Fully Active vs. Reactive AWD coupling systems. How much performance is really needed? Thomas Linortner Manager, Systems Architecture

Transcription

1 Fully Active vs. Reactive AWD coupling systems How much performance is really needed? Thomas Linortner Manager, Systems Architecture

2 Overview 1. Market requirements for AWD systems 2. Active and Reactive Systems two ways of achieving the compromise between performance and cost 3. Performance and costs what are their major drivers? 4. Control of active and reactive systems 5. System performance requirements 6. How do they compare? 7. The right selection Date: Sept 2011 Author: Linortner 2

3 Key market requirements Current main focus of OEMs How to achieve customer targets Cost Weight Reduce complexity Take energy out of the drivetrain Reduce complexity Optimize control strategy for low clutch wear / oil damage Efficiency Reduce number of bearings / bearing positions Performance Fulfill vehicle dynamics requirement (response / torque distribution) Fulfill torque accuracy requirement Compliance with ABS/ESP Today's market: Achieve sufficient performance at best costs! Date: Sept 2011 Author: Linortner 3

4 Systems overview Cost vs. Performance performance Competitive area Slightly better performance leads to significant cost increase Reactive Active Standard Passive market price 100% 175% 220% Slightly lower performance does not fulfill vehicle level requirements price Active: Reactive: Passive: electronic torque w/o delta speed/angle electronic torque with delta speed or angle slip dependent only without electronic control Date: Sept 2011 Author: Linortner 4

5 Systems under investigation Active System Wet multi-plate clutch creating torque with axial force Axial force controlled by the following: Motor/Pump unit Pressure sensor for feedback control Reactive system Wet multi-plate creating torque with axial force Axial force created by the following: Differential speed sensing pump Controlled by proportional pressure limiting valve Date: Sept 2011 Author: Linortner 5

6 Systems under investigation Functional differences Active Reactive Date: Sept 2011 Author: Linortner 6

7 Influencing parameters for performance & cost Active System High power of actuator Faster system requires stronger motor & ECU High accuracy of torque transfer High quality sensors, elaborate compensations Reactive System Performance of reactive element (=pump) and the possible need for preload Element for control (solenoid) Date: Sept 2011 Author: Linortner 7

8 Influencing parameters Active System Methods of reducing power requirements Reduce piston travel P = (F*s) / t Short force path Low compliance friction material Sinter Stiffer paper New material Alternative material Previous material Improve actuator efficiency Operate at best efficiency on Implement smart control logic No control with excessive dynamic if not required Active Torque off Efficient Torque off Date: Sept 2011 Author: Linortner 8

9 Influencing parameters Active System Methods of improving accuracy Short tolerance chain by direct pressure measurement For multi-plate wet clutch: T F a *m * R m N System must control Force * F P * a A Piston Direct pressure measurement m tolerance m vs temp m vs lifetime m vs slip speed Sensor Tolerance Piston Area 100 % Pressure Sensor Piston Pressure Axial Force at Clutch Clutch Torque Date: Sept 2011 Author: Linortner 9

10 torque,delta-n front-rear speed,accpedal,steeringangle Influencing parameters Reactive System Method of improving reactive element performance Definition of pump performance dv_tmax / V_th > 5 30deg delta angle at wheels handling sufficient track - STIFFto achieve max torque Control element performance Preload requirement Torque increase/decrease fl[km/h] fr[km/h] rl[km/h] rr[km/h] pedal[%] steer[ ] No significant need for delta speed Mean delta-n = 1.4rpm torque[nm] delta-n[rpm] time[s] Date: Sept 2011 Author: Linortner 10

11 Control of Active System Active systems are usually used in an overlocked condition unless operated at constant delta speed (ratio mismatch between front and rear axle) In this operating condition, the controlled torque capacity is always higher than the actual transferred torque Date: Sept 2011 Author: Linortner 11

12 Tire Stiffness dependent Control of Reactive System Reactive Systems are transferring the actual required torque according to the tire slip characteristic Transferrable force [N] Übertragbare Kraft PTU Front axle Driving force [N] Hypoid Clutch Rear axle Cornering force [N] Clutch Stiffness dependent Slip [%] In this operating condition, the controlled torque capacity is always equal to the actual transferred torque. Date: Sept 2011 Author: Linortner 12

13 Control of Active and Reactive System Rear axle torque transfer Torque Capacity vs. Actual Transferred Torque Torque Active System Torque Reactive System Torque capacity Clutch overlocked Vehicle driving torque Vehicle driving torque FA Torque RA Torque 50% 75% 90% 50% 75% 90% Torque capacity = actual torque Clutch not overlocked µra=1.0 Axle load: 50:50 µfa=1.0 µfa=0.5 µfa=0.2 µra=1.0 Axle load: 50:50 µfa=1.0 µfa=0.5 µfa=0.2 Date: Sept 2011 Author: Linortner 13

