Effects of Equivalence Ratio on Emission Characteristics of Ammonia–Hydrogen Dual-Fuel Engine

Abstract

Ammonia–hydrogen dual-fuel engines are regarded as a promising clean-energy power system, but nitrogen-based pollutants (NH₃, NO, N₂O, and NO₂) emitted during their operation restrict their popularization and application. To clarify the influence of equivalence ratio on the emission characteristics of ammonia–hydrogen dual-fuel engines, the emission laws and formation mechanisms of nitrogen-based pollutants under different equivalence ratios were studied and the evolution of the in-cylinder cross-sectional mole fraction of NO, N₂O and NO₂ was analyzed. The results show that equivalence ratio has a significant regulatory effect on in-cylinder temperature, combustion efficiency and pollutant formation. Unburned NH₃ emission first decreases and then increases, reaching the minimum at an equivalence ratio of 1.05. NO emission peaks at 0.74 due to high temperature and combustion efficiency and is nearly zero at 1.48 under fuel-rich conditions. N₂O emission is highest at 0.5–0.62 and is inhibited at high equivalence ratios. NO₂ is mainly generated under low equivalence ratios and disappears when the equivalence ratio exceeds 0.74. This study reveals the influence mechanism of equivalence ratio on nitrogen-based pollutants of ammonia–hydrogen dual-fuel engines and provides a theoretical basis and experimental support for the optimization of engine emission-control strategies.

1. Introduction

Energy is the cornerstone of human social progress and the driving force for industrial, technological, and economic development [1]. With the continuous innovation of industrial systems, energy demand has increased significantly. A large amount of fossil fuels is used in industrial development, and their combustion releases massive quantities of carbon dioxide, thereby affecting the global climate [2], exacerbating environmental degradation, and causing global warming. This has in turn raised societal concerns about the depletion of fossil resources and changing climate patterns [3]. Hydrogen is widely regarded as a promising candidate for future clean energy due to its high calorific value, low ignition requirements, and rapid flame propagation characteristics [4]. Nevertheless, the physical limitations of hydrogen, such as difficulty in liquefaction, susceptibility to permeation leading to metal hydrogen embrittlement, and its flammable and explosive nature, pose considerable challenges to its large-scale storage, transportation, and utilization [5]. Consequently, researchers have shifted their focus to ammonia fuel—another zero-carbon alternative energy source—in an attempt to overcome the technical bottlenecks of hydrogen fuel. As a zero-carbon fuel, ammonia has a high hydrogen content and benefits from a highly mature storage and transportation system, making it an optimal carrier for green hydrogen [6]. Below, ammonia, hydrogen, methanol, and LNG are compared in terms of environmental friendliness, combustion efficiency, storage and transportation costs, and safety [7].

Numerous scholars at home and abroad have laid a solid research foundation for ammonia/hydrogen blended fuel engines. Ammonia–hydrogen dual-fuel engines can be classified into three categories based on ignition methods: (1) compression ignition (CI); (2) spark ignition (SI); and (3) jet ignition (JI). Their combustion modes include homogeneous charge compression ignition (HCCI), homogeneous charge spark ignition (HCSI), port injection jet ignition (PIJI), and direct injection jet ignition (DIJI) [8]. This study adopts the spark ignition combustion mode. A spark plug generates an electric spark inside the cylinder to ignite the ammonia–hydrogen mixture [9]. Spark ignition enables precise control of ignition timing, bringing the combustion phase closer to the maximum brake torque (MBT) timing. Ammonia requires a high ignition energy and has a slow flame propagation speed, resulting in unstable combustion and low efficiency at low hydrogen blending ratios [10].

