Drivability Development Based on CoSimulation of AMESim ...

2009-01-0725

Drivability Development Based on CoSimulation of AMESim Vehicle Model and Simulink HCU Model for Parallel Hybrid Electric Vehicle

Jeongho Hong, Seokjoon Kim and Byungsoon Min Hyundai Motor Company

Copyright © 2009 SAE International

ABSTRACT

Parallel Hybrid Electric Vehicle consists of internal combustion engine, engine clutch, motor, automatic transmission, Integrated Starter Generator (ISG), and battery. Due to hybridizations such as using engine clutch to disengage the internal combustion engine and omitting torque converter from the automatic transmission to increase fuel economy, drivability will not be same as conventional vehicle. To ensure drivability comparable to conventional vehicle, dynamic simulation has been utilized to foresee the drivability issues for the proposed hybrid system and ideas for improvements are tested in simulation. CoSimulation of AMESim vehicle model and Simulink Hybrid Control Unit (HCU) model has been used to test and improve HCU logic.

INTRODUCTION

Fuel economy has been main driving force for the hybrid electrical vehicle (HEV) development. Early in the development, good fuel economy could ensure sales of HEVs even with somewhat less than comparable drivability to conventional vehicle. However, nowadays, there are lots of competitions among HEV manufacturers and customer’s drivability expectations are getting high and high. In order to be competitive in today’s various HEVs manufactured, not only good fuel economy is necessary, but also drivability has to be at least comparable to or better than conventional vehicle.

The proposed hybrid system in this paper is Parallel Hybrid Electric Vehicle. For fuel economy, torque converter has been removed and engine clutch has

been included. Due to those hybridization, drivability will suffer compared to conventional vehicle with automatic transmission. Expected drivability problems are as follows [1]:

Vehicle response and vibration as engine clutch engages/disengages

Vehicle response and vibration as gear shifts up/down

Tip-in/out shock/jerk

Fuel economy benefit of removing torque converter has been well established in [2]. And it was also shown in [2] that gear shift quality with and without torque converter can be comparable.

Usually, it is relatively easy to find and explain mechanism for the drivability problems in simulation rather than with vehicle hardware analysis [3]. Though, it would be more convincing if vehicle hardware analysis can be done. In this paper, to ensure drivability of Parallel HEV comparable to conventional vehicle, dynamic simulation has been utilized to analyze the drivability issues for the proposed hybrid system and ideas for improvements are tested in simulation. This processes of analyzing and testing in simulation will save time and cost when compared to vehicle hardware approaches.

The Engineering Meetings Board has approved this paper for publication. It has successfully completed SAE’s peer review process under the supervision of the session organizer. This process requires a minimum of three (3) reviews by industry experts. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE. ISSN 0148-7191 Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely responsible for the content of the paper. SAE Customer Service: Tel: 877-606-7323 (inside USA and Canada) Tel: 724-776-4970 (outside USA) Fax: 724-776-0790 Email: [email protected] SAE Web Address: http://www.sae.org

Printed in USA

*9-2009-01-0725*

Licensed to Jilin UniversityLicensed from the SAE Digital Library Copyright 2010 SAE International

E-mailing, copying and internet posting are prohibitedDownloaded Tuesday, December 07, 2010 11:39:40 PM

Author:Gilligan-SID:13402-GUID:40772422-202.98.18.39

MAIN SECTION

DRIVABILTITY OF PARALLEL HYBRID ELECTRIC VEHICLE

Parallel Hybrid Electric Vehicle Layout

Engine Motor TM FD

Clutch

Wheel

Wheel

ISG

Engine Motor TM FD

Clutch

Wheel

Wheel

ISG

Figure 1 Parallel Hybrid Electric Vehicle Layout

Parallel HEV consists of internal combustion engine, engine clutch, motor, automatic transmission, ISG, and battery as shown in Figure 1. Engine clutch disconnects internal combustion engine when it is not necessary, and thus improves fuel economy. The torque converter has been removed form the automatic transmission to increase fuel economy.

Parallel HEV has three modes to operate. First, as motor mode, engine clutch is disengaged and vehicle is driven by electric motor only. Second, as engine mode, engine clutch is engaged and vehicle is driven by internal combustion engine only. Last, as hybrid mode, engine clutch is engaged and vehicle is driven by both electric motor and internal combustion engine together.

