible to take the air-resistance into account because of a problem with the brake of the roller bench. The weight of the vehicle was estimated and it was not possible to set-up the bench for this specific value. So a set-up weight of 910 kg was taken to do the measurements. The maximum speed on the bench was 120 km/h. To get equal circumstances the tests were done twice, with a cold and a hot engine. Because of the problems with other traffic and the danger of doing these acceleration tests on the road, they were only done on the roller bench. 11

12 Fig 6: measurement results, acceleration at 60% throttle. In the table below you can see the servo positions of the drum actuators A and B at a certain gear. These positions are given in radians. These rotations are transposed into translation of the different shift forks by the drums, see Figure 1. Gear Drum / position A [rad] Drum position B [rad] 0/ Table 2: actuator positions [rad]. Looking at the figure of the drum position A and B (3 rd plot of Figure 6), it strikes that moving between two gears the actuator pauses for a small amount of time. At this time, the synchronization ring is pushing against its mesh to connect with the new gear. The controller reserves a little time to let this event go by in a smooth way. The same can be seen for the clutch actuator. The clutch is disengaged for an actuator position of 0 rad. and fully engaged for a position of 60 rad. In the disengagement of this actuator two stages can be distinguished, see Figure 8. The actuator firstly moves very fast to the point were the clutch-plate is almost loose (1). Then it moves in a more smooth way to a point where the clutch is fully disengaged (2). Engaging the clutch is done in three stages as can be seen in Figure 8. The first stage is fast and pushes the plate until it touches (3). Then there is the stage of slip, which is less fast (4). And finally when the clutch is engaged, the actuator moves fast to a safe position (5). At this position the diaphragm spring in the clutch is fully pressing the pressure plates and overclamps the clutchdisc. This actuator behaviour applies to all the shift events and will be simulated in chapter 7 for different tests that were done. 12

13 When one looks at the results of the measurements and the table you can see the car is shifting from 1 st to 5 th gear at 60% throttle position until it reaches 120 km/h. Figure 7 is a picture from the manufacturer which shows when the car shifts up at a certain speed depending on the accelerator pedal position. This does not entirely resemble our measurements, which can be caused by the setting of the roller bench (no road load). Furthermore a failure in the data-acquisition of the accelerator pedal position in these tests, this position in the 4 th plot in figure 6 and 9 shows 0 %. This should be 60 % and 100% respectively. Accelerator pedal (%) vehicle speed (km/h) Fig 7: up shifts by vehicle speed and acc. Pedal position. In figure 8 and 9 the results of the acceleration test with a 100% throttle position are shown. Because of the fact that the car was not able to drive faster than 120 km/h, it upshifts not further than third gear for 100% throttle, see also figure 7 above. Looking at the time that is needed to upshift at these two different tests, there is a difference for the shift from 1 st to 2 nd gear of 1 s., see the 1 st plot in Figure 6 and 9. At 100% throttle it takes 2 s. and at 60% it takes 3 s. before the maximum torque to the wheels is back. This seems contradictive because at 100% throttle there is higher torque level and a bigger difference in shaft speeds, therefore it should take more time to shift to the next gear, than at 60% throttle. At 100% throttle the clutch controller takes 2 s. for closing the clutch, see Figure 8. The clutch actuator takes about 1 s. to run to the three stages at throttle levels lower than 80%. (40% hot, 60% hot and cold: see appendix B.) But because of the throttle valve, which is not yet fully opened in these situations, it takes more time before the maximum torque is at the wheels. In appendix B can be seen that from 80% throttle the torque interruption is about 2 s. The clutch actuator then takes 2 s to run through the three stage of engaging. In Figure 8 the difference in engaging speed and therefore the difference in torque interruption is shown. 13

14 Fig 8: profile followed by clutch actuator. The two profiles, which are followed by the clutch actuator, are presented above. Here one can see that the clutch slip time, the second stage, takes more time for 100% throttle. But due to the fact that the throttle valve at 60% throttle has a slow response, it takes more time before the maximum torque is back to the wheels. For all the other shifts this takes about 1.5 s to shift to the next gear. See also appendix B. Furthermore you can see the difference in driveshaft torque for both experiments and the different speeds of engine and primary shaft when shifting. These are all very obvious and there are no further curiosities. The acceleration from 0 to100 km/h takes about 9.5 s. 14

