Cold Starting Performance Of A 42-volt Integrated Starter Generator System

Transcript

SAE TECHNICAL PAPER SERIES  2002-01-0523 Free For 30 Days Cold Starting Performance of aRead 42-Volt Integrated Starter Generator System Gerald T. T. Fattic, James E. Walters and Fani S. Gunawan Delphi Automotive Systems, Energenix Center  DISCOVER NEW BOOKS  READ EVERYWHERE  BUILD YOUR DIGITAL READING LISTS Reprinted Reprint ed From: 42 Volt Volt Technol Technology ogy 2002 (SP–1661) SAE 2002 World Congress Detroit, Michigan March 4-7, 2002 400 Commonweal Commonwealth th Drive, Drive, Warrendal Warrendale, e, PA 15096-0001 15096-0001 U.S.A. U.S.A. Tel: (724) 776-4841 776-4841 Fax: (724) 776-5760 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, 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 C ustomer Sales and Satisfaction Department. Read Free For 30 Days To request permission to reprint a technical paper or permission to use copyrighted SAE publications in other works, contact the SAE Publications Group.  DISCOVER NEW BOOKS  READ EVERYWHERE  BUILD YOUR DIGITAL READING LISTS 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. 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. Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group. Persons wishing 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 Performance of a 42-Volt 42-Volt Integrated Starter Star ter Generator System Gerald T. Fattic, James E. Walters and Fani S. Gunawan Delphi Automotive Systems, Energenix Center Read Free For 30 Days Copyright © 2002 Society of Automotive Engineers, Inc.  DISCOVER NEW BOOKS 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. system. Test results performed at –29 degrees centigrade for the cranking of a gasoline 4.0 liter, V-6 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.  READ EVERYWHERE  BUILD YOUR DIGITAL READING LISTS across the operating temperature range to enable the engine the opportunity to start. 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. engine. It must produce sufficient 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 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 specifications are as follows: Bore: 100.4 mm Stroke: 84.4 mm Compression Ratio: 9.7 The Transmission specifications are as follows: st 1  ge  gear ratio: 2.47 nd 2  ge  gear ratio: 1.86 rd 3  ge  gear ratio: 1.47 th 4  ge  g ear ratio: 1.00 th 5  ge  g ear ratio: 0.75 150 100   m    N   e   u 50   q   r   o    T 0 Peak Torque for the First Compression Stroke Peak Torque for the Second Compression Stroke -50 -100 Read Free For 30 Days -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 DETERMING DYNAMIC FRICTION TORQUE following compression strokes have lower peak torque values because the other cylinders are finishing their  To determine the dynamic friction torque of the engine power stroke when the compression stroke is occurring and transmission, the engine and transmission and the intake manifold pressure has been reduced by LISTS DISCOVER NEW BOOKS  READ EVERYWHERE  BUILD YOUR DIGITAL READING combination was mounted with an ISG and  soaked in a the earlier compression strokes. thermal chamber. The chamber chamber was cooled cooled to -29 -29 degrees centigrade. The spark spark plugs were removed CRANKSHAFT POSITION AT SHUTDOWN from the engine. This was done to release release any pressure from the cylinders when the crankshaft was rotated. The When the engine comes to rest after running, the torque required for turning the crankshaft several crankshaft position settles in the area of minimum rotations was recorded. The average value of 55 Nm crankshaft torque. The area where the crankshaft torque was recorded and used as the dynamic friction torque is less than the friction torque is the position where the value for low speed and low low temperature operation. The engine will come to rest after running. Figure 3 shows process was repeated at 25 degrees centigrade to the crankshaft angles where the crankshaft torque is record the dynamic dynamic friction torque. The average value of  less than the dynamic friction torque of 40 Nm. 40 Nm was recorded as the dynamic friction torque at room temperature. In order to understand the requirements for engine cranking, an understanding of the load characteristics of  the engine-transmission combination was required. 200 Nm Temperature -29 degrees centigrade 25 degrees centigrade 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   e   u   q   r 100 Nm   o    T    t    f   a    h   s    k   n 50 Nm   a   r    C 0 Nm -50 Nm 240 300 360 420 480 540 600 660 720 Cranshaft angle (Degrees) Figure 3 Predicted Crankshaft Stopping Positions Positions The position where the engine-starting event occurs is the position 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 Read Free 30 Days system consists of theFor electric machine, position sensor, current sensors, inverter, inverter, and machine controller. controller. Figure 6 shows a simplified electrical 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. DISCOVER  BUILD YOUR DIGITAL READING LISTS I b A digital sensing device was used to detect theNEW BOOKS  READ EVERYWHERE ∆θROTOR  movement of the 58-tooth wheel. The digital sensor was Ia 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  Figure 6 Simplified Electrical Drive System speed information to be used in the control algorithm. The 58X sensor system was chosen because it is a The machine controller receives a torque command from common system used for engine controls for many a vehicle level controller and applies the appropriate different engine models. The engine crankshaft sensor  voltage to the electric machine to create the desired can be used for the electric machine control, eliminating torque. The machine controller uses vector control the need for a separate sensor. techniques in order to have good dynamic performance and disturbance rejection capabilities across a wide ELECTRIC MACHINE DESIGN speed range. A simplified vector control strategy is shown in figure 7. 