Drivability due to engine clutch operation

Engine clutch has been utilized to stop internal combustion engine during vehicle operation for fuel economy purpose. Even though fuel saving is possible, engaging/disengaging engine clutch can cause vehicle response and vibration problems during mode change.

In order to have drivability comparable to conventional vehicle with automatic transmission, three items have to be considered for using engine clutch. First, starting internal combustion engine quickly and speeding it up rapidly. Second, synchronizing engine speed and motor speed and controlling engine clutch. Last, torque transition after engine clutch engagement. Internal combustion engine is always operating for conventional vehicle and thus there is no need to start and speed up engine during vehicle operation. This causes Parallel HEV response delay. Synchronizing engine speed and motor speed and controlling engine clutch during engaging/disengaging engine clutch can cause engine clutch engagement shock and engagement time delay, which causes vehicle response delay and vibration problems. At the torque transition phase, motor torque and engine torque have to be controlled accordingly in order to give smooth transition feel to driver.

When vehicle operate in motor mode and driver want to pass on highway, quick-down gear shift happens. Compared to conventional vehicle, Parallel HEV needs

one more operation, mode transition, which requires time. So thus vehicle response time will suffer for Parallel HEV in that situation, too.

Drivability due to torque converter omission

Toque converter used in conventional vehicle with automatic transmission has roles of launching, torque multiplication and driveline damping.

In order to have drivability comparable to conventional vehicle with automatic transmission, two items have to be considered for omitting torque converter. First, vehicle response has to be considered when vehicle launches and accelerate. Second, shock/jerk problems have to be considered when vehicle creeps, launches, gear shifts up/down, and tips-in and out. Shock/Jerk usually causes motor speed resonances, driveshaft (D/S) torque resonances, and vehicle fore-aft acceleration resonances.

COSIMULATION

Software used for the CoSimulation of drivability in this paper is AMESim and Simulink. AMESim is a one-dimensional simulation tool that provides a simulation environment such that mechanical/electric/hydraulic system can be modeled and executed together [4]. Vehicle is modeled in AMESim in this paper. Simulink is a supplementary product for Matlab, which has graphic user interface. HCU control logic is implemented in Simulink. Review and test of HCU control logic is done by CoSimulation of AMESim and Simulink as shown in Figure 2.

Vehicle Model(AMESim)

HCU Model(Simulink,Stateflow)

Figure 2 CoSimulation

Automatic Transmission

Vehicle

Engine/Motor

Automatic Transmission

Vehicle

Engine/Motor

Figure 3 AMESim Vehicle Model

Vehicle model is implemented in AMESim as shown in Figure3. Internal combustion engine is modeled as maps of engine speed and engine torque. ISG and motor are modeled in similar fashion. All of them have appropriate time response characteristics, which are obtained from




Motor Torque CMD Engine Torque TQI CMD

Engine TorqueMotor SpeedEngine Speed

ISG Torque CMD

D/S Speed D/S Torque

Engine Clutch Pressure

Engine Clutch Current CMD

Nm Nm Nm

Nm

NmRPM RPM

RPM

A

Bar

sec

Bench TestSimulation

Motor Torque CMD Engine Torque TQI CMD

Engine TorqueMotor SpeedEngine Speed

ISG Torque CMD

D/S Speed D/S Torque

Engine Clutch Pressure

Engine Clutch Current CMD

Nm Nm Nm

Nm

NmRPM RPM

RPM

A

Bar

sec

Bench TestSimulationBench TestSimulation

experiments, and inertia. Transmission is modeled in such a way that planet gear/clutch/brake are included with inertia. Output power from transmission goes to driveshaft and is interacting with vehicle body through tire model. This AMESim model is good enough to describe vehicle fore-apt motion under 10~30 Hz [5]. Engine clutch is actuated hydraulically and modeled with hydraulic circuit as shown in Figure 4.

VFS Current Control Signal

Pressure Control Part

Force Generating Part

Figure 4 Engine Clutch Hydraulic Circuit

Correlation result between simulation of engine clutch hydraulic circuit and experimental result is shown in Figure 5. They match reasonably well for our purpose.