15 Fig 9: measurement results, acceleration at 100% throttle. 15

16 Figure 10 shows the results of a kick-down test at 40 km/h in 3 rd gear. At kick-down the car shifts back to 2 nd gear in contrast with 100% throttle where there is no downshift. Furthermore one can see an oscillation for the engine speed in de 2 nd plot of figure 10. It looks that the allshift -controller judges the engine speed to be not high enough or the anticipated engagement too harsh in order to engage the clutch. The clutch is already slipping and transmitting torque at this moment. The controller measures this speed difference between the shafts and re-opens the clutch a bit to let the engine-speed flare for a half second more. This can be seen in the 3 rd graph of the figure at 3.5 s. This is in the second stage where the clutch is slipping. In table 3, the results for all performed downshifts are listed Figure 11 shows shifting decision schedule for downshifting. Only downshifts, which shift one gear at a time, are shown in this picture. For example, the downshift from 4 th to 3 rd gear at 60 km/h with 100% throttle. Fig. 10: measurement results, kick-down at 40 km/h. 16

17 Accelerator pedal (%) vehicle speed (km/h) Fig 11: down shifts by vehicle speed and acc. Pedal position. The acceleration test from 60 km/h was performed in both 4 th and 5 th gear. In table 3 it can be seen that for both gears the car shift back to 3 rd and 2 nd gear for respectively 100% throttle and kick-down. The difference between kick-down and 100% throttle obvious, at kick-down one more gear is shifted back than at 100%. The more gears you shift back the higher torque you get on the wheels after the downshift. So it is logical that when the accelerator pedal is pushed into his kick-down position the car shifts back one more gear. Shift time [s] Torque level [Nm] Overtake gear > 100% 40-> kickdown 60-> 100% 60-> 100% 60-> kickdown 60-> kickdown 80-> 100% 80-> kickdown 3 3-> > > > > > > Table 3: gear shifts, shift time and torque level for overtake tests. 17

18 This also occurs when driving in 6 th gear at 80 km/h. See table 3. Here the car shifts down to 4 th gear at 100% throttle and back to 3 rd gear at kick-down. Looking at torque level for this last example: From Nm in 6 th gear to 750 Nm in 3 rd at kick-down and to Nm in 4 th at 100% throttle. So there is a huge difference in torque you get on the wheels at this two different accelerator pedal positions. Looking at shift time for these tests, you see that a kick-down downshift takes more time than a 100% throttle downshift. This is obvious because the differences in torque level and shaft speeds are bigger for kick-down. The difference in shift time at 80 km/h can also be explained looking at the drumlayout. 6 th and 4 th are actuated by a different drum, which saves shifting-time as explained in chapter 3. This in contrast with 6 th and 3 rd gear, which are on the same drum. Fig. 12: measurement results, kick-down at 60 km/h in 4 th gear 18

19 Chapter 6 6 Measurements: Downshifts at cornering These cornering tests were firstly performed on the roller bench to see what happens at these unusual shift operations. Because these tests are events seen in practice frequently, the results of the tests done on the road are shown. Fig. 13: measurement results, corner shift 2->1. 19

20 In Figure 13, the downshift from 2 nd to 1 st gear is shown. As can be seen in the last figure we are dealing with very low speed in 2 nd gear. When you press the accelerator pedal, the car opens its throttle valve full in contrast with the overtake experiments. The driver then first gets torque for half a second after which the gearbox decides to shift down one gear. This is not a nice experience for the driver, because the torque interruption is accentuated even more in this way of control. The engine has reached his maximum torque already when the decision for the downshift is made. The downshift decision should be made immediately or not at all. Because of the time duration of 2 s. for this shift event there is an enormous amount of vehicle speed loss. This first test was done in wet road condition, which can be seen on the vibration in the driveshaft torque in 1 st gear, see the 1 st plot in Figure 13 after 6 s. The front wheels of the car are spinning due too less friction, which follows out of the measured tyre speeds. This vibration is not present in the test in 2 nd gear were the road was dry, see the 1 st plot in Figure 14 after 3.5 s. Looking at the profile, which is followed by the clutch actuator, plot 3 in Figure 13 and 14, the slip stage strikes. At the 2-1 gearshift the clutch takes more time in the slip area in contrast with the 3-2 gearshift. This is probably because the 1 st gear is actually to drive off only. Due to the small ratio the clutch needs more time to get both shaft speeds equal. This problem is less present for the 2 nd gear. The time needed to complete these two downshifts is equal, and about 2 s. 20

