Fuel 358 (2024) 130296

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Performance and emission analysis of ammonia-ethanol and

ammonia-methane dual-fuel combustion in a spark-ignition engine: An
optical study
Kalim Uddeen a, Qinglong Tang b, *, Hao Shi c, James Turner a
a
Clean Combustion Research Center, Physical Science and Engineering Division, King Abdullah University of Science and Technology, Thuwal 23955-6900, Saudi Arabia
b
State Key Laboratory of Engines, Tianjin University, Tianjin 300072, China
c
Reactive Flows and Diagnostics, Technische Universitat Darmstadt, Otto-Berndt-Str.3, Darmstadt 64287, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: In light of the pressing need to address climate change, it is imperative to take immediate action to reduce
SI engines conventional hydrocarbon fuel demand. In pursuit of achieving decarbonization goals, it has become apparent
Ammonia combustion that modern internal combustion engines need to adopt either carbon–neutral fuels or fuels with lower hydro­
Ammonia-ethanol blends
carbon content. To this end, there are several options available including alternatives such as ammonia (NH3),
Ammonia-methane blends
Flame speed
hydrogen (H2), methane (CH4), methanol (CH3OH), and ethanol (C2H5OH). These fuels offer the potential to be
Pollutant emissions produced in a manner that is either carbon–neutral (the latter two) or entirely free of carbon emissions (the
former two), making them an ideal choice for powering advanced IC engines with minimum climate impact. The
use of ammonia as a fuel has the potential to significantly reduce the demand for conventional fuels and decrease
the emission of harmful pollutants like CO, CO2, particulates, and unburned hydrocarbons during combustion.
However, as a combustion fuel, it poses several challenges, including a low burning velocity, narrow flamma­
bility range, and instabilities arising during the combustion process. This study compared the two different dual
fuel approaches, mixing ammonia-ethanol and ammonia-methane, and investigated their effect on engine per­
formance and emissions. The experiments were conducted on a light-duty, single-cylinder, and spark-ignition
four-stroke engine with an optically accessible cylinder head equipped with a port-fuel injection system. The
results show that blending ethanol and methane with ammonia significantly improves its combustion charac­
teristics due to the higher flame speed of the added components. Ammonia-ethanol blends produced reduced
combustion duration, lower combustion instability, and higher engine efficiency compared to ammonia-methane
blends due to the higher flame speed of ethanol. Additionally, ammonia-ethanol blends also produced higher
NOx and CO2 emissions due to the higher in-cylinder temperature and lower H/C ratio. The study also applied
high-speed natural flame luminosity imaging to observe the flame propagation speed for various fuel blending
cases. The results found that ammonia-ethanol blends produced higher flame propagation speed than ammonia-
methane blends.

1. Introduction environment, our communities, and the biodiversity of our planet [2].
The combustion of hydrocarbon (HC) fuels release harmful pollutants
As a responsible and conscientious society, we must acknowledge such as oxides of nitrogen (NOx), unburned hydrocarbon (UHC), carbon
and confront the harsh realities of climate change, the increasing costs of dioxide (CO2), carbon monoxide (CO), and particulate matter (PM),
fossil fuel, and the associated constraints of fuel supply. These pressing which strongly affect the environment [3,4]. Therefore, it is vital to find
issues are largely attributed to our heavy dependence on fossil fuels, clean alternatives to fossil fuels that are not only environmentally
highlighting the imperative to adopt alternative and sustainable energy friendly but also sustainable. Among the most promising clean fuel
sources [1]. It is a fact that the world’s reserves of fossil fuels are sources are hydrogen (H2) and ammonia (NH3), which produce no
dwindling, and their hazardous emissions have inflicted damage to our carbon emissions during combustion [5]. Hydrogen has emerged as a

* Corresponding author.
E-mail address: [email protected] (Q. Tang).

https://doi.org/10.1016/j.fuel.2023.130296
Received 5 September 2023; Received in revised form 17 October 2023; Accepted 2 November 2023
Available online 15 November 2023
0016-2361/© 2023 Elsevier Ltd. All rights reserved.
K. Uddeen et al. Fuel 358 (2024) 130296

Fig. 1. Application of NH3 as a near-zero emission fuel system.

