Czech Technical University in Prague

Faculty of Mechanical Engineering

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

Department of Automotive, Combustion Engine and Railway Engineering
Study program: Master of Automotive Engineering Field of study: Advanced Powertrains

GASEOUS FUEL SUPPLY FOR A GAS ENGINE WITH A SCAVENGED PRE-CHAMBER

DIPLOMA THESIS

Author : Mr. Nishanth Nithyanandham
Supervisor : Ing. Ji?í Vávra, Ph.D.
Specialist : Ing. Zbyn?k Syrovátka
Year : 2018

Disclaimer

I hereby declare that this thesis work is my independent work and I have used only the documents listed in the attachments and references. It contains no materials previously published or written by another person.

………………………

Date

………………………………….

Mr. Nishanth Nithyanandham

Acknowledgment

First, I would like to express my gratitude to my main supervisor Ing. Ji?í Vávra, PhD. for all the guidance, support and patience throughout my master thesis. I am equally thankful to my co-supervisor Ing. Zbyn?k Syrovátka for giving guidance in GT-power with a friendly attitude and for timely remarks which made complex decisions easier. I would like to thank my entire family for their love, support and understanding throughout the duration of my master’s studies.

Author:

Mr. Nishanth Nithyanandham

Title:

Gaseous fuel supply for a gas engine with a scavenged pre-chamber

Study program:

Master of Automotive Engineering

Field of study:

Advanced Powertrains

Academic Year:

2016/2018

Department:

U 12120

Supervisor:

Ing. Ji?í Vávra, Ph.D.

Czech Technical University in Prague

Faculty of Mechanical Engineering

Technická 4, 166 07 Praha 6 – Dejvice, Czech Republic

Josef Bozek Research Centre for Vehicles of Sustainable Mobility

Scientific and Technical Park

P?ílepská 1920, 252 63 Roztoky u Praha

Specialist:

Ing. Zbyn?k Syrovátka

Czech Technical University in Prague

Faculty of Mechanical Engineering

Technická 4, 166 07 Praha 6 – Dejvice, Czech Republic

Josef Bozek Research Centre for Vehicles of Sustainable Mobility

Scientific and Technical Park

P?ílepská 1920, 252 63 Roztoky u Praha

Abstract

The aim of this thesis is to design and develop an additional gaseous fuel supply for a 4-cylinder 4-stroke gas engine with scavenged pre-chamber for a light duty truck. The work presented in this thesis focuses on elimination of pressure pulsation in the fuel supply and ensure proper fuel flow to each pre-chamber in the engine
The previous experimental work where conducted on a 4-cylinder engine of a light duty truck which was converted into a single cylinder engine with a modified cylinder head with a pre-chamber. A 4-cylinder model of the experimental single cylinder engine with scavenged pre-chamber was designed in GT-POWER with additional gaseous fuel supply to the pre-chambers. The model was calibrated to match experimental data of single cylinder engine scavenged pre-chamber. Following the observation, it was inferred that the 4-cylinder GT-POWER model was well calibrated. This model was simulated to obtain full engine performance map data which is a valuable input to setup the experimental 4-cylinder engine with scavenged pre-chamber in the future and to test designed additional fuel supply.
Based on a sensitivity analysis of the size of the designed Gas Rail in GT- POWER for range of engine operation modes, an optimum dimension for the gaseous fuel supply line was deduced. The results of the simulation from the GT- POWER model serve as a base for setting up the additional gaseous fuel supply for the experimental engine in the future to improve the overall functionality of the scavenged pre-chamber.
Keywords: Scavenged pre-chamber, Gas rail, Pressure pulse, Mass flow rate, Fuel supply system, GT-POWER.

Contents

Nomenclature 8
Chapter 1 – Introduction 10
1.1 Problems in transportation sector 10
1.2 Natural gas as an alternative fuel 11
1.3 Research motivation and scope 12
Chapter 2-Literature search 13
2.1 Emissions in SI engine (air/fuel equivalence) 13
2.2 Lean burn concept 14
2.3 Stratified charged concept 15
2.4 Pre-chamber concept 16
2.4.1 Turbulence generating torch cell 17
2.4.2 Pre-chamber stratified charge engine flame jet ignition with auxiliary fuel injector and with no scavenging 17
2.4.3 Pre-chamber stratified charge engine with pre-chamber inlet valve and auxiliary carburettor. 18
2.5 Turbocharged engine 19
Chapter 3-Experimental setup and Methodology 22
3.1 Experimental setup 22
3.1.1 Experimental engine 22
3.1.2 Pre-chamber design 24
3.1.3 Fuel supply line to the Pre-chamber 25
3.2 Methodology 25
3.2.1 1-D model 26
3.2.2 Combustion model in GT-POWER 26
Chapter 4-Design Part 28
4.1 Idea of the base design 28
4.1.1 Plenum 29
4.1.2 intake runner 29
4.1.3 Runners 30
4.2 Implementation of the base design 30
4.2.1 Female tee 30
4.2.2 Male hex nipple 31
4.2.3 Male adapter 32
4.2.4 Pressure sensor 32
4.2.5 Relief valve 33
4.2.6 Assembled CR Unit 33
Chapter 5 GT-POWER Part 35
5.1 4-cylinder model 35
5.2 Modification of the 4-cylinder model 36
5.2.1 Checking additional fuel supply system working and concept 37
5.2.2 Standard Flow Control 39
5.2.3 Lambda control 41
5.2.4 MEP controller 42
5.2.5 EGR controller 42
5.3 Model calibration and comparison 43
5.3.1 Model calibration 43
5.3.2 Model comparison 44
5.4 Full load model simulation and comparison 49
5.4.1 Full load simulation of the model 49
5.4.2 Model comparison 53
5.4.3 Full load simulation with EGR 55
Chapter 6-Common rail analysis 57
6.1 discretization length setting 58
6.2 Base CR analysis 61
6.3 CR dimension analysis 62
6.3.1 Runner length analysis 62
6.3.2 Runner cross-section analysis 64
6.3.3 Plenum volume analysis 66
Summary and conclusion 68
References 70
List of Figures 72
List of Tables 74
Attachment 78
Nomenclature
2D Two dimensional
? Thermal efficiency
3D Three dimensional
? Specific heat ratio
rc Compression ratio
wt Turbine specific work
wc Compressor specific work
cp Specific heat capacity
? Pi
CNG Compressed Natural gas
CO Carbon Monoxide
CO2 Carbon Dioxide
CR Common rail
CH4 Methane
DI Direct injection
ECU Electronic control unit
EGR Exhaust Gas Recirculation
GHG Greenhouse gas
h Height
HC Hydrocarbon
LNG Liquefied Natural Gas
MEP Mean effective pressure
NG Natural Gas
NO Nitric Monoxide
NO2 Nitrogen Dioxide
NOx Oxides of nitrogen
Pgas Actual pressure of the gas
P0 Standard pressure
PID Proportional integral derivative
r Radius
RPM Revolutions per minute
SI Spark ignition
Tgas Actual temperature of the gas
To Standard temperature
TDC Top dead centre
Vgas Cumulative volume into the pre-chamber per cycle
VGT Variable geometry turbine

Chapter 1-Introduction
1.1 Problems in transportation sector
The transportation system is considered as a dominant factor influencing the socio-economic modern society. But even though it is a big advantage, it poses several threats like energy resource depletion and its effect on human health and climate change. From the recently published data 1 plotted in Figure 1, it shows that CO2 is a primary GHG (Greenhouse gas) emitted. 81% of the GHG is CO2. Methane and NO (Nitrous Oxide) also comprises 10% and 6% of the GHG. These GHG apart from depleting the ozone layer, also causes Problems to human health 2.

Figure 1 GHG emission composition 1

Figure 2 CO2 emission source 1
The combustion of fossil fuels like diesel and gasoline by the automotive sectors are one of the main sources of CO2, NO and HC (Hydrocarbon). The Figure 1 shows that contribution of the automotive sector to the total GHG emissions is around 34 %. The main challenge of the automotive manufacturers is to build a more efficient, economic vehicle with relatively less emissions.
1.2 Natural gas as an alternative fuel
NG (Natural gas) is an odourless gaseous mixture of HC mainly made of CH4. Two forms of natural gas are used in vehicles: compressed natural gas (CNG) and liquified natural gas (LNG). CNG is produced by compressing natural gas to less than 1% of its volume at standard atmospheric pressure. LNG is natural gas in its liquid form. LNG is produced by purifying natural gas and super-cooling it to -260°F to turn it into a liquid 3.
The composition of the natural gas can vary at each deposit. Even the composition of the natural gas affects the performance and emission characteristics of the engine 5. The manufactures must be able to develop an engine which must have fuel flexible. In 6 the influence of natural gas composition on engine performance has been assessed at laboratory of the Josef Bozek Research centre for a 4-cylinder turbocharged SI engine. From the research 6, it can be noticed that addition of gaseous higher HC slightly increases the engine power and efficiency but worsens the knock resistance of the fuel and slightly increases the CO2 emissions. The addition of hydrogen worsens the knock resistance of the fuel, decreases the engine power and efficiency. But decreases the emissions of CO2. Whatever the composition of the natural gas it maybe, still it has lower emissions than the conventional fuels. Natural gas has higher octane number, which means it is more resistant to knock. So, the natural gas can be compressed more without knock. Thus, natural gas can achieve a greater thermal efficiency than the conventional when used in an engine 17.
When energy consumption, pollutants and GHG emissions of gasoline, diesel and natural gas were assessed, it can be observed that natural gas are more efficient than the other two 4. Natural gas in vehicles has proved to achieve low emissions and it has the potential to be very environmentally friendly. Energy consumption of transport industries and fuel prices of gasoline and diesel are going up every year and growing awareness in cutting GHG, natural gas can be a prospective fuel in the future

1.3 Research motivation and scope
When a NG engine runs in extremely lean homogenous mixture, it leads to low combustion temperature and much further reduction in NOx emissions (lean burn concept) 8. Additional air in the lean mixture increases the specific heat ratio which leads to increase in thermal efficiency 9. when operating in extremely lean mixture it leads to lower burning speed, high HC emissions and poor combustion stability.
Many researches have been working to come up with new strategies to increase burning speed of a NG SI engine. One such strategy is use a different kind of ignition system. One of the ignition concept is the scavenged pre-chamber concept.
Therefore, the motivation for the research presented is connected to work of my predecessor 8, 10 which emphasis the application of scavenged pre-chamber ignition system in light duty truck gas engine as an alternative to the conventional method.
Work in this thesis as follow
To give an overview on the essential back ground concepts and highlight the important studies that support the thesis motivation.
To elaborate about the experimental setup, experimental engine and the Methodology used to in this thesis.
To create a basic idea of additional the fuel supply and a way to implement it in real time. To produce the 3-D models and drawings of the concept.
to create a GT-power model of a scavenged pre-chamber gas engine with pre-chamber fuel supply system. Modify the Model, so that it is similar to the experimental setup. Calibrate the GT-power model with data from the experimental engine. Calculate the engine performance map for the GT-power model.
to test the fuel supply system using GT-power for pressure pulsation which affect the performance, to estimate the working range of the Supply system and to do sensitivity analysis on the dimensions of the fuel system.
Chapter 2-Literature search
This chapter is divided into five sections. The first section gives an idea about the emissions in SI engines. The next three sections highlight the concepts like lean burn, stratified charge and pre-chamber concept. The last section explains about turbocharged engine.
2.1 Emissions in SI engine (air/fuel equivalence)
Air excess is an important factor in SI engine emissions. The factors that affect emissions which are controlled by the air excess ratio are the oxygen concentration and the temperature. Figure 3 shows the effect of lambda on SI engine. It indicates that if the engine can be operated in sufficiently lean condition, the engine could get a good fuel consumption and control engine emissions. However, such a lean combustion approach requires a fast burning combustion. The Figure 3 shows the variation of brake fuel consumption, NO, and HC with air/fuel and fuel/air equivalence ratio.