21 Fig. 14: measurement results, corner shift 3->2 At the downshift from 3 rd to 2 nd gear the engine speed is often flaring for the second time just after the clutch was almost closed the first time. This can be seen in Figure 15 after 4.5 s. This was seen earlier at the kick-down from 40 km/h, but now in a more pregnant manner. The engine speed seems to be too small or the anticipated engagement too harsh upon which the clutch opens a bit before it totally engages. This happens about one out of two times this test was done. Fig 15: shaft speeds at a 3->2 shift with engine speed peak. 21

22 Chapter 7 7 Simulations For different tests that were done with the car, resembling simulations are made. This is done to get a good comparison with the BST simulations that were done earlier in the development process. In this way a good understanding on the various actuator profiles can be gained. For example, the speed and acceleration of the clutch actuator. Only the AMT simulations are shown here. On this page some results of the simulation of the acceleration-test that were performed with 100% throttle are depicted. Fig. 16: simulation and measurement results of acceleration at 100% throttle. The upper left plot in Figure 16 shows the clutch and throttle profile of the simulation compared with the clutch actuator position and throttle valve angle of the car during the test. The clutch actuator position and throttle valve measurements are scaled in this figure in order to get a match with the simulation profiles. In this way it is easier to see the resemblance or deviation between the measurements and simulation. 22

23 Like the figure shows it is not possible to match the simulation to the measurements precisely. This is due to the fact that the relation between the torque and the throttle valve angle for the Colt is not exactly known. This also holds for the clutch position and the torque transmitted. Furthermore, the produced engine torque has a certain delay upon changing the throttle position. And the clutch torque transmitted during slipping of the clutch is actually unknown since the friction coefficient is not exactly known. We are especially interested in shift times and the slopes of the different profiles. In Figure 18 the results of the acceleration simulation for 60% throttle is presented. Figure 17 shows the different lookup tables used for the throttle and clutch parameters for these two accelerations. (100% and 60% throttle) If you compare the lookup tables used for the acceleration simulations for 60% and 100% throttle valve positions one can see that there is no big difference in the slopes of the clutch as well as the throttle profiles. Except for the engagement of the clutch in 2 nd gear, see Figure 17 for both clutch profiles at about 5 s. This difference in clutch engagement was already discussed at the acceleration-tests, presented in Figure 8. The simulation lookup tables also follow the different stages in which the clutch actuator is moving. One can see this at the different slopes in the clutch profiles in Figure 17. Fig 17: lookup tables for simulation of acceleration tests. 23

24 Fig 18: simulation and measurement results of acceleration at 60% throttle. In the table below you see for these two simulations the time needed to (dis)-engage the clutch for different gears. These times can be seen in Figure 17 for the clutch profile to get from 0 to 220 for engaging and from 220 to 0 for disengaging. 100% throttle 60% throttle Engage clutch in 1 st gear Disengage clutch in 1 st gear Engage clutch in 2 nd gear Disengage clutch 2 nd gear Engage clutch 3 rd gear Table 4: clutch actuator times. There is less difference for the times, the clutch needs to (dis)engage for both simulations, except for engaging 2 nd gear. 24

25 Next the result for the simulations for both the 2-1 and 3-2 downshifts at cornering are shown. In Figure 20 for these two situations the lookup tables, which are used in the simulations, for the clutch and throttle are drawn. This gives the opportunity to compare these downshift with each other and to gain a good understanding on both the clutch and throttle control which are used for these downshift actions. Fig. 19: simulation and measurement results of cornering test for 2->1 downshift. This picture shows there is less difference in disengaging time between both shifts, like at the acceleration tests. Again there is a difference in engaging time like at the corneringtests between these two downshifts. 2->1 3->2 Disengaging Engaging Table 5: clutch actuator times. 25

26 Fig 20: lookup tables for cornering tests. Fig. 21: simulation and measurement results of cornering test for 3->2 downshift. 26

27 During the cornering-tests another weird downshift occurred at driving in 3 rd gear. The car shifted back to 1 st gear when pressing the accelerator pedal fast to kick-down. The results of that measurement are shown below. The simulation for this event is also present in the figures. This is not a common downshift for a kick-down at 3 rd gear, because this occurred only one time during the many times this test was done. It could have something to do with the wet road and the friction of the tires. This can be seen at the vibration in the driveshaft in figure 21. But what you would expect is that the feeling for the driver is less annoying then for the 3-2 downshift. In the figure of the driveshaft torque can be seen that although this event takes more than 2 second the downshift was started earlier so the torque interruption is less sensible. This gives a more pleasant feeling for the driver. Fig. 22: simulation and measurement results of cornering test for 3->1 downshift. 27

