Upload
stanish-andromeda
View
893
Download
1
Embed Size (px)
Citation preview
48V HYBRIDIZATION OF A MID-SIZE VEHICLE
USING ELECTRIC MOTOR AND ELECTRIC
ASSISTED SUPERCHARGER
Powertrain Program
IFP School 2014/2015
Supervisor: S. Potteau
Authors: I.D.
Aggelos Zoufios E12792
Shixiong Zhao E12790
Stanish Gunasekaran E12593
Thomas Redlinger E12718
Date: 12th June 2015
I
ABSTRACT
In a context of growing demand for sustainable transportation worldwide, different
technical solutions for hybrid vehicles are nowadays investigated as effective ways to
improve efficiency of the driveline and thus to reduce CO2 emissions. As a matter of fact,
the CO2 emission targets set by EU (95 g/km in 2020 and 75 g/km in 2025) are extremely
demanding.
In order to reach the 2020 CO2 emission target with a spark ignition engine, solutions
needs to be developed or reinforced. The solution to be reinforced is downsizing and the
solution to be developed is hybridization. These two solutions would be able to lower CO2
emissions of spark ignition engines to at least 95 g/km.
By implementing mild hybridization (48V) in a vehicle equipped with a 1.2L downsized
turbocharged gasoline engine, it is possible to reduce fuel consumption and as a
consequence CO2 emissions comparing to the baseline diesel vehicle (Golf VII 1.6 TDI)
and meet the 2020 regulation. These gains in terms of CO2 can reach up to 10% (89 g/km).
At the same time, no compromise should be done between fuel consumption and fun-to-
drive. Turbocharged engines exhibit poor transient performance (“turbo lag”). In this
regard, some work has been done in incorporating air compressors driven by electric
machines in the air flow path of an engine. An electric supercharger allows improving fuel
consumption, turbo lag and increasing engine torque at low speed.
By implementing both an electric machine (15kW) and an electric supercharger (4kW),
the driving performance of the proposed hybrid has been maintained or even improved
reducing the turbo lag.
Finally although the price of a mild hybrid can be more expensive compared with a
conventional diesel, it has been proven that benefits can be gained back during the car
lifetime.
II
ACKNOWLEDGEMENT
We are using this opportunity to express our gratitude to everyone who supported us
through this project.
Foremost we would like to express our sincere gratitude to our supervisor, Mr Sebastien
Potteau from Valeo, for his continuous support and availability to answer any questions
we might have had. His trustful and insightful comments were appreciable all the time of
research and writing of this project.
Special thanks go out to Mr Prakash Dewagan from IFP School, for sharing his expertise
in Matlab Simulink ©.
Kind thanks also go out to Mrs Ouafae El Ganaoui-Mourlan and Mr Pascal Greau for giving
us the opportunity to work on this project.
III
TABLE OF CONTENTS
Abstract ................................................................................................................................................................ I
Acknowledgement.......................................................................................................................................... II
Introduction ....................................................................................................................................................... 1
1 Benchmarking ......................................................................................................................................... 3
1.1 Diesel .................................................................................................................................................. 3
1.2 Gasoline.............................................................................................................................................. 4
1.3 Hybrid ................................................................................................................................................. 5
1.4 Comparison and Benchmarking ............................................................................................... 7
1.5 Cycles .................................................................................................................................................. 8
1.6 Summary ........................................................................................................................................ 12
2 Engine Selection & Optimization .................................................................................................. 14
2.1 Downsizing and gear ratios .................................................................................................... 14
2.2 3-Cylinders Engine ..................................................................................................................... 17
3 Hybrid Architecture Selection........................................................................................................ 22
3.1 Degree Of Hybridization ........................................................................................................... 22
3.2 Hybrid Architecture ................................................................................................................... 23
3.2.1 Series Architecture ............................................................................................................ 23
3.2.2 Parallel Architecture ......................................................................................................... 23
3.2.3 Power Split Architecture ................................................................................................. 23
3.3 Component Sizing ....................................................................................................................... 24
3.3.1 Engine Selection .................................................................................................................. 24
3.3.2 Electrical System Selection ............................................................................................. 24
3.4 Energy Management Strategy ................................................................................................ 26
IV
3.4.1 Principle ................................................................................................................................. 26
3.4.2 Implementation .................................................................................................................. 27
3.5 Elaboration and Quantitative Rating of Possible Parallel Hybrid Topologies ..... 30
3.5.1 Conventional Architecture .............................................................................................. 30
3.5.2 Stop & Start Implementation ......................................................................................... 30
3.5.3 Mild-Hybrid Architectures .............................................................................................. 31
3.6 Final Architecture ....................................................................................................................... 35
4 Gear Selection Optimization ........................................................................................................... 37
4.1 Gear Shifting Strategy ................................................................................................................ 37
4.1.1 Principle ................................................................................................................................. 37
4.1.2 Implementation .................................................................................................................. 38
4.2 Simulations .................................................................................................................................... 38
4.2.1 NEDC Cycle ............................................................................................................................ 38
4.2.2 Real Driving Cycle In Paris .............................................................................................. 41
5 E-Supercharger Implementation .................................................................................................. 44
5.1 System Layout and Requirements ........................................................................................ 45
5.2 Simulation Methodology .......................................................................................................... 46
5.2.1 Engine Torque Calculation ............................................................................................. 46
5.2.2 Motor Torque ....................................................................................................................... 49
5.3 Performance Calculation Results .......................................................................................... 49
5.3.1 Time to Torque at Constant Engine Speeds ............................................................. 49
5.3.2 Overall Results on All Acceleration Ranges Considered ..................................... 51
5.3.3 Focus on 80-120km/h Acceleration ........................................................................... 52
6 Results ..................................................................................................................................................... 55
V
6.1 Fuel Consumption & CO2 Emissions ................................................................................... 55
6.1.1 NEDC Cycle ............................................................................................................................ 55
6.1.2 RDE Cycle............................................................................................................................... 55
6.2 System Overview ......................................................................................................................... 56
6.3 Additional Weight & Cost Estimation.................................................................................. 57
6.4 Cost to the OEM............................................................................................................................ 57
6.5 Total Cost Of Ownership Estimation ................................................................................... 58
Conclusion and Directions for further study ..................................................................................... 59
References ....................................................................................................................................................... 61
1
INTRODUCTION
Internal Combustion Engine has been the main machine for producing work and power
in the transport sector since Nicolaus Otto invented the first four stroke spark ignition
engine in 1876. Although, Otto’s engine was able in producing only a small amount of
power due to very small efficiency, huge steps have been achieved for increasing the
efficiency of an ICE (e.g. advanced materials, high pressure injection pumps, power
electronics …).
At the same time new European norms (Euro 5, Euro 6) have reinforced the limitations
for the control of the already regulated exhaust emissions coming from an ICE.
Considering an ICE, these CO2 emissions reduction can only be achieved by reducing as
much as possible the required amount of fuel to produce a certain power output, and thus
improving fuel economy.
Being the ICE inefficient enough at part loads, new techniques have been investigated in
order to eliminate the operating time of an ICE under these conditions. One solution is the
engine downsizing. That is to say forcing the engine with smaller displaced volume to
operate at higher loads and as a consequence under improved efficiency to produce the
same amount of power and torque. Another solution is the hybridization, nowadays a
promising solution for further improvement of fuel economy that contributes in the
further reduction of CO2 emissions.
The basic principle of hybridization is to implement an additional power source in the
vehicle driveline which will operate in behalf of the ICE when it has been proven to be
inefficient or in combination with the engine. The connection between the different power
sources of the vehicles can be done either in powertrain level (Parallel hybrids) or in
power level (Series hybrids). A combination between these two configurations can lead
to more complex hybrids. The main limitation as long as it concerns hybrid vehicles is the
energy source: the battery. Although battery technology has accomplished huge steps the
recent years, its incapability to store and deliver a big amount of energy is and will be
crucial for the future of the hybrid vehicles.
At the same time it is obvious that along with low fuel consumption main driver’s demand
is good performance in acceptable cost. In terms of performance, although the downsizing
2
of an engine with the use of turbocharging has been proven to be beneficial for fuel
economy, turbochargers exhibit poor transient performance especially at low engine
speeds. By implementing air compressors to the air loop, driven not by the exhaust gas
enthalpy of the turbine but by electric motors (electric superchargers), the low end torque
and as a consequence the feeling of the driver is really improved.
Regarding the economic aspect, hybrid vehicles are more expensive comparing to
conventional vehicles due to the usage of additional components (bigger battery, electric
motor, extra cables …). So the reduction of CO2 emissions with the usage of hybridization
should meet an acceptable cost threshold for both the final customer and the OEM.
The purpose of this study is to try to match or improve the fuel consumption and CO2
emissions emitted by a state of the art diesel vehicle (Golf VII 1.6L TDI) with a gasoline
hybrid vehicle without sacrificing the performance of the vehicle, and all these in an
acceptable economic package for the final customer.
3
1 BENCHMARKING
Having identified our target and our product goal, benchmarking is essential to define the
best in class performance matrices and hence define our functional requirements.
Benchmarking is the process of comparing one's business processes and performance
values to industry bests or best practices from other companies. Dimensions typically
measured are fuel consumption, performance and cost.
1.1 DIESEL
A set of various diesel cars in the C segment with comparable performance and price
ranges is compared for performance, consumption and emissions (Figure 1-1 and Figure
1-2). Considering the compromises on cost to company and the required performance
levels the top three vehicles will be shortlisted later.
The comparison provides a general overview over the range of engine power in the C
segment.
Figure 1-1 : Diesel Vehicle CO2 Emissions.
The diesel engines tend to be heavier than the gasoline counterparts while the CO2
emissions tend to be lower. However Diesel cars are generally associated with higher
soot/ smoke in the exhaust. This has led to emission norms emphasizing the installation
of DPF & after treatment devices which further increase the cost and weight of the vehicle.
1295
1395
1185
1385
99 99
84
104
80
90
100
110
120
130
140
150
800
900
1000
1100
1200
1300
1400
1500
Golf VII 1,6 TDI BMW 116dEfficient
DynamicsEdition
Peugeot 308Blue HDI 120
Alfa RomeoGiullieta 1,6JTDM-2 105
CO
2 e
mis
sio
n (
g/km
)
Cu
rb w
eig
ht
(kg)
Curb weight CO2 emission
4
Figure 1-2 : Diesel Vehicle Performances.
1.2 GASOLINE
A comparison of similar gasoline cars with respect to the Golf VII are selected for
benchmarking. Some of the performance indicators considered are: stand-still
acceleration (0-100km/h), roll-on acceleration (80-120km/h) and maximum speed
(Figure 1-4).
A common trend observed is that the average fuel consumption and CO2 emission levels
are higher than for the diesel engines (Figure 1-3). And observing the market trends
towards the gasoline vehicles it is a necessity to implement suitable strategies to reduce
these levels. These will be discussed later in this report.
The weight range is between 1050 kg (Ford Fiesta) to 1350 kg in the case of Peugeot 207.