highly versatile fuel option that can efficiently be utilized in both fuel combustion duration and increase engine performance by generating
cells (FCs) and internal combustion (IC) engines. Hydrogen has high multiple ignition jets inside the combustion chamber [19,20]. These
burning velocity, a wide flammability range, and high octane numbers. innovative solutions have the potential to achieve steady and reliable
Theoretically, when hydrogen is burned with oxygen, the only byprod­ combustion in spark-ignition (SI) engines and could have a significant
uct is water. However, if air is present during combustion, the increased impact on the future of automotive technology. In a study conducted by
temperature can result in the formation of nitrogen oxides (NOx) as well Lhuillier et al. [18], the combustion characteristics of ammonia were
[6]. Despite being a highly versatile element, hydrogen exhibits low examined in a spark-ignition (SI) engine. The results showed that, unlike
volumetric energy density, which poses a challenge when it comes to HC fuels, ammonia combustion required advanced spark timing (ST)
storing and transporting it [7]. Moreover, the wide range of its explosion and higher intake pressure to achieve a faster burning velocity. Duyn­
limits necessitates careful handling and strict adherence to safety mea­ slaegher et al. [21] also investigated the effect of charge air temperature
sures to avert potential danger [8]. As a solution to the challenges posed and pressure and found that increasing these factors led to an increase in
by hydrogen, many research efforts are focused on using ammonia as a the burning velocity of ammonia. Silva et al. [22] used three-
fuel for IC engines. dimensional numerical simulation to understand the combustion char­
Ammonia is a useful hydrogen transporter because it has a high acteristics of ammonia. They found that piston design along with tumble
hydrogen density of 17.8 % per unit weight when kept in a liquid state at motion have a great effect on leading faster combustion rate at advanced
1.1 MPa and 300 K [9]. This feature makes it an interesting carbon-free ignition timing.
fuel that has several advantages over traditional fossil fuels. It is Additionally, other studies have shown that increasing compression
completely free from CO2, oxides of sulfur (SOx), and soot emissions, ratio (CR) can also lead to faster flame development and improved flame
making it a much cleaner option for the environment [5]. It is producible propagation speed, resulting in a shorter combustion period [16,17]. To
from a wide range of sources, including renewable sources, fossil fuels, achieve stable ammonia combustion and improved engine performance,
and biomass. This versatility makes it an attractive option for industries multiple spark ignition within the combustion chamber may also be a
looking to transition to cleaner energy sources. It is also transportable viable solution [14].
and storable using existing facilities such as fuel tanks, ships, trucks, and Using a dual-fuel approach is an efficient method to improve the
pipelines. This means that it can be easily integrated into existing combustion properties of ammonia. Several more reactive fuels
infrastructure, making it a viable alternative fuel option [3]. Fig. 1 including gasoline, ethanol, methanol, hydrogen, and methane can
summarizes the production, transportation, and utilization of ammonia boost the reactivity of fuel mixtures containing ammonia. The Cooper­
in various sectors. ative Fuels Research (CFR) engines have been employed in several tests
Since the ammonia molecule contains both nitrogen (N) and using ammonia and gasoline mixes, and it was discovered that ammonia
hydrogen (H), theoretically the sole products of a full combustion with may successfully substitute gasoline, resulting in improved engine per­
air are nitrogen gas (N2) and water. Despite these advantages, there are formance and lower CO and UHC emissions [23,24]. Lu et al. [25]
several drawbacks to utilizing ammonia as fuel, which are detailed analyzed the kinetic mechanism of ammonia and methanol combustion
below. [10,11]: and discovered that blending methanol with ammonia in a combustion
system has great potential to increase the reactivity of the mixture due to
• Low flammability range its high flame speed. Lhuillier et al. [12] examined the effect of blending
• Low burning velocity hydrogen with ammonia and discovered that mixing up to 20 % (by
• High cyclic variability during combustion volume) H2 in the fuel mixture reduces misfires during combustion,
• High ignition energy lowers cycle-to-cycle variation, and provides good engine efficiencies
• Production of fuel-bound NOx and NH3 emissions close to stoichiometry due to higher heat transfer losses. Li et al. [26]
• Corrosive in nature investigated blends of ammonia and hydrogen with varying energy ra­
• Hazardous for health tios in an optical SI engine and discovered that, after reaching an opti­
mum value of 7.5 % by energy content, the hydrogen-ammonia energy
There are many ongoing research studies around the world to find ratio causes the reported thermal efficiency to first increase and subse­
ways to overcome these limitations and use it as an effective fuel for quently drop.
combustion. Some of the approaches that are being considered include Several studies have shown that mixing methane (CH4) and ethanol
dual-fuel strategies [12,13], using multiple spark ignition sites [14,15], (C2H5OH) with ammonia can enhance its combustion characteristics
advancing ignition timing [10], increasing compression ratio [16,17], and make it a better fuel combination. For instance, in a study by Kurien
and raising intake temperature and pressure [18]. Some studies showed et al. [27], an increase in methane fraction in the ammonia-methane
that pre-chamber combustion can be an effective way to reduce the blend resulted in improved engine performance due to the higher

2
K. Uddeen et al. Fuel 358 (2024) 130296

Table 1 ammonia into ethanol improved thermal and combustion efficiencies

Engine descriptions [36]. due to lower heat transfer losses. All these studies suggest that a dual-
Description Specification fuel approach could be an effective solution to overcome some of the
challenges associated with ammonia. However, very few studies have
Type Single-cylinder optical research engine
Displacement volume 511 cc been performed on an SI engine using ammonia-ethanol blends and then
Compression ratio (CR) 10.5 compared the results with the performance of ammonia-methane
Bore 85 mm blends.
Stroke 90 mm Ethanol is also a potential alternative fuel that can replace traditional
Piston geometry Flat
Number of valves Intake (2), Exhaust (2)
HC fuels and reduce harmful pollutant emissions. It is a sustainable
Valve gear type DOHC choice because it may be made from waste materials or natural re­
Intake valve Open 10◦ CA bTDC sources. Since ethanol has a greater octane rating than gasoline, it has
Close 50◦ CA aBDC better anti-knock properties and can run at higher compression ratios.
Exhaust valve Open 60◦ bBDC
Along with having a substantially higher heat of vaporization than
Close 0◦ aTDC
gasoline, it also has a higher volumetric efficiency and a stronger charge
cooling effect. Additionally, the increased oxygen level promotes a more
complete and clean combustion process, which reduces CO, HC, and
particle emissions [30]. Similarly, methane is another favorable fuel for
IC engines due to its physical and chemical properties. It has a higher
hydrogen-to-carbon ratio, which itself reduces CO2 emissions [30]. As
well as a lack of studies conducted on ammonia-ethanol blends
compared to ammonia-methane blends in SI engines, there is limited
research available on the high-speed imaging of ammonia blended with
more reactive fuels. Studies on the effects of hydrocarbon fuel addition
on the flame speed of ammonia are also rare.
Therefore, this research work investigates the combustion charac­
teristics of ammonia-ethanol and ammonia-methane blends in an optical
SI engine, reporting the results in terms of energy fraction. To do so,
high-speed natural flame luminosity (NFL) imaging was employed,
which allows the study of flame propagation for the pure ammonia case
as well as for various dual-fuel blending cases. Additionally, this work
compares the exhaust pollutant emissions, including NOx, UHC, and
CO2, for both ammonia-ethanol and ammonia-methane mixing cases
under similar operating conditions.

2. Experimental setup

2.1. Engine operating system

An optical research engine (AVL-5402) with a single cylinder was

managed using an AVL PUMA Open Automation System (V2012) [31].
Fig. 1 and Table 1 show, respectively, the engine schematic and details
[32]. A spark plug (ER8EH, NGK) positioned at the top of the cylinder
head initiated the combustion. Additionally, the top of the cylinder head
Fig. 2. Schematic of engine setup. was equipped with a pressure transducer (GU22CK, AVL), which
monitored and recorded the in-cylinder pressure [33]. As shown in
Fig. 2, a flat quartz (Suprasil 2 grade B) flask-bottomed piston crown
Table 2 connected to an extended Bowditch piston was used [34]. To take pic­
High-speed imaging system specification. tures of the flame in the combustion chamber from the bottom, a 45◦ UV
Description Specification mirror was positioned below the Bowditch piston [35].
Camera type SA4, Photron
Frame rate 36,000 fps 2.2. High-speed natural flame luminosity (NFL) imaging
Frames per crank angle 5
Exposure time 27.8 μs The Photron SA4 camera with a Nikon lens (50 mm, f 1.2), captured
Shutter speed 1/36000 s
the flame propagation visible through the bottom of the quartz piston, as
Image resolution 320 × 272
Recorded images per cycle 400 seen in Fig. 1. The camera was synced with the engine operation using
Camera trigger timing Same as spark timing trigger signals from the AVL Indicom. Each case involved firing the
engine continuously for 200 cycles. The camera had a resolution of 320
× 272 and was programmed to shoot at a rate of roughly 36,000 frames
flame speed of methane and thus it reduced cyclic variation. Another per second. The camera exposure time was 27.8 μs, and there were 400
study by Oh et al. [28] focused on the viability of a maritime dual-fuel, photos every cycle, with a crank angle of 0.2 CAD between two subse­
spark-ignited engine using natural gas and ammonia. When compared to quent photos. Table 2 lists the properties of the high-speed imaging
pure natural gas, the researchers discovered that using ammonia as a system.
replacement fuel dramatically lowered CO2 emissions by 18 %–28 %.
Moreover, Pelé et al. [29] used direct injection of ethanol combined with 2.3. Operating conditions
ammonia stored in a tank to study the combustion and exhaust char­
acteristics of a spark-ignition engine. They discovered that adding 25 % The experiment was conducted under specific conditions, including a

