SAE TECHNICALPAPER SERIES 2002-01-0523 Cold Starting Performance of a 42-Volt Integrated Starter Generator System Gerald T. Fattic, James E. Walters and Fani S. Gunawan Delphi Automotive Systems, Energenix Center Reprinted From: 42 Volt Technology 2002 (SP–1661) SAE 2002 World Congress Detroit, Michigan March 4-7, 2002 400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the paper may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay a per article copy fee through the Copyright Clearance Center, Inc. Operations Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department. To request permission to reprint a technical paper or permission to use copyrighted SAE publications in other works, contact the SAE Publications Group. All SAE papers, standards, and selected books are abstracted and indexed in the Global Mobility Database No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written permission of the publisher. ISSN 0148-7191 Copyright 2002 Society of Automotive Engineers, Inc. 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. A process is available by which discussions will be printed with the paper if it is published in SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group. Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE. Printed in USA 2002-01-0523 Cold Starting Performance of a 42-Volt Integrated Starter Generator System Gerald T. Fattic, James E. Walters and Fani S. Gunawan Delphi Automotive Systems, Energenix Center Copyright © 2002 Society of Automotive Engineers, Inc. ABSTRACT Over the next several years, vehicle manufacturers will begin to use a 42 volt based system to integrate the starter and generator into one unit known as an integrated starter-generator (ISG). The ISG and its associated electronics and battery pack form a system that has the ability to perform torque smoothing of the driveline, electrical launch assist, regenerative braking, high power generation, engine stop/start, and other features. One of the important tasks to be performed by the ISG is starting the internal combustion engine under extremely low temperature conditions. Traditionally, the 12-volt cranking motor has performed this solitary task over the last sixty years. The ISG system is capable of incorporating the cranking motor task and must be designed to perform this function over the full automotive temperature range. The cold starting requirements have a great influence on the design of any ISG system. This paper will examine how the cold starting requirements affected the design of the Delphi Energen10® ISG system. Test results performed at –29 degrees centigrade for the cranking of a gasoline 4.0 liter, V-6 powertrain are presented. A discussion of the electric motor control strategy used during the cold starting events with an ISG system is also included. across the operating temperature range to enable the engine the opportunity to start. As new demands for electrical power continue to increase the amount of required generator power, the industry is moving to a higher voltage system in order to reduce current levels. A 42 volt system has been specified to provide the opportunity to supply more power to various electrical loads on the vehicle. The 42 volt system provides sufficient voltage for one ISG machine to provide the starting and the generating functions. Delphi Automotive Systems has developed an ISG system, which was incorporated into a Ford Explorer. The installation is shown in figure 1. Trans ISG Engine INTRODUCTION The cranking motor is specifically designed to start the internal combustion engine. It must produce sufficient torque to turn the engine to a specified minimum speed Figure 1 The ISG System Installation. ENGINE AND TRANSMISSION SELECTED FOR THE SYSTEM DESIGN 250 200 In order to understand the requirements for engine cranking, an understanding of the load characteristics of the engine-transmission combination was required. DETERMING DYNAMIC FRICTION TORQUE To determine the dynamic friction torque of the engine and transmission, the engine and transmission combination was mounted with an ISG and soaked in a thermal chamber. The chamber was cooled to -29 degrees centigrade. The spark plugs were removed from the engine. This was done to release any pressure from the cylinders when the crankshaft was rotated. The torque required for turning the crankshaft several rotations was recorded. The average value of 55 Nm was recorded and used as the dynamic friction torque value for low speed and low temperature operation. The process was repeated at 25 degrees centigrade to record the dynamic friction torque. The average value of 40 Nm was recorded as the dynamic friction torque at room temperature. 