Power Distribution System Modeling and Simulation of an Alternative Energy
Testbed Vehicle
A thesis presented to
the faculty of
the Russ College of Engineering and Technology of Ohio University
In partial fulfillment
of the requirements for the degree
Master of Science
Yin Wu
November 2010
© 2010 Yin Wu. All Rights Reserved.
2
This thesis titled
Power Distribution System Modeling and Simulation of an Alternative Energy
Testbed Vehicle
by
YIN WU
has been approved for
the Department of Mechanical Engineering
and the Russ College of Engineering and Technology by
Gregory Kremer
Associate Professor of Mechanical Engineering
Dennis Irwin
Dean, Russ College of Engineering and Technology
3
ABSTRACT
WU, YIN, M.S., November 2010, Mechanical Engineering
Power Distribution System Modeling and Simulation of an Alternative Energy
Testbed Vehicle (141 pp.)
Director of Thesis: Gregory Kremer
This thesis deals with the comparison between two powertrain strategies for the
Alternative Energy Testbed Vehicle (AETV): a Fuel Cell-Battery (FC-BT) powertrain
strategy and a Fuel Cell-Battery-Ultracapacitor (FC-BT-UC) powertrain strategy. In
this thesis, a methodology of modeling hybrid vehicle configurations with three energy
devices in ADvanced VehIcle SimulatOR (ADVISOR) 2002 was developed. The
vehicle models, including the AETV physical model, the AETV Fuel Cell (FC) system,
and the power bus of the FC-BT-UC powertrain, were programmed in
MATLAB/Simulink. The models were then applied to the simulation and comparison
of the powertrain strategies. Large scale simulations were run in order to find the
optimized powertrain strategy. The cost function established in the thesis considered
the acceleration performance, gradeability, hydrogen consumption, and powertrain cost.
According to the cost function, if the AETV is equipped with the components in stock,
the FC-BT AETV outperforms the FC-BT-UC AETV.
Approved: ____________________________________________________________
Gregory Kremer
Associate Professor of Mechanical Engineering
4
TABLE OF CONTENTS
Page
Abstract ..................................................................................................................... 3
Abbreviations ............................................................................................................ 7
List of Tables ............................................................................................................. 9
List of Figures ......................................................................................................... 10
Chapter 1 Introduction ............................................................................................. 14
1.1 Environmental Concerns ............................................................................ 14
1.2 Technological Concerns ............................................................................. 14
1.3 Alternative Energy Testbed Vehicle Project Introduction ............................ 15
1.4 Thesis Objectives ....................................................................................... 17
1.5 Thesis Outline ............................................................................................ 18
Chapter 2 Literature Review .................................................................................... 19
2.1 Existing Configurations of Fuel Cell Electric Vehicle Powertrain ............... 19
2.2 Energy Devices in Powertrains ................................................................... 23
2.2.1 Fuel Cells ........................................................................................ 24
2.2.1.1 Proton Exchange Membrane Fuel Cell .................................. 24
2.2.1.2 Ammonia Electrochemical Reformer..................................... 25
2.2.2 Batteries .......................................................................................... 27
2.2.3 Ultracapacitors ................................................................................ 30
2.3 Testbed Vehicle .......................................................................................... 32
Chapter 3 Review on Simulation Tools .................................................................... 34
3.1 Vehicle Simulation Tools ............................................................................ 34
3.2 Introduction of ADvanced VehIcle SimulatOR .......................................... 35
3.2.1 ADVISOR 2002 .............................................................................. 35
3.2.2 ADVISOR Interface ........................................................................ 36
3.2.3 Models in ADVISOR ...................................................................... 37
Chapter 4 Powertrain Architecture and Component Sizing ....................................... 41
4.1 Architecture of the Alternative Energy Testbed Vehicle Powertrain............. 41
4.2 Vehicle Specifications ................................................................................ 42
5
4.3 Component Sizing ...................................................................................... 44
4.3.1 Fuel Cell Sizing ............................................................................... 44
4.3.2 Motor Sizing ................................................................................... 46
4.3.3 Energy Storage System Sizing ......................................................... 48
4.3.3.1 Battery Sizing ....................................................................... 49
4.3.3.2 Ultracapaicitor Sizing ........................................................... 49
Chapter 5 Components Modeling and Energy Management Strategies ..................... 51
5.1 Modeling of Fuel Cell ................................................................................ 51
5.1.1 Fuel Cell Operating Principle .......................................................... 51
5.1.2 Fuel Cell Operating Conditions ....................................................... 53
5.1.3 Fuel Cell Model in ADVISOR ......................................................... 54
5.1.4 Fuel Cell System Efficiency ............................................................ 55
5.1.5 Other Data for the Fuel Cell Model.................................................. 58
5.1.6 Auxiliary Units ................................................................................ 60
5.2 Battery Model ............................................................................................ 62
5.3 Ultracapacitor Model ................................................................................. 64
5.4 Motor Model .............................................................................................. 65
5.5 Energy Management Strategies .................................................................. 66
5.5.1 Fuel Cell Control ............................................................................. 67
5.5.2 Power Bus Control .......................................................................... 69
5.5.2.1 Power Bus Control for Fuel Cell-Battery System .................. 69
5.5.2.2 Power Bus Control for Fuel Cell-Battery-ultracapacitor System
......................................................................................................... 70
5.5.3 Transmission Shifting Strategy ........................................................ 76
5.6 Vehicle Models........................................................................................... 77
5.7 Discussion on Model Validation ................................................................. 78
Chapter 6 Simulation Results and Analysis .............................................................. 82
6.1 Driving Cycles ........................................................................................... 82
6.2 Constraints on Decision Variables .............................................................. 84
6.3 Results Comparison of the Fuel Cell-Battery System and the Fuel
Cell-Battery-Ultracapacitor Systems ................................................................ 88
6
6.4 Generalized Comparison ............................................................................ 98
6.5 Full-Size Case Study ................................................................................ 103
Chapter 7 Conclusions and Future Work ................................................................ 110
7.1 Conclusions ............................................................................................. 110
7.2 Future Work .............................................................................................. 111
References ............................................................................................................. 113
Appendix ............................................................................................................... 117
7
ABBREVIATIONS
ADVISOR: ADvanced VehIcle SimulatOR
AETV: Alternative Energy Testbed Vehicle
“bx-uy”: Energy storage system with number of x battery modules and number of y
ultracapacitor cells
ESS: Energy storage system
EV: Electric Vehicle
FC: Fuel Cell
FC-BT: Fuel Cell-Battery
FC-BT-UC: Fuel Cell-Battery-Ultracapacitor
HEV: Hybrid Electric Vehicle
IMA: Integrated Motor Assist (Honda)
IC: Internal Combustion (Engine)
Li-ion: Lithium ion
NEV: Neighborhood Electric Vehicle
NEVDC: Neighborhood Electric Vehicle Driving Cycle
Ni-MH: Nickel Metal hydride
NREL: U.S. National Renewable Energy Laboratory
OU: Ohio University
Pb: Lead Acid
PEM: Proton Exchange Membrane (fuel cell)
8
PM: Permanent magnet (motor)
SLPM: Standard Liters per Minute
SOC: State of Charge
VW: Volkswagen
9
LIST OF TABLES
Page
Table 2.1. Battery performance and cost comparison ............................................... 28
Table 2.2. Specifications of the Ni-MH battery pack ................................................ 30
Table 2.3. Maxwell BCAP3000 specifications. ........................................................ 32
Table 2.4. Vehicle specifications. ............................................................................. 43
Table 4.1. Vehicle performance constraints. ............................................................. 45
Table 4.2. Corresponding stoichiometric factors of certain currents of the AETV FC
system ..................................................................................................................... 57
Table 6.1. Simulation driving cycles parameters. ..................................................... 84
Table 6.2. Simulation variables and limits. ............................................................... 85
Table 6.3. Assumptions for the battery and ultracapacitor. ........................................ 86
Table 6.4. Cost(x) calculation example. ................................................................... 88
Table 6.5. Sum of squares
of the currents with different
................................... 90
Table 6.6. Specifications of PM16, PM25, and PM32. ............................................. 99
Table 6.7. Comparison results of group1 with PM16 Motor. .................................. 101
Table 6.8. Comparison results of group2 with PM25 Motor. .................................. 102
Table 6.9. Comparison results of group3 with PM32 Motor. .................................. 103
Table 6.10. Specifications of the full-size vehicles. ................................................ 104
Table 6.11. Performance and cost comparison of the full-size vehicles. .................. 109
10
LIST OF FIGURES
Page
Figure 1.1. Honda FCX ........................................................................................... 15
Figure 2.1. FCEV powertrain layout: FC direct supply systems ............................... 19
Figure 2.2. FCEV powertrain layout: FC-BT hybrid systems ................................... 19
Figure 2.3. FCEV powertrain layout: FC-UC hybrid systems ................................... 19
Figure 2.4 FCEV powertrain layout: FC-BT-UC hybrid systems .............................. 20
Figure 2.5. Mark9 SSL™ PEM FC stack ................................................................. 25
Figure 2.6. Complete assembly battery module. ....................................................... 29
Figure 2.7. Maxwell Energy Series BCAP 3000 ....................................................... 31
Figure 2.8. IMA Motor ............................................................................................ 33
Figure 3.1. ADVISOR GUI: Vehicle input screen .................................................... 37
Figure 3.2. ADVISOR backward facing approach .................................................... 38
Figure 3.3. ADVISOR Fuel Cell Vehicle Block Diagram ......................................... 40
Figure 4.1(a). FC-BT powertrain ............................................................................. 41
Figure 4.1(b). FC-BT-UC powertrain ....................................................................... 42
Figure 4.2. Motor characteristic curve ...................................................................... 46
Figure 5.1. Polarization fitted curve of a single Ballard PEM FC ............................. 52
Figure 5.2. Polarization curve of the 4.8 kW Ballard PEM FC stack ........................ 53
Figure 5.3. The FC net model in ADVISOR ............................................................. 55
Figure 5.4. FC system efficiency system efficiency vs. net FC power ...................... 58
11
Figure 5.5. Fuel consumption & Airflow rate vs. Stack current ................................ 59
Figure 5.6(a). Energy loss plot of the AETV with a 600W auxiliary unit .................. 61
Figure 5.6(b). Energy loss plot of the AETV with a 200W auxiliary unit .................. 61
Figure 5.7. Rint model ............................................................................................. 63
Figure 5.8. Battery Rint model in ADVISOR ........................................................... 64
Figure 5.9. Modified ultracapacitor model in ADVISOR ......................................... 65
Figure 5.10. The empirical motor efficiency map in ADVISOR ............................... 66
Figure 5.11. Power bus of the FC-BT system ........................................................... 70
Figure 5.12. New power bus of the FC-BT-UC system............................................. 72
Figure 5.13(a). Ultracapacitor SOC history of the FC-BT-UC powertrain ................ 74
Figure 5.13(b). Battery SOC history of the FC-BT powertrain ................................. 74
Figure 5.13(c). Battery SOC history of the FC-BT-UC powertrain ........................... 75
Figure 5.13(d). Battery current variation of the FC-BT powertrain ........................... 75
Figure 5.13(e). Battery current variation of the FC-BT-UC powertrain ..................... 75
Figure 5.14. Tractive forces vs. vehicle speed .......................................................... 77
Figure 5.15(a). Vehicle model of the FC-BT system in ADVISOR ........................... 78
Figure 5.15(b). Vehicle model of the FC-BT-UC system .......................................... 78
Figure 5.16. Battery SOC history of the validation case ........................................... 81
Figure 6.1(a). UDDS driving cycle .......................................................................... 82
Figure 6.1(b). FTP driving cycle .............................................................................. 83
Figure 6.1(c). Neighborhood electric vehicle driving cycle 1 (NEVDC1) ................. 83
12
Figure 6.1(d). Neighborhood electric vehicle driving cycle 2 (NEVDC2) ................ 83
Figure 6.2. Vehicle trace missed plot of the b20-u10 vehicle with
= 400 .............. 91
Figure 6.3. Battery current of the FC-BT and the FC-BT-UC vehicles with
= 400
over NEVDC1 ................................................................................................. 91
Figure 6.4. Battery current of the FC-BT and the FC-BT-UC vehicles with
= 600
over NEVDC1 ................................................................................................. 92
Figure 6.5. Battery current of the FC-BT and the FC-BT-UC vehicles with
= 800
over NEVDC1 ................................................................................................. 92
Figure 6.6. Battery current of the FC-BT and the FC-BT-UC vehicles with
= 1000
over NEVDC1 ................................................................................................. 93
Figure 6.7. Vehicle weights comparison ................................................................... 94
Figure 6.8. 0-40km/h acceleration times comparison................................................ 94
Figure 6.9. Gradeability at 15km/h comparison ........................................................ 95
Figure 6.10. 100 km fuel consumption comparison .................................................. 95
Figure 6.11. Cost(x) comparison .............................................................................. 96
Figure 6.12. Motor/controller efficiency map ........................................................... 98
Figure 6.13. FC output power of the pure FC vehicle over the first 500s of UDDS 105
Figure 6.14. FC output power of the FC-BT vehicle over the first 500s of UDDS .. 105
Figure 6.15. FC output power of the FC-BT-UC vehicle over the first 500s of UDDS
...................................................................................................................... 106
Figure 6.16. FC output power comparison between the pure FC vehicle, FC-BT vehicle
13
and the FC-BT-UC vehicle ............................................................................. 106
Figure 6.17. Battery SOC comparison between between the FC-BT vehicle and the
FC-BT-UC vehicle ......................................................................................... 107
Figure 6.18. Battery current comparison between the FC-BT vehicle and the
FC-BT-UC vehicle ......................................................................................... 107
14
CHAPTER 1 INTRODUCTION
1.1 ENVIRONMENTAL CONCERNS
Accelerating depletion of fossil fuels and serious global warming caused by increasing
pollution by gasoline vehicles urges the investigation of new energy devices for
vehicles. A hydrogen fuel cell (FC) is a practical alternative to propel the vehicle. It is
not only a clean and renewable energy source with significant reduction of energy use
and harmful emissions, but also an efficient onboard energy device which has the
potential for long range drive and excellent vehicle performance (Chan 2007).
1.2 TECHNOLOGICAL CONCERNS
A pure FC propulsion system has some limitations. FC powered systems have limited
transient performance because of long power response times. Due to its low energy
density, a FC with higher output power is heavy and has a large volume which will
sacrifice vehicle performance (Pede and Iacobazzi 2004). Besides, the unidirectional
current flow in a FC prohibits recovering regenerative braking energy. High initial
investment cost is also a restriction for implementing a pure FC vehicle. To overcome
these limitations it is possible to hybridize FC with other reversible energy storage
sources in order to utilize the advantages of each one and compensate for the
disadvantages. Batteries and ultracapacitors are conventional choices for reversible
energy storage.
15
Currently both commercial vehicle companies and university automotive labs have
been working on the implementation of hybrid FC vehicles. The Honda FCX Clarity
Fuel Cell Vehicle (Honda 2009) uses a 288V standard Lithium-ion (Li-ion) battery pack
as an electricity storage unit to assist a 148 lb 100kW FC in powering the vehicle
(Figure 1.1). The prototype vehicle in the Hy.Power project (Rodatz, Garcia and
Guzzella 2003), which has been road tested, implements ultracapacitors with a FC
system to achieve the transient power demand fluctuations. Both of them have shown
that the hybridized Energy Storage System (ESS) can meet general vehicle
performance requirements.
Figure 1.1. Honda FCX (Honda 2009).
1.3 AETV PROJECT INTRODUCTION
The motivation of the Alternative Energy Testbed Vehicle (AETV) project is to
construct a light weight FC electric prototype vehicle in order to demonstrate the
16
feasibility of implementing an ammonia fuel engine (consisting of an ammonia
electrolytic cell and a PEM FC) for automotive applications. The specific objective of
the AETV project is to design and optimize an ammonia fuel engine which can be
adapted to the car by implementing an appropriate energy management strategy. The
AETV project is a multidisciplinary project which involves chemical engineering
(ChE), electrical engineering (EE) and mechanical engineering (ME). Adapting and
implementing a proposed power supply system is one of the major tasks of ME, while
ChE is in charge of designing and building the ammonia fuel engine, and EE is in
charge of the control strategy for the power system. The AETV is not only supposed to
perform like a real car with sporty appearance and the capability of traveling at 25mph
with a 4.8kW FC, but its cost also must be minimized.
Currently, the Ohio University (OU) AETV has been set up in battery powered mode.
The FC is undergoing testing and the ammonia electrolytic cell is in development.
Some design level modifications on the powertrain are needed before installing the FC,
building and adapting the ammonia fuel engine to the vehicle. Therefore, this thesis
will determine a vehicle powertrain strategy that will be able to achieve the desired
vehicle performance and support the objectives of the AETV.
17
1.4 THESIS OBJECTIVES
The objective of this thesis is to use research and simulation modeling to design an
energy supply system for the AETV. The specific objectives of this thesis are:
a) Present two hybrid powertrain configurations: a Fuel Cell-Battery (FC-BT) hybrid
powertrain and a Fuel Cell-Battery-Ultracapacitor (FC-BT-UC) hybrid powertrain.
b) Size the energy components for each ESS in order to achieve the desired vehicle
performance, which includes acceleration, gradeability and cruising requirements.
More specifically, that means sizing the battery pack used in the FC-BT system, and
sizing the battery and ultracapacitor used in the FC-BT-UC system.
c) Utilize the existing vehicle modeling and simulation tool to build up hybrid FC
powertrain performance models. Simulate the performance of each hybrid powertrain
and verify the feasibility of the hybrid FC power systems for specific driving cycles.
d) Develop power distribution strategies for each powertrain based on power, energy,
efficiency, and durability characteristics of both battery and ultracapacitor, so that it
will make the best use of the power sources.
e) Carry out the comparison between two powertrain configurations considering
simulation results of acceleration time, gradeability, and fuel economy. Determine an
energy system strategy for the AETV based on the comparisons results.
f) Provide a complete system recommendation and all of the vendor and purchasing
details necessary for implementing the powertrain for the AETV.
18
1.5 THESIS OUTLINE
In this thesis, Chapter 1 introduces the significance of developing alternative fuel
vehicles, and the background of the OU AETV project. Chapter 2 presents the recent
power distribution strategies of FCEVs and the energy components that support the
strategies. Chapter 3 introduces the simulation package used in this thesis, ADVISOR
2002, and discusses the importance of customizing the software. Chapter 4 explains the
component sizing of the AETV. Chapter 5 explores the AETV FC power system model
and the new dual ESS model. On the foundation of these component models, the
FC-BT vehicle and the FC-BT-UC vehicle are modeled at the end of this chapter.
Chapter 6 presents the simulations of the AETV with different powertrain
configurations, the pure FC powertrain, the FC-BT powertrain and the FC-BT-UC
powertrain. The simulations results are presented and compared in 3 groups: the first
group of results is analyses of cases with the in-stock motor; the second group of results
is analyses of cases with potential motors; and the third group of results is analyses of
full-size vehicles with a larger FC system and ESSs. Both vehicle performance and cost
analysis are carried out and compared in evaluating the powertrain strategies. Finally,
Chapter 7 summarizes the work of this thesis, and points out the future work needed.
