Doctoral Thesis

Ultra-Lean Combustion of SI Engine Using

Alternative Stratification Principles

September 2020

Doctoral Program in

Advanced Mechanical Engineering and Robotics

Graduate School of Science and Engineering

Ritsumeikan University

NYAMSUREN Gombosuren
 Doctoral Thesis Reviewed
by Ritsumeikan University

Ultra-Lean Combustion of SI Engine Using

Alternative Stratification Principles
（新しい成層化原理を用いた
火花点火式エンジンの超希薄燃焼）
September 2020
2020年9月

Doctoral Program in
Advanced Mechanical Engineering and Robotics
Graduate School of Science and Engineering
Ritsumeikan University
立命館大学大学院理工学研究科
機械システム専攻博士課程後期課程

NYAMSUREN Gombosuren
ニャムスレン ゴンボスレン

Supervisor：Professor OGAMI Yoshifumi

研究指導教員：大上 芳文 教授

1
 Ultra-Lean Combustion of SI engine using
Alternative Stratification Principles

Department of Mechanical Engineering, Ritsumeikan University;

gr048072@ed.ritsumei.ac.jp

Abstract:
The demand for internal combustion engines remains high for mobile power sources in

all fields owing to low costs, running distance capacity, charging reliability, and vehicle

durability. Used vehicles with port-injected gasoline engines will continue to provide most

of our transportation needs until all vehicles become electric.

On the other hand, air pollution, efficiency, and environmental factors make this more

challenging. According to recent research, using a fueled prechamber can serve as an

alternative for lean combustion in the main chamber, resulting in increased efficiency,

reduced fuel consumption, and fewer toxic emissions.

However, there are challenges in developing a fueled prechamber for commercial

engines, including mixture and soot formation in the limited prechamber space. Moreover,

the charging possibility of unfueled prechambers remains unclear.

Therefore, I propose the use of an unfueled prechamber to enhance the lean burning

efficiency of a gasoline engine with a spark plug and explore the possibility of charging an

unfueled, unscavenged prechamber with a fuel-rich mixture without additional systems.

This was accomplished by investigating different prechamber modifications and

conducting a series of CFD simulations. The results showed that the proposed configurations

not only make it possible to charge the prechamber with a fuel-rich mixture, but also help

prevent the mixture from leaking into the main chamber during the compression stroke. I

believe that my study makes a significant contribution to the automotive field because the

study results demonstrate the potential of retrofitting these modifications on existing port-

2
injected engines in order to improve the combustion efficiency and overcome the

aforementioned challenges.

Keywords: prechamber; two-stage combustion; torch ignition; unscavenged prechamber; SI

engine

3
Declaration

I wish to submit a thesis titled “Ultra-Lean Combustion of an SI engine using Alternative

Stratification Principles” for the Ph.D. degree.

This study investigates the use of an unfueled prechamber with a fuel-rich mixture to improve
the efficiency of a gasoline engine. This was accomplished by investigating different
prechamber modifications and conducting a series of CFD simulations. The results showed
that the proposed configurations not only make it possible to charge the prechamber with a
fuel-rich mixture, but also help prevent the mixture from leaking into the main chamber
during the compression stroke. I believe that my study makes a significant contribution to the
literature because the study results demonstrate the potential of retrofitting these
modifications on existing port-injected engines in order to improve the combustion efficiency.

I confirm that, except when due recognition has been made, the work belongs to only one author;
the work has not been submitted previously, in whole or in part, to qualify for any other academic
award; the content of the dissertation is the result of work performed from the official start date
of the approved research program; any editorial work, paid or unpaid, performed by a third
party, is recognized; and ethical procedures and guidelines were followed.

Gombosuren Nyamsuren
Department of Mechanical Engineering, Ritsumeikan University
gr048072@ed.ritsumei.ac.jp

4
Acknowledgments

I would like to express my deep appreciation to my supervisor Dr. Ogami Yoshifumi for
guidance, encouragement, and support in the process of completing this Ph.D. I would like
to thank my teammates for heated discussions. In addition, I would like to express my
appreciation to the JICE for the scholarship and financial support throughout this study.
I would like to thank Mr. Sugimoto Hiroyuki of Ritsumeikan University for the technical
support and all the technicians involved in the project at both Ritsumeikan University and
Mongolian University of Science and Technology.
I am grateful to my mother, Mrs. Jambalsuren. G, and to my brothers and sisters who
supported me all this time. This thesis is dedicated to my father, Mr. Nyamsuren A.

5
Nomenclature

𝐼∙𝑠𝑎 Moment of inertia of fluid mass about swirl axis, kg/m2

𝐼∙𝑡𝑎 Moment of inertia of fluid mass about tumble axis, kg/m2
k Turbulent kinetic energy, m2/s2
𝐿∙𝑠𝑎 Magnitude fluid angular momentum with respect to swirl axis, kg·m2/s
𝐿∙𝑡𝑎 Magnitude fluid angular momentum with respect to tumble axis, kg·m2/s
N Engine operating speed, rpm
Rs Swirl ratio, dimensionless
Rt Turbulent ratio, dimensionless

Greek symbols
𝜔𝑠 Angular velocity rotating flow with respect to swirl axis, rad/s
𝜂𝑡ℎ Theoretical efficiency of Otto cycle
𝑟 Compression ratio
𝛾 Specific heat coefficient

Abbreviations
BDC Bottom dead center
CFD Computerized fluid dynamics
IC Internal combustion
TDC Top dead center
CAD Crank angle degree
PFI Port fuel injection
GDI Gasoline direct injection
BC Black carbon
THC Total hydrocarbon
PM Particulate matter
PN Particulate number

6
Table of Contents
Abstract: ................................................................................................................................. 2
1. Introduction .................................................................................................................. 13
1.1 Background ................................................................................................................ 13
1.2 Research purpose and scope ....................................................................................... 18
1.3 Research structure ...................................................................................................... 20
2. Literature review .............................................................................................................. 22
2.1 Lean combustion of SI engines .................................................................................. 23
2.2 Improvements to lean combustion ............................................................................. 25
2.3 Injection and ignition strategies in SI engines............................................................ 26
2.3.1 Homogeneous mixture ................................................................................... 27
2.3.2 Stratified mixture ............................................................................................ 28
2.3.3 Ignition principles........................................................................................... 30
3. Alternative approaches for mixture stratification ............................................................. 33
3.1. Historical background ............................................................................................ 33
3.2. Development of the prechamber ............................................................................ 38
3.3. Radical ignition ...................................................................................................... 42
3.4. Two-stage combustion systems ............................................................................. 43
3.5. Swirl prechamber spark plugs ............................................................................... 44
3.6. Prechamber pilot spark plug .................................................................................. 45
3.7. Unfueled Prechamber ............................................................................................ 46
3.8. Turbulent jet ignition ............................................................................................. 47
3.9. Direct injection jet ignition (DI-JI) ........................................................................ 49
4. Effect of pre-chamber systems on SI engines .............................................................. 50
4.1. Flammability limit ..................................................................................................... 50
4.2 Spark time................................................................................................................... 52
4.3. Start of combustion ................................................................................................... 53
5. Main principles and parameters of the pre-ignition chamber....................................... 54
5.1 Preliminary study of the prechamber ......................................................................... 56
5.2 Fluid flow in the prechamber ..................................................................................... 57
5.2.1 Turbulence and tumble ........................................................................................ 58

7
 5.2.2 Swirl and tumble ................................................................................................. 58
6. Methodology .................................................................................................................... 59
6.1 Modeling the engine ................................................................................................... 61
6.2 Decomposing and meshing ........................................................................................ 63
6.3. Grid test ..................................................................................................................... 65
6.4 Cycle-to-cycle variation ............................................................................................. 65
7. Results and discussion ...................................................................................................... 69
7.1 Geometry and parametric solutions ............................................................................ 69
7.1.1 Model development of the prechamber ............................................................... 69
7.1.2 Geometry of the main chamber ........................................................................... 70
7.2 Design of the prechamber........................................................................................... 71
7.2.1 Effect of inlet and outlet duct diameters of the prechamber ............................... 71
7.2.2. Geometry configuration of the model ................................................................ 75
7.3 Position of the prechamber ......................................................................................... 80
7.3.1 Tilted plane effect on the charge flow ................................................................. 80
7.3.2 Influence of angular position φ° on the charge flow ........................................... 82
7.3.3 Influence of inclination angle θ° ........................................................................ 87
7.3.4 Effect of the vertical position of the prechamber ................................................ 90
7.4 Comparison of wavelike fluctuation in the prechamber charge flows ....................... 92
7.4.1 Fluctuation phenomena ....................................................................................... 92
7.4.2 Charge and discharge of the prechamber ............................................................ 92
7.4.3 Engine speed........................................................................................................ 95
7.4.4 Smoothness of the fluctuation ............................................................................. 97
7.4.5 Model matching ................................................................................................... 99
7.4.6 Conclusion of the chapter .................................................................................. 102
8. Conclusions .................................................................................................................... 102

8
List of publications

[1] G. Nyamsuren, H. Asada, and Y. Ogami, “Computational study on using an unfueled

prechamber for lean burning in a gasoline engine,” 2019.
https://doi.org/10.1299/jsmekansai.2019.94.414.
[2] G. Nyamsuren, K. Fukudome, and Y. Ogami, “CFD investigation of self-scavenging
prechamber for the lean flammability of a gasoline engine,” 2018.
https://doi.org/10.1299/jsmekansai.2018.93.1012.
[3] N. Gombosuren, O. Yoshifumi, and A. Hiroyuki, “A charge possibility of an unfueled
prechamber and its fluctuating phenomenon for the spark ignited engine,” Energies, vol.
13, no. 2, p. 303, Jan. 2020, https://doi.org/10.3390/en13020303.
[4] N. Gombosuren, H. Asada, and Y. Ogami, “Computational fluid dynamics study for an
efficient passive prechamber,” The 29th International Symposium on Transport
Phenomena (ISTP29), vol. 29, Honolulu, USA, p. 4, Oct.2. 2018.
[5] G. Nyamsuren and Y. Ogami, “Fluctuating phenomena in the charge flow of un-fueled
prechamber of gasoline engine,” Sixteenth International Conference on Flow
Dynamics, vol. 16, Sendai, Miyagi, Japan, p. 2, Nov. 2019.
[6] G. Nyamsuren, Y. Ogami, and H. Asada, “Charge flow investigation of un-fueled
prechamber for the gasoline engine,” Fifteenth International Conference on Flow
Dynamics, vol. 15, Sendai, Miyagi, Japan, Nov. 2018.

9
List of Figures
Figure 1 Energy demand for EU ......................................................................................... 13
Figure 2 Emission target share of EU by 2025.................................................................... 14
Figure 3 Comparison of maximum gross indicated efficiency of various methods ............ 19
Figure 4 Number of studies on lean burn since 1970 .......................................................... 24
Figure 5 Number of studies related to NOx-SI engines since 1970 .................................... 25
Figure 6 Number of studies on cold-start SI engines over the years ................................... 27
Figure 7 Comparison of homogenous and stratified charges .............................................. 29
Figure 8 Partial stratification spark plug injector [46] ........................................................ 31
Figure 9 Concept of the prechamber patented by Ricardo [47] .......................................... 33
Figure 10 Number of studies conducted on prechambers over the years ............................ 34
Figure 11 Types of prechambers ......................................................................................... 35
Figure 12 Stratification principles ....................................................................................... 36
Figure 13 Turbulence-generating torch design 2Sy Toyota [33] ....................................... 38
Figure 14 Types of prechamber systems [11] ..................................................................... 40
Figure 15 Principles of a fueled prechamber ....................................................................... 41
Figure 16 Schematic and example of a pre-chamber spark plug [22] ................................. 44
Figure 17 Principle of pre-chamber spark plug with pilot injection [50] ............................ 46
Figure 18 Principle of unfueled prechamber combustion ................................................... 47
Figure 19 Spark ignition (SI) and pre-chamber turbulent jet [17]....................................... 47
Figure 20 Variation in 0–10% burn duration with λ for all TJI nozzle diameters [84] ....... 55
Figure 21 Principles of swirl and tumble flows................................................................... 59
Figure 22 Charge flow characteristics ................................................................................. 60
Figure 23 Configuration of an engine cylinder head with the prechamber located on the top
of the pent-roof-type combustion chamber (two valves each for inlet and exhaust). .......... 61
Figure 24 Decomposition model of the engine with prechamber: Inlet-A, and outlet-B, ducts
in the main body of prechamber-C ....................................................................................... 63
Figure 25 Comparison of tumble ratios for different grid cell numbers ............................. 65
Figure 26 Tumble ratio for four cycles of cold flow simulation. ........................................ 66
Figure 27 Comparison of tumble ratios for four cycles with and without the prechamber. 67
Figure 28 Valve lift profile .................................................................................................. 68
Figure 29 Comparison of the main chamber geometry ....................................................... 70
Figure 30 Variation in compression ratio with charge flow velocity .................................. 70
Figure 31 Model and mechanical configuration of the simulation...................................... 72
Figure 32 Comparison of the charge flow with the charge velocity for various inlet hole
diameters. ............................................................................................................................. 73
Figure 33 Velocity magnitude at 63.95° after TDC in the intake stroke ............................. 74
Figure 34 Prechamber. Dashed area indicates the difference between the designs. ........... 75
Figure 35 Contour of velocity magnitudes around the prechamber during the intake stroke at
64.1° from the top dead center (TDC). ................................................................................. 77

10
Figure 36 Mass flow rates of the outlet ducts of the prechamber with chamfered edges (grey
line) and sharp edge (black line). ......................................................................................... 78
Figure 37 Comparison of mass flow rate differences between inlet and outlet ducts of
chamfered design (gray line) and the sharp-edge design (black line) of the prechamber .... 79
Figure 38 Comparison of the design configurations of a different plane of the prechamber
inclination ............................................................................................................................. 80
Figure 39 Charge flow comparison of two designs ............................................................. 81
Figure 40 Angular position of the prechamber with respect to φ°. ..................................... 82
Figure 41 Velocity magnitude at various angular positions of the prechamber with respect to
φ°. ......................................................................................................................................... 83
Figure 42 Velocity magnitude at various angular positions of the prechamber with respect to
φ°. ......................................................................................................................................... 84
Figure 43 Mass flow rates of the inlet of the prechamber for rotation angles of 60° and 55°,
45° with the angular correction. ........................................................................................... 85
Figure 44 Mass flow rates of the outlet of the prechamber for rotation angles of 60° and 55°,
45° with the angular correction. ........................................................................................... 85
Figure 45 Tilt position of the prechamber ........................................................................... 87
Figure 46 Comparison of tumble ratios of different tilt positions ....................................... 88
Figure 47 Comparison of the charge flow velocity magnitudes at different CAD positions..89
Figure 48 Comparison of vertical positions of the prechamber .......................................... 90
Figure 49 Comparison of the mass flow rates at different vertical positions of the
prechamber ........................................................................................................................... 91
Figure 50 Pressure at different places of the prechamber. .................................................. 93
Figure 51 Net mass flow rates of the inlet and outlet ducts of the prechamber for chamfered-
edge design ........................................................................................................................... 94
Figure 52 Inlet (blue), outlet (black), and net (red) mass flow rates of the ducts of the
prechamber for chamfered-edge design at the 2.36 mm position. ....................................... 94
Figure 53 Comparison of the prechamber positions............................................................ 96
Figure 54 Fluctuation intensity and velocity vector ............................................................ 97
Figure 55 Comparison of the net mass flow rates with respect to engine speed. ................ 98
Figure 56 Net mass flow rate fluctuation model ................................................................. 99
Figure 57 Amplitude of the fluctuation related to engine speed ....................................... 100
Figure 58 Relationship between charge flow fluctuation frequency and engine speed .... 101
Figure 59 Timing of the charge mass flow rate fluctuation in the prechamber (y represents
the coefficient b; x represents the engine speed) ................................................................ 101

11
List of Tables

Table 1 Engine specifications .............................................................................................. 62

Table 2 Model zones ............................................................................................................ 63
Table 3 Prechamber position and inlet and outlet duct diameter......................................... 72
Table 4 Vertical position of the prechamber ....................................................................... 76
Table 5 Angular position of the prechamber ....................................................................... 80
Table 6 Angular position of the prechamber though φ° ...................................................... 83
Table 7. Position of the tilt angle of the prechamber ........................................................... 87
Table 8 Vertical position of the prechamber ....................................................................... 90
Table 9 Fluctuation coefficients .......................................................................................... 99

12
1. Introduction
1.1 Background
Fossil fuels currently make up 81% of the global primary energy consumption [1]. Scarce

resources and increasingly stringent emission requirements necessitate a stronger focus on

alternative fuels and systems [2].

Oil products are expected to take approximately 88% of the EU transportation sector

requirements by 2030 and 84% by 2050, despite the implementation of alternative fuels and

methods [3]. Looking at the total EU transportation energy demand, as shown in figure 1,

domestic, intra-EU, and intercontinental traffic and road transport by far consume the highest

energy (72.3% of the total) [3]. This is consistent with the growth in the number of light

vehicles in the years to come [4].

Natural
gas, 0.7%
Biofuels, Fuel oil,
3.7% 10%
Electricity,
1.4%
Diesel,
LPG, Gasoline,
49.8%
1.4% 20.7%

Kerosene,
12.3%

Figure 1 Energy demand for EU

Because of the need for cleaner transportation and greater fuel economy, many

researchers have focused on studying alternative methods to optimize the use of alternative

fuels in an internal combustion engine (ICE).

Extension in the lean burning limit of spark ignition (SI) engines allows to enhance the

range of stable lean fuel combustion and thermal efficiency [5]. The octane number indicates

the resistance of a fuel from detonating under given operating conditions. Figure 2 shows the

worldwide distribution of carbon emissions.

13
 Mid EV Scenario Low EV Scenario High EV Scenario
75 g CO2/km in 2025 75 g CO2/km in 2025 75 g CO2/km in 2025
(95% phase in) (90% phase in) (100% target)

Figure 2 Emission target share of EU by 2025

A few strategies have been proposed to achieve rapid burning in spark SI engines to

enable them to compete with conventional-fuel ICEs. This involves altering the shape of the

combustion chamber, employing dual-fuel combustion (such as natural gas and hydrogen),

and improving the injection approach; including the application of stratified methods [6].

Modifying the combustion chamber structure can improve the air flow and generate strong

turbulence, which expands the flame zone, thus resulting in rapid burning. For dual-fuel

applications, the combination of natural gas and hydrogen has been considered to increase

the flame speed, though greater nitrogen oxide (NOx) emissions have been reported when a

higher amount of hydrogen is used [7].

A direct injection technique has been developed through the injection boost approach. In

this method, a lean combustion is possible, which helps reduce emissions and increase fuel

economy. In recent years, gas engines have been modified into lean combustion engines by
14
sending excess air into the engine along with the fuel. This system is advantageous from two

aspects: 1) The excessive air decreases the amount of nitrogen oxide (NOx) as compared to

conventional natural gas engines; 2) The excessive air delivers additional oxygen, increases

the gamma value, and leads to lower pumping losses under partial loading. Consequently,

more efficient combustion and power can be produced from the same amount of fuel [8].

However, a higher air–fuel ratio (AFR) results in longer and incomplete combustion

processes, making the engine unstable and increasing hydrocarbon emissions [9].

Stratification is a lean combustion principle wherein a fuel-rich mixture is locally formed

adjacent to the spark plug within a lean mixture in the rest of the cylinder volume. The

ignition of this rich mixture generates sufficient energy to burst the lean mixture, resulting in

a more stable flame. Charge stratification permits the engine output to be monitored without

restricting the air flow into the cylinder, thus minimizing the intake pumping losses. Another

strategy to overcome the low ignitability of SI is the incorporation of a prechamber to achieve

a leaner mixture. Prechamber spark plugs have become popular in large gas engines with

wide cylinder bores [10].

The following are some of the features of a prechamber:

• It contains a spark plug, a prechamber, and numerous holes.

• A turbulent jet is utilized to ignite the lean mixture in the main chamber.

• The jets function as a distributed source of ignition; the combustion process is less

reliant on the total air–fuel ratio (AFR).