150 Torque Nm The engine and transmission selected for the system was the Ford 4.0-liter V-6 gasoline engine and five speed 5R55E automatic transmission. The engine specifications are as follows: Bore: 100.4 mm Stroke: 84.4 mm Compression Ratio: 9.7 The Transmission specifications are as follows: 1st gear ratio: 2.47 2nd gear ratio: 1.86 3rd gear ratio: 1.47 4th gear ratio: 1.00 5th gear ratio: 0.75 100 50 Peak Torque for the First Compression Stroke 0 Peak Torque for the Second Compression Stroke -50 -100 -150 0 60 120 180 240 300 Crankshaft angle (Degrees) Figure 2 Crankshaft Torque During Rotation. The first compression stroke results in an increase in the crankshaft torque, which is followed by a decrease during the power stroke. The second compression stroke is the first full compression stroke. This results in a peak torque higher than the first compression stroke. The following compression strokes have lower peak torque values because the other cylinders are finishing their power stroke when the compression stroke is occurring and the intake manifold pressure has been reduced by the earlier compression strokes. CRANKSHAFT POSITION AT SHUTDOWN When the engine comes to rest after running, the crankshaft position settles in the area of minimum crankshaft torque. The area where the crankshaft torque is less than the friction torque is the position where the engine will come to rest after running. Figure 3 shows the crankshaft angles where the crankshaft torque is less than the dynamic friction torque of 40 Nm. 200 Nm Dynamic Friction Torque 55 Nm 40 Nm GAS PRESSURE AND FRICTION TORQUE A calculation was performed for the torque required to turn the crankshaft at low speeds using a Delphi ® developed MATLAB model. The friction torque value of 55 Nm was used for the –29 degrees centigrade calculation. The speed was assumed to be constant so there were no effects caused by inertia of the components. The graph of the crankshaft torque is displayed in figure 2. These are the areas where the engine will stop when shutdown occurs 150 Nm Crankshaft Torque Temperature -29 degrees centigrade 25 degrees centigrade 100 Nm 50 Nm 0 Nm -50 Nm 240 300 360 420 480 540 600 660 720 Cranshaft angle (Degrees) Figure 3 Predicted Crankshaft Stopping Positions The position where the engine-starting event occurs is the position that the engine stops on shutdown. This position allows at least 30 degrees of crankshaft rotation on the V-6 engine before reaching the maximum torque of the first full compression stroke. SPEED AND POSITION SENSOR FOR THE ELECTRICAL MACHINE CONTROL The sensor strategy that was chosen for the electric machine control was the 58X sensor configuration. Figure 4 displays the sensor pulse profile that appears as 58 evenly spaced pulses with an open area. The placement of the open space indicates a particular position of the crankshaft. 58 consecutive pulses The motor torque available at speeds over 100 rpm is dependent upon the voltage available from the 42V system. The motor torque decreases as the system voltage decreases when the speed is above 100 rpm. This effect will be observed when the system voltage varies as the torque load of the system changes during the cranking event. ELECTRICAL DRIVE SYSTEM In the application of ISG systems, the electrical drive system consists of the electric machine, position sensor, current sensors, inverter, and machine controller. Figure 6 shows a simplified electrical drive system. Battery Open space the width of two pulses Torque* Machine Controller Inverter Electric Machine Figure 4 Pulse-train of a 58X Sensor. A digital sensing device was used to detect the movement of the 58-tooth wheel. The digital sensor was chosen so the system could operate at very low speeds. The electric machine controller will utilize the initial sensor pulses from the movement of the crankshaft for speed information to be used in the control algorithm. The 58X sensor system was chosen because it is a common system used for engine controls for many different engine models. The engine crankshaft sensor can be used for the electric machine control, eliminating the need for a separate sensor. ELECTRIC MACHINE DESIGN The electric machine of the ISG system is a three-phase induction machine. The electric machine was designed to produce 200 Nm of torque at 100 rpm at – 29 degrees centigrade. The electric machine torque characteristics are shown in figure 5. Ib ∆θROTOR Ia Figure 6 Simplified Electrical Drive System The machine controller receives a torque command from a vehicle level controller and applies the appropriate voltage to the electric machine to create the desired torque. The machine controller uses vector control techniques in order to have good dynamic performance and disturbance rejection capabilities across a wide speed range. A simplified vector control strategy is shown in figure 7. Torque* Iq Table Speed Id Vq Current Controller Coordinate Trans. Vd PWM Generation Electric Machine ∆θROTOR 250 Ib 24 Volts 28 Volts 32 Volts Coordinate Trans. 