19
CHAPTER 2 LITERATURE REVIEW
2.1 EXISTING CONFIGURATIONS OF FCEV POWERTRAIN
Based on energy storage devices used in propulsion systems, there are four basic energy
distribution configurations of Fuel Cell Electric Vehicle (FCEV) powertrain: FC direct
supply system, battery-hybrid system, ultracapacitor assisted system, and FC-BT-UC
hybrid system as shown in Figure 2.1-2.4 respectively. A FC is the only energy device
used in a direct supply system, while batteries and ultra-capacitors work as assistant
energy devices in the other three powertrain configurations.
H2 supply
system
Motor/
Inverter
Fuel cell
Figure 2.1. FCEV powertrain layout: FC direct supply systems.
H2 supply
system
Fuel Cell
DC converter
Motor/
Inverter
Battery
Figure 2.2. FCEV powertrain layout: FC-BT hybrid systems.
H2 supply
system
Fuel cell
DC link
Ultracapacitor
Figure 2.3. FCEV powertrain layout: FC-UC hybrid systems.
Motor/
Inverter
20
H2 supply
system
Battery
DC Converter
Fuel cell
DC converter
Motor/
Inverter
Ultracapacitor
Figure 2.4 FCEV powertrain layout: FC-BT-UC hybrid systems.
A lot of work has been done on sizing powertrain systems (Wipke, Markel and Nelson
2001) (Rodatz, Garcia and Guzzella 2003) (Gao 2005) (Ferreira and Pomilio 2008)
(Kim and Sohn 2008) (Bernard and Delprat 2009), and comparing different powertrain
configurations (Pede and Iacobazzi 2004) (Gao 2005) (Bauman 2008) (Sapienza 2008).
Based on these previous investigations and research, the FC-UC system can be
eliminated because the FC-BT hybrid powertrain have been shown to provide
acceptable vehicle performance with lower cost. Moreover, a FC hybrid powertrain
with both battery and ultracapacitor have been shown to provide excellent fuel
economy, energy efficiency, and system performance by exploiting the advantages of
each energy storage component but at a higher cost.
A FC stack with a capacity of 40 kW was hybridized with a 30 kW Nickel Metal
Hydride (Ni-MH) battery pack in (Kim and Sohn 2008) to power a mini-bus. The
simulation results based on the Urban-driving cycle demonstrated that the best
21
powertrain efficiency was achieved under a FC-BT hybridization degree of 4:3. The
paper (Wipke, Markel and Nelson 2001) gave an optimal component sizing on energy
distribution system design for a FC-BT SUV. The procedure of optimization illustrated
that degree of hybridization and selection of driving cycle greatly affected the level of
vehicle performance. For urban/highway composite driving cycle, the optimized
system consisted of a 66 kW FC stack and a battery pack with 28×50 Ah/module energy
capacity.
An overall drivetrain comparison of FC direct supply system, FC-BT system, FC-UC
system and FC-BT-UC system was presented in reference (Pede and Iacobazzi 2004).
From the perspective of the paper, FC full powered system was considered worse than
the other 3 systems with poor vehicle performance and high system cost. The paper also
demonstrated that the optimal component selection and sizing highly depended on the
primary tasks fulfilled by the vehicle, namely, the type of the vehicle and driving cycle.
But the conclusions of the paper were mostly based on theoretical analysis without
considering parametric study on component sizes and control strategies for simulation.
In (Gao 2005), the author presented the methodology of sizing the energy components
and a comparison for four cases of energy distribution system including two sizes of
FC-BT system and two sizes of FC-UC system. It was concluded that FC-UC hybrid
powertrain was better because of the advanced acceleration performance and fuel
22
economy. However, the FC-BT hybrid powertrain studied in the paper was equipped
with Lead Acid (Pb) battery pack, while Ni-MH battery with higher specific power and
specific energy had been considered as a better choice for FCEV applications (Burke
2007).
Reference (Bauman 2008) filled the gap of accurate parametric analysis on energy
distribution systems that few previous researchers had investigated. It gave a fair
comparison for FC-BT system, FC-UC system, and FC-BT-UC system because the
analysis was based on precise vehicle models including DC converter and motor.
Considering a weighted combination of vehicle performance, fuel economy and
powertrain cost for different strategies, the paper showed that FC-UC system was not
competitive with the other two systems because vehicle mass and cost increases
significantly as the output power of the ultracapacitor increases. On the other hand,
FC-BT system had better fuel economy and lower powertrain cost than FC-BT-UC
system, while FC-BT-UC system had better acceleration performance. The battery pack
used in the paper is Li-ion battery. Therefore, neither (Gao 2005) or (Bauman 2008)
directly addressed a powertrain hybridized with a Ni-MH battery pack, but they were
still important sources of comparison results between different powertrain
configurations.
23
The previous research and papers discussed above were based on specific assumptions.
Most assumptions are reasonable and valid for a wide range of practical situations. First
of all, they assumed that the FC system would provide the base power for the hybrid
electric vehicles (HEVs), and battery pack and ultracapacitors would provide the peak
power for the vehicles. Secondly, powertrain capital costs and fuel economy were
separately considered in each comparison, while the comparison results are sensitive to
both powertrain capital costs and the consumed hydrogen cost which is related to fuel
economy. Third, the durability and end of life costs of the various components were not
addressed. In addition, hydrogen was assumed to be stored on board in high-pressure
tanks and other types of hydrogen storage such as an ammonia fuel engine were not
considered. Moreover, except for Bauman’s research, the other works made the
assumption that each configuration used the same energy management strategy.
Without changing the energy management strategy shortened the design process, but it
was not accurate for a fair comparison.
2.2 ENERGY DEVICES IN POWERTRAINS
A thorough understanding of the characteristics and specifications of different energy
components is of great importance for designing the energy distribution system.
24
2.2.1 FUEL CELLS
With a developing history over 150 years, the FC does not become an appropriate
energy device for transportation purposes until recent decades. As an electrochemical
device, a FC produces electricity through reaction between hydrogen and oxygen.
Instead of thermal to mechanical energy conversion in Internal Combustion Engine
(ICE), a FC can achieve up to 3 times higher efficiency than an ICE, since it converts
free energy directly into electric energy (Mierlo, Bossche and Naggetto 2004).
2.2.1.1 PROTON EXCHANGE MEMBRANE FUEL CELL
According to the supplied electrolyte, FCs can be divided into 6 major types including
Alkaline FC, Proton Exchange Membrane (PEM) FC, Direct Methanol FC, Phosphoric
Acid FC, Molten Carbonate FC and Solid Oxide FC. Among the 6 types of FCs,
PEMFC is the primary choice for automotive purpose due to its operating conditions,
current density, operating efficiency and cost prospects (Husain 2003).
The 4.8 kW Ballard Mark9 SSL™ FC stack shown in Figure 2.5 applied on the AETV
is a PEM FC. The FC stack shows fast power response at acceleration and needs zero
start up time, although it need a “warm up” time to achieve the ideal operating
conditions for the best performance. This warm up time for a commercial PEM FC
vehicle would be very short, if it is equipped with appropriate components to consist a
gas/coolant supply system which can achieve ideal operating conditions quickly.
25
Figure 2.5. Mark9 SSL™ PEM FC stack.
However, a FC system cannot capture reversible braking energy to regenerate energy
for longer driving range and higher fuel economy. Thus, it can be beneficial to
hybridize the FC with an energy component that allows bidirectional energy flow.
Moreover, generally the more power drawn from the FC, the lower the efficiency of the
system. Therefore, it is also important to have a simple energy management strategy
which can maintain the FC system operating in high efficiency regions (Mierlo,
Bossche and Naggetto 2004). The high efficiency operating region of Mark9 SSL™
PEM FC will be measured and investigated in future work.
2.2.1.2 AMMONIA ELECTROCHEMICAL REFORMER
Being used as the direct fuel in a FC, hydrogen can be either stored in onboard pressure
vessels or generated by reformers installed on the vehicle. A variety of fuels can be
26
used as the feedstock for generating the hydrogen, such as natural gas (methane),
methanol, liquefied petroleum gas, and ammonia (Ehsani, Gao, et al., Fuel Cell
Vehicles 2004).
Reference (Zamfirescu and Dincer 2009) analyzed the potential benefits and technical
advantages of using ammonia as a sustainable fuel for power generation on vehicles.
The ammonia was compared with other conventional fuels (gasoline, compressed
natural gas, liquefied petroleum gas, methanol) as well as with pure hydrogen based on
the performance indicators including the system effectiveness, the driving range, fuel
tank compactness and cost of driving. Some indicators were evaluated by energy
efficiency, the energy storage density per unit volume and per unit mass, and the cost
per unit storage tank volume. The paper indicated that the ammonia was the cheapest
fuel in terms of the cost per 100 km driving range and in terms of the cost per gigajoule
stored onboard. The energy density of ammonia in pressurized tank was the third in
terms of gigajoule per unit volume, which was less than gasoline and liquefied
petroleum gas but much higher than pure hydrogen and other compared fuels.
In addition, an ammonia fuel processor also has the following advantages:
1) Ammonia is commercially available and viable.
2) Hydrogen can be easily generated by simply heating the ammonia.
27
3) Ammonia can be simultaneously used as a refrigerant to cool the operating system in
order to improve the system efficiency.
4) It is much safer than other strategies. The flammability region of ammonia is from 16%
to 25% volume percentage in air, which is a relatively small range compared to the
other fuels. The auto-ignition temperature is 651 degree Celsius, which is much higher
than the auto-ignition temperature for hydrogen, natural gas and gasoline.
5)
Due
to
its
carbon-free
chemical
essence,
reforming
Ammonia
is
environment-friendly with nitrogen as the only byproduct.
2.2.2 BATTERIES
Batteries have been a traditional choice as an energy storage device in electric vehicles
(EVs) and HEVs for years. Similar to a FC, a battery is also an electrochemical device
that converts chemical energy into electricity except that the battery is recharged by an
external electric power source (such as the electrical grid or an on board alternator or
generator) while the FC is fed a supply of fuel in order to sustain the electrochemical
reactions. An advantage over FCs is that batteries can recover regenerative braking
energy due to their bidirectional current flow. Performance of a battery can be evaluated
by behavior in its life cycle, specific energy, specific power, energy efficiency, and cost
(Mierlo, Bossche and Naggetto 2004). Table 2.1 shows a comparison of characteristics
of the common batteries for EV application. Cost is estimated based on retail price
published on www.batteryspace.com in April 2007.
28
Table 2.1
Battery performance and cost comparison (Batteryspace.com, 2007) (Ehsani, Gao, et
al., Energy Storage 2004).
Specific
Energy
(Wh/kg)
Peak
Power
(W/kg)
Energy
Efficiency
(%)
Pb
35-50
150-400
Ni-MH
70-95
Li-ion
80-130
System
Cycle Life
SelfDischarge
(% per 48h)
Cost
(/Wh)
>80
500-1000
0.6
$0.20
200-300
70
750-1200+
6
$1.00
200-300
>95
1000+
0.7
$1.20
Although Pb batteries cost much less than the others, they have a significantly bad
tradeoff on the specific power, the specific power and cycle life. Further, lithium is still
a relatively new and somewhat expensive battery technology. Therefore, Ni-MH
battery is the most common energy storage device for transportation application at
present. For example, all generations of Toyota Prius and Honda Insight use Ni-MH
batteries as their primary energy storage devices (Sears 2009).
Figure 2.6 is the battery module that is currently implemented in the AETV. It is a
Ni-MH battery module which is manufactured by Panasonic EV Energy and used on
Honda Insight EV ( Insight Central: Honda Insight Forum 2008).
29
Figure 2.6. Complete assembly battery module.
The mechanical layout and technical specifications of the battery pack are presented in
table 2.2.
30
Table 2.2
Specifications of the Ni-MH battery pack ( Insight Central: Honda Insight Forum
2008).
Model
Ni-MH Battery
Number of modules
20
Size (mm×mm×mm)
369×231×99
Weight (kg)
35
Number of cells (per module)
6
Terminal Voltage (per module, V)
144
Output Voltage (per module, V)
100.8
Rated Capacity (per module, Ah)
6.5
Storage Capacity (per module, kWh)
0.936
Resistance (Ohm)
11.4
Working Temperature ( )
-30 ~ +60
2.2.3 ULTRACAPACITORS
Generally speaking, a hybrid FC vehicle has to carry out the function of cruising,
grades and accelerating by using an energy device and a power device. Compared to a
31
battery or a FC, an ultracapacitor is characterized by a power density up to 3 kW/kg.
Hence, it is well suited for acceleration transients. Another characteristic of an
ultracapacitor is that it has a wider varying range of state of charge (SOC) compared
with a battery, which contributes to a longer life cycle. Nevertheless, the energy density
of an ultracapacitor is as much as twenty times lower than a battery (Ehsani, Gao, et al.,
Energy Storage 2004). Thus, vehicle curb weight will increase dramatically because a
large amount of cells are needed in an ultracapacitor module to possess a specific
energy capacity.
Figure 2.7 depicts the Maxwell Energy Series Boostcap 3000 ultracapacitor, which is
ideal for industrial and hybrid or EV applications. It can be used with a number of
primary power supplies, such as battery, FC, or generator. Table 2.3 lists the
specifications of the BCAP 3000 including the size and performance of the
ultracapacitor.
Figure 2.7. Maxwell Energy Series BCAP 3000 (Maxwell.com, 2007).
32
Table 2.3
Maxwell BCAP3000 specifications (Maxwell.com, 2007).
Model
Rth (C/W)
BCAP3000 E270
3.2
Isc (A)
4800
Capacitance (F)
3000
Emax (Wh/kg)
5.52
Pmax (W/kg)
11,000
Pd (W/kg)
4,200
Volume (L)
0.475
Mass (kg)
0.51
Length (mm)
138
Height (mm)
14
Diameter (mm)
60
2.3 TESTBED VEHICLE (Gregg. 2007)
The testbed vehicle based on a 1973 Volkswagen Super Beetle had been set up by
Christopher Gregg and Todd Steigerwalt (Gregg. 2007). The suspension, steering,
transmission, and differential are all stock components from the chassis of the 1973
VW Beetle. The vehicle had been converted to an EV by completing modifications
33
such as replacing the IC engine with the Integrated Motor Assist (IMA) (Figure 2.8)
and removing the entire braking system.
Figure 2.8. IMA Motor.
Considering the aerodynamic shape and light-weight of the vehicle, the auto body was
replaced by FiberFab Aztec GT body. The ammonia fuel engine mass used in this thesis
is given by assumption since it is not available from internet or previous works. The
other specifications of the vehicle are listed as following:
1) IMA Motor (Figure 2.8): DC brushless; 13 hp (9.7 kW) @ 1500 rpm; 58 lb·ft
(79 Nm) @ 1000 rpm
2) Transmission/Differential: 4 Speed manual transaxle (transmission and differential
are one piece)
34
CHAPTER 3 REVIEW ON SIMULATION TOOLS
3.1 VEHICLE SIMULATION TOOLS
Computer modeling and simulation will be used to reduce the design cycle and cost of
the AETV by testing configurations and energy management strategies before
prototype construction begins. Building a complete simulation model based on real
components is important for analyzing and evaluating the vehicle performance
accurately. An efficient and systematic vehicle simulation method is crucial to the
hybrid powertrain design. As one of the most popular vehicle simulation platforms,
MATLAB/Simulink is preferred due to its user-friendly programming environment and
accessible block diagrams representing single components and complex systems. For
example, in (Bauman 2008) and (Bauman 2009), the author built up a vehicle simulator
in MATLAB/Simulink based on precise real world FC, battery, and ultracapacitor
models to facilitate the customized vehicle simulations. Besides, some other vehicle
simulation packages such as V-Elph (Butler 1999) are also developed in
MATLAB/Simulink environment.
ADVISOR is one of the packages and widely used in electric and hybrid vehicle
simulations. In (Ogburn, M. 2000), a model of a FC HEV is successfully developed in
ADVISOR. The vehicle data and model results show that ADVISOR has the capability
to model FC HEVs with appropriate modification of the m-file programs. Designed to
“evaluate an energy management strategy for hybrid vehicle’s fuel converter”
35
(ADVISOR Documentation 2002), ADVISOR is selected as the simulation package for
this thesis and will be discussed in the following section.
3.2 INTRODUCTION OF ADvanced VehIcle SimulatOR (ADVISOR)
3.2.1 ADVISOR 2002
ADVISOR was initially developed by the U.S. Department of Energy’s National
Renewable Energy Laboratory (NREL) center in November 1994 as an Advanced
VehIcle SimulatOR for analyzing fuel economy, energy flow and power efficiency of
both conventional and new energy vehicles. It is capable to do the component selection
and sizing for conventional, hybrid and FC vehicles, energy management strategies and
optimization of the energy management strategy. The simulation package was
developed based on Simulink block diagrams and Matlab program files that contain
vehicle configuration, control, and performance data. As a flexible modeling tool with
empirical models, ADVISOR is widely used by auto companies, institutions and
individuals users. The software database is continuously updated with new components
and data provided by the users and periodic user conferences (Markel and Brooker
2002).
ADVISOR has been updated annually since its initial development. AVL published the
first commercial version of ADVISOR in 2004. ADVISOR 2002 is the latest version
36
with free access. Considering the budget of the AETV project, ADVISOR 2002 was
used in this project.
3.2.2 ADVISOR INTERFACE
More advanced than the other simulation packages using command line interfaces,
ADVISOR uses a well-defined user-friendly Graphic User Interface (GUI). The
graphical feature greatly enhances the design and simulation efficiency since it allows
relatively easy modification by manipulating icons rather than programming. For
example, the main system configuration window is the “Vehicle input” screen as shown
in Figure 3.1. As seen in Figure 3.1, users can select powering a vehicle with a
conventional IC engine drivetrain, a pure electric drivetrain, a series or parallel hybrid
electric drivetrain, or a FC based drivetrain by clicking on the corresponding icons
(Markel and Brooker 2002). According to which are applicable to the project vehicle,
other configurations including energy components, exhaust aftertreatment, motor,
transmission, wheel/axle, and powertrain control may be modified within each
drivetrain configuration (ADVISOR Documentation 2002).
37
Figure 3.1. ADVISOR GUI: Vehicle input screen.
3.2.3 MODELS IN ADVISOR
The GUI in ADVISOR is used to interact with the background program files in
MATLAB and control the selection of the component models to be used. These
component models are stored individually in a library in ADVISOR and represented by
block diagrams in Simulink. Each block can be assigned specified parameters which
will later be loaded into the MATLAB workspace as an input set. ADVISOR uses both
backward-facing and forward-facing simulation approaches. First, it employs a
backward-facing approach by receiving the speed requests from a specified driving
cycle and then converting to torque and speed request on wheel/axle. Then the torque
and speed request is converted to power demands of final drive block in which the
38
direct torque and speed for electric motor is calculated. However, the calculated power
or speed demand may exceed the capability of the drivetrain components. In this case,
the forward-facing approach is employed to overcome the limitation of the real
components and estimate the true vehicle performance from calculated practical flow
of power back through the blocks (Markel and Brooker 2002). Figure 3.2 shows the
backward facing approach of analyzing a FC vehicle in ADVISOR. Each component
model will be briefly introduced below.
Driving cycle
Actual performance
Vehicle physical model
Power demand from the motor
Power distribution strategy
Power requested from the fc system
MOTOR
MODEL
Power requested from the ESS
FUEL
CONVERTER
MODEL
Power available from the fc system
Power received by the motor
ESS
MODEL
Power available from the ESS
Figure 3.2. ADVISOR backward facing approach.
