Journal of the Energy Institute 102 (2022) 362–373

Contents lists available at ScienceDirect

Journal of the Energy Institute

journal homepage: www.elsevier.com/locate/joei

A comparison between low- and high-pressure injection dual-fuel modes of

diesel-pilot-ignition ammonia combustion engines
Tie Li a, b, Xinyi Zhou a, b, *, Ning Wang b, Xinran Wang b, Run Chen a, b, Shiyan Li a, b, Ping Yi a, b
a
State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, PR China
b
Institute of Power Plants and Automation, Shanghai Jiao Tong University, PR China

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

Keywords: Ammonia, as a carbon-free fuel, is drawing more and more attention due to its potential of decarbonization for
Ammonia marine engines. The dual-fuel mode that can compensate for the poor combustion properties of ammonia has
Carbon-free exhibited some prospects for practical application, but more relevant researches are urgent to promote the
Compression ignition engine
carbon-free fuel, especially for the high-pressure injection dual-fuel (HPDF) mode. In this paper, the ammonia
Dual-fuel
Emissions
spray experiments under the diesel-like conditions are conducted using the high-speed photography to calibrate
the spray sub-models. Then, the combustion and emission characteristics of the low- and high-pressure injection
dual-fuel modes of diesel pilot-ignition ammonia combustion engines are numerically studied and compared.
Particularly, the equivalent CO2 considering the global warming potential of CO2 and N2O are evaluated (i.e.
CO2 + 300 × N2O). The results show that the maximum ammonia ratio about 80% by energy is recommended for
the low-pressure injection dual-fuel (LPDF) mode and further increase the diesel replacement ratio will increase
the probability of misfire, while the HPDF mode has the potential to achieve a 97% diesel replacement ratio.
Compared with the pure diesel mode, the HPDF mode has similar indicated thermal efficiency, cooling and
exhaust loss, while it can significantly reduce the equivalent CO2 with little increase in the NH3 emissions. Owing
to the reduced heat transfer loss, the LPDF mode has the potential to achieve a higher indicated thermal effi­
ciency than the pure-diesel mode, but its ability to reduce the equivalent CO2 would be relatively limited because
the significant increase of N2O emissions if exhaust after treatments were not employed. The comparison be­
tween the LPDF mode and HPDF mode shows that the former case has the potential to achieve a higher indicated
thermal efficiency, while the latter case can significantly reduce the engine-out NH3, NOx and greenhouse gas
emissions.

1. Introduction international ships by at least 50% by 2050 compared to 2008, and (3)
phasing out GHG emissions from international ships as soon as possible
The compression ignition (CI) diesel engines dominate the markets in this century. Since diesel engines still account for approximately 99%
as prime movers for heavy-duty vehicles such as ships and trucks owing of international ship prime power in 2020 [3], one effective way to
to their excellent reliability and thermal efficiency, but the use of diesel achieve the above goals is to substitute diesel fuel through carbon-free
fuels inevitably produce the greenhouse gas (GHG) emissions [1,2]. fuel.
Recently, the reduction of GHG emissions such as carbon dioxide (CO2) Ammonia (NH3) has been deemed to be a promising carbon-free fuel
has been widely anticipated and discussed by all walks of life to prevent for CI engines [4–6]. It can be easily liquefied under pressure of 9 bar at
global warming since the Paris Agreement came into effect in 2016. room temperature, and the volumetric energy density of liquefied
Taking the international shipping industry as an example, the Interna­ ammonia is around 1.6 times higher than that of liquefied hydrogen,
tional Maritime Organization (IMO) adopted the ’’initial strategy on which makes ammonia often considered as hydrogen carrier. Because of
reduction of GHG emissions from ships’’ in April 2018, and aimed at (1) the well-established infrastructures for ammonia production, trans­
reducing carbon intensity of international ships by at least 40% by 2030 portation and storage, the direct use of ammonia as fuel becomes
compared to 2008, (2) reducing the total annual GHG emissions from attractive. Moreover, the possibility of producing ammonia using green

* Corresponding author. 800 Dongchuan Rd., Shanghai, 200240, PR China.

E-mail address: [email protected] (X. Zhou).

https://doi.org/10.1016/j.joei.2022.04.009
Received 11 March 2022; Received in revised form 16 April 2022; Accepted 25 April 2022
Available online 29 April 2022
1743-9671/© 2022 Energy Institute. Published by Elsevier Ltd. All rights reserved.
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