Figure 3 variation break specific fuel consumption of NO and HC emissions with air/fuel and fuel/air equivalence ratio 16
2.2 Lean burn concept
It can be clearly seen from the above section that if the engine is operated in leaner than stoichiometric ratio, then there is a reduction in NOx emissions. However, the NO rises initially and then reduce drastically. Lean burn occurs when the air/fuel ration or lambda is greater than 1. It means that additional air is added to the mixture. Adding additional air in the mixture increases the specific heat ratio of the mixture which leads to increase in thermal efficiency from the equation (1) 10 9.
?=1-1/?r_c?^((?-1)) (1)
Where,
? = Thermal efficiency %
rc = Compression ratio –
? = Specific heat ratio –
. The lean burn reduces the peak cylinder temperature which reduces heat losses 8. NOx is produced mainly due to high temperatures. Therefore, low temperatures of the combustion due to lean mixture produces lower level of NOx emissions. Lean burn concept improves the fuel consumption of the engine. So, it allows the engine to run less throttled with same power output by reducing the throttling and pumping losses. In addition to the fuel consumption, it decreases the probability of knocking and operation of engine in high compression ratio possible which results in higher thermal efficiency. SI engines have a low engine efficiency in part load operation condition 17.
Despite the advantages of the learn burn concept, there are several problems in the use of lean burn concept. The drawbacks are low burning rate, lower flame speed, which results in increased combustion duration. The slow burning doesn’t allow complete combustion before the opening of the exhaust. Due to this fact the misfire limit is reached. Due to the misfire in the engine, there is increase in the level of cycle to cycle combustion variation in the engine. This obviously decreases the efficiency, increase torque variations and increases the HC emissions in the engine 9.
The drawbacks of low burning rate can be overcome by making the burning fast and stable. This is possible by creating high turbulence, high compression, turbocharging and using advanced combustion techniques 11.

2.3 Stratified charged concept
There are 2 conditions involved with respect to charge mixture during combustion, homogeneous and stratified charge mixture. These types of mixture are dependent on the time of fuel injection. The homogeneous charge is injected earlier during the intake stroke so there is enough time for proper mixing of fuel. This is done by either port fuel injection or DI. The amount of air going inside are controlled by using throttle or variable intake valve according to the fuel injected, to maintain the right air to fuel ratio. This adds up to the pumping losses of the system. To overcome the pumping losses, the fuel is injected in the combustion chamber late in the compression stroke to have a readily ignitable mixture near the spark plug and a leaner mixture in the remaining combustion chamber. This is known as the stratified charge engine. Due to direct injection during later stages of compression strokes, the knock problem is avoided. This makes this concept more fuel tolerant and can work with a wide range of fuel.
There are many types of stratified charged engines. Few stratified charged engines have the piston top in the shape of bowl. This bowl-in-piston increases the degree of air swirl created during intake stroke and compression stroke. This design enhances the mixing of air and fuel. The fuel is injected tangentially into the bowl during the later stage of compression strokes which propagates the downstream flame and quick burning of the fuel air mixture. Figure 4 shows the examples of two stratified charged engines used in commercial practice: Texaco controlled combustion system and the M.A.N FM system 9 11 18

Figure 4 two stratified charged engines used in commercial practice: Texaco controlled combustion system and the M.A.N FM system 9
2.4 Pre-chamber concept
An alternative stratified charge concept is the Pre-chamber concept. This concept of the pre-chamber was first presented and patented by Sir Harry Ricardo in 1918. In this type of system, there is a small chamber (pre-chamber) connected to the main chamber which has its own supply of fuel air mixture. The spark plug is placed in the pre-chamber. During the intake stroke the pre-chamber is filled with rich mixture and main chamber is filled with lean mixture. When the compression starts the lean mixture moves into the mixture in the pre-chamber and to form an ignitable, slightly rich mixture near the spark plug. When the ignition starts in the pre-chamber, the flame developed comes as a jet into the main-chamber igniting the lean mixture in the chamber thus being called jet ignition. This increases the lean mixture operation range of the engine. Its operational principle is shown in Figure 5 9

Figure 5 operational principle of pre-chamber 9
The advantage of this type concept can be seen in the emission levels when compared to conventional method. Figure 6,7,8 shows comparison of conventional and pre-chamber engine emissions. It can be observed that there is reduction in the emissions of NOx, CO and HC.

Figure 6 comparison of conventional SI engine and pre-chamber engine NOx emissions 24

Figure 7 comparison of conventional SI engine and pre-chamber engine HC emissions 24

Figure 8 comparison of conventional SI engine and pre-chamber engine CO emissions 24

There are many different types of pre-chamber concept. The main types are briefly explained in the succeeding section. There are three main types:
Turbulence generating torch cell
Pre-chamber stratified charge engine flame jet ignition with auxiliary fuel injector and with no scavenging
Pre-chamber stratified charge engine flame jet ignition with pre-chamber inlet valve and auxiliary carburettor.
2.4.1 Turbulence generating torch cell
This system doesn’t have separate intake system for the pre-chamber. So, scavenging and pre-chamber fuel metering system is not present in this type. The focus is to increase the initial flame growth rate. The pre-chamber and the main chamber is connected through orifice. The critical design parameters are the orifice, flow pattern into the pre-chamber during compression and location of the park plug in the pre-chamber. The main disadvantage of this system is that the pre-chamber is not scavenged, so burned gas remain in the pre-chamber and mix with the fresh mixture in the main chamber between cycles. The pre-chamber volume is usually from 1 to 20 percent of the clearance volume. The Figure 6 shows the turbulence generating torch cell model 9

Figure 9 the turbulence generating torch cell model 9

2.4.2 Pre-chamber stratified charge engine flame jet ignition with auxiliary fuel injector and with no scavenging
This system is like the turbulence generating torch cell. But the pre-chamber is enriched by addition of fuel with an auxiliary fuel injector during the time of sparking. The pre-chamber is not scavenged as seen in the Figure 7. The pre-chamber volume is usually 20 to 25 percent of the clearance volume 9.

Figure 10 pre-chamber stratified engine with auxiliary fuel injector with no pre-chamber scavenging 9
2.4.3 Pre-chamber stratified charge engine with pre-chamber inlet valve and auxiliary carburettor.
This system is same as pre-chamber stratified engine with auxiliary fuel injector. But instead of auxiliary fuel injector, this system is provided with an auxiliary intake valve and carburettor for the pre-chamber. It helps in scavenging in between combustion in this system. With the help of separate intake system for the pre-chamber, it makes it easy to feed the with rich fuel mixture while the main intake system fills the main chamber with lean mixture. This approach usually has volumes of 2 to 3 percent of the clearance volume. The Figure 8 shows a pre-chamber stratified charge engine with pre-chamber inlet valve and auxiliary carburettor 9.

Figure 11 pre-chamber stratified charge engine with pre-chamber inlet valve and auxiliary carburettor 9
This system is sub-divided into 2 types based on the orifice design. First, with one or more small orifices for deep jet penetration and faster burning process. Second with large orifice for lower velocity jet and slower burning 10. Figure 9 shows two different approaches to pre-chamber orifice design with the pre-chamber stratified charge carburated and scavenged engine: (a) one or more small orifice(s) for deep jet penetration and faster burning process; (b) large orifice for lower velocity jet and slower burn 9.

Figure 12 two different approaches to pre-chamber orifice design with the pre-chamber stratified charge carbureted and scavenged engine: (a) one or more small orifice(s) for deep jet penetration and faster burning process; (b) large orifice for lower velocity jet and slower burn 9
2.4.4 Latest pre-chamber concepts
MULTITORCH GmbH introduced pre-chamber spark plug with housing in 2012. These have longer service life when compared to normal spark plug types. The ignition electrode comprises multiple electrode arms which creates an electric arc. This reasons to the advantages of the pre-chamber spark plug. Figure 13 shows the cross-section of MULTITORCH GmbH pre-chamber spark plug with housing 23.

Figure 13 the cross-section of MULTITORCH GmbH pre-chamber spark plug with housing 23

The latest development of the pre-chamber ignition system is done by Mahle Powertrain in 2014 and they named it as Turbulent Jet Ignition (TJI) 24. Studies 25 shows that TJI system has stable operations with ignition energies lower than that of the conventional spark ignition system. Figure 14 shows the cross-section of Mahle powertrain TJI system 24.

Figure 14 the cross-section of Mahle powertrain TJI system 24
2.5 Turbocharged engine
Turbocharged engines have pressurised intake, which increases the mass flow rate of air. The increase of mass flow air increases the fuel flow rate. This leads to increase in the power output of the engine. The air is compressed using a compressor which is driven by the exhaust gases turbine
The ideal compressor and turbine follows an isentropic process which is adiabatic and reversible. The real process is obviously irreversible. Expressions of work is uses the simplified version of steady flow equation. Figure 10 shows the temperature/entropy diagram for turbo charger. The suffix ‘s’ denotes an isentropic process. The specific work of the compressor and turbine is given below in the equation (2) and (3). where wc and wt are the compressor and turbine specific work. cp is the specific heat capacity 11.
wc = cp(T2-T1) (2)
wt = cp(T3-T4) (3)

Figure 15 the temperature/entropy diagram for turbo charger 11
The compressor is run by the turbine. So, the mechanically efficiency of the turbocharger is the ratio of the work done by the compressor and the work done by the turbine. The mechanical efficiency is given in the equation (4). Where ?_m,is the mechanical efficiency of the turbocharger.
?_m=(m_12 c_p12 (T_2-T_1))/(m_34 c_p34 (T_3-T_4)) (4)

There are two types of Turbocharger based on the compressor design, positive displacement and non-positive displacement type. Positive displacement compressor includes roots, sliding vane, screw, reciprocating piston and Wankel types. Non-positive displacement compressor includes axial and radial flow compressor. The non-positive displacement compressors are different due to the internal flow nature and the magnitude of the rotational speed when compared to the positive displacement compressors 11.
A typical automotive engine uses the radial flow compressor and turbine in the turbocharger. The radial turbocharger can be fitted with VGT (Variable geometry turbine) to increase or decrease the boost pressure according to the requirements. The VGT improved low speed performance of the turbocharger. The simplest type of VGT is the variable scroll area turbine shown in the Figure 11. If the exhaust gas flow into the turbine is slow the scroll area is changed and gives higher gas velocities. The Figure 12 shows another type of VGT where the rotating nozzle vanes are used to change the angle to control the effective flow area of the exhaust gas thus changing the boost pressure in the turbocharger.