In this approach, the combustion process proceeds in two stages: in the prechamber and

in the main chamber of the engine cylinder. The ignition starts in the prechamber, and when

the flame passes through the hole (duct that connects the main and prechamber), jets of flame

from the prechamber are introduced into the main chamber. Thus, a better flame propagation

rate and more stable combustion can be realized [11]. Ferrari recently introduced a turbulent

jet ignition (TJI) prechamber developed by MAHLE at the 2015 Canadian Grand Prix. Given

the evident benefits of a prechamber, this study focused on the use of a prechamber for SI

engines.

15
 Studies have shown that incorporation of the lean-burning concepts to the use of a lean

or an ultra-lean air–fuel mixture is a proven approach to reduce fuel consumption as well as

pollutant emissions [11]. The most promising alternative to port fuel injection (PFI) is

gasoline direct injection (GDI) owing to its clean-burning and fuel-saving advantages.

However, it has been reported that the GDI engines emit black carbon outputs, such as global-

warming pollutants, to a greater extent than PFI engines do [12]. It has been reported that the

GDI engines emit a higher total amount of hydrocarbon (THC), mass of particulate matter

(PM), and amount of solid particulate number (PN) at 30 °C [13,14]. These issues can be

overcome by employing a prechamber, which helps develop a higher stratified charge.

Prechamber combustion has been referred to as “prechamber combustor” [15], [16],

“prechamber ignition” [11], “torch ignition” [17], “jet ignition” [18], “two-stage combustion”

[19], and “scavenged prechamber” [2]. The results of these studies validate the advantages

of the fuel-added prechamber for SI engines and confirm the lean-burning and fuel-saving

characteristics of prechambers with an additional fuel delivery nozzle.

Although the benefits of a fueled prechamber with additional fuel or air supply have been

validated, the benefits of un-fueled prechambers have not been thoroughly studied.

Alternative terminologies for unfueled prechamber combustion include “passive prechamber

combustion” [20] and “unscavenged prechamber combustion” [21]. Benajes et al. reported

increased passive prechamber efficiencies with good combustion stability and high

combustion efficiencies under stoichiometric conditions. They studied a turbocharger to

amplify the injection of a fuel-rich mixture to the prechamber using different configurations

[20]. The study experimentally proved that the unfueled prechamber efficiency can be

increased, and that it may no longer be passive, provided that the charging method is viable.

With the use of a turbocharger, charging a prechamber with richer fuel can help increase the

efficiency. Therefore, an understanding of the charge flow phenomena is important for

determining the effectiveness of unfueled un-scavenged prechamber systems. However,

numerical studies in this field are lacking, indicating the need for further studies to advance

the concept of lean combustion with prechamber systems [11]. Some phenomena, including

charge and discharge or reverse flow before the ignition and flow exchange processes
16
between the prechamber and main chamber and inside the prechamber, remain unresolved.

Further efforts should be expended to develop the TJI concept via more sophisticated

measurement techniques, such as particle image velocimetry (PIV) and/or numerical methods

such as computational fluid dynamics (CFD) [20]. A drawback of passive pre-spark plugs is

the safe ignition of the mixture under idling and partial loading conditions, from low to

medium with a correspondingly high internal content of residual gas. Moreover, the dilution

of the charge in the cylinder with air makes it even more difficult to obtain a good air–fuel

mixture in the preliminary chamber. To ensure reliable ignition of the mixture in the

preliminary chamber even under such boundary conditions, it is necessary to develop a

suitable geometry for the preliminary chamber in combination with direct injection [22].

Since 2014, I along with my team have investigated the charging of an unfueled

prechamber with a mixture richer than that in the main chamber. Studies include

investigations on prechamber positions, combinations of the effects of precession and tilt

angles, and geometry of the inlet/outlet holes of the prechamber at the beginning of the intake

stroke [23]–[28]. The charge flow velocity was found to be the primary predictor of charge

flow intensification. A transient velocity pattern was considered at the inlet and outlet duct

locations of the prechamber. In terms of the piston movement, a prechamber with a fuel-rich

mixture can be charged in the compression and intake strokes. The second part of the charge

flow can be explained by the pressure difference produced during compression in the main

chamber via the piston movement. The first part of the charge flow has the most important

phenomenon since the addition of the rich mixture and discharging of the burnt gas are

realized without an additional turbocharger, fuel, or air injector.

In this study, the charge flow of an unfueled prechamber was investigated from the

beginning to the end of the intake stroke considerably. In contrast to studies on conventional

prechambers operating without fuel, different studies on tuning and separate assessments of

the input and output channels for their direction of charge flow are being conducted. The

methodology and simulation models are discussed in Section 3 and 5.

Through cold flow simulations, the mass flow rates through the prechamber inlet and

outlet ducts were determined without the influence of chemical or thermal reactions at the
17
beginning of the intake stroke. This helped clarify the characteristics of the first charge flow

and the effects of different inner edges of the prechamber configurations. The influence of

the design changes on the charge flow is described in Section 4.1.

The charge flow characteristics of the inlet and outlet mass flow rates were nonlinear.

The first charge amplitudes were useful in identifying whether the prechamber was well

charged. An effect similar to the inverse liquid bottle phenomenon [29], [30] is useful not

only for facilitating the charge flow but also for maintaining the charge without any leakage

during the second charge. The effect of fluctuation [29], [31] of the charge and discharge

flows is described in Section 4.2.

This charge flow phenomenon provides the reasoning behind the flow configuration in

addition to the possible condition for directing the richer part of the port-injected fuel into

the prechamber. Section 5 presents the conclusions regarding the effectiveness of the new

prechamber, along with directions for future work. The current design of the optimized

configuration not only makes it possible to charge the prechamber with a fuel-rich mixture

but also helps prevent the mixture from leaking into the main chamber during the

compression stroke.

1.2 Research purpose and scope

For the reasons discussed above, the prechamber is considered a potential alternative to

provide stratification. When an engine is running, particularly at partial loads, the energy is

reduced because of the pumping losses. At a partial load, injecting fuel directly into the

chamber after closing the intake valves can help eliminate the effect on the volumetric

efficiency and reduce pumping losses. With lean combustion, the fuel economy is improved

by reducing fuel consumption and emissions. In lean-burn SI engines, the emissions can be

reduced to almost zero; however, the main difficulty is related to the ignition. The

flammability of a lean mixture can be improved using a prechamber. A prechamber with a

small cavity, connected to the main chamber through small openings, enables a quick burning

of the lean mixture. A prechamber with or without fuel increases the flammability of these

18
lean mixtures. Most researchers have employed the direct injection method for the

prechamber, while unfueled prechambers have not been widely investigated.

In this work, the following research gaps have been addressed:

1. The inefficiency in the ignition system for lean combustion in SI engines with a

prechamber.

2. The effect of prechamber jet ignition and port-injection in SI engines on thermal

efficiency.

3. The formation of a mixture of fuel in the case of port-injection and prechamber jet

ignition.

4. Lack of study on the characteristics of unfueled prechambers, including in terms of

the geometry, position, and design characteristics of SI engines.

Figure 3 shows a comparison of the maximum gross indicated efficiency of various

ignition methods.

Speed = 1200 rpm

Load = 10 bar IMEP
48

47

46
Research target
45
λ=2.0
44

43 λ=1.7
42 λ=1.6
41
Spark plug Un-fueled Fueled prechamber
Prechamber

Figure 3 Comparison of maximum gross indicated efficiency of various methods

Since modern SI engines use only a part of the fuel energy, it is important to focus on the

possibilities to maximize the charging potential of prechamber methods. The research in this

dissertation focuses on three main objectives covering the field:

1. Exploring charge strategies for SI engines to expand lean flammability.

19
 2. Understanding the possibility of prechamber ignition in an SI engine.

3. Finding design solutions to unfueled prechambers using multidimensional software,

including the distribution of charge velocity near the ignition source, to understand

the behavior of charge flow characteristics in the passive prechamber of the engine.

The aim of the present study was to analyze the use of a prechamber ignition system with

an emphasis on lean combustion, given the known advantages in producing less exhaust gas

compared to using a stoichiometric mixture [8], [11], [20]. To this end, the charging

possibility of the unfueled prechamber is the main target for connecting port-injection and

prechamber ignition systems in terms of the performance of torch-ignition and spark-ignition

engines based on CFD modeling backed by laboratory experiments. Experiments were

conducted at the worldwide mapping point (WWMP) and wide-open throttle to understand

the requirements for operating the prechamber ignition system, in order to maximize the

thermal efficiency of engines with port injection. Since the scope of the research is

determined by the three objectives, a number of separate questions arise:

1. What are the implications of using a jet prechamber for a port-injected SI engine (in

terms of the performance)?

2. Is it possible to increase the lean flammability of port-injected SI with less pumping

losses when using a prechamber?

3. What are the ideal design characteristics for prechamber ignition in a port-injected

SI engine?

1.3 Research structure

The objective of this thesis was to study the use of a prechamber jet ignition system with and

without direct injection strategy in light-duty SI engines. Techniques to overcome the

disadvantages of conventional SI engines are proposed via port-injection and jet ignition

through a prechamber. On the basis of the concepts of this study, research subjects are

summarized in this chapter.

20
Chapter 2 presents the key concepts and explanations, and most importantly emphasizes the

state of art. This chapter shows what has been achieved and what remains to be studied, which

is the prospect of the present research. The chapter is divided into three sections: the first one

explains the lean combustion of SI engines; the second one details the methods employed to

improve lean combustion; the third one illustrates the ignition strategies used in SI engines.

Chapter 3 covers alternative strategies for mixture stratifications and reports the combustion

concept of prechamber jet ignition and direct injection with numerical and experimental

methods. The result show that jet ignition can realize a faster combustion. The calculated

burning rates are very high even with ultra-lean mixtures stratified near the jet nozzles. This

preliminary analysis demonstrated that the use of a prechamber with direct injection of

gaseous fuel can help generate faster combustion even with an ultra-lean concentration of the

net mixture.

Chapter 4 presents the main effects of parameter systems on engine operation. Detailed

descriptions of the flammability limit of SI engines and ignition-related findings are

provided. The strategies proposed for port-injection SI engines are explained.

Chapter 5 deals with research issues regarding port injection with prechamber application.

The theoretical aspects of the prechamber engine with the CFD approach for the numerical

analysis are discussed. This chapter consists of two sections: 1) a preliminary study of the

prechamber, and 2) fluid flow issues including tumble and swirl motion.

Chapter 6 describes a numerical model developed to understand the charge flow formation

inside a port-injected SI engine with a prechamber. The chapter begins with the fundamentals

of prechamber design with an expanded review of previous investigations on methods used

for the numerical modeling of prechamber systems. Flow validation cases with a grid

dependency study are included. First, a model is developed with the geometry of the main

and prechamber assemblies using the multidimensional CFD software; a test case study is

conducted on the holes in the combustion chamber to characterize the charge phenomena

under a given geometry constraint. Second, decomposition combines direct meshing with

prechamber geometry variations. The third section comprises a comprehensive analysis of

the grid test results, and the fourth section analyzes the cycle-to-cycle variations of the ICE.
21
The results of this analysis are then utilized for developing a numerical model of the

prechamber, which is defined thoroughly in this chapter.

Chapter 7 covers a comprehensive CFD analysis of the charge flows and fuel distribution in

the prechamber and combustion chamber based on the relevant engine operating conditions.

The partial load condition at 2000 rpm is applied to clarify the behavior of the unfueled

prechamber with the indirect injection concept. The chapter consists of four main sections.

First, the geometry of the main and prechamber are discussed. Second, the influence of

geometry on the charge flow of the prechamber is evaluated. The results show that a wider

prechamber duct diameter helps increase the charging possibility of the prechamber

depending on the pressure in the main chamber. In the third section, the position of the

prechamber is tested. The charge mass flow rate and its characteristics are discussed in the

last section.

Chapter 8 presents the main conclusions of this study. It also outlines the optimization and

best solutions for connecting the prechamber in SI direct ignition engines.

2. Literature review
This chapter has two main sections. The first section describes the operation and the fuel

supply system in the SI engine. The second section presents the development of alternative

mixture stratification systems for lean combustion in SI engines.

J. F. Thomas and R. H. Staunton [32] emphasized that the SI engine efficiency can be

improved via the following three methods:

a) by increasing the compression ratio,

b) by reducing the throttling of the intake airflow as a method of regulating power and

torque under partial loading,

c) by implementing a lean mixture concept.

They outlined the following promising technologies [32]:

1. Turbocharged lean burn and PFI: these can be applicable if stratification is possible.

22
 2. Lean combustion during refueling with early injection (homogeneous charge). The

use of a turbocharger is optional, with the capability for stratified charge.

3. Lean combustion during refueling (prechamber) indirect injected diesel (IDI)

(recommended at low pressures)

4. Late direct injected (DI) or IDI (only with high-pressure injection)

5. Driving of the engine to employ stoichiometric refueling under certain conditions

and to operate in lean mode at low loads.

The second part (if stratification is possible) of the first aspect of these technologies has not

been widely studied; the second technology aspect has only been partially studied for the

turbocharged case; the third technology aspect has mostly been explored for DI applications.

The fourth and fifth cases listed have been widely studied.

From the above review, there is some research gap related to the stratified charge in the

case of early injection and usage of inlet stratification.

2.1 Lean combustion of SI engines

Lean combustion is one of the most promising techniques to decrease emission levels and

increase fuel economy. Figure 4 shows the search results for the keyword “lean burn.”

According to the database of the Web of Science, the number of studies related to

prechambers has increased rapidly since 1990, with approximately 200 papers presented

yearly. A lean burn appears when the AFR is greater than the theoretical ratio

(stoichiometric), λ = 1, where λ is the ratio of the actual AFR to the stoichiometric AFR.

Operating an engine in a lean state with additional air increases the specific thermal

coefficient (γ), which leads to an increase in the thermal efficiency according to equation (1)

.
1
𝜂𝑡ℎ = 1 − 𝑟𝛾−1 (1)

Here:
𝜂𝑡ℎ is the theoretical efficiency of the Otto cycle, 𝑟 is the compression ratio, and 𝛾 is the

specific heat coefficient.

23
 Figure 4 Number of studies on lean burn since 1970

Regarding exhaust emissions, with excess air under conditions with lean mixture

operation, the combustion efficiency increases while reducing CO emissions and the total

amount of hydrocarbons. NOx is formed typically in the burnt gas as a result of complete

high-temperature combustion reactions after flame formation, and the rate of NOx formation

exponentially varies with the temperature of the burnt charge. Therefore, the low temperature

of the burnt charge in the lean mixture can reduce NOx emissions without causing fuel loss

[33]. Moreover, lean combustion reduces knocking and allows engine operation at a high

compression ratio, resulting in a better thermal efficiency. The method of using an air–fuel

mixture with additional air (lean mode) or exhaust gas recirculation (EGR) has been validated

for its effectiveness. Lean combustion increases the efficiency of engine driving cycles,

which reduces fuel consumption. It also allows the engine to work with less throttling while

maintaining the same road power, which significantly reduces pumping losses [34]. The

output is controlled by changing the amount of fuel injected into the combustion chamber,

and not via throttling as in conventional SI engines.

In addition, a significant reduction in NOx emissions can be realized with a lean

flammability λ ~ 1.4 in conventional SI engines, at higher λ owing to lower combustion

temperatures. I studied topics related to this research with the keyword NOx-SI engine in the

24
Web of Science database. Figure 5 shows the result of the research. As shown, the number

of studies related to this subject has increased year on year.

Figure 5 Number of studies related to NOx-SI engines since 1970

In addition, the temperature change with respect to the AFR during the lean

stoichiometric combustion of conventional SI engines decreases the efficiency of the three-

way catalyst, which is the main system used for controlling NOx, HC, and CO emissions.

This combustion instability leads to increased HC emissions because of the misfiring of

partial combustion cycles. Partial combustion is primarily due to the low speed of the laminar

flame of lean mixtures, which affects the growth and spread of the flame core. As the lean

mixture burns, the length of the combustion process is extended. Lean combustion is also

limited to cyclic variations, with increased mixture dilution.

2.2 Improvements to lean combustion

Numerous upgrades are required to overcome the problems associated with lean-mixture

burning, such as slow burning and low flammability, which have been discussed in [35] [36].

Some possible solutions are as follows:

a) Environmental solution: Design factors may concern valve reaction and variable

timing, turbulence and mixing, compression ratio (CR), fuel injection, and ignition

timing.

b) Enhancement of reliable control systems and devices. In natural gas engines, the

lean limit (misfire limit) is influenced by the concentration of water vapor

25
 (humidity) in the intake air [37]. Water vapor has a diluent effect with the change in

the O2 sensor output.

c) Switching modes between stoichiometric and lean burns: This is a feasible

approach to meet emission criteria while retaining some of the benefits of lean

burns.

d) Stratification: This is one of the best alternatives for increasing lean flammability

and reducing NOx emissions. A low-pressure fuel supply (<10 MPa) is vital

because a high pressure (35–40 MPa) can reduce the range of the vehicle. The

charge stratification technique, including injection at the end of the compression

cycle, requires a high gas pressure.

e) Ignition system improvement: The ignition lead time in the prechamber helps

reduce maintenance requirements and extend the application of the lean

flammability. Owing to the optimization of SI, the objectives achieved to improve

the quality of ignition, to create a more reliable ignition under conditions close to

the misfire limit, to expand the lean flammability, and to develop the core faster to

improve the rate of heat generation.

f) Increased turbulence and mixing in the combustion chamber: The heat release

rate increases with an increase in the turbulence and mixing rate in the combustion

chamber when fast burning is required.

g) Other: Avoiding throttling and use of partial stratification with alternative

approaches.

Along with the low laminar velocity of lean mixtures in natural gas engines, it is important

to develop a technology that is designed to take benefits of lean combustion while avoiding

the drawbacks.

2.3 Injection and ignition strategies in SI engines

The mixing of air in the cylinders is a key phenomenon in ICE. With direct injection, two

combustion conditions are formed: homogeneous and stratified mixtures. Different types of
26
mixtures can be made depending on the time of injection of the fuel into the cylinder.

Injecting fuel after closing the valves can produce a homogeneous mixture, whereas injection

near the top dead center creates a stratified charge. Figure 6 shows the related research on

cold-start SI engines conducted over the years.

The results from the Web of Science show that the cold start ability of SI engines has

been studied since as early as 1912. The number of studies related to this topic has rapidly

increased since the 2000s.

Figure 6 Number of studies on cold-start SI engines over the years

2.3.1 Homogeneous mixture

The homogeneous and stratified mixtures are crucial for SI engines, because they bring a

significant distinction in regulating the charge mixture composition of direct injection

engines compared to fuel injection through the inlet port. The direct injection approaches can

be categorized in terms of the mixture formation into homogeneous and stratified charges.

The primary fuel injection system generates a homogeneous charge, which is designed for SI

engines operating at medium and high loads [38]. Homogeneous mixtures improve the

indicated efficiency of the engine by reducing losses throughout the exchange process. This

minimizes the effect on the volumetric efficiency of the engine if the injection begins at the

end of the intake stroke [13]. The influence of injection and ignition timing on the combustion

characteristics and emissions of LPG and compressed natural gas direct injection (CNG DI)

has been reported in [8] [40]. Liu et al. showed that an improved fuel injection can lead to

27
better mixing of the air with fuel and accelerate flame propagation. Since the combustion

characteristics are affected by the composition and procedure of the mixture formation, early

injection contributes to the formation of a flame core, which then reduces the combustion

duration, with a slight increase in rapid combustion. The NOx emission and HC concentration

increase with the advance of injection and spark timings, whereas CO concentration varies

slightly with respect to these parameters. In [39], an In GASSP 2Sturbocharged combustion

system using a three-way catalyst was utilized to validate the mixing process of homogeneous

combustion in a direct injection CNG engine. An analysis of 50 planar laser induced

fluorescence (PLIF) images at 50° crank angle before top dead center (CA BTDC) under

partial loading showed that 100% of the mixture was flammable. The excellent mixing gave

inhomogeneous mixtures with the advance of injection timing. When fuel is injected early

enough, the time required to form homogeneous mixtures is sufficient until the ignition. The

lean limit for the homogeneous mixture was λ = 1.25. The lean limit (λ > 1.25) leads to

incomplete combustion, which contributes to an excessive release of hydrocarbons.

