1 of 13
`
`FORD 1355
`
`

`
`Design Innovations in
`Electric and Hybrid
`Electric Vehicles
`
`SP-1089
`
`
`
`GLOBAL MOBILITY DATABASE
`
`All SAE papers, standards, and selected
`books are abstracted and indexed in the
`Global Mobility Database.
`
`Published by:
`Society of Automotive Engineers, Inc.
`400 Commonwealth Drive
`Warrendale, PA 15096-0001
`USA
`
`Phone: (412) 776-4841
`Fax: (412) 776-5760
`February 1995
`FORD 1355
`
`2of13
`
`2 of 13
`
`FORD 1355
`
`

`
`Permission to photocopy tor internal or personal use. or the internal or personal use of specific
`clients, is granted by SAE for libraries and other users registered with the Copyright Clearance
`Center (CCC), provided that the base fee of $6.00 per article is paid directlyto CCC, 222 Rosewood
`Drive. Danvers, MA 01 923. Special requests should be addressed to the SAE Publications Group.
`1 -56091 -639-7/95$6.00.
`
`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.
`
`ISBN 1-56091 -639-7
`SAE/S P-95/1 089
`
`Library of Congress Catalog Card Number: 94-74742
`Copyright 1 995 Society of Automotive Engineers, inc.
`
`Positions and opinions advanced in this pa-
`per are those of the author(s) and not neces-
`sarily 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 it it is published in
`SAE Transactions. For permission to pub-
`lish this paper in full or in part, contact the
`SAE Publications Group.
`
`3of13
`
`Persons wishing to submit papers to be con-
`sideredforpresentation orpublication through
`SAE should send the manuscript or a 300
`word abstract of a proposed manuscript to:
`Secretary, Engineering Meetings Board, SAE.
`
`Printed in USA
`
`FORD 1355
`
`3 of 13
`
`FORD 1355
`
`

`
`PREFACE
`
`The papers in this SAE special publication, Design Innovations in Electric and Hybrid Electric
`Vehicles (SP—1089), cover technology for both electric and hybrid electric vehicles. As is
`well accepted, to have a good hybrid electric vehicle requires first having a good electric
`vehicle. Major manufacturers have initiated the effort required to take electric vehicle
`technology from the laboratory through the required development steps to provide an
`automotive product. This work will provide a foundation for the development of hybrid
`electric vehicles.
`
`Unique engines, unique operating strategies and unique packaging solutions will all be the
`hallmark of successful hybrid electric vehicles. Over the past several years, the hybrid-
`electric vehicle concept has been gaining attention as a possible way to reduce emissions
`and increase fuel efficiency compared to a conventional vehicle. Hybrid—electric vehicles
`contain a hybrid power supply system - one that incorporates a minimum of two
`independent power sources to supply the drivetrain. The main advantage of this concept is
`it permits flexibility in power system design and power distribution between sources. This
`versatility enables greater flexibility in designing the powertrain to meet the required
`performance of the vehicle. The challenge is to combine the different power sources such
`that the advantages outweigh the increased cost of this configuration. These papers cover
`some of the latest technical developments related to the engine aspect of hybrid-electric
`vehicle development. Topics included in this year's session are: development of hybrid-
`electric vehicle design code; optimization of vehicle and engine control strategies; and novel
`engines for hybrid-electric vehicles.
`
`Also critical to the automotive products of the future is the engineering talent required to
`produce the innovative designs. One of theprograms aimed at exciting students to the
`new automotive opportunities is the HEV Challenge. This program is well represented by
`papers in this book. Experience has shown that the HEV Challenge is not only motivating
`students, but also surfacing innovative automotive engineering solutions to difficult
`problems. We are pleased to be able to share some of this excitement through this
`publication.
`
`All of these subjects and the design methodologies required to achieve them, are covered
`by papers in this collection. We hope that this year's papers will trigger your imagination
`and provide the foundation for innovative developments that will help electric and hybrid
`electric vehicles play an important role in our transportation system.