All the engines are however front transversal and weight to power ratio lies between 55
– 60 kg/kW for the analyzed set of cars. Gasoline cars own a huge portion of the market
share outside European market and the European legislations show a proclivity to
gasoline cars with the progressing emission trends. Hence choosing a gasoline version of
the hybrid makes more sense with the legislation and technology trend.
189
195193
185
11
9,4
9,7
11,3
9
9,5
10
10,5
11
11,5
12
175
180
185
190
195
200
Golf VII 1,6 TDI BMW 116dEfficient
DynamicsEdition
Peugeot 308Blue HDI 120
Alfa RomeoGiullieta 1,6JTDM-2 105
0-1
00
(se
c )
Max
Sp
ee
d (
Km
/hr)
5
Figure 1-3 : Gasoline Vehicles CO2 Emissions.
Figure 1-4 : Gasoline Vehicles Performances.
1.3 HYBRID
The hybrid vehicles have smaller engines which are generally downsized and this
contributes to a lower emission level (Figure 1-5). The series and parallel hybrid
architectures are very popular among manufacturers and between both the parallel
architecture require smaller battery pack and is more suitable for mild hybrid application
1090
1200
1160
1205
110114
129
115
80
90
100
110
120
130
140
150
1000
1050
1100
1150
1200
1250
1300
Peugeot 308 II1.2 THP 130 PS
(2014)
Ford Focus III1.0 EcoBoost125 PS (2012)
Nissan Juke 1.2DIG-T 115 PS
(2014)
Hyundai i20 1.4100 PS (2015)
CO
2 e
mis
sio
ns
(g /
km)
Cu
rb w
eig
ht
(kg)
Curb weight CO2 emission
207
192
178
184
11
13,2
15,1
8,5
2
4
6
8
10
12
14
16
175
180
185
190
195
200
205
210
Peugeot 308 II1.2 THP 130 PS
(2014)
Ford Focus III 1.0EcoBoost 125 PS
(2012)
Nissan Juke 1.2DIG-T 115 PS
(2014)
Hyundai i20 1.4100 PS (2015)
80
-12
0 (
sec)
Max
Sp
ee
d
(Km
/hr)
800-120( sec) 80-120 acc
6
for a C segment vehicle. Parallel architecture allows thermal and electrical sources to
provide torque simultaneously and improve engine efficiency.
Figure 1-5 : Hybrid Vehicles CO2 Emissions.
Figure 1-6 : Hybrid Vehicles Performances.
The drawbacks however of the hybrid vehicle are as follows:
High Prices and Vehicle weight (battery pack)
Highway Consumption not improved
Difficulty when driving aggressively, lack of dynamism in acceleration, engine
reacts late on highway.
12461310
1370
1660
1295
101
84 8288
102
80
90
100
110
120
130
140
150
160
170
800
900
1000
1100
1200
1300
1400
1500
1600
1700
HondaInsight II
Toyota Auris136h
Lexus CT200h
Peugeot3008 Hybrid
Golf VII 1,6TDI
CO
2 e
mis
sio
n (
g/k
m)
Cu
rb w
eig
ht
(kg)
Curb weight CO2 emissions
182180 180
185
192
11,1
8,2 8,1
6,6
9,6
0
2
4
6
8
10
12
170
175
180
185
190
195
200
205
210
HondaInsight II
Toyota Auris136h
Lexus CT200h
Peugeot3008 Hybrid
Golf VII 1,6TDI
80
-12
0 a
cc (
sec)
Max
Sp
ee
d
(Km
/hr)
Max Speed 80 - 120 Km/hr
7
1.4 COMPARISON AND BENCHMARKING
Here are few noteworthy points to note:
• Gasoline cars have higher CO2 emissions hence our choice to hybridize a gasoline
vehicle as a market replacement of the Golf VII is interesting.
• The important customer expectations from the C segment car are observed to be
pick up acceleration, roll on acceleration, max speed and vehicle size.
• The conclusion from the benchmarking study is shown in Figure 1-7.
Figure 1-7 : Fuel Consumption (top) and CO2 Emissions (bottom) Summary.
3,9 3,8
4,8
3,23,4
5
44,4
5,4
0
1
2
3
4
5
6
Diesel Hybrid Gasoline
Fue
l co
nsu
mp
tio
n (
L/1
00
km)
Overall Fuel Consumption Comparison
99
84
110
84 88
114104 101
124
0
20
40
60
80
100
120
140
Diesel Hybrid Gasoline
CO2 Comparison
8
1.5 CYCLES
An emission test cycle is a protocol contained in an emission standard to allow consistent
comparable measurement of emissions for different engines. Test cycles specify the
conditions under which the engine is operating during the emission test.
Light duty vehicles and motorcycles are tested on a chassis dynamometer test bench using
standardised vehicle speed cycles to test for compliance with the exhaust emissions and
CO2 emissions. There are various legislations adopted throughout the world depending
on the driving conditions, political and technological scenario.
The various cycles today include:
FTP – Commonly known as the FTP 75 defined by the US EPA is a set of regulations to
measure the tailpipe emission and consumption of cars. The characteristics of the
cycle are:
Distance travelled: 17.77 km (11.04 miles)
Duration: 1874 seconds
Average speed: 34.1 km/h (21.2 mph)
Figure 1-8 : FTP Cycle.
JC08 – Japanese driving conditions are represented through JC08 chassis
dynamometer test cycle for light vehicles (< 3500 kg). The test represents driving in
congested city traffic, including idling periods and frequently alternating acceleration
and deceleration. Measurement is made twice, with a cold start and with a warm start.
The test is used for emission measurement and fuel economy determination, for
gasoline and diesel vehicles.
9
Figure 1-9 : JC08 Cycle.
NEDC – New European Driving cycle is supposed to represent the real driving
conditions in Europe (1997). It consists of four cycles of ECE -15 (urban) mode and
one EUDC (Extra urban or highway mode). The cycle is tested at 20-30°C conditions
on a roller test bench to represent flat road, and zero wind conditions. The cycle is
however criticized for not representing real motorway conditions and not necessarily
repeatable and comparable leading to cycle beating.
Figure 1-10 : NEDC Cycle.
WLTP - The Worldwide harmonized Light vehicles Test Procedures (WLTP) defines a
global harmonized standard for determining the level of pollutants and CO2
emissions, fuel consumption, and electric range from light-duty vehicles around the
globe including European Union, Japan, India and China under the UNECE vehicle
regulations. The cycle imposes strict constraints regarding road load (motion
resistance), gear shifting, total car weight (by including optional equipment, cargo and
10
passengers), fuel quality, ambient temperature, and tire selection and pressure. The
cycle is applied based on vehicle class which is derived from the car power to weight
ratio.
Figure 1-11 : WLTC Cycle.
ARTEMIS - The Common Artemis Driving Cycles are chassis dynamometer procedures
developed within the European Artemis project, based on statistical analysis of a large
database of European real world driving patterns. The cycles include three driving
schedules: Urban, Rural road and Motorway. The Motorway cycle has two variants
with maximum speeds of 130 and 150 km/h. Artemis cycle definitions also include
gear shifting strategies.
Figure 1-12 : Artemis Cycle.
The pros and cons of all the major cycles are enlisted in Table 1-1. It can be observed that
each cycle has its unique characteristic, very specific to the driving needs of the area it
represents. The choice of cycle is very important to fix realistic target model, in a given
driving condition. NEDC cycle though a common European benchmark is very steady state
11
and does not represent real driving conditions as shown later in this report. However it
must be taken in the stride of setting a common emission limit and also since it provides
a scale of comparison with the Volkswagen Golf VII Diesel engine which has emission data
based on the NEDC.
Pros Cons Remarks
NEDC Common European
benchmark
Not realistic Constant acceleration,
deceleration, speed
Cycle beating
No auxiliary power
consumption
WLTP Gear shifting, Weight
accounted May vary from the local conditions
DATA from 5 different countries
and 3 different segment cars Optimal shift points
Artemis Includes gear shifting Not used for certification
of pollutants or consumption
Statistical model
JC08 Very transient Does not represent
normal driving Congested city
traffic
Aggressive driving and air-conditioning included with US06
high speed and motorway
Still not completely representative of European driving
FTA
Table 1-1 : Comparison of Cycle Benefits.
This report shows that the Real driving emissions are far different from the Cycle chosen.
The upcoming EU6c Emission Regulation will implement RDE as an additional
requirement in the 2017 - 2020 timeframe. Compared to the current test environments,
which are designed and optimized for perfect reproducibility and a removal of external
influences, driving a vehicle on the road under "real-life" conditions will never be 100%
reproducible. The influence of the road profile, the ambient conditions, the traffic
situation and the behaviour of the driver itself will significantly influence the results. One
to one comparison of test results will not be possible, but it is crucial from the point of
view of real world global limits for pollutants. A summary of improvement in performance
due to hybridisation in the RDE cycle is also included in the report.
12
1.6 Summary
Figure 1-13 : Trend showing reduced fuel consumption.
The customers represent a huge diversity in the type of driving profile and the conditions
they drive in. Sustained efforts need to be taken to reduce the overall fuel consumption
whilst catering to customer demands (keeping fun to drive features, etc). This balance is
determined by developing a cycle that is robust and realistic. The different demands and
driving conditions translate to different energy management strategies and powertrain
utilisation as shown below. The overall trend has been to reduce the fuel consumption
and work is underway for OEMS and Suppliers.
182 g/km
172 g/km
Figure 1-14 : Different Utilization of Engine Energy Management in Different Cycles.
13
Figure 1-15 : Cycles Summary.
14
2 ENGINE SELECTION & OPTIMIZATION
2.1 DOWNSIZING AND GEAR RATIOS
The engine selected for this study is a 1.6L turbocharged GDI engine. In order to reduce
the fuel consumption, the engine was further downsized to 1.2L. In order to downsize the
engine from 1.6L to 1.2L the specific torque was used:
𝑇 = 𝑏𝑚𝑒𝑝 ∗ 𝑉𝑑
4 ∗ 𝜋 (Equation 2-1)
By checking (Equation 2-1 it is easy to understand that in order to obtain the same torque
from the 1.2L engine comparing to 1.6L, it is necessary for the engine to operate at higher
bmep (higher load) and as a consequence in areas with higher efficiency and lower brake
specific fuel consumption (bsfc). That is the main principle of downsizing.
The full load curve of the engine in terms of specific torque and specific power can be seen
in Figure 2-1.
The maximum torque and power delivered by the engine are respectively 205 Nm@1750
rpm and 80.7 kW@5500 rpm.
Figure 2-1 : Full Load Curve of the Engine Using Specific Torque and Specific Power.
In order to have the desired performance of the vehicle in terms of maximum vehicle
comparison with the reference vehicle of our study (Golf VII 1.6L TDI), the gear ratios
0,0
10,0
20,0
30,0
40,0
50,0
60,0
70,0
80,0
0
20
40
60
80
100
120
140
160
180
0 1000 2000 3000 4000 5000 6000 7000
Po
wer
[K
W]
Torq
ue
[Nm
]
Engine Speed [RPM]
Specific Torque (Nm/l) Specific Power (kW/l)
15
have to be selected. There are two main resistive forces applied to a vehicle during its
movement.