Although the equations used in ADVISOR are approximated first-order equations, they
are accurate for vehicle analysis. For example, Eq. 3.1 is the basic solid-body motion
39
equation for describing the straight-line dynamic motion of a vehicle. It is a
fundamental equation of ADVISOR and most other vehicle simulators.
,
where
(3.1)
is the tractive force of the vehicle, and the four terms in the equation are the
force for overcoming rolling resistance, the force for overcoming aerodynamic
resistance, the force for climbing hills, and the force for acceleration respectively. As
shown in Eq.3.1, the vehicle tractive force at each time step can be obtained according
to the speed request of a given driving cycle and the parameters of the vehicle physical
model including vehicle mass, drag coefficient, frontal area, and rolling resistance.
Notice the driving cycles used in this thesis do not have elevation; therefore, the term
that represents the force for propelling the vehicle on a non-zero grade is zero. As
illustrated in Figure 3.3, the tractive force is transformed to the real-time speed and
torque request on wheel/axle. The gearbox model transforms the speed and torque
request on wheel/axle through the reduction ratio to the speed and torque request on the
motor. Finally the motor model decides the power and energy needed for meeting the
driving cycle. The component models in ADVISOR are quasi-static and mostly
empirical models. The data were collected based on the steady state tests in the
laboratory, and corrected for transient performance (ADVISOR 2002).
40
Figure 3.3. ADVISOR Fuel Cell Vehicle Block Diagram.
A fuel converter model simulates the power source of the vehicle. In the case of the
AETV project, it is the device that converts energy in stored ammonia into electric
power usable for the drivetrain. ADVISOR has three different ways to enter data for an
empirical FC model, but only one will be used during each simulation: using the given
voltage-current curve, using the system efficiency vs. power output curve, or through a
GCTool which requires an external toolbox. ESS models in ADVISOR include Pb
battery model, Ni-MH battery model, Li-ion battery model and ultracapacitor model.
Also, several empirical motor/controller models and transmission models are preloaded
in ADVISOR.
The standard vehicle model in ADVISOR simulates powertrains with only one
assistant power device. However, the FC-BT-UC vehicle analyzed in this thesis has
two assistant power devices. A customized powertrain model will be developed in
chapter 5 in order to simulate the AETV with the FC-BT-UC powertrain.
41
CHAPTER 4 POWERTRAIN ARCHITECTURE AND COMPONENT SIZING
4.1 ARCHITECTURE OF THE AETV POWERTRAIN
Due to the dynamic characteristics of FCs, batteries, and ultracapacitors, the
hybridization of these energy components is the best strategy to build a powertrain that
can employ their advantages and tradeoff their limitations. A FC-BT powertrain and a
FC-BT-UC powertrain are the candidate hybrid powertrains for the AETV. Architecture
layouts of each powertrain are illustrated in Figure 4.1(a) and (b).
Wheel
Powertrain control strategy
Ballard fuel
cell system
Transmission
IMA motor/
controller
Energy flow
NiMH battery
Signal flow
Figure 4.1(a). FC-BT powertrain.
42
Wheel
Powertrain control strategy
Ballard fuel
cell system
Transmission
IMA motor/
controller
Energy flow
NiMH battery
Ultracapacitor
Signal flow
Figure 4.1(b). FC-BT-UC powertrain.
In the FC-BT powertrain, the energy flows from the FC stack both to drive the motor
and to charge the battery pack. Simultaneously, the battery pack also provides power to
the motor through the inverter to drive the vehicle. In the FC-BT-UC powertrain, the FC
system and battery pack operates the same way as in the FC-BT powertrain. The
ultracapacitor pack works as an alternative source to the battery pack to supply the bulk
power demand when rate of power change in hard acceleration is faster than a certain
set point.
4.2 VEHICLE SPECIFICATIONS
The AETV is briefly introduced in chapter 2. More detailed specifications are needed to
size the major energy components of the powertrain. These specifications are listed in
table 4.1.
43
Table 4.1
Vehicle specifications (Gregg. 2007).
Vehicle description
Parameters
Value
Unit
Base vehicle glider mass
550
kg
Vehicle cargo mass
136
kg
IMA motor mass
60
kg
Ammonia fuel engine mass
88
kg
Battery mass
35
kg
Transmission mass
41
kg
Rotational inertia mass factor,
1.035
-
Gravity, g
9.8
Transmission efficiency,
0.94
-
Motor average efficiency,
0.96
-
0.9
-
Aerodynamic drag coefficient, Cd
0.381
-
Frontal Area,
1.4
Wheel radius,
0.3111
m
Rolling resistance coefficient,
0.0136
-
Drivetrain efficiency,
(
)
44
Table 4.1 (continued)
Transmission Gear
Gear ratio
Overall reduction ratio
1st gear,
3.60:1
15.84
2nd gear,
1.94:1
8.536
3rd gear,
1.22:1
5.368
4th gear,
0.82:1
3.608
Final gear (differential),
4.4:1
-
4.3 COMPONENT SIZING
4.3.1 FUEL CELL SIZING
In both the FC-BT hybrid system and the FC-BT-UC hybrid system, it is assumed that
the FC will supply the base power required for achieving the top cruising speed and
gradeability. This specifies the minimum size of the FC for propelling the AETV. The
basic nominal vehicle performance is given in table 4.2.
45
Table 4.2
Vehicle performance constraints.
Performance description
Unit
Value
Maximum cruising speed
km/h (mph)
40 (25)
Gradeability @ 15km/h (9.3mph)
-
10%
0 – 40km/h (0-25mph),
s
8
The power supplied for flat road cruising, driving on grade, and acceleration is given
by:
,
where
is the vehicle total mass (assumed to be 910 kg);
resistance coefficient;
is the frontal area in
rotational inertia;
is the air density;
;
(4.1)
is the tire rolling
is the aerodynamic drag coefficient;
is the road grade;
is the mass factor including
is the overall drivetrain efficiency including average efficiency of
motor and transmission. The calculation results indicate that
is 1.48 kW at 32 km/h
(20mph) cruising speed on a flat road, 2 kW at maximum speed of 40 km/h (25mph) on
a flat road, and 4.8 kW for driving on a 10% maximum grade road at 15 km/h (9.3 mph).
Therefore, at its peak capacity the 4.8 kW Ballard PEM FC can support the vehicle to
achieve the nominal vehicle performance.
46
4.3.2 MOTOR SIZING
The motor used on the AETV is the IMA permanent magnet (PM) DC motor from the
HONDA Insight hybrid vehicle. Based on the information provided by the motor model
from ADVISOR, the motor/controller can output a 400A maximum current and has a
60V minimum voltage. The motor speed range is 0 to 8000 rpm, while the torque range
is 0 to 46.5 Nm (ADVISOR 2002). The power range of the motor can be associated
with the torque by multiplying the corresponding speed.
Power = Torque
Angular velocity
The characteristic curves of the motor, including torque vs. speed curve and power vs.
Torque(Nm)
50
12
10
8
6
4
2
0
40
30
20
10
0
0
1000
2000
3000
Motor speed(rpm)
4000
Figure 4.2. Motor characteristic curve.
Power(kW)
speed curve, are plotted in Figure 4.2.
Torque vs speed
Power vs speed
47
Figure 4.2 illustrates that the base speed
(It is the speed at which it operates while
delivering rated torque with rated armature voltage and field current applied. The area
below base speed is called the constant torque area of operation) of the motor is 2000
rpm. The estimated motor redline
(allowed maximum rpm of the traction
motor) is twice as much as the 2000 rpm base speed, that is, 4000 rpm. This is because
the speed ratio of the IMA motor is equal to or less than 2, per the characteristics of a
PM DC motor. The corresponding vehicle base speed and maximum vehicle speed are
determined respectively by:
,
where
is wheel radius;
is the first gear ratio of the transmission;
ratio of the final drive; and
vehicle base speed is
maximum
(4.2)
is the gear
is the top gear ratio. The results demonstrate that the
km/h, and the motor can support the vehicle travelling at a
km/h speed. It indicates that the IMA motor has the capability to
achieve the specified top speed of the AETV.
The DC motor used to propel the hybrid vehicle should satisfy the performance
requirements of maximum vehicle speed (P1), gradeability (P2) and acceleration (P3).
Thus, the rated motor power must be at least equal to or greater than max [P1, P2, P3].
Adapting the equation to 8 seconds acceleration time from 0 to 40 km/h (25 mph), the
calculation result of total power
is 9.33 kW:
,
(4.3)
48
where
is the mass factor including rotational inertia;
from zero speed to
in second;
is the acceleration time
is the final speed of the specified acceleration
performance in m/s, which equals the maximum vehicle speed of 40 km/h (24.85 mph);
is the base vehicle speed in m/s at corresponding IMA motor base speed
.
Therefore, with the 9.8 kW rated power, the Insight IMA motor is sufficient to achieve
the desired acceleration performance.
4.3.3 ENERGY STORAGE SYSTEM SIZING
To fully use the electric motor power capacity, the power supply system should possess
a total peak power that is greater than the rated maximum power of the electric motor.
Thus, the power capacity of the ESS can be denoted as:
,
(4.4)
Notice the ESS in the FC-BT vehicle represents the battery pack, and the ESS in the
FC-BT-UC vehicle represents and battery pack and the ultracapacitor pack. Based on
the 9.8 kW rated power of the motor and the 4.8 kW rated power of the FC system
determined by the mild gradeability and cruising requirements of the vehicle, the peak
power of the ESS is calculated to be at least equal to or greater than 5 kW.
49
4.3.3.1 BATTERY SIZING
The battery pack currently in the vehicle is taken from a HONDA Insight HEV. It
consists of 20 sealed Ni-MH battery modules rated at 6.5 amp-hours energy capacity
with 144V terminal voltage. The peak discharge power of the battery pack is 6 kW at
20
. (Matthew et al. 2001) This battery pack will be used in both FC-BT and
FC-BT-UC powertrain configurations.
4.3.3.2 ULTRACAPACITOR SIZING
The ultracapacitor pack used in the FC-BT-UC powertrain for the AETV is provided as
a peak power source at hard accelerations. Depending on the principle of the powertrain
design, there are different ways to distribute power between the three energy
components. In this thesis, the ultracapacitor substitutes the battery at hard
accelerations; in other words, the batteries and the ultracapacitors operate exclusively.
The energy distribution strategy will be explained in detail in chapter 5. Based on this
energy distribution strategy, the transient power of the ultracapacitor is sized to achieve
the maximum power that can be obtained by the battery pack.
Maxwell BCAP 3000 is selected for the AETV. Assume an ultracapacitor pack rated at
a 10 kW maximum power is used in the system; the number of ultracapacitors needed is
calculated as:
50
The hybrid vehicle with 10 ultracapacitors can only response the transient power
request, but the energy capacity is too small to maintain the vehicle running at a
relatively high velocity for long durations, as the battery can do. Therefore, the
simulations in chapter 6 will use 10 ultracapacitor cells as a starting off point.
In summary, a 4.8 kW FC and a 6.5Ah Ni-MH battery pack are used in each hybrid
powertrain. In addition, an extra ultracapacitor pack with a maximum power rating of
10 kW is used in the FC-BT-UC hybrid powertrain.
51
CHAPTER 5 COMPONENTS MODELING AND ENERGY MANAGEMENT
STRATEGIES
In this section, vehicle models of the AETV with two different power supply systems,
the FC-BT vehicle and the FC-BT-UC vehicle, are developed. More specifically, the
4.8kW Ballard FC implemented in the AETV is modeled for each vehicle; the Insight
Ni-MH battery model and the IMA model had been established by NERL; and the
ultracapacitor model is modified to adapt the new FC-BT-UC vehicle model. The data
presented in this section will be used in the simulations in Chapter 6.
5.1 MODELING OF FUEL CELL
5.1.1 FUEL CELL OPERATING PRINCIPLE
The basic operating principle of a PEM FC is the “reverse reaction” of water
electrolysis. A single cell consists of an anode that separates the hydrogen protons and
electrons as shown in equation 5.1, a cathode with the reaction shown in equation 5.2,
and a polymer membrane that only allows protons to pass through. The electrons
migrate through an external electrical circuit as a direct current. (O’Hayre, et al. 2006,
3-19)
,
(5.1)
,
(5.2)
52
In the process of converting chemical energy into electricity, 4 major types of FC losses
are generated: activation losses in electrochemical reaction, ohmic losses in ionic and
electronic conduction, concentration losses in mass transportation, and fuel crossover
and internal currents. These losses give rise to the shape of the polarization curve
illustrating the relationship between voltage and current of a FC, which is commonly
used for characterizing a FC (O’Hayre, et al. 2006, 3-19).
A single FC generates a tiny amount of direct current electricity. Hence, generally
several single FCs are assembled to a stack (Ballard Power Systems Inc. 2006). The FC
stack used in the AETV is a 4.8 kW Ballard PEM FC stack made up of 25 single cells
with the polarization curve shown in Figure 5.1.
1.2
Cell Voltage (V)
1
0.8
0.6
0.4
0.2
0
0
50
100
150
200
250
300
Current (A)
Figure 5.1. Polarization fitted curve of a single Ballard PEM FC.
350
53
The curve in Figure 5.1 is the 4th order approximated fitted curve based on the
polarization curve given in Mark9 SSL™ Product Integration Manual. Then, the
polarization curve of the 4.8 kW stack is derived and illustrated in Figure 5.2.
25
Stack Voltage (V)
20
15
10
5
0
0
50
100
150
200
250
300
350
Current (A)
Figure 5.2. Polarization curve of the 4.8 kW Ballard PEM FC stack.
5.1.2 FUEL CELL OPERATING CONDITIONS
The FC stack is sensitive to the operating conditions such as reactant inlet pressure,
stack operating temperature, and coolant temperature. For example, overall stack
performance improves with increasing reactant inlet pressure, and cell voltage
increases with stack operating temperature. However, considering stack integrity, cell
stability, operation safety and stack cost, the stack cannot be operated at the pressure
54
and temperature that exceeds certain limits. The FC model in this thesis is built up
based on the performance given at 206.8 kPa (30 psi) nominal inlet pressure and 61
degrees Celsius nominal inlet temperature, with 61 degrees Celsius nominal coolant
temperature. (Ballard Power Systems Inc. 2006)
5.1.3 FUEL CELL MODEL IN ADVISOR
Although a FC system is sensitive to the operating conditions as mentioned in the
previous section, its dynamic feature is dominating only when the power request from
the FC increases or decreases faster than a set rate. Moreover, since the major interest of
this project is not creating a FC system, the existing quasi-static Simulink model in
ADVISOR will be used to simulate the FC system.
ADVISOR uses 3 ways to model a FC: polarization curve model, net model (power
efficiency curve model) and GCTool. In this thesis, the power-efficiency curve model is
used to define the system. The power-efficiency model is shown in Figure 5.3.
According to the basic operating principles of a PEM FC, the model first receives the
signal of power request from the power bus. Then the subblock calculates the amount of
fuel that is supplied to the system. However, considering the efficiency and polarization
characteristics of the FC, the fuel supplied is not only used for generating power but
also consumed as thermal emission and system losses. Therefore, some functions are
used in the model. These functions represent relations between fuel consumption and
55
actual power output, between fuel consumption and thermal emission and other gas
emissions, and the temperature aspects on the FC working status.
Figure 5.3. The FC net model in ADVISOR (ADVISOR 2002).
5.1.4 FUEL CELL SYSTEM EFFICIENCY
As a device converting chemical energy to electric energy, the thermodynamic losses
are considerable during such a conversion process. In general, the practical system
efficiency accounts for thermal losses and nonthermodynamic losses, which include
voltage losses and fuel utilization losses. It is given by:
,
(5.3)
In equation 5.3, the thermal efficiency is defined by the Gibbs free energy relative to the
heating value of the energy conversion. For hydrogen FCs, the efficiency declines as
the temperature increases. But the variation is less than 5% at temperatures lower than
56
200℃. Therefore,
at standard temperature and pressure, namely, 0.83, is
selected for the AETV FC system (O’Hayre et al. 2006, 23-56).
The voltage efficiency of a FC stack is dependent on current and is also dependent on
temperature. It is the ratio of
(the cell voltage at a given current) versus
(the
cell voltage at an open circuit), as expressed in equation 5.4 (O’Hayre et al. 2006,
23-56). At the operating conditions described in section 5.1.2, the theoretical
is
given as 1.47V in the product manual. However, in practical testing and running, the
open circuit voltage cannot exceed 1.23V. Thus, the voltage efficiency is given by
equation 5.4, where the stack voltages at certain currents are collected from the
polarization curve in Figure 5.2 (Ballard Power Systems Inc. 2006).
,
(5.4)
Fuel supplied to the FC system usually cannot be fully used for generation of electricity.
The degree of fuel waste is reflected in the fuel utilization efficiency as:
,
where
(5.5)
is stoichiometric factor of the FCs, which is dependent of the stack current
(O’Hayre et al. 2006, 23-56). The stoichiometric factors of the AETV FCs
corresponding to the stack currents are presented in table 5.1.
57
Table 5.1
Corresponding stoichiometric factors of certain currents of the AETV FC system
(Ballard Power Systems Inc. 2006).
I(A)
15
5.6
30
3
60
1.8
120
1.6
240
1.6
300
1.6
Given the effects of the energy conversion efficiency, the stack voltage efficiency and
the fuel use efficiency, the total system efficiency can be derived from equation 5.3.
Figure 5.4 shows the relationship between the total system efficiency and the net
system power.
58
0.35
Fuel cell system efficiency
0.3
0.25
0.2
0.15
0.1
0.05
0
0
500
1000
1500
2000
2500
Power (kW)
3000
3500
4000
4500
Figure 5.4. FC system efficiency system efficiency vs. net FC power.
5.1.5 OTHER DATA FOR THE FUEL CELL MODEL
In a complete FC system model, fuel consumption and air flow rate are of great
significance. The fuel consumption and air flow rate in SLPM (Standard Liters Per
Minute) are given in the product manual, where
.
These two values are dependent on stack current and number of cells in the stack
(Ballard Power Systems Inc. 2006). Converting SLPM to SI unit (g/s), the fuel
consumption of the AETV FC is given by:
,
(5.6)
59
where
is 2.016 g/mol, the molecular weight of hydrogen. The stack airflow
requirement is given by:
,
where
(5.7)
is 28.966 g/mol, the molecular weight of air;
is 1.6, the air
stoichiometric factor. Figure 5.5 illustrates the fuel consumption map and airflow rate
0.09
4.5
0.08
4
0.07
3.5
0.06
3
0.05
2.5
0.04
2
0.03
1.5
0.02
1
Fuel consumption
0.01
0.5
Air flow rate
0.00
0
50
100
150
200
250
300
Airflow rate (g/s)
Fuel consumption (g/s)
map in regard to the stack current.
0
350
Current (A)
Figure 5.5. Fuel consumption & Airflow rate vs. Stack current.
Based on the data obtained in this section, a modified 4.8kW FC model is established.
“FC_Ballard_AETV.m” in the Appendix is the corresponding model m-file.
60
5.1.6 AUXILIARY UNITS
Auxiliary components, including a hydrogen circulating pump, an air circulating pump,
a coolant circulating pump, and a ventilation fan are implemented to support the
operation of the FC system. For such a stack with a 4.8 kW maximum power capacity,
the power needed to drive these auxiliaries is between 40 W and 540 W (Candusso et al.