3
K. Uddeen et al. Fuel 358 (2024) 130296

Table 3 monitored and maintained at 90 ± 1 ◦ C by the PUMA automation sys­

Operating parameters. tem. The IAV FI2RE system was used to control the spark ignition. For
Engine speed 1200 ± 1 rpm each case, 200 continuous fired cycles were recorded using the AVL data
acquisition system. Software programs AVL CONCERTO 5.0 and MAT­
Intake temperature 25 ± 1 ◦ C
Intake pressure 1 ± 0.15 bar LAB were used to process the data. Table 3 contains a complete list of
Oil and coolant temperatures 90 ± 1 ◦ C operational parameters.
Ammonia and methane injection timing During intake valve opening
Ammonia injection pressure 4 ± 0.2 bar
Ethanol injection timing − 330 CAD aTDC 2.4. Fuel supply system
Ethanol injection pressure 6 ± 0.1 bar
Global air–fuel equivalence ratio (λ) 1
Ammonia was kept in a liquid state in the tank and then supplied as
gas at 4 ± 0.2 bar via a pressure regulator. For the ammonia-methane
case, three different MFCs were employed to operate the flow of
Table 4 ammonia, methane, and air separately, and their blends were intro­
Fuel properties [9,21,25].
duced into a fuel line connected to the intake port. The gaseous fuel
Fuel Ammonia Methane Hydrogen Ethanol Gasoline entered the combustion chamber during the intake valve-opening
(gas) (gas) (gas) (liquid) (liquid) period. However, for the ammonia-ethanol case, the ammonia-air
Chemical NH3 CH4 H2 C2H5OH – mixture entered into the fuel line by MFCs, and the ethanol was injec­
formula ted at an injection timing of − 330 CAD aTDC by a port fuel injector at a 6
Density (kg/ 0.76 0.72 0.08 790 720–775
m3) ± 0.1 bar injection pressure, i.e. during the intake stroke. The energy
Lower heating 18.6 50 120 26.7 42–44
value (MJ/ fraction (EF) of hydrocarbon (HC) fuel of ethanol or methane is defined
kg) in equation (1). Additionally, detailed properties of various fuels are
Laminar 7.00 37 291 41 35 provided in Table 4, including those for hydrogen and typical gasoline
burning for comparison purpose
velocity-close
to Energy of HC fuel × 100
stoichiometry EF (LHV%) = (1)
Energy of HC fuel + Energy of ammonia
(cm/s)
Stoichiometric 6.05 17.2 34.3 9 14.7
air–fuel ratio 3. Results and discussion
by mass
Minimum 8.00 0.28 0.02 0.22 0.24
3.1. Combustion characteristics of ethanol mixed with ammonia
ignition
energy (mJ)
Minimum auto- 650 630 585 423 257 This study investigated the combustion characteristics obtained by
ignition blending ethanol with ammonia at different STs of − 20, − 25, and − 30
temperature
CAD aTDC. As shown in Fig. 3, the combustion phasing (CA50) and
(oC)
Research octane >130 >120 >100 >108 >90
combustion duration (CA10-CA90) are influenced by both the ethanol
number energy fraction and ST. The crank angle that corresponds to where 10 %,
(RON) 50 %, and 90 % of the mixture mass fraction is burned are represented
Flammability 15–28 5.3–14 4–75 4.3–19 1.4–7.6 by CA10, CA50, and CA90 respectively. Blending sufficient ethanol fuel
limits in air
reduced the combustion phasing and duration significantly by
(vol. %)
increasing the flame temperature and burning velocity, and the opti­
mum values can be achieved by adjusting the ignition timing. A signif­
fixed engine speed of 1200 ± 1 rpm, and an intake temperature and icant reduction in CA50 and CA10-CA90 can be seen by changing the ST
pressure that were naturally aspirated. To ensure accurate measure­ from − 20 to − 30 CAD aTDC because the flame generated by more
ments, an air heater and mass flow controller (MFC) were used to advanced ignition timing led to higher in-cylinder pressure and tem­
regulate the input air temperature and pressure at 25 ± 1 ◦ C and 1 ± perature earlier in the cycle, which increased the burning rate of the
0.15 bar, respectively. The oil and water temperatures were also closely charge. When testing ST of − 20 CAD aTDC, the CA50 values went down

Fig. 3. CA50 (a) and CA10-CA90 (b) variations for various ethanol energy fractions at different STs.

4
K. Uddeen et al. Fuel 358 (2024) 130296

Fig. 4. Pmax (a), IMEPnet (b), COVIMEP (c), and net indicated efficiency (d) variations for various ethanol energy fractions at different STs.

Fig. 5. In-cylinder pressure at ST of − 25 CAD aTDC (a) and ST of − 30 CAD aTDC (b) for different ethanol energy fractions.

5
K. Uddeen et al. Fuel 358 (2024) 130296

Fig. 6. HRR at ST of − 25 CAD aTDC (a) and ST of − 30 CAD aTDC (b) for different ethanol energy fractions.