Rolling Resistance Force
𝐹𝑡𝑦𝑟𝑒 [𝑁] =𝑀 [𝑘𝑔] ∗ 𝑔 ∗ 𝐶𝑟𝑟 [
𝑘𝑔𝑡 ]
1000 (Equation 2-2)
Aerodynamic Force
𝐹𝑎𝑒𝑟𝑜 [𝑁] =1
2∗ 𝜌 [
𝐾𝑔
𝑚3] ∗ 𝑆 [𝑚2] ∗ 𝐶𝑥 ∗ 𝑉2 [
𝑚2
𝑠2] (Equation 2-3)
In our study the third resisting force due to the slope of the road is considered equal to
zero. So the sum of these two forces will give us the total resisting force (𝐹𝑟𝑒𝑠𝑖𝑠𝑡 [𝑁] =
𝐹𝑡𝑦𝑟𝑒 [𝑁] + 𝐹𝑎𝑒𝑟𝑜 [𝑁]). The resistive power the vehicle has to overcome is given by the
following equation:
𝑃𝑟𝑒𝑠𝑖𝑠𝑡 [𝑊] = 𝐹𝑟𝑒𝑠𝑖𝑠𝑡 [𝑁] ∗ 𝑉 [𝑚
𝑠] (Equation 2-4)
When the power that is produced by the engine multiplied by the transmission efficiency
is equal to the maximum resisting power applied to the vehicle the vehicle has obtained
its maximum speed. Then the last gear ratio can be deduced by looking at the ratio
between the highest possible vehicle speed and the maximum power engine speed. The
final gear ratio is computed using the following formula:
𝐿𝑖 [
𝐾𝑚ℎ
1000𝑟𝑝𝑚] = 𝑉1000 =
𝑉𝑒ℎ𝑖𝑐𝑙𝑒 𝑆𝑝𝑒𝑒𝑑 [𝐾𝑚/ℎ]
𝑁 [𝑟𝑝𝑚]/1000 (Equation 2-5)
This last gear ratio is optimal regarding maximum vehicle speed.
Three types of gearbox were tested. The optimal one, a short one (-10% of the optimal
one) and a long one (+10% of the optimal one). The maximum vehicle speed with these
three types of gearbox is shown in Figure 2-2. The maximum vehicle speed with the
optimal, short and long gearbox are respectively 190, 171 and 188 km/h. These three
types of gearbox were also tested in order to obtain the roll-on acceleration from 80-120
km/h. The results can be seen in Figure 2-3. The time required with the longest gearbox
16
to reach 120 km/h is 12 s. This performance is close to the targeted one (11.6s) and since
a long gearbox is better considering fuel consumption, this gearbox with a final gear ratio
of 38 has been selected.
Figure 2-2 : Maximum Vehicle Speed for the three Gearboxes (Optimal, Short and Long).
Figure 2-3 : Roll-on Acceleration from 80-120 km/h for the three Gearboxes (Optimal, Short and Long).
0
10
20
30
40
50
60
70
80
0 50 100 150 200 250 300
Po
wer
(kW
)
Vehicle Speed (km/h)
Maximum Vehicle SpeedPresist (kW) optimal gearbox Short Gearbox Long Gearbox
80
85
90
95
100
105
110
115
120
125
0 2 4 6 8 10 12 14
Veh
ivle
Sp
eed
(km
/h)
Time (s)
Vehicle Speed = f(time)
17
After the selection of the final gear ratio, it is important to obtain the gear ratios associated
to the other gears. The deduced gear ratios are shown in Table 2-1.
Gear Number Gear ratio
1st 5.24
2nd 2.89
3rd 1.89
4th 1.33
5th 0.97
Table 2-1 : Gear Ratios.
It has to be noticed that the first gear ratio has been set to reach a take of performance of
4.4 m/s2. Then the intermediate ratios have been selected using a combined steeping
(average between geometric and arithmetic stepping).
2.2 3-CYLINDERS ENGINE
Heat transfer affects engine performance, efficiency and emissions. For a given mass of
fuel within the cylinder, an increase in heat transfer to the combustion chamber walls will
lower the average combustion gas temperature and pressure. This will lead in a work per
cycle transferred to the piston reduction. Thus specific power and efficiency are affected
by the magnitude of the engine heat transfer.
Heat transfer between the unburned charge and the chamber walls in spark–ignition
engines affects the onset of knock which, by limiting the compression ratio, also influences
power and efficiency. Moreover changes in gas temperature due to the heat transfer
impact the emission formation processes, both within the engine’s cylinder and in the
exhaust system where afterburning of CO and HC occurs. The exhaust temperature also
governs the power that can be obtained from exhaust energy recovery devices such as a
turbocharger turbine.
Friction is also affected by engine heat transfer and contributes to the coolant load. The
cylinder liner temperature governs the piston and ring lubricating oil film temperature,
18
and hence its viscosity. Some of the mechanical energy dissipated due to friction must be
rejected to the atmosphere by the cooling system. The fan and water pump power
requirements are determined by the magnitude of the heat rejected. Thus the importance
of engine heat transfer is clear. The energy balance in an engine can be seen in Figure 2-4.
Figure 2-4 : Energy Balance in an Internal Combustion Engine.
Generally the fuel energy released during combustion in an ICE is converted to the
following forms with the following distribution:
Mechanical work in the crankshaft ~ 30%
Heat losses in the exhaust ~ 30%
Coolant losses in the coolant ~ 30%
Miscellaneous losses (radiation, free convection, oil if separately cooled) ~10%
Nf = Ne + Ncool + Nexh + Nmisc (Equation 2-6)
The heat flux into the wall has in general both a convective and a radiative component.
The heat flux is conducted through the wall and then convected from the wall to the
coolant. Heat is transferred by forced convection between the in-cylinder gases and the
cylinder head, valves, cylinder walls, and piston during induction, compression, expansion
and exhaust processes. Heat is transferred by forced convection from the cylinder walls
and head to the coolant, and from the piston to the lubricant or other piston coolant.
Substantial convective heat transfer occurs to the exhaust valve, exhaust port, and exhaust
Air
Fuel
Exhaust losses (Nexh)
Coolant losses (Ncool) Different losses
(Nmisc)
ICE
Mechanical Work (Ne)
19
manifold during the exhaust process. However the biggest amount of heat transfer takes
place between the burned gases and the cylinder wall during the phase of combustion and
expansion [1]. The amount of heat losses to the liner is given by the (Equation 2-7.
�̇� = ℎ ∗ 𝐴 ∗ (𝑇𝑔 − 𝑇𝑤𝑎𝑙𝑙) (Equation 2-7)
In the (Equation 2-7, h is the heat transfer coefficient, A the surface of the liner being in
contact with the coolant, Tg the gas temperature and Twall the temperature of the liner.
It can be seen that the amount of heat rate is strongly dependent on the available area of
the cylinder liner. By reducing this area a benefit considering the heat losses could be
obtained.
By reducing the number of cylinders from 4 to 3 it is possible to reduce the available area
for heat transfer and as a consequence the amount of heat losses in the coolant. The basic
assumptions made for going from 4 to 3 cylinders are:
Exhaust losses remain constant at 30% of the total fuel energy
Stroke to Bore ratio (S/B) for conventional gasoline engines is 0.8
Constant mean wall temperature
Bmep unchanged
According to these assumptions any possible gain in terms of reduction of the available
heat transfer area and as a consequence a reduction of the heat losses will be in the benefit
of the brake specific fuel consumption of the engine. Indeed for the engine to maintain the
same load (bmep), with reduced heat losses the amount of air and fuel introduced in the
cylinder is lower.
The displaced volume of the engine is 1.2L. Displaced volume can be calculated from the
(Equation 2-8.
𝑉𝑑 =𝑝𝑖 ∗ 𝐵2
4∗ 𝑆 (Equation 2-8)
Where B is the piston diameter (Bore) and S the stroke of the engine.
20
Using the previous assumption and solving the (Equation 2-8 for the bore it is possible to
obtain a value for the piston diameter and then the stroke. The available area for heat
losses is given by the following equation:
𝐴 = 𝜋 ∗ 𝐵 ∗ 𝑆 (Equation 2-9)
The results and the benefit of the reduction in heat transfer area are below shown in Table
2-2.
4 cylinders 3 cylinders Percentage of area
reduction
S
[cm]
B
[cm]
Vd
[cm3]
A
[cm2]
S
[cm]
B
[cm]
Vd
[cm3]
A
[cm2] [%]
6.25 7.81 1200 614.3 6.88 8.6 1200 558 9.16
Table 2-2 : Difference in Heat Transfer Area Obtained Between 4 and 3 Cylinders.
According to [1], although the heat losses are such a substantial part of the fuel energy
input, elimination of heat losses would only allow a fraction of the heat transferred to the
combustion chamber walls to be converted to useful work. The remainder would leave
the engine as sensible exhaust enthalpy. Considering an automotive high – speed naturally
aspirated CI engine with a compression ratio of 15. The indicated efficiency is 45%, and
25 percent of the fuel energy is carried away by the cooling water. Of this 25%, about 2%
is due to the friction. Of the remaining 23%, about 8% is heat loss during combustion, 6%
heat loss during expansion, and 9% heat loss during exhaust. From the 8% lost during
combustion about half (4% of the fuel energy) could be converted into useful work on the
piston. From the 6% heat loss during expansion, about one – third (2%) could have been
utilized. Thus of the 25% lost to the cooling system, only about 6% could have been
converted to useful work on the piston, which would increase the indicated efficiency of
the engine from 45 to 51%. (Figure 2-5)
21
Figure 2-5 : Energy Flow Diagram for an Internal Combustion Engine. [1]
For a spark ignition engine, the conversion to useful work will be lower, because the
compression ratio is lower. However, the heat losses at part load (which constitute an
important operating regime for automobile use) are a substantially larger fraction of the
fuel heating value. Studies with computer simulations of the SI engine operating cycle
indicate that at typical part–load conditions a proportional reduction in combustion
chamber wall heat losses of 10% results in a proportional increase (improvement) in
brake specific fuel conversion of about 3 percent [1].
By subtracting the previously calculated 9.16% reduction in the available heat transfer
area a 2.75% improvement in bsfc is obtained.
22
3 HYBRID ARCHITECTURE SELECTION
3.1 DEGREE OF HYBRIDIZATION
Hybridization represents undoubtedly a valuable option to improve fuel savings thanks
to the following efficiency enhancements [2]:
Deleting idling losses by switching off the engine when vehicle stops,
implementing a Stop & Start strategy.
Using regenerative braking in order to recover part of the vehicle kinetic energy
during deceleration instead of dissipating it through braking system.
Operating the internal combustion engine closer to its best efficiency, trying
to avoid its use under highly inefficient operating conditions, thanks to the
additional degree of freedom provided by the electrical power source and energy
storage devices.