hydrogen from renewable energy sources supports ammonia as a sus­ main fuel, the valve overlap and quenching effects will result in methane
tainable fuel (i.e. green ammonia). to slip into the exhaust gas [19]. Recently, the high-pressure injection
The use of ammonia as engine fuels can be traced back to the dual-fuel (HPDF) engine have been deemed to have the potential to
beginning of the 20th century [7], and the subsequent researches can be substantially reduce the unburned hydrocarbons emissions and increase
divided into three main periods. The first period is around the Second the thermal efficiency [20–22]. For the HPDF engine, the high-reactivity
World War, the ammonia-fueled engines ushered in a rapid development pilot diesel fuel is injected into the combustion chamber at the end of the
owing to the shortage of fossil fuel [6]. The most representative is the compression stroke, and serves as an ignition source for the following
operation of ammonia-fueled buses for public transportations in injected low-reactivity main fuel such as methane. At this mode, the
Belgium in 1943 [8]. After the Second World War, the oil supplies were in-cylinder combustion processes are controlled by the fuel-air mixing,
restored and the interest in ammonia-fueled engines gradually which is similar to the conventional diesel engine. Since there is no risk
decreased. The second period is from the 1960s to the 1970s, new of knocking, the compression ratio of HPDF engine can be high, which
concepts of ammonia-fueled engines were proposed to face the energy helps to improve the thermal efficiency. Moreover, the combustion
crisis [4]. However, the cost of ammonia fuel could be up to ten times processes are complete and stable for the diffusive flame, which reduces
higher than that of hydrocarbon fuel for the same amount of energy in the slip of unburned fuel and widens the operation condition.
that era [9], which limited the practical application of ammonia-fueled Because of the toxicity of NH3, the potential of reducing the un­
engines. Over the past years, the ammonia-fueled engines have received burned NH3 emissions in HPDF engines is attractive. Unfortunately, the
widespread attention again owing to the urgent requirements of global use of ammonia in HPDF engine has almost not been reported so far,
economy decarbonization, but the relevant studies are premature, except the exploratory numerical simulations by Frankl et al. [23,24] in
especially for the CI engines. Technical University of Munich. In their study, the pilot diesel and liquid
The pure ammonia mode in CI engine is challenging because that the ammonia are injected into cylinder at − 2.5 ◦ CA and 1 ◦ CA after top dead
ammonia features the poor combustion characteristics such as narrow center (aTDC), respectively. They reported that the required amount of
flammability, high auto-ignition temperature and low flame speed [6]. pilot fuel in the HPDF engine is smaller than 5% of the total injected
For example, Pearsall et al. [9] rebuilt a AVDS-1790 Vee-Twin engine energy [24]. As a result, the HPDF engine seems to have the potential of
using a 30:1 compression ratio piston. They reported that no combustion achieving a higher ammonia fuel ratio compared with the LPDF engine,
at all was observed when the intake temperature was room temperature, and the direct injection of liquid ammonia can be employed for the
and the ignition can only occur in the first cycle when the intake tem­ HPDF engine to ensure the sufficient injection mass. Unfortunately,
perature was increased to 360 K. Gray et al. [10] reported that the pure Frankl et al. [23,24] mainly focus on the combustion characteristic such
ammonia mode in CI engine can only be achieved with the compression as heat release rate, in-cylinder pressure and temperature, important
ratio of 35:1 and intake gas and coolant temperature of 422 K. Recently, information such as thermal efficiency and N2O emissions are not
researchers have paid some attention to the low-pressure injection provided.
dual-fuel (LPDF) mode. In this mode, the ammonia and air are mixed Three-dimensional computational fluid dynamics (3-D CFD) simu­
before the compression stroke, then the diesel fuel is usually injected lation has been recognized as a reliable method for investigating the
into the combustion chamber at the late stage of compression stroke to spray combustion processes and emission formation of diesel engines
start the combustion processes. Reiter et al. [11] studied the LPDF mode [25], and the accuracy of 3-D CFD simulation for HPDF engines using
using a diesel engine with 106 mm bore diameter and 17:1 compression methane [19] and hydrogen [23] as main fuel has also been experi­
ratio. Ammonia was introduced into the engine air intake system mentally verified recently. The successful simulation of HPDF engine is
through a stainless steel pipeline, while the diesel fuel injection system inseparable from the accurate spray modeling and reliable chemical
was unchanged. They demonstrated that the brake thermal efficiency is reaction mechanism. Since the last century, the chemical reaction
around 33–38% when ammonia provides 40–80% of the total energy, mechanism for ammonia combustion has been continuously updated
and the brake thermal efficiency is around 19% when ammonia provides and revised based on the experimental laminar burning velocity [26].
95% of the total energy. With the same engine, Reiter et al. [12] re­ However, the researches on ammonia/diesel HPDF engine has just
ported that the exhaust hydrocarbon and carbon monoxide of the started in the past two years, to the best of the present authors’
dual-fuel mode is higher than those of the pure diesel mode for the knowledge, there is no available experimental data for high-pressure
constant power output, and the NOx emissions will increase significantly injection liquid ammonia spray under the thermodynamic conditions
if the ammonia accounts for the majority of the heat release. Niki et al. near the top dead center, which limits the calibration of spray
[13] conducted an experimental study on the combustion characteristics sub-models and therefore affects the accurate simulations of ammo­
of the ammonia/diesel LPDF engine with 112 mm bore diameter and nia/diesel HPDF engine.
18.5:1 compression ratio. In their experiments, ammonia gas was As reviewed in the above literatures, although the ammonia/diesel
injected into the intake pipe of engine by a gas injector. They exhibited LPDF engine has been experimentally and numerically studied to some
that the ignition delay increases when the heat release contribution of extent, the researches on ammonia/diesel HPDF mode have rarely been
ammonia increases from 0% to around 18%. In their subsequent reported, and the comparison between these two modes is still lacking.
experimental work, Niki et al. [14] reported that the unburned NH3 and Moreover, the experiment data for high-pressure injection liquid
N2O emissions increase with the ammonia supply increasing, and they ammonia spray under the thermodynamic conditions near the top dead
suggested advancing the diesel fuel injection timing to reduce the NH3 center is urgently needed to calibrate the spray models. It is well known
and N2O emissions [15]. The recent experimental researches on that nitrous oxide (N2O) has a high global warming potential around
ammonia/diesel LPDF engine have been listed above, and the numerical 300 times of CO2 for a 100-year timescale. As a result, the total green­
studies are even rarer. Tay et al. [16,17] conducted a numerical study on house gas including CO2 and N2O must be considered to evaluate the
the ammonia/diesel LPDF engine with bore diameter of 92 mm and potential of ammonia/diesel dual-fuel engine to mitigate the greenhouse
compression ratio of 18.5:1. They found that the NO emissions increase gas effect. Unfortunately, the related research has not been reported.
substantially when ammonia provides more than 60% of the total en­ Therefore, this study aims at bridging the gap between the insuffi­
ergy. Another numerical study on ammonia combustion in the LPDF cient relevant knowledge and the urgency of promoting optimal engine
mode was reported by Lasocki et al. [18]. They demonstrated that the designs and widespread application of the zero-carbon fuel. The high-
biggest difference in NO emissions between the LPDF mode and pure speed schlieren and diffused back-illumination methods are used to
diesel mode occurs at the high engine load for a engine with 105 mm obtain the spray vapor- and liquid-phase penetration evolutions of liquid
bore diameter. ammonia spray under diesel-like conditions. After calibrating the spray
According to the experience of LPDF engine using methane as the sub-models against the experimental ammonia spray data, the

363
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

Table 1 from the fuel tank with a pressure stabilized at 2 MPa. Then, the liquid
Comparison of main fuel properties between liquid ammonia and diesel [6,27]. ammonia was compressed by the piston moving in the accumulator and
Properties Ammonia Diesel supplied to the injector through the high-pressure fuel pipe. The tem­
perature of liquid ammonia was controlled by an oil heating system
Boiling point (K) 239.8 455.0–633.0
Autoignition temperature (K) 924 527 through a double-walled tube. The fuel injection pressure was 60 MPa,
Octane number 130 8–15 and the fuel temperature was maintained at 350 K in the experiments.
Lower heating value (MJ/kg) 18.8 42.5 Since the critical temperature and critical pressure of ammonia are
Latent heat of vaporization (kJ/kg) 1371 375 405.4 K and 11.3 MPa [23], the phase state of the injected ammonia
were liquid. The reason for determining the above injector diameter and
numerical models for both the ammonia/diesel LPDF engine and HPDF injection pressure was to ensure the similar energy input rate as the pure
engine are established. Then, the combustion and emission character­ diesel mode, and the single-hole injector was employed to simplify the
istics of these two modes are studied and compared. Especially, the calibration of numerical model. Moreover, the main fuel properties
equivalent CO2 considering the global warming potential of CO2 and between liquid ammonia and diesel are summarized in Table 1.
N2O are calculated. Finally, the major conclusions are reached. A constant volume combustion chamber (CVCC) with 100 mm-in-
diameter optical windows were employed to measure the spray char­
2. Experimental and numerical methods acteristics of liquid ammonia. The high-pressure and high-temperature
ambient conditions in the CVCC at the fuel injection timing was ob­
2.1. Experiments for spray model calibration tained by igniting a combustible mixture consisting of ethylene (C2H4),
nitrogen (N2) and oxygen (O2) with two spark plugs, and the transient
In the present study, the direct injection of liquid ammonia was pressure in the CVCC was obtained by a pressure transducer (Kistler
employed for the HPDF engine. For the HPDF mode, the heat release rate 6054B). The CVCC can bear the ambient pressure of 35 MPa and the
will be controlled by the fuel-air mixing rate and the liquid ammonia has ambient temperature up to 1300 K, more details about the chamber
the potential to provide more than 95% of the total injected energy [24]. design can refer our previous studies [28].
As a result, the accurate prediction of the spray mixture formation Fig. 1a shows the high-speed schlieren method for measuring the
processes of liquid ammonia is critical for the successful engine simu­ spray vapor-phase penetration. The parallel light was obtained by
lation. However, there is no available experimental data for coupling an in-house light emitting diode (LED) with the power of
high-pressure injection liquid ammonia spray under the thermodynamic 120W, a pinhole with 2 mm diameter and a concave mirror with 500-
conditions near the top dead center, while liquid ammonia has the rather mm focal length. Then, the parallel light passed through the ammonia
different properties such as significantly lower boiling point and higher spray in the CVCC and was recorded by a CMOS camera (NAC HX-6) and
latent heat than liquid hydrocarbon fuels, causing calibration of spray a 50-mm Nikon f/1.4 lens coupled with a 500-mm focal length concave
sub-models very critical for engine combustion simulation. mirror and a knife-edge at the focus. Details can about the optical setups
In order to provide the experimental spray data for model calibra­ can be found in Ref. [29]. Fig. 1b shows the diffused back-illumination
tion, an in-house pressure accumulator capable of the maximum fuel (DBI) method for obtaining the spray liquid-phase penetration [30,31].
injection pressure of 300 MPa and a single-hole diesel injector with the The same LED was adopted as the light source. After passing through a
hole diameter of 0.22 mm were used for ammonia injection. For the fuel 5000 mesh diffuser and the spray region, the light was captured by a
injection system, liquid ammonia was introduced into the accumulator CMOS camera (NAC HX-6) and a 50-mm Nikon f/1.4 lens. The imaging