200 Ia θ SYNC Temperature Torque Nm Inverter 150 Slip Angle Calculator θSLIP + Σ + θROTOR Angle Processing Speed 100 Figure 7 Vector Control Strategy ® 50 0 0 50 100 150 200 250 300 RPM Figure 5 Electric Machine Torque for Different System Voltages at –29 Degrees Centigrade. For this generation of Energen10 ISG systems, the machine technology is an induction machine that is operated with a wide field-weakened range. Since the motor is an induction machine, an incremental position sensor such as a 58X sensor can be used. Due to the high pole number of the electric machine, the controller’s torque regulation capability becomes more sensitive to Demux Controller / Stationary [T',Iqs'] Iqs_err V q s Vq Vq_lim PI Vxs Vd_lim [T',Ids'] ange_cont Ids_err V d s Vxs Vds_mot Vd Vqs_cont Mux IND_SFUNCII Vys Vys ange_mot Inverter Model Vds_cont Demux IM Motor Model Ids* Sum1 Post-Processing Stationary / Motor Vqs_mot Iqs* Sum Position Sensor and wb_flux_qs wb_flux_ds Mux1 PI1 theta_e wr theta_r Iqs Ids Theta_r Position Iqs Load Torque Theta_r Tload wr Saturation Model Lm1 f lux*wb qs Lls1 Motor / Stationary Stationary / Controller Ixs_current_sense Ixs Iqs_cont Ixs Iys Ids_cont Iqs* Ids* swe_sat Llr Lm swe calculator f lux*wb ds Llr1 Iqs_mot Ids_mot Iys_current_sense ange_cont Iys ange_mot 1 s slip angle Ang Figure 8 Simulink® Model for 58 Tooth Sensor Study ELECTRICAL DRIVE SYSTEM MODELLING TOOLS ® An electrical drive model was created in the MATLAB ® Simulink simulation environment to investigate the effect of various sensor resolutions and control strategies on the engine cranking performance and other operating modes. Models were formed for the electric machine, controller, inverter, position sensor and battery that capture the dominant effects of the system. The machine was modeled in a synchronous reference frame to minimize simulation time. Machine magnetic saturation as well as inverter voltage saturation were captured in the model due to the importance of these effects. In addition, the effects of the sensor resolution as well as controller and machine position error are modeled. If required, the effect of digital controls and sample delays can be included at the expense of complexity and increased simulation time. ® The Simulink based electrical drive system model for the 58X position sensor is shown in figure 8. From the electrical system model, the effect of operating with a normal 58X sensor was evaluated. The low sensor resolution was seen to make low speed operation problematic due to large torque ripple and reduction in the mean torque applied by the machine. The large torque ripple can excite driveline resonance while the reduction in the mean torque can, in the extreme, make the cranking event performance unacceptable. Delphi overcame these issues by developing and implementing algorithms that provides for near ideal performance by processing the information from the 58x sensor. Figure 9 shows the predicted dynamic starting torque, speed and the magnified mechanical position (times 10) for the crank event using the new sensor and sensor /control strategy. 250 Torque 200 Nm, RPM, 10*Degree the angular error caused by encoder resolution. The position sensor is normally selected based on the desired performance of the drive. Due to the cost and reliability challenges of this application and the desire not to have a separate sensor for machine and engine control, it was decided that the only acceptable sensor was the engine’s 58X crank position sensor. A simulation tool was used to examine and performance tradeoffs for a design that utilizes a sensor with this range of resolution, 150 RPM 100 Mech Angle x 10 50 0 0.29 0.3 0.31 0.32 0.33 Time (s) 0.34 0.35 0.36 0.37 Figure 9 Predicted Start Characteristic Using New Sensor and Control Strategy Commanded Motor Torque 250 200 200 150 150 100 100 COLD CRANKING PERFORMANCE WITH 58X SENSOR AT –29 DEGREES CENTIGRADE 50 50 0 Crankshaft Angle 200 1 180 0.75 160 0.5 140 0.25 120 0 100 -0.25 80 -0.5 60 -0.75 40 -1 20 -1.25 0 50 100 150 200 250 0 300 Time (Milliseconds) Figure 10 Phase Currents for the First 175 Degrees of Rotation Figure 10 displays the phase currents and crankshaft position of a cold cranking event. The crankshaft begins rotation at a time of 38 milliseconds. This is when the electric machine torque exceeds the static torque of the system. The engine then begins to increase in speed. The engine speed will continue through a profile of acceleration and de-acceleration as the crankshaft position propagates through a series of compression and power strokes. The values of commanded electric motor torque and engine speed were recorded during the cold cranking event. The maximum commanded torque during cranking was 203 Nm. The acceleration is uniform during the first few degrees of rotation indicating a constant load torque. The electric machine is operated in the field weakened region during the 150 to 225 millisecond time period. This is observed when the phase current is reduced during this time period. 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Time (seconds) Figure 11 Engine Speed and Commanded Electric Motor Torque for Cold Cranking The plots in figure 11 displays the engine speed and the commanded electric machine torque for the cold cranking event. The engine accelerates until the first compression stroke results in a significant deceleration. The minimum engine speed value is slightly less than 50 rpm for this compression stoke. This deceleration profile is the result of the load torque being greater than the motor torque. This corresponds with the calculated load torque of 225 Nm for the first compression stroke. The process repeats for the next compression stroke. The deceleration for the second compression stroke is more than the initial stroke. The peak load torque is slightly greater than the initial compression stroke. The following compression