Therefore, the hydrogen–ammonia blending ratio has a significant impact on the engine’s combustion performance and emission characteristics. C. Lhuillier et al. [11] investigated the engine’s combustion and emission characteristics at different equivalence ratios by controlling the hydrogen energy fraction from 0% to 54%. Their research showed that increasing the hydrogen energy fraction enhances flame propagation speed and shortens the ignition delay period. When the hydrogen fraction reaches 10%, the hydrogen energy ratio significantly affects early flame formation. Joo et al. [12] experimentally studied the flammability limits and nitrogen oxide (NO_x) emissions of premixed ammonia–hydrogen flames. The results indicated that blending hydrogen into premixed ammonia fuel broadens the flammability limits. Although the absolute value of NO_x emissions increases, the NO_x emission index remains nearly constant as the energy substitution rate of ammonia fuel rises. Additionally, elevating the intake temperature greatly influences expanding the flammability limits of premixed ammonia–hydrogen flames. Valera-Medina et al. [13] studied the combustion of 30% H₂/70% NH₃ and 40% H₂/60% NH₃ (by volume) under fuel-rich conditions. The results demonstrated that hydrogen addition improves ammonia’s chemical reactivity, ensuring good combustion stability without excess air, and ammonia–hydrogen blended fuel has considerable potential to replace conventional fuels. Dong et al. [14] evaluated the potential of ammonia and hydrogen as zero-carbon marine fuels, focusing on two-stroke and four-stroke engines on oil tankers. They experimentally compared the performance of marine propulsion systems using ammonia–hydrogen fuel versus conventional fuels. The findings revealed that ammonia and hydrogen offer substantial potential for mitigating global warming and are highly effective in reducing particulate matter emissions compared to traditional fuels. Wei et al. [15] investigated the effect of hydrogen blending on ammonia combustion flame characteristics and its flame stabilization mechanism in a gas turbine combustor and measured the flame structure and flow field. The results showed that blending 10% hydrogen (molar fraction) with ammonia produces a more stretchable flame, enhancing flame stability without increasing NO_x emissions.

Chen et al. [16] studied the combustion and emission characteristics of premixed ammonia and ammonia–hydrogen flames in a porous burner, analyzing the effects of porous burner parameters, hydrogen blending ratio, and equivalence ratio on ammonia flame characteristics, temperature distribution, and NO formation. The results indicated that ammonia combustion temperature rises with increasing hydrogen blending ratio, peaking at an equivalence ratio of 1. When the equivalence ratio exceeds 1, little nitrogen in the fuel is converted to NO in both pure ammonia and ammonia–hydrogen blended combustion. Tong et al. [17] conducted numerical simulations by constructing a three-dimensional ammonia–hydrogen combustion chamber model, analyzing the effects of intake pressure, equivalence ratio, and hydrogen blending ratio on temperature distribution, combustion chamber wall temperature, NO emissions, and efficiency. The results showed that an equivalence ratio in the range of 0.95–1 has a minor effect on improving thermodynamic performance but significantly reduces NO_x emissions. A hydrogen blending ratio below 15% markedly reduces NO_x formation. Cheng et al. [18] numerically studied the laminar flame propagation speed and NO_x emissions of hydrogen-rich ammonia under high-temperature and high-pressure conditions. Their research revealed that a higher initial temperature increases laminar flame propagation speed but also elevates emissions, while an increase in initial pressure suppresses the growth of laminar flame propagation speed and reduces emissions. Li et al. [19] investigated the knock characteristics of ammonia–hydrogen blended fuel using a dual-fuel optical spark ignition engine. Tests showed that, when the hydrogen/ammonia ratio is below 50%, the entire combustion process exhibits normal premixed flame propagation with no signs of end-gas autoignition. When the hydrogen content reaches 60%, spark-assisted compression ignition occurs, improving thermal efficiency without causing engine knock. However, when the hydrogen concentration exceeds 70%, the blended fuel begins to exhibit knock behavior, with knock intensity increasing as hydrogen content rises. Pengzhen Li et al. [20] explored the effects of injection timing and engine speed on the combustion and emission characteristics of an ammonia–hydrogen dual-fuel rotary engine. Their experiments indicated that, at the same injection timing, a higher ammonia fraction generally reduces peak pressure and delays its occurrence to after the piston’s top dead center. Retarding injection timing increases the fuel consumption rate of ammonia and hydrogen.