Fig. 1. Optical arrangements for (a) schlieren method and (b) diffused back-illumination imaging.

364
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

Table 2 measured by the diffused back-illumination imaging. It can be seen that

Experimental conditions for liquid ammonia spray. the LES can better solve to some degree the details of the turbulent
Parameters Value fluctuations at the spray boundaries, while both the LES and RANS can
well predict the overall spray morphology. Since the calculation cost is
Nozzle hole diameter (mm) 0.22
Fuel injection pressure (MPa) 60 too large when coupling the LES and chemical kinetic mechanism, the
Fuel temperature (K) 350 RANS will be used for the following engine simulations.
Injection duration (ms) 2.5 Fig. 3 shows the comparison of liquid ammonia spray penetrations
Ambient temperature (K) 900 between the experiments and simulations. The RANS is used to solve the
Ambient density (kg/m3) 18
Ambient oxygen concentration (%) 0
turbulence flow. It can be seen that the vapor- and liquid-phase pene­
trations of liquid ammonia spray are almost overlapped at the injection
startup. Then, the liquid-phase ammonia stops penetrating and the
speed was 15,000 fps, while the exposure time was 2 μs. Table 2 sum­ liquid length is almost a constant value, while the vapor-phase ammonia
marizes the experimental conditions for the liquid ammonia spray. still penetrates downstream. The maximum liquid penetration is the
position where the total fuel evaporation rate in the spray equals to the
2.2. Numerical models fuel injection rate [40]. It should be noted that the predicted liquid
ammonia spray penetrations are shorter than the experimental values at
2.2.1. Liquid ammonia spray the initial stage of injection, which is consistent with the previous diesel
The CONVERGE code package was used in the present study. Vali­ spray simulations. This phenomenon can be attributed to the error of the
dation of inert spray is very critical for HPDF ammonia spray combus­ fuel injection rate test at the initial stage of injection, and the methods to
tion, because the heat release rate will be controlled by the fuel-air revise this under-predicted spray penetration at the injection startup can
mixing rate. As a result, before the numerical study of the HPDF engine, be found in Ref. [38]. Generally, the present model settings can well
the sub-models for liquid ammonia spray were first calibrated against predict the spray vapor- and liquid-phase penetrations, and the calcu­
the experimental spray data under the inert condition in this section. lated results are almost independent of mesh size when the minimum
The computational domain was simplified to a 48 × 48 × 97 mm cuboid. mesh size is less than 0.45 mm. As a result, the minimum mesh size of
The turbulence flow was resolved using the large eddy simulation (LES) 0.45 mm was used in the following engine simulations to balance the
[32] or the Reynolds-Averaged-Navier-Stokes (RANS) equations with
the RNG k-ε turbulence model [33]. The liquid fuel injection was
modelled using the Lagrangian particles, with the Kevin
Helmholtz-Rayleigh Taylor (KH-RT) model [34] for spray break-up, NTC
model [35] for drop collision and Frossling model [36] for drop evap­
oration. Moreover, the Amsden model [37] was used for the wall heat
transfer. In order to study the effects of mesh size on the prediction
accuracy, the base grid for the overall computational domain was set to
change from 1.0 mm to 1.8 mm. The 2-level adaptive mesh refinement
(AMR) was used based on the temperature and velocity gradients, and
the 2-level fixed mesh refinement was employed in the spray developing
region. As a result, the minimum mesh size was ranged from 0.25 mm to
0.45 mm. The fuel injection rate profile was measured using an in-house
test rig based on the Bosch long-tube mothed [28], diesel fuel is used as
the test fuel but the fuel density was corrected during the data pro­
cessing. More details about the numerical models and methods to cali­
brate the spray sub-models can be found in our previous studies [38,39].
Fig. 2 shows the comparison of spray morphologies between exper­
iment and simulations with (a) LES and (b) RANS. The solid line in­
Fig. 3. The comparison of liquid ammonia spray penetrations between the
dicates the spray vapor-phase boundary measured by the schlieren
experiments and simulations with different minimum mesh sizes.
method, while the dashed line refers to the spray liquid-phase boundary

Fig. 2. Comparisons of spray morphologies between the experiment and simulations with (a) LES and (b) RANS.

365
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

Table 3 Table 4
Engine specifications. Operation conditions for the ammonia/diesel HPDF engine.
Parameters Value Parameters Value

Bore (mm) 95 Indicated mean effective pressure (MPa) 1.2

Stroke (mm) 102 Diesel ratio by energy (%) 1, 3, 10
Compression ratio 17.5 Diesel start of injection (◦ CA aTDC) − 8
Diesel nozzle hole diameter (mm) 0.127 Diesel injection duration (◦ CA) 1.8
Ammonia nozzle hole diameter (mm) 0.22 Ammonia start of injection (◦ CA aTDC) − 5
Diesel nozzle hole number 8 Ammonia injection duration (◦ CA) 15
Ammonia nozzle hole number 8 Excess air ratio 1.7
Intake pressure (bar) 1.3
Intake temperature (K) 322
Engine speed (rpm) 1000

Fig. 4. Schematic diagram of the ammonia/diesel dual-fuel injection: (a) the

angle between the ammonia spray and diesel spray is 22.5◦ (i.e. HPDF 22.5◦ );
(b) the angle between the ammonia spray and diesel spray is 0◦ (i.e. HPDF 0◦ ).

calculation accuracy and cost.