2.3.2 Stratified mixture

A stratified charge can be formed via late fuel injection near the top dead center. There is a

significant amount of fuel near the spark plug with a lean mixture in the rest of the combustion

chamber. A high compression ratio and an ultra-lean stratified mixture are crucial for

improving engine efficiency [8]. The concept of a stratified charge is unlike that of a

homogeneous mixture; therefore, some modifications and optimizations are required to

operate the engine in the stratified approach. Figure 7 shows a comparison of the

characteristics of homogeneous and stratified charges.

To accomplish stratification, Yadollahi and Boroomand conducted a computational study

on the effects of injector geometry and engine speed during the direct injection of natural gas

on AVL FIRE code. Two centrally located injectors were used: single-hole and multiple-

holes with inward opening injection needles [40]. They observed that a single-hole injector

provides the same proportion of flammable mass at the end of a compression stroke as that

provided by a multi-hole injector.

28
 Moreover, stratification becomes more difficult at higher speeds. In another study,

Baratta et al. [41] used STAR CD to examine the effects of injection approaches, protrusion

of the injector tip, and geometries of the piston and cylinder head on a centrally mounted

Mixture formation in a gasoline engine

Homogenous charge
If lean (less fuel + more air): Stratified mixture (less fuel + more air)
Less toxic, more economical Less toxic (complete combustion)
But difficult to ignite and low and
More economical (Less fuel)
power and instability
And
If rich (more fuel + less air):
Easy to ignite (sufficient energy)
More toxic and low economy
Sufficient power (complete
But easy to ignite and more
combustion)
powerful

Figure 7 Comparison of homogenous and stratified charges

CNG engine. A stratified charge was realized with a narrow-bowl piston, which could

deliver a better mixture.

Chiodi et al. [42] studied the effects of injection timing, types of injectors, and piston

geometry on the formation of a mixture in a direct-injection CNG turbocharged engine. Two

types of single-hole and multi-hole gas injectors were evaluated. Flat and modified piston

geometries with different nozzle types were compared. A stratified charge was realized with

both the geometries; a multitask injector produced the best fuel mixture near the spark plug.

For a homogeneous mixture, injectors with single and multiple holes have the same lean limit
29
(1.25); however, for a stratified mixture, an injector with several holes can create an ore-

convenience shape for the inevitable flame propagation [42]. A better-indicated efficiency

was obtained in the case of the stratified mixture compared to the homogeneous mixture [42].

In their study on the effects of injection and spark timing on combustion, Yadollahi and

Boroomand reported that stable combustion can be achieved with early injection in the case

of a stoichiometric mixture and that a delayed injection facilitates more stable combustion

for ultra-lean mixtures (λ > 2.0) [40]. Delayed injection timing in a single-cylinder direct-

injection CNG engine with a low compression ratio could improve the engine airflow by up

to 10%. However, this is not always a problem, particularly when the engine performance is

reduced by 4% at high speeds (5000 rpm). The late injection is limited because of the need

for injection advance to achieve appropriate mixture preparation at high speeds [43].

The influence of injection time was studied in the case of a CNG engine with a

compression ratio of 14:1 [44], and a similar trend was noted. It was concluded that late

injection (120° BTDC) gives a 20% better performance compared to early injection (300°

BTDC) at low engine speeds of up to 2750 rpm, whereas early injection gives a good

performance at engine speeds above 4500 rpm. The late injection gave a higher heat release

rate, better combustion efficiency, and a shorter combustion time owing to the better

stratification, at 2000 rpm. It leads to higher NOx and lower CO emissions, indicating a

higher combustion temperature and more complete combustion. It produced lower maximum

pressures compared to early injections, at high speeds. A high degree of stratification leads

to incomplete mixing and a significant cyclic difference, which explains the lower peak

pressure at low engine speeds.

2.3.3 Ignition principles

The fuel property directly affects the formation of the mixture in the combustion chamber.

In an engine with SI in the immediate vicinity of the spark plug, a sufficient amount of fuel

is required to ensure stable ignition and combustion [43]. Studies on natural gas combustion

using SI show that a change in the coefficient of excess air (λ) in the range of 1–1.4 in the

vicinity of the spark plug during ignition can provide stable combustion [43].
30
 Mixture formation is not only an important parameter for burning, but also for the ignition

system. A new direct fuel injection (DFI) approach, namely the spark plug fuel injector

(SPFI), has been developed [10] to overcome the drawbacks of lean combustion. The SPFI

concept involves connecting the spark plug and injector into the existing hole of the spark

plug, thus allowing minor engine modifications. The SPFI is a simple device used for

converting a PFI to a direct injection for gaseous or dual-fuel operation. Figure 8 shows the

configuration of the SPFI. The SPFI produces a higher burning rate, improved volumetric

efficiency, and better mixing owing to the high gas jet velocity compared to injecting into

the manifold. However, this leads to a loss of pressure in the SPFI fuel line and a decrease in

the effective compression ratio.

Cordiner et al. studied partial stratification (PSC) natural gas engines at the University of

Rome Tor Vergata for an experimental and numerical analysis of nitric oxide formation. The

experiment was performed on a Ricardo single-cylinder research engine with a PSC spark

plug. The idea of this method is to provide a stable ignition core in the combustion chamber,

providing a rich air–fuel mixture next to the spark plug (the amount of fuel was less than 5%

of the main charge) and an ultralight homogeneous charge to the main chamber. PSC offers

load control without throttling by increasing the lean flammability limit. This method

provides increased efficiency and fewer NO emissions under partial loading compared to

conventional lean-burn SI engines [45].

Figure 8 Partial stratification spark plug injector [46]

31
 Taking a step forward, using a prechamber for initiation is a potential method to improve

the ignition of a lean burn. A prechamber jet ignition system uses a chemically active

turbulent jet to initiate combustion in lean fuel mixtures in the main chamber.

In the case of a jet, which acts as a distributed ignition source, the charge combustion in

the main chamber is less dependent on the AFR. Multiple hot ignitions make the distance of

the flame relatively small, providing a short combustion duration, even in conventional slow-

burning lean mixtures.

32
3. Alternative approaches for mixture stratification
3.1. Historical background
The prechamber concept, as shown in figure 9, was first proposed and patented by Sir Harry

Ricardo in 1918 [47]. It is one of the first concepts of a stratified charge with a divided

chamber for SI engines.

Figure 9 Concept of the prechamber patented by Ricardo [47]

In this system, an enriched mixture is sent to the prechamber with one carburetor. Fresh air

is supplied to the main chamber, and the collector of the main chamber is unregulated. The

output power is controlled by changing the concentration and the amount of enriched mixture

supplied to the prechamber. Despite its good performance, the system has some drawbacks

under partial loading, such as low output and efficiency. The operation of the auxiliary inlet

valve was sensitive to the speed and load, leading to a low performance at certain speeds.

Figure 10 shows studies conducted on prechambers over the years.

33
 Figure 10 Number of studies conducted on prechambers over the years

In 1968, Gussak proposed an application for igniting a jet, called the LAG process

(Lavinia Aktivatisia Gorenia or avalanche activated combustion), based on the theory of

chain branching developed by Semenov, where he proposed using a rich mixture in the

prechamber which, when ignited, produced active species and “chain carriers” from the

incomplete combustion in the prechamber. These active particles and “chain carriers” are

then introduced into the main chamber, quickly promoting chain branching reactions [48].

During jet ignition, the active radicals present in the partially burnt product are ejected from

the prechamber, helping to initiate charge combustion in the main chamber via providing

several distributed ignition sites for quickly consuming the charge in the main chamber.

Gussak found that in LAG, the optimal size of the prechamber is 2–3% of the lumen volume

with a hole length-to-diameter ratio of 0.5 [48]. Owing to Gussak’s extensive research, the

importance of active radicals in this type of ignition system has been revealed, and LAG was

applied to the transmission of the Volga passenger vehicle in 1981, yielding an enriched (λ

= 0.5) ultralight (λ = 2) mixture in the cylinder [47]. Figure 11 shows the different types of

prechamber designs.

34
 Fuel Injector
Spark plug
Pre
chamber
Main Main
Spark
combustion combustion
plug

Piston Piston

Piston
a) Turbulence generating torch cell b) Auxiliary fuel injector

Auxiliary valve
Rich mixture inlet

Main
Spark plug combustion

Piston

c) With auxiliary air inlet valve

Figure 11 Types of prechambers[49]

Further LAG studies were conducted by Yamaguchi et al. [50] in the 1980s at the Nagoya

Institute of Technology in Japan, where the effects of prechamber size and diameters of the

holes on the ignition and combustion in a bomb with a divided chamber were investigated.

They argued that the ignition process can be divided into four types:

1. The well dispersed combustion with small diameter openings caused a long induction

period. The main chamber ignites and burns out quickly because of chemical reactions rather

than thermal reactions.

2. A composite ignition with a slightly larger hole than in (1) caused a shorter induction

period. Active radicals and thermal effects are responsible for ignition, where flame nuclei

contributed to combustion in the main chamber. This model has been identified to be the

most effective for lean conditions.

3. The ignition of the burner with a flame core was further expanded relative to that in (2)

but was large enough to create a turbulent jet that absorbed most of the mixture in the main

35
chamber. The ignition is only due to flame nuclei. Figure 12 shows the stratification

principles.

STRATIFICATION PRINCIPLES

Flow
GDI Prechamber optimization

Durability
Fueled Un-Fueled
issue

Cost
Expensive
effective

Figure 12 Stratification principles

4. The ignition flame of the front burner with the largest hole: The prechamber flame enters

the main chamber in the form of a burner, and the combustion of the main chamber occurs

as normal flame propagation.

The study confirmed the possibility of effective ignition of a combustible mixture

exclusively with active radicals. However, it was concluded that for lean combustion,

composite ignition was the most efficient method of combustion [50].

Another study used torch-cell engine designs that simplified the earlier design of a split

chamber stratified by Harry Ricardo, eliminating the need for additional refueling in front of

the chamber. In the constructions of the burner cell, the prechamber containing the spark plug

is filled with a fresh charge intended for the main chamber during the compression stroke.

During ignition, a turbulent burn is created, which then ignites the charge in the main

chamber. The ignition of the burner produces more than 20 times the energy required to ignite
36
air/fuel with a ratio in the range of 12:1–18:1, allowing an extremely lean mixture to be

ignited in the combustion chamber. Because of the high turbulence in the prechamber during

ignition, a rapid combustion occurs. The turbulence due to the chaotic motion produced an

unburned mixture as a source of ignition. Therefore, as the intensity of turbulence increases,

there is more contact between the unburned mixture and ignition resources, while the average

temperature of the combustion process decreases. The burner ignition system provides the

following combustion characteristics in the combustion chamber: a) The presence of unstable

and chemically active products provides enough energy to ignite lean mixtures. b) The jets

due to the combustion in the prechamber create turbulence, leading to further turbulence and

a decrease in the temperature gradient in the chamber. Figure 13 shows the design of a

turbulent burner engine produced by Toyota. The torch cell system extends the operating

range of the engine and was introduced and developed not only by Toyota, but also by several

other car manufacturers such as Ford and Volkswagen [33].

Unlike burner cells, an engine with a divided chamber and a stratified charge comprises

an additional fuel source in the prechamber. In earlier designs, the prechamber fuel was

supplied through a third valve and an additional carburetor, whereas in existing designs, an

additional fuel injector is used to supply fuel to the prechamber. Two separate interconnected

chambers provide charge stratification, particularly a rich mixture near the spark plug to

increase flammability. The initial designs of the divided chambers were characterized by

large prechambers and openings, in which case, when ignition starts in the prechamber, a

regular flame front slowly passes through the hole into the main chamber. Honda developed

a system based on this technology, called the compound vortex-controlled combustion

(CVCC), which meets the 1975 emission standards, with a catalytic converter. This system

relies mainly on a flame torch for igniting the main mixture, leading to normal flame

propagation [48].

37
 Figure 13 Turbulence-generating torch design 2Sy Toyota [33]

Several methods have been developed with a similar combustible mixture around the

ignition source in the prechamber and an ultra-lean mixture in the main chamber, including

reactive igniters. Jet ignitors form a subgroup of the concept of divided chambers with a split

charge, but with a much smaller hole(s) connecting the prechamber and the main chamber.

A smaller hole produces a high-speed burning jet, which quickly passes through the hole and

penetrates deeper into the main chamber. Because of the high speed of the jet, the chamber

volume should be relatively small so as to avoid contacting the walls of the combustion

chamber.

3.2. Development of the prechamber

Prechamber ignition systems are common in stationary gas engines with SI, where this

ignition system can be optimized for certain operating conditions of the chamber. Unlike a

typical SI engine, a gas engine with a prechamber exhibits significant improvement owing to

the stable ignition and high propagating speed in the prechamber with a high flaming rate

and a short combustion phase in the main chamber. In a mobile device, there are several

issues when using a prechamber ignition system for lean combustion [51], including:

1. After combustion cycles, some amount of unburned gas remains in the

prechamber as residue.

2. High heat transfer over the elements of the prechamber

3. Losses during scavenging

38
 Studies on prechambers for SI engines have focused on overcoming these shortcomings.

Owing to the research and development of prechamber applications, these shortcomings have

been alleviated [51].

In France, Robinet et al. [52] introduced another firing concept called APIR, which means

“Auto-inflammation Pilotée par Injection de Radicaux” or spontaneous combustion initiated

by radical injection using a homogeneous propane–air injection mixture. The APIR design

uses a highly stratified mixture launched with either gas–air or gasoline–air near the upper

limit of flammability into the prechamber. This ultra-rich mixture is injected directly into the

prechamber with a slightly leaner mixture made from the flow of a certain lean mixture from

the main chamber during the compression stroke. In addition, the combustion residue in the

prechamber does not affect the start of ignition, since the concentration of the air–fuel

mixture is not sufficient. An incomplete combustion of the enriched mixture leads to the

formation of an intermediate combustion residue and contributes to an increase in the

prechamber pressure. The only difference between APIR and pulsed jet combustion (PJC)

lies in the diameter of the holes linking the prechamber and the main chamber; the hole size

in the case of APIR is less than 1 mm. Smaller diameters were used for the following reasons:

1) Eliminating the spread of flame and avoiding the re-occurrence of combustion in the jet

vortex; 2) Formation of multiple holes in the narrow space, causing multiple radical seedings

in the main chamber; 3) Restricting backflow and delivering high-pressure radicals away

from the head of the prechamber. The configuration of the APIR device is shown in figure

14(e). Figure 15 shows the principles of a fueled prechamber.

39
 a) Jet ignition Stratified Charge b) Ignition by Avalanche activation c) Torch ignition (TI).
Engines, Davis, 1974 (LAG), Gussak et al,1975 Adams,1979

d) Hydroren Assisted Jet Ignition e) Self Ignition Trigered by Radical f) Bowl Prechamber
(HAJI), Lumsden and Watson, 1995 Injection, Robinet et al, 1999 ignition(BPI), Kettner and Rothe,
2005

g) Turbulent Jet Ignition(TJI), h) Compound vortex controlled i) Stable Kernel of Combustion

Toulson et al., 2012 combustor(CVCC), Date et al 1974 (SKC), Garret, 1975
H2

j) Volksvagen PCI system, k) Ford FROCO engine, Scussel, l) Turbulence Generatinng Pot,
Brandsteter, 1975 1978 Konishi, 1978

Figure 14 Types of prechamber systems [11]

40
 Figure 15 Principles of a fueled prechamber

An API prechamber (sub-chamber) is installed at the location of a typical spark plug. The

prechamber size was set approximately 1% of the main chamber with heads, with one to nine

holes with a diameter ranging from 0.5 to 0.8 mm. The ignition timing for the maximum

brake torque (MBT) of a conventional spark plug is approximately 50°, immediately reduced

to 25° for a device with a pulsed jet chamber (PJC) and up to 20° for the APIR device, which

implies a 20% less ignition time than a PJC device.

The APIR prechamber exhibits a rapid decrease in the cycle-to-cycle variation compared

to a regular spark plug. With an indicated mean effective pressure (IMEP) of 0.3 MPa, the

APIR prechamber used in the lean burning strategy showed a 95% reduction in the fuel

consumption and a 63% reduction in NOx and CO emissions compared to a regular spark

plug. The drawback of this device is the discharge of unburned hydrocarbons, which

increases by 145% compared to using a regular spark plug, primarily because of the trapped

hydrocarbons in the sub-chamber. Nevertheless, the captured unburned hydrocarbons

(UHCs) are more localized in the sub-chamber of the APIR, making it simpler to reduce them

[52].

41
3.3. Radical ignition
The radical ignition (RI) method uses two combustion chambers: a main chamber and a sub-

chamber. The sub-chamber is located above the main chamber. Figure 14(e) shows a

schematic of RI for SI engines. The rapid combustion is accomplished by emitting

combustion products with high temperature and energy density, including various types of

active radicals in the sub-chamber, through the openings in the main chamber to increase the

flammability of the mixture in the main chamber. The turbulence, which leads to multipoint

ignition and expansion of the early flame, allows a fast burn of the lean mixture in the main

chamber. This method provides a noticeable improvement in the burning rate and

flammability limit compared to the SI system [53], [54]. The disadvantage of this method is

that it cannot clear the residual gas from the previous cycles in the sub-chamber, giving rise

to cycle instability.

The radical ignition method was further validated by Park et al. using a constant volume

chamber (CVC). The progress in the lean burning of the sub-chamber was examined by

introducing the active radicals generated in the CVC sub-chamber by changing the amount,

total cross-sectional area, and diameter of the passage openings. They claimed that the

optimum size of the sub-chamber was approximately 0.11 cm–1 in relation to Ah/Vs, i.e., the

ratio of total area of the holes (Ah) to the volume of the sub-chamber (Vs). In comparison to

the SI technique, the RI method facilitated improved flammability and burning rate [53].

Ha et al. demonstrated the RI technology with the sub-chamber for converting a single-

cylinder CNG-powered diesel engine with direct fuel injection. The sub-chamber contains a

spark plug and an injector, and CNG fuel is injected into the sub-chamber. The sub-chamber

body functions as a ground electrode during the ignition. The RI-CNG engine overcomes the

disadvantages of the RI-gasoline engine in that it can drain out the residual gas from the

previous cycles in the sub-chamber and increase the cycle stability of the engine. However,

this method presents its own drawbacks because of the stoichiometric characteristic: it results

in a high emission of CO and NOx and an increase in the coefficient of variation in pressure

42
(COVp) at high engine speeds. Different performances in the lean state lead to a decrease in

COVp, but an increase in the combustion duration [55].

3.4. Two-stage combustion systems

A stratified mixture in the prechamber of two-stage combustion systems is an effective way

to improve the burning process of a lean mixture. The combustion of an engine with a

homogeneous mixture proceeds at λ = 1.6 because of the detonation limit. In the case of a

stratified (heterogeneous) mixture, the rich mixture in the prechamber and the lean mixture

in the main chamber help extend the range of lean combustion (λ = 2) in both the chambers.

This process includes two sections: the main chamber and the prechamber. An ultra-lean

mixture is sucked into the engine inlet, while a rich mixture is injected into the prechamber.

Through the compression cycle, the lean mixture in the main chamber is delivered to the

prechamber to attenuate the enriched mixture. Thus, a stratified charge concept wherein a

rich fuel mixture is injected near a spark plug with a lean mixture in the combustion chamber

is applied. Three key approaches are used in these systems: 1) Direct injection of fuel as an

ignition source to avoid detonation; 2) Use of a spark plug as an ignition source when the

fuel is mixed with air for ignition time monitoring; 3) Regulation of engine power by

changing the amount of fuel injected per cycle [49]. An experiment was conducted to

facilitate two-stage combustion of a gaseous engine converted from a diesel engine (CR 8)

to take place with the use of a spark plug. The results showed complete lean combustion with

an average combustion coefficient of up to 2.0 and a reduction in NOx emissions. However,

there was a decrease in the IMEP and an increase in the HC emissions in this case [56]. The

investigation was conducted by A. Jamrozik and W. Tutak [57] on two-stage liquefied

petroleum gas in comparison with single-stage gasoline fuel. A stratified gas charge two-

stage combustion system was proposed by Jarnicki at al., [58] for a sectional chamber with a

constant volume to study the ignition of gaseous fuel. A jet of fuel gas was introduced by

means of a prechamber into the main combustion chamber through openings. The

43
experimental and numerical analyses showed that the prechamber helped overcome the

difficulty in igniting a gaseous fuel with direct injection.

