FACULT Y OF MEC HANICS AND TECHNOLOGY DEPARTMENT AUTOMOBILES AND TRANSPORT MASTER “AUTOMOTIVE ENGINEERING FOR SUSTAINABLE MOBILITY” THESIS Scientific… [616339]

UNIVERSITATEA din PITEȘTI
Facultatea de Mecanică și Tehnologie
Departamentul Autovehicule și Transporturi
Fondat în 1969

UNIVERSITY OF PITESTI
FACULT Y OF MEC HANICS AND TECHNOLOGY
DEPARTMENT AUTOMOBILES AND TRANSPORT
MASTER
“AUTOMOTIVE ENGINEERING FOR SUSTAINABLE MOBILITY”

THESIS

Scientific supervisor:
Prof. univ. dr. CLENCI ADRIAN -CONSTANTIN

Graduate:
SAMAREA NU GEORGE

Pitesti
2020

George Samareanu Dissertation thesis, Pitesti 2 020
– 3 –
Table of Contents

1. Planning ………………………….. ………………………….. ………………………….. ………………………….. ……4
2. Introduction ………………………….. ………………………….. ………………………….. ………………………….. 7
3. Literature review ………………………….. ………………………….. ………………………….. ………………….. 8
3.1 Software and HIL models development ………………………….. ………………………….. …………….. 8
3.2 European regulation ………………………….. ………………………….. ………………………….. …………. 12
3.3 Fuel consumption and pollution ………………………….. ………………………….. ……………………… 15
4. Experimental investigation ………………………….. ………………………….. ………………………….. …..22
4.1 Test procedur e ………………………….. ………………………….. ………………………….. …………………. 22
4.2 Equipment used ………………………….. ………………………….. ………………………….. ……………….. 25
4.2.1 Chassis Dynamometer ………………………….. ………………………….. ………………………….. …25
4.2.2 HIL environment ………………………….. ………………………….. ………………………….. ……….. 26
4.3 Results and discussion ………………………….. ………………………….. ………………………….. ………. 30
5. Conclusion and future work s ………………………….. ………………………….. ………………………….. ..34
6. Annexes ………………………….. ………………………….. ………………………….. ………………………….. …..35
Annex 1: Abbreviation s and notation ………………………….. ………………………….. ……………………. 35
Annex 2: EURO norm s values ………………………….. ………………………….. ………………………….. …36
Annex 3: ICE roadmap ………………………….. ………………………….. ………………………….. …………… 37
Anne x 4: HIL control desk interface ………………………….. ………………………….. …………………….. 38
Annex 5: Aver age specific consumption ………………………….. ………………………….. ………………. 39
Annex 6: Fuel injected mass ………………………….. ………………………….. ………………………….. ……40
Annex 7: Gear engaged ………………………….. ………………………….. ………………………….. ………….. 41
Annex 8: Engine speed ………………………….. ………………………….. ………………………….. …………… 42
Annex 9: Outputted Torque ………………………….. ………………………….. ………………………….. ……… 43
Annex 10: Vehicle speed ………………………….. ………………………….. ………………………….. ………… 44
7. References ………………………….. ………………………….. ………………………….. ………………………….. .45

George Samareanu Dissertation thesis, Pitesti 2 020
– 4 –
1. Plannin g
As with a ny other project, I needed a planning of all the stages shown in figure 1, in order
to have an overview of the project as well as to solve the problems that arise immediately. In
order to minimalize the problems that can arise, I decided to include in the project planning three
milestones which are: SA, BE and PP .1

Figure 1 Project planning
The PP milestone was put it in the weeks in which at University of Pitesti we hold the
SCSS event in every year. Also, based on the Gantt chart, the following flowchar t presented in
figure 2 was created in order to exemplify the whole process behind .
In order to carry out the work , in addition to the Gantt diagram shown in Figure 1, a
Kanban technique was used in which the chapters were sorted as follow:
 To do, used for chapter that didn’t started
 In progress, were used for the chapter which started at the end of the previous
chapter.
 Postponed, used for the chapter in which some info’s were vague in which I need
to read again the documents
 Done, used for finished chapte r in which I’ll don’t have to make additional
changes

1 SA – supervisor approval; BE – begin experiment; PP – paperwork pre sentation

George Samareanu Dissertation thesis, Pitesti 2 020
– 5 –
In order to be similar as project, we need to know how efficient the steps are done as it
shown in figure 2 below. So, for this purpose I calculated the flow efficiency (Pe) as in any agile
project , based on working days (W d) and paused days (P d). This represents an KPI indicator and
it was calculated with the formula:

(1)

Figure 2 Project efficiency
Each of the 9 months, from November 2019 to July 2020, had the efficiency
presented in the table 1 below and , overall, the project had 83% efficiency2 with 210 days of
effective wo rk and 43 days of rest . Also, based on the Gantt chart, the flowchart presented in
figure 3 was created in order to exemplify the whole process behind.
Table 1 Project efficiency by month
Mont h Efficiency Work ed days
November 2019 90% 27
December 2019 65% 20
January 2020 87% 27
February 2020 83% 24
March 2020 90% 28
April 2020 87% 26
May 2020 94% 29
June 2020 70% 21
July 2020 80% 8