Despite the extensive research conducted by scholars both domestically and internationally on ammonia–hydrogen dual-fuel engines, the following critical research gaps still remain: The existing studies have predominantly focused on the analysis of total NO_x emissions, with very few systematically revealing the synergistic effects of equivalence ratio on the four nitrogen-based pollutants—NH₃, NO, N₂O, and NO₂. In particular, the formation mechanism of N₂O, which has an extremely strong greenhouse effect, remains insufficiently investigated. Most studies have only focused on the final emission results at the exhaust outlet, lacking a spatiotemporal evolution analysis of the entire process of pollutant formation, diffusion, and consumption within the cylinder. The coupling relationship between combustion characteristics and emission characteristics under different equivalence ratios has not been fully elucidated, making it difficult to guide practical engine operation optimization.

This study is based on the Caterpillar 3401 heavy-duty diesel engine as the prototype. By retrofitting a hydrogen injector and a spark plug, the original premixed compression ignition mode was converted to a spark ignition mode. A three-dimensional numerical simulation model was established using CONVERGE V3.0 software, and its reliability was validated through comparison with experimental data from the prototype engine. This paper focuses exclusively on the effects and formation mechanisms of equivalence ratio on the combustion and emission characteristics of an ammonia–hydrogen dual-fuel engine under a fixed 20% hydrogen energy share. The emission characteristics and in-cylinder spatiotemporal evolution of four major nitrogen-based pollutants (NH₃, NO, N₂O, and NO₂) across a wide equivalence ratio range (0.5–1.48) were systematically analyzed.

2. Numerical Model and Validation

2.1. Engine Operating Parameters

The Caterpillar 3401 engine was selected as the prototype for the following three main reasons: This engine model is widely used internationally as a standard heavy-duty diesel engine for alternative fuel research, with a large amount of publicly available experimental data for model validation, ensuring the comparability of the research results. Its single-cylinder displacement of 2.44 L and compression ratio of 16:1 are highly similar to the parameter characteristics of marine medium-speed diesel engines, giving the research results good engineering applicability. The geometric structure of this engine is publicly available and easily accessible, facilitating the establishment of a high-precision three-dimensional simulation model.

To investigate the combustion and emission characteristics of ammonia–hydrogen dual fuel, this study refers to the experimental research on an ammonia–diesel dual-fuel engine conducted by Amin Yousefi et al. [20]. On this basis, the engine model is modified in the present work: ammonia is injected into the intake port, mixes with air to form a premixed mixture, and then enters the cylinder, with the NH₃/N₂/O₂ ratio determined by mass fraction. A hydrogen injector is installed on the cylinder head, through which hydrogen is directly injected into the cylinder at high pressure, realizing a combined injection strategy of port fuel injection of ammonia and direct injection of hydrogen. A spark plug is mounted at the center of the cylinder head to ignite the in-cylinder mixture.

A three-dimensional engine simulation model, including intake port, exhaust port, hydrogen injector and spark plug, is established to simulate the intake, compression and power strokes of the ammonia–hydrogen dual-fuel engine. The 3D simulation model of the ammonia–hydrogen dual-fuel engine is shown in Figure 1 below. The main technical parameters of the Caterpillar 3401 engine are listed in

Table 1. Main structural parameters of Caterpillar 3401 single-cylinder diesel engine.

Parameter	Value
Engine Model	Caterpillar 3401 [21,22,23]
Cylinder Bore (cm)	13.72
Piston Stroke (cm)	16.51
Connecting Rod Length (cm)	26.3
Single-Cylinder Rated Speed (r·min−1)	1600
Number of Cylinders	1
Number of Strokes	4
Compression Ratio	16
Engine Displacement (L)	2.44
Exhaust Valve Opening/Closing Timing (° CA ATDC)	145/348
Intake Valve Opening/Closing Timing (° CA ATDC)	−385/−169
Hydrogen Nozzle Diameter (mm)	2
Angle Between Hydrogen Nozzle and Cylinder Head Plane (° CA)	50

.