3.5. Swirl prechamber spark plugs

In the early 1980s, in an attempt to simplify the LAG method by eliminating the additional

prechamber for fuel–air mixture, Reinhard Latch at Bosch Stuttgart examined a swirl

prechamber spark plug. This approach relied on a piston for compressing the air–fuel mixture

from the main chamber to the prechamber [22], [51], [59]. Figure 16 shows a prechamber

spark plug [60]. The miniature volume swirl chamber (red in figure 16) is situated within the

14 mm spark plug.

Spark plug

Prechamber

Figure 16 Schematic and example of a pre-chamber spark plug [22]

Latsch et al. extended the swirl prechamber to the concept of chamber ignition, for which the

term “bowl prechamber” is used [59]. This method involves a direct injector, a piston bowl,

and a prechamber spark plug. Injection occurs in two stages: the first is in the intake stroke

to obtain a lean mixture in the cylinder. The second stage is in the compression stroke, where

approximately 3% of the total fuel amount is injected into the piston bowl for transportation

to the prechamber by the piston. The rich prechamber mixture is delivered to the prechamber

and ignited by the spark plug, creating a jet of flame that burns the lean mixture in the cylinder

of the main chamber. The bowl-prechamber ignition reduced NOx emissions and partial load

fuel consumption with a decrease in the detonation at full load.

44
 Geiger et al. evaluated the performance and capability of swirl prechamber spark plugs

[61]. They concentrated on various ignition systems for highly diluted mixtures in SI engines.

They argued that this structure is not conducive for improving combustion near the lean

flammability compared to a conventional spark plug. Although flame ignition using a swirl

prechamber spark plug contributes to a rapid conversion of energy, the trapped residual gas

along with the air–fuel mixture inside the swirl prechamber makes ignition difficult around

the lean limit.

In a research on highly diluted mixtures for an SI engine, where the aim was to improve

combustion, a spark plug with a swirl prechamber scavenged with methane was studied, and

a comparison was made with a typical spark plug and a swirl prechamber without scavenging

[61]. Three systems were compared in terms of the performance at 2000 rpm and IMEP of

280 kPa; the ignition of the jet using a swirl prechamber spark plug led to quick energy

conversion; however, there was no improvement in the flame ignition because of the residual

gas inside the swirl prechamber. With the swirl prechamber scavenged with methane, the

lean limit was increased. The ignition enhancement was attributed exclusively to the

stratification of the rich charge mixture in the prechamber, with the lean limit increased to λ

= 1.96.

3.6. Prechamber pilot spark plug

Latsch et al. [51], [59] performed another study on a prechamber spark plug where the

prechamber spark plugs were combined with pilot injection. The main objective of this

system was to produce an ignition region with a low requirement for ignition energy, in order

to minimize the constraints of a typical ignition system while igniting the lean mixture by

stroking the flame in the cylinder. This idea was realized thanks to the discovery made by

Bowing et al. [61], who stated that the earlier design exhibited an ineffective purification of

the residual gas in the prechamber, leading to an inadequate arrangement of the mixture in

the prechamber. The main difference between this concept and other methods is that fuel is

45
injected in the intake stroke and not in the compression stroke, aimed at pushing the residual

gases out of the chamber.

Spark plug
Pilot injector

Prechamber

Figure 17 Principle of pre-chamber spark plug with pilot injection [51]

Figure 17 schematically shows the operation of the ignition system with a small amount

of fuel injected into the prechamber. Before pilot injection, a huge amount of residual gas

remains in the prechamber at the end of the combustion cycle. The pilot fuel is then

introduced in the intake stroke, which helps blow the residual gas into the main chamber.

During the compression stroke, the fresh mixture enters the chamber, forming a flammable

mixture.

The high pressure in the small prechamber directs the flame in the form of burners into

the main chamber and ignites the lean mixture [51]. Pilot injection in the prechamber can

stabilize and extend the lean limit, thus significantly reducing the fuel consumption and nitric

oxide emissions compared to the absence of pilot injection in the prechamber.

3.7. Unfueled Prechamber

In the prechamber, the spark plug electrode is located at the rear of the prechamber to

direct flame propagation towards the vias to facilitate ignition. Figure 18 shows the principle

of unfueled prechamber combustion. The main principle involved here in that the unfueled

prechamber is charged with the inlet flow of the main cylinder during the inlet and

compression stroke of the engine cycle.

46
 Pre
chamber

Main chamber

Figure 18 Principle of unfueled prechamber combustion

3.8. Turbulent jet ignition

Bunce et al., developed a replacement for the conventional spark plug used in a port-

injected engine [62], and this engine is termed the turbulent jet ignition. It was designed to

overcome the drawbacks associated with lean burning. It uses hydrogen as the fuel in front

of the chamber and is compatible with gaseous as well as liquid fuels such as gasoline. Figure

19 shows a cross-sectional illustration of a jet igniter used in a PFI engine.

DI Inductive
gaseous (35mj)
PFI gaseous

Prechamber

Figure 19 Spark ignition (SI) and pre-chamber turbulent jet [17]

47
 The prechamber nozzle generates 2% of the total energy in the prechamber, with the

remaining 98% being supplied through the filling channel for the main chamber. The spark

plug inductive shown in figure 19 is used as an ignition source in the turbulent jet igniter. A

tiny prechamber with a volume of ~2% of the clearance space (main chamber) was chosen

to decrease the volume, HC emissions, heat loss, and effect of the surface on the volume and

residual gas in the chamber. It is linked to the main chamber through single or multiple holes

of diameter ~1.25 mm, which contributes to the extinguishing of the flame and penetration

into the main chamber, and because of the chemical, thermal, and turbulent effects, the

combustion products in the prechamber initiate a combustion in the main chamber at several

locations.

The characteristics of an engine with gasoline injection using a turbulent reactive igniter

with various prechamber fuels at 2500 rpm in the WOT mode. The lean limit was extended

via the ignition of a turbulent jet with further upgrades, with increasing compression ratio. A

comparison of the turbulent jet ignition with a conventional spark plug showed an increase

in the peak thermal efficiency of 11% and is predicted to reach an indicated net thermal

efficiency of 45% (19% relative growth) at a higher compression ratio (~14). In addition,

when a reactive turbulent igniter runs on gasoline (main chamber) and propane (prechamber)

and operates under WWMP conditions at 1500 rpm and IMEP of 3.3 bar, the prechamber

combustion system can provide up to a 54% mass fraction diluent while maintaining

satisfactory combustion stability and ensuring an 18% enhancement in fuel consumption

associated with a gasoline ignition system [63]. Jet turbulent ignition of gasoline and natural

gas demonstrated an identical improvement in fuel consumption, with a further increase of

up to 20.6% using pure natural gas at an IMEP of 4.7 bar compared to the gasoline model.

When comparing ignition systems, the turbulent jet ignition pre-chamber system exhibits

significant advantages over the SI system, with optical images revealing stable combustion

past lambda 1.8 resulting in the near elimination of in-cylinder NOx emissions and significant

improvement in terms of the efficiency and fuel economy [17].

48
3.9. Direct injection jet ignition (DI-JI)
Direct injection helps reduce the negative effects on the volumetric efficiency, which

eliminates power loss and increases the calorific value of the mixture while providing higher

resistance from detonation, better efficiency, and flammability. The combination of direct

injection and jet ignition has become another promising approach for SI engines [64]–[67].

Professor Harry Charles Watson and Alberto Boretti established and presented a relationship

between direct fuel injection in the main chamber and volumetric ignition of several jet hot

gases from a small prechamber, in which a second fuel injector and spark plug are placed.

The direct injection combined with the jet ignition in an SI engine allowed more efficient

and complete combustion of gaseous and liquid fuels. This combined approach consists of a

direct fuel injector in the main chamber and one preliminary jet ignition chamber in each

cylinder for a multi-cylinder engine. The concept of direct injection and jet ignition in a four-

stroke engine is shown in figure 14(i). The preliminary jet ignition chamber contains a spark

plug and one prechamber injector with six equally spread nozzles with a diameter of 1.25

mm and a net volume of less than 1.5 cm3.

The direct injector in the main chamber injects fuel into the cylinder to obtain a lean

mixture, which consists of air and residues from previous cycles. In the prechamber, fuel is

injected through a second direct injector and ignited by a spark plug. The jets are then ignited

by an inhomogeneous mixture in the main chamber through openings. The authors [64]

emphasized the benefits of combining direct injection and jet ignition. Parallel reviews on

homogeneous DI or PFI and stratified DI and jet ignition indicated that this model offers a

more complete, greater lean limit. The approach increases the combustion rate with high

ignition energy, because of the huge amount of partially burnt combustion products at several

locations in the main chamber, and decreases the heat loss to the wall of the main chamber

[64]. Heat losses can be reduced with better distribution of the fuel in the lean mixture of the

main chamber, combusting mixture in the main cylinders in the entire volume, and the

presence of high ignition energy in the main chamber provided by the multi-point sides of

the ignitor. The combination of jet ignition and direct injection provided higher efficiency:

49
approximately 50% at full load with minor imperfections at partial load and a more ecological

ultra-lean combustion of the gaseous fuel mixture.

This advanced system can improve the full load process of a stationary or mobile engine

owing to the high brake efficiency (the ratio of engine brake power to total fuel energy),

which decreases the specific fuel consumption (ratio of engine fuel consumption to brake

power). At partial load, the engine runs at almost throttle-free power, as the load is controlled

by the amount of fuel injected. This results in an efficient combustion of the fuel mixture in

the range from close to stoichiometric to ultra-lean [65]–[67].

4. Effect of pre-chamber systems on SI engines

Some studies [68]–[72] have confirmed that the requirement for more energy to initiate

burning and low flame propagation speeds are the main problems faced when working with

lean mixtures. Prechamber ignition systems can improve the lean ignition because of the high

energy in the main chamber at the start of combustion. The increase in the available energy

for igniting the mixture directly influences the lean limit, ignition time, start of combustion,

speed of flame propagation, and rate of heat release. These essential combustion parameters

are discussed in the following sections.

4.1. Flammability limit

Most of the benefits of prechamber ignition systems are related to the increased flammability

limit of the air–fuel mixture, supporting combustion quality and repeatability in

stoichiometric mixtures when using dry mixtures in the main chamber. Moreover,

prechamber systems facilitate a significant increase in the fuel economy and demonstrate low

nitrogen emissions [51]. Experimental results were obtained in a study using the cooperative

fuel research (CFR) engine. Wimmer and Lee stated that the prechamber system shows

markable improvements in the lean combustion both before and after warm up [73]. The

50
lean combustion can be achieved outside of the normal flammability limit using a split-

chamber, according to Yamaguchi et al. [50].

Toulson et al. [51] studied a hydrogen-assisted jet ignition (HAJI) system for a CFR

engine with a gas mixture in the main chamber and various gaseous fuels in the prechamber.

They reported an increase in the flammability of this system. The results showed that

hydrogen (H2) extends the main chamber depletion limit to an air–fuel equivalence ratio (k)

of 2.5, whereas liquefied petroleum gas (LPG), compressed natural gas (CNG), and carbon

monoxide (CO) extend the lean limit to 2.35, 2.25, and 2.15, respectively. This range of

permissible operating limit of the engine indicates that an increase in the fuel ignition level

in the prechamber depends on several factors, including the speed of flame propagation and

the formation of chemically active products of combustion, and these are not only related to

the amount of energy in the fuel [17]. Later, Toulson et al. [74] discussed the use of LPG in

both the main and prechambers, allowing a single fuel system. The effects of LPG or gasoline

in the main chamber and H2 or LPG in the pre-chamber were analyzed to determine the effect

of fuel on the depletion limit of the mixture, emission levels, and combustion characteristics.

LPG–LPG and LPG–gasoline systems gave very similar results, and the difference between

them was within the experimental error.

Toulson et al. [75] reported that lean threshold variations were higher at low absolute

reservoir pressures (MAPs) than at high MAPs; for all fuel combinations, the lean limit was

extended to a value slightly higher than k = 2.5.

Attard et al. [76] compared the characteristics of ignition and combustion in an optical

single-cylinder engine running on natural gas in the main chamber. They studied the use of

a stratified TJI system and the provision of information about the ignition process. The base

lambda was increased from 1.3 to 1.8 with the introduction of the TJI system, which allowed

achieving an acceptable combustion stability. The results showed that the use of TJI with

more fluid mixtures resulted in a brighter and more intense blue flame than in the base engine.

The authors attributed this to the increase in the rate of heat release, which indicates a more

stable combustion.

51
4.2 Spark time
Combustion gases require a period for expansion, and angular or engine rotation speed may

increase or decrease the time frame in which combustion and expansion should occur [77].

Hence, the need to advance the sparking time is associated with unburned fuel at the time of

spark fires. The system reduces the progress of the ignition, which can be explained by an

improvement in the flame propagation speed.

Some of the combustion characteristics of a torch-ignited engine were compared with

those of a conventional one. An experimental analysis was performed to determine the effects

of prechamber inlet air–fuel ratios and nozzle diameters; these are respectively the main

operating factor and main design factor of torch-ignited engines. These factors control the

torch combustion process and determine the engine performance and emission characteristics

[78]. The combustion rate was significantly increased when using the prechamber, where the

ignition advance for maximum power was 2 to 3 times lower than that in an engine without

a prechamber.

Ryu et al. [16] studied the effects of a prechamber on the spark synchronization of a

uniform ignition system, and compared it with the direction of inlet jet in the MBT state. The

findings indicate that the ignition time was minimum when the angle between the jet and the

horizontal piston surface was 90°. Robinet et al. [52] studied a stratified APIR pre-ignition

system by comparing the operation of an SI base engine and a system with the APIR device.

The optimization of the ignition timing for a typical combustion case for an MBT at 32

BTDC, down to 10 BTDC with APIR. Roethlisberger and Favrat [79] observed one of the

most significant consequences of changing the ignition delay—a sharp decrease in the

combustion temperature. Pishinger [9] reported a lower combustion temperature; however,

the decrease had a negative effect on the average value of the braking efficiency pressure

(BMEP) and specific fuel consumption during braking (BSFC). The author studied the effect

of ignition changes on the IMEP and combustion stability at 1500 rpm, with k = 1.8, using

the TJI system with a constant flow of air and fuel into a single-cylinder engine [80]. The TJI

combustion system was insensitive to the ignition changes compared to a conventional SI

52
engine. A low IMEP variation was observed, approximately 1% in the range of 30° from the

moment of ignition, and combustion stability in the range of 40° from the moment of ignition.

4.3. Start of combustion

The length of time, expressed in terms of the angle of rotation of the crankshaft (CA) from

the moment of ignition to the moment when the pressure increases due to combustion, was

considered by Sakai et al. [81] as the ignition delay.

Experimental work on torch-ignited engines demonstrated that this type of engine

provided stable engine performance and low exhaust emissions owing to the overall lean

mixture. Some of the combustion characteristics of the torch-ignited engine were compared

with those of a conventional one. An experimental analysis was performed to determine the

effects of prechamber inlet AFR and nozzle diameters. As mentioned previously, these are

respectively the main operating factor and main design factor of torch-ignited engines. These

factors control the torch combustion process and determine the engine performance and

emission characteristics [84]. Ryu and Asanuma [82] noticed that the ignition delay in a

prechamber ignition system was 11 CA in all A/F ranges, whereas the ignition delay in

conventional engines is longer and gradually increases with increasing A/F. The onset of

combustion can also be analyzed using the mass fraction of burned (MFB). It represents the

fraction of energy released from the combustion of the fuel to the total energy at the end of

the combustion process, determined from the analysis of the pressure in the cylinder.

53
5. Main principles and parameters of the pre-
ignition chamber
The ability of a jet engine pre-ignition chamber to act as an ignition system depends on

various key parameters, including the volume, quantity, diameter, and length of the hole(s)

and the presence of auxiliary fuel injection in the prechamber. The volume of the prechamber

has a significant influence on the speed and depth of penetration of the jet and therefore on

the speed of turbulence in the main chamber. Large volumes give a greater impulse, thereby

increasing the penetration rate. Shah et al. [83] noted a decrease in the angle of flame

development and combustion duration in the main chamber with an increase in the volume

of the prechamber from 1.4 to 2.4%. An increase of more than 2.4% in the prechamber

volume decreased the efficiency of the prechamber as an ignition device. They argued that

the tilt limit in the main chamber increases with the prechamber volume without significantly

affecting the nozzle diameter; however, for a given prechamber volume, the smallest

diameter (1 mm) of the nozzle resulted the best ignition characteristics in terms of the

minimum combustion duration and the smallest angle of flame development.

They compared and simplified the main characteristics of the prechamber. Most

prechambers occupy a volume of approximately 2% of the gap or less, with a multi-hole

prechamber design.

Moreover, a single-hole prechamber requires a larger hole to allow the air–fuel mixture

to be supplied from the main chamber to the prechamber during compression in the

prechamber without any auxiliary fuel.

In an attempt to understand the characteristics of an effective prechamber, Shah et al.

studied the effects of nozzle volume and diameter through experimental and simulation

analyses [83]. They concluded that increasing the prechamber volume increases the absolute

mass flow rate from the prechamber while decreasing the nozzle diameter and increasing the

jet speed. A comparison of the nozzle diameter at a constant volume showed that a smaller

nozzle causes an earlier jet ejection, likely because of the effect of better mixing in the

prechamber with fresh air during the compression stroke. A larger nozzle diameter leads to
54
the formation of a more multilayer mixture in the prechamber, and combustion occurs at a

lower speed after SI. It is concluded that a smaller nozzle diameter leads to a faster

penetration into the combustion chamber solely because of the higher velocity and

momentum of the jet. In addition, the nozzle diameter has a direct effect on the average

turbulence in the main chamber, with a higher maximum level of turbulence being achieved

with a smaller nozzle diameter, and at a controlled volume, the maximum level and speed of

turbulence increase with decreasing nozzle diameter [83]. A single-nozzle prechamber

experiment without auxiliary fuel injection was conducted by Gentz et al. [84] in a rapid

compression machine (RCM) for analyzing the effect of prechamber nozzle diameter on the

burning time (0–10% and 10–90%) of a premixed propane/air mixture in terms of the AFR.

For three different nozzle diameters (1.5, 2, and 3 mm), a hole size of 1.5 mm resulted in a

shorter burning time from 0 to 10%, with fastest ignition of the flame, as shown in Fig. 20.
0% to 10% Burn Duration, (ms)

0.75 1 1.25 1.5 1.75 2

0
2
4
6
8 SI
D3.0
10 D2.0
12 D1.5

14
16
λ

Figure 20 Variation in 0–10% burn duration with λ for all TJI nozzle diameters [84]

It was concluded that the diameter of the hole has a negligible effect on the burning time

in the range of 0–90% under stoichiometric conditions. The effect was significant only under

lean conditions, in which case the smallest hole resulted in a more rapid flame spread

compared to employing the largest hole. They found that the ignition of a turbulent jet does

not extend the misfire limit compared to the basic SI test. This is because of the combination

of additional heat transfer losses of hot combustion products to the hole wall and additional

cooling of the jet via mixing with the cold unburned mixture in the main chamber.
55
5.1 Preliminary study of the prechamber
A comprehensive survey on available literature regarding ignition and mixture formation

strategies applied to SI engines showed that employing a prechamber can improve the fuel

conversion efficiency of SI engines. In this dissertation, the charging possibility of an

unfueled prechamber is analyzed using CFD to achieve efficient stratification for SI engines.