Current Command

Pressure: Experiment

Pressure: Simulation

Current A Pressure Bar

time (sec)

Figure 5 Engine Clutch Hydraulic Circuit Correlation

HCU control logic is implemented in Simulink as shown in Figure 6 and cosimulated with AMESim.

ECU, MCU, TCU, ETC => CAN

CAN ⇔ HCU ⇔CAN

CAN =>ECU, MCU, TCU, AMESim

Figure 6 HCU Control Logic in Simulink and AMESim

Figure 7 Correlation Result of CoSimulation and Bench Experiment

Interface




Correlation result between CoSimulation of AMESim vehicle model and Simulink HCU control logic and Bench experimental result is shown in Figure 7 and shows reasonably well matched result for our purpose. System Bench layout for the experiment is same as in Figure 1 except that tires are replaced by Dynamometers. In CoSimulation, the same HCU control logic as used in real Vehicle application is used, but other control logic like ECU/TCU/MCU are implemented with only basic functionality.

TIP-IN/OUT SHOCK/JERK IMPROVEMENT BY COSIMULATION

Parallel HEV has less driveline damping than conventional vehicle due to lack of torque converter, which can cause drivability problem. We take Tip-in/out shock/jerk case as an example to show the dynamic simulation process for the development of HEV drivability. In this paper, drivability will be measured by driveshaft torque peak-to-peak magnitude, which resembles vehicle fore-apt acceleration peak-to-peak magnitude as shown in Figure 8.

sec

Nm

Acceleration D/S Torque

g

Figure 8 Acceleration vs. Driveshaft Torque

Effect due to driveline damping

The effect of driveline damping for vehicle tip-in/out maneuver is shown in Figure 9. While conventional vehicle with automatic transmission show little bit of oscillation in transmission input speed when vehicle undergoes Tip-in/out maneuver, Parallel HEV shows much worse and longer oscillation of transmission input speed under similar driving condition.

Motor Torque

T/M Input Speed

APS

Tip-in/outShock/Jerk

Conventional Vehicle Parallel HEV

Engine Torque

Figure 9 Tip-in/out Shock/Jerk of Conventional Vehicle and Parallel HEV

Tip-in/out shock/jerk mechanism analysis by CoSimulation

For the tip-in maneuver during tip-out (coasting), Bench test result of Parallel HEV is shown Figure 10. There are oscillations in motor speed and driveshaft torque, which are called as Tip-in/out Shock/Jerk.

sec

Motor TorqueCommand (Nm)

Motor Speed(RPM)

Half D/STorque (Nm)

Figure 10 Tip-in/out Shock/Jerk Bench Test

For the transmission with planet gears, power is transferred through the gear contact. There always exists gear backlash between gear teeth of contacting gears due to manufacturing accuracy. The effect of gear backlash to the tip-in/out shock/jerk is simulated and the result is shown in Figure 11. As gear backlash is increased from zero to 5degree, the magnitude oscillation in driveshaft torque is increased and are similar to the Bench test result as shown in Figure 12.

Nm

Half D/S Torque

Gear Backlash- None- 1mm- 3mm (5.2°)

sec

Figure 11 Driveshaft Torque Oscillations Due to Varying Degree of Gear Backlash During Tip-in/out

Nm

Half D/S Torque

- Bench Test- Simulation

sec

Figure 12 Driveshaft Torque Oscillation Comparison During Tip-in/out




Gear backlash and Shock/Jerk

Gear Backlash

Gear Motor-Side

Gear Tire-Side

Figure 13 Gear Backlash

Schematically, gear backlash from input to transmission to tire can be drawn as in Figure 13. During coasting, vehicle slows down due to various frictions and tire-side gear drives motor-side gear. During this moment motor-side gear stands at left-side of tire-side gear as shown in Figure 13. At this moment, if driver steps on acceleration pedal, motor-side gear moves to the right side of tire-side gear. After moving the distance of about the size of backlash (3mm), motor-side gear transmits power to the tire-side gear. During this movement of motor-side gear for the size of backlash, there is no power transfer from motor to tire until contacts between gears are made but motor torque is keep increasing as shown in Figure 14. At the gear contact, the amount of about 30Nm of motor torque is suddenly transferred to tire and this produces almost step response at the driveshaft torque (magnitude of around 300 Nm) as shown in Figure 14. This abrupt torque created at the moment of gear contacts on driveshaft creates Shock/Jerk.