2001). The simulation results of the identical FC-BT AETV with a 200W auxiliary load
and with a 600W auxiliary load indicate that the auxiliary power level impacts on the
fuel economy, the total energy usage of the fuel converter, and the energy losses of the
vehicle. The vehicle with a 200W auxiliary load decreases the fuel economy to
32.7L/100km from 52L/100km of the vehicle with a 600W auxiliary load. The total
energy usage of the FC system with a 600W auxiliary unit is 6674kJ, which is larger
than a 4198 kJ energy usage of the vehicle with a 200W auxiliary unit. Figure 5.6(a)
and (b) illustrate the energy loss plots of the vehicle with the two different auxiliary
units. The figures show that the energy losses, including losses during overcoming
rolling and aerodynamic resistance, efficiency losses of the battery, wheel/axle, motor
and gearbox, are independent to the auxiliary power level. But because the energy used
for supporting the auxiliary load increases along with the increase of the auxiliary
power level, the energy loss of the FC system increases when the auxiliary power
increases; more specifically, it is 2978kJ with the 200W auxiliary unit and 4766kJ with
the 600W auxiliary unit.
61
However, there is limited information to determine an assured auxiliary power for the
AETV at present. In this thesis, the assumption of a 200W constant auxiliary power is
made for all the simulations based on the specification of the Honda Insight EV. A
precise auxiliary unit power will be determined when more information is available in
the future.
Figure 5.6(a). Energy loss plot of the AETV with a 600W auxiliary unit.
Figure 5.6(b). Energy loss plot of the AETV with a 200W auxiliary unit.
62
5.2 BATTERY MODEL
As a device with complicated electrochemical reactions, the electrochemical
characteristic of a Ni-MH battery is a nonlinear function of several variables including
internal resistance, open circuit voltage, number of cells, weight and temperature, etc.
(Mierlo et al. 2003). At present, various mathematical models have been constructed to
describe the battery, such as Peukert’s equation and the Shepherd model (Rosario,
2007).
The battery model in ADVISOR is simplified to an internal resistance (Rint) equivalent
circuit model shown in Figure 5.7, or a resistance capacitance (RC) (Brooker et al.
2002). Both the Rint model and the RC model use the backward-facing modeling
approach to handle the battery performance limits, and employ the same lumped
capacity thermal model. The Rint model comprises a perfect open circuit voltage
source and an internal resistance that simulates the coulombic efficiency of the battery,
while the RC model consists of two parallel connected, resistance-capacitance branch
circuits (Johnson 2002). Each model defines the battery with different parameters that
are collected from the battery with different testing procedures. It means that the two
models require the different specification aspects of the battery.
63
Rint
I
E (SOC, T)
V
Figure 5.7. Rint model.
In this thesis, the Rint model shown in Figure 5.8 is implemented in the simulations,
because the data source of the Honda Insight battery pack as a Rint model is known.
The model calculates parameters including equivalent current, terminal voltage, power
losses, and temperature variations, given the power request of the motor and current
SOC of the batteries at the initial time step. Then the model compute the actual
achieved power and SOC at the next time step, and feed the signals back to the energy
management strategy between multiple power supply systems.
64
ess_max_pwr
max pack
pwr (W)1
Goto <cs>1
ess_max_chg_pwr
Goto <cs>2
enable_stop
SOC
SOC
Goto <sdo>, <cs>
ess_pwr_out_a
SOC
algorithm
stop sim
To Workspace9
1
power req'd
into bus (W)
1
pack
Voc, Rint
compute
current
bus_voltage
power available
to bus (W)
Goto <mc>, <gc>, <sdo>ess_th_calc
Tess
Mux
limit power
f(u)
Qess_gen
Qair
Mux
ess_mod_tmp
ess_air_tmp
Tair
ess_air_th_pwr
ess_pwr_loss_a
ess_tmp
Goto <cs>
coul eff
Block
NOTES
Figure 5.8. Battery Rint model in ADVISOR (ADVISOR 2002).
5.3 ULTRACAPACITOR MODEL
The ultracapacitor in ADVISOR is also a Rint model. As illustrated in Figure 5.9, some
modifications are made on the ultracapacitor model to accommodate the dual ESS: first,
change the blocks named with “ess_” to “ess2_” in order to avoid the model clashing
with the blocks in the battery model; and second, add the SOC of ultracapacitor as an
output port of the system so that the SOC signal can be fed back to the power bus for
energy distribution control. The ultracapacitor in this thesis is modeled based on the
specifications and test data for Maxwell 3000 ultracapacitor provided in the product
manual. The modified duel ESS m-file is attached in the Appendix.
65
power req'd
into bus (W)
ess2_pwr_out_a
1
power available
to bus (W)
1
f(u)
Voc = soc*(Vmax-Vmin) + Vmin
Preq Iout
Voc
Voc*k
tmp
SOC Rs
Power
u[2] - u[1]*u[3]
ess2_voltage
Voc-I*R
Calc Iout
coul eff
Saturate I
Scope2
SOC2
Voc
2
Iout SOC
C
SOC2
Calculate SOC
C
(Farads)
ess2_th_calc Saturate T
ess2_mod_tmp
Tess
f(u)
Qess_gen
ess2_air_tmp
Tair
ess2_air_th_pwr
Qair
ess2_pwr_loss_a
Figure 5.9. Modified ultracapacitor model in ADVISOR.
5.4 MOTOR MODEL
The efficiency losses of PM DC motors are caused by copper losses, iron losses,
windage losses and motor constant losses (Larminie and Lowry, 2003). Eq. 5.8
combines the 4 different power losses.
,
where
,
,
(5.8)
, are the coefficients of copper losses, iron losses, and windage losses,
respectively. Knowing the
, the motor efficiency is expressed as:
,
(5.9)
Eq. 5.9 demonstrates that the motor efficiency is a nonlinear function of motor torque
and angular velocity . Thus, in the simulation analysis, the motor efficiency is not a
66
constant but varies according to the operating conditions. Figure 5.10 illustrates the
experiment efficiency map of the IMA PM motor model in ADVISOR.
Figure 5.10. The empirical motor efficiency map in ADVISOR (ADVISOR 2002).
5.5 ENERGY MANAGEMENT STRATEGIES
The two powertrains discussed in this thesis are hybrid systems with 2 or 3 energy
components, in which the requested terminal power is extracted from multiple ESSs.
As discussed in chapter 2, the FC, the battery and the ultracapacitor each has
advantages and disadvantages in different performance aspects.
The energy
management strategy can be considered as the power distribution scheme, which not
only satisfies the power for road request but also ensures a good power match between
67
multiple energy systems. Thus, this scheme should explore a maximum use of the
advantages of each component. The strategies discussed in this section include the FC
control, the power bus control for the FC-BT system, the power bus control for the dual
energy storage FC-BT-UC system, and the transmission shifting strategy.
5.5.1 FUEL CELL CONTROL
A FC HEV only uses the series hybrid strategy for powertrain distribution. Therefore,
the FC control in ADVISOR uses the strategy with conventional series hybrid vehicles
as a reference (Wipke et al. 2001). There are two commonly used strategies for
conventional series hybrid vehicles in ADVISOR: thermostat control and power
follower control. When the strategies are applied in a FC-BT system, the operation of
the FC system is controlled instead of the operation of an IC engine in a conventional
series hybrid vehicle.
In the thermostat control, the FC system is restricted to work at its highest efficiency
point or remain shut off in order to achieve the best fuel economy, while the batteries
are the primary power source to propel the vehicle. Because of this restriction, the
operating mode of the FCs in this strategy is easy to achieve. However, the FC system is
repeatedly turned on and off, which may damage the parts of the FC stack and shorten
its service life. Furthermore, the batteries become the only energy device to supply the
power when the FCs are turned off. The dynamic behavior of the batteries makes it
68
difficult to provide enough power and necessary transient response at vehicle
acceleration. The batteries are frequently charged and discharged in this driving
condition, and the service life of the battery pack is also affected.
The power follower control used in this thesis can overcome the disadvantages of the
thermostat control. The major principle of this strategy is to maintain the FC system
operating in its high efficiency range. As illustrated in Figure 5.4, one criterion used for
system efficiency boundary is the output power of the FC stack. The high efficiency
region of the FCs defined in this thesis is from 25% to 95% maximum output power.
Another principle of this strategy is to protect the batteries from overcharge and
overdischarge. Therefore, the FC stack is subject to the constraints of the SOC of the
batteries. More specifically, the strategy functions in the following ways: when the
vehicle starts up only the batteries start to work, and the FC system may then work
when the system arrives at the suggested operating temperature, depending on the SOC
of the batteries; when the battery SOC is below the low limit and the vehicle is on light
duty road request, which means the power requested by the motor is lower than the
maximum output power of the FCs, the FC system not only outputs the power that
satisfies the road request but also charges the battery pack until it reaches the ideal SOC
region; when the SOC is higher than its high limit, the FC system turns off and the
batteries work independently; when the battery SOC is in the ideal region, the vehicle is
driven by the FC system as the primary energy source and the batteries as the assistant
69
power source; finally when the vehicle decelerates, the FC system turns off and the
batteries capture the regenerative braking energy.
5.5.2 POWER BUS CONTROL
As the core of the entire vehicle control, the power bus receives the power demand and
the power capacity feedback from each subsystem, and determines the distribution of
the energy flow between the energy systems.
5.5.2.1 POWER BUS CONTROL FOR FC-BT SYSTEM
In the FC-BT powertrain, the FC system is the primary energy source. First, the
controller of the FC system obtains the signal of the power requested by the motor, and
calculates the power available from the FCs. Next, the power bus shown in Figure 5.11
receives the value of achieved FC power. Based on this feedback, the power bus
determines if more power needs to be requested from the batteries and how much
power the batteries should provide. Then, the battery model receives the signal of
requested battery power from the power bus and calculates the power available from
the batteries. The actual achieved power from the batteries is fed back to the power bus,
which finally determines the actual output power that can be utilized by the motor.
70
pb_pwr_out_r
1
1
req'd power
output (W)
power req'd from
generator (W)
power req'd from
energy storage
(W)
2
2
power available
from generator
(W)
Sum1
ess_on
0
pb_pwr_out_a
3
Sum2
output power
available at bus
3
not(and(u1,u2))
|u|
power available
from energy storage
(W)
<=
NAND
Abs
1e-9
0
Switch
|u|
Abs1
1e-11
<=
Block
NOTES
Figure 5.11. Power bus of the FC-BT system (ADVISOR 2002).
5.5.2.2 POWER BUS CONTROL FOR FC-BT-UC SYSTEM
For the vehicle with the FC-BT powertrain, if the regenerative braking happens when
the SOC of the battery is close to 1, the battery will be overcharged and damaged by this
instant power. Furthermore, a large power requested in a short time demands a large
discharge current from the battery, which will shorten the lifecycle of the battery
significantly, and add to the operating cost of the vehicle. In order to protect the
batteries, an ultracapacitor pack is added to the FC-BT-UC powertrain as a “peak load
shifting” device.
71
The energy distribution strategy for the FC-BT-UC system needs to deal with the power
distribution between the three energy systems: the FC, the battery and the ultracapacitor
systems. However, ADVISOR 2002 only applies to simulations of the powertrains with
at most two energy components. Hence, this section will describe the development of a
Simulink model that handles the FC-BT-UC powertrain.
In the new powertrain model, the combination of the batteries and the ultracapacitors
can be considered as a “dual ESS”. In this case, the power distribution between three
energy components can be simplified to the power distribution between two energy
components: the FC system and the dual ESS. By adding the ultracapacitors, the current
variation and the SOC of the battery should be smaller and smoother compared to the
FC-BT system under the same driving cycle. The new power bus model aimed to
achieve such effects is illustrated in Figure 5.12.
72
1
1
req'd power
output (W)
power req'd from
generator (W)
pb_pwr_out_r
Scope1
ess_on
Sum1
2
>
AND
power available
from generator
(W)
0
Switch1
0
du/dt
|u|
Derivative
Abs2
Relational
Operator2
Logical
Operator
OR
2
Logical
Operator2
power req'd from
energy storage
(W)
0
<=
ess_pwr_out_r
0
Relational
Operator3
1000
3
power available
from energy storage
(W)
output power
available at bus
3
not(and(u1,u2))
|u|
Sum2
<=
Abs
1e-9
NAND
0
Switch
4
0
pb_pwr_out_a
power available
from energy storage2
(W)1
Block
NOTES
5
SOC2
0.95
|u|
Abs1
1e-11
<=
Relational
Operator5
AND
Scope
power req'd from
ultracapacitor
(W)
4
>
Logical
Operator3
ess2_pwr_out_r
0
<
Relational
Operator4
Modified section
Figure 5.12. New power bus of the FC-BT-UC system.
73
The power distribution strategy of this power bus control is described as thus: the
model sends the fast response power requests to the ultracapacitors, in other words,
when the derivative of the power (dPess/dt) is larger than a preset value, the
ultracapacitors start to work instead of the batteries; when the power request for the
dual ESS is less than 0 and the absolute value of this power is larger than the preset
value, which means that a fast changing regenerative braking energy is generated or a
large extra power is generated by the FC system, the ultracapacitors start to work and
capture the energy, but if the SOC of the ultracapacitors reach the maximum allowed
SOC, the batteries - instead of the ultracapacitors - will start to capture the energy or be
charged; when the absolute value of dPess/dt is smaller than the preset value, the power
requested from the dual ESS will be provided by the batteries, and the power charged to
the dual ESS will be accepted by the batteries. The battery current changes slowly
because of the nature of the battery; therefore, if there is a sudden drop in the requested
power, for instance, when the vehicle decelerates quickly, the battery current will stay
larger than the load current for a while. In that case, the batteries may start to charge the
ultracapacitors, which should not happen since the purpose of the ultracapacitors is to
protect the battery system and increase its service life. Hence, the new power bus model
sets the power of the ultracapacitors to 0 in this situation. Before the battery current
decreases to the value of the load current, the battery will send the extra power to drive
the vehicle.
74
To verify the functions of the new power bus model, both the battery SOC and the
battery current of the FC-BT system (with the 4.8 kW FC and the 20 cells battery) can
be compared to the results of the FC-BT-UC system (with the 4.8 kW FC, the 20 cells
battery and a 10 cells ultracapacitor pack) under the same driving cycle. The
ultracapacitor pack is utilized as expected with the new power bus model, which can be
verified by the SOC history of the ultracapacitors as shown in Figure 5.13(a). When the
ultracapacitors are added to the powertrain, the SOC of the batteries changes smoother
as shown in Figures 5.13(b) and (c), and the current of the batteries changes less
frequently as shown in Figures 5.13(d) and (e). In addition, the operation time of the
ess2-soc-hist
batteries is shorter with the FC-BT-UC powertrain than with the FC-BT powertrain.
1
0.5
0
0
200
400
600
800
time (s)
1000
1200
1400
ess-soc-hist
Figure 5.13(a). Ultracapacitor SOC history of the FC-BT-UC powertrain.
0.7
0.6
0.5
0
200
400
600
800
time (s)
1000
1200
Figure 5.13(b). Battery SOC history of the FC-BT powertrain.
1400
ess1-soc-hist
75
0.7
0.6
0.5
0
200
400
600
800
time (s)
1000
1200
1400
Figure 5.13(c). Battery SOC history of the FC-BT-UC powertrain.
ess-current(A)
60
40
20
0
-20
-40
0
200
400
600
800
time (s)
1000
1200
1400
Figure 5.13(d). Battery current variation of the FC-BT powertrain.
ess1-current(A)
60
40
20
0
-20
-40
0
200
400
600
800
time (s)
1000
1200
1400
Figure 5.13(e). Battery current variation of the FC-BT-UC powertrain.
In summary, the power bus model applied to the FC-BT-UC powertrain works well and
achieves the expected improvements on the battery protection. Notice that two
76
thresholds will influence the simulation results: the power derivative (dPess/dt) and the
maximum allowed SOC of the ultracapacitors.
5.5.3 TRANSMISSION SHIFTING STRATEGY
The transmission used in the AETV is the 4 speed manual gearbox which is the stock
transmission of the 1973 Volkswagen Beetle. The performances of the vehicles with a
manual transmission are affected by the shifting strategy of the transmission. To
conduct a fair comparison between two powertrain strategies, the same shifting strategy
which makes the best use of the tractive effort of the vehicle will be utilized in this
thesis. The tractive effort is determined by the motor torque
ratio , transmission efficiency
and the tire radius
,
, the overall reduction
as shown in equation 5.10.
(5.10)
The characteristic curve of the IMA motor is given in section 4.3.2. It is known that the
motor torque is constant before the motor angular velocity reaches the base speed, but is
dependent on its speed when the angular velocity exceeds the base speed. Consequently,
the tractive efforts of each gear are piecewise functions which are dependent on the
motor torque.
Relationship between the tractive forces and the vehicle speed is plotted in Figure 5.14.
It illustrates that the 1st gear leads to the largest tractive effort before the vehicle speed
77
reaches 24.4 km/h. The 2nd gear leads the largest tractive effort when the vehicle speed
is between 24.4 km/h and 40 km/h.
Tractive effort (N)
2500
2000
1500
1st gear
1000
2nd gear
500
3rd gear
0
0
10
20
30
40
50
Vehicle speed (km/h)
Figure 5.14. Tractive forces vs. vehicle speed.
Therefore, the first two gears are enough for the AETV with a 40 km/h maximum
vehicle speed. In the shifting strategy, the velocity range of the first gear is set to be [0,
24.4] km/h, and the velocity range of the second gear is set to be [24.4, 40] km/h.
5.6 VEHICLE MODELS
The vehicle model with the FC-BT powertrain in ADVISOR is illustrated in Figure
5.15(a). Based on the modifications done to the ultracapacitor model and the power bus
78
model, the new vehicle model with the FC-BT-UC powertrain is established in Figure
5.15(b).
gal
total fuel used
(gal)
time
Clock
Goto<sdo>
-C-
<vc> fuel cell
fuel cell
control stategy
f uel conv erter <f c> net model
<cs>
AND
-C-
<sdo> fuel cell
emis
HC, CO,
NOx, PM (g/s)
drive cycle
<cyc>
exhaust sys
<ex>
vehicle <veh>
wheel and
axle <wh>
final drive <fd>
gearbox <gb>
motor/
controller <mc>
Version &
Copyright
ex_cat_tmp
<cs>
electric acc
loads <acc>
power
bus <pb>
energy
storage <ess>
Altia_off
Battery ESS
Figure 5.15(a). Vehicle model of the FC-BT system in ADVISOR.
gal
time
Clock
Altia_off
total fuel used
(gal)
-C-
Goto<sdo>
<sdo> fuel cell
fuel cell
control stategy
f uel conv erter <f c> net model
<cs>
<vc> fuel cell
AND
-Cemis
HC, CO,
NOx, PM (g/s)
ex_cat_tmp
drive cycle
<cyc>
Version &
Copyright
vehicle <veh>
wheel and
axle <wh>
final drive <fd>
gearbox <gb>
motor/
controller <mc>
electric acc
loads <acc>
energy
storage <ess>
exhaust sys
<ex>
<cs>
power
bus <pb>
Dual ESS
Ultracapacitor
System
Figure 5.15(b). Vehicle model of the FC-BT-UC system.