28 Chapter 8 8 Animation To show what would be the advantage of the BST-module in daily traffic, two animations are made. The first one is an animation of a crossroad situation were the car shifts back from 2 nd to1 st gear at a speed of 40 km/h. The figure shows the difference for a car with the BST module and the test car. Simulated acceleration profiles are used to make this animation. Fig 23: animation of crossroad situation Figure 23 shows the first animation. This is a situation where the driver approaches a crossroad and can not see clearly what is coming from both direction. He approaches at 40 km/h in 2 nd gear, thinks the crossroad is clear, when suddenly a car is coming from the left. He pushes the accelerator pedal and the car shifts back to 1 st gear. From the figure it can be seen that the difference in distance is about 10 m between both cars. The torque interruption and the extra time for downshifting can cause dangerous situations for a car, which is not equipped with the BST- module. The other simulation is an overtake situation where the car drives at 60 km/h in 4 th gear and shifts back to 3 rd gear. From the other side another car approaches and a dangerous situation can be the result of this overtake-action. When the downshifting take too long and the torque interruption causes less acceleration then needed this car can cause a hit. 28

29 From the figure below can be seen that the difference in distance for both cars is 7 m, which can make the difference for an occasional car accident. Fig 23: animation of overtake situation. 29

30 Chapter 9 9 Conclusions The test-programme enables mapping the driving characteristics of the test vehicle s Allshift programme. Acceleration performance and driveability are the main issues in this programme. First the acceleration-tests from 0 km/h with different accelerator pedal positions were performed. Minor differences in shift times for gearshift from 3 rd gear on are seen. For the gearshift between 1 and 2 the clutch actuator can follow different profiles to engage the clutch. This phenomenon, together with the throttle valve control, is also responsible for the difference in torque interruption for these tests. The Allshift -controller judges the engine speed to be not high enough or the anticipated engagement too harsh in order to engage the clutch. At this moment the clutch needs two attempts to engage. The driver can feel this as a long torque interruption. So the torque interruption can be controlled by the clutch actuation. Further these tests can be used to compare the acceleration performance of the new BST-module with the old AMT. At the different overtake tests, which were performed from 40 km/h, 60 km/h and 80 km/h, the kick-down button under the accelerator pedal involves the shift operation. This determines how many gears are shifted back, during such an overtake action. The time needed for this downshift is longer for the kick-down action, which is obvious because of the bigger differences in torque and shaft speeds. Although this operation takes longer, the torque on the wheels and therefore the acceleration is higher than a 100% throttle overtake action. The chosen gear at 100% throttle will namely be higher. Another thing that can involve the shift times is the layout of the actuator drums. A gearshift from 6 th to 4 th gear, which is actuated by different drums, is faster then a gearshift from 6 th to 3 rd gear, which is actuated by rotating the same drum. A real driveability issue is the cornering behaviour of the test vehicle. Some annoying downshifts are made when deeply pressing the accelerator pedal while driving low speeds in 2 nd and 3 rd gear. The timing of the shifts is misplaced, which causes a long torque interruption. After pressing the pedal it takes some time before the downshift actually takes place so one is driving at maximum engine-torque already. Making the subsequent torque interruption even more pregnant. These downshifts should therefore start before the throttle valve is fully opened or should not take place at all. These operations are not expected and therefore can cause dangerous situations. For calibration and tuning of the shift schedule in an AMT it is advisable to choose another strategy. Animations show that the BST module can improve a lot on the performance of the AMT gearbox. Both the driveability and acceleration performance will be improved importantly. The 0 to 100 km/h sprint will take less time and at overtake actions valuable acceleration time is saved. The cornering behaviour can be improved not only with the elimination of the torque interruption but also with a good controller, which can take more appropriate and safe decisions when it comes to back-shifting. 30

31 Appendix A 31

32 Appendix B test Shift time Torque level [Nm] acceleration gear % cold % hot % cold 1- > % hot > % cold > % hot > % cold > % hot > % cold >3 100% hot 1- > Cornering gear Shift time Torque level [Nm] Corner 2- >1 Corner 3- >2 Corner 3- >1 2- >1 3- >2 3- >

33 Shift time [s] Torque level [Nm] Overtake gear > 100% 3 40-> kickdown 3-> > 100% 4-> > 100% 5-> > kickdown 4-> > kickdown 5-> > 100% 6-> > kickdown 6->

34 Literature [1] J.D.W. de Cock: Vibration Analysis and Synthesis of the Brake/Impulse Shift Transmission Technology [2] [3] 34