Modeling is carried out using a simple cold flow simulation. A cold flow analysis was

employed in this study because the detailed chemistry is much simpler than that in other

calculation methods. The purpose of the simulation study was to demonstrate the cold flow

due to the connection of the prechamber with the port injector, represented in terms of the

angle of rotation (CA) of the airflow in the main combustion chamber. In addition, the results

presented in this chapter demonstrate the potential of the chamber qualitatively, rather than

the engine; therefore, quantitative data on the operation are not given.

The modeling was performed using SpaceClaim, a solid modeling CAD software. The

CAD model was then imported to ANSYS [85], which is a widely used CFD platform that

can simulate an entire CFD process from CAD to post-processing in a single integrated

software environment. It is incorporated with the advantages of a built-in surface transfer

feature, advanced auto-binding, and adoption of a single code with minimal user effort.

ANSYS IC engines feature automatic meshing and comprehensive selections of the physical

models, whereby accurate solutions are provided in a convenient environment. SpaceClaim

is a detailed design package that can be connected to the ANSYS environment. There are

four methods involved in the CFD approach for the flow analysis of ICEs: in-cylinder

combustion, full-cycle, port flow, and cold flow simulations. These methods can differ

depending on the process being studied [86].

The flow in the port was analyzed by keeping the engine geometry frozen at a critical

angle of rotation selected by the user. This analysis is similar to the static analysis for any

case solved using CFD. The combustion process in the cylinders is analyzed only during the

expansion stroke. Full-cycle simulations provide a complete picture of the engine process,

56
including airflow, fuel injection, combustion, and formation of exhaust gases from a chemical

reaction.

Cold flow analysis is a transient simulation. A complete engine cycle can be simulated

but without a chemical reaction. The cold flow simulation can help visualize the airflow

throughout the cycle, capturing the induction of air and predicting the swirl formation,

tumble, and mass flows through specific cross sections. The mixing of the air with the

injected fuel can also be simulated but without the reaction. Since the thermodynamic

changes in the engine are neglected in the simulation, the flow characterized by the power

stroke and exhaust does not reflect the real conditions.

The mass, momentum, energy, k-epsilon of turbulence, and tumble ratios without fuel

injection or chemical reaction are described in the following equations:

𝜕𝑝
⃗)=0
+ ∇(𝑝𝑢 (2)
𝜕𝑡

𝜕𝑝 2
(𝑝𝒖
⃗ ) + ∇(𝑝 𝑢
⃗𝑢⃗ ) = ∇𝑃 − ∇ (3 𝜌𝑘) + ∇ ∙ 𝜎
⃡ + 𝜌𝑔 (3)
𝜕𝑡

𝜕𝑝 2
(𝑝𝐼) + ∇(𝑝𝐼 𝑢
⃗ ) = −𝑃(∇ ∙𝑢
⃗ ) − ∇ ( 𝜌𝑘) + ∇ ∙𝐽⃡ +𝜌𝜀 (4)
𝜕𝑡 3

𝜕𝑝 2 𝜇
(𝜌𝑘) + ∇(𝑝 𝑢
⃗ 𝑘) = − 3 ρk∇𝑢
⃗ +𝜎
⃡ ∙ ∇𝑢
⃗ + ∇ [(𝑃𝑟) ∇𝑘] − 𝜌𝜀 (5)
𝜕𝑡

𝜕𝑝 2 𝜇 𝜀
(𝜌𝜀) + ∇(𝑝 𝑢 ⃗ + ∇ [(𝑃𝑟) ∇𝜀] + 𝑘 [𝐶𝑧1 𝜎
⃗ 𝜀) = − (3 𝐶𝑧1 − 𝐶𝑧3 ) 𝜌𝜀∇ ∙ 𝑢 ⃡∙ ∇ 𝑢
⃗ −𝐶𝑧2 𝜌𝜀] (6)
𝜕𝑡

∫ 𝜌𝑣 ∙ 𝑑𝐴 = ∑𝑛𝑖=1 𝜌𝑖 𝑣𝑖 ∙ 𝐴𝑖 (7)

where ρ is the density, u is the velocity vector, P is the pressure, σ is the turbulent viscous

stress tensor, I is the specific internal energy, j is the heat flux tensor, v is the overall velocity

vector, and A is the area [85], [86], [87].

5.2 Fluid flow in the prechamber

It is important to understand the movement of fluid in the combustion chambers, since it has

a significant effect not only on the homogeneous mixing of the air and fuel, but also on the

57
possible stratification and efficiency of the engine. Heywood discussed some of the general

parameters including the mass flow, drop, and swirl in a fluid flow [88].

5.2.1 Turbulence and tumble

Engine turbulence occurs because of the high-speed process during each cycle of the engine.

Air enters from the engine in repeated cycles at a relatively high speed, forming a turbulent

flow inside the cylinder. Designers require engines that can operate under high turbulence

when the piston is near the TDC. Since ignition occurs near the TDC, the high turbulence

makes it possible to rapidly break up and quickly propagate the flame compared to low

turbulence. Previous researchers studied the geometry of the main combustion chamber and

the shape of the piston to create turbulence in the engine; these factors sequentially affect the

combustion process. Other key factors contributing to the turbulence are the piston speed,

inlet flow condition, and intake manifold design. When considering the speed of all parts and

flows, the turbulent kinetic energy is one of the useful parameters that can be calculated via

simulation.

5.2.2 Swirl and tumble

A swirl is defined as the rotational movement of air about the vertical axis of a cylinder

(Figure 21(a)). Tumble flow occurs about a circumferential axis near the outer edges of the

piston bowl. Tumble is typically measured experimentally using a steady flow rig at a

selected valve lift; however, the tumble ratio is specific to the installation design (Figure

21(b)). In an ANSYS IC engine, the tumble ratio is calculated using the following equation:
𝐿∙𝑡𝑎 2𝜋𝑛
𝑅𝑡 = ⁄ 60 (8)
𝐼∙𝑡𝑎

where 𝐿∙𝑡𝑎 is the magnitude of the fluid angular momentum with respect to the tumble axis,

and 𝐼∙𝑡𝑎 is the moment of inertia of the fluid mass about the tumble axis. In addition to the

tumble ratio, the cross tumble ratio affects the rotational flow along the axis perpendicular to

the tumble axis [39][14].

58
 a) Swirl flow b) Tumble flow

Figure 21 Principles of swirl and tumble flows

6. Methodology
The following are the objectives of this work:

a) Finding the possibility of directing richer mixture into the prechamber;

b) Ensuring a richer mixture inside the prechamber during compression;

c) Determining the influence of the design and position of the prechamber on charge

flow.

With the cold flow method, the changes in the temperature and pressure in the engine cylinder

can be analyzed without any reaction due to the fuel injection or SI. This gives us an idea of

the real charge flow state prior to the combustion process. A modification is made to an

existing engine model to develop two new inner edge designs while maintaining the

geometries of the other parts of the engine (Figure 22). This design adjustment was made on

previously modified models that can be fitted into the spark plug hole without changing the

design of the main chamber of the original engine. For the cold flow analysis, the initial

axisymmetric engine model was modified into a full-size 3D engine model to investigate the

charge flow through the inlet and outlet ducts of the prechamber.

Finally, the simulation results were compared with the experimental results [20].
59
 Piston
movement

Figure 22 Charge flow characteristics

Hence, the richer part of the mixture of the port-injected engine should be sent to the

prechamber during the inlet-valve opening period, considering that the simulation period

depends on this period. Because of the gravity and tumble motion driven by the piston

movement, the mainstream inlet char flow must move downward (blue arrow in figure 22).

The parameters and values are taken from an actual naturally aspirated 0.5-L single-cylinder

engine, which highly resembles the geometry of the initial model. The parameter values were

those that had to be inputted at the preliminary stage of the cold flow simulation before

inputting to the engine model.

60
6.1 Modeling the engine
The engine studied in this research was a four-valve pent-roof-type four-cycle gasoline

engine, as shown in figure 23, with its specifications listed in Table 1. The initial model used

was taken from the ANSYS tutorial, which was modified into two unfueled prechamber

models with different inner wall configurations while retaining the geometry of the other

engine parts.
Outlet valve temperature
Inlet valve temperature
27 °C
φ° 40 °C

Outlet pressure 100 kPa

Inlet Pressure 80 kPa Temperature 40 °C
Temperature 27 °C

γ°
Cylinder head Cylinder wall
Temperature 27 °C θ° temperature 45 °C

°
Piston temperature 45 °C

Figure 23 Configuration of an engine cylinder head with the prechamber located on the top
of the pent-roof-type combustion chamber (two valves each for inlet and exhaust). The
prechamber is highlighted in yellow. The angular positions of the prechamber are
represented by the precession angle (φ°), tilt angle (γ) in the Y-Z plane, and tilt angle (γ) in
the X-Z plane. The horizontal position is changed with respect to the y and x axes, and the
vertical position is changed with respect to the z-axis.

The prechamber design was previously modified based on the ease of retrofitting the

prechamber into the spark plug hole without changing the cylinder head shape of the real

engine. The computational model of the engine analysis includes the engine intake valve,

exhaust valve, and cylinder head with a prechamber, cylinder, and piston.

61
 Table 1 Engine specifications

Parameter Value
Connecting rod length (mm) 144.3
Crank radius (mm) 45
Piston offset (mm) 0
Engine speed (rpm) 2000
Minimum lift (mm) 0.5
Bore (mm) 84
Stroke (mm) 90
Clearance volume (mm3) 47,290
Prechamber volume (mm3) 635.197
Compression ratio 6:1

The initial engine model was symmetrical, but the modified position of the prechamber

makes it asymmetric. This is because the prechamber of the current design location is

eccentrically shifted closer to the inlet valves and tilted in the symmetry plane of the engine

with a precession angle related to the prechamber centerline.

All the parameters were inputted at the preliminary stage of the cold flow simulation

before inputting to the engine model. Figure 24 shows the simulation model of the

prechamber system. The prechamber consists of three parts: a prechamber body (pink-

colored area), an inlet duct (blue-colored area), and an outlet duct (green-colored area).

Figure 23 shows the boundary conditions for the inlet and outlet ducts of the engine. The

ANSYS workbench FLUENT ICE, developed by ANSYS Inc., was used for the numerical

analysis in this study.

62
6.2 Decomposing and meshing
To accurately visualize the flow inside the prechamber, the engine model was divided into

different zones, which provided better control of the meshing of the model. Figure 24 shows

that different zones have different qualities and shapes of the mesh cell.

Figure 24 Decomposition model of the engine with prechamber: Inlet-A, and outlet-B,
ducts in the main body of prechamber-C

The grid cell is particularly fine in the volume of the main chamber and prechamber of the

combustion chamber and the valve seat. For accuracy and computation speed, the moving

zones were associated with a hexagonal cell and static zones with tetrahedral cells.

Table 2 Model zones

Zone No. Decomposed Zone

1 Intake port
2 Exhaust port
3 Intake valve seat
4 Exhaust valve seat
5 Combustion chamber with prechamber
6 Cylinder layer
7 Piston

For the engine boundary conditions, the inlet and outlet temperatures were set to 313 and

333 K, respectively. The inlet and outlet pressures were 80 and 100 kPa, respectively. The

63
temperatures of the cylinder head, piston, and cylinder wall were 348, 318, and 318 K,

respectively. The model zones shown in figure 24 are explained in Table 2.

Once all the parameters were determined, the turbulent model k-ɛ with an enhanced wall

function was used as the flow model. When heat was transferred from the main cylinder to

the prechamber, the mass flowrate and velocity through the inlet and outlet ducts were the

primary output parameters to consider for charging the prechamber. The inner edge

configuration and the cross-sectional change in the inlet and outlet ducts of the prechamber

are considered the input parameters.

64
6.3. Grid test
A grid independence test was performed to determine the minimum number of grid cells

required to ensure that the cells are not too small or too large, resulting in a significant

deviation from the correct result. Figure 25 shows the grid independence test results for the

chamfered design of the prechambers, where it is compared in terms of the tumble ratios for

three different numbers of grid cells in the range of 1,569,146 to 5,894,299.

Grid cell 5,894,299

Grid cell 3,514,368

Grid cell 1,569,146

Figure 25 Comparison of tumble ratios for different grid cell numbers

Figure 25 shows that the tumble ratios are similar even when the number of cells varies

significantly. This is particularly true at the beginning of the intake stroke, which ranges from

45 to 90° with respect to the TDC. The tumble ratio for 1,569,146 cells gives a deviation of

5% when compared to the other two models with different numbers of cells.

6.4 Cycle-to-cycle variation

A correlation was found between the stronger tumble observed during induction and

turbulence levels at the time of ignition, resulting in a faster flame development [88].

To obtain the appropriate result, a simulation was required to compare the effect of the

cycle-to-cycle variations on the tumble flow without the prechamber. Figure 26 shows the

65
tumble ratio for four cycles. The second stage of the tumble was increased. However, the

peak tumble flow reduces from time to time.

1st cycle 2nd cycle 3rd cycle 4th cycle

Figure 26 Tumble ratio for four cycles of cold flow simulation.

From figure 26, a cycle-to-cycle variation is observed without the prechamber. The

second stage of the tumble is increased.

Figure 27 shows a comparison of the cycle-to-cycle variations with (red in figure 27) and

without (black in figure 27) the prechamber. The cycle-to-cycle variation can be observed

from −30° (IVO) to 380° (after the compression stroke). The influence of the tumble ratio on

the cycle-to-cycle variations in the prechamber was analyzed.

66
 1st cycle 2nd cycle 3rd cycle 4th cycle

Figure 27 Comparison of tumble ratios for four cycles with and without the prechamber.

The tumble ratio with the prechamber is greater than that without the prechamber. The

result of the cycle-to-cycle variation analysis shows that the influence of the prechamber on

the cycle-to-cycle varying tumble flow is not significant (Figure 27).

This study presents an inlet flow analysis, starting from −30° CAD to TDC, where at this

crankshaft rotation angle, the intake opening overlaps with the closing of the exhaust valve.

If the remaining exhaust fumes are unaffected by the condition inside the cylinder, the

exhaust valve remains completely closed from 45° CAD to 720° CAD.

The cold flow simulation involved an intake stroke and a compression stroke. The

simulation was performed from 5° until the inlet valve opens, while the exhaust valve was

completely shut off. The mass flow rate of the injected air at the inlet was 0.0103 kg/s, and

the initial pressure at the inlet was 98,900 Pa. This initial condition was obtained from a real

engine model, which has geometric properties close to those of the original computational

model. In the cold flow simulation, the standard k-epsilon model was applied to the engine
67
analysis, with improved wall treatment. An enhanced wall treatment involves employing a

two-layer model where the law of laminar wall is combined with the law of turbulent flow.

As an integral part of the enhanced wall treatment, the two-layer model determines the

turbulent viscosity and ε for the mesh layer on the wall. The application of the enhanced wall

treatment is similar to that of a standard two-layer model if the mesh layer is small enough

to allow a laminar sublayer (when y+ < 30) [14]. For this study, y+ = 16. For accuracy, a

second-order upwind was set to determine the density, momentum, and turbulent kinetic

energy.

This study presents an analysis of the engine starting from 329.4 CAD; at this crankshaft

rotation angle, the inlet opening overlaps with the closing of the exhaust valve. If the residual

exhaust does not affect the condition inside the cylinder chamber, the exhaust valve is

completely closed between 330 CAD and 720 CAD. Figure 28 shows the valve lift process

throughout the engine cycle, involving the opening and closing of the exhaust and intake

valves.

Valve lift, m
0.01
0.009
0.008
0.007
0.006
0.005
0.004
0.003
0.002
0.001
0
0 100 200 300 400 500 600 700 800

lift exhaust lift inlet

Figure 28 Valve lift profile

68
7. Results and discussion
7.1 Geometry and parametric solutions
7.1.1 Model development of the prechamber

The simulation model was developed in two steps including engine and modification

prechamber modification

The engine used in this CFD preliminary study is not identical in geometry to the engine used

for the experimental analysis in this project. A simulated engine was available at an earlier

stage of this project, and a prechamber concept was also modeled. The simulated and

experimental engines were quite similar, including in terms of the combustion chamber type

and the number of valves; the differences were in terms of the injector position and

compression ratio. The simulated one has a compression ratio of 6:1, whereas the

compression ratio of the engine used in the experimental method, presented in subsequent

chapters, is slightly lower at 10:5:1. Despite the difference in the compression ratio, the

purpose here is not to evaluate the efficiency, but to observe the behavior of the prechamber

in a port injection engine. Different cylinder sizes and operating conditions can lead to

different quantitative values of the engine. The change in the thermal efficiency is

insignificant in the compression ratio range of 10.5–11. According to [49], an increase in the

thermal efficiency leads to an increase in the compression ratio, depending on the size of the

cylinder and operating conditions.

A small amount of fuel is necessary to create a scenario close to the stoichiometric

mixture in the prechamber. Details of this port injector are not included in the model. The

combustion chamber contains a bowl-in-piston combined with a pent-roof-type combustion

chamber; this is similar to the design of the single-cylinder engine in the laboratory used for

the experimental analysis of this project.

To analyze the influence of the compression ratio on the charge flow, the geometry of the

initial design was enhanced without significant design changes being made with regard to

different compressibilities. The compression ratio of the engine can be increased in multiple

ways. However, the easiest way is to increase the piston stroke. I did not increase the piston
69
stroke by avoiding volumetric increment, which would have reduced the efficiency. I reduced

the length of the main chamber wall by 2.03 mm, without changing the design of the piston,

while retaining the rest of the geometries of the engine. As a result, the engine compression

ratio was increased from 7.2 to 7.68.

7.1.2 Geometry of the main chamber

According to the literature review, the influence of compression ratio is significant. The

influence of the main cylinder geometry on the charge flow velocity was studied. The piston

dome was decreased from 14.74 to 12.71 mm, and the compression ratio of the engine was

increased from 7.2 to 7.68. The reduced spaces are compared in figure 29.

(a) (b)

Figure 29 Comparison of the main chamber geometry: (a) Space of the main chamber

(compression ratio-7.2, (b) Space of the main chamber (compression ratio-7.68)

There is no significant flow that can be used as a richer charge flow, neither in the intake

stroke nor in the charge stroke.

λ= 7.2
Velocity Z, m/s

λ= 7.68

Crank angle, degrees

Figure 30 Variation in compression ratio with
charge flow velocity
70
The charge flow velocity at the intake stroke is slightly higher than the velocity in the

compression stroke in both cases. The velocity in both cases decreases with increasing

compression ratio.

The simulation results can be summarized as follows:

1- With increasing compression ratio of the engine, the charge velocity of the

prechamber is reduced in the inlet and compression strokes (at the point W, in Fig.

30). When the compression ratio is increased by 6.6%:

a. The first part of the charge flow is decreased by 15–20%.

b. The second part of the charge is increased by 5–10% at the beginning of the

compression stroke and decreased by 10–15% at the end of the compression

stroke.

2- In the second part of the charge flow, the velocity magnitude and flow are directed

from both sides of the inlet and outlet holes of the prechamber in both cases (Figure

30).

7.2 Design of the prechamber

I rotated the prechamber by 45° relative to the Z-axis from the symmetry plane of the main

cylinder. At this position, the volume of the main chamber was varied by changing the

piston design.

7.2.1 Effect of inlet and outlet duct diameters of the prechamber

One of the main parameters determining the charge performance is the inlet duct diameter. I

examined the directional velocity at some points of the inlet and outlet ducts and at three

other positions (Figure 31) inside the prechamber.

71
 w

Figure 31 Model and mechanical configuration of the simulation

The arrows in figure 31 indicate the flow direction. Table 3 lists the geometrical
parameters.

Table 3 Prechamber position and inlet and outlet duct diameter

Parameter Design 2 Design 1.5 Design 1

X axis length (mm) −7.0 −7.0 −7.0
Y axis length (mm) −3.0 −3.0 −3.0
Z axis length (mm) 4.36 4.36 4.36
d (mm) 2.0 1.5 1.0
φ° clockwise 45° 45° 45°

Figure 32 shows a comparison of the flow velocities for inlet hole diameters of 2.0, 1.5, and

1.0 mm at a single point in the middle of the inner wall of the prechamber.

The highest velocity was obtained at 2.0 mm and 18° from the beginning of the intake stroke.