Fig. 7. Flame propagation images from a typical cycle for various ammonia-ethanol blending cases at ST of − 20 CAD aTDC.

from 49.7 to 25.2 CAD as the combustion shifted from pure ammonia to The results in Fig. 4 demonstrate that the maximum in-cylinder
a 50 % blend of ethanol with ammonia. This data is displayed in Fig. 3 pressure (Pmax), net indicated mean effective pressure (IMEPnet), co­
(a). However, at ST of − 30 CAD aTDC, the CA50 decreased from 38.6 to efficient of variation (COV) of IMEP, and net indicated efficiency can all
7.7 CAD respectively. Fig. 3 (b) also displays CA90-CA10 values, which be greatly raised by mixing ethanol and ammonia and utilizing earlier
demonstrated a similar reduction trend. At a 50 % ethanol blend, the spark timings (STs). The faster burning rate of the charge associated
minimum values of 56.4, 51.2, and 38.1 CAD were observed at STs of with the ammonia-ethanol mixture led to higher in-cylinder pressure,
− 20, − 25, and − 30 CAD aTDC respectively. with a maximum pressure of 37.2 bar observed for the 50 % ethanol

6
K. Uddeen et al. Fuel 358 (2024) 130296

300 100
0% Ethanol 0% Ethanol
25% Ethanol 90 25% Ethanol
30% Ethanol 30% Ethanol
250

Flame Area Proporation (%)

35% Ethanol 80 35% Ethanol

Mean Flame Intensity (a.u)

40% Ethanol 40% Ethanol
50% Ethanol 70 50% Ethanol
200
60

150 50

40
100
30

20
50
10

0 0
-20 -10 0 10 20 30 40 -20 -10 0 10 20 30 40
Crank Angle (CAD aTDC) Crank Angle (CAD aTDC)

(a) Mean flame intensity (b) Flame area proportion

Fig. 8. Flame intensity (a) and flame area proportion (b) for different ammonia-ethanol energy fractions at ST of − 20 CAD aTDC.

more than 20 % for pure ammonia combustion even at advanced ST,

because the low burning velocity of ammonia produced a longer com­
bustion duration, which caused unstable combustion. The COV was used
to define the combustion stability of the engine, and a maximum value of
5 % was chosen as the acceptable range. Due to pure ammonia’s slow
flame speed, which results in a prolonged combustion duration and
unstable combustion, the results showed a COV value of more than 20 %
for pure ammonia combustion even at advanced ST. However, a sig­
nificant decline in COVs can be observed by adding ethanol to the
ammonia fuel mixture, which increased its reactivity and burned the
charge more effectively. The ST of − 30 CAD aTDC, with an ethanol
blend of greater than 30 %, produced the lowest COVs. The impact of
ammonia-ethanol mixes on net indicated efficiency with different STs is
also shown in Fig. 4 (d). When ST was modified from − 20 to − 30 CAD
aTDC, the efficiency for the pure ammonia case increased from 10.8 % to
18.3 %. However, the 40 % ammonia-ethanol combination had peak
efficiencies of 27.9 % and 29.2 % at ST of − 25 and − 30 CAD aTDC,
respectively.
In Fig. 5, there is a clear representation of the impact that different
blends of ammonia-ethanol have on the in-cylinder pressure for two
different STs. The graph shows the average in-cylinder pressure for 20
Fig. 9. Flame propagation speed for different ammonia-ethanol energy frac­
tions at ST of − 20 CAD aTDC.
consecutive firing cycles. It is evident from the data that the addition of
30 % ethanol to ammonia resulted in a noticeable rise in pressure.
Furthermore, increasing the ethanol fraction beyond this point led to
energy fraction case at ST of − 30 CAD aTDC, as shown in Fig. 4 (a). The
faster flame propagation and stimulated higher in-cylinder pressure
results also showed that adding ethanol to the fuel mixture led to higher
during the combustion process, as can be seen in Fig. 5 (a). Additionally,
IMEP values, with a near-linear increase in IMEP values from 2.4 bar to
it is worth noting that advancing the ignition timing from − 25 to − 30
5.8 bar seen when the combustion changed from pure ammonia to 50 %
CAD aTDC had a significant impact on the pressure rise rate. The highest
ethanol blend for an ST of − 20 CAD aTDC, as shown in Fig. 4 (b).
peak pressure was achieved for the 50 % ethanol fraction, as shown in
However, peak IMEP values of 6.0 bar to 6.3 bar were observed at 40 %
Fig. 5 (b).
ethanol energy fraction for ST of − 25 and − 30 CAD aTDC, with a slight
Similarly, Fig. 6 presents the heat release rate (HRR) for different
decrease in IMEP value seen when further adding ethanol amount in the
ammonia-ethanol energy ratios at two distinct STs. Pure ammonia
mixture due to overly advanced ignition causing increased negative
combustion produced a lower peak of HRR and longer combustion
work or higher heat transfer losses, etc.
duration, because of its slower burning rate. Ethanol addition acceler­
In addition, Fig. 4 (c) and (d) present the COV and net indicated
ated ammonia flame propagation, which in turn instigated higher HRR
efficiency results which are defined in equations (2) and (3) as:
leading to a shorter combustion duration. It is very evident from the
σ (IMEPnet ) results that ST has a strong effect on HRR, the maximum HRR being
COV of IMEPnet = × 100 (2)
μ (IMEPnet ) found at the earlier ST of − 30 CAD aTDC. These findings suggest that a
blend of ethanol and ammonia could be a promising fuel for improving
ηnet =
IMEPnet × Vd
× 100 (3) engine performance, particularly in terms of peak pressure and HRR.
Mf × QLHV
3.1.1. Flame propagation characteristics for various ammonia-ethanol
Where σ and μ represent the standard deviation and mean values and
blending cases
Vd, Mf, and OLHV indicate the swept volume of the engine, mass of fuel,
Fig. 7 shows a clear illustration of the effects of various ammonia-
and lower heating value of fuel, respectively. The engine combustion
ethanol blends on the high-speed NFL images. Specifically, the images
stability was defined by COV and a maximum value of 5 % was selected
display the combustion process for blending ratios ranging from 0 % to
as acceptable combustion stability. The results showed a COV value of
50 % ethanol with ammonia. For each blending case, a single cycle with

7
K. Uddeen et al. Fuel 358 (2024) 130296

Fig. 10. Flame probability distribution map for various ammonia-ethanol blending cases at ST of − 20 CAD aTDC (a) and ST of − 30 CAD aTDC (b).

an IMEP value close to the mean IMEP value was selected, and the high- to its lower flame speed and the 50 % ethanol blend showed the
speed images were processed with a fixed crank angle window of − 5 to maximum flame intensity due to the faster flame propagation rate
30 CAD aTDC. The results showed that pure ammonia combustion caused by its higher heat release rate. Furthermore, Fig. 8 (b) presents
produced a significantly slower flame propagation rate; as a result, the the flame area proportion for various ethanol-blending cases. The slower
combustion duration was prolonged. However, by increasing the burning rate of pure ammonia occupied only about 63 % of the flame
ethanol fraction in the mixture the combustion rate was dramatically area proportion up to 30 CAD aTDC. However, mixing ethanol at about
improved. For the 50 % ethanol case, a major portion of the charge had 25 % to 30 % by energy led to a significant rise in the flame area pro­
been efficiently burned by 30 CAD aTDC. These findings highlight the portion at 30 CAD aTDC, and covered about 83 % to 91 % respectively.
potential benefits of utilizing ammonia-ethanol blends in spark-ignition Moreover, adding ethanol to ammonia at 50 % energy fraction further
combustion processes. improved the reactivity of the mixture and stimulated faster flame
Fig. 8 shows the effect of ammonia-ethanol blends on mean flame propagation, so that most of the combustion chamber area was engulfed
intensity and flame area proportion for the same NFL images as dis­ by about 25 CAD aTDC.
played in Fig. 7. An increase in flame intensity can be seen at about − 10 Fig. 9 presents the flame propagation speed by processing the images
CAD aTDC as the flame started growing, and the rate of increase is shown in Fig. 7 for different ammonia-ethanol energy fraction cases at
higher for the ethanol blending cases, with the results being illustrated ST of − 20 CAD aTDC. Here, the flame speed is defined as the first de­
in Fig. 8 (a). The pure ammonia case produced lower flame intensity due rivative of the equivalent flame radius curve obtained from the high-