Figure 16 variable scroll area turbine scheme 19

Figure 17 variable nozzle vane turbine 20
Chapter 3-Experimental setup and Methodology
This chapter explains about the experimental setup for thesis and the Methodology. The experimental setup part of the chapter explains the experimental engine and how the system works. Second part of the Chapter explains the Method used to obtain the data from the experiments.
3.1 Experimental setup
The experimental setup was set in Czech technical university Lab in Josef Bozek Research Centre for Vehicles of Sustainable Mobility. The experimental setup has already been developed by the predecessors 68. All the modifications for the engine is made in-house.
3.1.1 Experimental engine
The engine in the setup is a 4-cylinder 4 stroke light duty truck natural gas engine. The engine has a bore and stroke of 102mm and 120 mm. The displacement volume of the cylinder is 3.92 dm3 with a compression volume of 80cm3. The engine is turbo charged with an intercooler and a compression ratio of 12. The Turbo charger model is CZ-C14 which has VGT controllable by the ECU. There is a central mixing unit for feeding the gaseous fuel into the compressor. It is possible to control the lambda using the conventional oxygen sensor which is integrated with the ECU. The throttling is controlled controlling the throttle valve actuated by the stepper motor which is controlled by the ECU. A capacitive ignition system (UNIMA TC+) allows independent varying of the spark timing. All the actuators and the sensors are coupled to the ECU developed by the predecessor 6 using a field programmable gate array as a platform. The engine is coupled to a DC dynamometer with which the engine can be loaded. The main parameters of the original SI engine arrangement are given in the Table 16 12

No of cylinders 4
Bore 102 mm
Stroke 120 mm
Displacement 3.92 dm3
Compression ratio 12:1
Maximum Power 125 KW @ 2400 – 2800 RPM
Maximum Torque 600 Nm @ 1500 – 1600 RPM
Charging VGT turbo charger with intercooler

Table 1 main parameters of the original SI engine arrangement 6

The experiments were performed by converting the Original engine into a single cylinder engine by closing the intake and exhaust runners with metal plates. Figure 13 shows the scheme of the test engine layout with redlines indicating the intake and exhaust runners closed with plates.

Figure 18 the scheme of the test engine layout 8

3.1.2 Pre-chamber design
The module of pre-chamber assembly was designed and manufactured by the predecessor 8 in a such a way that the size, geometry and number of orifices of pre-chamber can modified and experimented. The pre-chamber module is cylindrical in shape. The pre-chamber housing protects the gas pipeline inside, the spark plug and ball check valve. Standard spark plug (Brisk CR10YS) is used. The pre-chamber is installed with a uncooled AVL GH15D pressure transducer. The cylinder head of the engine was modified in such a that the pre-chamber module can be fitted and taken out. The pre-chamber module cross-section and cross-section of the cylinder head with pre-chamber module shown in Figure 14.

Figure 19 The pre-chamber module cross-section and cross-section of the cylinder head with pre-chamber module 21

Different pre-chamber geometry and jet nozzles were tested in the engine under steady state operation by the predecessor 8. The pre-chamber with the largest volume and greatest cross-section area of the nozzle showed the best efficiency and low HC and NOx emissions. The pre-chamber parameter and the nozzle arrangement is given in the Table 2. For experimental work in this thesis the pre-chamber was replaced by a pre-chamber with a bigger volume. The parameters of the pre-chamber and nozzle is given in the table and with more details in Attachment I.

Volume 1.92 cm3
Fraction of compression volume 2.4%
Number of holes 12
Hole diameter 1.2mm

Table 2 pre-chamber parameters and jet nozzle arrangement 8
Volume 4.1 cm3
Fraction of compression volume 5.12%
Number of holes 12
Hole diameter 1.2mm

3.1.3 Fuel supply line to the Pre-chamber
The additional fuel is supplied by CNG gas line in the laboratory with pressure going up to 200 bar in the line. The pressure from the gas line is reduced by a pressure regulator. Fuel flow to the pre-chamber is controlled and measured by a mass flow meter OMEGA FMA2610. The fuel then passes through a rotameter with which the average volumetric flow can be inspected. The fuel then goes through a damping vessel and finally into the pre-chamber through a check valve. The schematics of the additional fuel supply line to the pre-chamber can be seen in Attachment A.
3.2 Methodology
GT-POWER is an industry level one dimensional simulation software to model and analyse the performance of an engine.GT-POWER is part of the GT-SUITE family.GT-POWER is used in this thesis to obtain the values of the physical quantities which are not possible using direct measurement.
The Gt-power uses a 1-D approach to analyse the flow in the model according to the dimensions. The combination of 1-D and the real working process analysis is the best method for development of an engine related system. The combustion model used in Gt-POWER is based on the Wiebe function Wiebe function.
3.2.1 1-D Model
1-D modelling is a mathematical representation of a flow through a component and its dynamics. 1-D model describes the spatial variations of the thermodynamic properties along the component, which is very important to determine the overall performance of components like manifolds, pipes, ports etc. 1-D models takes into consideration only changes in one direction. 1-D model takes shorter calculation time and is inaccurate at places where there are mixing, swirling and combustion, which are highly complex.
It is considered there is a flow through a straight duct. The area of the straight duct is assumed to change over the length. The models use the mass momentum and energy conservation equation for the unsteady compressible flow. The assumptions taken are there is no external volume, no heat conduction and only wall friction is considered. The equation used to study the dynamic are given below 9.
Mass conservation equation,
??/?t=-??w/?x-?w/A ?A/?x (5)

Momentum conservation equation,
??w/?t=-???w?^2/?x-??/?x-??w?^2/A ?A/?x (6)

Energy conservation equation,
??e/?t=-???wh?_0/?x-??wh?_0/A ?A/?x (7)
Where, ‘?’ is density, ‘t’ is the time, ‘x’ is the length of the pipe, ‘A’ is the area of cross-section of the pipe, ‘w’ is the velocity of the flow, e is the specific energy and ‘h0’ is the enthalpy.
3.2.2 Combustion model in GT-POWER
The GT-POWER combustion model is to estimate the mass fraction burned as a function of engine crank angle using the Wiebe function. Here combustion works as a function of crank angle. The combustion in model is based on the Wiebe combustion curve which is a S-shaped curve. The mass fraction burned moves from 0 which means no mass is burnt which indicates the start of combustion and grows exponentially to reach 1, which means all mass is burnt reaching the end of combustion. The time taken from the start of combustion to the end of combustion in term of crank angle is the combustion duration 14. Wiebe function is given below in equation (8) 9.
x_b=1-exp?-a (?(?-??_0))/???^(m+1) (8)

Where ? is the crank angle, ?0 is the crank angle for the start of the combustion, ?? is the total combustion duration which from xb = 0 to xb =1 and a and m are adjustable factors which determine the shape of the Wiebe curve.
Major parameters in the Wiebe function are the combustion duration, start of the combustion and the exponential factors ‘a’ and ‘m’. From the equation (5), it is seen that the mass fraction burned never actually reaches one. To optimize the Wiebe curve and fit it with the real time engine combustion. The factor ‘m’ is defined as a function of CA10 and CA90, which are crank angle at which 10% and 90% fuel is burned. This is more accurate as it is hard to determine the star and end of the combustion exactly. The start of the combustion is taken as a function of ‘m’ and CA50 which is crank angle at which the 50% of the fuel is burned. The GT-POWER takes CA50, combustion duration between 10% and 90% fuel burned, and the Wiebe exponent from the user to simulate the combustion in terms of Wiebe function in the system 14.

Chapter 4-Design Part
There are some basic requirements the base additional fuel supply model must fulfil. The engine dimensions and the placement of the fuel line into the additional fuel supply are the main consideration taken to fix the basic dimensions of the base model. The additional fuel supply system must have a safety to act in case of very high pressure in the system and a device to measure the pressure inside the system. It must be designed using components from the instrumentations fittings manufacturers which are easily available. The chapter is divided into 2 main part, the first part gives the basic idea for the design and the second shows how the design is implemented by using mass manufactured instrumentation fittings and sensors with 3D Model and drawings of the additional fuel supply system are created using the PTC Creo software.
4.1 Idea of the base design
. This fuel supply system is based on the common rail (CR) concept. The CR has to be fitted with a pressure sensor to sense the pressure in the rail and with relief valve which releases the pressure from the system in case the threshold safety pressure level is reached. The CR consists of 3 main parts, which include plenum, runner and intake runner. The fuel comes into the plenum through the intake runner, plenum stores the fuel and then sends it to the pre-chamber through the runners. The basic layout of the CR is given below in the Figure 15 with the nomenclature for the CR in the following sections.

Figure 20 basic layout of the CR
4.1.1 Plenum
It is a storage device which is placed between the intake runner and the runners. The plenum volume is larger than the intake runner and the runners. The function of the plenum is to store the fuel and equalize the pressure because of irregular demand by the pre-chamber. The plenum helps in even distribution of the fuel to the chamber. The plenum is highlighted in red in Figure 16 of the basic layout of CR.

Figure 21 basic layout of the CR (plenum highlighted in red)
4.1.2 intake runner
It is the part that delivers the fuel in to the system. The fuel from the laboratory setup comes to the plenum area through the intake runner. The intake runner is highlighted in red in Figure 17 of the basic layout of CR.

Figure 22 basic layout of the CR (intake runner highlighted in red)

4.1.3 Runners
The runners are the part of the CR which delivers the fuel to the pre-chamber from the plenum area. The runner governs the equal distribution of fuel by eliminating the pressure pulsation created due to irregular demand by pre-chamber from reaching the plenum area. The runner is highlighted in red in Figure 18 of the base layout of the CR.

Figure 23 basic layout of the CR (runners highlighted in red)
4.2 Implementation of the base design
The idea of the base design is implemented using mass manufactured instrumentation fittings which are easily available, robust and cheap. The CR and the experimental engine is in the developing stage, so it is better to use instrumentation fittings because it can be removed and different size can be implemented easily. If each time the CR with different dimension has to be manufactured, it will cost a lot and consume more time.
4.2.1 Female tee
This part is a T-split placed in 5 places in the CR. Four female tees act as the junction point to connect the plenum and the runner. The remaining female tee was used to connect the plenum and the intake runner. The female threading was chosen as the pressure sensor and the relief valve which are going to be connected in two of the female tee has male threading. So, it will be easier to connect. The female tee which has the runners connected must be aligned to the pre-chamber additional fuel supply placement for better performance of the CR 22. The dimensions required where close to the female tee made by manufacturer Superlok. The component and model name female tee NPT thread and IFT-8N. This part is made up of the material SS316, steel. It can withstand a pressure up to 351bar which is well under the operating pressure the CR is going to experience. All dimensions and details of this part are presented in Attachment B. The female tee and 3-D design of the female tee can be seen in the Figure 19.

Figure 24 female tee (left) and 3-D design of the female tee (right)

4.2.2 Male hex nipple
This part is placed in 4 places. The main function of this part is to connect the female tee together. The male threading is used as the female tee has female threading. His component was chosen in such a way that the thread pattern is same as that in the female tee. The male hex nipple with the required dimensions was found in Superlok make. The component and model name Hex nipple male NPT Threads and IHN-8N. This part is made up of the material SS316, steel. It can withstand a pressure up to 530bar which is well under the CR operating pressure. All dimensions and details of this part are presented in Attachment C. The male hex nipple and 3-D design of male hex nipple can be seen in the Figure 20.