`
`Bradford Bates
`
`Ford Motor Co.
`
`Chairman, Electric Vehicle Committee
`
`Frank Stodolsky
`Argonne National Laboratory
`
`Session Organizers
`
`4 of 13
`
`‘
`
`FORD 1355
`
`4 of 13
`
`FORD 1355
`
`

`
`9501 76
`
`9501 77
`
`9501 78
`
`9501 79
`
`950491
`
`950492
`
`950493
`
`950495
`
`950955
`
`950957
`
`TABLE OF CONTENTS
`
`Technical Analysis of the 1994 HEV Challenge ......................................1
`Nicole M. LeBlanc, Michael Duoba, Spencer Quong,
`Robert P. Larsen, and Marvin Stithim
`Argonne National Lab.
`William Rimkus
`Ford Motor Co.
`
`Testing Hybrid Electric Vehicle Emission and Fuel Economy
`at the 1994 DOEISAE Hybrid Electric Vehicle Challenge ..................... 13
`Michael Duoba, Spencer Quong, Nicole LeBlanc, and Robert Larsen
`Argonne National Lab.
`
`Electric Vehicle Performance In 1994 DOE Competitions ....................23
`Spencer Quong, Michael Duoba, Robert Larsen, Nicole LeBlanc,
`Richard Gonzales, and Carlos Buitrago
`Argonne National Lab.
`
`Design and Analysis of a Hybrid Electric Vehicle chassis ..................31
`John G. Aemi
`
`Prince Corp.
`Clark J. Radcliffe and John L. Martin
`
`Michigan State Univ.
`
`A Hybrid Vehicle Evaluation Code and its Application to
`Vehicle Design ............
`.........................
`..................
`.....................43
`Salvador M. Aceves and J. Ray Smith
`Lawrence Livermore National Lab.
`
`Controlling a CVT-Equipped Hybrid Car ................................................53
`Andreas Schmid, Philipp Dietrich, Simon Ginsburg, and Hans P. Gearing
`Swiss Federal Institute of Technology (ETH)
`
`The Effects of APU characteristics on the Design of Hybrid
`Control Strategies for Hybrid Electric Vehicles ....................................65
`Catherine Anderson and Erin Pettit
`Aerovironment
`
`ECTAMW, A Continuous Combustion Engine for Hybrid
`Electric Vehicles .................................................
`.................
`W. Robert Palmer and J. Dale Allen
`
`............ ..73
`
`Spread Spectrum, Inc.
`
`Computerized Speed Control of Electric Vehicles ................................89
`Zheiun Fan, Yoram Koren, and David Wehe
`The University of Michigan
`
`The Effect of Regenerative Braking on the Performance and
`Range of the AMPhibian ll Hybrid Electric Vehicle ...............................95
`Gregory W. Davis and Frank C. Madeka
`United States Naval Academy
`
`5of13
`
`FORD 1355
`
`5 of 13
`
`FORD 1355
`
`

`
`950958
`
`Fuel Economy Analysis for a Hybrld Concept Car Based on a
`Buffered Fuel-Engine Operating at an Optimal Point .........................103
`Marc Ross and Wei Wu
`
`University of Michigan
`
`950959
`
`A comparison of Modeled and Measured Energy Use In Hybrid
`Electric Vehlcles ............................................
`............................ 119
`
`Matthew Cuddy
`National Renewable Energy Lab.
`
`951069
`
`Comparison Between On-Road and Simulated Performance of
`the KEV Electrlc Vehicle ....................................................................... 129
`
`C.H. Kim, S. B. Koh, and E. NamGoong
`Kia Motors Corp.