The weakest response was observed in the 1.0-mm-diameter case. The velocity direction

was in line with the charge flow direction. The initial velocity increment occurred much later

than in the other cases.

The negative velocity increased in the late compression stroke in the 1.5-mm-diameter

case. This shows the possibility of inserting a mixture into the prechamber at the beginning

of the intake stroke.

The charge flow begins in the range of 50–60° and increases rapidly to its maximum

value at approximately half the intake stroke. Subsequently, it decreases and then increases

72
again, after which it decreases gradually. The velocity in the Z-direction remains positive

until the second half of the compression stroke. The flow simulation showed that the charge

flow of the prechamber can be divided into two parts. The first part initiates immediately

after the top dead center, and it can be used as the main part of the charge. This is the richest

part of the air–fuel mixture and can be sent to the main chamber during this charge flow.

Inlet Compression Combustion

Design 2
Design 1.5
Design 1

Figure 32 Comparison of the charge flow with the charge velocity for various inlet hole
diameters.

73
 Figure 33 Velocity magnitude at 63.95° after TDC in the intake stroke

The second part of the charge flow continues through the compression stroke. This part

of the charge is drawn by compression with an increment from the piston motion, and it can

be used as an additional charge with a well-mixed stoichiometric-type mixture.

The charge velocity magnitude of the first quarter of the crank angle was observed at

63.95° after the TDC (at the beginning of the intake stroke), as shown in figure 33. This is

consistent with the timing of the Vz-velocity increment shown in figure 32 (the examined

velocity at point W at the inside wall of the prechamber).

Accordingly, the analysis of the gas temperature diagrams provides useful local

information for combustion modeling in terms of the optimum nozzle diameter for preventing

reverse flow.

- The first charge flow increased owing to the increment in the inlet hole diameter;

however, the second part of the charge flow increased to 10 m/s in the 1.5-mm-diameter

case and decreased to −20 m/s in the 1.0-mm diameter case; it was stable at approximately

0 m/s in the 2.0-mm diameter case.

74
The velocity magnitude around the inlet valve drastically varied with the position; it was

more than 160 m/s at a distance of 1.0 mm but decreased to less than 20 m/s at the same

position. Further, it was approximately 0 m/s at a distance of 5.0 mm.

The objectives of this stage of the research were:

a) to predict the charging possibility of the unfueled prechamber in the input

stroke;

b) to obtain a better understanding of the design influence of the inner edge of

inlet and outlet ducts on the charge flow;

c) to reduce the cost of experiments conducted for the most relevant design.

7.2.2. Geometry configuration of the model

An analysis of the best geometry of the prechamber shows that the diameters of the inlet and

outlet ducts have a significant influence on the charge flow. Therefore, I performed a more

in-detail search for the design of the inner edge of the prechamber relatively to its charge

flow. The inner edge design was changed, as shown in figure 34. The diameters of the inlet

a) b)

Figure 34 Prechamber. Dashed area indicates the difference between the designs. a) Design
2C (chamfered inner edge of the prechamber R = 1 mm); b) Design 2S (Dashed area inner
edge of the prechamber is sharp)

75
and outlet ducts and relative location to the main chamber and angular positions of the

prechamber were chosen based on the best results of my analysis, as listed in Table 4.

As mentioned in the previous section, the most effective position of φ° was chosen as the

angular position, and the closest location to the inlet valve is chosen as the horizontal

position, as listed in Table 4.

Table 4 Vertical position of the prechamber

Parameter Design 2C Design 2S

X-axis length (mm) −7.0 −7.0
Y-axis length (mm) −3.0 −3.0
Z-axis length (mm) 4.36 4.36
φ° clockwise 45° 45°
Inner edge of ducts Chamfered Sharp-edged
Design 2C uses a chamfered design, and design 2S uses a sharp inner edge design (Figure
34). Designs 2C and 2S have a similar shape except for the inner ends of the inlet and outlet
ducts, which are chamfered with a 1 mm radius. These parameters were taken from a
naturally aspirated 0.5 L single cylinder, which resembled the geometry of the initial model.
A full-sized model was used to analyze the effect of various geometric prechamber properties
on the charge flow while keeping the diameter and depth of the prechamber constant.
The primary reason for directing a fuel-rich mixture into the prechamber was to determine
the timing that is compatible with the geometry solution to obtain enough pressure difference
between the prechamber and main chamber of the engine. Because the engine considered in
this research is naturally aspirated, the only way to realize such a pressure difference is to
vary the geometry with limited piston speed. To understand the effect of geometry, the engine
speed was varied while the other parameters were kept as steady as possible.
Two designs of the inner edge for the inlet and outlet ducts of the prechamber were
considered: a chamfered design (red circled area in figure 35(a)) and a sharp-edged design
(red circled area in figure 35(b)).

76
 The contours of the magnitude velocity at the same crank angle of 64° from the beginning
of the intake stroke are compared in figure 35. The velocity direction (white traces) in figure
35 is a zoomed-in version of what is shown in figure 34. Additionally, the flow separation
region in the sharp-edged design is wider (black line in figure 35(b)) than the chamfered-
edged (black line in figure 35(a)) design. This suggests that the possibility of charging in the
inner sharp-edge-type inlet and outlet ducts is less than that in the chamfered inner edge
design.

a) b)

Figure 35 Contour of velocity magnitudes around the prechamber during the intake stroke

at 64.1° from the top dead center (TDC). a) The chamfered inner edge (red circled area) of

inlet and outlet ducts of Design 2C; b) The sharp inner edge (red circled area) of inlet and

outlet ducts of Design 2S

The mass flow rates for the designs shown in figure 35(a) and 35(b) are compared in
figure 36. Figure 36(a) shows the mass flow rate in the inlet duct. The sharp-edge design
(black line) causes a lower mass flow rate at crank angles between 45 and 80° and between
90 to 180°. However, the mass flow rates are similar at other crank angles (from 180 to 750°).
The positive and negative mass flow rate values represent the charging and discharging
of the prechamber (Figures 36 to 41), respectively.

In figures 36 and 32, large regions of positive flows (representing a charge flow) can be

observed twice between one ignition and the next one. The first positive mass flow rate region,

through the inlet duct of the prechamber, is observed at crank angles ranging from 45 to 180°

in the intake stroke (red-circled area in figure 36(a)). The second region of the positive mass

77
 flow rates through the same duct was observed at crank angles ranging from 180 to 360° in the

compression stroke (blue-circled area in figure 36(a)).

Mass flow rate, kg/s

Mass flow rate, kg/s

Crank angle, degrees Crank angle, degrees
a) b)

Figure 36 Mass flow rates of the outlet ducts of the prechamber with chamfered edges (grey

line) and sharp edge (black line). a) Mass flow rates in inlet ducts for both designs, b) Mass

flow rates in outlet ducts for both designs

The mass flow rates of the inlet and outlet ducts fluctuate at the intake stroke. The
direction is indicated in red (charge increment direction) and blue (charge decrement
direction) arrows in figure 36. The results show that the increment or decrement durations of
the inlet and outlet ducts are the same. The steps of pulsations for both inlet and outlet ducts
correspond to each other. When the inlet mass flow rate increases, the outlet mass flow rate
decreases. This indicates that the charge flow in the prechamber is unlikely to be stable during
the intake stroke. The total mass flow rate of the discharge exhibits fluctuations.
The mass flow rates in the outlet duct of the prechamber for the chamfered and sharp-
edged designs are shown in figure 36(b). The value of the discharge mass flow rate for the
sharp-edge design is lower than that for the chamfered design, from the TDC (0°) to the BDC
(180°) during the intake stroke.
Figure 37 shows a comparison of the total mass flow rates for the chamfered and sharp-
edge designs. The designs have similar total mass flow rates; however, there are deviations
at some locations. This difference is prominent between the end of the intake and the
beginning of the compression strokes (at crank angles ranging from 135° to 225°, indicated
by the red circle). The positive mass flow rate for the chamfered design occurs from the
beginning of the intake stroke to the middle of the compression stroke. However, some
discharge mass flow rates are present during the intake stroke. This indicates that the

78
chamfered design is more suitable than a sharp-edged one when late remixing in the main
chamber.

Mass flow rate, kg/s

Crank angle, degrees

Figure 37 Comparison of mass flow rate differences between inlet and outlet ducts of

chamfered design (gray line) and the sharp-edge design (black line) of the prechamber

The charging possibility is slightly higher for the sharp edge design than for the
chamfered design. Nonetheless, a discharge mass flow is observed at BDC and at the
beginning of the compression stroke.

The mass flow rate configurations of the two designs were similar, except for the slight

difference at the end of the intake stroke (red-circled area in figure 37). The discharged mass

flow rate from the prechamber is neglected in the case of the chamfered design (gray line

inside the red-circled area in figure 37), which is useful for keeping the charge-rich mixture

separate from the well-blended mixture in the main chamber. For the sharp-edge design case,

the discharge flow can be detected (black line inside the red-circled area in figure 37), which

allows for the possibility of exchanging a charged air–fuel mixture between the prechamber

and main chamber. Consequently, this flow (black line in the red-circled area of Figure 37)

will reduce the difference in the mixture concentrations between these chambers. Some

extremely low-velocity areas can be observed at the positions of the valve openings. This

could indicate the extreme positions of the self-charge process in the self-scavenging

prechamber.

79
7.3 Position of the prechamber
7.3.1 Tilted plane effect on the charge flow

The influences of angular inclinations θ and γ relative to the vertical axis Z are analyzed at

different planes Y–Z and X–Z relatively.

The plane, shown in figure 29, is the plane of the prechamber tilt angle.

a) b)

Figure 38 Comparison of the design configurations of a different plane of the prechamber

inclination; a) Design 2CY; The tilt angle is located in the parallel plane of Y-Z parallel to

the symmetry plane; b) Design 2CX; The tilt angle is located in the parallel plane of Z-X;

The tilt angle of the prechamber is changed due to the symmetry (design 2CY in figure 38)

and gross symmetry (design 2CX in figure 38) plane of the experimental engine. Therefore,

the geometry is changed to conform with the experimental engine (Table 5).

Table 5 Angular position of the prechamber

Parameter Design 2CY Design 2CX

X axis length (mm) −7 −7
Y axis length (mm) 0 0
Z axis length (mm) 0 0
φ° clockwise 0° 90°
θ clockwise 11° 0
γ clockwise 0° 11°

80
 According to the simulation, the same tilt angle of 11° in Designs 2CX and 2CY shows

a significant difference. Figure 39 shows the result of the simulation. The charge flow in the

intake stroke is negligible in case of design 2S, whereas a noticeable charge flow is observed

in Design 2CY. In the case of Design 2CX, there is no significant usable charge flow, neither

1in the inlet nor in the compression strokes.

Design 2CY
Velocity Z, m/s

Design 2CX

Figure 39 Charge flow comparison of two designs.

The charge flow velocity of the intake stroke is much higher than that of the velocity of

the compression stroke in Design 2CY. From here, it can be observed that the influence of

angular position φ° relative to the vertical axis is significant.

Based on the simulation result, the following can be assumed:

The angular position of the prechamber significantly influences the charge flow (Figure 30).

a) The same tilt angle in the different planes has a different effect on the charge

flow.

b) The second part of the charge velocity is also neglected in the case of X-Z

plane tilted position.

81
 The geometric configuration of the current engine must be modified to meet with the best

result of this simulation or a new method should be developed to detect some charge flow in

the vicinity of the tilt angle.

7.3.2 Influence of angular position φ° on the charge flow

As discussed in the previous chapter, the angular effect of φ° was significant. Thus, I

proposed to analyze this design parameter in more detail. To obtain the appropriate result, I

set the geometry of the prechamber position through the angular change relative to the

symmetry plane Y-Z of the main chamber. Figure 40 shows the geometry development and

design configuration. Table 9 lists the parameter settings of the geometry.

The CAD model was developed in SpaceClaim and converted into the Design Modeler

of ANSYS.

φ°

φ°

Figure 40 Angular position of the prechamber with respect to φ°°.

Owing to the different inclination angle of the original engine spark plug hole, the tilt angle

of the prechamber is dispositioned in a different plane. Therefore, we attempted to adjust

the geometry to conform with the experimental engine.

The charge velocity in the z-direction at point W (Figure 41) is computed in accordance

with various angular positions (Table 6) as compared in figure 41 and 42.

82
 Table 6 Angular position of the prechamber though φ°

Parameter Position 2Ca Position 2Cb Position 2Cc Position 2Cd

X axis length (mm) −7 −7 −7 −7
Y axis length (mm) 0 0 0 0
Z axis length (mm) 0 0 0 0
φ° clockwise 30° 40° 60° 80°

400
500
300 w

600
80°

CAD, degrees

Figure 41 Velocity magnitude at various angular positions of the prechamber with respect
to φ°.

Velocity along the Z axis was typically positive during the inlet and compression stroke

as shown in figure 41. This means that charge flow was headed to the upper part of the

prechamber in the vicinity of the prechamber wall. An increment by up to times was observed

for angular positions of φ°: one in the intake stroke and one in the compression stroke.

The angular effect of the prechamber of the current design was significant. The first part

of the charge velocity at point W, shown in figure 41 was in the range of 16–17 m/s, and the

second part of the charge flow was in the range of 5–11 m/s, in the angular position range of

40–60°.

The following can be assumed from the simulation result:

The charge flow velocity can be divided into two parts.

a. The first part of the charge velocity is shown in blue (Figure 42).

83
 b. The second part of the charge velocity is shown in red (Figure 42).

2- Charge flow of the intake stroke deep inside the prechamber is much higher than

that of the compression stroke, according to the current design configuration.

Inlet 85 degree
Velocity, m/sec 0 Comp350 degree
20 1020
30
15 40
50
10 60
70
5 γ0 80
0 90
270

180

Figure 42 Velocity magnitude at various angular positions of the prechamber with respect

to φ°.

The highest value of the charge velocity magnitude was observed at approximately 40°

turn, and the lowest value of the charge velocity was at a position most distant for the design

2S. This phenomenon occurred throughout the first half of the intake stroke, as shown in

figure 41. Some increments are observed in the second half; however, they are lower than

those in the first half of the stroke. From figure 42, the charge flow of the intake stroke is 2–

3 times higher than the charge flow of the compression stroke. The highest velocity

84
magnitude is observed in design 2S and the lowest velocity value is observed in Design D in

both the inlet and compression strokes.

The outflow at precession angles of 60° (in black line), 55° (in blue), and 45° (in red) with

the angular correction is shown in figure 43 (the difference at a precession angle of 55° with

the angular correction and the mass flow rates of inlet and outlets are shown in figure 44).

Figure 43 Mass flow rates of the inlet of the prechamber for rotation angles of 60° and

55°, 45° with the angular correction.

Figure 44 Mass flow rates of the outlet of the prechamber for rotation angles of 60° and

55°, 45° with the angular correction.

However, the mass flow rates are similar at the remainder of the crank angle (from 180
to 750°) positions.
The positive and negative mass flow rate values represent the charging and discharging
of the prechamber (Figures 43 and 44), respectively.

85
 In figures 36 and 41, large regions of positive flows (representing a charge flow) were observed

twice between one ignition and the next one. The first positive mass flow rate region, through the

inlet duct of the prechamber, is observed at crank angles ranging from 45 to 180° in the intake

stroke (red-circled area in figure 36(a)). The second region of the positive mass flow rates, through

the same duct, was observed at crank angles ranging from 180 to 360° in the compression stroke

(blue- the circled area).

From the simulation results, we can assume the following:

- The possibility of charging the prechamber during the intake stroke (The

first part of the charge)

- The possibility of maintaining the inserted charge during the compression

stroke (The second part of the charge)

The velocity inside the prechamber has been determined at point W, as shown in figure 41.

The graphics in blue represent the highest velocity in the prechamber with respect to the Z-

axis.

- Richer part of the inlet charge can be directed into the prechamber.

- The rich mixture inside the prechamber can be conserved during the

compression charge

- An experimental research should be performed with the current design

configuration.

- The geometry solution should be optimized with respect to the change in the

main parameters of the inlet and outlet hole tilt angles.

- The effect of the vertical location on the velocity in the next stage of the

simulation should be evaluated.

86
7.3.3 Influence of inclination angle θ°
The inclination angle of the prechamber is set as θ° in the X–Z plane (figure 45). The

beginning of the cycle was simulated in ANSYS Fluent.

a. Center, θ = 0° b. Inclined, θ = 11° c. Inclined, θ = 18°

Figure 45 Tilt position of the prechamber

The initial position of the spark plug hole of the mainstream engine is θ = 0°; however,

the original position of the spark plug hole of the single-cylinder engine used in the

experimental setup was tilted by 30°. Thus, the position of the prechamber was tilted to θ =

11° and θ = 18° in the clockwise direction and aligned with the center of the main cylinder.

The three positions were compared in terms of the tumble ratios.

Geometry setting:

Table 7 lists the values of the prechamber setting.

Table 7. Position of the tilt angle of the prechamber

Parameter Center Tilted 11 Tilted 18

X axis length (mm) 0 0 0
Y axis length (mm) 0 0 0
Z axis length (mm) 0 0 0
φ° clockwise 0° 0° 0°
γ clockwise 0° 0° 0°
θ clockwise 0° 11° 18°

87
 • The initial model position of the spark plug hole was tilted and not positioned at the

center of the main chamber.

• Design change principles were found.

• The prechamber can now be moved in the Y-Z plane.

• CAD designs are made using SpaceClaim and converted to the Design Modeler.

Figure 46 shows a comparison of the tumble ratios of the related design.

Center
Tilted 11
Tilted 18

Figure 46 Comparison of tumble ratios of different tilt positions

As shown in figure 46, the tumble ratio is not significantly affected by the prechamber

tilt angle θ at the beginning of the inlet cycle. From the 0 to 30° from the TDC of the intake

stroke to the end of the compression stroke TDC, the tumble ratio effect was observed from

the inclination of the prechamber.

As shown in figure 47, the velocity at the inlet and outlet of the prechamber is

significantly affected by the prechamber tilt angle θ at the beginning of the inlet cycle.

88
 a)

b)

c)

a. Center b. First position c. Tilted 18°

Figure 47 Comparison of the charge flow velocity magnitudes at different CAD positions;

a-418 in the intake stroke, b-544 at the beginning of the compression stroke, c-627

at the end of the compression stroke

89
7.3.4 Effect of the vertical position of the prechamber

The main coordinates and angular positions were demonstrated in previous parts. The only

location that was not evaluated is the vertical location. Thus, the vertical location of the

prechamber is discussed in this section. Figure 48 shows a comparison of the design settings.

4.36 mm 2.36 mm

a. Design 2C-2 (shifted down by 2.36 b. Design 2C-4 (shifted down by 4.36
mm from the top of the cylinder head); mm from the top of the cylinder
head).

Figure 48 Comparison of vertical positions of the prechamber

As mentioned in the previous section, the most effective position of φ° was chosen as the

angular position, and the closest location to the inlet valve was chosen as the horizontal

position, as listed in Table 8.

Table 8 Vertical position of the prechamber

Parameter Design 2C-2 Design 2C-4

X axis length (mm) −7 −7
Y axis length (mm) −3 −3
Z axis length (mm) 2.36 4.36
φ° clockwise 45° 45°

90
 The position of the prechamber is changed through the Z-axis while keeping the rest of

the geometrical parameters constant. Charge flow can occur in the inlet and compression

strokes. The duration of the simulation is limited from the beginning to the end of the

compression stroke, to save computing time.

4.36 mm position

2.36 mm position

Figure 49 Comparison of the mass flow rates at different vertical positions of the prechamber

The mass flow rates of the inlet and outlet ducts of the prechamber are calculated based on

the disposition vertical positions. The result of the two designs is compared in figure 49. The

positive and negative marks in the graph represent the charge and discharge mass flow rates,

respectively. In the case of Design 2C-4, it has more positive charge flow after the inlet valve

is closed (45° after the TDC). However, the net mass flow rate is negligible after the exhaust

valve is closed. In the case of Design 2C-2, there is no significant usable charge flow, even

in the intake stroke.