14 System performance requirements Fulfill vehicle dynamics requirement Sufficient torque transferred to the secondary axle Response increase and decrease torque fast enough Fulfill torque accuracy requirement if this is an interface with OEM Date: Sept 2011 Author: Linortner 14

15 acceleration [ m/s 2 ] Sufficient torque transfer Possible torque transfer to the rear E T µ=const. 5 B D PTU 4 C D 3 A Open Center Differential (50:50) 2 1 A B Torque Distribution according Dynamic Axle Loads Active systems possible Control Range Torque on Front Axle [ % ] Torque on Rear Axle [ % ] Date: Sept 2011 Author: Linortner 15

16 Increase and decrease torque fast enough Increase: Controls algorithms have feedback loop via wheel speeds Active System Reactive System T T T 2 T 2 T 1 T 1 n 1 n>> n 1 (!) n n 1 n 2 n Transition from T1 to T2 when slip detected Fast system required Transition from T1 to T2 is a movement on the characteristic Fast response = system feature Date: Sept 2011 Author: Linortner 16

17 Fulfill torque accuracy requirement Common specification Typical SORs state +/- 10% accuracy What are the real requirements for AWD couplings? [Torque actual / skid torque] 1.4 Limit torque +/-10% 1 Eliminate slip at high load Limit binding Low slip for vehicle dynamics [Torque request / skid torque] Date: Sept 2011 Author: Linortner 17

18 How do they compare? Key challenges for active system Low power consumption Key challenges for reactive system Fast response Accuracy / Active behavior Date: Sept 2011 Author: Linortner 18

19 Comparison Power consumption Motor Pump base plate Integrated Pump Motor assembly Optimized for efficiency Pump mid plate T,eta Motor shaft Pump inner rotor Pump outer rotor Operating range n Date: Sept 2011 Author: Linortner 19

20 Torque request / actual [Nm] Comparison Power consumption Actuator Power Measurement in customer driving cycle Customer driving cycle over time Offroad Requested Highway torque Actual torque City Offroad 11.5 W Highway <2 W City 4.2 W Date: Sept 2011 Author: Linortner 20

21 Comparison Response Vehicle measurement Reactive system 120 dn[rpm] mkupp[nm*10] dn[rpm] mkupp[nm*10] Same dynamic as active system (<150ms) Clutch open Slip control Required torque controlled Time [s] Date: Sept 2011 Author: Linortner 21

22 torque [Nm] Comparison Accuracy Rig test evaluation on torque vs. delta-n Significantly worse than active system C 70 C Is the same vehicle dynamic with such a variation possible? Delta ~50% delta-n [rpm] Date: Sept 2011 Author: Linortner 22

23 Comparison Accuracy WOT start simulation Reactive System WOT Acceleration 30 C/70 C, orifice % pedal [%] nmot/10 [rpm] velocity [km/h] delta n [rpm] x10 Axletorque front/2.77 [Nm] Axletorque rear/2.77 [Nm] Distribution [%] x Nm vs. 135Nm slip front slip rear Variation between the characteristics 10% absolute variation 3% change in torque distribution Date: Sept 2011 Author: Linortner 23

24 torque [Nm] Comparison Accuracy Rig test evaluation on torque vs. delta-n Significantly worse than active system C 70 C Is the same vehicle dynamic with such a variation possible? Delta ~50% Yes, reactive systems works only together with the tire! delta-n [rpm] Date: Sept 2011 Author: Linortner 24

25 acceleration [ m/s 2 ] Comparison Torque transfer Required torque transfer µ=const. B C 5 µ=1 dry A Open Center Differential (50:50) 4 B Torque Distribution according Dynamic Axle Loads 3 µ=0.5 wet C Active systems possible Control Range Reactive torque transfer 2 A 1 µ=0.2 snow Torque on Front Axle [ % ] Torque on Rear Axle [ % ] Date: Sept 2011 Author: Linortner 25

26 Summary Active and Reactive systems are almost indistinguishable in the car, also to experienced drivers Same appropriate torque transfer Comparable dynamics Comparable power consumption More important requirements for the end customer are: Acoustics / comfort Durability Availability There is no right selection from a performance perspective The right selection is driven by other variables Packaging Integration / Interface Customer preference Cost advantages Date: Sept 2011 Author: Linortner 26

27 Date: Sept 2011 Author: Linortner 27 Open Questions?