Figure 25 male hex nipple (left) and 3-D design of male hex nipple(right)

4.2.3 Male adapter
This part is placed in 5 places in the CR. 4 male adapter acts as a connection between plenum. The remaining male adapter was used to connect the plenum and the inlet from the laboratory fuel supply line. The part has male threading as it is compatible with the female threading of the female tee. The required dimension was found in Superlok make with component and model name male adapter and SMA 8-8N. This part is made up of the material SS316, steel. All dimensions and details of this part are presented in Attachment D. The male adapter and 3-D design of male adapter can be seen in the Figure 21.

Figure 26 male adapter (left) and 3-D design of male adapter (right)

4.2.4 Pressure sensor
Pressure sensor is used to measure the pressure in the CR. It is connected to one end of Female tee. The chosen pressure sensor has a male threading which is compatible with the female tee. The pressure range of the sensor is from 0 to 60 bar which is optimum range to sense the pressure in the CR. The measuring accuracy is from 0 to 100 mbar. It gives a standard output 2 wire output signal from 4 to 20 mA which is compatible with the experimental setup. The make of the pressure sensor is BD sensors and the model are DMP 331. All details of the pressure sensor can be seen in the Attachment E. The pressure sensor and 3-D design of pressure sensor can be seen in the Figure 22.

Figure 27 pressure sensor (left) and 3-D design of pressure sensor (right)
4.2.5 Relief valve
The relief valve helps to relieve the pressure in the CR when the pressure goes over the safety limit. If for example, the check valve breaks and there is flow of pressure from the pre-chamber into the CR, the relief valve protects the CR from breaking and causing damage. The relief valve is connected to one end of the CR to the female tee. The relief valve has male thread which is compatible with the female tee. The maximum working pressure and cracking pressure range are 20.6 bar and 0.69 to 17.2 bar. The make of the valve is Superlok and the valve model name is SRVL-MS-8N-8-Y-SS. All details of the relief valve can be seen in Attachment F. The relief valve and 3-D design of relief valve can be seen in the Figure 23.

Figure 28 The pressure sensor (left) and 3-D design of pressure sensor (right)
4.2.6 Assembled CR Unit
All parts mentioned above are assembled together. The five male adapters are fixed with perpendicular part of all the female tees. The female tees are connected in line with the help of four male hex nipple. The pressure sensor is fixed with assembly in one end and the relief valve is fixed with the assembly in the other end. Female tee with the male hex nipple which supply fuel to the pre-chamber must be aligned and inline. The length of the without the pressure sensor and relief valve is approximately 530 mm which is suitable to the engine dimensions. Each runner is almost aligned with the pre-chamber inlet. The volume of each part of the CR is given below:
Volume of the female tee = 19.8 cm3
Volume of the male hex nipple = 4.9 cm3
Volume of the male adapter = 3.5cm3
The plenum has a volume of 119.16 cm3. The runners have a length and diameter of 50.8mm and 9.4 mm. The total volume of the CR is calculated to be around 136 cm3.The Figure 24 shows the picture of a 3-D drawing of assembled CR and the drawing is attached in Attachment G. The red arrow in the Figure indicates the flow of fuel in to the CR from the laboratory fuel supply. The blue arrows indicate the flow of fuel to the pre-chambers.

Figure 29 3-D drawing of assembled CR indicating the flow of fuel in to the CR from the laboratory fuel supply (red arrow) and the flow of fuel to the pre-chambers (blue arrows)
Chapter 5 GT-POWER Part
A 1-D GT-POWER model of the single cylinder was given by the consultant which can be seen in the Attachment H. The settings, boundary conditions and dimensions of the model is tuned almost close to that of experimental engine by the predecessor 12. This chapter has the following parts:
make a 4-cylinder model with pre-chamber from the given base model. The CR is also added to the model with the designed dimension, layout and proper boundary conditions.
Exploring and understanding the model and making Certain modifications, for example PID controller are added to the model to make the calibration of the model easier.
The 4-cylinder model is calibrated with data from the experimental engine and then compared.
The Full load characteristics of the created model is compared with data from unmodified SI version of the experimental engine. Finally, the full performance curve for the model is plotted.
The final 4-cylinder model with the modifications described in this chapter can be seen in the attachment J for future reference.
5.1 4-cylinder model
The given single cylinder model was converted into a 4-cylinder model. All the prevailing dimensions of the model were changed, as it is same as the experimental engine. 3 new cylinders and pre-chambers where added to the model with initial state object, heat transfer object, flow object, combustion model and other characteristics which were same as that of the base model. The volume of the added pre-chambers was 4.1 cm3.The dimensions of the pre-chamber can be found in below in the Attachment I. The designed CR was added to the model with the dimensions that are mentioned in chapter 3. The supply to the CR is from the end environment which represents the supply line from the laboratory. The parameters that can be changed in the 4-cylinder model by the user for testing different operating setup can be seen in Table 3.

Parameter Unit Description
Pressure initial bar The initial pressure (absolute) for the experiment setup
Temperature initial K The initial temperature for the experiment setup
Fuel supply pressure bar The fuel pressure given to the CR from the end environment which the laboratory fuel supply line
Intake Lambda Fuel ratio in the intake manifold
CA-50 Pre-chamber degree The specified angle is the degree between TDC and typically the 50% combustion point of the Wiebe curve (for pre-chamber ).
CA-50 Main degree The specified angle is the degree between TDC and typically the 50% combustion point of the Wiebe curve (for Main chamber).
Burn duration pre-chamber degree Duration of the Wiebe combustion curve of the pre-chamber
Burn duration Main chamber degree Duration of the Wiebe combustion curve of the main chamber
Engine RPM RPM Engine speed
Throttle angle degree Throttle angle degree (0-closed to 90-fully open)
Turbine rack position Turbine Rack position (0-full boost and 1-no boost)

Table 3 The parameters that can be changed in the 4-cylinder model by the user
5.2 Modification of the 4-cylinder model
There are some criteria the model has to full fill to work like the experimental engine. This section’s aim is to make some modification in the model to full fill the criteria. The criteria as follows:
There must be a standard flow of 0.3 m3/h of fuel through the check valve to each pre-chamber at all operating case and conditions.
The lambda in the main chamber must be controllable by the user in case setup
There must be a provision to control the IMEP output of the Model and switch off or on this control according to user’s wish.
There must be a provision to control the exhaust gas recirculation percentage in the model
5.2.1 Checking additional fuel supply system working and concept
The first simulation was performed to study the sensitivity of the additional fuel flow into the pre-chamber, depending on the fuel pressure for different engine speed. The model was simulated at three speeds -1200, 1800, and 2400rpm at fuel pressure of 1.5, 1.7, 1.9 and 2.1 bar. The initial temperature and pressure is set as 298.15 K and 1 bar for all simulations in this thesis. The readings are taken from one of the check valve to the pre-chamber assuming that the flow through CR to all the check valves will be almost same, which will be analysed in the next chapter. The Model simulation case setup for additional fuel supply sensitivity is mentioned below in Table 4.
Parameter Unit value
Lambda 1
CA-50 Pre-chamber degree -10
CA-50 Main degree 10
Burn duration pre-chamber degree 15
Burn duration Main chamber degree 30
Fuel supply pressure Bar 1.5, 1.7, 1.9, 2.1
Engine RPM RPM 1200, 1800, 2400
Throttle angle degree 90 (full open)
Turbine rack position 1 (no boost)

Table 4 Model simulation case setup for additional fuel supply sensitivity
From the results of the above simulation, it was noticed that the mass flow into the pre-chamber is due to the pressure difference between the pre-chamber and CR fuel pressure which is given by the laboratory supply line. If the pressure in the pre-chamber is more than the fuel pressure, the flow from the pre-chamber into the CR is restricted by the one-way check valve installed in the system. Figure 25 presents the plot of the mass flow rate and fuel pressure against crank angle. It is noticed that When the pre-chamber pressure goes below the fuel pressure in the CR, the mass flow starts. It is evident that in the exhaust stroke, when the pressure in pre-chamber goes up a bit there is a sink in the mass flow rate.

Figure 30 plot of the mass flow rate and fuel pressure against crank angle
From the simulation, a graph is plotted with the volumetric flow rate to the pre-chamber against the crank angle in Figure 26. Keeping the engine speed constant and if the fuel pressure is raised, it can be noticed that the flow rate of increases. If the pressure is high in the CR, then eventually the difference is also high which creates a greater flow rate. The cumulative volume in to the pre-chamber decreases as the RPM increases when the fuel pressure is constant. This phenomenon is because when the engine speed is low, there is more time for the fuel to flow inside the pre-chamber. Therefore, there is an increase in the cumulative volume into the pre-chamber as the engine speed decreases.

Figure 31 volumetric flow rate to the pre-chamber against the crank angle
5.2.2 Standard Volumetric Flow Control
There must be a standard volumetric flow of 0.3 m3/h of gas through the check valve to the pre-chamber. Standard flow formula according to the experimental Setup. The formula for standard flow ‘Q’ for the fuel into the pre-chamber is given below in the equation (9).

Q=(3?.P?_GAS ?.V?_GAS ?.T?_GAS.RPM)/(T_0 P_0 ?10?^5 ) (9)
Where,
Pgas (bar) = The actual pressure of the gas
Vgas (cm3) = Cumulative volume into the pre-chamber per cycle
Tgas (K) = the actual temperature of the gas
RPM = revolutions per minute
To(K) = standard temperature
P0 (bar) = standard pressure
Figure 27 is plot between fuel pressure and standard flow for various RPM with required standard flow marked. When the engine speed is constant, it is noted that the fuel pressure increases the standard flow increases. This is supported by the fact that cumulative volume into the pre-chamber increases with fuel pressure as discussed above. When the fuel pressure is constant, it is evident in the graph that the standard flow decreases as the RPM increases.

Figure 32 plot between fuel pressure and standard volumetric flow for various RPM with required standard flow marked
From the results, The Standard volumetric flow to pre-chamber is not constant at all RPM range. The Fuel pressure must be continuously varied to maintain the standard flow to the pre-chamber. To vary the fuel pressure during the simulation of model to get the required standard flow, a PID controller was added to the model. The PID controller works with the following equations (10) (11) (12).
?dx?_1/dt=u (10)
?dx?_2/dt=?u-x?_2/t (11)
y=(K_P+K_D/t)u+K_I x_1-(K_D x_2)/t (12)
Where ‘KP’ is the proportional gain, ‘KI’ is the integral gain, ‘KD’ the derivative Gain, ‘t’ is the Derivative time constant, ‘y’ is the controller output, ‘u’ is the difference between the reference signal value and the input signal value and ‘x1’ and ‘x2’ are the initial state variables 13.
For Standard flow PID controller, the controller output is the fuel pressure value and the target for the input signal (reference signal) is the required standard flow value which is given by the user at the start of the simulation. The input signal value is developed by an RLT sensor. The RLT sensor gives the output developed by the equation (6). The cumulative volume into the pre-chamber per cycle for the equation is given by an integrator function implemented in the model. The proportional gain, integral gain and Derivative gain was manipulated to get steady output value close to the required value in a short time according to the manual in 13. The final values for proportional, integral gain and derivative gain are 0.1, 0.12 and 0.
The simulation is performed again with standard flow PID controller with the case setup in the Table 4 and with one additional input, standard flow required is 0.3 m3/h. The Figure 28 shows a plot of fuel pressure and the standard volumetric flow for different engine speeds. From the Figure 28 we can observe that the Standard flow PID control is working properly as the fuel pressure is continuously changed by the PID controller to maintain the required standard flow for different engine speeds.