`
`6 of 13
`
`FORD 1355
`
`6 of 13
`
`FORD 1355
`
`

`
`The Effects of APU Characteristics on the Design of
`Hybrid Control Strategies for Hybrid Electric Vehicles
`
`Catherine Anderson and Erin Pettit
`Aerovlronment
`
`950493
`
`ABSTRACT
`
`A hybrid control strategy is an algorithm that determines
`when and at what power level to run a hybrid electric vehicle's
`auxiliary power unit (APU) as a function of the power demand
`at the wheels, the state of charge of the battery, and the current
`power level of the APU. The design of this strategy influences
`the efficiency of the overall system. The strategy must
`balance the flow of power between the APU, the battery, and
`the motor, with the intent of maximizing the average fuel
`economy without ovcrstressing the battery and curtailing its
`life.
`
`The development of a system's powertrain components
`and the design of an optimum control strategy for that system
`should be concurrent to allow tradeoffs to be made while the
`designs are still fluid. An efficient optimization process must
`involve all aspects of the system, including costs, from the
`beginning.
`In this paper, we explore the methodology behind the
`design of a hybrid control strategy. We also discuss the APU
`and battery design characteristics that are crucial to the
`strategy design, focusing on the interdependence of these
`design characteristics within the entire system. Finally, we
`propose a process for the development of an optimized hybrid
`powertrain and the corresponding control algorithm.
`
`INTRODUCTION
`
`A "hybrid" vehicle usually refers to one that incorporates
`a minimum of two independent power sources to supply the
`drivetrain. One of the primary advantages of this dual power
`supply system is it allows flexibility in power distribution
`between sources. This versatility enables greater optimization
`of the vehicle powertrain to meet the required performance of
`the system. In order to profit from such system flexibility, one
`must integrate into the system an intelligent control strategy
`that uses each component to the overall system's best
`advantage.
`A hybrid control strategy is an algorithm that determines
`how power in a hybrid powertrain should be distributed as a
`function of the vehicle parameters (power demand, battery
`state of charge (SOC), component temperatures, etc.) and of
`7of13
`
`component characteristics. One must develop this strategy
`carefully as part of the vehicle design process from the
`beginning. While the strategy determines the best operating
`points for the components, the range of available component
`characteristics provides the limits within which the strategy
`must operate.
`This paper explores the iterative process of concurrent
`powertrain component and control strategy design with an
`emphasis on optimizing the system as a whole. We focus
`primarily on the auxiliary power unit and the characteristics of
`the powertrain components that drive the strategy design.
`
`HYBRID VEHICLE CONCEPT
`
`Hybrid vehicles can be divided into two main categories:
`parallel, in which both systems are used to mechanically drive
`the wheels; and series, where the power supply systems are
`coupled directly to a power bus which then transfers power to
`the wheels.
`
`SERIES SYSTEM - The philosophy behind a series
`hybrid vehicle lies in its combination of a primary and a
`secondary energy conversion. In the primary conversion, an
`APU converts a highly transportable, stable chemical fuel to
`mechanical energy (or directly to electrical energy in certain
`cases) and, subsequently, to electrical energy. The most
`frequently considered APUs for hybrid systems include
`various internal and external combustion engines and fuel
`cells. This primary conversion device can be decoupled from
`the wheel power demand (unlike the engine in a conventional
`car) as a Load Leveling Device (LLD), which acts as an
`energy buffer, is included in the system. This LLD alternately
`stores energy (either directly from the primary conversion at
`low wheel power requirements or from the kinetic energy of
`the decelerating vehicle) and provides the propulsion motor
`with energy when the demand exceeds the APU power output.
`Some LLDs that have been proposed for use in hybrid vehicles
`include batteries, supercapacitors, hydraulic andlor pneumatic
`storage devices, and flywheels.
`The secondary conversion, occurring in the inverter and
`motor, transforms the electrical energy from either source into
`the mechanical energy that drives the vehicle. Figure 1 is a
`schematic of the energy flow within the vehicle.
`
`65
`
`FORD 1355
`
`7 of 13
`
`FORD 1355
`
`

`
`a range extender is small compared with the power demand, it
`is most often run at its maximum power level, and hybrid
`control strategies are fairly simple.
`The primary disadvantage of the series hybrid system in
`most cases is the extra inefflciencies included in converting
`the mechanical power output from the APU into electrical
`power and then back into mechanical power. Often, however,
`the increased flexibility of the system offers more optimized
`components that overcome this disadvantage.