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. Torque* Iq  Table Speed Id Vq Current Controller  Vd Coordinate Trans. PWM Generation Inverter  Electric Machine ∆θROTOR  250 I b 24 Volts 28 Volts 32 Volts Coordinate Trans. 200 Ia θ SYNC Temperature   m150    N   e   u   q   r   o    T100 Slip Angle Calculator  θSLIP + + Σ θROTOR  Angle Processing Speed 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 with a wide wide field-weakened 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 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 Vqs Vq Vq_lim Vqs_mot Iqs* Sum PI Vd Vd_lim ange_cont Ids_err Vds Ids* Sum1 Post-Processing Stationary / Motor Vx s Vx s Vds_mot [T',Ids'] Position Sensor an d wb_flux_qs wb_flux_ds Vqs_cont M ux IND_SFUNCII Vy s Vy s ange_mot Inverter Model Vds_cont Demux IM Motor Model Mux1 PI 1 theta_e wr theta_r Iq s Id s Theta_r Position Iqs Load Torque Theta_r Tload wr Saturation Model Lm 1 f lux*wb qs Lls1 Ixs_current_sense Ix s Iqs_cont Ix s Iy s Ids_cont Iqs* Ids* swe_sat Llr Lm swe calculator f lux*wb ds Llr1 Read Free For 30 Days Motor / Stationary Stationary / Controller Iqs_mot Ids_mot Iys_current_sense ange_cont Iys ange_mot 1 s slip angle An g Figure Figure 8 Simulink Model for for 58 Tooth Sensor Sensor Study Study  DISCOVER NEW BOOKS the angular angular error caused by encoder resolution. resolution. The position sensor is normally selected based on the desired performance performance of the 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, 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 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.  READ EVERYWHERE  BUILD YOUR DIGITAL READING LISTS From the electrical system model, the effect of operating with a normal 58X sensor 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 applied by the machine. 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   e 150   e   r   g   e    D    *    0    1  ,    M    P    R  , 100   m    N RPM Mech Angle x 10 50 0 0. 29 0 .3 0. 31 0. 32 0. 33 0. 34 0. 35 0. 36 0.37 Time (s) Figure 9 Predicted Start Characteristic Using New Sensor and Control Strategy 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 drive with minimal modification. 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. COLD CRANKING PERFORMANCE WITH 58X SENSOR AT –29 DEGREES CENTIGRADE Motor_RPM 250 200 200    )   m    N    (   e 150   u   q   r   o    T    d   e    d 100   n   a   m   m   o    C 50 150    M    P    R 100 50 Read Free For 30 Days 0 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 was commanded commanded to 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. Phase A Phase B Crankshaft Angle 1.25 1 0.75    t   n 0.5   e   r   r   u    C 0.25    d   e    t 0   a    R    t    i -0.25   n    U   r   e -0.5    P -0.75 Commanded Mo Motor To Torque 250 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0 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 withYOUR the calculated load LISTS  DISCOVER NEW BOOKS READ EVERYWHERE  BUILD DIGITAL READING torque of 225 Nm for the first compression stroke. The process repeats for the next compression stroke. The 200 deceleration for the second compression stroke is more 180 than the initial stroke. The peak load torque is slightly 160    )   s greater than the initial compression stroke. The following   e 140   e   r   g compression strokes will have lower peak load torque   e 120    D    ( values. This is due to two factors. factors. The power stoke from   e    l   g 100   n the previous compression stroke is contributing more    A    t    f acceleration and the intake manifold pressure is 80   a    h   s    k dropping due to to the previous previous compression strokes. The 60   n   a   r lower manifold pressure will lower the amount of torque    C 40 required for the next compression strokes. -1 20 -1.25 0 0 50 100 150 200 250 300 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. 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 Pa Pack_Amps Battery Po Power (w (watts) 400 11000 350 10000   s   p   m    A250   y   r   e    t    t 200   a    B   r   o 150    M    P    R100 300 9000 50 4000 8000 7000   s    t    t   a    W 6000 5000 0 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. 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 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 in all but one of the cold cold cranking events that were recorded. P h as e A Phase B Crankshaft Angle Acceleration Error from Gap with Missing Teeth 1 0.75 44    ) Rassem Henry of Delphi Research performed the modeling of the gas pressure and the friction torque for  the 4.0 liter engine. CONTACT Gerald T. Fattic holds a BS degree in Electrical Engineering from Purdue University. University. He has worked for  for  29 years in the automotive field. field. Gerald currently currently works as a development engineer in the Advanced Systems Read Free For 30 Days 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]  [email protected] 42   s   e    t 0.5    i   n    U   r   e 0.25    P    t   n   e 0   r   r   u    C   e -0.25   s   a    h    P   e   r   g   e    D    (   e    l   g   n    A    t    f   a    h   s    k   n   a   r    C Fani S. Gunawan holds BS and MS degrees in Electrical Engineering from Ohio State University. He joined Delphi 38  DISCOVER NEW BOOKS  READSystems EVERYWHERE  BUILD DIGITAL READING  Automotive in 1998. Fani YOUR currently holds the LISTS 36 position of Project Engineer, developing control algorithms and testing electric machines. Fani can be 34 contacted by e-mail: fani.gunawan@delphi [email protected] auto.com 40 -0.5 -0.75 -1 150 46 ACKNOWLEDGMENT 32 160 170 180 190 30 200 Time (Milliseconds) (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 disturbance was small 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.