5.7 DISCUSSION ON MODEL VALIDATION
The researches at NREL put the accuracy of the models as the primary consideration at
the beginning of developing ADVISOR. Validation and benchmarking exiciese were
conducted by NREL and other project participators. With the same inputs, the
79
simulation results obtained by ADVISOR “closely matches” the test results in the
vehicle industry (Wipke, Markel and Nelson 1998). Therefore, ADVISOR is a reliable
vehicle simulator for the simulations conducted in this thesis.
The powertrain subsystems which were modeled in ADVISOR and presented in this
chapter include the 10kW HONDA IMA motor, the 20-module Panasonic Ni-MH
battery pack, the 4.8kW Ballard PEM FC system, a proposed BCAP 3000
ultracapacitor pack, and the energy management strategies.
The IMA motor had been modeled in ADVISOR as “MC_INSIGHT_draft.m” by
NREL. It was created based on data source from published HONDA Insight Articles.
The maximum torque points were derived from the published speed vs. torque for the
Insight engine and the engine with IMA which were presented in UC Davis
Ultra-Clean Vehicle Workshop by HONDA (ADVISOR 2002).
The 20-moduel battery pack had been modeled as “ESS_NIMH6.m” in ADVISOR by
NREL, according to the data obtained through tests performed at 25 Celsius degree
following the PNGV Hybrid Pulse Power Characterization (HPPC) Procedure and at
NREL in the Battery Thermal Management Lab (ADVISOR 2002).
80
The 4.8kW FC system was modeled as “FC_Ballard_AETV.m” in this thesis according
to the data from the product manual and based on assumptions on certain FC operating
conditions. The model has not been validated with any tests in OU at current stage.
However, the overall system efficiency in the operation zone of the AETV FC closely
matches to the “Overall system efficiency vs. net power demand figure” given in the
paper (Bhargav, et al. 2006), which presented a systematic study on modeling of the
4.8kW Ballard FC.
The ultracapacitor pack is a proposed energy device in this thesis. Therefore, it has not
been purchased by the OU auto Lab yet. “UC3000_Maxwell_temp.m” was created
based on the Maxwell PC2500 Ultracapacitor model, which was established by NREL
according to the testing at NREL’s Battery Thermal Management Testing facility
(ADVISOR 2002). The data are corrected by the product manual of Maxwell
BCAP3000. The model has not been validated by component tests, however, the largest
output power of the modeled ultracapacitor pack is close to the 10kW rated peak power
presented in the product manual.
According to the energy management strategy selected in this thesis, the battery is
charged with the extra power from the FC system when the FC system operates at a
power level that is larger than the power for supporting the vehicle performance.
However, most cases studied in this thesis cannot observe the increase in the battery
81
SOC, because the power level of the FC system used in the thesis is relatively low for
supporting both vehicle driving and battery charging. To verify that the FC system
does charge the battery as described, a special case is studies in this section. It has a
100kW AETV FC system and a 75kW motor on the AETV FC-BT vehicle. Figure
5.16 illustrates the battery SOC history of the special case. It shows that the battery
SOC goes above the 70% initial SOC and does not exceed 95%. It demonstrates that
the battery can be charged by the FC system as expected; and the SOC can go above
80% because the “SOC limits” are provided as an optimal SOC buffer range.
1
ess-soc
0.9
0.8
0.7
0
200
400
600
800
time (s)
1000
1200
1400
Figure 5.16. Battery SOC history of the validation case
To sum up, at current stage, the models used and programmed in this thesis are reliable
for analysis of the AETV performance, and the energy management strategy operates
the way it describes in this chapter. However, the validation of the FC system and the
ultracapacitor system is necessary for more accurate analysis in the future.
82
CHAPTER 6 SIMULATION RESULTS AND ANALYSIS
With the components and systems modeled in chapter 5, the simulations of the hybrid
powertrains over certain driving cycles will be presented and analyzed in this chapter.
6.1 DRIVING CYCLES
The driving cycles under which the vehicle runs is dependent on the specified vehicle
category. The AETV is a HEV designed for local neighborhood driving. Urban
Dynamometer Driving Schedule (UDDS) in Figure 6.1(a) and American Federal Test
Procedure (FTP) in Figure 6.1(b) are most commonly used for evaluating vehicle
performance of general electric and hybrid vehicles. Two “neighborhood driving
cycles”, NEVDC1 and NEVDC2 illustrated in Figure 6.1(c) and (d), are implemented
in this thesis. They are obtained by cutting the UDDS and the FTP by half and then
replacing all the speed larger than 40km/h with 40km/h. The important parameters that
indicate the driving cycle characteristics are presented in table 6.1.
Speed (km/h)
100
80
60
40
20
0
0
200
400
600
800
time (s)
1000
Figure 6.1(a). UDDS driving cycle.
1200
1400
83
100
Speed (km/h)
80
60
40
20
0
0
500
1000
1500
2000
2500
time (s)
Figure 6.1(b). FTP driving cycle.
Speed (km/h)
50
40
30
20
10
0
0
200
400
600
800
time (s)
1000
1200
1400
Figure 6.1(c). Neighborhood electric vehicle driving cycle 1 (NEVDC1).
Speed (km/h)
50
40
30
20
10
0
0
500
1000
1500
2000
2500
time (s)
Figure 6.1(d). Neighborhood electric vehicle driving cycle 2 (NEVDC2).
84
Table 6.1
Simulation driving cycles parameters.
Time duration
(s)
Distance
(km)
Max speed
(km/h)
NEVDC1
1369
5.94
40.23
0.76
NEVDC2
2477
8.77
40.23
0.76
Driving Cycle
Max acceleration
(m/ )
6.2 CONSTRAINTS ON DECISION VARIABLES
The simulation results include the vehicle performance and components performance.
The vehicle performance is evaluated by the system efficiency, acceleration time,
gradeability, fuel economy, and ability to match the driving cycles. The performance of
the energy components are evaluated by SOCs and current variations, and efficiency
maps during the driving cycles. Before the simulation, it is necessary to assign the
important parameters and constraints to the variables that can be controlled, for
instance, vehicle mass, initial SOC of the battery pack, initial SOC of the ultracapacitor
pack, etc..
The simulations with the rough parameters calculated in chapter 4 are not accurate
enough for a fair comparison between the two hybrid powertrains. The different
combinations of the power control parameter
(dPess/dt) in the controller and sizes of
energy components obtain various simulation results. Table 6.2 lists the variables that
85
affect the simulation results and the parameter boundaries of them. Based on the
combination of component strategies in this table, 485 different systems will be
analyzed: 1 FC-BT case and 484 FC-BT-UC cases.
Table 6.2
Simulation variables and limits.
Powertrain type
Variables
Lower
limit
Interval
Upper
limit
FC-BT
Number of battery modules
20
-
20
FC-BT-UC
Number of battery module
10
1
20
Number of ultracapacitor cells
10
1
20
400
200
1000
(W/s)
Although the vehicle mass is assumed to be 910kg in Section 4.2, the simulations will
use different override vehicle mass depending on the battery modules and
ultracapacitor cells actually used. The assumptions, including unit weight, performance
specifications and unit cost, for the battery and the ultracapacitor are presented in table
6.3.
86
Table 6.3
Assumptions for the battery and ultracapacitor.
Weight
Specific power
(W/kg)
Specific energy
(Wh/kg)
Market price in
2010
Battery
1 (kg/module)
1093
46
54 ($/module)
Ultracapacitor
0.51 (kg/cell)
4200
5.52
90 ($/cell)
Note: The battery module price listed in this table is based on the market retail price of
Ni-MH D-cell batteries, which is 9$ per cell in 2010. The ultracapacitor cell price is
based on the market retail price of Maxwell BCAP3000 in 2010. Notice the comparison
results may vary if the costs vary.
The acceleration performance is evaluated by 0-40km/h acceleration test. The
gradeability is scored by the largest grade achieved by the vehicle at 15km/h cruising
speed, and the time duration of all grade tests is 10 seconds. The fuel economy is
evaluated by hydrogen consumption in liters per 100km based on the NEVDC1 driving
cycle. The initial SOC of the batteries and the ultracapacitors is 0.7 for all simulations.
The upper limit of the battery SOC is 0.8; the lower limit of the battery SOC is 0.4. The
ultracapacitor SOC is controlled not to exceed 0.95 for all simulations. The FC power is
fixed at 4.4kW in each simulation.
87
To include every item that evaluates a powertrain, a result function Cost(x) is built up in
eq. 6.1 to score the powertrains:
,
(6.1)
The terms in the function represent acceleration time, gradeability, fuel consumption
(liters per km) and the cost of purchasing the energy components, respectively. Given
the same coefficient for each evaluated terms, every term is equally weighted in the
comparison. The array - [
] - of the FC-BT vehicle (b20) is selected to be the
baseline array for calculating the Cost(x). The index arrays of the 484 FC-BT-UC
vehicles presented in table 6.2 are represented by [
]. The term
is
determined based on the highest ESS cost in this thesis, 16800 dollars. The baseline
array gives Cost(x) = 0. Table 6.4 gives an example of calculating the Cost(x) of the
FC-BT-UC vehicle with a b20-u10 and 400W/s as the value of . The maximum score
of the Cost(x) yields the optimal powertrain choice for the AETV as compared to the
FC-BT powertrain, and with the configurations and energy management strategies as
implemented.
88
Table 6.4
Cost(x) calculation example.
Index
Value
8.8s
8.9s
24.3% 24.1% 32.7lpm 34.1lpm
Cost(x)
0$
900$
-2.9%
6.3 RESULTS COMPARISION OF THE FC-BT SYSTEM AND THE FC-BT-UC
SYSTEMS
Now the simulation results can be gathered for the FC-BT system and the FC-BT-UC
systems listed in table 6.2 based on the models established in chapter 5 and the
assumptions and constraints given in section 6.2. To effectively compare these two
types of powertrains, only the FC-BT-UC systems that obtain the first 3 largest Cost(x)
value with each control variable
will be compared with the FC-BT system.
For the FC-BT-UC powertrains, apparently the larger the power control parameter ,
the more the power is demanded from the ultracapacitors. The battery current wave is a
good way to observe the utilization of the batteries. It reflects the intensity of the
current variations and the output energy of the batteries by the amplitude of the waves.
Generally the square of the amplitude is proportional to the intensity of a wave. To
evaluate the effects of the control parameters , , the sum of squares of the current
89
wave amplitude at every second, is calculated and compared between each system as
listed in table 6.5. The powertrain with the smallest value of
batteries. However, compared to the
FC-BT-UC vehicle, the
maximally protects the
difference between the FC-BT vehicle and each
difference between each FC-BT-UC vehicle is insignificant.
Therefore, the FC-BT-UC vehicles with any
listed in the table display great
advantages than the FC-BT vehicle in terms of the battery current variation. The battery
currents of the FC-BT-UC (b20-u20) vehicles with different control parameters are
individually plotted in Figure 6.3-6.6 in comparison with the battery current of the
FC-BT vehicle over NEVDC1. For the Figures illustrating the performance comparison
between two or three vehicles, Figure 6.3-6.6 and Figure 6.13-6.18: the dotted lines
indicate the performance of the FC-BT vehicle; the solid lines indicate the performance
of the FC-BT-UC vehicle; and the dashed lines indicate the performance of the pure FC
vehicle.
90
Table 6.5
Sum of squares
When
of the currents with different
FC-BT
FC-BT-UC
= 400
FC-BT-UC
= 600
FC-BT-UC
= 800
FC-BT-UC
= 1000
120030
66078
62423
73895
78266
is set at 400W/s, the controller sends the most power requests to the
ultracapacitors. The 2 highest scored powertrains of this group are vehicles with the
b19-u10 and the b20-u10 ESS. They rank in the top 5 of all powertrains in terms of the
Cost(x) score; however, all of them missed the traces of the NEVDCs in varying
degrees, which is caused by the low energy capacity of the ultracapacitors. For instance,
the vehicle with the b20-u10 ESS scores the highest, but it missed the required trace
greater than 2 mph for 7.74% time of the NEVDC1 as shown in Figure 6.2. Moreover,
although the
of this system is the second lowest, and Figure 6.3 shows that the
battery operates for longer duration in the FC-BT vehicle, the absolute value of the
current of the FC-BT-UC vehicle overshoots the current of the FC-BT vehicle
frequently in the first 700s of NEVDC1, which means the FC-BT-UC system with
400W/s control parameter have higher current charges and discharges more frequently
than the FC-BT system in the first part of NEVDC1. Thus,
= 400 is not an ideal
power control parameter for the vehicle. In the following comparison, vehicles with
equal to 600, 800 and 1000 will be analyzed in groups.
91
Vehicle speed (km/h)
50
Cycle trace
Vehicle trace
Missed trace
40
30
20
10
0
0
200
400
600
800
1000
1200
1400
time (s)
Figure 6.2. Vehicle trace missed plot of the b20-u10 vehicle with
= 400.
80
FC-BT-UC vehicle
FC-BT vehicle
ess-current(A)
60
40
20
0
-20
-40
0
200
400
600
800
1000
1200
1400
time (s)
Figure 6.3. Battery current of the FC-BT and the FC-BT-UC vehicles with
over NEVDC1.
= 400
92
80
FC-BT vehicle
FC-BT-UC vehicle
ess-current(A)
60
40
20
0
-20
-40
0
200
400
600
800
1000
1200
1400
time (s)
Figure 6.4. Battery current of the FC-BT and the FC-BT-UC vehicles with
over NEVDC1.
= 600
60
FC-BT vehicle
FC-BT-UC vehicle
50
ess-current(A)
40
30
20
10
0
-10
-20
-30
0
200
400
600
800
1000
1200
1400
time (s)
Figure 6.5. Battery current of the FC-BT and the FC-BT-UC vehicles with
over NEVDC1.
= 800
93
60
FC-BT vehicle
FC-BT-UC vehicle
50
ess-current(A)
40
30
20
10
0
-10
-20
-30
0
200
400
600
800
1000
1200
1400
time (s)
Figure 6.6. Battery current of the FC-BT and the FC-BT-UC vehicles with
over NEVDC1.
= 1000
The independent simulation results for the 3 vehicles which have the highest score of
Cost(x) in each group are compared with the FC-BT baseline case in Figure 6.7-6.11.
94
FC-BT
b20
b20 u13
FC-B-UC1000
b20 u12
b20 u14
b19 u12
FC-B-UC800
b18 u13
b18 u12
b19 u10
FC-B-UC600
b20 u11
b20 u10
855
860
865
870
Vehicle mass (kg)
Figure 6.7. Vehicle weights comparison.
FC-BT
b20
b20 u13
FC-B-UC1000
b20 u12
b20 u14
b19 u12
FC-B-UC800
b18 u13
b18 u12
b19 u10
FC-B-UC600
b20 u11
b20 u10
4
5
6
7
Acceleratoin time(s)
8
9
Figure 6.8. 0-40km/h acceleration times comparison.
10
95
b20
FC-BT
b20 u13
FC-B-UC1000
b20 u12
b20 u14
b19 u12
FC-B-UC800
b18 u13
b18 u12
b19 u10
FC-B-UC600
b20 u11
b20 u10
22
22.5
23
23.5
Grade (%)
24
24.5
25
Figure 6.9. Gradeability at 15km/h comparison.
FC-BT
b20
b20 u13
FC-B-UC1000
b20 u12
b20 u14
b19 u12
FC-B-UC800
b18 u13
b18 u12
b19 u10
FC-B-UC600
b20 u11
b20 u10
30
31
32
33
34
Fuel consumption (L/100km)
35
Figure 6.10. 100 km fuel consumption comparison.
36
96
FC-BT
b20
FC-B-UC1000
b20 u12
b20 u13
b20 u14
b19 u12
FC-B-UC800
b18 u13
b18 u12
b19 u10
FC-B-UC600
b20 u11
b20 u10
-0.1
-0.08
-0.06
-0.04
-0.02
Cost(x)
0
0.02
0.04
Figure 6.11. Cost(x) comparison.
Figure 6.7 shows that all the vehicles with the top ranking FC-BT-UC powertrains are
heavier than the vehicle with the FC-BT powertrain. Figure 6.8 shows that the
acceleration time is around 8.8s with no matter how many the ultracapacitors are added
in the ESS. The 10 compared vehicles can ascend 24.1%-24.3% highest grade for 10
seconds duration while maintaining at 15km/h vehicle speed, which implies that the
ultracapacitors cannot increase the gradeability of the vehicle. Compared to the FC-BT
vehicle, the FC-BT-UC vehicle with b20-u13 ESS and
FC-BT-UC vehicle with the b20-u14 ESS and
= 1000 W/s, and the
= 1000 W/s show slightly better fuel
economy, while the other FC-BT-UC vehicles have larger fuel consumptions. In
97
addition, the powertrain cost of the FC-BT AETV is 0 since all the ESS components are
in-stock. The powertrain cost of the FC-BT-UC AETVs can be estimated by the number
of ultracapacitor cells and the cost/cell given in table 6.3. Figure 6.11 shows that the
cost (x) values of all the FC-BT-UC vehicles are smaller than that of the FC-BT vehicle.
In summary, although the FC-BT-UC vehicle greatly extends the battery life cycle by
reducing the current variation intensity of the battery, adding the ultracapacitors to the
ESS does not improve the vehicle performance but raises the cost and mass of the
AETV. Hence, the FC-BT-UC powertrain cannot compete with the FC-BT powertrain
in the overall evaluation.
However, the empirical comparison cases show that a vehicle with ultracapacitors
added should perform better on acceleration than the original vehicle. The simulation
results of acceleration time gathered in section 6.3 violated the rule of thumb. This is
caused by the dynamic performance limitations of the IMA motor. Reviewing the motor
sizing in chapter 4, a 9.8kW motor is sized based on a 8s 0-40km/h acceleration time
and a constant 96% nominal motor efficiency. However, as mapped in Figure 6.12, the
94% efficiency is the highest efficiency the motor/controller can achieve in a driving
cycle, and the average motor efficiency is 91.5%.
98
Motor/controller Efficiency (driving only, not regen)
1
0.9
0.8
efficiency
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
200
400
600
800
time (s)
1000
1200
1400
Figure 6.12. Motor/controller efficiency map.
Therefore, the acceleration performance obtained by the ultracapacitors is restricted by
the dynamic nature of the motor. The conclusions summarized above only apply to the
vehicles with the AETV in-stock components. To generalize the comparison results, the
vehicles with different motors will be analyzed in section 6.4.
6.4 GENERALIZED COMPARISON
In this section, 3 motors, PM16, PM25, and PM32, will be implemented to the AETV
respectively. The specifications of these motors are listed in table 6.6.
99
Table 6.6
Specifications of PM16, PM25, and PM32.
Weight (kg)
Peak efficiency
Max power (kW)
PM16
21
92%
14
PM25
45
90%
25
PM32
41
90%
33
To ensure fair comparisons, the descriptions for vehicle and simulation constraints are
listed below:
1) Simulations are divided into 3 groups: the FC-BT vehicle and the FC-BT-UC vehicle
with the PM16 motor; the FC-BT vehicle and the FC-BT-UC vehicle with the PM25
motor; and the FC-BT vehicle and the FC-BT-UC vehicle with the PM32 motor.
2) The default vehicle components of the AETV, including the vehicle body, the Ballard
FC stack, the transmission, the wheel/axle, and the accessories, are used in all the
vehicles.
3) The FC control described in chapter 5 is used for all the vehicles.
4) The FC-BT vehicle uses the default 20 module battery pack; the FC-BT-UC vehicles
in each group use b20-u20 dual ESS.