`
`PARALLEL SYSTEM- In a parallel hybrid vehicle,
`there is a direct mechanical connection between the APU and
`the wheels through a transmission. As shown in figure 3, the
`electric propulsion system may either drive the same set of
`wheels as the APU through the transmission (Option 1), or
`drive the other set of wheels directly (Option 2).
`
` 4 4 9
`
`+ e
`
`v + o
`{4}1+3
`
`
`v‘~’o*r+v+o"o"+*+¢‘+__+_
`.
`....;.;.‘
`-.:9
`
`9 F
`
`igure 3: Parallel hybrid vehicle component
`configuration.
`
`The main advantages of the parallel configuration over the
`series is that the power from the engine is used directly by the
`drivetrain with no alternator or inverter losses. However,
`because the APU is directly coupled to the wheels, the APU
`speed is determined by the vehicle speed and the transmission
`ratio. This direct coupling limits the flexibility of hybrid
`strategy design, and (without a novel clutched transmission)
`forces the APU to idle when the vehicle is at rest.
`A parallel hybrid does have an efficiency advantage when
`the vehicle spends a majority of its driving time at a
`substantial cruise, but a vehicle that is operated on a "city
`driving" profile will lose this transmission efflciency
`advantage to inefficiencies in the APU engine. In addition, if
`the front and rear axles of the vehicle are driven by different
`power sources, the vehicle may exhibit changes in handling
`characteristics as the power distribution between the sources is
`adjusted.
`
`The thought processes presented in this paper are
`sufficiently general that they can be applied to any type of
`vehicle. To fully explore the flexibility allowed by the hybrid
`system, we focus on the design of a strate|g_y for the most
`ORD 1355
`
`66
`
`Since all the power sources and sinks are directly coupled
`by a DC power bus, control of the entire system can be
`achieved by simply commanding the APU output. The
`accessory and wheel loads pull required power off the bus
`with the LLD supplying the balance of power in the system.
`
`
`
`Figure 1: Series hybrid vehicle component configuration.
`
`ICE Vehicle
`
` Power Assist
` IncreasingAPUPower
` Pure Electric
`
`Range Extender
`
`Vehicle
`
`Increasing Battery Storage -Zf
`
`Figure 2: Comparison Chart of Power Assist and Range
`Extender Series Hybrid Vehicles
`
`In addition, the series hybrid design may fall into one of
`two categories: "power assist" or "range extender" (see figure
`2). A power assist hybrid uses the LLD to manage the power
`output from the APU to maximize efficiency and emissions in
`the APU. The usable storage capacity of the LLD is quite
`small (on the order of 1-5 kWh), and the APU must be capable
`of providing the maximum sustained power the vehicle is
`expected to need, with the LLD providing peak powers and
`transients. A range extender hybrid uses a very small APU
`with a substantial LLD such that the vehicle will perform
`similarly to a pure electric vehicle, with the additional small
`power source simply extending the range. Since the APU for
`
`8of13
`
`8 of 13
`
`FORD 1355
`
`

`
`versatile layout: the power assist hybrid. For simplicity, we
`use the example of a generalized IC engine and Pb-Acid
`battery for the APU and battery, respectively, as a focus for
`the discussion.
`
`Constant APU Mode
`
`APU Power
`
`HYBRID CONTROL STRATEGIES
`
`There are two distinct extremes in the spectrum of control
`strategies. One is a system that uses a "thermostat" algorithm
`to command the APU (i.e. the APU is turned on to a constant
`power level when the SOC of the LLD is below a certain
`lower threshold, and off when the SOC exceeds an upper
`threshold). In this mode, the LLD must accommodate all the
`transient power requirements. Although the APU may be
`operating at its most efficient point, the losses in the LLD
`from excessive cycling may surpass the savings from an
`optimized APU. For the example wheel power curve shown in
`figure 4, figure 5 shows the corresponding APU and LLD
`power requirements generated by a thermostat mode.