2 The project efficiency was above the 80% target set at the beginning of the work

George Samareanu Dissertation thesis, Pitesti 2 020
– 6 –

Figure 3 Project flowchart

George Samareanu Dissertation thesis, Pitesti 2 020
– 7 –
2. Introduction
Since emissions regulation become stricter and stricter, th e electronic control in
automotive industry had to keep up with it in order to insure environmental protection. With
electronic engine management we can control and calibrate an engine in order to maintain top
performance, fuel economy, system diagnostics and lower emissions.
In the last decade, automotive manufacturers realized that raw materials such as oil are
limited .3 They become aware of this scenario in which we consume our resources and they want
to preserve it for next generations. Therefore, they began to manufacture hybrid cars which have
less emission that normal cars, began to encourage people to buy and use electric vehicles more
and more because at the exha ust pipe we have zero emissions.
The ICE is the same as 30 years ago for working prin ciple level. The difference exist s
and appears in electric and electronic management of the engine and other components due to
safety and comfort such as:
 “additional features” of the engine such as the catalyst, particle filter, etc. which have
one single goal and that is to reduce as much as possible the exhaust emissions;
 ADAS systems to improve occupant and pedestrian safety and to improve the comfort
inside the vehicle.
Another improvement in the automotive sector is the use of HIL simulation to valid ate
different system and functionalities that will be used on a real vehicle like electrical behavior of
sensors and actuators, ADAS systems, testing calibration strategies, implement new
functionalities or correct issues that appear in serial life, fuel s trategies etc. as it shown in chapter
4.2.
This paperwork wants to make a comparative analysis between the measured and
recorded data at chassis dynamometer bench versus the data measured and recorded at H IL
bench about fuel economy. For this, the paperwor k is structured as follow:
1. Planning in order to achieve the results
2. Introduction to sustain the paperwork goal and subject
3. Literature review in which I present the technical information about software
development, European regulation and fuel consumption
4. Experimental investigation in which I reproduced on HiL bench the WLTP
driving cycle done at chassis dynamometer and also , detailed technical
information about the equipment used and the test results
5. Conclusion about test result and future work s with HiL b enches

3 Not only oil used for fuels but also, iron, rubber, sand and quarts used for glass. In the same way, not only this are
limited resources but also, they can pollu te very much if are not dispose in the right manner.

George Samareanu Dissertation thesis, Pitesti 2 020
– 8 –
3. Literature review
3.1 Software development
Software development is a process in which the final product (application) is conceived
based on the initial requirements. The process of software development includes designing,
documenting , programming , testing and validation, etc. for a system inside the project.4
In software development life cycle, there are two main considerations, one is to
emphasize on process and the other is the quality of the software and process itself. [1]
Software development in automotive industry is a combination between traditional
development which consist in pre -organized stages and more complex metho dologies like agile
development with Kanban boards. Those methodologies and the agile concept u se iterative and
incremental phases used in development. In figure 4 the agile development concept in
automotive industry is exemplified by the 7 stages that contribute to the final product which
reach the customer.

Figure 4 Agile development concept
Every stage in the development phase has a well -defined role and purpose as follow:
a) Brainstorm is the stage used to define and analyze the initial requirements for the
software ;
b) Design is the phase in which the documentation for that functionality and/or syst em is
made ;
c) Development is where the functionality of the system (ECU software) is integrated and
validated in order to ensure proper work ;

4 These general steps are used in many fields not only in automotive and can be used not only for specific system
but also for finding and implementing a solution to an issue, developing new and enhanced functionalities, etc.

George Samareanu Dissertation thesis, Pitesti 2 020
– 9 –
d) Assurance is used to resolve errors for that functionality of the system ;
e) Deployment is the last step before the fin al product reach the customer ;
f) Deliver to client is the last step but not the least because some problem in different cases
reach the customer and new iteration of the software is needed in order to solve the
problems;
g) Next iteration is used in case some p roblems reach the customer or they are well known
problems with no impact on customer side but can affect newer functionalities inside the
software.5
These incremental requirements refinement further refines the design, coding and testing
at all stages of production activity. In this way, the requirements work product is as accurate and
useful as the final software itself. [2]
Kanban board is a development method which manage s and improve s work across
human interactions by balan cing the available capacity and handling of blocking points inside
the devel opment by using several milestones and statuses .
Also, besides agile development, in automotive industry we need to have a transparency
between conception and validation for a sys tem requirement. For this, the W development6
showed in figure 5 is used.

Figure 5 W development cycle [3]

5 Also, in this case, some problems at software levels can appear when the vehicle reached the customer like
unintended acceleration was experienced as in Toyota case from 2009 -2010
6

George Samareanu Dissertation thesis, Pitesti 2 020
– 10 –
Embedded applications tend to grow and complexity and require sophisticated test
methods. One of these methods is Hard ware -in-the-loop (HIL) simulation, an approach that has
been introduced by the aerospace and defense industries in the 1950s. [4]
The techniques used in software development can also be used in the development of HIL
models and environments with few modifications or even as they are . Developing a new HIL
model need to take into consideration a various list of factors as: sensors and actuators used,
engine type, injection type, additional functionalities like stop & start, cruis e control, etc.
HIL tests have become a standard in automotive industry for EMS validation. These
validations are the last step in EMS V -cycle development before vehicle tests; however,
promises that HIL tests will completely replace vehicle tests have not yet been accomplished. [5]
In figure 6 it is shown the structure of the ECU and the software inside it, which contain a
platform and an applicative software. The applicative software is composed from the runnable
software, interface and basic software. The platform is composed from ECU hardware and basic
software which are made by the supplier of the ECU. Also, the supplier provide the runnable
software. With all this “contraing” the interface scope it is to maintain and ensure a good
compatibility between car manufacturer software and the hardware and software provided by the
supplier.

Figure 6 ECU hardware and software structure [3]

George Samareanu Dissertation thesis, Pitesti 2 020
– 11 –
HIL models consists of numerical model for powertrain made in A MESim® and real
time model made in Matlab® and Simulink®. This combination allows real time model to call
power -train models from AMESim® as Simulink S -function. Another reason for splitting and
using two different model s is that modeling the powertrain in side AMESim® allows to calculate
much easier a list of parameters as: volumetric efficiency, mechanical efficiency, combustion
validation, engine components7, etc. This mixed language between Matlab®, Simulink® and
AMESim® uses a master -slave co -simulation .8
In figure 7 below, it is shown the flowchart for developing new HIL models from scratch
to finish.

Figure 7 HIL development flowchart

7 Such as turbochargers, valves, injectors, cata lysts, etc. Or even combination between different domain like
mechanics, fluids, thermodynamics, electric, etc.
8 Simulink® is used for the control part and AMESim® is used for the engine components and working principle.
This mixed connection allows also the use of two separate solver, one for Simulink® and other for AMESim®.