Figure 33 plot of fuel pressure and the standard volumetric flow for different engine speeds
5.2.3 Lambda control
The Lambda in the main chamber must be controllable by the user in the model. In the above simulation, we can see that the cumulative volume per cycle is not the same in various engine speed. when the lambda in the intake is kept at 1 and additional fuel is added in the pre-chamber the fuel becomes richer. As the cumulative volume changes with engine speed, the lambda is not constant in various engine speed which can be observed in the Figure 29. The lambda in the main chamber is measured in the exhaust. Figure 29 shows the plots of cumulative volume per cycle and the lambda in the main chamber for various engine speed.

Figure 34 plot of cumulative volume per cycle and the lambda in the main chamber for various engine speeds

To control the lambda in the main chamber a lambda PID controller was added which worked on the equations (7)(8)(9). The proportional gain, integral gain and Derivative gain was manipulated to get steady output value close to the required value in a short time according to the manual in 13. The final values for proportional, integral gain and derivative gain are 0.3, 0.2 and 0. The controller output was fuel ratio to injector in the intake, the target for the input signal (reference signal) is the required lambda in the main chamber which is given by the user at the start of the simulation. The input signal value is developed by the moving area function implemented in the model that takes the average of the Lambda in the main chamber per cycle. The lambda PID controller changes the lambda in the intake by changing the fuel flow in injector to get the required Lambda in the main chamber. The case setup parameter ‘intake lambda’ is not required anymore, as only the lambda in the man chamber is only required. The case setup has new parameter ‘main chamber lambda’ which must be set by the user before the simulation to the required value.

5.2.4 MEP controller
If the Engine must be tested in different load conditions either the throttle position or the MEP (mean effective pressure) must be changed. If the throttling is increased the volumetric flow increases. Therefore, MEP and throttling are related by the equation (13) 9.
MEP= ?_v ?_f Q_hv ?_(a,i) F/A (13)
Where, ?_v is the volumetric efficiency, ?_f is the fuel conversion efficiency, Q_hv is the heating value of the fuel, ?_(a,i) is the density of the inlet mixture and F/A is the fuel air ratio.
In the case setup of the model the throttle position can be changed by the user. The model should be simulated several times to operate the engine in the demanded MEP by changing the throttle position manually. To avoid this long process a MEP PID controller with an on/off switch was incorporated which works on the principle equations (7)(8)(9). The proportional gain, integral gain and Derivative gain was manipulated to get steady output value close to the required value in a short time according to the manual in 13. The final values for proportional, integral gain and derivative gain are 1, 1.2 and 0. The controller output was the throttle angle, the target for the input signal (reference signal) is the demanded MEP which is given by the user at the start of the simulation. The input signal value is developed by the RLT sensor which gives the MEP of the cylinder. There is on/off switch installed so that it can be controlled by the user whether the MEP PID controller should be used or not. There are 2 additional case setup parameters, one is ‘MEP PID controller switch’ to switch the controller on (1) or off (0) and the second ‘Demanded MEP’ to specify the Required MEP. if the controller is on, the ‘demanded MEP’ must be set by the user and the throttle angle set by the user is of no use. As the MEP PID controller will alter the throttle angle to get the required MEP. If the controller is off, the demanded MEP need not be specified by the user. Instead the throttle angle must be set.
5.2.5 EGR controller
The EGR (exhaust gas recirculation) is a technique to control the SI engine NOx emissions by adding a fraction of exhaust gases to the engine intake system. This Exhaust gas act as a diluent to the intake mixture and reduces the peak temperature. To control the EGR in the model a Controller EGR valve was implemented in the model. The controller EGR valve controls the EGR by controlling the throttling angle or orifice diameter of the EGR connection. The desired EGR is given as an input by the user. The Controller EGR valve acts as a PID controller. It controls the throttling or orifice diameter of the EGR correction according to the EGR percent present in the throttle valve in the intake system until the target EGR specified is reached.

5.3 Model calibration and comparison
The experimental data from the single cylinder pre-chamber data was given by the consultant to Calibrate the model. The data was for constant RPM of 1800 and was for 3 different lambdas 1.048, 1.542, 1.994 with fully open throttle. The section aims in the calibration of the model with the experimental data and comparing the results.
5.3.1 Model calibration
The combustion model in the GT-POWER model is not same as the experimental engine. Some data like burn rate, residual fraction, trapping ration cannot be measured directly in the experimental model. So, to get the data from the experimental engine regarding the combustion model, TPA (Three pressure analysis) method was performed by my predecessors. To extract the data from the experimental setup a TPA model was created in the GT-POWER like the experimental setup and the experimental pressure data and heat release data from the experimental engine was given as an input. Reverse cycle analysis was done to calculate the burn rate. This step was iterated to get a steady value 15. The combustion model extracted from the experimental engine was given as an input in the created model. The Figure 30 shows a plot between CA10-90 and operating lambda for different engine speeds. The CA50 for the main chamber is 10 degrees and the Wiebe exponent for the main chamber 1. The input of CA10-90 for main chamber is given as a XYZ map in the model. The Figure 31 shows the plot between the burned fraction of the fuel and the operating lambda which is input in the model.

Figure 30 plot between CA10-90 and operating lambda for different engine speeds

Figure 31 plot between the burned fraction of the fuel and the operating lambda which is given as a input in the model
5.3.2 Model comparison
The aim was to compare created model and the experimental engine pressure and combustion characteristics. So, the model was operated with same parameters as that of the experimental engine. The model was simulated with 1800 RPM in different lambda. The throttle is full open in the experimental data. But, if the throttle is fully open in the model the MEP will much higher in the model than that in the experimental data, as experimental data is from a single cylinder engine and the created model is a 4-cylinder engine 9. So, the MEP of the experimental data in different lambda was given as the ‘demanded MEP’ to simulate the same load conditions in the created model. Three different cases were simulated, case 1 with main chamber lambda 1.048 and MEP 10.18 bar, case 2 with main chamber lambda 1.542 and MEP 7.3 bar and case 3 with main chamber lambda 1.994 and MEP 5.75 Bar. Main parameters of the case setup for model comparison simulation is listed in the Table 5.
Parameter Unit value
PID switch 1(on)
Demanded MEP Bar 10.18, 7.3, 5.75
Main chamber Lambda 1.048, 1.542, 1.994
Standard flow required m3/h 0.3
Engine RPM RPM 1800
Turbine rack position 1 (no boost)

Table 5 Main parameters of the case setup for model comparison simulation
The Figure 32 is a plot of the main chamber pressure against the crank angle for the case 1. It can be observed that the maximum main chamber pressure for experimental data is than less than the model, which is 60.48 bar and 62.38 bar. The maximum difference in main chamber pressure along the crank angle can be seen only at peak pressure points. When the engine is operated at lambda 1.048 the maximum pressure difference between the model and the experimental data is 1.9 bar which is around 3%, which is comparable considering it occurs only at the peak point.

Figure 32 plot of the main chamber pressure against the crank angle for the case 1
The Figure 33 shows a plot of the main chamber pressure against the crank angle for case 2. It can be observed that the maximum main chamber pressure for experimental data is less than that of the model, which is 47.9 bar and 52.14 bar. Same as the previous case, the maximum difference in main chamber pressure along the crank angle can be observed only at peak pressure points. When the engine is operated at lambda 1.542 the maximum pressure difference between the model and the experimental data is 4.24 bar which is around 9%, which is 3 times more difference that was calculated in case 1.

Figure 33 plot of the main chamber pressure against the crank angle for the case 2
.
The Figure 34 shows a plot of the main chamber pressure against the crank angle for case 2. It can be observed in case 3 that the maximum main chamber pressure for experimental data is less than that of the model, which is 38.48 bar and 40.68 bar. Like above cases, the maximum difference in main chamber pressure along the crank angle can be observed only at peak pressure points. When the engine is operated at lambda 1.994 the maximum pressure difference between the model and the experimental data is 2.2 bar which is around 6 %, which is lower than case 3 and higher than case 1.

Figure 34 plot of the main chamber pressure against the crank angle for the case 3
Some variations are observed in the peak pressures in all the cases. The variations in the peak pressure is not because of the variations lambda and MEP because it is observed to be less than 0.3 percent, which will not cause any noticeable changes in the peak pressures. When the combustion model was examined it was evident that there was some considerable difference between the model and experimental engine. Figure 35 presents plot of the CA50 and CA-10-90 against lambdas. The CA50 which gives information about start of combustion and CA10-90 shows the burned duration. These influence the peak pressure and the angle at which the peak pressure occurs 9. The Figure 35 shows that in case 1 there is a decrease in CA50 by 0.9 %, which means there is a change in start of combustion. Peak pressure variations in case 1 is lowest as it has least variation in CA 50 when compared to case 2 and case 3 which is around 4 % and 8 %. There is a bigger variation in case 2 when compared to the other 2 cases as it has considerable variations in both the CA50 and CA10-90.

Figure 35 plot of the CA50 and CA-10-90 in main chamber against lambda

From the above simulation it is understood that change in combustion in each case affect the pressure in the main chamber. The pressure difference between the chamber and CR pressure is one of the main criteria for additional fuel flow to the pre-chamber. The cylinder to cylinder pressure variations was also observed. The Figure 36,37and 38 compares the main chamber pressure in each cylinder against crank angle for case 1, 2 and 3. From the Figure it was observed the cylinder to cylinder pressure variations are within 0.8 percent change in all the cases.