`
`Federal Urban Cycle
`
`Wheel Power
`
`
`
`Power(kW)
`
`aésoass
`
`1500 1700 1900 2100 2300 2500 2700 2900 3100 3300 3500
`
`LLD Power
`
`(kW)sssosss
`Power
`
`1500 1700 1900 2100 2300 2500 2700 2900 3100 3300 3500
`'I'1me(s)
`
`Figure 5: The APU and LLD power outputs that satisfy
`the wheel requirements using a constant APU thermostat
`strategy.
`
`APU follower Mode
`
`APU Power
`
`I500 I700 I900 2100 2300 2500 2700 2900 3100 3300 3500
`Time(s)
`
`Figure 4: The power required at the wheels for a segment
`of the federal urban drive (LA4).
`
`(kW) oU:SG8U.
`ll
`Power
`
`1500 1700 1900 2100 2300 7500 2700 2900 3100 3300 3500
`
`The other extreme commands the APU to follow the
`
`actual wheel power whenever possible (similar to a
`conventional automobile). Using this strategy, the LLD
`cycling will diminish, and the losses associated with charge
`and discharge will be minimized. The APU, however, must
`then operate over its entire range of power levels and perform
`fast power transients, both of which can adversely affect
`engine efficiency and emissions characteristics. Figure 6
`shows the APU and LLD power requirements generated by
`this "following" mode for the same wheel power curve shown
`in figure 4. It should be noted that this is the mode a parallel
`hybrid vehicle always uses.
`For most of the APUs and LLDs under consideration,
`neither of these strategies would be the optimum strategy. The
`ideal hybrid control strategy is one that minimizes the
`combined inefficiencies of both the APU and the LLD while
`
`meeting the desired performance and the emission limits (as
`well as any other specific system characteristics that are being
`used as measures of design merit). The optimum strategy is
`highly dependent on the characteristics of the powertrain
`components and the planned use of the vehicle. Unfortunately
`as one attempts to optimize a system, the characteristics of the
`components begin to eonflict, driving the strategy in different
`directions.
`
`LLD Power
`
`0
`§—2o
`A’. 40
`-so
`
`1500 1700 1900 2100 2300 2500 2700 2900 3100 3300 3500
`Time (s)
`
`Figure 6: The distribution of power for a load following
`APU.
`
`COMPONENT CHARACTERISTICS AND DESIGN
`TRADE-OFFS
`
`LLDS - The LLD (in this case a battery) must be the most
`accommodative element in the powertrain. When there is a
`large power demand or production from the wheels (as during
`hard acceleration or braking), it must supply or accept the
`power required. In a hybrid application, the battery pack
`generally has lower capacity than it would have for a pure
`electric vehicles (particularly for a power assist hybrid where
`
`9of13
`
`67
`
`FORD 1355
`
`9 of 13
`
`FORD 1355
`
`

`
`the APU is of considerable size). To maintain the same
`performance, therefore, the power density must be greater. In
`addition, the state of charge of the battery can be significantly
`affected during a short acceleration or deceleration so that the
`srnall-scale charge/discharge period (or "microcycling") that
`the pack sees is a more significant percentage of its capacity.
`Differing control strategies can place varying demands on the
`cycling of the battery. Using the thermostat APU strategy, the
`battery would be required to cycle at the frequency of the
`wheel power demand, while the follower APU strategy would
`only require the battery to cycle when the wheel power
`demand exceeds the APU power capability.
`There are several characteristics of the battery that one
`must keep in mind when trying to quantify tradeoffs between
`the battery and the rest of the system:
`the charge/discharge
`efficiency, the total capacity of the pack, the transient
`capabilities, and, the hardest to determine, the life of the
`batteries.