George Samareanu Dissertation thesis, Pitesti 2 020
– 12 –
3.2 European regulation
In European Union, emissions from light duty and heavy -duty vehicle like nitrogen
oxides ( NO x), particulate matter (P M), particle number (P N), hydrocarbons unburned (HC),
carbon monoxide (CO) and carbon dioxide (CO 2) are regulated based on different stages known
as EURO norm .9
Emissions can be categorized as primary pollutants and secondary pol lutants. Primary
pollutants are dangerous and harmful to human and environment immediately and, the secondary
pollutants form different compound when release into the atmosphere. 10
Air pollution can be defined as addition of any material, which will have a deleterious
effect on life upon our planet. Automobile pollution of the due to various toxic gases and
particula te matter has become a global problem. [6]
The problem of air pollution and emissions in most of the cities and town around the
word is due to rapid growth in socio -economic activities. This growth led to increased
degradation of the environment which need to be protected as much as we can.
In figure 8 is exemplified the context of EURO 6 with all its variation based on the
driving cycles NEDC, WLTC and RDE

Figure 8 EURO 6 context [3]
In figure 9 is exemplified the theoretical assumption of EURO 7 as the next regulation
inside European Union based on the driving cycles NEDC WLTC and RDE.

9 For light duty vehicle the stage EURO is followed by Arabic number like EURO 1, EURO 2, EURO 3, EURO 4, EURO 5
and EURO 6. For heavy -duty vehicle the same concept is used but inste ad of Arabi c numbers, we use roman
umbers.
10 Primary emissions are substances which are emitted directly from the sour into the atmosphere and haven’t
been in any chemical reaction. Beside these, there are secondary emission which after were release from t he
source they will create different compounds (i.e. NO2)

George Samareanu Dissertation thesis, Pitesti 2 020
– 13 –

Figure 9 EURO 7 norm assumption [3]
Many countries inside EU decided to ban or to sanction the ICE vehicle which have an
emission norm lesser than EURO 3, EURO 4 and/or even EURO 5. Countries like France and
Germany start to b an vehicle with lower emission standards right away in different cities while
Norway wanted to completely ban ICE vehicle by 2025 by encouraging people to buy EV
powered vehicle .
In Norway, a “polluter tax” was introduced to discourage people from buying gasoline
and/or diesel vehicles and even decided to increase the price for fossil -based fuel.11
Also , in 2020 in Romania , the local government in Bucharest tried to introduce a
pollution tax for vehicle with a pollution norm lower than EURO 4 called “Oxygen Vignette” .
This tax allow vehicle to enter and be used in the central area in Bucharest.
For passenger cars, a comparation between emission norm and the test procedure that is
used is shown in table 2 in which from 2020 all new vehicle will be “judged” by EURO 6 D-full
and the test procedure for homologation will be WLTP.
Table 2 Comparation between EURO norms and homologation test procedure
2016 2017 2018 2019 2020 2021
Emission
norm Euro 6 B Euro 6 D -temp Euro 6 D -full
Test
procedure NEDC WLTP

11 Beside those drastic changes for ICE power vehicle, the future owners of EV cars have different benefits like sales
taxes removed from EV cars.

George Samareanu Dissertation thesis, Pitesti 2 020
– 14 –
Europ ean Union has started to put more and more pressure on car manufacturers in order
to lower emissions and GHG effect by adopting the CAF E indicator in 2009 by CO2
regulation12. This measure is necessary because the level of CO 2 emissions is high enough, and
this pollutant is the main gas off greenhouse effect.
European Environment Agen cy (EEA) gathers data about environmental pollution from
different sectors like automotive industry, agriculture, naval and maritime sector, etc., in order to
better improve our environment and to move forward to sustainability.
CAFE impose s to car manufacturers to have an average level of CO2 equal to 95
gCO2/km in the previous year .13 This indicator can be found also in figure 10 as a 2020 target.

Figure 10 Average CO2 level for passenger cars gathered by EEA [7]
In case of, car manufacturers are unable to have and average level of CO2 equal to 95
grams per kilometers, another pollution tax will be applied . If a car manufacturer exceed ed the
CAFE indicator, it will be charged with a penalty equals to 95 euros for each gCO2/km and car
sold.
As an example, if a car manufacturer sold 1 million vehicles with an average level of CO 2
equal to 100 g/km he will be penalized as follow:

12 CO2 regulation by EC No 443/2009
13 In the past, ca r manufacturers achieved average CO2 per vehicle segment such as passengers, SUV, LCV, etc. but
now, this average will be realized on the models instead of the segment.

George Samareanu Dissertation thesis, Pitesti 2 020
– 15 –
3.3 Fuel consumption and pollution
In order to understand fuel consumption, it is important to know the process by which the
various types of fuel are made. In figure 1 1 it is shown the composition of the f uels.

Figure 11 Fuel composition
Beside the general fuel composition shown in Figure 11 , in automotive industry the fuels
have the following compound added to it in order to modify its properties:
 Hydrocarbons from 70% up to 99,9%
 Biofuels14 between 0% and 3 0%
 Additives15 in proportion of less than 1%
The sulfur content present in the fuel has a harmful effect on the process of post –
treatment of the exhaust gases. This is present in the fuel due to distillation and refining process
of the oil.
In figure 1 2 is presented the process of oil refining in order to create different fuels, oils,
etc. This process starts with the raw material called petroleum which is introduced into a furnace
and the heated up to 370 °C.

14 Biofuels has 3 generation in which the raw material used is different from generation to generation as follow: 1st
generation of biofuels used sugar cane and sugar beet, 2nd generation use wood and straw and finally, the 3rd
generation use algae.
15 These substances are added in order to improve different perimeters like combustion, lubrica tion, cleaning, etc.