Figure 36 the main chamber pressure in each cylinder against crank angle for case 1

Figure 37 the main chamber pressure in each cylinder against crank angle for case 2

Figure 38 the main chamber pressure in each cylinder against crank angle for case 3

From the simulation, it was found that there were some pressure and combustion variations in the chamber and some marginal cylinder to cylinder pressure variations. Since the thesis mainly focuses on developing a CR fuel system for the pre-chamber, these variations which were noticed in the above simulation are tolerable. It doesn’t affect the parameters considered to design the CR much. From the magnitude of the variations, the created model was considered optimally calibrated for further proceedings.
5.4 Full load model simulation and comparison
Based on the results from the previous simulation in section, the model is well calibrated with the experimental engine. Since the model represents created to represent a 4-cylinder setup, it would be pointless to compare the, torque, turbine, compressor, intake and exhaust properties with part load conditions. As the MEP PID controller was used in the previous simulation, the model’s full load operation wasn’t tested. There is a lot of pressure, temperature and flow difference in full, part and low load conditions in an engine 11. We get the maximum torque and power output form the engine in full load condition, this can be considered as the extreme operating case of the engine. To test the working of the designed CR system, it is recommended to simulate the engine in the full load condition because the temperatures, pressures and flow will be in a much higher order when compared to part load or low load conditions. The aim of this section:
Calculating the full load characteristics of the model while considering the physical limitations of the components in the experimental setup
Comparing the full load torque of the model with the 4-cylinder original engine.
Studying the effect of EGR on the full load torque curve of the model.
The simulation in this section was simulated with the following case setup parameters shown in the table.
Parameter Unit value
PID switch 0(off)
Throttle angle degree 90 (full open)
Main chamber Lambda 1
Standard flow required m3/h 0.3
Engine RPM RPM 800-2600

5.4.1 Full load simulation of the model
In the future, when the experimental engine is converted to 4-cylinder engine with pre-chamber like the model setup, the parts in experimental setup will experience much higher physical loads like high pressure, temperature, flow etc than the present experimental setup. The working parts of laboratory engine setup has some physical limits which must be taken into consideration. The experimental engine can withstand a maximum cylinder pressure of 120 bar. The Turbo charger will function without damage is up to 160000 RPM. The piping before the turbine has a thermal limitation of 800 C.
The VGT turbocharger is used in the experimental engine. The VGT helps to alter the turbocharger RPM for same engine speed and give a high compressor boost pressure. The VTG turbo charger has rack position from 0 to 1. The position 0 gives the maximum compressor boost pressure and 1 means there is no compressor boost pressure. The model was simulated in the full load condition by switching of the MEP PID controller and positioning the throttle valve in full open. Model was simulated 2 times, with maximum boost pressure position and minimum boost pressure of the VTG turbocharger. it was noticed that there was a noticeable difference in the output of the engine. The Figure 39 shows a plot between the full load torque output of the engine against engine speed for VTG rack positions 0 and 1. It is evident that the torque output of the engine is greater when operated in maximum compressor boost pressure in all RPM. When the compressor boost pressure is maximum, the amount of air/fuel going inside the engine is more compared to the amount at no compressor boost pressure condition. If the amount of air/fuel mixture going inside engine increases, the volumetric efficiency of the engine increases which gives more torque /power output 9. Maximum output of the engine can be exploited when it runs with maximum compressor boost pressure.

Figure 39 plot between the full load torque output against engine speed for VTG rack positions 0 and 1
The data from the above simulation, it is seen that when the engine is operated in maximum compressor boost pressure the physical limits of components is crossed at certain engine speeds. The torque and the maximum cylinder pressure are related. The cylinder pressure and compressor boost pressure relation are same as the phenomenon behind the torque and compressor boost pressure relation which is mentioned above. Figure 40 shows the maximum cylinder pressure against the engine speed with a red line indicating the maximum cylinder pressure the engine can withstand. The pressure is well within limit when the engine is operated in no compressor boost pressure in the simulated engine speed range. Even though the engine output during maximum compressor boost pressure is high, the maximum cylinder pressure goes across the limit in engine speeds above 1600 RPM.

Figure 40 the maximum cylinder pressure against the engine speed with a red line indicating the maximum cylinder pressure the engine can withstand
Turbo charger RPM vs engine speed with a red line marking the turbocharger RPM limit is plotted in Figure 41. The trend is same as the above result, the turbocharger RPM for no compressor boost pressure is well within the its physical limitations. The maximum cylinder pressure has a direct impact on the speed of the turbocharger RPM. As there is an increase in the maximum cylinder pressure, there will be a rise in the exhaust pressure. This which will increase the flow of exhaust gases through the turbine resulting in the increase in turbocharger RPM. Unlike the maximum cylinder pressure limit, the turbocharger RPM limit is reached when the engine speed goes over 1800 RPM during maximum compressor boost pressure operation.

Figure 41 Turbo charger RPM vs engine speed with a red line marking the turbocharger RPM limit
The temperature before turbine is also influenced by the maximum cylinder pressure, as the cylinder pressure increases the temperature of the exhaust gases also increases. Eventually, this rises the temperature before the turbine. The Figure 42 shows the temperature before turbine at different engine speed with thermal limit before turbine marked in red. the limit temperature before the turbine is crossed only when the engine speed goes over 2200 RPM during maximum compressor boost pressure operation.

Figure 42 the temperature before turbine at different engine speed with thermal limit before turbine marked in red
. Before proceeding further, it is important to examine the full load torque curve without crossing the physical limits of the components. Since it was inferred that the cylinder pressure influences both the turbocharger RPM and the temperature before the turbine. The maximum cylinder can be brought under the limits by retarding the ignition timing9. The maximum cylinder pressure was controlled from crossing the limits and the other 2 physical parameters were observed. The VTG rack position was tuned manually to operate with maximum compressor boost pressure possible within the physical barrier limits. After several iteration of the position of VTG, the maximum cylinder pressure was brought under the limit and it was noticed the other two parameters also came under the limits. The Figure 43 is a plot of full load torque and Tuned VTG rack positions for different engine speed.

Figure 43 plot of full load torque and Tuned VTG rack positions for different engine speed
5.4.2 Model comparison
The comparison of the model and a 4-cylider original engine is presented in the full load torque vs engine speed plot shown in Figure 44. The torque of the model is observed to be less in low engine speeds when compared 4-cylinder engine. However, the torque of the model increases more than the 4-cylinder engine when high engine speeds are reached. The above phenomenon is absurd as the torque output of the model must be more or closer to the data from the 4-cylinder engine without pre-chamber because when the combustion stability is better in pre-chamber engine, it gives more engine efficiency 15. As Combustion model and the pressure inside the chamber was calibrated in section 5.3.1, it was concluded that the low torque problem was based on the intake and exhaust system characteristics.

Figure 44 full load torque vs engine speed plot

When the intake and exhaust data of the above experiments where Examined it was found that some parameters of the model and the 4-cylinder original engine were not comparable. The turbocharger RPM for the model and engine is plotted for different engine speed in the Figure 45. It is observed that the Turbo RPM of the model is lesser than the 4-cylinder engine in all engine speed. If the turbocharger RPM is less, then the intake pressure will be less which will reduce the engine output. The turbo RPM can be less due to turbocharger efficiency difference or maybe because turbo charger efficiency mapping is not same as that of the turbocharger used in the original engine

Figure 45 turbocharger RPM plotted for different engine speed

To determine whether the change in turbo RPM is due to turbocharger efficiency difference because of the pressure and temperature of the exhaust gases entering the turbine, Plots between exhaust temperature ratios (temperature before turbine/temperature after turbine) across turbine and engine speed are shown in Figure 46. The relation between the turbocharger efficiency and the exhaust temperature and pressure can be determined by the equation (4) 11. It was observed from the Figure that there is notable difference in the temperature ratios across the turbines for the model and the original engine. It was minimized to check the exact reason for the turbocharger RPM difference. The difference in temperature ratios across the turbine was reduced by matching the exhaust temperatures before the turbine of the model and the 4-cylinder engine. This was achieved by iterating the model in different CA50 in the combustion model which determines the start of combustion. When the difference was reduced to negligible values, it was still evident there was difference in the turbocharger speeds. So, it was concluded that the efficiency mapping of turbocharger were not same for the model and the 4-cylinder engine.

Figure 46 Plots between exhaust temperature ratios (temperature before turbine/temperature after turbine) across turbine and engine speed

From the above explanation, if the difference in the torque is just due to the mismatched turbocharger efficiency mapping, Then the torque must be lower in all cases. But in engine speeds more 2000 RPM, it can be noticed torques in the model are higher than the 4-cylinder original engine. This can be accounted for the difference in the intercooler performance in the model and the original engine. The Figure shows a plot of the temperature ratios (temperature before intercooler/temperature after intercooler) across the intercooler in different engine speeds.

Figure 47 plot of the temperature ratios (temperature before intercooler/temperature after intercooler) across the intercooler in different engine speeds
From the analysis so far, it is apparent that the difference in torque of the model and the 4-cylinder engine is because of the different turbocharger efficiency mapping and dissimilar intercooler performance. For further proceedings, that is CR optimizing, the torque difference in low engine speed is not that Important. As the maximum conditions will be experienced by the CR at high engine speeds. The torque of the model is much greater than the 4-cylinder engine, so if the CR is optimized in the high engine speed range it is certain that it will work fine for the low engine speeds also.
5.4.3 Full load simulation with EGR
The model was simulated with same case setup as above with different EGR percentage. The exhaust gas mixes with the fresh fuel-air mixture. It acts as a diluent which reduces the peak temperatures in the cylinder. The reduction in peak temperatures reduce the NOx formation and exhaust temperature which is shown in the Figure 48 and 49. Increase in the EGR reduces the combustion rate and stability there by decreasing the combustion stability. The reduction in exhaust temperature affects the turbocharger efficiency from the equation (4). The reduction in turbocharger efficiency and combustion stability due to increase in EGR percentage brings down the performance of the engine. The Figure 50 shows the torque vs engine speed for different EGR %.

Figure 48 NOx emissions vs engine speed for different EGR %

Figure 49 Exhaust temperature vs engine speed for different EGR %

Figure 50 Torque vs engine speed for different EGR %
Chapter 6-Common rail analysis
When designing a CR for as 4-cylinder engine, taken into consideration the pulsed nature of the fuel flow. The system should not send pulse from separate cylinder into the common rail. This will lead to the irregularities in the fuel flow. When there is more than one chamber for which the fuel must be supplied, it is better to have a plenum area where the flow pulsation is reduced and damped. The pulsation in the CR is the major factor which affects the fuel flow. The parameter used to measure the effectiveness of the CR is the volumetric efficiency ?_v. The cumulative mass of air going inside the pre-chambers per cycle by the volume of the pre-chamber gives volumetric efficiency given below in equation (14) 9: –
?_v= m_a/(?_(a,i) V_d ) (14)
Where, m_a is the mass of inducted mixture, ?_(a,i) is the density of the mixture and V_d is the volume of the pre-chamber. As the volume of the pre-chamber is constant, the main factor influencing the CR efficiency is the cumulative volumetric flow to the pre-chamber.
There are two ways to calculate the frequency in the CR system by considering the system as an organ pipe or the Helmholtz resonator The Helmholtz resonator is shown in the Figure 11. And the frequency is given by equation (15) .
f_H=c/2? ?(A/lV) (15)
Where, f_H is the resonant frequency, c is the speed of sound, A is the pipe cross section area, l is the pipe length and V is the resonator volume.

Figure 51 Helmholtz resonator
The frequency for the organ pipe model is given by equation (17) (16).
f_p= c/4L (16)
L=l+0.3d (17)
Where, L is effective pipe length, l is pipe length and d is the pipe diameter. When the intake system is for a single cylinder engine it is better to use the organ pipe theory. When multi cylinder engine intake system is designed it is better to design the system based on the Helmholtz equation. Since the CR design is for the fuel supply to multiple chamber, the design optimization of the CR is based on the Helmholtz equation 11.
6.1 discretization length setting
The Discretization length is the length into which the CR will be subdivided for the calculations in the GT-power software. A discretization length of 10mm – 20mm is usually recommended for the modelling a fuel intake system 13. Smaller discretization lengths should be used to get higher resolution and will take much longer simulation time. For systems that doesn’t need much accuracy and needs faster run times, larger discretization length is used. The aim of this section is to analysis and find out which discretization length setting is optimum for further experiments in this chapter.
. The simulation is done for discretization lengths of 10mm, 15mm and 20mm respectively. The model is simulated in full load condition, with the tuned VTG setting and for an engine speed range of 800 – 2600 RPM. The details of the case setup of simulation for this experiment is given below in Table 6.
Parameter Unit value
MEP PID switch 0(off)
Standard flow required m3/h 0.3
Engine RPM RPM 800-2600
Turbine rack position Tuned setting
Exhaust Lambda 1
Throttle position degree 90 (full open)

Table 6 Main parameter for case setup of discretization length simulation
Figure 52. shows the plot of CR pressure vs crank angle different discretization length. From the plot we can see that the pulses for all the discretization length have a relatively low amplitude. Even though the pulse for length 20 mm have relatively low pulse the pressure wave is much coarser than the other two.