`
`Qharggznjsgharge Efficiencies - A battery is most
`efficient within a range of SOCs that minimizes its charge and
`discharge resistances. In figure 7, one can see the general
`shape of a Pb-Acid battery's internal resistance versus state of
`charge curves for charging and discharging the battery. A
`balance point must be chosen on these curves to minimize
`resistive losses, yet still leave room for power peaks (both
`motoring and regenerative braking) at the wheels. This tends
`to push the strategy design to keep the SOC within the 50-70%
`region for minimum losses in both charge and discharge. This
`leaves enough capacity to handle an extended period of battery
`discharge (such as during a long hill climb) and enough
`"headroom" to absorb a long period of charging such as that
`which occurs during a long downhill. If the SOC is not
`maintained within the 50-70% region, the performance may be
`compromised. This diminished performance may take the
`form of lost regenerative "energy or limited power output
`during accelerations.
`
`Capacigg - The capacity of the pack is comparatively easy
`to measure, and the effects of the change of capacity on the
`strategy are fairly intuitive. (It should be noted that the
`capacity at one rate of discharge is different from the capacity
`at another rate, and therefore the definition of "capacity" is
`subject to discussion.) In general, the larger the battery pack
`capacity, the more the vehicle can be run like an pure electric
`vehicle with the APU providing supplemental power. With a
`large capacity, it is easier to achieve the power required for
`standard driving, and the pack does not have to be so rigidly
`constrained to a small window of states of charge. A small
`pack, however, must be used almost exclusively as a short-
`term energy buffer without significant energy storage.
`
`Transient gapabjljtjgs - A battery can change power levels
`almost instantaneously, unlike the APU which is limited by its
`mechanical inertia. When the APU cannot respond quickly
`enough to fluctuations in power demand, the battery must
`make up the difference. The battery must be able to sustain
`output at a peak power during these transients until the APUs
`power output reaches the commanded power.
`
`Life - Unfortunately, most available data on battery life is
`of limited applicability to hybrid systems. The complexities of
`the reactions within batteries make it almost impossible to
`predict battery life except as a questionable extrapolation of
`empirical data. Although few quantitative predictions of
`10 of13
`
`battery life are available, some qualitative statements can be
`made:
`
`1. A lead acid battery will degrade more (per a
`throughput kWh) if cycled deeply (cycled through a
`wide range of SOCs) than shallowly. The long term
`effects of rnicrocycling (cycling over a small range of
`SOCs) are not fully known.
`2. A battery will last longer if it has lower energy
`throughput.
`3. Hard cycling (high power cycling), even hard
`microcycling, will shorten the life.
`
`Charge
`
`Discharge
`
`0.5
`
`0.45
`
`.9as
`
`.°wU!
`
`.°to
`
`.oSnou.
`
`
`
`
`
`InternalResistance(Ohms)
`
`.0 G
`
`0
`
`0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
`Slateof Charge (95)
`
`1
`
`Figure 7: The charge and discharge internal resistances
`versus the state of charge of a battery.
`
`Figure 8 shows the difference between the SOCs of the
`thermostat and follower extremes discussed above (see figures
`3-5) over multiple repetitions of the federal urban driving
`cycle. In the thermostat mode, the APU power output is
`greater than the average power for the cycle causing the state
`of charge to continue to increase until it reaches a defined
`maximum state of charge (in this case 80%) requiring the APU
`to turn off. The follower mode, on the other hand, provides
`only a slight constant increase in SOC due to the battery's
`absorption of regenerative energy during the cycle.
`The deep cycled battery might only last half as long as the one
`kept within a tight SOC window. However, the costs of
`replacing the battery versus the cost of building an APU
`capable of fast transient response(that can protect the battery)
`must be weighed.
`
`APUs - Because the APU is decoupled from the
`drivetrain, there is greater flexibility in its design. The design
`need not be performance driven as in conventional IC engines,
`but can be focused on other characteristics, such as emissions,
`that may be more important for the specific vehicle being
`designed. Most importantly, however, the APU characteristics
`must be chosen to complement the LLD requirements; thus,
`the need for a working strategy throughout the design process.
`Characteristics crucial to the design include maximum power
`output, transient capabilities, fuel efficiency, emissions
`characteristics, engine noise vibration harshness (NVH), and
`service life.