8
K. Uddeen et al. Fuel 358 (2024) 130296

Fig. 11. IMEPnet, COV of IMEP, Pmax, and net indicated efficiency for various ammonia-methane energy fractions at ST of − 20 CAD aTDC.

dR
S= (5)
dt
Where, R, A, and S represent the equivalent radius, projected flame
area, and flame speed, respectively. It is clear from the graph that
blending ethanol with ammonia led to a significant increase in flame
propagation speed. The pure ammonia case exhibited a peak flame
speed of 7.5 m/s but adding 25 % to 30 % ethanol gave an increase of 12
% to 18.7 % in flame speed, respectively. Furthermore, 40 % and 50 %
ethanol fraction cases gave a maximum flame speed of 10.8 m/s and
11.1 m/s, respectively.

3.1.2. Flame probability distribution

The probability distribution (Pd) approach was used to evaluate the
properties of flame propagation. For each case, 20 cycles at the same
crank angle were converted into a grayscale image using the “gray­
thresh” function in MATLAB programming. After that, using the same
crank angle, the images were binarized and merged. The probability
distribution of each pixel at the same crank angle was then calculated
using equation (6) as follows:
∑N x
x=1 n(i,j)
Pd(i,j) = (6)
Fig. 12. Combustion duration variations for various ammonia-methane energy N
fractions at ST of − 20 CAD aTDC.
Where n is the intensity of the binarized visuals, which only display
0 or 1, and x is the cycle number. N is the total number of cycles. The
speed NFL images and then smoothed appropriately. Ref. [9] described a coordinates of each pixel are represented here by i and j. A jet color map
detailed method of evaluating flame propagation speed. The equivalent was also used, with red denoting the flame zones (probability 1) and
flame radius and flame propagation speed are defined by equations (4) blue denoting the unburned zone (probability 0). In Fig. 10, the distri­
and (5) as: bution of flames is displayed for different blends of ammonia and
√̅̅̅ ethanol at two distinct STs of − 20 and − 30 CAD aTDC. For the − 20 CAD
A
R= (4) aTDC ST case, a crank angle window of − 10 to 20 CAD aTDC was
π

9
K. Uddeen et al. Fuel 358 (2024) 130296

35 20
0% Methane 0% Methane
30% Methane 30% Methane
30 40% Methane 40% Methane
50% Methane 50% Methane
16

In-cylinder Pressure (bar)

60% Methane 60% Methane
25

Spark Timing

HRR (J/CAD)
12

Spark Timing
20

15
8

10

4
5

0 0
-40 -30 -20 -10 0 10 20 30 40 50 -40 -30 -20 -10 0 10 20 30 40 50
Crank Angle (CAD aTDC) Crank Angle (CAD aTDC)

(a) In-cylinder pressure (b) HRR

Fig. 13. In-cylinder pressure (a) and HRR (b) variations for different ammonia-methane energy fractions at ST of − 20 CAD aTDC.

Fig. 14. Flame propagation images from a typical cycle for various ammonia-methane blending cases at ST of − 20 CAD aTDC.

selected to display the flame development for each ethanol energy also considered a beneficial fuel for IC engines in terms of lower
fraction, as seen in Fig. 9 (a). The probability map indicates that small pollutant emissions than other HC fuels. It is the main constituent of
flame kernels were evolving in the chamber at − 10 to − 5 CAD aTDC for natural gas. Since it is a gaseous fuel, mixing methane with ammonia
each case. However, increasing the ethanol fraction in the mixture gives a good homogeneous mixture for combustion. As per the prior
significantly improved the flame propagation rate, and most of the discussion with ethanol, this work investigated the effect of blending
charge was burned at about 20 CAD aTDC, which was not observed for methane with ammonia in terms of various energy fractions such as 0 %
pure ammonia. Furthermore, Fig. 10 (b) shows the flame distribution (pure ammonia), 30 %, 40 %, 50 %, and 60 %. Fig. 11 shows net IMEP,
map obtained with an ignition timing of − 30 CAD aTDC. This case had a COV of IMEP, Pmax, and net indicated efficiency results obtained at an ST
crank angle window of − 20 to 10 CAD aTDC due to its earlier flame of − 20 CAD aTDC with different methane energy fractions. A significant
generation. Increasing the ethanol fraction from 25 % to 50 % resulted in increase in IMEP values can be seen by mixing more methane in the
a significant increase in the burning rate, and the flame could mixture because the higher burning velocity of methane led to higher in-
completely occupy the combustion chamber area by about 10 CAD cylinder pressure and temperature inside the combustion chamber. The
aTDC. IMEP values were improved from 2.4 bar to 5.7 bar as the methane
fraction changed from 0 % to 60 % respectively. In addition, blending
3.1.3. Combustion characteristics of methane mixed with ammonia methane with ammonia showed a dramatic decrease in COVs, and stable
Methane (CH4) is a single carbon atom-containing molecule, which is combustion at 60 % methane energy fraction in the mixture for a

10
K. Uddeen et al. Fuel 358 (2024) 130296

Fig. 15. Mean flame intensity (a) and flame area proportion (b) for different ammonia-methane energy fractions.

increase in HRR with increasing methane fraction in the ammonia-

methane mixture.