100
5) For the FC-BT-UC vehicles, the power bus control described in chapter 5 is applied.
The power control parameter
is set to be 600W/s, because the system with this value
maximally protects the batteries and obtains the highest Cost(x) value at the same time.
6) The value of Cost(x) is used to evaluate all the vehicles as explained in section 6.2.
The comparison results of the 3 groups are collected in table 6.7-6.9. Comparing every
item evaluated in each group, the difference between acceleration times has the largest
influence on the difference between the Cost(x) values. It shows that the motor type has
little effect on the Cost(x) value of the FC-BT vehicles, but considerably affects the
value of the FC-BT-UC vehicles. This is because the acceleration performance of the
vehicles is not only affected by the motor type but also limited by the specific power
density of the batteries in the FC-BT vehicles, while the use of ultracapacitors for “peak
power shifting” highly improves the acceleration performance of the FC-BT-UC
vehicles. The FC-BT-UC vehicle with the PM 25 motor in group 2 shows great
advantage in the Cost(x) value. In group 2 and 3, the FC-BT-UC misses the trace less
than 1% time of the NEVDCs due to the low specific energy of the ultracapacitors
compared to the Ni-MH battery. Despite that the FC-BT-UC powertrains are heavier
than the FC-BT powertrains, the improvements on the acceleration performance of the
FC-BT-UC vehicles significantly override the shortage on gradeability and powertrain
cost. The fuel economy of the vehicle is also increased since the ultracapacitors filters
the majority of the fast response power requests for the fuel converter.
101
Table 6.7
Comparison results of group1 with PM16 Motor.
FC-BT
FC-BT-UC
Vehicle mass (kg)
823
833
0-40km/h acceleration time (s)
9.4
9.5
12.8%
12.6%
34.9
35.5
15.5%
15.3%
NEVDC1
Complete
Complete
NEVDC2
Complete
Complete
-15.2%
-18.8%
Max grad @ 15km/h for 10s
Fuel economy based on NEVDC1 (L/100km)
Overall system efficiency
Cost(x)
102
Table 6.8
Comparison results of group2 with PM25 Motor.
FC-BT
FC-BT-UC
Vehicle mass (kg)
847
857
0-40km/h acceleration time (s)
8.9
3.7
22.2%
21.9%
48.9
48.6
11.6%
11.8%
NEVDC1
Complete
0.876% time missed
NEVDC2
Complete
0.484% time missed
-14.8%
-2.8%
Max grad @ 15km/h for 10s
Fuel economy based on NEVDC1 (L/100km)
Overall system efficiency
Cost(x)
103
Table 6.9
Comparison results of group3 with PM32 Motor.
FC-BT
FC-BT-UC
Vehicle mass (kg)
840
850
0-40km/h acceleration time (s)
7.9
4.2
24.3%
24%
54.6
51.9
10.4%
10.9%
NEVDC1
Complete
1.68% time missed
NEVDC2
Complete
0.928% time missed
-14.2%
-4.6%
Max grad @ 15km/h for 10s
Fuel economy based on NEVDC1 (L/100km)
Overall system efficiency
Cost(x)
In conclusion, implemented the motor that can fully explore the capacity of the ESSs,
the FC-BT-UC powertrain achieves large improvements in overall vehicle performance
over the FC-BT powertrain according to the Cost(x) value.
6.5 FULL-SIZE CASE STUDY
Currently, the AETV project is aimed to develop a FC hybrid vehicle that satisfies
neighborhood driving conditions. From a long-term point of view, it is necessary to
extend the comparison conclusions to full-size vehicles. In this section, 3 full-size
104
baseline vehicles with a 20kW FC system are analyzed: a pure FC vehicle, a FC-BT
hybrid vehicle and a FC-BT-UC hybrid vehicle. The vehicle specifications that affect
the simulation results, including a 49kW PM motor and a 1 speed gearbox, are
presented in table 6.10.
Table 6.10
Specifications of the full-size vehicles.
Pure FC
Motor
FC-BT
FC-BT-UC
PM49
Transmission
TX_1SPD_IDEAL
FC
20kW Ballard PEM FC
Battery (number of modules)
N/A
20
20
Ultracapacitor (number of cells)
N/A
N/A
20
Power control parameter
N/A
N/A
1000
(W/s)
The vehicle simulations are conducted over UDDS driving cycle. The FC output power
of the 3 baseline full-size vehicles over the first 500s of UDDS is individually plotted in
Figure 6.13- 6.15. The area enclosed by the power curves reveals the energy produced
by the FC system. Figure 6.16 shows the FC-BT-UC vehicle has the smallest enclosed
105
area, which means the utilization ratio of the FC is lowest in the FC-BT-UC vehicle and
highest in the pure FC vehicle. Thus, the FC-BT-UC vehicle exceeds the other two
vehicles in terms of the protection of the FC system. Figure 6.17 and 6.18 respectively
compares the battery SOC over UDDS and the battery current variation of the FC-BT
vehicle and the FC-BT-UC vehicle. The figures demonstrate that the battery operation
time in the FC-BT-UC vehicle is reduced by using the ultracapacitors.
4
x 10
FC output power (W)
2
1.5
1
0.5
0
0
50
100
150
200
250
time (s)
300
350
400
450
500
Figure 6.13. FC output power of the pure FC vehicle over the first 500s of UDDS.
4
x 10
FC output power (W)
2
1.5
1
0.5
0
0
50
100
150
200
250
time (s)
300
350
400
450
500
Figure 6.14. FC output power of the FC-BT vehicle over the first 500s of UDDS.
106
4
x 10
FC output power (W)
2
1.5
1
0.5
0
0
50
100
150
200
250
time (s)
300
350
400
450
500
Figure 6.15. FC output power of the FC-BT-UC vehicle over the first 500s of UDDS.
4
x 10
FC-BT vehicle
Pure FC vehicle
FC-BT-UC vehicle
FC output power (W)
2
1.5
1
0.5
0
0
50
100
150
200
250
time (s)
300
350
400
450
500
Figure 6.16. FC output power comparison between the pure FC vehicle, FC-BT vehicle
and the FC-BT-UC vehicle.
107
0.75
FC-BT-UC vehicle
FC-BT vehicle
ESS-SOC
0.7
0.65
0.6
0.55
0.5
0
50
100
150
200
250
time (s)
300
350
400
450
500
Figure 6.17. Battery SOC comparison between the FC-BT vehicle and the FC-BT-UC
vehicle.
80
60
ess-current (A)
40
20
0
-20
-40
-60
-80
FC-BT-UC vehicle
FC-BT vehicle
0
50
100
150
200
250
time (s)
300
350
400
450
500
Figure 6.18. Battery current comparison between the FC-BT vehicle and the
FC-BT-UC vehicle.
The cases analyzed above are the basic sized vehicles. To generalize the conclusions,
table 6.11 lists 5 vehicle cases, namely, 3 baseline vehicles and 2 full-size vehicles with
varied hybridization degree. The table presents the performance and cost comparison of
108
the vehicles. The data collected in the table is concluded in the following:
1) The hybridization degree of the energy devices will impact the overall vehicle
performance. In this thesis, the hybridization degree is selected according to the Mazda
compact car (Pede and Iacobazzi 2004). It is necessary to optimize the hybridization
degree in order to find an optimum vehicle powertrain in the future.
2) The 0-60km/h acceleration performance of the FC-BT-UC vehicles greatly surpasses
the other two types of vehicles.
3) The fuel economy of the FC-b25-u15 vehicle is the best among the 5 vehicles.
Implementation of the ultracapacitors also results in the highest overall system
efficiency.
4) The estimated powertrain cost should consider the budget of a 15kW FC stack since
the OU automotive lab only have a 4.8kW Ballard PEM FC in stock. The unit cost of
the FC system includes the capital cost and operating cost. According to the price
published by Ballard company in 2010, the FC system cost is 2800 $/kW net.
Considering a 30% system efficiency, it is assumed to be 1000$/kW in this thesis.
Compared with the unit cost of the ultracapacitors and the batteries, the high unit cost of
the FCs dominates the powertrain cost. Therefore, although the analyzed FC-BT-UC
vehicles have a higher powertrain cost than the other two types, the powertrain of the
FC-BT-UC vehicle only costs no more than 7% of the FC-BT vehicles and no more
than 10% of the pure FC vehicle.
109
Table 6.11
Performance and cost comparison of the full-size vehicles.
Pure FC
b20
b30
b20-u20
b25-u15
Vehicle mass (kg)
907
943
953
952
954.5
0-60km/h acc. (s)
9.9
7.4
6.7
5.4
5.8
Fuel economy based
on UDDS (L/100km)
92.9
52.4
50.5
49.4
47.8
Overall system eff.
9.5%
16.9%
17.4%
18%
18.4%
% time of UDDS
trace missed
0.073%
0
0
0
0
Powertrain cost ($)
15,000
15,000
15,540
16,800
16620
110
CHAPTER 7 CONCLUSIONS AND FUTURE WORK
7.1 CONCLUSIONS
In this thesis, an approach for modeling hybrid vehicle configurations with three energy
devices in ADVISOR has been developed. In order to accurately evaluate the Ohio
University (OU) Alternative Energy Testbed Vehicle (AETV) with two different
powertrains, the vehicle models, including the vehicle physical model, the 4.8 kW
Ballard Proton Exchange Membrane (PEM) Fuel Cell (FC) system, the energy
management strategy were programmed in MATLAB/Simulink. The powertrains used
were
the
Fuel
Cell-Battery
(FC-BT)
powertrain
and
the
Fuel
Cell-Battery-Ultracapacitor (FC-BT-UC) powertrain. The AETV with a FC-BT
powertrain was modeled with the standard FC vehicle configuration in ADVISOR. In
the case of the FC-BT-UC powertrain, the AETV was modeled with the modified FC
vehicle configuration.
Large-scale simulations were run based on the established vehicle models. A detailed
comparison between the FC-BT vehicle and the FC-BT-UC vehicles was presented.
The evaluated objects include acceleration time, fuel economy, gradeability, the degree
of battery protection, and powertrain cost.
111
The results demonstrate the following conclusions:
1. The battery service life in the FC-BT-UC AETV was extended due to the decreased
charge and discharge current based on the established vehicle models.
2. If the AETV uses the vehicle components in-stock, the FC-BT powertrain achieves
better vehicle performance than the FC-BT-UC powertrain, at a lower cost.
3. The motor type significantly affected the results. Implementing a suitable motor that
can fully explore the capacity of the Energy Storage Systems (ESSs), the FC-BT-UC
powertrain is preferable to the FC-BT powertrain if high acceleration performance is
needed.
4. If in the future a larger FC stack is available to install on the AETV, the FC-BT-UC
vehicle may achieve better vehicle performance and fuel economy than the FC-BT
vehicle with a reasonable powertrain cost, and protect the FC and the battery systems.
Therefore, with the current vehicle components in the OU auto lab, it is not necessary to
purchase ultracapacitors.
7.2 FUTURE WORK
1. Optimization of the energy management strategy will be necessary in the future.
2. The validation of the vehicle models should be accomplished by the tests of the FC
operation under road conditions in the future.
3. The extended battery life cycle can be quantified into dollars.
112
4. The hybridization degree optimization based on large-scale simulations is needed to
find an optimal powertrain size of the full-size vehicle.
3. The powertrain design of the AETV has to consider the parameters and constraints
on component installation. In the future, the feasibility of particular powertrain
components arrangements also needs to be analyzed.
113
REFERENCES
1. A. Brooker et al., ADVISOR Documentation (2002): National Renewable Energy
Lab., Apr. 30, 2002.
2. Battery space.com. 2007. "Comparison between NiMH, Lead Acid and Li-Ion
Battery at Similar Energy Level (360Wh)." (accessed October 26, 2009). Available at
www.batteryspace.com/prod-specs/ComparisonPoly.pdf
3. Bauman, Jennifer. "A comparative study of fuel-cell-battery, fuel-cell-ultracapacitor,
and fuel-cell-battery-ultracapacitor vehicles." IEEE transactions on vehicular
technology 57 (2008): 760-769.
4. Bauman, Jennifer. "An analytical optimization method for improved fuel
cell-battery-ultracapacitor powertrain." IEEE transaction on vehicular technology
(2009): 3186-3197.
5. Bernard, Jérôme, and Sebastien Delprat. "Fuel-cell hybrid powertrain: toward
minimization of hydrogen consumption." IEEE transaction on vehicular technology 58
(2009): 3168-3176.
6. Bhargava A., et al.. “PEM Fuel Cell System Modeling with Liquid Fuel Processing
and Hydrogen Membranes.” Paper presented at the 210th Meeting of Electrochemical
Society (ECS), Cancun, Mexico, October 29-November 3, 2006.
7. Burke, Andrew. "Batteries and ultracapacitors for electric, hybrid, and fuel cell
vehicles." Proceedings of the IEEE 95 (2007): 806-820.
8. Butler, Karen. "A matlab-based modeling and simulation package for electric and
hybrid electric vehicle design." IEEE transaction on vehicular technology (1999):
1770-1778.
9. Candusso D., E. Rullière, and E. Toutain. “A fuel cell hybrid power source for a
small electric vehicle.” Power engineering (2001): 85-92.
10. Chan, C.. "The state of the art of electric, hybrid, and duel cell vehicles."
Proceedings of the IEEE (2007): 704-718.
11. Ferreira, Andre, and P. Jose. "Energy management fuzzy logic supervisory for
electric vehicle power supplies system." IEEE transactions on power electronics 23
(2008): 107-115.
114
12. Gao, Wenzhong. "Performance comparison of a fuel cell-battery hybrid powertrain
and a fuel cell-ultracapacitor hybrid powertrain." IEEE transactions on vehiclular
technology 54 (2005): 846-855.
13. Gregg. C. 2007. Alternative energy testbed electric vehicle and thermal
management system investigation. Master thesis, Ohio University.
14. Honda FCX Clarity-How FCX Works-Official Web Site. 2009. (accessed October 26,
2009). Available at http://automobiles. honda.com/fcx-clarity/how-fcx-works.aspx
15. Insight Central: Honda Insight Forum. 2008. (accessed October 26, 2009).
Available at http://www.insightcentral.net/encyclopedia/enbattery.html
16. Iqbal Husain. 2003. Alternative Energy Sources. In Electric and hybrid vehicles:
design fundamentals, 81-94. Florida: CRC Press, LLC.
17. Iqbal Husain. 2003. Vehicle Mechanics. In Electric and hybrid vehicles: design
fundamentals, 17-38. Florida: CRC Press, LLC.
18. James Larminie and John Lowry. 2003. Electric machines and their controllers. In
Electric vehicle technology explained, 143-152. New Jersey: John Wiley & Sons, INC.
19. J. Van Mierlo, P. Van den Bossche, and G. Maggetto, “Models of energy sources
for EV and HEV: fuel cells, batteries, ultracapacitors, flywheels and engine-generators.”
Journal of power sources (2003): 76-89.
20. Keith Wipke, Tony Markel, and Doug Nelson, “Optimizing energy management
strategy and degree of hybridization for a hydrogen fuel cell SUV.” EVS 18 Berlin
(2001).
21. K. Minjin and Young-Jun Sohn. "Fuzzy control based engine sizing optimization
for a fuel cell/battery hybrid mini-bus." Journal of powerer sources 178 (2008):
706-710.
22. “Mark9 SSL™ Product Manual and Integration Guide.” (2006). Canada: Ballard
Power Systems Inc.
23. Markel, T. and A. Brooker. "ADVISOR: a systems analysis tool for advanced
vehicle modeling." Journal of power sources (2002): 255-266.
115
24. Maxwell.com. 2007. "MC Energy Series BOOSTCAP Ultracapacitors." (accessed
February 28, 2010). Available at
www.maxwell.com/pdf/uc/datasheets/eol/MC_Cell_Energy_1009323_rev3.pdf
25. M. Ehsani, Y. Gao, S.E. Gay, and A. Emadi. 2010. Energy Storage. In Modern
Electric, Hybrid Electric, and Fuel Cell Vehicles: Fundamentals, Theory, and Design,
299-332. Florida: Taylor and Francis Group, LLC.
26. M. Ehsani, Y. Gao, S.E. Gay, and A. Emadi. 2010. Fuel Cell Vehicles. In Modern
Electric, Hybrid Electric, and Fuel Cell Vehicles: Fundamentals, Theory, and Design,
347-373. Florida: Taylor and Francis Group, LLC.
27. Mierlo, Joeri Van, Peter Bossche, and Gaston Naggetto. "Models of energy sources
for EV and HEV: fuel cells, batteries, ultracapacitors, flywheels and
engine-generators." Journal of power sources 128 (2004): 76-89.
28. Ogburn, M. 2000. Systems intergration, modeling, and validation of a fuel cell
hybrid electric vehicle. Master thesis, Virginia Polytechnic Institute and State
University.
29. O’Hayre Ryan, Suk-Won Cha, Whitney Colella, and Fritz B. Prinz. 2006. Fuel cell
performance. In Fuel cell fundamental, 3-19. New Jersey: John Wiley & Sons, INC.
30. O’Hayre Ryan, Suk-Won Cha, Whitney Colella, and Fritz B. Prinz. 2006. Fuel cell
thermodynamics. In Fuel cell fundamental, 23-58. New Jersey: John Wiley & Sons,
INC.
31. Pede, G. and A. Iacobazzi. "FC vehicle hybridisation: an affordable solution for an
energy-efficient FC powered drive train." Journal of power sources 125 (2004):
280-291.
32. Rodatz, Paul, Olivier Garcia, and Lino Guzzella. "Performance and operational
characteristics of a hybrid vehicle powered by fuel cells and supercapacitors." SAE
transactions 112 (2003): 692-703.
33. Rosario, Leon C. 2007. Power and energy management of multiple energy storage
systems in electric vehicles. PhD thesis, Cranfield University.
34. Sapienza, C. "Batteries analysis for FC-hybrid powertrain optimization."
International Journal of hydrogen energy 33 (2008): 3230-3234.
116
35. Sears, Jesse. 2009. “Lithium Ion Vs. Ni-Mh Battery.” Available at
www.ehow.com/about_5415114_lithium-ion-vs-nimh-battery.html
36. V.H. Johnson. “Battery performance models in ADVISOR.” Journal of power
sources (2002): 321-329.
37. Wipke, Keith, Matthew Cuddy, and Steven Burch. "Advisor 2.0: A
Second-Generation Advanced Vehicle Simulator for Systems Analysis." Paper
presented at the North American EV & Infrastructure Conference and Exposition
(NAEVI 98), Phoenix, Arizona, December 3-4, 1998.
38. Wipke, Keith, Matthew Cuddy, and Steven Burch. "ADVISOR 2.1: A user-friendly
advanced powertrain simulation using a combined backward/forward approach." IEEE
transactions on vehicular technology (1999): 1751-1761.
39. Zamfirescu, C. and I. Dincer. “Ammonia as a green fuel and hydrogen source for
vehicular applications.” Fuel processing technology 90 (2009): 729–737.
40. Zolot, Matthew D., Kelly, K., Keyser, M., Mihalic, M., and Persaran, A. “Thermal
evaluation of the HONDA Insight battery pack.” Paper presented at the 36th Intersociety
Energy Conversion Engineering Conference (IECEC.01), Savannah, Georgia, July
29-August 2, 2001.