Enabling internal combustion engine downsizing while still maintaining
acceptable vehicle performance thanks to the additional boosting which can be
provided by the electric power source.
There are different degree of hybridization: Micro, Mild, Full and Plug-in hybrids.
The Plug-in hybrids are out of the scope of this study. The Micro-Hybrid cars are only able
to save fuel consumption through stop and start implementation (~5% fuel savings). The
Full-Hybrid vehicles such as the Toyota Prius utilize expensive high capacity motor-
generator integrated alongside the drive train and high capacity battery packs (more than
48V – high voltage lead to safety measures) [3]. Mild-Hybrid systems have the benefit of
incorporating fewer changes in the system architecture in comparison to Full-Hybrid
architecture and consequently being more ready for conversion from a conventional to a
hybrid powertrain. In addition it offers the following advantages over strong hybrids [3]:
A lower cost to benefit ratio
Higher power to weight ratio
Easy mechanical integration on production vehicles.
Lower risk factor owing to lower operating voltage.
No requirement for a complex energy management control strategy.
23
Thus it appears clearly that the Mild-Hybrid topology is the most suitable for this study.
3.2 HYBRID ARCHITECTURE
It exists, three different ways to arrange a hybrid architecture: series, parallel and power
split architecture.
3.2.1 SERIES ARCHITECTURE
In this kind of architectures, all of the energy from the engine goes through the electric
system before reaching the wheels. This allows the engine to operate on maximum
efficiency point but requires high energy storage capacity (large batteries and therefore
high cost). Also this architecture induces additional electric losses through the whole
electric system. [5]
3.2.2 PARALLEL ARCHITECTURE
In this architecture the propulsion at the wheel can be ensured by either the engine or the
electric system or both. The required size of the electric system is thus smaller reducing
the cost.
Moreover unlike the series hybrid, the energy from the engine does not have to go through
the electric system lowering losses when the engine is operating efficiently. But control of
engine operating points is more difficult than in a series architecture. This architecture
has been applied in the Honda hybrids. [5]
3.2.3 POWER SPLIT ARCHITECTURE
This architecture combines features of the two previous ones. One part of the engine’s
power is connected to the wheels through a series branch (power going through electric
system), and a second part through a parallel branch (engine power going directly to the
wheels). The series branch allows controlling the engine more efficiently. The size of the
batteries and electric losses are, however, kept relatively small compared with a series
architecture thanks to the parallel branch. The most known application of the power split
architecture are the Toyota hybrids. [5]
In comparison to parallel hybrids the battery size for a power split hybrid is larger at iso-
total peak power due to electrical losses, leading to an increase in price.
24
As the parallel hybrid require fewer changes when converting from a conventional to a
hybrid powertrain and lower cost thanks to smaller battery, the parallel architecture has
been chosen for this study.
3.3 COMPONENT SIZING
3.3.1 ENGINE SELECTION
The vehicle propulsion being assisted by an electric motor it is then possible to size the
engine smaller than normally required. In this way the engine will be loaded closer to its
maximum efficiency points (close to maximum torque curve). The downsizing will lead to
a fuel consumption improvement thanks to better efficiency without compromising the
fun to drive since the E-Machine will assist the propulsion during acceleration.
The E-Machine used in a Mild-Hybrid system is typically more powerful than a
conventional 12V starter and can quickly start the engine without the driver experiencing
an uncomfortable delay. The system efficiency can be further improved by charging the
batteries when the engine is operating at low load condition. [3]
3.3.2 ELECTRICAL SYSTEM SELECTION
The electric motor power capacity ranges from 1.5kW to 15kW while operating voltage
of a mild hybrid systems vary from 12volts to 64volts. [3]
The characteristics of the E-Machine used to quantitatively rate the possible hybrid
architecture are summarized in Table 3-1.
Rated Power (kW) 15
Max Torque (Nm) 70
Max Speed (RPM) 18 000
Efficiency 0.9
Table 3-1 : E-Machine Characteristics.
To size the electric motor a sensitivity study is conducted by using the machine map
(Figure 3-1) and adapting the maximum torque. Although the efficiency is engine speed
dependent it is assumed constant and equal to 90%.
25
Figure 3-1 : E-Machine Efficiency Map. [2]
Regarding the battery, the voltage is one of the constraint of the project: it is fixed to 48V.
The batteries are sized for power. More specifically, the output power of the batteries is
the peak power of the motor divided by the average motor efficiency. Thus In this case the
battery pack power should be at least 16.7kW. A combination of 17 Ultra-High Power Cell
(characteristics summarized in Table 3-2) in series is used to quantitatively rate the
possible hybrid architecture providing 18.9kW.
Nominal Capacity (Ah) 15
Mass (kg) 0.440
Nominal Voltage (V) 2.7
Maximum Power (W) 1110
Energy (Wh) 55.5
Charge-Discharge Efficiency (%) 100
Table 3-2 : Ultra-High Power Cell Characteristics. [4]
A pack of high power cell battery has been chosen since in this study the aim is to have
high power available during short periods (e.g. use of electric supercharger, engine
cranking by E-Machine).
The previously optimized engine and the previously described E-Machine give a
hybridization ratio of:
𝑅ℎ 𝑝𝑎𝑟𝑎𝑙𝑙𝑒𝑙 =𝑃𝐼𝐶𝐸
𝑃𝐼𝐶𝐸 + 𝑃𝑒𝑙𝑡= 84% (Equation 3-1)
26
3.4 ENERGY MANAGEMENT STRATEGY
3.4.1 PRINCIPLE
The degree of freedom of a parallel HEV is the engine torque (Te), this architecture
combining mechanical power through a mechanical node (Figure 3-2). The energy
management strategy has thus to select the engine torque that will minimize the fuel
consumption, this implies using the battery as a buffer. As shown on Figure 3-2, the
gearbox ratio can also be set as a degree of freedom, this will be investigated later in the
gear optimization section.
Figure 3-2 : Parallel HEVs Degree of Freedom. [6]
Let’s introduce now the Hamiltonian function (H) and the equivalence factor (S0):
𝐻 = 𝑃𝑓 + 𝑆0 × 𝑃𝑒𝑐ℎ (Equation 3-2)
Where:
𝑃𝑓 = 𝐻𝐿𝐻𝑉 × 𝑚𝑓̇ is the fuel power,
𝑃𝑒𝑐ℎ = −𝑆𝑂𝐶̇ × 𝑈𝑜𝑐 × 𝑄𝑏 is the electrochemical power,
𝑈𝑜𝑐 is the open-circuit voltage of the battery,
𝑄𝑏 is the battery capacity,
S0 is a non-dimensional parameter. [6]
This function allows optimizing the overall fuel consumption by introducing an equivalent
consumption for the power taken from the battery. It estimates how much fuel
correspond to the battery consumption. As at any time the Hamiltonian function depend
27
on the degree of freedom (engine torque) the optimal engine torque regarding fuel
consumption is the one that minimizes H.
This strategy generally allows engine operating points either to be suppressed
(powertrain goes full electric for cranking or to avoid inefficient low load points) or
moved to better efficiency points (recharge mode when battery is depleting, boost mode)
(Figure 3-3).
Figure 3-3 : Energy Management Strategy. [5]
NB: The function to crank the engine via the E-Machine will only be implemented on the
optimized energy management strategy (see section 4).
3.4.2 IMPLEMENTATION
The energy management strategy is implemented via Matlab/Simulink ©. As previously
described the aim of this submodel is to compute the engine and motor torque to generate
the required torque at the wheels while minimizing fuel consumption. It consists in
defining an array of possible engine torque candidates and select the one that minimizes
the Hamilton function seen in the previous section while respecting the following
constraints:
Engine speed ⋲ [idle speed, max speed]
Engine torque ⋲ [friction torque, max torque = f(RPM)]
Motor speed ⋲ [min speed, max speed]
Motor torque ⋲ [max regenerative torque, max motoring torque]
The architecture chosen cannot be recharged from the grid (not plug-in), it will
therefore operate in a charge sustaining mode: 𝑆𝑂𝐶𝑓𝑖𝑛𝑎𝑙 = 𝑆𝑂𝐶𝑖𝑛𝑖𝑡𝑖𝑎𝑙
Battery SOC ⋲ [min SOC, max SOC]
28
The input of this submodel are the vehicle speed, gear index, powertrain required torque
and the battery state of charge. (Figure 3-4)
Figure 3-4 : Energy Management Submodel.
The gear index and the vehicle speed are fixed by the cycle and the torque demand is
estimated as follow:
𝑇𝑃𝑊𝑇 = (𝐹𝑎𝑒𝑟𝑜 + 𝐹𝑡𝑦𝑟𝑒 + 𝐹𝑠𝑙𝑜𝑝𝑒) × 𝑅𝑡𝑦𝑟𝑒 +𝐽 × 𝑎𝑣𝑒ℎ
𝑅𝑡𝑦𝑟𝑒 (Equation 3-3)
Where :
𝐹𝑎𝑒𝑟𝑜 =1
2𝜌𝑆𝐶𝑥 × 𝑉2 is the aerodynamic force,
𝐹𝑡𝑦𝑟𝑒 = 𝑀 ∗ 𝑔 ∗𝐶𝑟𝑟
1000 is the rolling resistance force,
𝐹𝑠𝑙𝑜𝑝𝑒 = 𝑀 ∗ 𝑔 ∗ sin (𝛼) is the slope resistance force.
SCx (m²) 0.66
(kg/m3) 1.225
M (kg) 1295
Crr (kg/t) 8.5
Table 3-3 : Vehicle Characteristics.
The main outputs of the submodel are the engine and motor torque that minimize the fuel
consumption. The sum of these torque should equals the powertrain required torque at
the mechanical node that link the engine and the E-Machine. (Figure 3-2)
The array of possible candidates for the engine torque is composed of 10 values equally
spaced between the min and the max possible engine torques at the current engine speed
plus one value corresponding to a purely ICE mode corresponding to the actual gear ratio
29
and a 0 to represent the full electric mode. Thus there are 12 possible engine torques.
Knowing the required torque at the wheel (input) it is then possible to deduce the 12
corresponding value of the motor torque taking into account the gear ratio of the electric
pathway.
Figure 3-5 : Engine (left) and Motor (right) Torque Candidates on NEDC Cycle as a Function of Time.
From the previous torques calculations it is then possible to estimate the Hamiltonian
function by evaluating the fuel power and electrochemical power. To prevent any issue
with the constraints previously set unfeasibility flags have been implemented to identify
which of the solution candidates are not admissible. Finally all that remains is to find the
candidate that minimizes the Hamiltonian and that has not raised any unfeasible flag.
The model used for the E-Machine, thermal engine and battery were provided during a
supervised classwork on hybrid management strategy. [6]
In order to satisfy the charge sustaining mode the equivalent factor S0 has to be tuned.