sec

mm

Nm

Nm

Gear Backlash

Motor Torque

Half D/S Torque

Figure 14 Shock/Jerk Due to Gear Backlash

Shock/Jerk Improvement by motor control

The key idea behind shock/jerk improvement is the mechanism for shock/jerk due to gear backlash. Motor needs to be controlled such a way that motor-side gear moves close to the tire-side gear when driver tips in but the magnitude of motor torque at the very moment of contact is near ‘0’, which will transmit almost ‘0’ Nm to driveshaft resulting prevention of shock/jerk. Such a motor control profile is created and applied in simulation and the result shows little shock/jerk as shown in Figure 15.

sec

Nm

Half D/S Torque

- Base- Backlash Control

Figure 15 Shock/Jerk Improvement by Backlash Motor Control (Simulation)

Bench test also shows similar result as shown in Figure 16. Robustness for the proposed motor control profile is shown in Table 1. The result in Table I shows that the controller is robust to APS (acceleration position sensor) and motor speed, but it would give better result if control parameters are mapped according to APS and motor speed.

Base Motor Control Backlash Motor Control

- D/S Torque- Motor Speed

- D/S Torque- Motor Speed

Toru

e(N

m)

Speed (

RPM

)

sec sec

Figure 16 Shock/Jerk Improvement by Backlash Motor Control (Bench Test)

CONCLUSION

Drivability issues of Parallel Hybrid Electric Vehicle due to hybridization like omitting torque converter and inserting engine clutch are explained. Tip-in/out Shock/Jerk problem has been analyzed and ideas for improvements are given by means of Dynamic Simulation. Proposed motor control profile is tested in simulation and system bench. By using CoSimulation, Parallel HEV drivability can be developed without much of hardware experiments, which saves time and costs.




REFERENCES

1. Colvin, D. and Masterson, B., “Challenges of Engine Starts and Drivability in a Parallel Hybrid-Electric System”, SAE Paper No. 2004-01-0063.

2. Kao, C. K., Smith, A. L. and Usoro, P. B., “Fuel Economy and Performance Potential of a Five-Speed 4T60-E Starting Clutch Automatic Transmission Vehicle”, SAE Paper No. 2003-01-0246.

3. Sinnamon, J. F., “Co-Simulation Analysis of Transient Response and Control for Engines with Variable Valvetrains”, SAE Paper No. 2007-01-1283.

4. Garcia, O. M. and Kargar, K., “Simulation Tool for Transmission and Driveline Systems Design”, SAE Paper No. 2000-01-0832.

5. Capitani, R., Delogu, M. and Pilo, L., “Analysis of the Influence of a Vehicle’s Driveline Dynamic Behaviour Regarding the Performance Perception at Low Frequencies. SAE Paper No. 2001-01-3333.

1900RPM1800RPM1700RPM1600RPM1500RPM

APS(%)

50

40

30

20

940

860

680

470

Before

80

80

150

240

After

91%

91%

78%

49%

Improvement(%)

910

870

720

530

Before

75%

59%

96%

88%

Improvement(%)

170

260

20

50

After

110

50

60

100

After

75%

94%

87%

67%

Improvement(%)

910

860

700

510

Before

250

50

50

65

After

89%97023073%670

94%8803094%630

92%7209093%550

170

After

Motor Speed

BeforeImprovement(%)


420 88% 490 80%

1900RPM1800RPM1700RPM1600RPM1500RPM

APS(%)

50

40

30

20

940

860

680

470

Before

80

80

150

240

After

91%

91%

78%

49%

Improvement(%)

910

870

720

530

Before

75%

59%

96%

88%

Improvement(%)

170

260

20

50

After

110

50

60

100

After

75%

94%

87%

67%

Improvement(%)

910

860

700

510

Before

250

50

50

65

After

89%97023073%670

94%8803094%630

92%7209093%550

170

After

Motor Speed



420 88% 490 80%

Table 1 Robustness of Backlash Motor Control to APS and Motor Speed – Driveshaft Torque Peak-to-peak Improvement




Drivability Development Based on CoSimulation of AMESim ...

Documents

Transcript of Drivability Development Based on CoSimulation of AMESim ...