The simulation results show that the influence of the vertical position of the prechamber

is significant. The charge flow at the beginning of the intake stroke fluctuates until the

exhaust valve is closed.

91
7.4 Comparison of wavelike fluctuation in the prechamber

charge flows
7.4.1 Fluctuation phenomena
Designers require an engine that can operate with high turbulence when the piston is near the
TDC. This is because, near the TDC, ignition occurs, and high turbulence allows for a more
rapid break up and faster flame spread compared to the low turbulence. Previous researchers
evaluated the combustion chamber geometry and piston shape required to generate
turbulence in the engine; these factors sequentially affect the combustion process. Other
important factors that contribute to the turbulence generation are the piston speed, inlet flow
condition, and intake manifold design. In relation to the speed of all parts and flows, the
turbulent kinetic energy is one of the useful parameters that can be determined via simulation.

7.4.2 Charge and discharge of the prechamber

From the simulation, it was observed that the first part of the charge flow to the
prechamber occurred immediately after the TDC of the intake stroke and can be considered
the central part of the charge. It can be used as the richest part of the air–fuel mixture when
fuel is injected into the chamber while the inlet valve is open, and the air–fuel mixture is not
homogeneously mixed. During this period, the charge flow of the prechamber fluctuates
continuously, and the highest part of its amplitude at the beginning of the intake stroke was
a self-scavenging pulse. Figures 36 and 41 show the alternating net charge inflow and outflow
in the prechamber, respectively. This alternation results in a pulse, which is important not
only to establish a charge and discharge order but also to distinguish between the prechamber
and the main chamber during any specific period of the combustion cycle.
Figure 50 shows a comparison of the pressures at the inlet (black line), outlet (red line),
middle side of the prechamber (green line), and top of the prechamber (blue line). The results
show that the pressure at the inlet and outlet ducts are mostly higher than the pressure in the
middle, whereas the pressure at the top side of the prechamber (blue line) is much lower
during the intake stroke (0–180°). This incremental pressure difference is the reason for the
first charge of the prechamber. The pressure differences between the ducts and inner areas of
the prechamber are shown in figure 50.
The next charge configurations occur right after the BDC when all the pressures from
various locations become similar from the beginning to the middle of the compression stroke
(180 to 270°). All the pressure and pressure differences increased to push the charge from
92
the main high-pressure chamber to the low-pressure prechamber (Figure 50). The directions
of the mass flow rates from both the ducts were positive, indicating that the charge flowed
into the prechamber from both sides of the inlet and outlet during this period, for both the
chamfered and sharp edge designs (Figure 36).

Crank angle, degrees

Figure 50 Pressure at different places of the prechamber.

When the mass flow rate at the inlet duct increases in the positive direction, the mass
flow rate of the outlet duct decreases in the negative direction, combined with the pulsations
of inlet and outlet ducts, and some fluctuations occur, as shown in figure 51. The trends in
the mass flow indicate the charge and discharge directions. The negative mass flow rate
dominates for both the inlet and outlet ducts during the first 40° of the intake stroke (Figure
51); this is evidence of discharge flow occurring at the beginning of the intake stroke.
After the discharge mass flow rate at the TDC, there is a positive mass flow rate in the
direction from the inlet valve to the prechamber inlet duct. The highest mass flow rate
occurred at 45° after the beginning of the intake stroke. The weakest charge flow occurred at
220° after the TDC. The mass flow direction fluctuated throughout the intake stroke. The
increment in the positive mass flow rate following the velocity increment in the direction,
shown in figure 5, indicates the potential to introduce a mixture into the prechamber at the
beginning of the intake stroke.
The charge flow begins at crank angles between 50 and 60° and increases rapidly to its
maximum value around the halfway point of the stroke. Thereafter, it fluctuates, before
gradually decreasing after the second increment.

93
 Mass flow rate, kg/s

Figure 51 Net mass flow rates of the inlet and outlet

ducts of the prechamber for chamfered-edge design

The experimental results show a wave-like fluctuation in the total mass flow rate in the
inlet and outlet ducts of the prechamber as well as pressure fluctuation [12]. The pulse
amplitude of the fluctuation decreases with increasing crank angle.
Figure 52 shows the phase duration of the pulsation. Based on the trend in this figure, the
pulse step (the number of pulses from the beginning, indicating the difference between pulse
durations) changes relative to the crank angle. Therefore, it is assumed that the charge flow
during the intake stroke is not continuous but is a combination of the alterations between the
discharge and charge processes.
Mass flow rate, kg/s

Figure 52 Inlet (blue), outlet (black), and net (red) mass flow rates of the ducts of the
prechamber for chamfered-edge design at the 2.36 mm position.

94
 At the beginning of the intake stroke, both the valves are opened, and the pressure
difference required to direct the charge flow to the prechamber is not stable if the charged
mass enters the prechamber. The charged mass changes the density of the prechamber and
creates a pressure difference on the opposite side. This pressure increment causes a discharge,
and then the pressure drop reoccurs, resulting in charge flow addition. The net mass flow rate
characteristics resemble an “inverted liquid bottle effect” [30].
In figure 52, the mass flow rate of the inlet duct fluctuates from the beginning of the
intake stroke and the outlet mass flow rate (the positive and negative values represent the
charging and discharging of the prechamber, respectively). The steps and directions of the
fluctuations are identical. As the inlet mass flow increases, the outlet mass flow decreases.
Consequently, the mass flow rate into the prechamber increases because the outlet mass flow
rate decreases. This indicates the charging process. When the charge flow decreases, the
outlet flow increases. This indicates a discharge process. The delta between the discharge
and charging process results in a net mass flow rate fluctuating with the same frequency as
the inlet and outlet mass flow fluctuations. Furthermore, the charge flow and outlet mass flow
rates increase from 20 CAD. In this case, there are no fluctuations, and both the inflow and
outflow increase smoothly. Therefore, the net mass flow rate is smooth, and there is no
difference between the charge and discharge flows. The inflow is moves out at the same rate
as it enters, and there is no difference in the mass of the prechamber volume.

7.4.3 Engine speed

If there is some fluctuation in the net mass flow rate, the prechamber will be charged and
discharged (Figure 53 and 54). If there is no pulsation in the net mass flow rate, the charging
possibility is negligible. It is particularly important to charge the fresh mixture and discharge
exhaust smoke from the prechamber. Finally, it is important to pre-clean the prechamber
from the exhaust smoke. The fluctuation ends early when engine RPM decreases.

95
 1000 rpm
Mass flow rate, kg/s

120 rpm
Mass flow rate, kg/s

Figure 53 Comparison of the prechamber positions: 1000 RPM (Top graph) and 120 RPM

(Bottom graph).

Like the “upside-down liquid bottle phenomenon,” without enough pressure difference
or vibration fluid inside the prechamber, nothing will come out. If there is enough pressure
difference to trigger the pulsation, the mass flow will go into the prechamber until the
pressure is high enough to push the exhaust gas out from the prechamber. This is because the
vortices in the prechamber and main chamber interact through the inlet and outlet ducts of
the prechamber. The phase of the intense vortex interaction during the vortex pairing is
followed by a “calming” period that corresponds to the vortex roll-up after the coalescence
[31]. This process is repeated until the pressure or velocity difference between the
prechamber and the main chamber is negligible.

96
 Figure 54 Fluctuation intensity and velocity vector [22]

The flame front moves in the opposite direction during the same period (2° CA) of time.

This flow behavior is graphically represented by the velocity vectors [22].

The frequency of the fluctuation decreases with increasing engine speed (1000 RPM case
in figure 53), whereas it increases with decreasing speed (120 RPM case in figure 53). This
is because the two time-dependent parameters, i.e., the speed and frequency of the charge
flow fluctuations, have the same variables of time. Therefore, the characteristics of the
fluctuation depend on the engine speed.

7.4.4 Smoothness of the fluctuation

A further increase in the speed fluctuation is observed in the middle part of the intake
stroke after 45° from the TDC.

Figure 49 shows that the position of the prechamber significantly influences the charge
flow as well as the fluctuation of the charge flow itself.
Smooth fluctuations appear at a high position and uneven fluctuations occur at the deep
positions of the prechamber.

Figure 55 shows a comparison of the result different RPM s in terms of the fluctuations.
Smoother fluctuation can be observed in the higher RPM case, after the 45°.

97
 2000 rpm
Mass flow rate, kg/s

Crank angle, degrees

1000 rpm
Mass flow rate, kg/s

Crank angle, degrees

Figure 55 Comparison of the net mass flow rates with respect to engine speed.

98
7.4.5 Model matching
The advantage of the present model is its flexibility and simplicity, which allows studying
the effects of several factors, such as the mixture distribution and fuel components, on the
engine performance, which are more practical than other types of modeling.

The following assumptions were made for the fluctuation model:

-First, the amplitude of the fluctuation decreases as time goes by.

-Second, the fluctuation frequency is related to some trigonometric functions.

Total mass flowrate

0.00004
0.00003
massflow
0.00002
0.00001
0
0 50 100 150 200 250 300 350 400 450
-0.00001
-0.00002
model
-0.00003
-0.00004
-0.00005
-0.00006

M=0.0002𝑒 −0.014𝑡 Cos(0.14t+1.2)

Figure 56 Net mass flow rate fluctuation model

Table 9 shows the comparison of the fluctuations at different RPMs.

Table 9 Fluctuation coefficients

RPM 120 1000 2000

Net mass flow rate (M) 4.14×20−6 0.00002 0.0002
Timing (b) −0.11 −0.055 −0.014
Frequency (omega) 1 0.6 0.14
Timing (phi) 0 0.6 1.2

99
 As shown in figure 56, the amplitudes of the fluctuation of both the simulated and

predicted results decrease exponentially. The waveform of the fluctuation is sinusoidal.

Moreover, the frequency of fluctuation is reduced relative to the time factor of CAD. From

the observation of the mass flow rate fluctuation of different designs, the coefficient before

the exponential function is close to the maximum amount of the net mass flow rate of the

inlet and outlet ducts of the prechamber. Moreover, the degree of the exponential factor

represents the amplitude reduction in accordance with the CAD. The angular coefficient of

the sinusoidal function represents the frequency of the fluctuation, and the constant number

represents the beginning point of the fluctuation according to the CAD. The validity of the

equation is the number of aforementioned parameters was altered in accordance with the

change in the different parameters of the prechamber and engine design.

Net Mass flow rate (M)

0.00025

0.0002

0.00015
y = 3E-06e0.0021x
R² = 0.995
0.0001

0.00005

0
0 500 1000 1500 2000 2500

Figure 57 Amplitude of the fluctuation related to engine speed

The amplitude of the net mass flow rate fluctuation model changed in relation to the

change in the engine speed. As shown in figure 57, the charge mass flow amplitude increases

exponentially in accordance with the increment in the engine RPM.

100
 Frequency (omega)
1.2

1
y = -0.0005x + 1.0558
0.8 R² = 1

0.6

0.4

0.2

0
0 500 1000 1500 2000 2500

Figure 58 Relationship between charge flow fluctuation frequency and engine speed

From the above graph it can be observed that the frequency of the fluctuation has an inverse

linear relationship with the increment of the RPM.

Timing (b)
0
0 500 1000 1500 2000 2500
-0.02

-0.04

-0.06

-0.08 y = -1E-08x2 + 8E-05x - 0.1189

R² = 1
-0.1

-0.12

Figure 59 Timing of the charge mass flow rate fluctuation in the prechamber (y represents
the coefficient b; x represents the engine speed)

From figure 59, it can be assumed that the charge flow is limited by the engine speed.

This is because the Timing coefficient of the charge flow fluctuation is limited by the number

b = −0.1189. From the above graph, it can be predicted that the charge mass flow rate of the

current design can result in charging under any conditions when b > −0.014.

101
It can be concluded that every design of the prechamber must has its Timing graph as shown

in figure (56 -59). The charging possibility of the prechamber can be evaluated with this

methodology.

7.4.6 Conclusion of the chapter

When fluctuation occurs only in one part of the inlet and outlet of the prechamber, the
charge flow can be directly observed. If there are no fluctuations, the charge flow cannot be
observed, however, there is some increment in the mass flow rate. These characteristics show
reasonable agreement with the pressure chart estimates (Figure 50).

8. Conclusions
In this research, different prechamber modifications were investigated by conducting a

series of CFD simulations. Since the flow conditions inside the cylinder influence the

combustion quality and mixing process of air and fuel, a fueled prechamber was found to be

an essential factor influencing the engine efficiency. The efficiency of an unfueled

prechamber relies on the charging possibility of the prechamber with a fuel-rich mixture. It

was important to understand how the charge flow characteristics and fluctuations can cause

a difference in the density and pressure profiles.

A cold flow analysis was conducted to gain insights into the cylinder pressure and

temperature development, which is a crucial factor for pre-combustion conditions, providing

an understanding of the influence of the design parameters on flow-related physical

conditions independent from the influence of chemical reactions. In this thesis, I focused on

the charge flow of the prechamber in the input stroke; however, I found another possibility

in the compression stroke. The pressure and temperature of the analyzed engine were similar

to those of other engines. The peak pressure and temperature were 690 kPa and 535 K,

respectively, which are comparable to those of engine systems studied by other researchers.

Two different inner edge configurations of the prechamber were compared in terms of the

mass flow rate. The simulation results show that the inner edge shape of the prechamber inlet

and outlet ducts do not considerably influence the charge flow at the beginning of the intake

102
stroke. The difference in the charge flow is only within a short gap of the crank angle from

125 to 265° and is approximately 15%.

The results indicate the possibility of charging an unfueled, unscavenged prechamber

with a fuel-rich mixture. The improved configuration not only makes it possible to charge

the prechamber with a fuel-rich mixture but also helps prevent the rich mixture from leaking

into the main chamber during the compression stroke.

A similar principle is known as the “inverted liquid bottle effect.” If the flows in the inlet

and outlet ducts of the prechamber are interrupted, the discharging possibility of the

prechamber is negligible. This is the main concept of keeping a richer charge in the

prechamber during the compression stroke. However, if the mass flow rate through the ducts

increases, the charging possibility becomes the net value of the charge and discharge mass

[31]. If the mass flow rates in the inlet and outlet ducts fluctuate, the net mass flow will also

fluctuate. The amplitude of this fluctuation reduces gradually until the pressure difference

becomes negligible.

The influences of the inner edge on the flow state are significant from the end of the

intake stroke to the beginning of the compression stroke. The general configurations of the

chamfered and sharp edge designs were similar. As the piston approaches the bottom dead

center, the chamfered inner edge design caused a decrease in the discharge flow rate, which

enhanced the charge flows on both sides of the inlet and outlet ducts and was compromised

in the middle of the prechamber during the compression stroke. The prechamber with a sharp

inner edge design had some discharge flow from the end of the intake stroke, and the charge

flows from the inlet duct dominated during the compression stroke.

The following can be concluded from the CFD simulation results:

• The practical relevance of this study is identification and classification of the flow
potential that can be used to charge and recharge unfueled unscavenged

prechamber.

• The charge flow into the unfueled prechamber can be divided into two parts: The

first is in the intake stroke (from TDC to BDC) for charge, and the second is in the

compression stroke (from BDC to TDC) for recharge.

103
 • The first part of the charge flow is not continuously directed into the prechamber
during the intake stroke.

During the intake stroke, the charge flow fluctuates in a pulsed manner. When the charge

flow dominates, there is no discharge flow. When discharge flow dominates, the charge flow

is reduced. This type of pulse effect is useful for making density and pressure differences,

which lead to inlet charge mass and discharge of the used gas. It is also useful for keeping

the rich mixture inside the prechamber. The theoretical achievement of this research is

determination of the fluctuation effect and its amplitudes of the first charge. These are useful

to:

• Predict the charging possibilities of the prechamber for further development

• Maintaining charge without leakage during the second charge owing its valve-like
fluctuation.

The charge flow of the prechamber depends on the design and position of its inlet and

outlet ducts. Storing a charged-enriched mixture inside the prechamber during a compression

stroke is as crucial as filling the prechamber with the richest fuel mixture. The charge flow

for the unfueled prechamber can be fueled, and the local fuel enrichment in the prechamber

can be used as a jet igniter for the lean mixture in the main combustion chamber using a

richer portion of the fuel that is injected when the inlet valve is open. This allows for minor

modifications to be done on the existing port-injected engine to operate as a gasoline direct-

injected engine.

Further, there exist immense practical applications of this study to port-injected SI

engines that are still in use.

References

[1] H. Khatib, “IEA World Energy Outlook 2011—A comment,” Energy Policy, vol. 48,
pp. 737–743, Sep. 2012, https://doi.org/10.1016/j.enpol.2012.06.007.
[2] L. S. Baumgartner, S. Wohlgemuth, S. Zirngibl, and G. Wachtmeister, “Investigation
of a Methane Scavenged Prechamber for Increased Efficiency of a Lean-Burn Natural

104
 Gas Engine for Automotive Applications,” SAE Int. J. Engines, vol. 8, no. 2, pp. 921–
933, Apr. 2015, https://doi.org/10.4271/2015-01-0866.
[3] “2015-07-alter-fuels-transport-syst-in-eu.pdf.” Accessed: Dec. 10, 2019. [Online].
Available:
https://ec.europa.eu/transport/sites/transport/files/themes/urban/studies/doc/2015-07-
alter-fuels-transport-syst-in-eu.pdf.
[4] “Global Automotive Supplier Study 2018,” Roland Berger.
https://www.rolandberger.com/en/Publications/Global-Automotive-Supplier-Study-
2018.html (accessed May 19, 2020).
[5] M. Bunce, H. Blaxill, W. Kulatilaka, and N. Jiang, “The Effects of Turbulent Jet
Characteristics on Engine Performance Using a Pre-Chamber Combustor,” SAE
International, Warrendale, PA, SAE Technical Paper 2014-01–1195, Apr. 2014.
https://doi.org/10.4271/2014-01-1195.
[6] T. Iskandar, “Compressed Natural Gas Direct Injection (Spark Plug Fuel Injector),” in
Natural Gas, P. Potocnik, Ed. Sciyo, 2010.
[7] Doğan, Kutlar, Javadzadehkalkhoran, and Demirci, “Investigation of Burn Duration and
NO Emission in Lean Mixture with CNG and Gasoline,” Energies, vol. 12, no. 23, p.
4432, Nov. 2019, https://doi.org/10.4271/2014-01-1195.
[8] A. Jamrozik, “Lean combustion by a pre-chamber charge stratification in a stationary
spark ignited engine,” J Mech Sci Technol, vol. 29, no. 5, pp. 2269–2278, May 2015,
https://doi.org/10.1007/s12206-015-0145-7.
[9] S. Gail, T. Nomura, H. Hayashi, Y. Miura, K. Yoshida, and V. Natarajan, “An Intake
Valve Deposit (IVD) Engine Test Development to Investigate Deposit Build-Up
Mechanism Using a Real Engine,” SAE Int. J. Fuels Lubr., vol. 10, no. 3, pp. 728–740,
Nov. 2017, https://doi.org/10.4271/2017-01-2291.
[10] T. I. Mohamad, A. Yusoff, S. Abdullah, M. Jermy, M. Harrison, and H. H. Geok, “The
Combustion and Performance of a Converted Direct Injection Compressed Natural Gas
Engine using Spark Plug Fuel Injector,” Sep. 2010, pp. 2010-32–0078,
https://doi.org/10.4271/2010-32-0078.
[11] C. E. C. Alvarez, G. E. Couto, V. R. Roso, A. B. Thiriet, and R. M. Valle, “A review of
prechamber ignition systems as lean combustion technology for SI engines,” Applied
Thermal Engineering, vol. 128, pp. 107–120, Jan. 2018,
https://doi.org/10.1016/j.applthermaleng.2017.08.118.
[12] N. Zimmerman, J. M. Wang, C.-H. Jeong, J. S. Wallace, and G. J. Evans, “Assessing
the Climate Trade-Offs of Gasoline Direct Injection Engines,” Environ. Sci. Technol.,
vol. 50, no. 15, pp. 8385–8392, Aug. 2016, https://doi.org/10.1021/acs.est.6b01800.
[13] R. Zhu et al., “Tailpipe emissions from gasoline direct injection (GDI) and port fuel
injection (PFI) vehicles at both low and high ambient temperatures,” Environ. Pollut.,
vol. 216, pp. 223–234, Sep. 2016, https://doi.org/10.1016/j.envpol.2016.05.066.
[14] X. Zheng et al., “Characteristics of black carbon emissions from in-use light-duty
passenger vehicles,” Environ. Pollut., vol. 231, pp. 348–356, Dec. 2017,
https://doi.org/10.1016/j.envpol.2017.08.002.