Fig. 5. The comparison of ignition delay between the experiments and simu­
2.2.2. High-pressure injection dual-fuel (HPDF) engine lations. The ignition delay is defined as the duration from the start of pilot
The ammonia/diesel HPDF engine used in this study was derived diesel fuel injection to the start of heat release, and the experimental data is
from a diesel engine with a bore diameter of 95 mm, and the engine from Ref. [15]. For the experiments and simulations, the bore diameter is 112
specifications are shown in Table 3. Fig. 4 shows the schematic diagram mm, stroke is 110 mm, compression ratio is 18.5, engine speed is 1500 rpm,
pilot diesel injection mass is 24 mg, diesel ratio by energy is 55%, and the
of the ammonia/diesel dual-fuel injection from the center of the cylinder
naturally aspirated is employed.
head. The injector consists of eight holes for diesel injection and eight
holes for liquid ammonia injection. The angle between the diesel spray
and ammonia spray was set to be 0◦ and 22.5◦ , as shown in Fig. 4. It including the compression, fuel injection, combustion and expansion
should be noted that for the HPDF 0◦ , the ammonia and diesel fuel are processes. It should be noted that the accurate prediction of the pilot
injected from different nozzle holes, as shown in the upper right corner diesel ignition under the ammonia atmosphere is very important for the
of Fig. 4b. In order to take into account the pure diesel mode, the diesel LPDF mode, and this will be validated against the experimental engine
nozzle hole diameter of the dual-fuel injector was set to be constant with data in Section 2.2.3.
that of the base diesel engine (i.e. 0.127 mm). Consistent with the spray The operational conditions for the ammonia/diesel HPDF engine are
experiment in Section 2.1, the ammonia nozzle hole diameter was 0.22 shown in Table 4. The diesel ratio by energy was varied from 1% to 10%
mm. to study the minimum pilot fuel required, while the total fuel energy and
For the HPDF mode, accurate predictions of the spray mixture for­ the overall excess air ratio were kept constant. Here, the diesel ratio by
mation processes are very critical, because the heat release rate will be energy is defined as the lower calorific value of the input diesel fuel
controlled by the fuel-air mixing rate. As a result, the turbulence and divided by that of the total input fuel. Except the operation conditions in
spray models for the HPDF engine simulation were constant with those Table 4, the pure diesel mode with the same start of injection, total fuel
for the well calibrated liquid ammonia spray simulation in Section 2.2.1. energy and overall excess air ratio was also studied for comparison. In
The combustion processes were calculated using the detailed transient order to achieve a relatively fair comparison, the engine speed, intake
chemistry solver SAGE, and the chemical kinetic mechanism was a pressure and intake temperature were set to be constant for all the
combination of GRI3.0 [41] and the n-heptane mechanisms, which operating conditions.
consisted of 464 reactions and 76 species. The reason for this combi­
nation is that the GRI3.0 was found to give the best predictions for 2.2.3. Low-pressure injection dual-fuel (LPDF) engine
ammonia laminar burning velocity in the previous studies of the present For the LPDF mode in this study, the port fuel injection processes
authors [26], and the n-heptane mechanism has been widely accepted were not considered and the ammonia and air were set to be fully mixed
for prediction of diesel combustion [23]. It should be noted that the in the cylinder at the intake valve closing timing, other numerical
ammonia laminar burning velocity was measured under the ambient models and chemical kinetic mechanism of the ammonia/diesel LPDF
pressure ranging from 1 bar to 5 bar in Ref. [26], and the experiments engine was similar with those of the HPDF engine. The accurate pre­
and validation of the existing chemical kinetic mechanisms at higher diction of the pilot diesel ignition under the ammonia atmosphere is
ambient pressure are excepted in the near future. The simulations were very important for the LPDF engine simulation, and this is validated
conducted from the intake valve closing to exhaust valve opening, against the experimental engine data by the National Maritime Research
Institute [15], as shown in Fig. 5. It can be seen that the calculated

366
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

Table 5 delay of the pilot diesel fuel is same as that of the pure diesel mode (i.e.
Operation conditions for the ammonia/diesel LPDF engine. around 3◦ CA), as the increase of heat release rate (HRR) occurs at
Parameters Value around − 5◦ CA aTDC. The ammonia spray penetrates into the pilot diesel
combustion products at − 3◦ CA aTDC, but the heat release rate is smaller
Diesel ratio by energy (%) 10, 20, 30
Diesel injection duration (◦ CA) 2, 4, 6 than 30 J/deg until around − 1◦ CA aTDC. In addition to the reduced
Excess air ratio before diesel injection 1.60 local oxygen caused by pilot diesel combustion, the longer ignition delay
Excess air ratio after diesel injection 1.43, 1.26, 1.10 of ammonia of about 4◦ CA may be attributed to its low chemical reac­
Diesel start of injection (◦ CA aTDC) − 8 tivity as well as high vaporization energy. The HRR of the HPDF mode
Intake pressure (bar) 1.3
Intake temperature (K) 322
shows a two-peak shape: a small peak by pilot diesel combustion fol­
Engine speed (rpm) 1000 lowed by a sharp peak by ammonia combustion. Since the interaction
between the ammonia spray and the diesel combustion products is
enhanced, the combustion phasing is slightly advanced when the angle
ignition delay are in good agreement with the experimental data under
different start of pilot diesel injection timing, further indicating the
effectiveness of the selected mechanism. It should be noted that the
discrepancy between the simulation and experimental results decreases
with the diesel fuel injection timing retarding. This might because that
the well verified n-heptane mechanism plays a more important role at
the late fuel injection timing case. Table 5 shows the operation condi­
tions for the ammonia/diesel LPDF engine. It should be noted that the
start of diesel injection, engine speed, intake pressure and intake tem­
perature here were set to be constant with the HPDF mode for the
effective comparison.

3. Results and discussion

3.1. High-pressure injection dual-fuel (HPDF) engine

3.1.1. Effects of dual-fuel injector hole arrangement

Fig. 6 shows the effects of dual-fuel injector hole arrangement on the
spray flame evolutions. The information of pure diesel mode is added for
comparison. The diesel injection timing is kept at − 8◦ CA aTDC, total
input energy of about 1950 J, excess air ratio of about 1.7, engine speed
of 1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K.
The ammonia ratio by energy is 97% for the HPDF modes. It can be seen
that the diesel fuel is injected into the cylinder at − 8◦ CA aTDC for all the
pure diesel and HPDF modes, and the ammonia is injected into the
cylinder at − 5◦ CA aTDC for the HPDF mode. For the pure diesel mode,
the duration from the start of injection to the start of heat release (i.e. the
Fig. 7. Effects of dual-fuel injector hole arrangement on the heat release rate,
ignition delay) is around 3◦ CA, as shown at the bottom of Fig. 7. The in-cylinder pressure and temperature. The diesel injection timing is kept at
heat release rate (HRR) profile of the pure diesel mode shows a sharp − 8◦ CA aTDC, total input energy of about 1950 J, excess air ratio of about 1.7,
peak by premixed combustion followed by the diffusion combustion that engine speed of 1000 rpm, intake pressure of 1.3 bar and intake temperature of
is controlled by the fuel-air mixing. For the HPDF modes, the ignition 322 K. The ammonia ratio by energy is 97% for the HPDF modes.

Fig. 6. Effects of dual-fuel injector hole arrangement on the spray flame evolutions. The diesel injection timing is kept at − 8◦ CA aTDC, total input energy of about
1950 J, excess air ratio of about 1.7, engine speed of 1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K. The ammonia ratio by energy is 97% for
the HPDF modes. The temperature information is superimposed on the isosurfaces with excess air ratio of 1.3.