`
`F0 RD 1 355
`
`10 of 13
`
`FORD 1355
`
`

`
`Thermostat Mode
`
`@100
`" 80
`E 60
`2 40
`3 20
`g 0 1500
`
`Follower Mode
`
`33100
`‘J 80
`E so
`‘_ 40
`3 20
`:9:
`0 1500
`
`2500
`Time (5)
`
`2500
`Time (s)
`
`3500
`
`3500
`
`Figure 8: The state of charge during repetitions of the
`federal urban driving cycle with a constant APU at 20 kW
`in the thermostat mode (see figures 2,3) and in the
`follower mode(see figure 4).
`
` - The maximum power output of
`the APU will affect strategy design choices in a similar
`manner to the capacity of the battery. With a high power
`capability, one may design the strategy to operate more or less
`like a conventional car engine in a power following mode,
`whereas a low power capability will force the strategy to run
`the engine at its highest power level so that it can keep up with
`current demands and store extra energy for periods of high
`demand.
`
`AEL1 'I‘ran§i'entQapabj|i1ies - Mechanically. the transient
`capabilities of an engine are limited by the inertia involved in
`increasing or decreasing the engine speed. Although slower
`transients are desirable for reducing emissions, slow transients
`can curtail the life of the battery or potentially harm the
`engine. For example, slow transients can be a serious problem
`during a transition from a hard acceleration to a hard braking.
`If the APU has been commanded by the control strategy to
`supply a high power during an acceleration, and suddenly full
`regenerative braking is required, the LLD may not be able to
`accept the total power coming to it, unless the APU can reduce
`its power quickly. This limitation will cause a loss of much of
`the regenerative energy available. In an extreme case, the
`APU may be unloaded by an over-voltage condition, leading
`to potential overspeed. The APU control strategy must be
`robust, such that no combination of driver actions will result in
`damage to any drivetrain component.
`
`Fuel Efficiency - The fuel efficiency of an APU generally
`varies as a function of the power level. The specific fuel
`consumption (SFC) of an engine is typically best at middle
`power levels and worst at the low and high power extremes.
`The APU operating strategy that will maximize fuel efficiency
`is one that runs the APU primarily in the range of powers over
`which the SFC is best (often termed the engine's "sweet spot").
`The ratio of the highest power level to the lowest power level
`11 of13
`
`69
`
`used in the strategy is called the turn down ratio. The
`narrower this sweet spot is, and, thus, the smaller the most
`efficient turn down ratio is, the more the fuel efficiency
`requirements constrain the strategy toward a thermostat mode
`(see figure 9, a series of SFC graphs showing varying sweet
`spots.). Increasing the range of high efficiency and thus the
`turn down ratio and the ability of the strategy to follow drive
`power more closely (therefore relieving some stress on the
`battery) can increase the complexity and cost of the engine or
`lower the peak fuel efficiency. Tradeoffs must be made
`between engine complexity, cost, fuel efficiency, and battery
`lifetime. For example, if a long battery lifetime is the most
`important aspect of the system, then a large sweet spot is
`desired, possibly sacrificing engine simplicity, efficiency, or
`low engine emissions. In a situation where the average power
`is fairly constant a smaller sweet spot may be the most
`efficient solution.
`
`Emi_r§i_gns - Frequently, one of the principle aims of a
`hybrid vehicle is to reduce vehicle emissions to ULEV (Ultra
`Low Emission Vehicle) levels. Consequently, APU emissions
`are very important for system success. In general, emissions
`are minimized when a stoichiometric air to fuel ratio is
`maintained by a closed loop feedback system (using an oxygen
`sensor for feedback). In some operating regimes, such as
`engine starts and transients, the stoichiometric ratio is very
`difficult to maintain resulting in an increase in emissions.