105
[15] M. Bunce and H. Blaxill, “Methodology for Combustion Analysis of a Spark Ignition
Engine Incorporating a Pre-Chamber Combustor,” SAE Technical Papers, vol. 2014,
Oct. 2014, https://doi.org/10.4271/2014-01-2603.
[16] H. Ryu, A. Chtsu, and T. Asanuma, “Effect of torch jet direction on combustion and
performance of a prechamber spark-ignition engine,” Society of Automotive
Engineers,Warrendale, PA, CONF-870204-, Jan. 1987. Accessed: Dec. 17, 2019.
[Online]. Available: https://www.osti.gov/biblio/6513906-effect-torch-jet-direction-
combustion-performance-prechamber-spark-ignition-engine.
[17] E. Toulson et al., “Visualization of Propane and Natural Gas Spark Ignition and
Turbulent Jet Ignition Combustion,” SAE Int. J. Engines, vol. 5, no. 4, pp. 1821–1835,
Oct. 2012, doi: 10.4271/2012-32-0002.
[18] G. Gentz, M. Gholamisheeri, and E. Toulson, “A study of a turbulent jet ignition system
fueled with iso-octane: Pressure trace analysis and combustion visualization,” Applied
Energy, vol. 189, pp. 385–394, Mar. 2017,
https://doi.org/10.1016/j.applthermaleng.2015.02.026.
[19] A. Jamrozik, W. Tutak, A. Kociszewski, and M. Sosnowski, “Numerical simulation of
two-stage combustion in SI engine with prechamber,” Applied Mathematical
Modelling, vol. 37, no. 5, pp. 2961–2982, Mar. 2013,
https://doi.org/10.1016/j.apm.2012.07.040.
[20] J. Benajes, R. Novella, J. Gomez-Soriano, P. J. Martinez-Hernandiz, C. Libert, and M.
Dabiri, “Evaluation of the passive pre-chamber ignition concept for future high
compression ratio turbocharged spark-ignition engines,” Applied Energy, vol. 248, pp.
576–588, Aug. 2019, https://doi.org/10.1016/j.apenergy.2019.04.131.
[21] A. Roubaud, R. Röthlisberger, and D. Favrat, “Lean burn cogeneration biogas engine
with unscavenged combustion prechambers: comparison with natural gas,” Int. J.
Applied Thermodynamics, pp. 169–175, 2002.
[22] M. Blankmeister, M. Alp, and E. Shimizu, “3.3 Passive Pre-Chamber Spark Plug for
Future Gasoline Combustion Systems with Direct Injection,” p. 26.
[23] G. Nyamsuren, H. Asada, and Y. Ogami, “Computational Study on Using an Unfueled
Prechamber for Lean Burning in a Gasoline Engine,” 2019,
https://doi.org/10.1299/jsmekansai.2019.94.414.
[24] G. Nyamsuren, K. Fukudome, and Y. Ogami, “CFD Investigation of Self-Scavenging
Prechamber for the Lean Flammability of a Gasoline Engine,” 2018,
https://doi.org/10.1299/jsmekansai.2018.93.1012.
[25] N. Gombosuren, O. Yoshifumi, and A. Hiroyuki, “A Charge Possibility of an Unfueled
Prechamber and Its Fluctuating Phenomenon for the Spark Ignited Engine,” Energies,
vol. 13, no. 2, p. 303, https://doi.org/10.3390/en13020303.
[26] N. Gombosuren, H. Asada, and Y. Ogami, “Computational fluid dynamics study for an
efficient passive prechamber,” The 29th International Symposium on Transport
Phenomena (ISTP29), vol. 29, no. Honolulu, USA, p. 4, Oct. 2018.
[27] G. Nyamsuren and Y. Ogami, “Fluctuating Phenomena in the Charge Flow of Un-
Fueled Prechamber of Gasoline Engine,” Sixteenth International Conference on Flow
Dynamics, vol. OS2-54, no. Sendai, Miyagi, Japan, p. 2, Nov. 2019.
106
[28] G. Nyamsuren, Y. Ogami, and H. Asada, “Charge Flow Investigation of Un-Fueled
Prechamber for the Gasoline Engine,” Fifteenth International Conference on Flow
Dynamics, Sendai, Miyagi, Japan, vol. 15, no. Sendai, Miyagi, Japan, Nov. 2018.
[29] J. Liang, X. Luo, Y. Liu, X. Li, and T. Shi, “A numerical investigation in effects of inlet
pressure fluctuations on the flow and cavitation characteristics inside water hydraulic
poppet valves,” International Journal of Heat and Mass Transfer, vol. 103, pp. 684–
700, Dec. 2016, https://doi.org/10.1016/j.ijheatmasstransfer.2016.07.112.
[30] P. Gao, T. Liu, T. Yang, and S. Tan, “Pressure drop fluctuations in periodically
fluctuating pipe flow,” J. Marine. Sci. Appl., vol. 9, no. 3, pp. 317–322, Sep. 2010,
https://doi.org/10.1007/s11804-010-1014-5.
[31] S. A. Karabasov and V. M. Goloviznin, “Direct Numerical Simulations of Compressible
Vortex Flow Problems,” Advanced Fluid Dynamics, Mar. 2012,
https://doi.org/10.5772/26473.
[32] J. F. Thomas and R. H. Staunton, “What Fuel Economy Improvement Technologies
Could Aid the Competitiveness of Light-Duty Natural Gas Vehicles?,” May 1999, pp.
1999-01–1511, https://doi.org/10.4271/1999-01-1511.
[33] M. Noguchi, S. Sanda, and N. Nakamura, “Development of Toyota Lean Burn Engine,”
Feb. 1976, p. 760757, https://doi.org/10.4271/760757.
[34] E. J. Tully and J. B. P. Heywood, “Lean-Burn Characteristics of a Gasoline Engine
Enriched with Hydrogen from a Plasmatron Fuel Reformer,” 2003,
https://doi.org/10.4271/2003-01-0630.
[35] R. Bahreini et al., “Characterizing emissions and optical properties of particulate matter
from PFI and GDI light-duty gasoline vehicles,” J. Aerosol. Sci., vol. 90, pp. 144–153,
Dec. 2015, https://doi.org/10.1016/j.jaerosci.2015.08.011.
[36] A. A. Quader, “Lean Combustion and the Misfire Limit in Spark Ignition Engines,”
SAE Transactions, vol. 83, pp. 3274–3296, 1974.
[37] A. Wimmer and E. Schnessl, “Effects of Humidity and Ambient Temperature on Engine
Performance of Lean Burn Natural Gas Engines,” Jan. 2006,
https://doi.org/10.1115/ICEF2006-1559.
[38] R. D. Reitz, R. Hanson, D. Splitter, and S. Kokjohn, “High-Efficiency, Ultra-Low
Emission Combustion in a Heavy-Duty Engine via Fuel Reactivity Control,” p. 30.
[39] M. Baratta, N. Rapetto, E. Spessa, A. Fuerhapter, and H. Philipp, “Numerical and
Experimental Analysis of Mixture Formation and Performance in a Direct Injection
CNG Engine,” Apr. 2012, pp. 2012-01–0401, https://doi.org/10.4271/2012-01-0401.
[40] B. Yadollahi and M. Boroomand, “The effect of combustion chamber geometry on
injection and mixture preparation in a CNG direct injection SI engine,” Fuel, vol. 107,
pp. 52–62, May 2013, https://doi.org/10.1016/j.fuel.2013.01.004.
[41] M. Baratta, et al., “Multi-Dimensional Modeling of Direct Natural-Gas Injection and
Mixture Formation in a Stratified-Charge SI Engine with Centrally Mounted Injector -
Google Search,” SAE Int. J. Engines, vol. 1, no. 1, pp. 607–626,
https://doi.org/10.4271/2008-01-0975.

107
[42] M.Chiodi, H. J.Berner, and M.Bargende M, “Investigation on different injection
strategies in a direct-injected turbocharged CNG-engine,” SAE Technical Paper Series,
No. 2006-01-3000, Sep. 14 2006, https://doi.org/10.4271/2006-01-3000.
[43] Y. Goto, “Mixture Formation and Ignition in a Direct Injection Natural Gas Engine,”
JSME International Journal Series B, vol. 42, no. 2, pp. 268–274, 1999,
https://doi.org/10.1299/jsmeb.42.268.
[44] A. A. R. Aziz, F. Firmansyah, and R. Shahzad, “Combustion Analysis of a CNG Direct
Injection Spark Ignition Engine,” International Journal of Automotive and Mechanical
Engineering, vol. 2, pp. 157–170, Jul. 2010,
https://doi.org/10.15282/ijame.2.2010.5.0013.
[45] S. Cordiner, G. Simone, and V. Mulone, “Experimental-Numerical Analysis of Nitric
Oxide Formation in Partially Stratified Charge (PSC) Natural Gas Engines,” SAE
International Journal of Engines, vol. 2, pp. 326–340, Oct. 2009,
https://doi.org/10.4271/2009-01-2783.
[46] T. Mohamad, M. Harrison, M. Jermy, and H. Heoy Geok, “The structure of the high-
pressure gas jet from a spark plug fuel injector for direct fuel injection,” Journal of
Visualization, vol. 13, pp. 121–131, May 2010, https://doi.org/10.1007/s12650-009-
0017-2.
[47] H. R. Ricardo, “Internal-combustion engine,” US1271942A, Jul. 09, 1918.
[48] L. A. Gussak, V. P. Karpov, and Yu. V. Tikhonov, “The Application of Lag-Process in
Prechamber Engines,” Feb. 1979, p. 790692, www.jstor.org/stable/44699063.
[49] J. B. Heywood, Internal combustion engine fundamentals. New York: McGraw-Hill,
1988.
[50] S. Yamaguchi, N. Ohiwa, and T. Hasegawa, “Ignition and burning process in a divided
chamber bomb,” Combustion and Flame, vol. 59, pp. 177–187, Feb. 1985,
https://doi.org/10.1016/0010-2180(85)90023-9.
[51] J. Getzlaff, J. Pape, C. Gruenig, D. Kuhnert, and R. Latsch, “Investigations on Pre-
Chamber Spark Plug with Pilot Injection,” Apr. 2007, pp. 2007-01–0479,
https://doi.org/10.4271/2007-01-0479.
[52] C. Robinet, P. Higelin, B. Moreau, O. Pajot, and J. Andrzejewski, “A New Firing
Concept for Internal Combustion Engines: ‘I’APIR,’” Mar. 1999, pp. 1999-01–0621,
https://doi.org/10.4271/1999-01-0621.
[53] J. Park, J. Yeom, G. Han, and S. Y. Chung, “Fundamental Study on the Radical Ignition
Technique Using Constant Volume Chamber ( CVC ),” 2006.
[54] S.-S. Chung, J.-Y. Ha, J.-S. Park, M.-J. Lee, and J.-K. Yeom, “Rapid Bulk Combustion
of Lean Premixture by Using Radical Injection Method and an Application to an Actual
Engine,” Oct. 2003, pp. 2003-01–3212, https://doi.org./10.4271/2003-01-3212.
[55] D. Ha, J. Park, J. Yeom, J. Ha, and S. Chung, “Study on Combustion and Emission
Characteristics of CNG Fueled RI-Engine,” Aug. 2007, pp. 2007-01–3621,
https://doi.org/10.4271/2007-01-3621.
[56] K. Cupiał, A. Jamrozik, and A. Spyra, “Single and two–stage combustion system in the
SI test engine,” 2002.

108
[57] A. Jamrozik and W. Tutak, “A study of performance and emissions of SI engine with a
two-stage combustion system,” CHEMICAL AND PROCESS ENGINEERING, vol. 32,
pp. 453–471, Dec. 2011, doi: 10.2478/v10176-011-0036-0.
[58] R. Jarnicki, P. Bocian, and T. J. Rychter, “Theoretical analysis of ignition of a gas fuel
jet with the use of an ignition chamber,” 2001.
[59] R. Latsch, “The Swirl-Chamber Spark Plug: A Means of Faster, More Uniform Energy
Conversion in the Spark-Ignition Engine,” Feb. 1984, p. 840455,
https://doi:.org/10.4271/840455.
[60] M. Kettner, M. Rothe, A. Velji, U. Spicher, D. Kuhnert, and R. Latsch, “A New Flame
Jet Concept to Improve the Inflammation of Lean Burn Mixtures in SI Engines,” SAE
Technical Papers, Oct. 2005, https://www.jstor.org/stable/44722103.
[61] J. Geiger, S. Pischinger, R. Böwing, H.-J. Koß, and J. Thiemann, “Ignition Systems for
Highly Diluted Mixtures in SI-Engines,” Mar. 1999, pp. 1999-01–0799,
https://doi.org/10.4271/1999-01-0799.
[62] M. Bunce et al., “Efficiency and Emissions Benefits of Ultra-Lean Engine Operation as
Enabled by Jet Ignition,” p. 19.
[63] E. Toulson, H. C. Watson, and W. P. Attard, “Modeling Alternative Prechamber Fuels
in Jet Assisted Ignition of Gasoline and LPG,” Apr. 2009, pp. 2009-01–0721,
https://doi.org/10.4271/2012-32-0002.
[64] A. Boretti, “Direct Injection and Spark Controlled Jet Ignition to Convert A Diesel
Truck Engine to LPG,” SAE Technical Papers, Oct. 2010, https://doi.org/10.4271/2010-
01-1976.
[65] A. Boretti, H. Watson, and A. Tempia, “Computational analysis of the lean-burn direct-
injection jet ignition hydrogen engine,” Proc. Institution of Mechanical Engineers, Part
D: Journal of Automobile Engineering, vol. 224, no. 2, pp. 261–269, Feb 1, 2010,
https://doi.org/10.1243/09544070jauto1278.
[66] A. A. Boretti and H. C. Watson, “The Lean Burn Direct-Injection Jet-Ignition Flexi Gas
Fuel LPG/CNG Engine,” Nov. 2009, pp. 2009-01–2790, https://doi.org/10.4271/2009-
01-2790.
[67] A. A. Boretti and H. C. Watson, “The lean burn direct injection jet ignition gas engine,”
International Journal of Hydrogen Energy, vol. 34, no. 18, pp. 7835–7841, Sep. 2009,
https://doi.org/10.1016/j.ijhydene.2009.07.022.
[68] B. C. Thelen, D. Chun, E. Toulson, and T. Lee, “A Study of an Energetically Enhanced
Plasma Ignition System for Internal Combustion Engines,” IEEE Transactions on
Plasma Science, vol. 41, no. 12, pp. 3223–3232, Dec. 2013,
https://doi.org/10.1109/TPS.2013.2288204.
[69] F. A. (Ferran A. Ayala, “Combustion lean limits fundamentals and their application to
a SI hydrogen-enhanced engine concept,” Thesis, Massachusetts Institute of
Technology, 2006.
[70] O. A. Uyehara, “Prechamber for Lean Burn for Low NOx,” Feb. 1995, p. 950612,
https://doi.org/10.4271/950612.

109
[71] E. Toulson, H. C. Watson, and W. P. Attard, “The Effects of Hot and Cool EGR with
Hydrogen Assisted Jet Ignition,” Aug. 2007, pp. 2007-01–3627, doi: 10.4271/2007-01-
3627.
[72] J. Hynes, “Turbulence effects on combustion in spark ignition engines,” phd, University
of Leeds, 1986.
[73] D. B. Wimmer and R. C. Lee, “An Evaluation of the Performance and Emissions of a
CFR Engine Equipped with a Prechamber,” Feb. 1973, https://doi.org/10.4271/730474.
[74] E. Toulson, H. C. Watson, and W. P. Attard, “Gas Assisted Jet Ignition of Ultra-Lean
LPG in a Spark Ignition Engine,” Apr. 2009, pp. 2009-01–0506, doi:
https://doi.org/10.4271/2009-01-0506.
[75] E. Toulson, H. C. Watson, and W. P. Attard, “The Lean Limit and Emissions at Near-
Idle for a Gasoline HAJI System with Alternative Pre-Chamber Fuels,” Sep. 2007, pp.
2007-24–0120, doi: https://doi.org./10.4271/2007-24-0120.
[76] W. P. Attard, E. Toulson, A. Huisjen, X. Chen, G. Zhu, and H. Schock, “Spark Ignition
and Pre-Chamber Turbulent Jet Ignition Combustion Visualization,” Apr. 2012, pp.
2012-01–0823, https://doi.org/10.4271/2012-01-0823.
[77] B. V. Lande and S. Kongre, “The effect of advanced ignition timing on ethanol-gasoline
blended spark ignition engine,” in 2016 International Conference on Electrical,
Electronics, and Optimization Techniques (ICEEOT), Mar. 2016, pp. 19–26,
https://doi.org/10.1109/ICEEOT.2016.7755071.
[78] L. A. Gussak, V. P. Karpov, and Yu. V. Tikhonov, “The Application of Lag-Process in
Prechamber Engines,” Feb. 1979, p. 790692, https://doi.org/10.4271/790692.
[79] R. P. Roethlisberger and D. Favrat, “Comparison between direct and indirect
(prechamber) spark ignition in the case of a cogeneration natural gas engine, part I:
engine geometrical parameters,” Applied Thermal Engineering, vol. 22, no. 11, pp.
1217–1229, Aug. 2002, https://doi.org/10.1016/S1359-4311(02)00040-6.
[80] W. P. Attard, N. Fraser, P. Parsons, and E. Toulson, “A Turbulent Jet Ignition Pre-
Chamber Combustion System for Large Fuel Economy Improvements in a Modern
Vehicle Powertrain,” SAE International journal of engines, May 2010,
https://doi.org/10.4271/2010-01-1457.
[81] Y. Sakai, K. Kunii, S. Tsutsumi, and Y. Nakagawa, “Combustion Characteristics of the
Torch Ignited Engine,” SAE Transactions, vol. 83, pp. 3504–3512, 1974.
[82] H. Ryu and T. Asanuma, “Combustion analysis with gas temperature diagrams
measured in a prechamber spark ignition engine,” in Symposium (International) on
Combustion, 1985, vol. 20, pp. 195–200.
[83] A. Shah, P. Tunestal, and B. Johansson, “Effect of Pre-Chamber Volume and Nozzle
Diameter on Pre-Chamber Ignition in Heavy Duty Natural Gas Engines,” Apr. 2015,
pp. 2015-01–0867, https://doi.org/10.4271/2015-01-0867.
[84] G. Gentz, B. Thelen, P. Litke, J. Hoke, and E. Toulson, “Combustion Visualization,
Performance, and CFD Modeling of a Pre-Chamber Turbulent Jet Ignition System in a
Rapid Compression Machine,” SAE Int. J. Engines, vol. 8, no. 2, pp. 538–546, Apr.
2015, https://doi.org/10.4271/2015-01-0779.

110
[85] “Internal Combustion Engines Tutorial Guide,” ANSY S, Inc., vol. Release 18.0, no. ISO
9001: 2008, p. 380, Jan. 2017, doi: ansy sinfo@ansys.com.
[86] T. Ahmad, S. L. Plee, and J. P. Myers, “Fluent Theory Guide,” p. 814.
[87] W. Kurniawan, S. Abdullah, and A. Shamsudeen, “A Computational Fluid Dynamics
Study of Cold-flow Analysis for Mixture Preparation In a Motored Four-stroke Direct
Injection Engine,” Journal of Applied Sciences, vol. 7, Dec. 2007, doi:
10.3923/jas.2007.2710.2724.
[88] J. C. Wall and J. B. Heywood, “The Influence of Operating Variables and Prechamber
Size on Combustion in a Prechamber Stratified-Charge Engine,” Feb. 1978, p. 780966,
https://doi.org/10.4271/780966.

111