367
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

Fig. 8. Effects of dual-fuel injector hole arrangement on the energy balances.

The diesel injection timing is kept at − 8◦ CA aTDC, total input energy of about
1950 J, excess air ratio of about 1.7, engine speed of 1000 rpm, intake pressure
of 1.3 bar and intake temperature of 322 K. The ammonia ratio by energy is
97% for the HPDF modes.

Fig. 10. Effects of dual-fuel injector hole arrangement on the temporal evo­
lutions of NOx, NH3 and NH. The diesel injection timing is kept at − 8◦ CA aTDC,
total input energy of about 1950 J, excess air ratio of about 1.7, engine speed of
1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K. The
ammonia ratio by energy is 97% for the HPDF modes.

interactions near the top dead center of the pure diesel mode are
stronger than those of the HPDF mode, as shown in Fig. 6, the heat
transfer loss of the pure diesel mode is slightly higher. Generally, the
indicated thermal efficiency and exhaust loss are similar among the pure
diesel and HPDF modes.
Fig. 9 shows the effects of dual-fuel injector hole arrangement on the
emission characteristics. Since N2O is 300 times as potent as CO2 in GHG
emissions for a 100-year time scale [42], the equivalent CO2 considering
the global warming potential of CO2 and N2O are calculated (i.e. CO2 +
300 × N2O). For the HPDF modes, since 97% diesel is replaced by
ammonia, the CO2 emissions are significantly reduced while the NH3
emissions are obviously increased. Although the HPDF modes produce
more N2O, their equivalent CO2 is still significantly lower than the pure
diesel mode, indicating the potential of HPDF engine to mitigate the
greenhouse gas effect. Fig. 10 shows the effects of dual-fuel injector hole
arrangement on the temporal evolutions of NOx, NH3 and NH. It can be
seen that the HPDF modes produce more NOx emissions from − 8 to 5
◦
CA aTDC, although their combustion temperature is lower than the
pure diesel mode during this stage. This is because that the nitrogenous
Fig. 9. Effects of dual-fuel injector hole arrangement on the emission charac­ radicals and oxygenated species react to form fuel-borne NOx under the
teristics. The diesel injection timing is kept at − 8◦ CA aTDC, total input energy HPDF modes, as the NH concentration of the HPDF mode is obviously
of about 1950 J, excess air ratio of about 1.7, engine speed of 1000 rpm, intake
higher than that of the pure diesel mode. After around 5 ◦ CA aTDC, the
pressure of 1.3 bar and intake temperature of 322 K. The ammonia ratio by
NOx emissions of the HPDF modes start to decrease while those of the
energy is 97% for the HPDF modes.
pure diesel mode exhibit a negligibly small change, which can be
attributed to the thermal DeNOx processes (i.e. NH+NO→N2+OH,
between the ammonia spray and diesel spray is 0◦ in the case of HPDF
NH+NO→N2O+H) for the HPDF modes. Generally, the HPDF mode has
0◦ . It should be noted that the in-cylinder temperature is almost constant
the potential to reduce greenhouse gas and NOx simultaneously, espe­
from − 5◦ CA aTDC to TDC for the HPDF 22.5◦ case, this is because that
cially for the situation where the angle between the ammonia spray and
the high vaporization energy of liquid ammonia counteracts the effects
diesel spray is 0◦ . In the following sections, the angle between the
of piston moving.
ammonia spray and diesel spray will be kept at 0◦ .
Fig. 8 shows the effects of dual-fuel injector hole arrangement on the
energy balances. Here, the unburned loss is defined as the difference
3.1.2. Effects of diesel ratio by energy
between the total input energy and the cumulative heat release. Owing
Fig. 11 shows the effects of the diesel ratio by energy on the heat
to the relatively poor flammability of ammonia, the unburned loss of the
release rate, in-cylinder pressure and temperature of the HPDF engines.
HPDF mode is higher than that of the pure diesel mode. Moreover, the
The information of the pure diesel mode is also plotted for comparison.
quenching distance of ammonia is longer than that diesel fuel [6],
The diesel injection timing is kept at − 8◦ CA aTDC, total input energy of
leading to more trapped fuel in the crevice volume. Since the flame-wall

368
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

Fig. 11. Effects of the diesel ratio by energy on the heat release rate, in-
cylinder pressure and temperature of the HPDF engines. The diesel injection Fig. 13. Effects of the diesel ratio by energy on the emission characteristics of
timing is kept at − 8◦ CA aTDC, total input energy of about 1950 J, excess air the HPDF engines. The diesel injection timing is kept at − 8◦ CA aTDC, total
ratio of about 1.7, engine speed of 1000 rpm, intake pressure of 1.3 bar and input energy of about 1950 J, excess air ratio of about 1.7, engine speed of
intake temperature of 322 K. 1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K.

ignition delay of liquid ammonia is defined as the duration from start of

liquid ammonia injection to the timing when heat release rate is larger
than 30 J/deg. Since 1% diesel case leads to misfire, approximately 3%
diesel by energy is required to serve as an ignition source for the
following injected low-reactivity ammonia under the HPDF modes.
Fig. 13 shows the effects of the diesel ratio by energy on the emission
characteristics of the HPDF engines. Owing to the increased diesel
fraction, the CO2 emissions of the 10% diesel case is higher than those of
the 3% diesel case. However, the equivalent CO2 of the former case is
lower, which can be attributed to the reduced N2O emissions. It should
be noted that the NOx emissions of HPDF modes are lower than those of
the pure diesel mode, although the former cases are believed to produce
more fuel-borne NOx. In addition to the reduced thermal NOx by the
lowered in-cylinder temperature in the HPDF modes, this can be
attributed to the thermal DeNOx processes of the HPDF modes by direct
injection NH3 as a reductant. Generally, compared with the pure diesel
mode, the HPDF modes can significantly reduce the greenhouse gas
Fig. 12. Effects of the diesel ratio by energy on the energy balances of the
emissions while maintaining the indicated thermal efficiency.
HPDF engines. The diesel injection timing is kept at − 8◦ CA aTDC, total input
energy of about 1950 J, excess air ratio of about 1.7, engine speed of 1000 rpm,
intake pressure of 1.3 bar and intake temperature of 322 K. 3.2. Low-pressure injection dual-fuel (LPDF) engine

about 1950 J, excess air ratio of about 1.7, engine speed of 1000 rpm, The spray flame evolutions, heat release rate, in-cylinder pressure
intake pressure of 1.3 bar and intake temperature of 322 K. For the HPDF and temperature of the LPDF modes with different diesel injection mass
modes with different diesel injection mass, the injection timing and in­ are shown in Figs. 14 and 15, and the information of pure diesel mode is
jection duration of diesel and ammonia are maintained constant. It can also added for comparison. The diesel injection timing is kept at − 8◦ CA
be seen that the ignition delays of pilot diesel fuel are very similar aTDC, engine speed of 1000 rpm, intake pressure of 1.3 bar and intake
among the pure diesel mode and HPDF modes, except for the misfire temperature of 322 K. The diesel injection mass is varied while the diesel
case of 1% diesel by energy. For the HPDF mode with 10% diesel by injection timing, intake temperature, intake pressure and engine speed
energy, the ammonia begins to release heat immediately after it is are kept constant for all the cases. For the LPDF modes, the ammonia and
injected into the cylinder, leading to a reduced unburned loss compared air are fully mixed in the cylinder at the intake valve closing timing, and
with the 3% diesel case, as shown in Fig. 12. However, the increased the excess air ratio before the diesel injection is 1.60. As shown in
pilot diesel also results in an increased heat transfer loss. The ignition Fig. 15, the in-cylinder temperature and pressure at the end of
delay of ammonia of 3% diesel case is longer than that of the 10% diesel compression stroke of the pure diesel mode are slightly higher than
case, resulting in an increased peak of HRR for the former case. Here, the those of the LPDF modes, which can be attributed to the fact that the
ammonia-air mixture has a smaller specific heat ratio than air. The