`During a cold-start, the engine must run rich to achieve
`sufficient vaporization of the fuel. Rich running results in
`high hydrocarbons (HC) and carbon monoxide (CO)
`emissions, but low nitrogen oxides (NOx) emissions. A hot-
`start has many of the same problems as a cold start, but the
`time duration before the engine and catalytic converter warm
`up is much shorter. A hybrid strategy which minimizes engine
`cycling will minimize start-related emissions, but that may
`require that the engine have a higher tum-down ratio.
`Transients present an emissions problem that is largely
`related to the speed of the transient. The closed loop feedback
`system that maintains the stoichiometric air fuel ratio is
`sufficient during quasi—steady state modes, however, it can
`only react as fast as the 02 levels can be sensed. If the
`transient is too fast, the engine may run rich, increasing C0
`and HC emissions, or lean, increasing NOx emissions. Some
`of this effect can be reduced using a hybrid strategy that only
`allows slow transients, but this places greater strain on the
`LLD.
`As a series hybrid vehicle decouples both the speed and
`the power of the APU from the speed and power requirement
`at the wheels, this extra degree of freedom can also be used to
`reduce emissions. For a given required power outpitrfthere
`are many combinations of speed and torque that could be used
`to provide that power. If the engine is run in a low speed, high
`load state, the fuel efficiency, noise, and hydrocarbon
`emissions are all improved. At high loads, however, the NOX
`emissions are high and traditional NOX reducing measures
`such as Exhaust Gas Recirculation (EGR) are more difficult.
`Choosing this optimum engine operating point as a function of
`power is an important design consideration but it is not
`necessarily part of the hybrid strategy design.
`
`- Engine noise is not
`much of an issue as far as the performance of a drivetrain is
`concerned, but to avoid customer distress, it must be
`considered as an influencing factor on a hybrid strategy. For
`example, a strategy that has the engine on at full power while
`
`FORD 1355
`
`11 of 13
`
`FORD 1355
`
`

`
`the vehicle is at a complete stop could be extremely
`disconcerting for the driver. Fortunately, the periods of low
`ambient (masking) noise are mostly concurrent with low
`power demands, so throttling back an engine at low vehicle
`speeds is not too much of a compromise in performance.
`
`Life - The life of an APU can generally be extended by
`running it at low, constant power levels. Constant running at
`a sweet spot (for emissions and fuel economy) during low
`power demand driving, however, may cause the battery SOC
`to rise and the engine to be shut off. Depending on the control
`strategy this on/off cycling can be quite frequent. Numerous
`hot starts may shorten the life of an engine unless it is
`designed for multiple starts per trip.
`
`0.7
`0.6
`
`E 0.5
`“0.4
`
`8 0.3
`51 0.2
`0,1
`
`0
`
`0.7
`0.6
`
`E 0.5
`0.4
`50.3
`g 0.2
`0.1
`
`0
`
`0.7
`0.6
`
`0-5
`0.4
`
`0-3
`0.2
`0.l
`
`0
`
`0
`
`10 20 30 40 50
`Power Output (kW)
`(8)
`
`0
`
`10 20 30 40 50
`Power Output (WV)
`(b)
`
`0.7
`0.6
`
`0.5
`0.4
`0.3
`0.2
`0.1
`
`0
`
`0 1020 30 40 S0
`Power Output (kW)
`(0)
`
`0 1020304050
`Power Output (kW)
`(d)
`
`Figure 9: Four examples of engine efficiency curves.
`Specific Fuel Consumption versus Power Output. The
`"sweet spot" is smallest in (a) and largest in (d).
`
`BRAKING CONTROL SYSTEMS - The only other
`system component whose specific characteristics are crucial to
`the optimization of the powertrain is the braking system. The
`added feature of regenerative braking ("regen") can improve
`fuel economy greatly. Unfortunately, because a typical
`vehicle must use all four wheels for braking to maximize
`control, a front wheel drive vehicle with balanced braking will
`not be able to capture all available regen energy. A scheme
`must be devised to maximize the amount of regen captured
`without destablizing the car's decelerations. An optimum
`division of braking power between the front and the rear is a
`function of the degree of deceleration and the desired handling
`of the car. For example, a braking strategy may use regen for
`
`all declarations up t