George Samareanu Dissertation thesis, Pitesti 2 020
– 16 –

Figure 1 2 Oil refining and distil lation pro cess [3]
After heating, the oil is separated into fractions inside a c olumn , by a process called
fractional distillation16. This process will create different compounds like gasoline, LPG, diesel,
etc., based on the carbon numbe rs, boiling temperature as shown in table 3.
Table 3 Fractional distillation compounds
Product Carbon number17 Boiling tempe rature [°C]
Gas 1-4 0-40
Petrol (gasoline) 5-12 40-205
Kerosene 10-18 175-325
Diesel 12-20 250-350
Heavy fuels 20-70 370-600
Residue >70 >600

16 This process will vaporize the chemical compounds, by heating the crude oil, then, every fraction will condensate
inside the distillation column forming different compounds with various utilities.
17 Carbon numbers reflect the product obtained and as exxaple, for gases with carbon number between 1 and 4
could be: methane, ethane, butane, propane

George Samareanu Dissertation thesis, Pitesti 2 020
– 17 –
The fuel used in automotive industry enter a cycle in which the vehicle creates emissions,
then the emissions react with different compounds, then the chemical produced dissipate into the
atmosphere and finally build up high amount of tha t product.
In 2020, the challenges of car manufacturers regarding internal combustion engines
remained focused on fuel consumption, pollution as well as maintaining and improving
performance.
The perimeter with great concern today is, emissions and fuel economy. Those perimeter s
need to be correlated with the performance of the vehicle in order to gain a balance between
environmental protection and top performance.
This move ment , beside the E URO norm, pushed the car manufacturers to rese arch and
refine hybrid vehicles, as well as to encourage customers to buy 100% electric vehicles.
Within the emissions we find regulated emissions as well as emissions that are not
regulated, however, they affect the environment in the long term. This damage destroys the
natural microclimates, destroys the natural ozone layer, melts the glaciers, etc. All these effects
are due to the global warming caused by the greenhouse effect.
The greenhouse effect18 is a nature phenomenon, without which, the average temperature
would be -18°C . [8]
In figure 13 it is shown, the GHG e ffect and how it work s on our planet. Also, t able 4
presents the primary compounds that exist in Earth atmosphere ad contribute to the increase of
the greenhouse effect as well a s their lifetime.

Figure 13 Green House Gas effect [9]

18 Since the Industrial Revolution , mankind has increased the concentration of carbon dioxide, from 280 ppm in
1750 to 415 ppm in 2019 .

George Samareanu Dissertation thesis, Pitesti 2 020
– 18 –
Table 4 Main compounds of GHG and their lifetime [8]
Compound Lifetime [years] GWP / 100 years
CO 2 50 – 200 1
H2O < 2 8
CH 4 7 – 12 23
N2O 120 – 150 296
CFmCln 50 – 60 4600 – 14000
CH mFn 14 – 16 12 – 12000

In which:
 CO 2 – Carbon dioxide
 H2O – Water vapors
 CH 4 – Methane
 N2O – nitrous oxide
 CFmCln – Chlorofluorocarbons19
 CH mFn – Hydrofluorocarbons20

The regulated emissions are CO, NOx, PM, PN , and HC. Beside those, in day to day life of
an automobile, we have and create emissions from different sources, from simple to comp lex, as
shown below in figure 14 . In the same figure it is exemplified the efficiency of a vehicle based
on the losses that appear in day to day routine.

Figure 14 Vehicle emission and efficiency [10]

19 Chlorofluorocarbons (CFC) have been used widely as refrigerants R12 and they contribute to ozone depletion in
the upper atmosphere
20 Hydrofluorocarbons (HCFC ) also used as refrigerants R21 but they don’t harm the ozone layer. These compounds
contribute at global warming with a warming potential thousands of times greater than carbon dioxide

George Samareanu Dissertation thesis, Pitesti 2 020
– 19 –
The sources of air pollution by origin, are divided into natural and anthropogenic
(anthropogenic, artificial) sources. Anthropogenic sources (anth ropogenic / artif icial) are made
up of the whole human actions that affe ct the atmospheric environment. Human actions pollute
the air when due to their constituents normal atmosphere is added or subt racted, resulting in
alteration of the properties of the air in a human or environmental way. [9]
With all this constrains, it was necessary to think “how the car will work in the future ?”
and, for this we have different solution which are:
 LPG fuel21
 CNG fuel22
 Ethanol fuel23
 Traction b attery (electric vehicle)
 Hybrid vehicle
When designing and calibrating a vehicle engine, several axes are taken into
consideration such as: power and torque, fuel consumption, pollu tion, and not least the cost . This
"action" gives birth to a compromise th at the vehicle will have all its life. These compromise
represent finding a balance between fuel consumption, power and torque, pollution and cost; a
vehicle must be efficient (to be desired by the customer by producing the fun to drive effect),
protect th e environment as much as possible, be accessible at a price and not consume excess
fuel.
In order to achieve this balance, different strategies implemented in ECUs such as
injection cutting , regenerative braking, fuel mixture and many others have been crea ted. N ot only
can the software methods implemented reduce fuel consumption, but also the mechanical
strategies and modification as well . Methods for improving fue l consumption based on
mechanical strategies are presented in t able 5 below as well the estima ted savings in percentage .