2.2. Simulation Sub-Model Settings

This study employs the ammonia oxidation mechanism proposed by Zhang et al. [24], comprising 38 species and 263 elementary reactions, as the detailed chemical kinetic mechanism for ammonia–hydrogen combustion. The selection of this mechanism is primarily based on the following considerations: this mechanism was specifically developed for ammonia–hydrogen blended fuels and has been extensively validated under various experimental conditions, including jet-stirred reactors (JSRs) and laminar flame speeds, enabling accurate prediction of the key reaction characteristics of NH₃/NO/H₂ systems.

Compared with the classic ammonia combustion mechanism by Glarborg et al., this mechanism provides a more accurate description of the N₂O formation pathway, which is critically important for the comprehensive analysis of all nitrogen-based pollutant species, which is the central focus of this study. The computational scale of this mechanism is moderate, satisfying the computational efficiency requirements of three-dimensional engine simulations while maintaining accuracy. This mechanism particularly emphasizes the regulatory role of H₂ on the oxidation rate of NO_x, providing an accurate predictive foundation for the formation and evolution of NO_x. The decomposition mechanism of NH₃ and the formation pathway of NO_x are illustrated in Figure 2.

It can be seen from Figure 2 that HNO is an important intermediate species for the formation of NO and NO₂. Meanwhile, the main equations involved in the reaction pathways during ammonia–hydrogen combustion are given in the following formulas. NO emissions mainly consist of thermal NO and fuel NO. Thermal NO is primarily formed under high-temperature and oxygen-rich conditions. When the temperature exceeds 1800 K, nitrogen and oxygen react to generate NO via the Zeldovich mechanism [24], with the main reaction pathways as follows:

$$\begin{array}{c}{\mathrm{N}}_2+\mathrm{O}\rightleftharpoons \mathrm{N}\mathrm{O}+\mathrm{N}\end{array}$$

(1)

$$\begin{array}{c}\mathrm{N}+{\mathrm{O}}_2\rightleftharpoons \mathrm{N}\mathrm{O}+\mathrm{O}\end{array}$$

(2)

$$\begin{array}{c}\mathrm{N}+\mathrm{O}\mathrm{H}\rightleftharpoons \mathrm{N}\mathrm{O}+\mathrm{H}\end{array}$$

(3)

This is the first step of ammonia oxidation and also the rate-controlling step of the entire combustion process. The increased concentration of active radicals significantly enhances the ammonia decomposition rate, thereby reducing unburned ammonia emissions. Compared with the simulation results of a pure ammonia engine under the same operating conditions, 20% hydrogen blending improves the combustion efficiency from 72.3% to 90.8% (at Φ = 0.62). Fuel NO formation is influenced by the oxidation level of NH₃ and mainly proceeds via the reaction pathway NH₃→NH₂→HNO/NH/N→NO [25]. Among numerous reaction routes, HNO is predominantly converted into NO through its decomposition or interactions with H/OH radicals, and this conversion process accounts for a considerable proportion. The key reaction equations are listed below, in which M represents the third-body species involved in the reaction process. It is reasonable to characterize the evolution of fuel NO by the evolution of HNO [25]. NO consumption reactions are generally important under moderate-temperature conditions ranging from 1200 K to 1400 K and fuel-rich environments [26].

$$\begin{array}{c}\mathrm{H}\mathrm{N}\mathrm{O}+\mathrm{M}=\mathrm{N}\mathrm{O}+\mathrm{H}+\mathrm{M}\end{array}$$

(4)

$$\begin{array}{c}\mathrm{H}\mathrm{N}\mathrm{O}+\mathrm{H}=\mathrm{N}\mathrm{O}+{\mathrm{H}}_2\end{array}$$

(5)

$$\begin{array}{c}\mathrm{H}\mathrm{N}\mathrm{O}+\mathrm{O}\mathrm{H}=\mathrm{N}\mathrm{O}+{\mathrm{H}}_2\mathrm{O}\end{array}$$