117
APPENDIX
This appendix outlines the modified m-files used for modeling and simulation of the
AETV. In order to keep the data formats compatible with ADVISOR, the m-files are
programmed based on the existing component files in ADVISOR 2002 to fit the new
components. The modified data files and other codes include:
1) FC_Ballard_AETV.m;
2) VEH_FiberFab.m;
3) ESS_Dualess.m;
4) PTC_AETV.m;
5) TX_4SPD_AETV.m.
1) FC_Ballard_AETV.m;
% ADVISOR Data file: FC_Ballard_AETV.m
% Data source: Data compiled and derived from the Ballard fuel
cell product manual
% Data confidence level:
%
% Notes: Modeling results for a 4.8kW net fuel cell system
operating on hydrogen.
%
% Created on: Apr/05.2010
% By: Yin Wu, Ohio University, yw111408@ohio.edu
%
% Revision history at end of file.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% FILE ID INFO
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
fc_description='Model - 5kW Ballard PEM Fuel Cell System';
fc_version=2002; % version of ADVISOR for which the file was
118
generated
fc_proprietary=1; % 0=> non-proprietary, 1=> proprietary, do
not distribute
fc_validation=0; % 0=> no validation, 1=> data agrees with
source data,
% 2=> data matches source data and data collection methods
have been verified
disp(['Data loaded: FC_Ballard_AETEV.m ',fc_description]);
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% FUEL USE AND EMISSIONS MAPS
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% (g/s), fuel consumption indexed vertically by fc_map_spd
and
% horizontally by fc_map_trq
fc_pwr_map=[0.3 0.6 1.1 1.5 2.16 3.7 4.4]*1000; % W (net)
including parasitic losses
fc_eff_map=[0.1 0.18 0.304 0.316 0.303 0.272 0.25]; %
efficiency indexed by fc_pwr
% one low power point is estimated
% create fuel use map (g/s)
fc_fuel_lhv=120.0*1000; % (J/g), lower heating value of the
fuel
fc_fuel_map=fc_pwr_map./fc_eff_map./fc_fuel_lhv; % used in
block diagram
% create fuel consumption map (g/kWh)
fc_fuel_map_gpkWh=(1./fc_eff_map)/fc_fuel_lhv*3600*1000; %
used in gui_inpchk plots
% (g/s), engine out HC emissions indexed vertically by
fc_map_spd and
% horizontally by fc_map_trq
fc_hc_map=zeros(size(fc_fuel_map));
% (g/s), engine out HC emissions indexed vertically by
fc_map_spd and
% horizontally by fc_map_trq
fc_co_map=zeros(size(fc_fuel_map));
119
% (g/s), engine out HC emissions indexed vertically by
fc_map_spd and
% horizontally by fc_map_trq
fc_nox_map=zeros(size(fc_fuel_map));
% (g/s), engine out PM emissions indexed vertically by
fc_map_spd and
% horizontally by fc_map_trq
fc_pm_map=zeros(size(fc_fuel_map));
% (g/s), engine out O2 indexed vertically by fc_map_spd and
% horizontally by fc_map_trq
fc_o2_map=zeros(size(fc_fuel_map));
fc_emis=0;
% boolean 0=no emis data; 1=emis data
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Cold Engine Maps
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
fc_cold=0;
% boolean 0=no cold data; 1=cold data exists
fc_cold_tmp=20; %deg C
fc_fuel_map_cold=zeros(size(fc_fuel_map));
fc_hc_map_cold=zeros(size(fc_fuel_map));
fc_co_map_cold=zeros(size(fc_fuel_map));
fc_nox_map_cold=zeros(size(fc_fuel_map));
fc_pm_map_cold=zeros(size(fc_fuel_map));
%Process Cold Maps to generate Correction Factor Maps
names={'fc_fuel_map','fc_hc_map','fc_co_map','fc_nox_map'
,'fc_pm_map'};
for i=1:length(names)
%cold to hot raio, e.g. fc_fuel_map_c2h =
fc_fuel_map_cold ./ fc_fuel_map
eval([names{i},'_c2h=',names{i},'_cold./(',names{i},'+eps
);'])
end
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% DEFAULT SCALING
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
fc_pwr_scale=1.0;
% -- scale fc power
%the following variable is not used directly in modelling and
120
should always be equal to one
%it's used for initialization purposes
fc_eff_scale=1.0; % -- scale the efficiency
fc_trq_scale=1.0; % -- required only for autosize and
optimization routines
fc_spd_scale=1.0; % -- required only for autosize and
optimization routines
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% OTHER DATA
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
fc_fuel_cell_model=2; % 1--> polarization curves, 2--> pwr
vs. eff, 3--> gctool model
fc_fuel_type='Hydrogen';
fc_fuel_den=0.018*1000; % (g/l), density of the fuel
fc_fuel_lhv=120.0*1000; % (J/g), lower heating value of the
fuel
fc_max_pwr=5; % kW
peak engine power
%fc_base_mass=5.0*fc_max_pwr;
% kg
mass of the
fuel cell stack, assuming mass penalty of 5 kg/kW (DOE 2000
status)
fc_base_mass=1.6364*fc_max_pwr;
% kg
mass of the
fuel cell stack, assuming mass penalty of 1.6364kg/kW (Based
on Manual of Ballard Fuel Cell)
fc_acc_mass=2.5*fc_max_pwr;
% kg
mass of fuel
cell accy's, electrics, cntrl's - assumes mass penalty of 2.5
kg/kW ESTIMATE
target_range=80; % mi
target_fe=80; % mpgge
gas_lhv=42.6*1000; % J/g
gas_dens=0.749*1000; % g/l
target_fuel_storage_spec_energy=2000; % Wh/kg from DOE
targets
fc_fuel_mass=target_range/target_fe*3.785*gas_lhv*gas_den
s/3600/target_fuel_storage_spec_energy;
% (kg), fuel storage mass assuming specified range and
specified fuel economy
fc_mass=fc_base_mass+fc_acc_mass+fc_fuel_mass; % kg total
engine/fuel system mass
%fc_mass=88;
%fc_mass=max(fc_map_spd.*fc_max_trq)/132; % (kg), mass of
the engine, assuming specific power of 132 W/kg
121
clear gas_lhv gas_dens target_range target_fe
target_fuel_storage_spec_energy
% variables not applicable to a fuel cell but needed for use
of engine block diagram
fc_tstat=80;
% C
engine coolant
thermostat set temperature (typically 80 +/- 5 C)
fc_cp=500;
% J/kgK ave cp of engine
(iron=500, Al or Mg = 1000)
fc_h_cp=500;
% J/kgK ave cp of hood &
engine compartment (iron=500, Al or Mg = 1000)
%fc_ext_sarea=0.3;
% m^2
exterior surface
area of engine
fc_ext_sarea=2*(0.3*0.3)+4*(0.3*0.6);
% m^2
exterior surface area of engine
fc_hood_sarea=1.5;
% m^2
surface area of
hood/eng compt.
fc_emisv=0.8;
%
emissivity of
engine ext surface/hood int surface
fc_hood_emisv=0.9;
%
emissivity hood ext
fc_h_air_flow=0.0;
% kg/s
heater air flow rate
(140 cfm=0.07)
fc_cl2h_eff=0.7;
% -ave cabin heater HX
eff (based on air side)
fc_c2i_th_cond=500;
% W/K
conductance btwn
engine cyl & int
fc_i2x_th_cond=500;
% W/K
conductance btwn
engine int & ext
fc_h2x_th_cond=10;
% W/K
conductance btwn
engine & engine compartment
%fc_ex_pwr_frac=[0.40 0.30];
% -frac of waste heat that goes to exhaust as func of engine speed
fc_ex_pwr_frac=[0.20 0.10];
% -frac of waste heat that goes to exhaust as func of power level,
(SAE 2000-01-0373)
%fc_exflow_map=fc_fuel_map*(1+14.5);
% g/s
ex gas flow map: for SI engines, exflow=(fuel use)*[1 +
(stoic A/F ratio)]
fc_exflow_map=fc_fuel_map*(1+91);
% g/s ex
gas flow map: for fuel cell exflow=(fuel use)*[1 + (A/F
ratio)], where
1.5*H2+2.0*(O2+3.774*N2)==>H20+0.5H2+0.5O2+2.0*3.774*N2,
122
where 1.5=anode stoich, and 2.0=cathode stoich
fc_waste_pwr_map=fc_fuel_map*fc_fuel_lhv - fc_pwr_map;
W
tot FC waste heat = (fuel pwr) - (mech out pwr)
fc_ex_pwr_map=zeros(size(fc_waste_pwr_map));
% W
initialize size of ex pwr map
for i=1:length(fc_pwr_map)
%
fc_ex_pwr_map(i)=fc_waste_pwr_map(i)*interp1([min(fc_pwr_
map) max(fc_pwr_map)],fc_ex_pwr_frac,fc_pwr_map(i)); % W
pwr map of waste heat to exh
end
%fc_extmp_map=fc_ex_pwr_map./(fc_exflow_map*1089/1000) +
20; % W
EO ex gas temp = Q/(MF*cp) + Tamb (assumes engine
tested ~20 C)
fc_extmp_map=fc_ex_pwr_map./(fc_exflow_map*1145/1000) +
20; % W
EO ex gas temp = Q/(MF*cp) + Tamb (assumes engine
tested ~20 C) (cp based on exhaust composition listed above)
% user definable mass scaling function
fc_mass_scale_fun=inline('(x(1)*fc_trq_scale+x(2))*(x(3)*
fc_spd_scale+x(4))*(fc_base_mass+fc_acc_mass)+fc_fuel_mas
s','x','fc_spd_scale','fc_trq_scale','fc_base_mass','fc_a
cc_mass','fc_fuel_mass');
fc_mass_scale_coef=[1 0 1 0]; % coefficients of mass scaling
function
%%%%%%%%%%%%%%%%%%
% auxiliary systems
%%%%%%%%%%%%%%%%%%
% air compressor/blower
% air flow rate provided by compressor (g/s)
% NOTE: this is defined as SLPM and converted to g/s
fc_air_comp_map=[0.0427 0.356 0.854 1.423 1.708 1.993 2.42
4.27]; % Air flow entering compressor, g/s
% air compressor power requirements indexed by
fc_air_comp_map
fc_air_comp_pwr=[0.46 1.9 2.21 2.42 3.15 4.01 5.37
7.44]*100; % air flow power, W, scaled down from the 50kw fuel
cell model in ADVISOR
% ratio of fuel to air (g/s)/(g/s)
%
fc_fuel_air_ratio=mean((fc_fuel_map*fc_cell_num)./fc_air_
123
comp_map); % average of ratio of two input streams
% fuel pump (for liquid fuels only)
% fuel flow rate provided by fuel pump (g/s)
% NOTE: this is defined as SLPM and converted to g/s
% fc_fuel_pump_map=fc_fuel_map*fc_cell_num; % g/s
% fuel pump power requirements indexed by fc_fuel_pump_map
% fc_fuel_pump_pwr=[0.647 0.757 0.787 0.811 0.880 0.942 1.02
1.443]*100; % W, scaled down from the 50kw fuel cell model
in ADVISOR
% NOTE: the above auxiliary load represents the radiator fan
and coolant pump load - data provided indexed by fuel usage
% coolant pump
% coolant flow rate provided by pump (g/s)
% NOTE: this is defined as SLPM and converted to g/s
fc_coolant_pump_map=[0 1000]; % g/s, % not applicable
% coolant pump power requirements indexed by
fc_coolant_pump_map
fc_coolant_pump_pwr=[0 0]*1000; % W, % not applicable
% fixed coolant flow rate
fc_coolant_flow_rate=0; % g/s, % not applicable
% coolant specific heat capacity
fc_coolant_cp=4.1843; % J/(g*K) water
%fc_coolant_cp=(0.5*4.1843+0.5*2.474); % J/(g*K) 50/50
water/ethylene glycol blend
% water pump (for alcohol fuels only)
% water flow rate provided by water pump (g/s)
% NOTE: this is defined as SLPM and converted to g/s
fc_water_pump_map=[0 1000]; % g/s, % not applicable
% water pump power requirements indexed by fc_water_pump_map
fc_water_pump_pwr=[0 0]*1000; % W % not applicable
% supply ratio of fuel to water
fc_fuel_water_ratio=32/1000; % not applicable
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% REVISION HISTORY
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% 04/10/2010 (yw): file created based on FC_ANL50H2.m
% 04/20/2010 (yw): scaled down fc_air_comp_map,
124
fc_fuel_pump_pwr, and fc_air_comp_pwr for AETEV case from the
50kw fuel cell
% 04/20/2010 (yw): replaced target_range and target_fe for
the fuel cell
% implemented in AETEV.
% 04/20/2010 (yw): replaced fc_fuel_map calculation with the
case of 4.8kW
% ballard fuel cell
2) VEH_FiberFab.m;
% ADVISOR data file: VEH_FiberFab.m
%
% Data source:
%
% Data confirmation:
%
% Notes: Defines road load parameters for FiberFab Aztec GT
body.
%
% Created on: 05-Apr-2010
% By: Yin Wu, Ohio university, yw111408@ohio.edu
%
% Revision history at end of file.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% FILE ID INFO
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
veh_description='OU AETEV';
veh_version=2002; % version of ADVISOR for which the file was
generated
veh_proprietary=0; % 0=> non-proprietary, 1=> proprietary,
do not distribute
veh_validation=0; % 0=> no validation, 1=> data agrees with
source data,
%
2=> data matches source data and
data collection methods have been verified
125
disp(['Data loaded: VEH_FiberFab - ',veh_description])
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% PHYSICAL CONSTANTS
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
veh_gravity=9.81;
% m/s^2
veh_air_density=1.2; % kg/m^3
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% VEHICLE PARAMETERS
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Note on vehicle mass:
%
If you wish to accurately set your total vehicle mass
%
in the A2 GUI, you should use the mass override
%
checkbox and enter in the value you need.
%
The glider mass below is just an estimate that gives
910kg for a 4.8 kW
%
scaled fuel cell system in a hybrid vehicle with
4-speed transmission.
veh_glider_mass=550; % (kg), vehicle mass w/o propulsion
system (fuel converter,
% exhaust aftertreatment, drivetrain,
motor, ESS, generator)
veh_CD=0.381; % (--), coefficient of aerodynamic drag
veh_FA=1.4;
% (m^2), frontal area
% for the eq'n:
rolling_drag=mass*gravity*(veh_1st_rrc+veh_2nd_rrc*v)
%veh_1st_rrc=0.009; % (--)
%veh_2nd_rrc=0;
% (s/m)
% fraction of vehicle weight on front axle when standing still.
Data from a
% 1970 VW beetle: weight distribution F/R = 42%.
veh_front_wt_frac=0.42; % (--)
% height of vehicle center-of-gravity above the road,data
from a 1970 VW beetle.
veh_cg_height=0.381;
% (m)
% vehicle wheelbase, from center of front tire patch to center
of rear
% patch. Data from a 1970 VW beetle.
veh_wheelbase=2.4; % (m)
126
veh_cargo_mass=136; %kg
cargo mass
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% REVISION HISTORY
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3) ESS_Dualess.m;
% ADVISOR data file: ESS_Dualess.m = ESS1_NIMH6.m +
ESS2_UC2_Maxwell_temp.m
%
% ESS1 Data source:
%
% Insight file created from NREL lab test data
%
NREL test data from testing entire Insight Battery
Pack Jan.2001 (Insight Model Year 2000)
%
%
Insight pack is reported to be same technology as
Japanese Prius (1998) with 20 modules instead of 40
%
Battery Type: NiMH Spiral Wound
%
Nominal Cell Voltage: 1.2V
%
Total Cells: 120 (6 cells x 20 modules) (40
modules for Japanese Prius)
%
Nominal Voltage: 144 V (288 V for Japanese
Prius)
%
Published Capacity: 6.5 Ah
%
%
Tests performed at 25 deg C following the PNGV Hybrid
Pulse Power Characterization (HPPC) Procedure
%
% Data confirmation:
%
This data comes from Testing at NREL in the Battery
Thermal
%
Management Lab
%
% Created on: 2/6/01
% By: KJK, NREL
127
%
% ESS2 Data source:
% Testing at NREL's Battery Thermal Management Testing
facility.
%
% Data confirmation:
%
% Notes:
% These parameters describe the Maxwell BCAP3000
Ultracapacitor.
%
% Created on: 11/01/01
% By: Tony Markel and Matt Zolot, NREL
%
% Revision history at end of file.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%
ess_module_num=20; %20 modules in INSIGHT pack
ess2_module_num=20; % The upper bound is cells, the lower
bound is 11 cells
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% FILE ID INFO
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ess_description='Spiral Wound NiMH Used in Insight & Japanese
Prius and Maxwell PC3000 Ultracapacitor';
ess_version=2002; % version of ADVISOR for which the file was
generated
ess_proprietary=0; % 0=> non-proprietary, 1=> proprietary,
do not distribute
ess_validation=0; % 0=> no validation, 1=> data agrees with
source data,
% 2=> data matches source data and data collection methods
have been verified
disp(['Data loaded: ESS1_NIMH6&ESS2_UC2_Maxwell_temp ',ess_description])
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% SOC RANGE over which data is defined
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ess_soc=[0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1];
% (--)
128
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Temperature range over which data is defined
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% The following data was obtained at 25 deg C.
values are the same for all temperatures
ess_tmp=[0 25]; % (C) place holder for now
Assume all
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% LOSS AND EFFICIENCY parameters (from ESS_Prius_pack)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Parameters vary by SOC horizontally, and temperature
vertically
% the average of 5 discharge cycles at 6.5A at 25 deg C was
5.995Ah.