Figure 52 CR pressure vs crank angle for different discretization length
Comparing Figure 53, Figure 54 and Figure 55 which shows the average cumulative volumetric flow per cycle for the discretization lengths 10mm, 15mm and 20mm for engine speed points 800, 1800 and 2600 RPM. It is noted that the there is significant change in the volumetric flow in the 20mm discretization length model when compared to the other 2 models. The percentage of change of cumulative volumetric flow between the discretization length 20mm and 10 mm are around 0.04% ,0.6 and 2.8% for engine speeds 800, 1800 and 2600 RPM. But we can see that the change percentage between 10mm and 15mm discretization length for engine speeds 800, 1800 and 2600 RPM are just around 0.02%,0.03% and 0.3 %.

Figure 53 the average cumulative volumetric flow per cycle for the discretization lengths at engine speed of 800 RPM

Figure 54 the average cumulative volumetric flow per cycle for the discretization lengths at engine speed of 1800 RPM

Figure 55 the average cumulative volumetric flow per cycle for the discretization lengths at engine speed of 2600 RPM
From the analysis so far, it is evident that the readings between discretization length 10mm and 15mm are more accurate than the length 20mm. This is supported by the fact that the readings in 20mm length is much lower when compared to other 2 . The readings of 10mm and 15 mm are comparable, which means the values are converging and accurate. If the model is simulated with 10mm it is marginally accurate and takes a little longer run time than the simulations with 15mm. Therefore, a discretization length of 12mm is which gives considerable accuracy and relatively shorter run time is decided. All following experiments in this chapter is carried out with discretization length of 12 mm.

6.2 Base CR analysis
All the previous experiments where conducted with the base CR design. To study the effect of varying dimensions of the CR, the base CR design was set as a reference. The model was simulated with case setup in the Table 6 with discretization length of 12 mm for all succeeding experiments. The Average cumulative volumetric flow per pre-chamber was measured for different engine speeds in the Figure 56. From the plot, it was observed that the volumetric efficiency of the CR is lower decreases as the engine speed increases. It is known that the Cumulative volumetric flow and volumetric efficiency is related from the equation (11). As it was discussed above, the amount of volume flowing inside the chamber is greater if the engine speed is low, as there is more time for the fuel to flow.

Figure 56 Average cumulative volumetric flow per pre-chamber for different engine speeds

To analyse the CR design, the cylinder to cylinder variation in the flow is also an important factor. The pulsation in the CR not only alters the volumetric flow but has a influence on the cylinder to cylinder flow. The average percentage of variations in the flow to the cylinders for different engine speeds are plotted in the Figure. It was noted the flow cylinder variations are more at higher engine speeds. This trend can be pointed to the fact that higher engine speeds will had more disturbances in the CR pulsation.

Figure 57 average change in the chamber to chamber cumulative volumetric flow for different engine speed
From the results, it was clear that the two important parameters to optimize the CR are improving the average cumulative volumetric per cycle per cylinder and decreasing the cylinder to cylinder variations in the fuel flow. From the above discussed equation (12), it is possible to change the frequency in the pulse by varying the dimensions of the CR. Hence this motivated to study the effects of CR dimension optimizing on its performance
6.3 CR dimension analysis
The lengths from the inlet for the CR to the outlet to the chambers, the plenum volume and the dimensions of the runner are the design parameters which has major impact on the cylinder to cylinder flow variations. The runner dimensions play a role in the change in the pulse formation as given by the equation (12). The runner dimensions can alter the frequency of the pulse formation and time it the pulse takes to reach the plenum area
6.3.1 Runner length analysis
The aim was to investigate the effect of varying length of the runner and compare how it impacts the performance of the rail.3 different lengths were simulated. The lengths were increased by factors of 2,3 and 4 from the base CR. The 3 different CR was designated by names L1, L2 and L3 with L1 being the shortest and L3 being the longest. The runner lengths for each case is given below in the Table 7.
Case name Runner length (mm) CR volume (cm3)
Base CR 50.8 136.6
L1 100 153.8
L2 150 171.1
L3 200 188.3

Table 7 CR runner length and respective CR volume for different cases
With the data from the experiments, a plot of variation in the cumulative volumetric flow per for each case for different engine speeds was show in the Figure 58 . In All the Three cases, the has increase in the cumulative volumetric flow when compared to the base CR. From an engine speed range of 800 to 1200 RPM, the increase in the cumulative flow is the highest in L2 with around 1.4% increase and L3 being lowest with around 0.7%. As the engine speed moves from 1200 to 2200 there is a general decrease in volumetric flow increase trends. L2 has the lowest gradient fall in the increased percentage of volumetric flow. In high engine speed ranges, that is from 2200 to 2600 RPM, there is again a general increase in the volumetric flow increase percentage. The values are comparatively significant in high and low engine speeds.

Figure 58 Average Volumetric flow change when compared to the Base CR vs engine speed for different runner lengths

Average chamber to chamber variation change is plotted for different engine speed in Figure 59. The case L1, L2 and L3 have within 14% change, except for the case of L3 where there is 90% increase in cylinder to cylinder. In All the cases the cylinder to cylinder flow variations are higher than the variations in the Base CR except in 3 points, L1 at 1200Rpm and L2 and L3 at 1800 RPM.

Figure 59 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different runner lengths
In summary of the above data in this section, it can be concluded that there is an increase in the runner lengths increase the cumulative volumetric flow per cylinder per cycle but with minor increase in the chamber to chamber flow variations. The length in case L1 can be considered as a trade of between the increase in the cumulative volumetric flow and chamber to chamber flow variations. The optimal lengths depend on the case specific requirements of either increase in the cumulative volumetric flow or decreased chamber to chamber flow variations.
6.3.2 Runner cross-section analysis
The aim was to investigate the effect of varying cross-section of the runner by changing the diameter and compare how it impacts the performance of the CR. models with 3 different cross sections were simulated. The diameters of the runners were increased by factors of 2,3 and 4 from the base CR. The 3 different CR was designated by names A1, A2 and A3 with A1 being the smallest and L3 being the largest. The runner cross sections for each case is given below in the Table 8.
Case name Runner diameter (mm) Runner Cross section area (mm2) CR volume (cm3)
Base CR 9.4 69.39 136.6
A1 20 314.15 198.9
A2 30 706.8 298.6
A3 40 1256.6 438.6

Table 8 CR runner diameter, cross section and respective CR for different cases
With the data from the experiments, a plot of change in the average cumulative volumetric flow per chamber for each case for different engine speeds was shown the Figure 60. In only case A2 there was an increase in the cumulative volumetric flow when compared to the base CR at all engine speeds. At low engine speeds there is around 1.2% increase in cumulative volumetric flow. when the engine comes to mid-range engine speed there is a fall in cumulative volumetric. At high engine speeds, again there is increase in the cumulative volumetric flow.

Figure 60 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different runner diameters
Average chamber to chamber variation change is plotted for different engine speed in Figure 61. the chamber to chamber variation change is maximum 7%, except for the case of A1 where there is 13% increase in cylinder to cylinder variations al low engine speeds. Cases A2 and A3 has variations lower than that of the base CR at low and mid-engine speed range and at high speeds the increase in the variation change are less than 2%.

Figure 61 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different runner diameters
In summary of the above data in this section, it can be concluded that if there is an increase in the cross-section area of the runner, there is an increase in the cumulative volumetric flow per cylinder per cycle but minor decrease in the chamber to chamber flow variations. The cross-section area in case A3 can be considered as optimal option among the 3 cases as it is the only case where there was an increase in the cumulative volumetric flow when compared to the base CR at all engine speeds and with lower cylinder to cylinder volume flow variations at low and mid-engine speeds.
6.3.3 Plenum volume analysis
The aim was to investigate the effect of varying volume of the plenum by changing the diameter and compare how it impacts the performance of the CR. Models with 3 different volume were simulated. The diameters of the runners were increased by factors of 2,3 and 4 from the base CR. The 3 different CR was designated by names V1, V2 and V3 with V1 being the smallest and V3 being the largest. The Plenum volume for each case is given below in the Table 9 .
Case name Plenum volume (cm3) CR volume (cm3)
Base CR 119.6 136.6
V1 239.6 257.1
V2 359.6 377.1
V3 481.2 498.7

Table 9 CR plenum volume and respective CR volume for different cases
With the data from the experiments, a plot change in the average cumulative volumetric flow per chamber for each case for different engine speeds was shown the Figure 62. In low engine speeds there is a marginal increase in the average cumulative flow of around 0.6 % and in high engine speeds there is a marginal decrease in the flow of around 0.5 %.

Figure 62 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different plenum volumes
Average chamber to chamber variation is plotted for different engine speed in Figure 61. the chamber to chamber variation change is maximum 6%, except for the case of V1 where there is 13% increase alt low engine speeds. Case V3 has chamber to chamber variations lower than that of the base CR at low and mid-engine speed range and at high speeds the increase in the variations are less than 4%.

Figure 63 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different plenum volume
In summary of the above data in this section, it can be concluded that if there is an increase in the volume of the plenum, there is an increase in the cumulative volumetric flow per chamber and minor decrease in the chamber to chamber flow variations in low engine speeds. The plenum volume in V1 case can be considered as a trade of between the increase in the cumulative volumetric flow and chamber to chamber flow variations. The optimal plenum volume depends on the case specific requirements of either increase in the cumulative volumetric flow or decreased chamber to chamber flow variations.
Summary and conclusion
The main aim of the thesis was to support the development a pre-chamber engine as an alternative to the conventional engines to improve the performance, improve fuel efficiency and reduce emissions. The thesis had two main parts:
To create a 4-cylinder model with pre-chamber like the experimental setup.
To design and analyse the additional fuel supply system for the pre-chamber.
Studies were conducted on the created 4-cylinder GT-POWER model. The model showed some contradictions when compared with the experimental setup. Some improvements were made in the model for making it much comparable to the experimental setup. The model was calibrated based on the experimental data. The full load characteristic was analysed from the 4-cylinder model with a tuned VGT such that the values under the components physical limits. The full load curve of the model and the original 4-cylinder were compared. The dissimilarities in the character were reasoned. The effect of EGR in the model where studied.
The CR for additional fuel supply to the pre-chamber were designed with the instrumentation fittings. The CR design was added to the model. Simulations were performed to study the discretization length sensitivity on the pressure pulse and cumulative volumetric flow. The base design was tested to check the cumulative volumetric flow and chamber to chamber variations. Sensitivity analysis was performed on the CR dimensions to see how the cumulative flow in the rail changes.
Some key conclusions in this regarding the GT-POWER model and the CR fuel system are given below:
The cumulative volumetric and mass flow into the pre-chamber per cycle increases as the engine speed decreases
The cumulative volumetric and mass flow into the pre-chamber per cycle increases as the fuel pressure increases
The cumulative flow into the pre-chamber affects the lambda in the main chamber.
The created GT-POWER model without proper turbocharger mapping and intercooler performance as the original 4-cylinder will not produce the same full load torque curve.
The EGR decreases the performance of the pre-chamber engine. But leads to much better NOx emissions and fuel economy.
The CR designed is the best possible option for this experimental setup, if the CR must be designed from instrumentation fittings from mass producers. If the CR needs to br optimized, some dimensional changes can be done and should be manufactured in-house.
Contribution of this thesis for the scavenged pre-chamber experimental setup:
User friendly 4-cylinder GT-power model of the pre-chamber engine which is same as the experimental setup.
Tuned VGT rack position setting data to run the to be developed 4-cylinder pre-chamber version of the model in the physical limits of the components in the experimental setup.
3-D and manufacturable design of the CR with relatively low-pressure pulse for the to be developed 4-cylinder pre-chamber version of the model.
The list of the testing range and operational condition of the common rail.
References
United States Environmental Protection Agency.
Available at:
The Effect of Vehicular Emissions on Human Health by Ronni Esther Rossman
Available at:
Available at:

Anne Prieur., Richard Tilagone., “A Detailed Well to Wheel Analysis of CNG Compared to Diesel Oil and Gasoline for the French and the European Markets” SAE technical paper No. 2007-01-0037

Toshiyuki Suga., Tadashi Muraishi., Thomas Brachmann., Fumio Yatabe., “Potential of a Natural Gas Vehicle as EEV” SAE technical paper No. 2000-01-186

Vavra, J., Takats, M., Klir, V., and Skarohlid, M., “Influence of Natural Gas Composition on Turbocharged Stoichiometric SI Engine Performance,” SAE Technical Paper 2012-01-1647, 2012, doi:10.4271/2012-01-1647.