21 When burning, LPG fuel release CO 2 but in smaller quantities than oil -based fuel and higher than CNG fuels. A
disadvantage of this type is the storage itself in which, LPG need to be stored under pressure.
22 CNG is less pollutant that gasoline or diesel fuels and, also emits lower carbon dioxide which will decrease GHG
effect. As LPG fuel, CNG need to be stored under pressure which will prevent spilling.
23 Fuel based on ethanol can have from 10% up to 85% ethanol combine with gasoline. Ethanol is a particulate -free
fuel type that combust with oxygen which will form carbon dioxide, carbon monoxide, water and aldehydes

George Samareanu Dissertation thesis, Pitesti 2 020
– 20 –
Table 5 Methods for fuel economy [8]
Fuel saving method Estimated fuel saving [%]
Downsizing of engine with turbocharger 10 – 15
Direct injection 10 – 13
Light weighting 10
Electric motor assist 7
Stop&Start with regenerative braking 7
Variable valve activation 5 – 7
Dual clutch transmission 4 – 5
Reducing mechaical friction 3 – 5
Improved aerodynamics 2 – 4

Due to these constraints, the internal combustion engine has undergone continuou s
improvement, as presented in a nnex 3, from the traditional form to that of its use in pHEV .
The main substances with harmful action, both on air quality and human health, from the
vehicle and related activities , are the following :
 Sulfur dioxide (SO 2)
 Nitrogen oxides (NO X)
 Volatile organic compounds (VOCs )24
 Carbon oxide (CO)
 Hydrogen sulphide (H 2S)
 Ammonia (NH 3)
 Carbon dioxide (CO 2)
Also, the pollution produces negative effects on the environment such as:

24 The most representative volatile organic compounds a re petroleum products: gasoline, petroleum ethers,
benzene, acetone, chloroform, esters, phenols, etc

George Samareanu Dissertation thesis, Pitesti 2 020
– 21 –
 Acid rain is formed when sulfur dioxide or nitrogen oxides are mixed in the
atmosphere with water molecules.
 Smog25 is formed when ozone, nitrogen oxides and volatile organic components react
in the presence of sunlight
 greenhouse effect
To limit pollution, software strategies are implemented such as injecti on cutting, regenerative
braking, etc. At the mechanical level, the presence of the catalyst, the particulate filter, and other
equipment such as oxygen sensor and NOx trap , allowed the compliance with the legislation in
force .
In table 6 below shows the effect of noble materials within TWC as well as their efficiency
on emissions.
Table 6 TWC materials and their efficiency [9]
HC CO NO x
Pt (pl atinum ) + + + + + +
Pd (palladium ) + + + + + + + +
Rh (r hodium ) + + + + + + +
+ + + High efficiency ; + + Medium efficiency ; + Low efficiency

25 This effect is also called p hotochemical smog and belongs to the modern industry. Classic smog is the result of
burning coal and is a mixture of smoke a nd sulfur dioxide. As an example, in Beijing China we have photochemical
smog while classic smog was produced in 1952.

George Samareanu Dissertation thesis, Pitesti 2 020
– 22 –
4. Experimental investigation
4.1 Test procedure
The development of the WLTC has been carried out under a program laun ched by the
World Forum for the Harmonization of Vehicle Regulations (WP.29) of the United Nations
Economi c Commission for Europe (UNECE) through the working party on pollution and energy
transport program (GRPE). The aim of this project was to develop a harmonized light duty test
cycle, that represents the average driving chara cteristics around the world and to have a
legislative world -wide -harmonized TA procedure put in place from 2017 onwards. [11]
The test procedure takes into account the conditions of the dynamometer, the gear
changes, the weight of the vehicle, the amb ient temperature, etc . The WLTC test is applied on
three classes26 of vehicles, differentiated by a variable called specific power (PWr) which
represents the power divided by the vehicle weight and measured in kW / kg, as follo ws:
 Class 1 – low power vehicles
 Class 2 – medium power vehicles
 Class 3 – high power vehicles
Today, passenger vehicles belong to class 3 having the specific power ratio between 34
and 100 W / kg respectively 0.034 and 0.1 kW / kg. Also in table 7 are pre sented the main
parameters for the WLTP test class 3. [12]
Tabel 7 WLTP class 3 parameters [12]
Low Medium High Extra high Total
Duration 589 433 455 323 1800
Stop duration 150 49 31 8 235
Dista nce 3095 4756 7162 8254 23266
% of stops 26.50% 11.10% 6.80% 2.20% 13.40%
Maximum speed 56.5 76.6 97.4 131.3 131.3
Average speed
without stops 25.3 44.5 60.7 94 53.5

The test procedure was carried out, both on chassis dynamometer and hardware in the
loop. For performing the test , the same engine and calibration was used, both for chassis
dynamometer and HIL, having the following specifications presented in the table 8.

26 Each vehicle class is divided into urban, suburban, rural and highway, with a division between urban and extra –
urban in proportion of 5 2% and 48% respectively.

George Samareanu Dissertation thesis, Pitesti 2 020
– 23 –
Table 8 Engine specification
Engine layout 4 stroke – straight 3
Fuel Gasoline
Production Year 2018 – present
Displacement 999 cm3
Engine type Aspirated
Power output 53 kW
Engine speed at max power 6300 rpm
Torque output 95 Nm
Engine speed at max torque 3500 rpm
Firing order 1–2–3
Injection type MPI
Engine block Aluminium
Cylinder bore 71 mm
Piston stroke 84 mm
Compression ratio 11.0
Valves 12 (4 per cylinder)
Vehicles using the engine Renault Clio V, Dacia Logan, Dacia
Sandero, Dacia Logan MCV

The difference occurred because on HIL it is impos sible to do the test manu ally (with the
help of a mouse) due to a lot of dead time between changes like: shifting the gears, accelerate,
braking, etc. In order to limit the errors and dead time, on HIL it was used an auto test.
The purpose of the test is to see what differences t here are between chassis dynamometer
and HIL, both in numerical value and in percentage. The signals of interest are average specific
consumption ( ASC ) for fuel measured in
, fuel injected mass ( FIM ) measured in
, gear
engaged, engine speed (N) measured in , torque (T) measured in and vehicle speed
(Vs) measured in
. This method with automatic test has some advantages over the t est that is
done manually like repeatability of the test and test during night.
Figure 15 shows the simplified automatic test procedure represented as flowchart and, in
annex 4 it is shown the interface for control desk used on HIL.

George Samareanu Dissertation thesis, Pitesti 2 020
– 24 –

Figure 15 Automatic test flowchart

George Samareanu Dissertation thesis, Pitesti 2 020
– 25 –
4.2 Equipment used
4.2.1 Chassis Dynamometer
A chassis dynamometer27 is a device used to test and develop a vehicle that uses a set of
rollers to simulate a real road in a controlled environment. This configuration allows the
measurement of vehicle emissions and fuel consumption influenced by various factors such as
traction force, forward resistance, aerodynamic coefficient, vehicle mass, etc.
A chassis dynamometer is a device that can simulate the resistance imposed to the wheel
of a vehi cle according to different driving cycles. [13]
Figure 1 6 shows schematically the general architecture of a test cell equipped with
chassis dynamometer .