The optimal equivalence factor is chosen considering the following algorithm:
Figure 3-6 : Flowchart of the offline Energy Management Strategy. [6]
0 200 400 600 800 1000
0
50
100
150
200
Time [s]
Ten
g [N
m]
Engine Torque candidates
0 200 400 600 800 1000-80
-60
-40
-20
0
20
40
60
80
Time [s]
Tm
ot [
Nm
]
Motor Torque candidates
30
3.5 ELABORATION AND QUANTITATIVE RATING OF POSSIBLE PARALLEL
HYBRID TOPOLOGIES
In order to evaluate the potential benefit of such architectures it is necessary to assess the
fuel consumption of the conventional gasoline powertrain. Comparison between different
architectures are based on NEDC cycle.
3.5.1 CONVENTIONAL ARCHITECTURE
For purely ICE powertrain there is no energy management strategy. Indeed the engine
torque has to provide the torque required at the wheels; there is no degree of freedom.
The thermal engine chosen for this study is a 3-cylinder 1.2L gasoline engine with direct
injection. It is assumed that 10% of crankshaft output power is lost due to friction and
auxiliaries. The transmission is a 5-gear manual transmission without any mechanical
loss.
Figure 3-7 : Engine Operating Points on NEDC (Conventional Engine).
The fuel consumption on NEDC cycle for this particular engine is 5.38 L/100km.
Considering a gCO2/km ⇄ L/100km conversion coefficient of 23.81 [7] this engine emits
128 gCO2/km, this is far above the target emission of the diesel Golf VII (99 gCO2/km).
3.5.2 STOP & START IMPLEMENTATION
This is an ideal Stop & Start which only considers that fuel consumption is 0 when vehicle
is at idle. Thus this architecture has the same engine operating point as the conventional
1000 2000 3000 4000 5000 6000
0
50
100
150
200
w [RPM]
T [
Nm
]
Engine Operating Points
31
one (Figure 3-7). The consumption on NEDC cycle is 5.09 L/100km, thus Stop & Start
implementation can achieve a fuel benefit of 5.4% and emits only 121 gCO2/km. But as
expected this reduction of CO2 emissions is not enough. It is mandatory to implement a
Mild-Hybrid architecture.
3.5.3 MILD-HYBRID ARCHITECTURES
In this section is investigated the potential benefits of the different parallel hybrid
architectures (P1, P2, P3 and P4, described in Figure 3-8)
Figure 3-8 : Different Parallel Hybrid Architecture. [8]
The energy strategy management previously described and the battery pack and electric
machine previously sized will be used to quantitatively rate the fuel savings and emission
reduction. To simulate the 4 different architectures the global architecture displayed in
Figure 3-9 is employed.
Figure 3-9 : Global Hybrid Architecture. (Similar than HOT Software)
The gear ratios Rg, Rn, Rc, Rm and Rd are used to simulate clutches and transmission
ratios. Rg is always set to 0 since it is only useful to simulate a series hybrid architecture.
Rn allows simulating a clutch between the engine and the electric machine which is typical
of a P2 architecture. Rc represents the ratio of the manual transmission and Rm the ratio
between the e-machine shaft and the manual transmission output shaft (typical of P3
architectures with double shaft). Finally Rd is the final drive ratio.
32
For all the architecture tested it is assumed that:
200W electric power is dedicated to ECU and auxiliaries.
All the braking torque can be recovered by the electric machine.
Engine resistive torque (when run by e-Machine) is 5Nm.
No use of electric machine power for cranking.
The gear shifting is imposed on NEDC cycle.
3.5.3.1 P1 ARCHITECTURE
The first architecture to be tested is the P1 architecture. (Figure 3-10)
Figure 3-10 : P1 Hybrid Architecture.
As shown on Figure 3-10 the Motor (M) is directly linked to the ICE, so it is not possible
to recover energy when the driver is pressing the clutch pedal. In this way the E-Machine
can only run full electric and regenerate energy during braking when the clutch is
engaged, so the E-Machine will undergo the engine resistive power due to friction. But as
no additional transmission is required the extra-weight implied by the hybridization is
limited. Although this configuration has obvious drawbacks the fuel consumption is
improved since the bad efficiency points at low load are removed (see green box on Figure
3-11) compared to the conventional gasoline engine (Figure 3-7).
Figure 3-11 : Engine and Electric Machine Operating Points on NEDC (P1 Architecture).
1000 2000 3000 4000 5000 6000
0
50
100
150
200
w [RPM]
T [
Nm
]
Engine Operating Points
0 2000 4000 6000 8000 10000-80
-60
-40
-20
0
20
40
60
80
w [RPM]
T [
Nm
]
Electric Machine Operating Points
33
The fuel consumption of this architecture on NEDC cycle is 3.99 L/100km, which
represents a fuel saving of 25.8% compared to the conventional gasoline engine. The CO2
emissions are thus lowered to 96 gCO2/km. This challenges the Golf VII diesel, but is not
enough to reach the 2020 target.
3.5.3.2 P2 ARCHITECTURE
The P2 architecture is very similar to the P1 architecture excepted that an ICE side clutch
is added to make possible to physically decouple ICE and e-Machine. (Figure 3-12)
Figure 3-12 : P2 Hybrid Architecture.
Thus unlike the P1 architecture it is possible to run full electric or regenerate energy
during braking by disengaging the engine via the ICE side clutch. This architecture is
therefore more efficient than the previous one. Indeed as shown on Figure 3-13 the e-
machine is more used and even point at mid load are upshift to operate at better efficiency
(green box).
Figure 3-13 : Engine and Electric Machine Operating Points on NEDC (P2 Architecture).
The fuel consumption of this architecture on NEDC cycle is 3.52 L/100km, which
represents a fuel saving of 34.6% compared to the conventional gasoline engine. The CO2
1000 2000 3000 4000 5000 6000
0
50
100
150
200
w [RPM]
T [
Nm
]
Engine Operating Points
0 2000 4000 6000 8000 10000-80
-60
-40
-20
0
20
40
60
80
w [RPM]
T [
Nm
]
Electric Machine Operating Points
ICE Side Clutch
34
emissions are thus lowered to 86 gCO2/km. This architecture allows reaching the 2020
target. But the additional clutch required increases the space needed to implement this
topology. As the Golf VII has a transverse engine it is very difficult to make the side clutch
fit in the vehicle overhang, thus although this architecture presents high fuel benefits it is
not the most suitable for this project.
3.5.3.3 P3 & P4 ARCHITECTURE
The last architectures to be investigated are the P3 and P4 (Figure 3-14). These 2
topologies are studied together since they present the same pros and cons.
Figure 3-14 : P3 (left) and P4 (right) Hybrid Architecture.
As the motor is located after the manual gearbox, it is possible to perform regenerative
braking with engine totally decoupled. It is possible to set the E-Machine directly on the
transmission output shaft to save the additional weight induced by a double shaft, but this
will limit the torque at the wheels provided by the motor. But although a double shaft will
take more space and induce more weight it can fit with a transverse engine and can
enhance the motor torque. The advantage of the P3 architecture over the P4 architecture
is that the additional transmission can be downsized since the motor output torque will
be multiply by the final drive ratio.
To quantify the potential benefit of such topologies the P3 architecture will be studied
with a single shaft (ratio 1 with manual transmission output shaft) and with a double shaft
(ratio 3 with manual transmission output shaft).
Fuel Consumption (L/100km)
CO2 emissions
(g/km)
Fuel Savings (%)
P3 Ratio=1 4,04 96 25,0
P3 Ratio=3 3,89 93 27,7
Table 3-4 : P3 Architectures Fuel Consumption and Emissions.
35
As shown in Table 3-4 an increasing ratio between the motor shaft and the manual
transmission output shaft improves the fuel benefits thanks to a multiplication of the
motor torque at the wheels. As for the previous hybrid topologies the implementation of
an E-Machine allows avoiding the engine working at low load inefficient points (green box
on Figure 3-15).
Figure 3-15 : Engine and Electric Machine Operating Points on NEDC (P3 Ratio=3 Architecture).
As highlighted on Table 3-4 the fuel improvement realized with a P3 architecture and a
ratio of 3 is enough to meet the 2020 regulation.
3.6 FINAL ARCHITECTURE
Thanks to the previous investigation it is now possible to choose the most suitable
configuration for our project. The results of the previous study are summarized in Figure
3-16 and Figure 3-17.
Figure 3-16 : Hybrid Architecture Selection – Fuel Consumption.
1000 2000 3000 4000 5000 6000
0
50
100
150
200
w [RPM]
T [
Nm
]
Engine Operating Points
0 2000 4000 6000 8000 10000-80
-60
-40
-20
0
20
40
60
80
w [RPM]
T [
Nm
]
Electric Machine Operating Points
0
1
2
3
4
5
6
ConventionalGasoline
Stop & Start P1 P2 P3 R=1 P3 R=3
5,385,09
-5.7%3,99
-25.9% 3,52-34.6%
4,04-25.0%
3,89-27.7%
Fue
l Co
nsu
mp
tio
n (
L/1
00
km)
Hybrid Architectures
FUEL CONSUMPTION
36
Figure 3-17 : Hybrid Architecture Selection - CO2 Emissions.
These two figures clearly show that the best architecture in terms of fuel consumption
and emission is the P2 hybrid. But as previously stated this topology does not fit with a
transverse engine, which is the case for the Golf VII reference vehicle, due to the additional
clutch required. Thus since a double shaft is suitable for a transverse engine, the P3
architecture seems more legitimate to avoid a complete modification of the vehicle layout.
Moreover as shown on Figure 3-17 this architecture is able to reach the emission target
(95 gCO2/km) by playing with the additional transmission ratio.
The architecture being selected, it is now possible to optimize it by performing some
sensitivity studies on the E-Machine power and by implementing a new gear strategy in
order to further decrease the fuel consumption. As the vehicle is equipped with an E-
Machine able to provide additional torque to the wheel it is possible to choose the gear
shifting on NEDC and thus optimize it regarding fuel consumption.
0
20
40
60
80
100
120
140
ConventionalGasoline
Stop & Start P1 P2 P3 R=1 P3 R=3
128121
96
84
96 93
CO
2Em
issi
on
(g/
km)
Hybrid Architectures
CO2 EMISSIONS
2020 Regulation
37
4 GEAR SELECTION OPTIMIZATION
In this section is developed an optimized energy management strategy taking the gear
engaged as a second degree of freedom in addition to the engine torque. Moreover more
realistic assumptions will be considered: the Stop & Start function will be disabled during
the first 150 s of the cycle to allow the engine to warm up and the E-Machine will be used
to ensure the thermal engine cranking.
4.1 GEAR SHIFTING STRATEGY
As the transmission is manual, on real driving the gear shifting strategy optimized for fuel
consumption will only be seen as a gear shifting indicator (green box on Figure 4-1). The
driver can therefore choose if he desires to change gear or not.
Figure 4-1 : Gear Shifting Indicator.
4.1.1 PRINCIPLE
The gear shifting strategy implemented on Matlab/Simulink © is the following:
If the vehicle speed is constant or increase a gear has to be engaged. This condition
is required for safety, the vehicle cannot accelerate when the driver is not pressing
the acceleration pedal. This means that the E-Machine cannot run Full Electric with
thermal engine decoupled.