3.1.4. Flame propagation characteristics for various ammonia-methane

blending cases
The NFL images in Fig. 14 display various ammonia-methane
blending cases. A single cycle was selected to process the images, and
a fixed crank angle window of − 5 to 30 CAD aTDC was used for each
blending case, which produced an IMEP value close to the mean IMEP
value. The images demonstrate that burning pure ammonia resulted in
lower flame propagation and longer combustion duration. However,
faster HRR was observed by blending methane with ammonia, which
accelerated charge burning and thus reduced the combustion duration.
Increasing the methane energy fraction from 30 % to 60 % caused a
significant increase in flame propagation, and most of the charge was
burned by around 30 CAD aTDC.
In addition, Fig. 15 exhibits the effect of ammonia-methane blends
on mean flame luminosity intensity and flame area proportion for the
NFL images shown in Fig. 14. Mixing methane with ammonia increased
the flame luminosity because the faster burning rate of methane insti­
gated higher in-cylinder pressure and HRR, which generated higher
Fig. 16. Flame propagation speed for different ammonia-methane energy
luminosity; the results are shown in Fig. 15 (a). 60 %, methane energy
fractions at ST of − 20 CAD aTDC.
fraction produced the maximum flame intensity due to higher HRR
generated. Similarly, Fig. 15 (b) shows the flame area proportion for
particular operating condition was achieved. Furthermore, the effi­ different methane energy factions. It is evident that increasing the
ciency was improved from 10.8 % for the pure ammonia case up to 26.6 methane fraction in the fuel mixture advances the flame area curves and
% for the 60 % methane energy fraction case. shows higher proportions for any given crank angle. The 30 % to 40 %
Fig. 12 presents the various combustion stage durations and CA50 methane blend cases engulfed flame area proportions of 79 % to 89 %
positions for the different energy fraction cases. Here CA10-CA50, respectively at 30 CAD aTDC. However, at 30 CAD aTDC 93 % to 99 % of
CA50-CA90, and CA10-CA90 define the flame propagation phase, fast the flame area was covered when the methane fraction was increased
combustion phase, and the total combustion duration, respectively. The from 50 % to 60 % respectively.
CA50 position was greatly advanced by blending methane with Fig. 16 presents the effect of ammonia-methane blends on flame
ammonia, and the values reduced from 49.7 CAD to 22.8 CAD as propagation speed by processing the images shown in Fig. 14 at ST of
methane fraction changed from 0 % to 60 % respectively. It is also − 20 CAD aTDC. A dramatic rise in flame speed can be observed when
apparent from the figure that the effect of methane addition on CA10- mixing methane with ammonia. In comparison with the pure ammonia
CA50 is more pronounced than on CA50-CA90, which could be case, increasing the methane energy fraction from 30 % to 40 % led to an
because of early faster flame propagation at a higher compression increase in flame speed of about 12 % to 16 % respectively. However,
pressure part of the cycle (due to piston motion). In addition, methane further increasing the methane fraction from 50 % to 60 % showed a
blending greatly reduced the overall combustion duration, and when the flame speed of 9.1 m/s to 10.2 m/s respectively for the same operating
methane fraction increased from 0 % to 60 % the CA10-CA90 shortened conditions.
from 99.7 CAD to 58.1 CAD respectively. The pure ammonia case exhibited a peak flame speed of 7.5 m/s,
As shown in Fig. 13, when the methane fraction in the ammonia- which is more than 100 times higher than the ammonia laminar flame
methane mixture is increased, there is a corresponding increase in speed at 298 K and 1 bar conditions (about 7 cm/s). This is mainly due to
chamber pressure and heat release rate. This is because the higher the higher temperature and turbulence intensity of the in-cylinder
reactivity of the air–fuel mixture caused by the increased methane charge that enhances the flame speed. The peak flame speed increases
content leads to faster burning and higher in-cylinder pressure. The when adding methane or ethanol to ammonia. The ethanol case presents
results in Fig. 13 (b) further confirm this trend, showing a significant

11
K. Uddeen et al. Fuel 358 (2024) 130296

Fig. 17. Flame probability distribution map for various ammonia-methane blending cases at ST of − 20 CAD aTDC.

Fig. 18. Peak of HRR (a) and CA50 (b) variations for different methane energy fractions at ST of − 20 CAD aTDC.

a higher flame speed under the same energy ratio due to the faster because of the lower burning speed of ammonia. However, a significant
laminar flame speed of ethanol than the methane (41 cm/s vs 37 cm/s at increase in the flame region can be seen when adding methane to the
298 K and 1 bar). mixture because it improved the reactivity of the mixture, and in the
Fig. 17 illustrates the flame probability distribution (Pd) map ob­ methane addition cases most of the charge was burned by 20 CAD aTDC.
tained for various ammonia-methane blending cases. For each case, 20
continuous firing cycles were processed as described in Section 3.1.3. A
crank angle window of − 10 to 20 CAD aTDC was chosen to show the 3.2. Comparison of ethanol and methane mixed with ammonia
flame evolution for each case. A small kernel can be seen growing at − 10
CAD aTDC, and increasing the methane fraction in the mixture boosts This study also compares the combustion characteristics obtained by
the flame evolution process, which is more evident at − 5 and 0 CAD blending ammonia-ethanol and ammonia-methane at the same oper­
aTDC. The 0 % methane case showed a slower flame propagation rate ating conditions and common ST of − 20 CAD aTDC. Fig. 18 shows the
peak of HRR and the CA50 obtained for various energy fraction cases. It

12
K. Uddeen et al. Fuel 358 (2024) 130296

Fig. 19. IMEPnet (a), COV of IMEP (b), and Net indicated efficiency (c) variations for different energy fractions at ST of − 20 CAD aTDC.

is apparent from the graphs that blending the HC fuels with ammonia engine power output. At 50 % energy fraction, ammonia-ethanol and
leads to higher HRR and reduced CA50 values because they improve the ammonia-methane cases produced 26.7 % and 22.8 % net indicated
reactivity of the air–fuel mixture; the results can be seen in Fig. 18 (a) efficiency respectively, the results being shown in Fig. 19 (c).
and (b) respectively. Since ethanol has a higher burning velocity than
methane, as a result, blending ethanol with ammonia exhibited higher 3.3. Pollutant emissions
HRR values than the ammonia-methane cases, which further advanced
combustion phasing. This study also compares various pollutant emissions produced when
In Fig. 19, the effect of blending ethanol and methane with ammonia blending ethanol and methane with ammonia under the same operating
on net IMEP, COV of IMEP, and net indicated efficiency is illustrated. conditions. Fig. 20 illustrates the NOx, UHC, and CO2 emissions in the
Fig. 19 (a) displays that the ammonia-ethanol blend case generated exhaust for different dual-fuel blending cases. In the case of pure
higher IMEP values than the ammonia-methane blend case. This is ammonia, most of the NOx originates from fuel-bound nitrogen. How­
because ethanol’s faster flame speed produces higher in-cylinder pres­ ever, mixing ethanol and methane with ammonia instigated higher in-
sure and temperature. Additionally, mixing methane with ammonia cylinder pressure and temperature, which could be a reason for gener­
caused longer combustion duration, leading to higher heat transfer ating more thermal NOx during combustion. Fig. 20 (a) shows that
losses, resulting in higher COV values than the ethanol blending case, as increasing the ethanol and methane fractions led to a dramatic increase
shown in Fig. 19 (b). Comparing the blending of ethanol and methane in NOx emissions. The ammonia-ethanol case generated higher NOx than
with ammonia, ethanol produced higher engine efficiency due to higher the ammonia-methane blends because the oxygen present in ethanol