Figure 16 Chassis dynamometer testing cell general architecture [8]
Different factors are taken into account when generating emissions such as the traffic
situation, driving style, environmental conditions, vehicle characteristics and emission control
technologies. To simulate real life conditions, from the test cell to the operating room, the
following parameters are sent and received: torque, roller speed, vehicle speed, vehicle
acceleration, traveled distance .

27 Beside measuring emissions, distance and fuel consumption, the chassis dynamometer can also measure
vibrations, noise and harshness.

George Samareanu Dissertation thesis, Pitesti 2 020
– 26 –
4.2.2 HIL environment
Hardware in the loop is a technique used in developing and testin g hardware and software
all together. At hardware level, we have connected to the bench the ECU, wiring and additional
tools. In the same way, for testing and developing the software, different tools are used like
Simulink® from Matlab, INCA from ETAS, CAN alyzer from Vector, DDT2000, etc.
Calibration engineers u se tools such as INCA (ETAS) or CANape (Vector) to connect to
a running ECU and to online tune its para meters during operation. Online calibration is based on
the XCP protocol. The simulation models generated by chip simulation implement that protocol.
Therefo re, INCA and CANape can both be connected to a chip simulation running on PC, to
onlinetune parameters. This way, calibration tasks can be moved from road and test rigs t o
highly available and ch eap PC platforms . [14]
The simple working principle for hardware in the loop is presented in the figure 17
below.

Figure 17 Simple working principles for HIL
The controller28 compares the signal from the sensors and the set po int of that signal and
then it will send an electrical input to the actuators. The actuators will transform the electric
signal into physical value and then the real time simulated system will evolve dynamically. The
job of sensors is to measure the signal from the RT Ss and the send the information to controller
which will repeat this cycle, called closed loop.

28 On HIL both ECU and TCU can be connected together or separately

George Samareanu Dissertation thesis, Pitesti 2 020
– 27 –
For software developing, testing and validation, alongside with HIL bench it is necessary
to have the following: ECU, wirings, CAN , virtual model a nd auxiliary tools. All of thi s are
presented in the figure 1 8 below which exemplify the working principle with connected
equipment .

Figure 18 HIL working principle with connected equipment [3]
in which:
 ECUbox29 allow to conn ect the controller to the bench
 HIL bench keep the assembly all together
 Model is made up with Simulink®, AMESim® and other tools and then the diagrams and
block are converted into C code.
 Loadrack have real components inside it based on the project
 CAN 30is used to connect different microcontrollers in order to communicate between
them
 ControlDesk is the graphic user interface in which the user interact with the bench
 DDT2000 is the diagnostic tool used by Renault
 CANalyser is used to send custom CAN messag es, analyze frames or messages send by
microcontrollers
 INCA is used to make modification to the variables and calibration inside the ECU
 Remote PC

29 In the past years, this too l was called Harness but, in order to increase productivity and help people it was
created the ECUbox
30 just like on the vehicle , beside CAN we can have LIN connection for actuators and sensors.

George Samareanu Dissertation thesis, Pitesti 2 020
– 28 –
The HIL bench architecture and ports are presente d in figure 19 .

Figure 19 HIL architecture and ports [3]
At the bottom we have the real time processor, also known as computer on rack, where
the model is running in real time. Above the RT processor is the I/O board that translate s the
input and output signals. The simulated batter power all of the systems with 14 volts just like on
the car. Signal conditioning is used to make electrical faults if the user want s to diagnose a
system. On LoadRack we have real components like injectors, valves, lights, etc. which can be
pulled out and replace based on the project.
Hypertonic connector is used to connect the ECU or TCU to the bench using ECU box.
On Break out boxes we have individual connector for each ECU signal in case we need to make
special interaction with the bench. Diagnostic con nector is used to connect DDT200 for OBD
tasks.
This whole assembly is powered by the main source of the bench and also this is the first
generation of HIL benches that was used. This connectivity evolves in time in order to keep up
with the industry.

George Samareanu Dissertation thesis, Pitesti 2 020
– 29 –
The block diagram created with Simulink is used inside the ECU after it is transform ed
into C code31. This technique to write codes helps increasing the productivity and ensure that
every strategy works correctly when interacts with other strategies . In figur e 20 it is shown how
a C and Simulink code are for the air fuel ratio formula.

Figure 20 Air fuel ratios represented as C code and Simulink diagram
Hardware in the loop was used alongside with automatic test to reproduce the data from
chassis dynamometer by following the curves for vehicle speed, engine speed and gear engaged.

31 The C code can be obtained by auto coding or manual coding. Auto coding is used when the strategies are well
known like: combustion, fuel cut, EGR, etc. Manual coding is used for strategies that are relative new in order to
ensure proper coding like: high voltage systems, battery management, etc.