As the P3 architecture has been chosen, if the vehicle decelerates it is
recommended to shift to neutral to decouple the engine from the motor in order
to recover the maximum energy.
During take-off phase the E-Machine runs in Full Electric (1st gear) until the engine
launching speed is reached (750 RPM).
A fuel penalty is added to a gear shift to avoid oscillating gear shifting.
38
A fuel penalty is added to engine switch on/off to avoid oscillating behaviour.
The engine is not allowed to run below 1000RPM when the 2nd, 3rd, 4th or 5th gear
is engaged.
NB: In further developments it can be interesting to implement an e-clutch in order to be
able to disengage the ICE when running in full electric mode. In this way the engine
resistive torque will be deleted, without expecting the driver to press the clutch pedal.
4.1.2 IMPLEMENTATION
The previous Energy Management Strategy is used as a basis for this new strategy.
However instead of defining a vector of 12 candidate for the imposed gear, 6 vectors of 12
candidates are defined one vector for each of the gear. Then as in the previous
management strategy the solution that minimizes the Hamiltonian function will be chosen
and the gear related to this solution recommended.
Flags are raised to notify that a solution is not acceptable if it does not respect the
conditions stated in the section 4.1.1 .
4.2 SIMULATIONS
Simulations are performed on NEDC cycle and on a Real Driving Cycle.
4.2.1 NEDC CYCLE
The optimized gear shifting strategy is displayed on Figure 4-2. As expected, it is visible
on this graph that to save fuel the driver has to upshift as soon as possible.
39
Figure 4-2 : Gear Optimization on NEDC.
Figure 4-3 highlights the fuel flow required for the designed thermal engine on NEDC
Cycle. As previously stated during the first 150s the engine has an idle fuel consumption
due to the disabling of the Stop & Start function. The engine cranking phases are also
visible. Indeed at each vehicle take off there is a delay during which the E-Machine is
running full electric to allow the thermal engine reaching its launching speed.
Figure 4-3 : Fuel Flow on NEDC.
Regarding the battery state of charge, the final value is 0.58 which is close enough to 0.6
(initial SOC) to ensure the charge sustaining mode (Figure 4-4). During the first 150s the
battery is depleting constantly since there is no use of E-Machine during cranking phase
but electric power has to be supplied to auxiliaries (200W).
0
1
2
3
4
5
0 200 400 600 800 1000
Ge
ar (
-)
Time (s)
Gear Optimization on NEDC
Fix. Gears Opt. Gears
0,0E+00
5,0E-04
1,0E-03
1,5E-03
2,0E-03
2,5E-03
0
20
40
60
80
100
120
0 200 400 600 800 1000
Fuel
Flo
w (
kg/s
)
Veh
icle
Sp
eed
(km
/h)
Time (s)
Fuel Flow on NEDC
V_veh d_fuel
Engine cranking phase E-Machine runs Full Electric
40
Figure 4-4 : Battery State of Charge on NEDC.
The fuel consumption on NEDC for this new energy management strategy is 3.74
L/100km and the associated CO2 emissions are 89 gCO2/km. This represents a fuel
economy of 3.9 % compared to the same hybrid architecture (P3) but with the previous
energy management strategy gear shifting being imposed. This is due to the fact that this
new strategy allows engine operating points to be moved to better efficiency points
(green box on Figure 4-5).
Figure 4-5 : Engine and Electric Machine Operating Points on NEDC (P3 - Ratio=3 Architecture with
Optimal Gear Shifting).
This architecture emitting 89 gCO2/km on the NEDC cycle fulfill the requirement of 2020
regulation concerning pollutant emissions and is 10 gCO2/km better than the target state-
of-the-art Diesel Golf VII.
0,35
0,4
0,45
0,5
0,55
0,6
0,65
0
20
40
60
80
100
120
0 200 400 600 800 1000
SOC
(-)
Veh
icle
Sp
eed
(km
/h)
Time (s)
Battery State of Charge
V_veh SOC
1000 2000 3000 4000 5000 6000
0
50
100
150
200
w [RPM]
T [
Nm
]
Engine Operating Points
0 2000 4000 6000 8000 10000-80
-60
-40
-20
0
20
40
60
80
w [RPM]
T [
Nm
]
Electric Machine Operating Points
41
To continue the optimization process it could be interesting to perform a sensitivity study
on motor power since so far the power has been set to the maximum power of mild hybrid
vehicle (15 kW) and on transmission ratio between the motor shaft and manual
transmission output shaft.
These two parameters are actually linked since to ensure engine cranking with the motor
a sufficient torque has to be provided by the motor. Thus if motor power is reduced the
transmission ratio has to be increased to raise the torque multiplication factor between
the motor shaft and the wheels. And if the transmission ratio is reduced the motor power
has to be increased.
It turns out that to ensure engine cranking with E-Machine on NEDC cycle the
transmission ratio need to be at least 3 and the E-Machine Power at least 15 kW, so these
quantities can only be increased. As the E-Machine power cannot be further increased
(15kW max for Mild-Hybrid applications) the only lever to play with is the transmission
ratio. However if the transmission ratio increases too much at maximum vehicle speed
the E-Machine can be driven at too important rotational speed leading to failure. The
maximum rotational speed allowed by the motor is 18 000 RPM and the maximum vehicle
speed is around 190 km/h. Thus the upper limit for the transmission ratio is 3.3.
But to limit the size and the weight of the additional transmission, as a ratio of 3 is enough
to meet the project targets, it has been decided not to increase this ratio and keep a value
of 3.
Let’s now consider a real driving cycle to observe if the CO2 emissions are also acceptable
on day to day life rides.
4.2.2 REAL DRIVING CYCLE IN PARIS
The driving cycle chosen is displayed on Figure 4-6. The journey links IFP School to
Bobigny. It includes urban and extra-urban driving parts but the maximum speed is
limited to 90 km/h which is far below the maximum speed allowed on French motorways.
42
Figure 4-6 : Real Driving Cycle in Paris.
The vehicle speed has been recorded using a GPS watch. There is therefore some
uncertainties linked to the measuring device.
Looking at the cycles speed profile (Figure 4-7), it is clearly visible that the RDE cycle is
much more transient than NEDC cycle and this is especially why this cycle is criticized.
Figure 4-7 : NEDC & RDE Cycles Speed Profiles.
To estimate the benefits of the designed hybrid architecture with the optimized gear
shifting strategy, the fuel consumption of the conventional gasoline vehicle has been
assessed with a simple gear shifting strategy based on vehicle speed. The conventional
gasoline architecture has a fuel consumption of 4.60 L/100km on this cycle.
0
20
40
60
80
100
120
0 500 1000 1500 2000 2500 3000
Veh
icle
Sp
eed
(km
/h)
Time (s)
Cycles Speed Profiles
NEDC RDE
43
By simulating the final hybrid architecture along with the optimized gear shifting strategy
it is possible to reach 3.29 L/100km on this cycle. This represents a fuel saving of 28.5%
compared to the conventional gasoline vehicle. In the same way that on the NEDC cycle,
the E-Machine allows engine operating points to be moved to better efficiency points or
suppressed (Figure 4-8).
Figure 4-8 : Engine and Electric Machine Operating Points on NEDC Conventional Gasoline (Left) /
Proposed Hybrid Architecture (Right).
On this RDE cycle the proposed hybrid architecture emits 79 gCO2/km which is also
below the 2020 regulation limit (95 gCO2/km).
Thus at the end of this study it clearly appears that the proposed hybrid architecture is
able to meet the project target in terms of emission, that is to say having CO2 emissions
lower than 95 gCO2/km, on both NEDC and RDE cycles.
1000 2000 3000 4000 5000 6000
0
50
100
150
200
w [RPM]
T [
Nm
]
Engine Operating Points
1000 2000 3000 4000 5000 6000
0
50
100
150
200
w [RPM]
T [
Nm
]
Engine Operating Points
Hybridization
44
5 E-SUPERCHARGER IMPLEMENTATION
Through downsizing and turbocharging the conventional gasoline engine, hybridization
of the vehicle with an electric motor and optimization of the gear selection strategy, it is
possible to reduce the fuel consumption considerably in order to reach the CO2 emission
target that has been set for the project. However this improvement should not be achieved
in sacrifice of the driving performance of the vehicle. The fun-to-drive should be either
maintained or improved.
Turbocharger is a good enabler regarding improving the specific power of the engine. A
larger turbocharger leads to a greater specific power, while reducing the specific fuel
consumption. However, as the size of the turbocharger gets bigger, the response time of
the engine will be longer, then it will take more time for the driver to feel the torque he
wants when pressing the acceleration pedal. This means degradation of the driving
performance [9]. The potential of the electric supercharger technology to solve this
drawback has thus been explored in this project.
According to the research done by automotive companies such as Valeo, potential gains
through integrating an electric supercharger to the engine include, but are not limited to,
the following aspects:
• Allows extreme downsizing;
• Enables good transient response at any engine speed;
• Allows downspeeding without decreasing the maximum torque that could be
achieved;
• Enables low cost hybridization when coupled with the regenerative system.
A picture of an electric supercharger from Valeo is shown in Figure 5-1. The axle of the air
compressor is driven by a specifically designed electric motor, which is powered by a
battery pack on the vehicle.
Figure 5-1 : Electric-Assisted Supercharger from Valeo.
45
In this phase, simulations have been performed to predict the roll-on acceleration
performance of the vehicle, in the cases of implementing an e-Supercharger or not, and
also using an e-motor during the acceleration or not.
Due to limited data availability and the complexity of air-loop modeling, an e-
Supercharger of 4kW will be explored directly without further consideration on sizing.
5.1 SYSTEM LAYOUT AND REQUIREMENTS
The inlet temperature and pressure of an e-Supercharger is often limited, thus restricting
its location to the upstream of the turbocharger, as shown in Figure 5-2 [10]. The e-
Supercharger is not capable of handling full airflow at high engine loads/speeds, thus a
bypass route for the e-Supercharger might be required. Other layout or solutions also
exist, for example, a motor could be integrated with the turbocharger directly [11].
Figure 5-2: Layout of the 2-Stage Air Charging System with an E-Supercharger. [10]
As mentioned before, the e-Supercharger has to be powered by a battery pack on the
vehicle. Since an alternate current motor is often used, an AC/DC inverter is necessary to
achieve the current frequency inversion. While this AC/DC inverter is often integrated on
the e-Supercharger itself, a DC/DC converter might still be needed to achieve a voltage
step-up or –down from the electric network on the vehicle. A generator is needed to
generate the necessary energy which is then stored in the battery for consumption later.
In the present study, since a 48V battery pack has been adopted and a 48V e-Supercharger,
a DC/DC converter is not requested between them. On the other side, the generator’s role
could be played by the e-motor. A dedicated control module will be needed to control the
T
CA
C
C
eSC
Bypass
valve
Turbocharger
C
46
operation of the e-Supercharger and its cooperation with the turbocharger, but this
control issue will not be tackled in this project due to lack of time.