Li, G., et al.,”Experimental Study on the Ignition and Combustion Mechanisms of a Methane–Air Mixture in a Divided Constant-Volume Combustion Chamber. Energy ; Fuels”, 2012. 26(8): p. 4696-4702.

Vávra J, Syrovátka Z, Takáts M, Barrientos E. “Scavenged Pre-Chamber on a Gas Engine for Light Duty Truck.” ASME 2016 Internal Combustion Engine Division Fall Technical Conference: V001T03A014. doi:10.1115/ICEF2016-9423.

Heywood, J., B., „Internal Combustion Engine Fundamentals.” New York: McGraw-Hill, 1988. Print.

Baumgartner, L., Wohlgemuth, S., Zirngibl, S., and Wachtmeister, G., “Investigation of a Methane Scavenged Pre-chamber for Increased Efficiency of a Lean-Burn Natural Gas Engine for Automotive Applications,” SAE Int. J. Engines 8(2):921-933, 2015, doi:10.4271/2015-01-0866.

Ricardo, H.R., Internal-combustion engine. 1918, Google Patents.

Zbynek Syrovatka, Michal Takats, and Jiri Vavra “Analysis of Scavenged Pre-Chamber for Light Duty Truck Gas Engine” SAE Technical Paper 2017-24-0095, 2017, doi:10.4271/2017-24-0095.

GT-Suite manual

Yeliana, C. Cooney, J. Worm, D. Michalek, J. Naber “WIEBE FUNCTION PARAMETER DETERMINATION FOR MASS FRACTION BURN CALCULATION IN AN ETHANOL-GASOLINE FUELLED SI ENGINE” Journal of KONES Powertrain and Transport, Vol. 15, No. 3 2008

Ashish Shah ‘Improving the Efficiency of Gas Engines using Pre-chamber Ignition’ Lund University

Emissions Control: In-Cylinder

Siti Khalijah Mazlan “Study of Direct Injection and Pre-Chamber Application in Light Duty Gaseous Fuel Engines” School of Engineering, College of Science Engineering and Health, RMIT University

ANDERS N. JOHANSSON “Challenges and Advantages of Stratified Combustion in Gasoline Direct-Injected Engines” CHALMERS UNIVERSITY OF TECHNOLOGY Department of Applied Mechanics

https://www.dieselnet.com/tech/air_turbo_vgt.php

http://faq.outclub.ru/download/pajero_iv/maintenance/Service_Manual_2008_2013/2013/15/html/M215000900040200ENG.HTM

Š?OURAL, Martin. DP 2014 – SM06: Zapalovací kom?rka pro plynový motors nep?ímým zážehem. Praha 2014. Diplomová práce. ?VUT v Praze. Vedoucí práce Ing. Ji?í Vávra, PhD.

Amit Kumar Gupta Abhishek Mishra “Design and Development of Inlet Manifold for Six Cylinder Engine for Truck Application” ISSN – 2250-1991
https://www.researchgate.net/publication/274925435_Design_and_Development_of_Inlet_Manifold_for_Six_Cylinder_Engine_for_Truck_Application
United States Patent number: US8324792B2 https://patents.google.com/patent/US8324792
United States Patent number: US8857405B2 https://patents.google.com/patent/US8857405#patentCitations
Fraser-N. Parsons P. Attard, W. and E. Toulson. A turbulent jet ignition pre-chamber combustion system for large fuel economy improvements in a modern vehicle powertrain. SAE Int. J. Engines 3(2):20-37, 2010. doi:10.4271/2010-01-1457

List of Figures
Figure 1 GHG emission composition 1 10
Figure 2 CO2 emission source 1 10
Figure 3 variation of NO, HC and CO emissions with air/fuel and fuel/air equivalence ratio 16 13
Figure 4 two stratified charged engines used in commercial practice: Texaco controlled combustion system and the M.A.N FM system 9 15
Figure 5 operational principle of pre-chamber 9 16
Figure 6 The Figure shows the turbulence generating torch cell model 9 17
Figure 7 pre-chamber stratified engine with auxiliary fuel injector with no pre-chamber scavenging 9 18
Figure 8 pre-chamber stratified charge engine with pre-chamber inlet valve and auxiliary carburettor 9 18
Figure 9 two different approaches to pre-chamber orifice design with the pre-chamber stratified charge carbureted and scavenged engine: (a) one or more small orifice(s) for deep jet penetration and faster burning process; (b) large orifice for lower velocity jet and slower burn 9 19
Figure 10 the temperature/entropy diagram for turbo charger 11 20
Figure 11 variable scroll area turbine scheme 19 21
Figure 12 variable nozzle vane turbine 20 21
Figure 13 the scheme of the test engine layout 8 23
Figure 14 The pre-chamber module cross-section and cross-section of the cylinder head with pre-chamber module 21 24
Figure 15 basic layout of the CR 28
Figure 16 basic layout of the CR (plenum highlighted in red) 29
Figure 17 basic layout of the CR (intake runner highlighted in red) 29
Figure 18 basic layout of the CR (runners highlighted in red) 30
Figure 19 female tee (left) and 3-D design of the female tee (right) 31
Figure 20 male hex nipple (left) and 3-D design of male hex nipple(right) 31
Figure 21 male adapter (left) and 3-D design of male adapter (right) 32
Figure 22 pressure sensor (left) and 3-D design of pressure sensor (right) 33
Figure 23 The pressure sensor (left) and 3-D design of pressure sensor (right) 33
Figure 24 3-D drawing of assembled CR indicating the flow of fuel in to the CR from the laboratory fuel supply (red arrow) and the flow of fuel to the pre-chambers (blue arrows) 34
Figure 25 plot of the mass flow rate and fuel pressure against crank angle 38
Figure 26 volumetric flow rate to the pre-chamber against the crank angle 38
Figure 27 plot between fuel pressure and standard flow for various RPM with required standard flow marked 39
Figure 28 plot of fuel pressure and the standard flow for different engine speeds 40
Figure 29 plot of cumulative volume per cycle and the lambda in the main chamber for various engine speeds 41
Figure 30 plot between CA10-90 and operating lambda for different engine speeds 43
Figure 31 plot between the burned fraction of the fuel and the operating lambda which is given as a input in the model 44
Figure 32 plot of the main chamber pressure against the crank angle for the case 1 45
Figure 33 plot of the main chamber pressure against the crank angle for the case 2 46
Figure 34 plot of the main chamber pressure against the crank angle for the case 3 46
Figure 35 plot of the CA50 and CA-10-90 in main chamber against lambda 47
Figure 36 the main chamber pressure in each cylinder against crank angle for case 1 48
Figure 37 the main chamber pressure in each cylinder against crank angle for case 2 48
Figure 38 the main chamber pressure in each cylinder against crank angle for case 3 48
Figure 39 plot between the full load torque output against engine speed for VTG rack positions 0 and 1 50
Figure 40 the maximum cylinder pressure against the engine speed with a red line indicating the maximum cylinder pressure the engine can withstand 51
Figure 41 Turbo charger RPM vs engine speed with a red line marking the turbocharger RPM limit 51
Figure 42 the temperature before turbine at different engine speed with thermal limit before turbine marked in red 52
Figure 43 plot of full load torque and Tuned VTG rack positions for different engine speed 52
Figure 44 full load torque vs engine speed plot 53
Figure 45 turbocharger RPM plotted for different engine speed 53
Figure 46 Plots between exhaust temperature ratios (temperature before turbine/temperature after turbine) across turbine and engine speed 54
Figure 47 plot of the temperature ratios (temperature before intercooler/temperature after intercooler) across the intercooler in different engine speeds 55
Figure 48 NOx emissions vs engine speed for different EGR % 56
Figure 49 Exhaust temperature vs engine speed for different EGR % 56
Figure 50 Torque vs engine speed for different EGR % 56
Figure 51 Helmholtz resonator 57
Figure 52 CR pressure vs crank angle for different discretization length 59
Figure 53 the average cumulative volumetric flow per cycle for the discretization lengths at engine speed of 800 RPM 59
Figure 54 the average cumulative volumetric flow per cycle for the discretization lengths at engine speed of 1800 RPM 60
Figure 55 the average cumulative volumetric flow per cycle for the discretization lengths at engine speed of 2600 RPM 60
Figure 56 Average cumulative volumetric flow per pre-chamber for different engine speeds 61
Figure 57 average change in the chamber to chamber cumulative volumetric flow for different engine speed 62
Figure 58 Average Volumetric flow change when compared to the Base CR vs engine speed for different runner lengths 63
Figure 59 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different runner lengths 64
Figure 60 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different runner diameters 65
Figure 61 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different runner diameters 65
Figure 62 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different plenum volumes 66
Figure 63 Average Chamber to chamber volumetric flow change compared to Base CR vs engine speed for different plenum volume 67
List of Tables
Table 1 main parameters of the original SI engine arrangement 6 23
Table 2 pre-chamber parameters and jet nozzle arrangement 8 25
Table 3 The parameters that can be changed in the 4-cylinder model by the user 36
Table 4 Model simulation case setup for additional fuel supply sensitivity 37
Table 5 Main parameters of the case setup for model comparison simulation 44
Table 6 Main parameter for case setup of discretization length simulation 58
Table 7 CR runner length and respective CR volume for different cases 63
Table 8 CR runner diameter, cross section and respective CR for different cases 64
Table 9 CR plenum volume and respective CR volume for different cases 66
Attachment
Attachment A

Attachment B

Attachment C

Attachment D

Attachment E
Attachment F

Attachment G

Attachment H

Attachment I

Attachment J

x

Hi!
I'm Katy

Would you like to get a custom essay? How about receiving a customized one?

Check it out