% Data (Ah): 6.030 5.973 5.990 5.989 5.995
ess_max_ah_cap=[
6.0
6.0
]; % (A*h), max. capacity at 6.5 A, indexed by ess_tmp
% average coulombic (a.k.a. amp-hour) efficiency below,
indexed by ess_tmp
% Coulombic Efficiency - 6.5A discharge, 3A charge to dV/dt
of 0.035V across the pack (120 cells)
% DATA: 90.991 90.308 90.470 90.386 90.327
ess_coulombic_eff=[
.905
.905
]; % (--);
% module's resistance to being discharged, indexed by ess_soc
and ess_tmp
% The discharge resistance is the average of 4 tests from 10
to 90% soc at the following
% discharge currents: 6.5, 6.5, 18.5 and 32 Amps
% The 0 and 100 % soc points were extrapolated
ess_r_dis=[
0.0377 0.0338 0.0300 0.0280 0.0275 0.0268
0.0269 0.0273 0.0283 0.0298 0.0312
0.0377 0.0338 0.0300 0.0280 0.0275 0.0268
0.0269 0.0273 0.0283 0.0298 0.0312
129
];
% module's resistance to being charged, indexed by ess_soc
and ess_tmp
% The discharge resistance is the average of 4 tests from 10
to 90% soc at the following
% discharge currents: 5.2, 5.2, 15 and 26 Amps
% The 0 and 100 % soc points were extrapolated
ess_r_chg=[
0.0235
0.0220 0.0205 0.0198 0.0198 0.0196
0.0198 0.0197 0.0203 0.0204 0.0204
0.0235 0.0220 0.0205 0.0198 0.0198 0.0196
0.0198 0.0197 0.0203 0.0204 0.0204
];
% module's open-circuit (a.k.a. no-load) voltage, indexed by
ess_soc and ess_tmp
ess_voc=[
7.2370 7.4047 7.5106 7.5873 7.6459 7.6909
7.7294 7.7666 7.8078 7.9143 8.3645
7.2370 7.4047 7.5106 7.5873 7.6459 7.6909
7.7294 7.7666 7.8078 7.9143 8.3645
];
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% LIMITS (from ESS_Prius_pack)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ess_min_volts=6;% 1 volt per cell times 6 cells lowest from
data was 255V so far 8/26/99
ess_max_volts=9; % 1.5 volts per cell times 6 cells highest
from data so far was 361V 8/26/99
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% OTHER DATA (from ESS_Prius_pack except where noted)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ess_module_mass=(44*.4536)/20; % (kg), mass of Insight pack
(44 lb Automotive News, July 12) divided by 20 modules
%Module mass including external case
%ess_acc_mass=(35-44*.4536)/20;
130
%ess_base_mass=(44*.4536)/20;
%ess_module_num=20; %Honda Insight
%ess_module_mass=ess_acc_mass+ess_base_mass; % (kg), mass
of Insight pack (including the fundation case)
ess_cap_scale=1; % scale factor for module max ah capacity
% user definable mass scaling relationship
ess_mass_scale_fun=inline('(x(1)*ess_module_num+x(2))*(x(
3)*ess_cap_scale+x(4))*(ess_module_mass)','x','ess_module
_num','ess_cap_scale','ess_module_mass');
ess_mass_scale_coef=[1 0 1 0]; % coefficients in
ess_mass_scale_fun
% user definable resistance scaling relationship
ess_res_scale_fun=inline('(x(1)*ess_module_num+x(2))/(x(3
)*ess_cap_scale+x(4))','x','ess_module_num','ess_cap_scal
e');
ess_res_scale_coef=[1 0 1 0]; % coefficients in
ess_res_scale_fun
%battery thermal model
ess_th_calc=1;
% -0=no
ess thermal calculations, 1=do calc's
ess_mod_cp=800;
% 800 J/kgK
ave heat capacity of module from calorimeter test
ess_set_tmp=35;
% C
thermostat temp of module when cooling fan comes on
%ess_area_scale=1.6*(ess_module_mass/11)^0.7;
% -if module dimensions are unknown, assume rectang shape and
scale vs PB25
ess_dia=0.0322;% m
ess_length=0.374; %m
ess_mod_sarea=pi*ess_dia*ess_length;
% m^2
total
module surface area exposed to cooling air (typ rectang
module)
ess_mod_airflow=0.01;
% kg/s
cooling air mass flow rate across module (20 cfm=0.01 kg/s
at 20 C)
ess_mod_flow_area=2*0.00317*ess_length;
% m^2
cross-sec flow area for cooling air per module (assumes 10-mm
gap btwn mods)
131
ess_mod_case_thk=.1/1000;
% m
thickness of module case (typ from Optima)
ess_mod_case_th_cond=0.20;
% W/mK
thermal conductivity of module case material (typ polyprop
plastic - Optima)
ess_air_vel=ess_mod_airflow/(1.16*ess_mod_flow_area); %
m/s ave velocity of cooling air
ess_air_htcoef=30*(ess_air_vel/5)^0.8;
% W/m^2K
cooling air heat transfer coef.
ess_th_res_on=((1/ess_air_htcoef)+(ess_mod_case_thk/ess_m
od_case_th_cond))/ess_mod_sarea; % K/W tot thermal res key
on
ess_th_res_off=((1/4)+(ess_mod_case_thk/ess_mod_case_th_c
ond))/ess_mod_sarea; % K/W tot thermal res key off (cold
soak)
% set bounds on flow rate and thermal resistance
ess_mod_airflow=max(ess_mod_airflow,0.001);
ess_th_res_on=min(ess_th_res_on,ess_th_res_off);
clear ess_dia ess_length ess_mod_sarea ess_mod_flow_area
ess_mod_case_thk ess_mod_case_th_cond ess_air_vel
ess_air_htcoef ess_area_scale
%------------------------------------------------------------------------%------------------------------------------------------------------------%UC info
% Maxwell 3000 reference: Maxwell UC specifications and test
data for PC2500 by NERL
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% SOC RANGE over which data is defined
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ess2_soc=[0:.1:1]; % (--) % not used in model
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Temperature range over which data is defined
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ess2_tmp=[0 25 40]; % (C) % from test data
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
132
% Current range over which data is defined
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ess2_i=[-225 -112.5 -56.3 56.3 112.5 225];
internal resistance data
% (C) indexes
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% LOSS AND EFFICIENCY parameters
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% average coulombic (a.k.a. amp-hour) efficiency below,
indexed by ess_tmp (row) and ess_i (col)
ess2_coulombic_eff=[
0.999314173 0.997716624 0.99587234 0.99587234
0.997716624 0.999314173
0.998253057 0.996273292 0.990403543 0.990403543
0.996273292 0.998253057
0.996265099 0.99481399 0.993372248 0.993372248
0.99481399 0.996265099]; % (--); from test data
% module's resistance, indexed by ess_i and ess_tmp
ess2_r=[
0.29 0.29 0.29 0.29 0.29 0.29
0.29 0.29 0.29 0.29 0.29 0.29
0.29 0.29 0.29 0.29 0.29 0.29]/1000; % (ohm) from test data
% module's capacitance, indexed by ess_i and ess_tmp
ess2_cap=[
3000.00 3000.00 3000.00 3000.00 3000.00 3000.00
3000.00 3000.00 3000.00 3000.00 3000.00 3000.00
3000.00 3000.00 3000.00 3000.00 3000.00 3000.00]; %
(Farads) from test data
%ess2_voc=[1.25:((2.5-1.25)/10):2.5;1.25:((2.5-1.25)/10):
2.5;1.25:((2.5-1.25)/10):2.5];
%use eps instead of 1.25 so
voltage range can be set by SOC range
ess2_voc=[eps:((2.5-eps)/10):2.5;eps:((2.5-eps)/10):2.5;e
ps:((2.5-eps)/10):2.5];
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% LIMITS
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ess2_max_volts=3; % from manufacturer recommendations
133
ess2_min_volts = 0.001; % common practice is 50% of
ess_100soc_volts, but really depends on power
electronics/motor
% drive minimum voltage requirements, so should be set to (a)
50% of V(100%soc) by convention, or (b)
mc_min_volts/ess_module_num, or (c) set to zero for no
influence.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% OTHER DATA
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ess2_module_mass=0.55; % (kg), mass of a single 2.7 V cell
% user definable mass scaling relationship
ess2_mass_scale_fun=inline('(x(1)*ess2_module_num+x(2))*(
x(3)*ess2_cap_scale+x(4))*(ess2_module_mass)','x','ess2_m
odule_num','ess2_cap_scale','ess2_module_mass');
ess2_mass_scale_coef=[1 0 1 0]; % coefficients in
ess_mass_scale_fun
% user definable resistance scaling relationship
ess2_res_scale_fun=inline('(x(1)*ess2_module_num+x(2))/(x
(3)*ess2_cap_scale+x(4))','x','ess2_module_num','ess2_cap
_scale');
ess2_res_scale_coef=[1 0 1 0]; % coefficients in
ess_res_scale_fun
ess2_parallel_mod_num=1; % a default value for number of
modules placed in parallel, model treats ideally as a higher
capacitance ultracap.
ess2_cap_scale=1; % scale factor for module max ah capacity
% uc thermal model
ess2_th_calc=1;
% -0=no
ess thermal calculations, 1=do calc's
ess2_mod_cp=(1471.3+1614.6)/2;
% J/kgK ave
heat capacity of module (40 to 25C, and 10 to 25C)
ess2_set_tmp=35;
% C
thermostat temp of module when cooling fan comes on
ess2_mod_sarea=(60.7*2*pi*138)/1000000; % m^2
total
module surface area exposed to cooling air
134
((width+depth)*2*height) from spec sheet
ess2_mod_airflow=0.07/5;
% kg/s
cooling air mass flow rate across module (140 cfm=0.07 kg/s
at 20 C)
ess2_mod_flow_area=
((12.5+60.7)^2-pi*(60.7/2)^2)/1000000; % m^2
cross-sec
flow area for cooling air per module (12.5-mm gap btwn mods)
ess2_mod_case_thk=2/1000;
% m
thickness of module case
ess2_mod_case_th_cond=0.20;
% W/mK
thermal conductivity of module case material (estimate)
ess2_air_vel=ess2_mod_airflow/(1.16*ess2_mod_flow_area); %
m/s ave velocity of cooling air
ess2_air_htcoef=30*(ess2_air_vel/5)^0.8;
% W/m^2K
cooling air heat transfer coef.
ess2_th_res_on=((1/ess2_air_htcoef)+(ess2_mod_case_thk/es
s2_mod_case_th_cond))/ess2_mod_sarea; % K/W tot thermal
res key on
ess2_th_res_off=((1/4)+(ess2_mod_case_thk/ess2_mod_case_t
h_cond))/ess2_mod_sarea; % K/W tot thermal res key off (cold
soak)
% set bounds on flow rate and thermal resistance
ess2_mod_airflow=max(ess2_mod_airflow,0.001);
ess2_th_res_on=min(ess2_th_res_on,ess2_th_res_off);
clear ess2_mod_sarea ess2_mod_flow_area ess2_mod_case_thk
ess2_mod_case_th_cond ess2_air_vel ess2_air_htcoef
% for stand-alone debugging, normally defined elsewhere
%ess_on=1;
%mc_min_volts=0;
%ess2_init_soc=0.001;
ess2_mod_init_tmp=20;
%cyc_mph=[0 35; 1 35];
air_cp=1200;
amb_tmp=20;
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% REVISION HISTORY
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% 07/01/2010: file created from ess_nimh6.m and
uc_maxwell_temp.m by Yin Wu, Ohio university
% 07/14/2010: ess2_max_volt and ess2_module_mass is updated
135
based on UC_maxwell_BCAP_3000 spec sheet
% thermal data based on module spec are waiting to be updated.
% 07/15/2010: ess2 module surface area, cross-sec flow area,
are updated.
% 07/16/2010: ess2 module number and parallel module number
need to be updated based on a 10kW peak power request.
4) PTC_AETV.m;
% ADVISOR data file: PTC_AETV.m
% Data source: AETV transmission specifications.
%
% Data confirmation:
% Notes:
% Defines all powertrain control parameters, including
gearbox and hybrid
% controls for the AETV to ensure proper operation.
%
% Created on: 28-April-2010
% By: Yin Wu, Ohio Unviersity, yw111408@ohio.edu
% Revision history at end of file.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%
if ~exist('update_cs_flag')
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% FILE ID INFO
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
ptc_description='Powertrain control of AETEV with the fuel
cell-battery hybrid powertrain';
ptc_version=2002; % version of ADVISOR for which the hfile
was generated
ptc_proprietary=0; % 0=> non-proprietary, 1=> proprietary,
do not distribute
ptc_validation=0; % 1=> no validation, 1=> data agrees with
source data,
% 2=> data matches source data and data collection methods
136
have been verified
disp(['Data loaded: PTC_OUFB - ',ptc_description])
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% GEARBOX CONTROL
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% fractional engine load {=(torque)/(max torque at speed)}
above which a
% downshift is called for, indexed by gb_gearN_dnshift_spd
gb_gear1_dnshift_load=[2 2]; % (downshift is requested in
gear1)
gb_gear2_dnshift_load=[2 2]; % (downshift is requested in
gear2)
% fractional engine load {=(torque)/(max torque at speed)}
below which an
% upshift is called for, indexed by gb_gearN_upshift_spd
gb_gear1_upshift_load=[-1 -1]; % (fraction of max load at
current speed at which upshift is requested in gear1)
gb_gear2_upshift_load=[-1 -1]; % (fraction of max load at
current speed at which upshift is requested in gear2)
gb_gear1_dnshift_spd=[0 1000]; % (rad/s)speed at which
downshift is requested in gear1
gb_gear2_dnshift_spd=[0 1000]; % (rad/s)speed at which
downshift is requested in gear2
gb_gear1_upshift_spd=[0 1000]; % (rad/s)speed at which
upshift is requested in gear1
gb_gear2_upshift_spd=[0 1000]; % (rad/s)speed at which
upshift is requested in gear2
% convert old shift commands to new shift commands
gb_upshift_spd={gb_gear1_upshift_spd; ...
gb_gear2_upshift_spd}; % (rad/s)
gb_upshift_load={gb_gear1_upshift_load; ...
gb_gear2_upshift_load}; % (--)
gb_dnshift_spd={gb_gear1_dnshift_spd; ...
gb_gear2_dnshift_spd}; % (rad/s)
gb_dnshift_load={gb_gear1_dnshift_load; ...
gb_gear2_dnshift_load}; % (--)
clear gb_gear*shift* % remove unnecessary data
137
% fixes the difference between number of shift vectors and
gears in gearbox
if length(gb_upshift_spd)<length(gb_ratio)
start_pt=length(gb_upshift_spd);
for x=1:length(gb_ratio)-length(gb_upshift_spd)
gb_upshift_spd{x+start_pt}=gb_upshift_spd{1};
gb_upshift_load{x+start_pt}=gb_upshift_load{1};
gb_dnshift_spd{x+start_pt}=gb_dnshift_spd{1};
gb_dnshift_load{x+start_pt}=gb_dnshift_load{1};
end
end
% duration of shift during which no torque can be transmitted
gb_shift_delay=0; % (s)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% HYBRID CONTROL STRATEGY
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%ess_init_soc=0.7; % (--), initial battery SOC; now this
is inputed from the simulation screen
cs_hi_soc=0.8; % (--), highest desired battery state of
charge
cs_lo_soc=0.4; %(--), lowest desired battery state of
charge
cs_fc_init_state=0; % (--), initial FC state; 1=> on, 0=>
off
cs_min_pwr=calc_max_pwr('fuel_converter')/fc_pwr_scale*10
00*0.25;% (W), minimum operating power
cs_max_pwr=calc_max_pwr('fuel_converter')/fc_pwr_scale*10
00*.95;% (W), maximum operating power (exceeded only if
SOC<cs_lo_soc)
cs_charge_pwr=4000;% (W), extra power output when
(cs_lo_soc+cs_hi_soc)/2-SOC=1
cs_min_off_time=90;% (s), minimum time fc remains off,
enforced unless SOC<=cs_lo_soc
cs_max_pwr_rise_rate=2000;% (W/s),
cs_max_pw_rise_rate*fc_spd_scale*fc_trq_scale is the
fastest the fuel converter power command can increase
cs_max_pwr_fall_rate=-3000;% (W/s),
cs_max_pwr_fall_rate*fc_spd_scale*fc_trq_scale is the
138
fastest the fuel converter power command can decrease (this
number < 0)
cs_charge_deplete_bool=1; % boolean 1=> use charge deplete
strategy, 0=> use charge sustaining strategy
fc_fuelcell_warmup_bool=0; % boolean 1=> enforce warmup pwr
limitations, 0=> do not enforce warmup pwr limitations
%ess_ultracap_bool=0; % boolean 1=> use ultracapacitor
parameters, 0=> use
%battery parameters
end
%%%%%%%%%%%%%%% START OF SPEED DEPENDENT SHIFTING
INFORMATION %%%%%%%%%%%%%
% Data specific for SPEED DEPENDENT SHIFTING in the (PRE_TX)
GEARBOX CONTROL
% BLOCK in VEHICLE CONTROLS <vc>
% --implemented for all powertrains of AETEV
%
tx_speed_dep=1; % Value for the switch in the gearbox control
%
If tx_speed_dep=1, the speed dependent gearbox
is chosen
%
If tx_speed_dep=0, the engine load dependent
gearbox is chosen
%
% speeds to use for upshift transition (shifting while
accelerating)
tx_spd_dep_upshift = [
0/3.6, 1
24.4/3.6, 1
24.4/3.6, 2
40/3.6, 2];
% speeds to use for downshift transition (shifting while
decelerating)
tx_spd_dep_dnshift = [
0/3.6, 1
24.4/3.6, 1
24.4/3.6, 2
40/3.6, 2];
%%%%%%%%%%%%%%% END OF SPEED DEPENDENT
139
SHIFTING %%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% CLEAN UP
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% REVISION HISTORY
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% 04/20/2010: Build up the new powertrain control strategy
for OU fuel cell
% battery hybrid vehicle based on PTC_FUELCELL.m
% 04/25/2010: modified variables for speed dependent shifting
5) TX_4SPD_AETV.m.
% ADVISOR data file: TX_4SPD_AETV.m
%
% Data source:
%
% Data confirmation:
%
% Notes: 94% efficient 4 spd gearbox
%
% Created on: Apr/05/10
% By: Yin Wu, Ohio University, yw111408@ohio.edu
%
% Revision history at end of file.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%
%Description of type of transmission(important in
determining what block diagram
%to run in gui_run_simulation)
tx_type='manual 4 speed';
tx_version=2002;% version of ADVISOR for which the file was
generated
140
tx_description = 'manual 4 speed transmission with 94%
efficiency gearbox for AETEV';
disp(['Data loaded: TX_4SPD_AETEV - ',tx_description]);
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% INITIALIZE
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% The tested transmission had four gears, with the gear ratios
listed
% as the first four entries in 'gb_ratio,' below.
gb_ratio=[3.6 1.94 1.22 0.82]*4.4;
gb_gears_num=4;
%TX_VW % FILE ID, LOSSES
gb_mass=41; % (kg), mass of the gearbox
%the following variable is not used directly in modelling and
should always be equal to one
%it's used for initialization purposes
gb_eff_scale=1;
gb_inertia=0;
% (kg*m^2), gearbox rotational inertia
measured at input; unknown
% trq and speed scaling parameters
gb_spd_scale=1;
gb_trq_scale=1;
%final drive variables
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% LOSSES AND EFFICIENCIES
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
fd_loss=0;
% (Nm), constant torque loss in final drive,
measured at input
% 94% efficient over all torques, speeds, and gears
% note: normally, the tx_eff_map matrix is 3-D (one
"z-direction" level per gear); however,
% ...in this case, only 1st gear needs to be specified. When
higher level gears are asked for,
% ...the algorithm responds by giving the best/closest data
141
it has--1st gear. See lib_transmission/gearbox/loss
% ...for more details. --mpo, 9-July-2001
tx_map_spd=[0 10000]; % speed of transmission shaft output
(wheel-side of transmission) in rad/s
tx_map_trq=[-10000 10000]; % torque of transmission shaft
output (wheel-side of transmission) in Nm
tx_eff_map=[0.94 0.94;0.94 0.94]; % transmission efficiency;
row index is tx_map_spd, col index is tx_map_trq
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% OTHER DATA
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
fd_ratio=1;
% (--), =(final drive input speed)/(f.d.
output speed)
fd_inertia=0; % (kg*m^2), rotational inertia of final drive,
measured at input
fd_mass=0; % (kg), mass of the final drive
tx_mass=gb_mass+fd_mass;% (kg), mass of the gearbox + final
drive=(transmission)
% user definable mass scaling relationship
tx_mass_scale_fun=inline('(x(1)*gb_trq_scale+x(2))*(x(3)*
gb_spd_scale+x(4))*(fd_mass+gb_mass)','x','gb_spd_scale',
'gb_trq_scale','fd_mass','gb_mass');
tx_mass_scale_coef=[1 0 1 0]; % coefficients for mass scaling
relationship
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% REVISION HISTORY
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%