369
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

Fig. 14. Effects of the diesel ratio by energy on the spray flame evolutions of the LPDF modes. The diesel injection timing is kept at − 8◦ CA aTDC, engine speed of
1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K. The temperature information is superimposed on the isosurfaces with excess air ratio of 1.3.

Fig. 16. Effects of the diesel ratio by energy on the energy balances of the LPDF
engines. The diesel injection timing is kept at − 8◦ CA aTDC, engine speed of
1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K.

cases, and there is a trend that the volume of high-temperature region

increases with the diesel fraction increasing. Moreover, for the LPDF
modes, since the high diesel fraction case injects more diesel fuel before
the ignition, the peak of HRR, in-cylinder maximum pressure and tem­
Fig. 15. Effects of the diesel ratio by energy on the heat release rate, in-
perature increase with the diesel fraction increasing.
cylinder pressure and temperature of the LPDF engines. The diesel injection Fig. 16 shows the effects of diesel ratio by energy on the energy
timing is kept at − 8◦ CA aTDC, engine speed of 1000 rpm, intake pressure of 1.3 balances of the LPDF engines. The heat transfer loss of the LPDF mode is
bar and intake temperature of 322 K. obviously smaller than that of the pure diesel mode, this is because that
the flame-wall interactions near the top dead center are obviously
diesel fuels are injected into cylinder at − 8◦ CA aTDC for both the pure stronger for the pure diesel mode, as shown in Fig. 14. Moreover, as
diesel and LPDF modes, and they are not ignited at − 5◦ CA aTDC. At shown in Fig. 15, more heat is released near the top dead center for the
− 4◦ CA aTDC, the auto-ignition occurs at the spray downstream for the pure diesel mode, which is also a reason for the increased heat transfer
pure diesel mode, while the pilot diesel fuels are still not ignited for the loss. There is a general trend that the unburned loss increases with the
LPDF modes. In addition to the reduced in-cylinder temperature at the ammonia fraction increasing, which can be attributed to the relatively
start of injection timing, the longer ignition delay of the LPDF mode can poor combustion characteristics of ammonia. It should be noted that the
be attributed to the fact that the ammonia in the air suppresses the long quenching distance of ammonia should also be considered to
autoignition of pilot diesel. At TDC, the auto-ignition occurs for all the explain the increased unburned loss. The unburned loss of the 30%
diesel case is larger than that of the 20% diesel case. This can be

370
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

Fig. 17. Effects of the diesel injection timing on the energy balances of the Fig. 19. Comparison of the energy balances among the pure diesel, HPDF and
LPDF engines. The diesel injection timing is kept at − 8◦ CA aTDC, engine speed LPDF modes. The diesel injection timing is kept at − 8◦ CA aTDC, engine speed
of 1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K. of 1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K.

Fig. 20. Comparison of the flame-wall interactions and piston heat flux among
the pure diesel, HPDF and LPDF modes. The diesel injection timing is kept at
− 8◦ CA aTDC, engine speed of 1000 rpm, intake pressure of 1.3 bar and intake
temperature of 322 K. The temperature information is superimposed on the
isosurfaces with excess air ratio of 1.3, and the heat flux information is
superimposed on the piston wall.

Fig. 18. Effects of the diesel ratio by energy on the emission characteristics of LPDF mode with 10% diesel fraction has the smallest CO2 emissions, it
the LPDF engines. The diesel injection timing is kept at − 8◦ CA aTDC, engine produce more greenhouse gas (i.e. the equivalent CO2 emissions here)
speed of 1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K. than the pure diesel mode. The LPDF modes with 20% and 30% diesel
fraction can achieve a smaller equivalent CO2 emissions than the pure
attributed to the reduced excess air ratio and increased unburned carbon diesel mode, but they produce more NOx emissions. The increased NOx
monoxide for the fornmer case. Since the unburned loss is 7.7% for the emissions of these two cases can be attributed to the increased fuel-
LPDF mode with diesel fraction of 10%, a further reduction in the diesel borne NOx and increased in-cylinder temperature, as shown in
fraction at this diesel injection timing will result in an unacceptable Fig. 15. Moreover, the NH3 emissions increases significantly when diesel
indicated thermal efficiency and unburned emissions. As a result, the ratio by energy is reduced to 10%, which also limits the diesel
effects of diesel injection timing on energy balances are studied to replacement ratio for the LPDF mode. Generally, the LPDF with 20%
evaluate the minimum pilot fuel required for LPDF mode, as shown in diesel exhibits the highest indicated thermal efficiency (i.e. 47.0%) and
Fig. 17. It can be seen that further advancing the diesel injection timing the smallest equivalent CO2 emissions (i.e. 74.8% of the pure diesel
will lead to misfire when the diesel ratio by energy is 10%. Generally, mode), and it will be compared with the HPDF mode in Section 3.3.
approximately 10%–20% diesel by energy is required for the stable
operation of ammonia/diesel LPDF engine. 3.3. Comparison of HPDF and LPDF
The emission characteristics of the LPDF engines with different diesel
injection mass is shown in Fig. 18, and the pure diesel mode is added for Fig. 19 shows the comparison of the energy balances among the pure
comparison. It can be seen that the CO2 emissions increase with the diesel, HPDF and LPDF modes. Here, the HPDF mode with 3% diesel and
diesel injection mass increasing, while the increased diesel injection LPDF mode with 20% diesel are selected for comparison because they
mass will lead to the reduced N2O emissions. As a result, although the have the high indicated thermal efficiency and the low greenhouse gas

371
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

higher than those of the HPDF and pure diesel modes. As a result,
although 80% diesel is replaced by ammonia, the LPDF mode only re­
duces the equivalent CO2 emissions by around 25% compared to the
pure diesel mode. Since N2O is mainly formed by the reaction NH + NO
→ N2O + H [43], the fuel injection strategy should be optimized to
reduce the combustion temperature and NO formation as well as N2O
emissions for the LPDF mode. Generally, the LPDF mode has the po­
tential to achieve an increased indicated thermal efficiency, while the
HPDF mode exhibits a better emission characteristics.