strokes will have lower peak load torque values. This is due to two factors. The power stoke from the previous compression stroke is contributing more acceleration and the intake manifold pressure is dropping due to the previous compression strokes. The lower manifold pressure will lower the amount of torque required for the next compression strokes. Figure 12 shows that the power from the battery pack will vary as the engine progresses though the successive compression and power strokes. The battery pack current and power was recorded during the cold cranking event. Motor_RPM Battery Pack_Amps Battery Power (watts) 400 11000 350 10000 300 9000 250 8000 200 7000 150 6000 100 5000 50 4000 0 Watts Phase B Crankshaft Angle (Degrees) Per Unit Rated Current Phase A 1.25 0 0.0 RPM or Battery Amps The complete ISG system was tested at –29 degrees centigrade. The engine, transmission, electric machine, and batteries were soaked overnight in the cold chamber. The electric machine was commanded to rotate the engine and transmission to determine cranking performance. The engine was not fueled during these cranking tests. The engine is normally fueled after the first complete crankshaft rotation. The phase currents and crankshaft angle were recorded during the cranking event. The control system tracks the rotation of the crankshaft from the sensor position pulses. Commanded Torque (Nm) Motor_RPM 250 RPM Based on the engine thermal test data as well as the electric drive simulations, a starting strategy was formed that could be easily incorporated into a standard vector controlled drive with minimal modification. The starting algorithm was incorporated into Delphi Automotive’s electric machine controller. Using the machine control strategy developed from simulation, testing of the actual engine crank events were performed over a range of temperatures. 3000 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Time (seconds) Figure 12 Engine Speed, Battery Current and Battery Power During Cold Cranking. The engine speed, battery current, and the battery power during the cold cranking event is shown in figure 12. The battery pack power had a peak value of 9700 watts and the battery pack current had a peak value of 370 amperes. ACKNOWLEDGMENT Several cranking events were performed at –29 degrees Centigrade. During one of those events, the calculation error for the area of the two missing teeth was observed. The sensor algorithm will project the rotor position pulses for the area of the two missing teeth from the acceleration profile of the two previous teeth. If the speed profile does not change appreciably during the 18 mechanical degrees of rotation where the gap of the two missing teeth reside, then the calculation error will be undetected. This was the case in all but one of the cold cranking events that were recorded. CONTACT 1 Phase Current Per Unit Phase B Crankshaft Angle Acceleration Error from Gap with Missing Teeth 46 0.75 44 0.5 42 0.25 40 0 38 -0.25 36 -0.5 34 -0.75 32 -1 150 160 170 180 190 Crankshaft Angle (Degrees) Phase A 30 200 Time (Milliseconds) Figure 13 Cold Cranking Event with Acceleration Error The acceleration error shown in figure 13 caused the crankshaft angle count to be held constant for 3.8 milliseconds. This caused some disturbance in the phase current control. The disturbance was small and brief enough that the torque production of the electric machine did not affect the cranking event significantly. CONCLUSION The Energen10® ISG system was designed to perform the cold cranking function at –29 degrees centigrade. The test results verify that the system did perform the cold cranking task and maintain control of the electrical machine during operation. The ISG system has demonstrated the ability to accelerate the engine, motor and transmission to a sufficient speed to ensure the engine combustion process for starting at very low temperatures. The simulation results predicted that the system would be able to perform the cold cranking function. The electric machine control algorithm was designed to use the 58X sensor that is widely used in engine control applications which reduces the ® Energen10 ISG system cost. Rassem Henry of Delphi Research performed the modeling of the gas pressure and the friction torque for the 4.0 liter engine. Gerald T. Fattic holds a BS degree in Electrical Engineering from Purdue University. He has worked for 29 years in the automotive field. Gerald currently works as a development engineer in the Advanced Systems Control group at the Energenix Center of Delphi Automotive Systems. He can be contacted at by e-mail:
[email protected] James E. Walters holds a BSEE from Purdue University and MSEE degree from the University of Wisconsin. James is the group leader of Advanced Electric Machine Controls at the Energenix Center of Delphi Automotive Systems. James can be contacted by e-mail:
[email protected] Fani S. Gunawan holds BS and MS degrees in Electrical Engineering from Ohio State University. He joined Delphi Automotive Systems in 1998. Fani currently holds the position of Project Engineer, developing control algorithms and testing electric machines. Fani can be contacted by e-mail:
[email protected]