5.2 SIMULATION METHODOLOGY
Roll-on acceleration performance is a good indicator to represent the fun-to-drive of a
vehicle. The range of roll-on acceleration can be varied. Typical ranges on which different
vehicles are compared include 30-60km/h, 60-100km/h and 80-120km/h. Also, the gear
on which an acceleration is realized can be selected depending on the range. For these
study all the three ranges mentioned above have been investigated, while the 5th gear has
been adopted mainly but other possibilities have also been explored.
The torque requested during the roll-on acceleration could be supplied by both the
internal combustion engine and the electric motor. The logic of simulations for a
turbocharged engine and an engine with both a turbocharger and an e-Supercharger is
the same except for the need of using different data on turbo lags.
5.2.1 ENGINE TORQUE CALCULATION
During an acceleration, the torque supplied by the engine is limited by the maximum
torque at certain engine rotation speeds but also depends on the turbo lag. In this project,
the simulation is implemented through a step-by-step procedure, in which the torque in
each time step is recalculated according to the torque in the previous step, and the torque
increase during the time of a step. Figure 5-3 gives an illustration of the initial condition
of the roll-on acceleration and the different steps of the calculation.
Figure 5-3: Illustration of the Step-by-step Calculation Procedure.
47
Initial condition: N0 corresponds to the engine speed in the selected gear (e.g., the 5th
gear) at the initial speed (e.g., 80km/h). T0 is the torque able to run the vehicle at constant
speed of 80km/h, after which a roll-on acceleration is initiated. N1 is the engine speed
before which the engine torque is kept at a constant level T1, which is the torque that could
be reached instantaneously after the driver presses the acceleration pedal without any
delay. Since the torque request to run at constant speed is relatively small compared to
an acceleration starting from that speed, T1 is usually larger than T0.
Typical calculation step (duration of dτ = 0.05s): At the beginning of a step, the engine is
running at the speed Ni, giving the torque Ti.
Ni+1 at the end of this step can be obtained by recalculating the vehicle speed, based on:
The speed of the previous step;
The acceleration enabled by the torque of the previous step;
The gear engaged.
Ti+1 can be calculated based on Ti and the increase of torque during this step (of duration
dτ):
Ti+1 = Ti + dTi+1 (Equation 5-1)
Where dTi+1 = dτ * T’(Ni+1, T i).
T’(Ni+1, T i) is the slope at torque Ti on the Time to Torque curve at speed Ni+1. However,
the evolution of the Time to Torque Curve has to be set first.
Figure 5-4: Typical Shape of the Time to Torque Curve at a Constant Engine Speed. [12]
48
Modeling of the Time to Torque curve: The typical shape of the Time to Torque curve
obtained by experiments on engine test benches is shown in Figure 5-4 [12]. It can be
divided into two phases. The first phase is the constant torque phase, which is quite short
and results from the lag of the air loop. In this phase the engine torque is instantaneously
increased to a certain level and then kept for a duration tt0. The second phase reflects the
ramp of the engine torque up to the maximum torque that could be reached at this
particular engine speed. As the torque gradient increase in the beginning of this phase is
usually large and tends to decrease with time until reaching zero, the exponential function
might be a good choice to approximately model this behavior. An exponential function has
thus been adopted for this study.
The complete function implemented to model the Time to Torque curve is given below:
0
0 0
0
0
, *
,
,* * *( )
i
i
max i max i i
T NT t N
a T N a T N
t tt
T N exp c t ttt tt
(Equation 5-2)
Where a and c are constants that indirectly set the time needed to reach full load tt.
This seems to be an explicit function which can approximately match the trend of the
actual turbo lag phenomenon. To keep it simple, it can be assumed that the overshoot
parameter a is set to 1. Then only c needs to be determined.
Figure 5-5: Curves Obtained by Changing Parameter c in the Time to Torque Function (a=1, tt0=0).
In practice, the data known are the initial torque 0 iT N , the duration of the constant
torque period, the maximum torque and the time needed to reach 90% maximum torque
at fixed certain engine rotation speeds. Based on this point of 90% maximum torque, the
constant c can be estimated for each engine speed:
49
0 00.9 * * ( ) ( ) ( ( ) ( )) * *( )max i max i max i iT N a T N a T N T N exp c tt tt
max max max 0 0ln(( * 0.9 ) / * ( )( ) ( ) ( ( ) ( ))) /i i i ic a T N T N a T N T N tt tt
Once c is known, the Time to Torque curve gradient at any point can be calculated. This
gradient is used in the step-by-step methodology previously described to calculate the
torque increase at each step.
5.2.2 MOTOR TORQUE
The torque coming from the e-motor can be obtained by interpolation using the motor
map, knowing the motor speed. As the motor is powered by the battery pack, attention
needs to be paid at each step of calculation to the State of Charge level of the battery. It
should always be kept above the specified lower limit to prevent any premature aging.
Figure 5-6: Interpolation on the Motor Characteristic Curve to Obtain the Motor Torque.
5.3 PERFORMANCE CALCULATION RESULTS
By implementing the methodology described in the previous section, the roll-on
acceleration performance on different ranges is obtained. Comparisons will be focused on
the difference between implementing or not an e-Supercharger, and also the difference
between implementing or not an e-motor.
5.3.1 TIME TO TORQUE AT CONSTANT ENGINE SPEEDS
The value of tt for each engine speed can be interpolated from the Time to Torque data at
several engine speeds. Interpolations have been done respectively for the turbocharger
and a 4kW e-Supercharger from Valeo, as shown in Figure 5-7.
50
Figure 5-7: Time to torque vs engine speed.
Using the formulas proposed in the previous section, the torque and thus the BMEP
evolution paths at 1250rpm and 3000rpm are plotted respectively in Figure 5-8 and
Figure 5-9. It can be seen that at 1250rpm, there is a big gap between the implementation
with and without an e-Supercharger, while at 3000rpm the two torque evolution paths
are actually quite close. This indicates that there is a negligible impact of the e-
Supercharger when the engine is running at high rpm.
Figure 5-8: Torque Evolution at 1250rpm.
Figure 5-9: Torque Evolution at 3000rpm.
0,0
0,5
1,0
1,5
2,0
2,5
3,0
3,5
4,0
4,5
1000 1500 2000 2500 3000 3500
Tim
e to
max
imu
m t
orq
ue
(s)
Engine speed (rpm)
Response time without eSC
Response time with eSC
0
5
10
15
20
25
0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0
BM
EP (
bar
)
t (s)
TC+Esc
TC
0
5
10
15
20
25
0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0
BM
EP (
bar
)
t (s)
TC+Esc
TC
51
5.3.2 OVERALL RESULTS ON ALL ACCELERATION RANGES CONSIDERED
Figure 5-10 gives a summary of all calculation results for the different ranges of roll-on
acceleration, i.e. from 80-120km/h, 60-100km/h and 30-60km/h. All these accelerations
are performed on the 5th gear with a reduction ratio of 3.05 from the engine to the wheels
(V1000=38km/h/krpm).
Figure 5-10 : Roll-on Performance of the Vehicle with Different Architectures.
Roll-on 80-120km/h: When the e-motor is not available or turned off during the
acceleration, a 4% improvement is observed for the architecture with an e-Supercharger
compared with the solution with only a turbocharger. If the e-motor is turned on, the time
needed to reach 120km/h could be shorten considerably, while an e-Supercharger can
still achieve around 4% improvement additionally.
Roll-on 60-100km/h: The e-Supercharger helps saving around one second without any e-
motor assistance, and can give a reduction delay of 6.5% when the e-motor is used.
Roll-on 30-60km/h: The gain is maximum for this test. When no torque is asked from the
e-motor, 1.35 seconds could be saved by implementing the 4kW e-Supercharger, which
means around 10% improvement.
To summarize, the Electric-assisted supercharger is able to considerably reduce the time
to torque, and this improvement is especially interesting when the engine is running at
low engine speed. On the other side, the e-motor is another good enabler of better
acceleration performance, however as we will see later, it requires rapid depletion of the
battery capacity.
14.0113.42
14.0513.46
12.42 12.70
9.82
8.88
5.97
9.46
8.30
5.52
5
6
7
8
9
10
11
12
13
14
15
80-120km/h 60-100km/h 30-60km/h
Du
rati
on
(s)
Ranges Of Roll-on Acceleration
TC w/o EM
eSC w/o EM
TC w/i EM
eSC w/i EM
Diesel Golf VII
52
To conclude implementing a 4kW e-Supercharger alone is not able to gain enough
advantage over the Golf VII 1.6 TDI model, the 15kW e-motor is needed additionally. This
might be attributed to the short gears we have selected and to the fact that the interaction
between the e-Supercharger and the rest of the air loop components are not taken into
account.
5.3.3 FOCUS ON 80-120KM/H ACCELERATION
In this section detailed information will be given for roll-on acceleration from 80 to
120km/h, to enable some further analysis and discussion.
Figure 5-11 gives the engine torque evolution paths during the acceleration from 80 to
120km/h, respectively for the architecture with and without an e-Supercharger. Notice
that the concerned range of engine speed is between 2000rpm and 3200rpm. Recalling
Figure 5-7, this means the difference in time to reach maximum torque is between 0.25s
and 1.2s. Thus the final difference in the performance is limited to 1.2s, and should be in
fact lower. This explains why we did not gain so much with an e-Supercharger (Figure
5-10).
Figure 5-11: Engine Operating Point Evolution (without Motor).
It is also possible to plot the vehicle acceleration evolution for different cases (Figure
5-12). This gives a more clear view on where the differences come from. Velocity
evolution profiles are plotted in Figure 5-13 to illustrate the roll-on experiences for
various cases.
50,0
70,0
90,0
110,0
130,0
150,0
170,0
190,0
210,0
1000 1500 2000 2500 3000 3500
Torq
ue
(Nm
)
Engine speed (rpm)
TC+eSC
TC
Max Torque [Nm]
53
Figure 5-12: Vehicle Acceleration Evolution.
Figure 5-13: Vehicle Speed Evolution.
Let’s now consider the state of charge (SOC) of the battery pack during the accelerations.
As highlighted on Figure 5-14, the main electricity consumption comes from the operation
of the e-motor. When the e-motor is on, the SOC of the battery can decrease by 10% in less
than 10 seconds. This might not be good for needs to run in pure-electric mode later to
enable better fuel economy. In contrast, the e-Supercharger leads to only 3% of SOC
decrease when the target speed is reached. This is consistent with the sized power levels
of the e-motor (15kW) and the e-Supercharger (4kW).
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Acc
eler
atio
n (
m/s
)
t (s)
TC w/o motor
eSC w/o motor
TC w/i motor
eSC w/i motor
75,0
80,0
85,0
90,0
95,0
100,0
105,0
110,0
115,0
120,0
125,0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
velo
city
(m
/s)
t (s)
TC w/o motor
eSC w/o motor
TC w/i motor
eSC w/i motor
54
Figure 5-14: Battery SOC Evolution.
Finally, the potential impact of changing the gear ratio has been investigated. In Figure
5-10 it can be seen that with the e-motor, the roll-on performances are actually far better
than the state-of-art diesel engine. This means the gear ratios can be further increased in
order to reach further fuel-economy benefit, while keeping a good fun-to-drive.