4. Summary and conclusions

In this paper, the experiments of ammonia spray under diesel-like

conditions are conducted to calibrate the spray models, and the com­
bustion and emission characteristics of the ammonia/diesel low-
pressure injection dual-fuel (LPDF) mode and high-pressure injection
dual-fuel (HPDF) mode are compared and analyzed in detail. The main
conclusions are listed below:

• The sub-models such as KH-RT for spray break-up and Frossling for
drop evaporation, which have been widely used for diesel spray
simulations, can also be used to predict the vapor- and liquid-phase
penetration evolutions of ammonia spray after proper calibration
against the experimental data.
• The maximum ammonia ratio around 80% by energy is recom­
Fig. 21. Comparison of the emission characteristics among the pure diesel, mended for the LPDF mode, while the HPDF mode has the potential
HPDF and LPDF modes. The diesel injection timing is kept at − 8◦ CA aTDC, to achieve a 97% diesel replacement ratio.
engine speed of 1000 rpm, intake pressure of 1.3 bar and intake temperature of • Compared with the pure diesel mode, the HPDF mode has similar
322 K. indicated thermal efficiency, cooling and exhaust loss, while it can
significantly reduce the greenhouse gas (i.e. CO2 + 300 × N2O) with
emissions as well as the high replacement ratio of diesel fuel under their a little increases in the NH3 emissions. Moreover, the HPDF mode has
respective modes. The diesel injection timing is kept at − 8◦ CA aTDC, the potential to reduce NOx emissions owing to the thermal DeNOx
engine speed of 1000 rpm, intake pressure of 1.3 bar and intake tem­ processes.
perature of 322 K. It can be seen that the energy balances are very • Owing to the reduced cooling loss, the LPDF mode has the potential
similar between the pure diesel and HPDF modes, while the LPDF mode to achieve a higher indicated thermal efficiency than its counterpart
has the smallest heat transfer loss. The reduced heat transfer loss of the pure-diesel mode. However, since the increase of N2O emissions
LPDF mode is closely related to the reduced flame-wall interactions near offsets the benefit of reducing the CO2 emissions, the ability to
the top dead center, as shown in Fig. 20. Owing to the reduced cooling reduce the greenhouse gas emissions is relatively limited for the
loss, the LPDF mode shows the highest indicated thermal efficiency and LPDF mode. As a result, the optimization of the fuel injection strategy
exhaust energy. and combustion chamber geometry as well as the exhaust after­
Fig. 21 shows the comparison of the emission characteristics among treatment are excepted to reduce the N2O for the LPDF mode.
the pure diesel, HPDF and LPDF modes. The NH3 emissions of the LPDF Moreover, since the ammonia in the clearance volume is difficult to
mode are significantly higher than those of the pure diesel and HPDF burn, the LPDF mode significantly increase the NH3 emissions, which
modes. This can be attributed to the fact that the ammonia in the are expected to be cleaned up by the exhaust DeNOx technology such
clearance volume under the LPDF mode is difficult to burn, as shown in as selective catalyst reduction.
Fig. 22. Both the HPDF and LPDF modes can significantly reduce the CO2 • The comparison between the LPDF mode and HPDF mode shows that
emissions, especially for the HPDF mode with 97% ammonia ratio by the former case has the potential to achieve a reduced cooling loss
energy. However, the N2O emissions of the LPDF mode are obviously and therefore an increased indicated thermal efficiency, while the

Fig. 22. Comparison of the NH3 mass fraction among the pure diesel, HPDF and LPDF modes. The diesel injection timing is kept at − 8◦ CA aTDC, engine speed of
1000 rpm, intake pressure of 1.3 bar and intake temperature of 322 K. The isosurfaces refer to the NH3 mass fraction of 0.01.