13
K. Uddeen et al. Fuel 358 (2024) 130296

Fig. 20. NOx (a), UHC (b), and CO2 emissions for different energy fractions at ST of − 20 CAD aTDC.

could more readily oxidize nitrogen present in ammonia. Fig. 20 (b) ethanol or methane improves the overall flame speed and increases the
presents the effect of HC fuel addition to ammonia on total UHC emis­ peak cylinder temperature, which gives rise to more NOx emissions.
sions. The graphs show a monotonic increase in UHC emissions when Since the flame speed of ethanol is higher than that of methane, the
increasing the ethanol or methane fraction in the fuel mixture. ammonia-ethanol cases improve the combustion better, providing lower
Ammonia-methane blends exhibited larger UHC emissions because of COV and UHC emissions than the ammonia-methane cases. More CO2 is
the low combustion efficiency and high COV, which results in more HC formed in the ammonia-ethanol cases compared to the ammonia-
at the exhaust. Furthermore, Fig. 20 (c) compares the CO2 emissions methane cases due to the higher combustion efficiency.
originating from the different blending cases. Mixing HC fuels with
ammonia led to a considerable jump in CO2 levels. Additionally, ethanol 4. Conclusions
has a lower stoichiometric air/fuel ratio and higher C/H ratio than
methane, which means that the ammonia-ethanol case consumes more This study examined the feasibility of dual-fuel combustion using
fuel for the same airflow rate and produces more CO2 emissions. ammonia-ethanol and ammonia-methane in an optical SI engine under
Moreover, the ammonia-ethanol case exhibited a better combustion ef­ identical operating conditions. The flame development for various dual-
ficiency, which resulted in higher CO2 emissions. Therefore, ammonia- fuel blends was observed by high-speed optical imaging. The following
ethanol combustion produced higher CO2 emissions than the is the summary of the main conclusions of this research work:
ammonia-methane case. At 50 % of energy fraction, ammonia-ethanol
combustion produced about 25 % higher CO2 level than ammonia- 1. The combustion of pure ammonia is highly unstable and results in
methane combustion. lower engine efficiency due to its slow flame propagation speed.
In summary, the main issue of ammonia combustion is the low flame However, engine performance was significantly improved by
speed which results in low combustion efficiency and high COV. Adding blending ethanol with ammonia. By changing the combustion from