George Samareanu Dissertation thesis, Pitesti 2 020
– 30 –
4.3 Results and discussion
To analyze the test results32, I chose to use the average values of the signals because in
the recordings, both at the roller bank and in the HIL simula tion, the operating points exceeded
1.2 million.
Based on vehicle speed, engine speed and gear engaged, the measured signals were
average specific consumption (AS C) in liters per hour and fuel injected mass (FIM) in
milligrams per stroke. Beside those, it was necessary to record also the outputted torque (T), in
order to be able to calculate various signals. Those measured signals are listed into table 9 below.
Table 9 Average value for measured signals
Average value for measured signals on Vehicle
Signal Average WLTP Urban Suburban Rural Highway
ASC 1.47 1.36 1.35 1.37 1.80
FIM 11.96 6.39 12.42 12.5 16.51
N 1798.91 1257.72 1471.44 1835.07 2631.39
Vs 52.33 19.2 40.05 57.27 92.8
Gear 3 2 3 4 5
T 42.21 22.72 42.91 44.62 58.59
Average value for measured signals on HIL
Signal Average WLTP Urban Suburban Rural Highway
ASC 1.59 1.51 1.49 1.46 1.85
FIM 12.70 6.36 13.81 13.51 17.11
N 1828.03 1284.66 1497.99 1856.8 2672.66
Vs 52.55 19.27 40.14 57.47 93.33
Gear 3 2 3 4 5
T 44.20 22.71 46.765 48.43 58.89

With the measured signals, the effective power, the average effective pressure, the
specific fuel consumption, the effective efficiency, the hourly fuel consumption and the fuel
consumption per 100 kilometers were calculated with the formulas from (1) to (6) below and
shown in table 10 .
A. Effective power (Pe)

(1)

32 The analysis of the test was performed using the program Measurement Data Analysis (MDA) from ETAS. Every
interval for the WLTP test, was put it with corresponding time interval at which the average value was calculated.
Also, in MDA, the average value for the whole WLTP test was calculated.

George Samareanu Dissertation thesis, Pitesti 2 020
– 31 –
where:
 n – engine speed in [rpm]
 T – outputted torque in [Nm]
 π – numerical constant

B. Brake Mean Effective Pressure (B MEP)

(2)
where:
 T – outputted torque in [Nm]
 π – numerical constant

C. Brake Specific Fuel Consumption (BSFC)

(3)
where:
 FIM – fuel injected mass in

 n – engine speed in [rpm]
 Pe – effective power in [kW]

D. Effective efficiency (η)

(4)
where:
 cp – calorific power for gasoline is 43500

 BSFC – brake specific fuel consumption in

E. Hourly fuel consumption (Ch)

[
] (5)

George Samareanu Dissertation thesis, Pitesti 2 020
– 32 –
F. Fuel consumption per 100 km ( )

(6)
where:
 Vs – vehicle speed in

Table 10 Average value for calculated s ignals
Average value for calculated signals on Vehicle
Signal Average WLTP Urban Suburban Rural Highway
C100km 6.97 6.70 7.30 6.41 7.49
P 8.58 2.99 6.61 8.57 16.14
BMEP 5.30 2.86 5.39 5.61 7.36
BSFC 325.01 324.43 331.68 321.02 322.91
η 25.47% 25.51% 24.95% 25.78% 25.63%
Ch 2.79 0.97 2.19 2.75 5.21
Average value for calculated signals on HIL
Signal Average WLTP Urban Suburban Rural Highway
C100km 7.46 6.78 8.25 6.98 7.84
P 9.07 3.06 7.34 9.42 16.48
BMEP 5.55 2.85 5.88 6.09 7.40
BSFC 327.66 320.9 2 337.13 319.66 332.94
η 25.27% 25.79% 24.55% 25.89% 24.86%
Ch 2.97 0.98 2.47 3.01 5.49

For the signals who exceeded the accepted error threshold, it is necessary a fine tunning
of the software on HIL benches in order to a strategy as real and functional as possible . Some
factors that influenced this exceeded, can be: modeling of forward resistances, running
coefficient, real weather conditions (conditions that are unpredictable), condition of the vehicle
and parts and subassemblies, etc.
For the care of the size to make the a ccepted error, it is necessary for finer calibration of
the thresholds and strategy. Some factors that influence this increase can be: modeling of
resistance s, the coefficient of road, real weather conditions (conditions that are unpredictable),
the condit ion of the vehicle and parts and subassemblies, etc.
However, the errors for WLTP sub cycles and WLTP whole test are below the accepted
error threshold as shown in figure 21 below .

George Samareanu Dissertation thesis, Pitesti 2 020
– 33 –

Figure 21 Experiment errors

George Samareanu Dissertation thesis, Pitesti 2 020
– 34 –
5. Conclusion and future work s
The results of the WLTP test performed on HIL versus the test performed at the roller
bench satisfy the conditions imposed on the errors, as the errors occurred are below the value
imposed in most cases. For signals that have exceeded the error threshold, this is beca use no
matter how much you want HIL to have a dynamic system such as in everyday life, small
differences in environmental conditions, roads, vehicle (weight, resistances, aerodynamic
coefficient , etc.) and other disturbing factors, change the obtained val ues.
Also, for the WLTP test divided into four cycles (urban, suburban, rural and highway)
their errors were below the maximum accepted value of 2%.
The HIL equipment can be used to see the driving scenario in the WLTP test and how the
signals inside the E CU evolve as well as to test various problems encountered in the next
iteration of the software, problems encountered by the customer (which is not intended to
happen). , strategies and ideas, etc. From this point of view, the HIL equipment is more suitabl e
on the validation side and not on the vehicle approval side (consumption and emissions).
Also, the use of HIL equipment allows it to come with some quite important advantages
such as:
 Is cheaper than a prototype vehicle
 The productivity is increased
 The development time is shorter
 Is more flexible than a real vehicle
 Suited for repetitive and automatic test
 Is precise for software validation
 Is safer than a real vehicle
The use of HIL also allowed the use of the Simulink® language for the realization of
strategies, a language that came in turn with different advantages such as:
 List of build -in libraries
 Custom libraries and blocks
 Simple interface
 Complex algorithms using stateflow
 Wide range of controllers
With the help of HIL, for future works, I will t ry to use it to validate the engine software used in
hybrid and electric vehicles as well as for the battery management system.