372
T. Li et al. Journal of the Energy Institute 102 (2022) 362–373

latter case can significantly reduce the engine-out NH3, NOx and [21] H. Liu, J. Li, J. Wang, C. Wu, B. Liu, J. Dong, et al., Effects of injection strategies on
low-speed marine engines using the dual fuel of high-pressure direct-injection
greenhouse gas emissions.
natural gas and diesel, Energy Sci. Eng. 7 (2019) 1994–2010.
[22] S. Gleis, S. Frankl, D. Waligorski, M. Prager, G. Wachtmeister, Investigation of the
Credit author statement High-Pressure-Dual-Fuel (HPDF) Combustion Process of Natural Gas on a Fully
Optically Accessible Research Engine, 2019. SAE paper 2019-01-2172.
[23] S.G. Frankl, S. Gleis, S. Karmann, M. Prager, G. Wachtmeister, Investigation of
Tie Li: Supervision, Funding acquisition, Writing - review & editing. ammonia and hydrogen as CO2-free fuels for heavy duty engines using a high
Xinyi Zhou: Conceptualization, Methodology, Model, Experiment, Data pressure dual fuel combustion process, Int. J. Eng. Res. 22 (10) (2021) 3196–3208.
curation, Investigation, Writing - original draft. Ning Wang: Experiment. [24] S.G. Frankl, A.D. Gelner, S. Gleis, M. Härtl, G. Wachtmeister, Numerical study on
renewable and sustainable fuels for HPDF engines, in: Proceedings of the ASME
Xinran Wang: Experiment. Run Chen: Writing - review & editing. Shiyan 2020 Power Conference, 2020. NO. POWER2020-16438.
Li: Writing - review & editing. Ping Yi: Writing - review & editing. [25] X. Zhou, T. Li, Y. Wei, S. Wu, Scaling spray combustion processes in marine low-
speed diesel engines, Fuel 258 (2019), 116133.
[26] N. Wang, S. Huang, Z. Zhang, T. Li, P. Yi, D. Wu, et al., Laminar burning
Declaration of competing interest characteristics of ammonia/hydrogen/air mixtures with laser ignition, Int. J.
Hydrogen Energy 46 (2021) 31879–31893.
The authors declare that they have no known competing financial [27] J.S. Cardoso, V. Silva, R.C. Rocha, M.J. Hall, Eusebio D. CostaM, Ammonia as an
energy vector: current and future prospects for lowcarbon fuel applications in
interests or personal relationships that could have appeared to influence internal combustion engines, J. Clean. Prod. 296 (2021), 126562.
the work reported in this paper. [28] X. Zhou, T. Li, Z. Lai, B. Wang, Scaling fuel sprays for different size diesel engines,
Fuel 225 (2018) 358–369.
[29] Z. Zhang, T. Li, R. Chen, N. Wang, Y. Wei, D. Wu, Injection characteristics and fuel-
Acknowledgements
air mixing process of ammonia jets in a constant volume vessel, Fuel 304 (2021),
131408.
The supports by the Major International (Regional) Joint Research [30] X. Zhou, T. Li, Modeling of the entire processes of diesel spray tip penetration
Project of National Natural Science Foundation of China (52020105009) including the start- and end-of-injection transients, J. Energy Inst. 98 (2021)
271–281.
are gratefully acknowledged. [31] X. Zhou, T. Li, P. Yi, Modeling of diesel spray tip penetration during start-of-
injection transients, Int. J. Eng. Res. 22 (9) (2021) 3013–3029.
References [32] M. Jangi, R. Solsjo, B. Johansson, X.-S. Bai, On large eddy simulation of diesel
spray for internal combustion engines, Int. J. Heat Fluid Flow 53 (2015) 68–80.
[33] Z. Han, R.D. Reitz, Turbulence modeling of internal combustion engines using RNG
[1] R.D. Reitz, H. Ogawa, R. Payri, T. Fansler, S. Kokjohn, Y. Moriyoshi, et al., IJER κ-ε models, Combust. Sci. Technol. 106 (4–6) (1995) 267–295.
editorial: the future of the internal combustion engine, Int. J. Eng. Res. 21 (1) [34] R.D. Reitz, F. Bracco, Mechanisms of breakup of round liquid jets, in: Encyclopedia
(2020) 3–10. of fluid mechanics 3, 1986, pp. 233–249.
[2] X. Zhou, T. Li, P. Yi, The similarity ratio effects in design of scaled model [35] D.P. Schmidt, C. Rutland, A new droplet collision algorithm, J. Comput. Phys. 164
experiments for marine diesel engines, Energy 231 (2021), 121116. (1) (2000) 62–80.
[3] Clarkson Research, Alternative Fuels Installations by Vessel, 2020. Sep. [36] A.A. Amsden, P.J. O’rourke, T.D. Butler, KIVA II-A Computer Program for
[4] M.-C. Chiong, C.T. Chong, J.-H. Ng, S. Mashruk, W.W.F. Chong, N.A. Samiran, et Chemically Reactive Flows with Sprays. Los Alamos National Lab Report No, LA-
al., Advancements of combustion technologies in the ammonia-fuelled engines, 11560-MS, 1989.
Energy Convers. Manag. 244 (2021), 114460. [37] A.A. Amsden, M. Findley, KIVA-3V: A Block-Structured KIVA Program for Engines
[5] P. Berwal, S. Kumar, B. Khandelwal, A comprehensive review on synthesis, with Vertical or Canted Valves, Los Alamos National Lab report no. LA-13313-MS,
chemical kinetics, and practical application of ammonia as future fuel for 1997.
combustion, J. Energy Inst. 99 (2021) 273–298. [38] X. Zhou, T. Li, P. Yi, Z. Zhang, N. Wang, Y. Wei, On the fuel injection rate profile as
[6] P. Dimitriou, R. Javaid, A review of ammonia as a compression ignition engine boundary conditions for diesel spray combustion simulations, Fuel 276 (2020),
fuel, Int. J. Hydrogen Energy 45 (2020) 7098–7118. 118026.
[7] D. Clerk, Gas or Other Internal-Combustion Engine, 1904, US20803504A. United [39] R. He, P. Yi, T. Li, X. Zhou, Y. Gu, Evaluations of the KH-RT breakup and dynamic
States. structure SGS models for evaporating sprays under diesel engine-like conditions,
[8] E. Kroch, Ammonia - a fuel for motor buses, J. Inst. Petrol. 31 (1945) 213–223. Atomization Sprays 30 (3) (2020) 189–212.
[9] T.J. Pearsall, C.G. Garabedian, Combustion of Anhydrous Ammonia in Diesel [40] X. Zhou, T. Li, Y. Wei, N. Wang, Scaling liquid penetration in evaporating sprays
Engines, 1967. SAE paper 670947. for different size diesel engines, Int. J. Eng. Res. 21 (9) (2020) 1662–1677.
[10] J.T. Gray, E. Dimitroff, N.T. Meckel, R.D. Quillian, Ammonia Fuel - Engine [41] G.P. Smith, D.M. Golden, M. Frenklach, N.W. Moriarty, B. Eiteneer, M. Goldenberg,
Compatibility and Combustion, 1966. SAE paper 660156. et al., GRI-Mech 3.0, http://combustion.berkeley.edu/gri-mech/version30/file
[11] A.J. Reiter, S.-C. Kong, Demonstration of compression-ignition engine combustion s30/grimech30.dat.
using ammonia in reducing greenhouse gas emissions, Energy Fuel. 22 (2008) [42] A. Yousefi, H. Guo, S. Dev, B. Liko, S. Lafrance, Effects of ammonia energy fraction
2963–2971. and diesel injection timing on combustion and emissions of an ammonia/diesel
[12] A.J. Reiter, S.-C. Kong, Combustion and emissions characteristics of compression- dual-fuel engine, Fuel 314 (2022), 122723.
ignition engine using dual ammonia-diesel fuel, Fuel 90 (2011) 87–97. [43] C. Duynslaegher, H. Jeanmart, J. Vandooren, Flame structure studies of premixed
[13] Y. Niki, D.-H. Yoo, K. Hirata, H. Sekiguchi, Effects of ammonia gas mixed into ammonia/hydrogen/oxygen/argon flames: experimental and numerical
intake air on combustion and emissions characteristics in diesel engine, in: investigation, Proc. Combust. Inst. 32 (2009) 1277–1284.
Proceedings of the ASME 2016 Internal Combustion Engine Fall Technical
Conference, 2016. NO. 2016-9364.
[14] Y. Niki, Y. Nitta, H. Sekiguchi, K. Hirata, Diesel fuel multiple injection effects on Nomenclature
emission characteristics of diesel engine mixed ammonia gas into intake air, J. Eng.
Gas Turbines Power-Trans. ASME 141 (6) (2019), 061020.
[15] Y. Niki, Reductions in unburned ammonia and nitrous oxide emissions from an aTDC: after top dead center
ammonia-assisted diesel engine with early timing diesel pilot injection, J. Eng. Gas HRR: heat release rate
Turbines Power-Trans. ASME 143 (9) (2021), 091014. CI: compression ignition
[16] K.L. Tay, W. Yang, J. Li, D. Zhou, W. Yu, F. Zhao, et al., Numerical investigation on GHG: greenhouse gas
the combustion and emissions of a kerosene-diesel fueled compression ignition IMO: International Maritime Organization
engine assisted by ammonia fumigation, Appl. Energy 204 (2017) 1476–1488. LPDF: low-pressure injection dual-fuel
[17] K.L. Tay, W. Yang, S.K. Chou, D. Zhou, J. Li, W. Yu, et al., Effects of injection HPDF: high-pressure injection dual-fuel
timing and pilot fuel on the combustion of a kerosene-diesel/ammonia dual fuel 3-D: three-dimensional
engine: a numerical study, Energy Proc. 105 (2017) 4621–4626. CFD: computational fluid dynamics
[18] J. Lasocki, M. Bednarski, M. Sikora, Simulation of ammonia combustion in dual- CVCC: constant volume combustion chamber
fuel compression-ignition engine, in: IOP Conference Series: Earth and LED: light emitting diode
Environmental Science, vol. 214, 2019, 012081. DBI: diffused back-illumination
[19] S. Frankl, S. Gleis, Development of a 3D-computational fuid dynamics model for a RANS: Reynolds-Averaged-Navier-Stokes
full optical high-pressure dual-fuel engine, SAE Int. J. Engines 13 (2) (2020) KH-RT: Kevin Helmholtz-Rayleigh Taylor
241–252. AMR: adaptive mesh refinement
[20] Y. Dong, O. Kaario, G. Hassan, O. Ranta, M. Larmi, B. Johansson, High-pressure
direct injection of methanol and pilot diesel: a non-premixed dual-fuel engine
concept, Fuel 277 (2020), 117932.

373