14
K. Uddeen et al. Fuel 358 (2024) 130296

pure ammonia to a blend of 40 % ethanol energy fraction at an [5] Chai WS, Bao Y, Jin P, Tang G, Zhou L. A review on ammonia, ammonia-hydrogen
and ammonia-methane fuels. Renew Sustain Energy Rev 2021;147:111254.
earlier ST of − 30 CAD aTDC, efficiency improved relatively by 59.6
[6] Verhelst S, Wallner T. Hydrogen-fueled internal combustion engines. Prog Energy
%. Combust Sci 2009;35(6):490–527.
2. Similarly, adding methane to ammonia stimulated higher in-cylinder [7] White C, Steeper R, Lutz A. The hydrogen-fueled internal combustion engine: a
pressure with faster HRR than the pure ammonia case. The result technical review. Int J Hydrogen Energy 2006;31(10):1292–305.
[8] Karim GA. Hydrogen as a spark ignition engine fuel. Int J Hydrogen Energy 2003;
showed that combustion duration was reduced by a significant 41.7 28(5):569–77.
% when utilizing a fuel mixture with a 60 % methane energy [9] Uddeen K, Tang Q, Shi H, Magnotti G, Turner J. Multiple spark ignition approach
fraction. to burn ammonia in a spark-ignition engine: An optical study. SAE Technical Paper;
2023-01-0258.
3. Ammonia-ethanol blends produced reduced combustion duration, [10] Zhang R, Chen L, Wei H, Li J, Chen R, Pan J. Understanding the difference in
lower combustion instability, and higher engine efficiency compared combustion and flame propagation characteristics between ammonia and methane
to ammonia-methane blends due to the higher burning velocity of using an optical SI engine. Fuel 2022;324:124794.
[11] Mørch CS, Bjerre A, Gøttrup MP, Sorenson SC, Schramm J. Ammonia/hydrogen
ethanol. At 50 % energy fraction, the ethanol-blending case pro­ mixtures in an SI-engine: Engine performance and analysis of a proposed fuel
duced 17.1 % relatively higher engine efficiency than the methane- system. Fuel 2011;90(2):854–64.
blending case. [12] Lhuillier C, Brequigny P, Contino F, Mounaïm-Rousselle C. Experimental study on
ammonia/hydrogen/air combustion in spark ignition engine conditions. Fuel 2020;
4. Ammonia-ethanol blends produced higher NOx emissions than the 269:117448.
ammonia-methane blends because the faster flame speed produced [13] Grannell SM, Assanis DN, Gillespie DE, Bohac SV. Exhaust emissions from a
by the ethanol addition generated higher in-cylinder temperature, stoichiometric, ammonia and gasoline dual fueled spark ignition engine. Internal
combustion engine division spring technical conference 2009;43406:135–41.
which instigated higher thermal NOx. In addition, at 50 % energy
[14] Uddeen K, Tang Q, Shi H, Magnotti G, Turner J. A novel multiple spark ignition
fraction, ethanol blends produced 25 % higher CO2 emissions. strategy to achieve pure ammonia combustion in an optical spark-ignition engine.
However, ammonia-methane blends exhibited larger UHC emissions Fuel 2023;349:128741.
due to the low combustion efficiency. [15] Tang Q, Shi H, Uddeen K, Sharma P, Yao M, Turner JW, et al. Study of engine
knocking combustion using simultaneous high-speed shadowgraph and natural
5. Flame propagation characteristics from high-speed NFL imaging flame luminosity imaging. Appl Therm Eng 2023;235:121440.
presented a slow flame propagation speed of less than 7.5 m/s with a [16] Dinesh M, Pandey JK, Kumar G. Study of performance, combustion, and NOx
long combustion duration for pure ammonia combustion. Blending emission behavior of an SI engine fuelled with ammonia/hydrogen blends at
various compression ratio. Int J Hydrogen Energy 2022;47(60):25391–403.
ethanol and methane with ammonia increased the flame propagation [17] Uddeen K, Almatrafi F, Shi H, Tang Q, Parnell J, Peckham M, et al. Investigation
speed significantly. At 50 % energy fraction, the ammonia-ethanol into Various Strategies to Achieve Stable Ammonia Combustion in a Spark-Ignition
blend showed a 22 % higher flame speed than the ammonia- Engine. SAE Technical Paper; 2023-24-0040.
[18] Lhuillier C, Brequigny P, Contino F, Rousselle C. Combustion characteristics of
methane blend. ammonia in a modern spark-ignition engine. SAE Technical Paper; 2019-24-0237.
[19] Almatrafi F, Uddeen K, Kenkoh K, Aljabri H, Parnell J, Peckham M, et al.
Experimental Study of Fuel Mixture Limitations of Ammonia and Gasoline in a
CRediT authorship contribution statement
Passive Pre-Chamber Engine. SAE Technical Paper; 2023-32-0106.
[20] Almatrafi F, Uddeen K, Ben Houidi M, Cenker E, Turner J. Direct injection strategy
Kalim Uddeen: Data curation, Investigation, Methodology, Writing to extend the lean limit of a passive pre-chamber. Internal Combustion Engine
– original draft, Formal analysis, Validation. Qinglong Tang: Data Division Fall Technical Conference. 86540. American Society of Mechanical
Engineers; 2022:V001T03A4.
curation, Investigation, Methodology, Supervision, Validation, Writing – [21] Duynslaegher C, Jeanmart H, Vandooren J. Ammonia combustion at elevated
original draft, Writing – review & editing, Formal analysis. Hao Shi: pressure and temperature conditions. Fuel 2010;89(11):3540–5.
Data curation, Methodology. James Turner: Funding acquisition, Su­ [22] Silva M, Almatrafi F, Uddeen K, Cenker E, Sim J, Younes M, et al. Computational
Assessment of Ammonia as a Fuel for Light-Duty SI Engines. SAE Technical Paper;
pervision, Project administration. 2023-24-0013.
[23] Grannell SM, Assanis DN, Bohac SV, Gillespie DE. The fuel mix limits and efficiency
of a stoichiometric, ammonia, and gasoline dual fueled spark ignition engine.
Declaration of Competing Interest Journal of Engineering for Gas Turbine and Power 2008;130(4):042802.
[24] Ryu K, Zacharakis-Jutz GE, Kong S-C. Effects of gaseous ammonia direct injection
on performance characteristics of a spark-ignition engine. Appl Energy 2014;116:
The authors declare that they have no known competing financial 206–15.
interests or personal relationships that could have appeared to influence [25] Lu M, Dong D, Wei F, Long W, Wang Y, Cong L, et al. Chemical mechanism of
ammonia-methanol combustion and chemical reaction kinetics analysis for
the work reported in this paper. different methanol blends. Fuel 2023;341:127697.
[26] Li J, Zhang R, Pan J, Wei H, Shu G, Chen L. Ammonia and hydrogen blending
Data availability effects on combustion stabilities in optical SI engines. Energ Conver Manage 2023;
280:116827.
[27] Kurien C, Varma PS, Mittal M. Effect of ammonia energy fractions on combustion
Data will be made available on request. stability and engine characteristics of gaseous (ammonia/methane) fuelled spark
ignition engine. Int J Hydrogen Energy 2023;48(4):1391–400.
[28] Oh S, Park C, Kim S, Kim Y, Choi Y, Kim C. Natural gas–ammonia dual-fuel
Acknowledgments combustion in spark-ignited engine with various air–fuel ratios and split ratios of
ammonia under part load condition. Fuel 2021;290:120095.
This research work is supported by the Clean Combustion Research [29] Pelé R, Brequigny P, Bellettre J, Mounaïm-Rousselle C. Performances and pollutant
emissions of spark ignition engine using direct injection for blends of ethanol/
Center (CCRC) at the King Abdullah University of Science and Tech­ ammonia and pure ammonia. Int J Engine Res 2022;14680874231170661.
nology (KAUST) and the National Natural Science Foundation of China [30] Catapano F, Di Iorio S, Magno A, Sementa P, Vaglieco B. A comprehensive analysis
(NSFC) through its project 52206166. of the effect of ethanol, methane and methane-hydrogen blend on the combustion
process in a PFI (port fuel injection) engine. Energy 2015;88:101–10.
[31] Shi H, Uddeen K, An Y, Johansson B. Experimental study on knock mechanism with
References multiple spark plugs and multiple pressure sensors. SAE Technical Paper; 2020-01-
2055.
[32] Uddeen K, Shi H, Tang Q, Magnotti G, Turner J. The effects of compression ratio
[1] Kalghatgi G. Is it the end of combustion and engine combustion research? Should it
and combustion initiation location on knock emergence by using multiple pressure
be? Transport Eng 2022::100142.
sensing devices. Int J Engine Res 2023;24(3):1121–39.
[2] Uddeen K, Shi H, Tang Q, Magnotti G, Turner JW. Using multiple ignition sites and
[33] Uddeen K, Shi H, Tang Q, Turner JW. Investigations into the effects of spark plug
pressure sensing devices to determine the effect of air-fuel equivalence ratio on the
location on knock initiation by using multiple pressure transducers. SAE Technical
morphology of knocking combustion. SAE Technical Paper; 2022-01-0433.
Paper; 2021-01-1159.
[3] Li J, Lai S, Chen D, Wu R, Kobayashi N, Deng L, et al. A review on combustion
characteristics of ammonia as a carbon-free fuel. Front Energy Res 2021;9:760356.
[4] Shi H, Tang Q, Uddeen K, Johansson B, Turner J, Magnotti G. Effects of multiple
spark ignition on engine knock under different compression ratio and fuel octane
number conditions. Fuel 2022;310:122471.

15
K. Uddeen et al. Fuel 358 (2024) 130296

[34] Zhang J, Shi H, Luong MB, Tang Q, Uddeen K, Magnotti G, et al. Prediction of [36] Uddeen K, Shi H, Tang Q, Magnotti G, Turner J. Optical study of knocking
knock intensity and validation in an optical SI engine. Combust Flame 2023;254: phenomenon in a spark-ignition engine by using high-speed OH*
112854. chemiluminescence imaging: A multiple ignition sites approach. Int J Engine Res
[35] Shi H, Tang Q, Uddeen K, Magnotti G, Turner J. Optical diagnostics and multi- 2023;14680874231166268.
point pressure sensing on the knocking combustion with multiple spark ignition.
Combust Flame 2022;236:111802.

16