Let’s now change the 5th gear ratio (V1000) from 38 to 47.5km/h/krpm. Table 5-1
highlights that after lengthening the final gear ratio, the performance of the architecture
with an e-Supercharger can still reach the target derived from the competitor, while
implementing a turbocharger only is no longer able to meet the target. The benefit
brought by the e-Supercharger is enlarged now compared with the shorter gear ratio,
since the engine is now running at lower speeds. This gear ratio increase is not in vain
since it could bring even more fuel consumption gain, which is the first-priority concern.
V1000 5th gear (km/h/krpm) 38 47.5
Architecture Turbo only Turbo + eSC Turbo only Turbo + eSC
Time needed from 80-120km/h (s) 9.82 9.46 12.16 11.54
Table 5-1: Effects of Varying the Gear Ratio (E-Motor Active).
Thus if more time was allocated to the project it could be interesting to perform an
optimization on the manual transmission (gear length, addition of a 6th gear).
40,0%
45,0%
50,0%
55,0%
60,0%
65,0%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
SOC
t (s)
TC w/o motor
eSC w/o motor
TC w/i motor
eSC w/i motor
55
6 RESULTS
The proposed gasoline hybrid needs to fulfill emission standards, cost to the OEM, and
satisfy the customer expectations in terms of performance and cost better than the
existing Diesel Golf VII. The comparison in each domain is individually discussed.
6.1 FUEL CONSUMPTION & CO2 EMISSIONS
6.1.1 NEDC CYCLE
Figure 6-1 : NEDC Cycle Fuel Consumption & Emissions.
From the above plot we can observe a 32.5% decrease in CO2 emissions compared to the
conventional gasoline and a 10 % decrease in emission compared to the baseline Diesel
Golf VII. Hence from a legislative point of view a clear benefit is observed and is good in
terms of marketing strategies.
6.1.2 RDE CYCLE
The RDE cycle is implemented in the model and the emission values obtained show a 30%
reduction between the conventional gasoline and the final proposed hybrid architecture.
5,5
3,8 3,74
132
9989
0
20
40
60
80
100
120
140
0
1
2
3
4
5
6
Conventional Gasoline Diesel Proposed Hybrid
CO
2Em
issi
on
s (g
/km
)
Fue
l Co
nsu
mp
tio
n (
L/1
00
km)
NEDC CYCLE
Fuel Consumption ( L/100km) C02 emission
10%
56
Figure 6-2 : RDE Cycle Fuel Consumption & Emissions.
6.2 SYSTEM OVERVIEW
Figure 6-3 : System Overview.
The system is a P3 hybrid architecture implemented in a transverse frontal gasoline
engine layout. The system consists of a 5 speed manual transmission and an e-motor is
linked to the transmission through a gear (in our case ratio: 3). There are two battery
packs (12V and 48 V) to the rear of the car and the circuit is built using a DC/DC inverter.
The other conventional layout remains the same. An e-Supercharger is linked to the
engine in the front and is powered by the e-machine.
Tank
48V Battery
DC/DC
Inverter
12V Battery
12V Electric
Loads
E-Supercharger
(4kW)
1.2L 3-cylinder
Gasoline Engine
Electric Motor
(15kW)
5-speed Manual
Gearbox
57
With respect to the conventional gasoline vehicle the main additional components are:
• E-Supercharger (4kW)
• E-motor (15 kW)
• Battery pack (48V)
• DC/DC converter
• Cables
The weight and cost addition due to the above mentioned are discussed in the next
section.
6.3 ADDITIONAL WEIGHT & COST ESTIMATION
The weight of individual components described above and a rough indicated market price
are listed. The price is and indicative market price for 2020 for mass production (> 200
000 pieces/year). These give a total of a 43 kg weight addition and a 1285 € cost addition.
However these are just approximate values and may differ with OEM and supplier.
6.4 COST TO THE OEM
The CO2 reduction cost is an important value to the OEM and the lower CO2 emissions
cost justifies the choice of implementing a gasoline hybrid version for a Diesel car. The
assumptions made in this synthesis include that the Diesel engine is 40 kg heavier than a
gasoline version 3-cylinder and cost about 1000 euros more.
Comparison: Proposed
Hybrid
Conventional
Gasoline
vs. Diesel (Golf VI 1.6TSI)
Weight Increase + 43 kg + 3 kg
Additional Cost + 1285 € + 285 €
CO2 Benefit (g/km) 43 g/km 10 g/km
CO2 Reduction Cost 32.1 €/gCO2/km 28.5 €/gCO2/km
Table 6-1 : Cost to the OEM.
58
6.5 TOTAL COST OF OWNERSHIP ESTIMATION
In Table 6-2 are listed the assumptions used to estimate the total cost of ownership.
Gas Hybrid Diesel
Mean distance in France (km/year)
12 700 12 700
Ownership duration (year)
5 5
Fuel rate (€/L) (Super 98) 1.28 1.17
Fuel consumption (L/100km)
3.72 3.9
Table 6-2 : Total Cost of Ownership Assumptions.
By statistical surveys the mean ownership duration in Europe for a car is found to be a
period of 5 years. And the present day average fuel rate throughout Europe for gasoline
is 1.28 €/L and diesel is 1.17 €/L. From these values the total cost spent by the customer
on fuel can be estimated. The values of gasoline and diesel are averaged and are
considered for the year 2015.
Cost Calculation Gas Hybrid Diesel
Insurance (€/year) 385 550
Maintenance (€/year) 900 1000
Fuel consumption (€/year) 605 560
Total cost (€/5years) 9450 10550
Table 6-3 : Total Cost of Ownership Calculation.
The rough average amount a customer will tend to spend on the vehicle throughout its
lifetime is found to be lesser for the proposed final hybrid architecture with a net saving
of 1100 €/ 5 years. This data is based on the fact that market in Europe is Hybrid friendly
meaning the insurance is lesser for hybrid cars. The maintenance is assumed lesser than
Diesel because of better engine operating conditions and lesser soot.
59
CONCLUSION AND DIRECTIONS FOR FURTHER STUDY
Through this project, mild hybridization and e-Supercharger technology have been
applied to a mid-size gasoline powered vehicle in order to reach both the emissions and
driving performance level of a state-of-art diesel vehicle model (Golf VII 1.6 TDI).
To reach these targets a 1.6L 4-cylinder gasoline engine has first been downsized to
decrease the engine displacement while keeping the output power necessary to reach
good performance level. Then the fuel consumption gain of decreasing the number of
cylinders to three has been explored by using an explicit heat losses model. The
advantages and disadvantages of various hybrid architectures have been compared, and
main efforts have been deployed on the sub-categories of the parallel hybrid. A Simulink
© model has been built to compare the fuel consumption gain related to each hybrid
topology. At the end of this study a non-coaxial P3 architecture has been selected, since it
features both good fuel consumption gain and proper compatibility with the transverse
layout of the C-segment cars. At the same time, sizing of the electric motor and battery has
been done to satisfy various constraints. Optimization of gear selection and energy
management strategy has been implemented, in order to achieve better fuel economy on
the traditional NEDC cycle and a Real Driving Cycle established by the team. To keep or
even improve the fun-to-drive, the potential of e-Supercharger technology has been
investigated. Explicit models have been built to model the time to torque behavior and a
step-by-step methodology has been implemented to simulate the roll-on performance of
the proposed hybrid on different ranges of acceleration. Finally, a cost estimation has
been done to evaluate the balance between gain and cost.
Main findings and conclusions are as follows:
• Through Mild Hybridization (48V) and optimization of the energy management
strategy, it is possible to achieve good CO2 emissions (89 gCO2/km), better than both
conventional gasoline and state-of-the-art diesel vehicle and thus meet the 95
gCO2/km target.
• By implementing both an electric machine (15kW) and an electric supercharger
(4kW), the driving performance of the proposed hybrid can be maintained or even
improved.
60
• Although the price of a mild hybrid can be more expensive compared with a
conventional diesel, benefits can be gained back during its lifetime.
Thus this study indicates that the solution of 48V hybridization of a mid-size gasoline
vehicle using an electric-assisted supercharger is promising, and poses good
competitiveness over a state-of-art diesel.
Due to the time limit, the optimization process has not been tackled in an exhaustive way.
Further works remain to be done to make the results more accurate and reliable. This
includes:
• A detailed modeling of the air-loop to predict more accurately the performance of the
combination of a turbocharger and an e-Supercharger. This will allow to take into
account the interaction between the e-Supercharger and the turbocharger and to
simulate the BSFC improvement due to engine back pressure reduction when using an
e-Supercharger.
• An iteration process to further optimize the gear ratios and find the best trade off in
terms of fuel consumption and fun to drive.
• Fuel benefit estimation for the proposed hybrid architecture on other cycles such as
WLTC.
61
REFERENCES
[1] Heywood, John, B. (1988) “Internal Combustion Engine Fundamentals”, McGRAW-HILL
INTERNATIONAL EDITIONS.
[2] Millo, F., Badami, M., Ferraro, C.V., Lavarino, G. and Rolando, L. (2010), “A Comparison
Between Different Hybrids”, SAE Technical Papers.
[3] Rajendran Vallur, A., Khairate, Y. and Awate, C. (2015), “Prescriptive Modeling,
Simulation and Performance Analysis of Mild Hybrid Vehicle and Component Optimization”,
SAE Technical Papers.
[4] Nesscap (2015), “Ultracapacitor Data Sheet”, available at:
http://www.nesscap.com/ultracapacitor/EDLC/Supercapacitor/Large_cell_supercapaci
tor_family/cylindrical_supercapacitor_cell.jsp [accessed on 30/05/15]
[5] Kasseris, E. P. and Heywood, J. B. (2007), “Comparative Analysis of Automotive Powertrain”, SAE Technical Papers. [6] Sciarretta, A. (2015), “Hybrid Energy Management relevant to Powertrain Control and Mechatronics (TU10)”, IFP School: unpublished. [7] Levassor, W., Wysoki, L. (2015), “Longitudinal Dynamics Practical Works relevant to Engine Technology (TU4)”, IFP School: unpublished. [8] Ravello, V. (2015), “Hybrid Drive Train relevant to Transmissions and Alternative Drive Trains (TU8)”, IFP School: unpublished.
[9] Yamashita, Y., Ibaraki, S., Sumida, K. (2010), “Development of Electric Supercharger to Facilitate the Downsizing of Automobile Engines”, Mitsubishi Heavy Industries Technical Review, Vol. 47 No. 4.
[10] Speaker from MAHLE Powertrain (2015), “eSupercharging for Heavily Downsized Gasoline Engines”, SIA Conference Report.
[11] Prasad, S. D., Beshah, A. and Robert, P. (2010), “Coordinated Electric Supercharging and Turbo-Generation for a Diesel Engine”, SAE Technical Papers.
[12] Petitjean, D. (2015), “Turbocharger technology and intake supercharging for engines
(TU5)”, IFP School: unpublished.