George Samareanu Dissertation thesis, Pitesti 2 020
– 35 –
6. Annexes
Annex 1: Abbreviations and notations

Acronym Description
KPI Key performance indicator
ICE Internal combustion en gine
ADAS Advance driving assistance system
HIL Hardware in the loop
WLTP Worldwide harmonized light -duty test procedure
ECU Electronic control unit
EMS Engine management system
Nox Nitrogen oxides
PM Particulate matter
PN Particle number
HC Hydro carbons unburned
CO Carbon monoxide
CO2 Carbon dioxide
OBD Onboard diagnostics
GPF Gasoline particle filter
GWP Global Warming Potential
AT Automatic transmission
GSI Gear shift indicator
RDE Real driving emissions
GHG Green Hous Gas
CAFÉ Corpora te average fuel economy
EEA European Environmental Agency
PLG Liquefied petroleum gas
CNG Compressed natural gas
Phev Plugin hybrid electric vehicle
twc Three way catalyst
pmi Multi point injection
tcu Transmission control unit
RTSs Real time simul ated system
CAN Controller area network
SED Small electric device
HEV Hybrid electric vehicle
BEV Battery electric vehicle
Edct Electronic dual clutch transmission
S&S Stop & Start
EGR Exhaust gas recirculation

George Samareanu Dissertation thesis, Pitesti 2 020
– 36 –
Annex 2: EURO norm values [15]

EURO 1 EURO 2 EURO 3 EURO 4 EURO 5 EURO 633 Diesel CO [g/km] 2.72 1.0 0.64 0.50 0.50 0.50
NOx [g/km] – – 0.50 0.25 0.180 0.080
HC+NOx
[g/km] 0.97 0.7 0.56 0.30 0.230 0.170
PM [g/km] 0.14 0.08 0.05 0.025 0.005 0.005
PN [g /km] 6×1011 6×1011 Gasoline CO [g/km] 2.72 2.2 2.3 1.0 1.0 1.0
NOx [g/km] 0.15 0.08 0.060 0.060
HC+NOx
[g/km] 0.97 0.5 0.20 0.10 0.10 0.10
PM [g/km] 0.005 0.005
PN [#/km] 6×1011

33 The EURO 6 norm is divided into three main parts known as: EURO 6b, EURO 6c, EURO 6d. For EURO 6d it is divided again into two other parts: Euro 6d temp
and EURO 6d full

George Samareanu Dissertation thesis, Pitesti 2 020
– 37 –
Annex 3: ICE roadmap [3]

George Samareanu Dissertation thesis, Pitesti 2 020
– 38 –
Annex 4: HIL control desk interface

George Samareanu Dissertation thesis, Pitesti 2 020
– 39 –
Annex 5: Aver age specific consumption

George Samareanu Dissertation thesis, Pitesti 2 020
– 40 –
Annex 6: Fuel injected mass

George Samareanu Dissertation thesis, Pitesti 2 020
– 41 –
Annex 7: Gear engaged

George Samareanu Dissertation thesis, Pitesti 2 020
– 42 –
Annex 8: Engine speed

George Samareanu Dissertation thesis, Pitesti 2 020
– 43 –
Annex 9: Outputted t orque

George Samareanu Dissertation thesis, Pitesti 2 020
– 44 –
Annex 10: Vehicle speed

George Samareanu Dissertation thesis, Pitesti 2 020
– 45 –
7. References
[1] S. Sheetal, S. Dorothi and D. Gupta, "Agile Processes and Methodologies: A Conceptual
Study," International Journal on Computer Science and Engineering, vol. 4, no. 5, pp. 892 -898,
2012.
[2] G. Kumar and P. K. Bhatia, "Impact of Agile Methodology on Software," International Journa l
of Computer Technology and Electronics Engineering, vol. 2, no. 4, pp. 46 -50, 2012.
[3] Renault intranet network, 2020.
[4] M. Schlager, W. Elmenreich and I. Wenzel, "Interface Design for Hardware -in-the-Loop
Simulation," in IEEE Symposium on Industrial Informa tics, Montreal, 2006.
[5] D. v. Wissel, J. -M. Quelin, J. -M. Basset and J. Serra, "Industrial use of HIL Engine
Management System validation," in IAV SAPC, Berlin, 2012.
[6] P. Kalita, “AUTOMOBILE EMISSION CONTROL AND ITS SERVICING”, Jorhat, Kalita,
Porag, 2013, page 35.
[7] E. E. Agency, "Average carbon dioxide emissions from new passenger cars,". [Online].
Available: https://www.eea.europa.eu/data -and-maps/daviz/average -emissions -for-new-cars-
5#tab -chart_. [Accessed 15 02 2020].
[8] A. Clenci, "Economicitatea autovehicu lelor si protectia mediului," course notes UPIT, Pitesti,
2017.
[9] I. Florian, D. LIȚĂ and A. BUȘOI, “Environmental problems in automotive engineering”,
Pitesti, MATRIX ROM, 2013, page 16, 30 and 127
[10] E. E. Agency, "Vehicle emissions and efficiency," 10 12 2 019. [Online]. Available:
https://www.eea.europa.eu/media/infographics/vehicle -emissions -and-efficiency -1/view.
[Accessed 8 3 2020].
[11] Biagio Ciuffo and all, "The development of the World -Wide Harmonized Test Procedure for
Light Duty Vehicles (WLTP) and the Pathway for its Implementation into the EU Legislation,"
Transportation Research Board (TRB), 11 01 2015. [Online]. Available:
https://www.researchgate.net/publication/274958503_The_development_of_the_World –
Wide_Harmonized_Test_Procedure_for_Light_Duty_Veh icles_WLTP_and_the_Pathway_for_
its_Implementation_into_the_EU_Legislation. [Accessed 12 04 2020].
[12] Wikipedia, "Worldwide Harmonised Light Vehicles Test Procedure," [Online]. Available:
https://en.wikipedia.org/wiki/Worldwide_Harmonised_Light_Vehicles_Test_P rocedure.
[Accessed 12 04 2020].
[13] Zhuo Yang and all, "An overview of chassis dynamometer in the testing of vehicle
emission," in MATEC Web of Conferences, Guangzhou, 2018.
[14] J. Mauss, "Chip simulation used to run automotive software on PC", in ERTS, Toulouse ,
2014.
[15] The Automobile Association , "European emission standards," [Online]. Available:
https://theaa.com/driving -advice/fuels -environment/euro -emissions -standards / [Accessed 5 1
2020].