(6)

$$\begin{array}{c}\mathrm{N}\mathrm{O}+\mathrm{N}{\mathrm{H}}_2={\mathrm{N}}_2+{\mathrm{H}}_2\mathrm{O}\end{array}$$

(7)

$$\begin{array}{c}\mathrm{N}\mathrm{O}+\mathrm{N}\mathrm{H}={\mathrm{N}}_2+\mathrm{O}\mathrm{H}\end{array}$$

(8)

$$\begin{array}{c}\mathrm{N}\mathrm{O}+\mathrm{N}={\mathrm{N}}_2+\mathrm{O}\end{array}$$

(9)

N₂O is mainly formed under high pressure, moderate and low temperatures (between 800 and 1200 K), and lean-burn conditions [27], with the reaction formula shown in Equation (10). Its consumption reactions, however, occur primarily at high temperatures [28], where N₂O is converted to N₂ via direct decomposition or reaction with H, as illustrated in reaction pathways (11) and (12).

$$\begin{array}{c}\mathrm{N}\mathrm{H}+\mathrm{N}\mathrm{O}={\mathrm{N}}_2\mathrm{O}+\mathrm{H}\end{array}$$

(10)

$$\begin{array}{c}{\mathrm{N}}_2\mathrm{O}+\mathrm{H}={\mathrm{N}}_2+\mathrm{O}\mathrm{H}\end{array}$$

(11)

$$\begin{array}{c}{\mathrm{N}}_2\mathrm{O}+\mathrm{M}={\mathrm{N}}_2+\mathrm{O}+\mathrm{M}\end{array}$$

(12)

The formation of NO₂ mainly occurs during the exhaust phase and in low-temperature regions near the combustion chamber walls where the temperature is below 1000 K, where NO reacts with excess O₂ and HO₂ to form NO₂. The reaction pathways are as follows:

$$\begin{array}{c}\mathrm{N}\mathrm{O}+{\mathrm{O}}_2\rightleftharpoons \mathrm{N}{\mathrm{O}}_2+\mathrm{O}\end{array}$$

(13)

$$\begin{array}{c}\mathrm{N}\mathrm{O}+\mathrm{H}{\mathrm{O}}_2\to \mathrm{N}{\mathrm{O}}_2+\mathrm{O}\mathrm{H}\end{array}$$

(14)

$$\begin{array}{c}2\mathrm{N}\mathrm{O}+{\mathrm{O}}_2\to 2\mathrm{N}{\mathrm{O}}_2\end{array}$$

(15)

When simulating the in-cylinder combustion process of the engine using the CONVERGE V3.0 software, a series of key sub-models must be employed to ensure the accuracy and reliability of the simulation results. These mainly include the in-cylinder turbulence model, combustion model, spray model, emission model, and ignition model. The sub-models adopted in the simulation and calculation process are listed in

Table 2. Selection of sub-models for simulation and calculation.

Model Category	Selected Model
Turbulence Model	k-ε model [29]
Heat Transfer Model	Han and Reitz [30]
Spark Ignition Model	Energy
Combustion Model	SAGE [31]
Combustion Reaction Mechanism	Zhang [24]

.

The rationale for selecting each sub-model is as follows:

k-ε turbulence model: This is the most widely used turbulence model in engine flow simulations, capable of accurately predicting the in-cylinder turbulent flow characteristics with high computational efficiency. For the large-scale in-cylinder flow and mixing processes of interest in this study, the k-ε model provides sufficient accuracy.

Han and Reitz heat transfer model: This model was specifically developed for in-cylinder heat transfer processes in engines, taking into account the effects of variable-density turbulence on heat transfer, enabling more accurate prediction of cylinder wall heat transfer losses. Compared with the conventional wall function approach, this model offers higher prediction accuracy under engine operating conditions.

Energy ignition model: This model simulates the spark ignition process by adding an energy source in the spark plug region, accurately predicting ignition delay and early flame development. The model is simple yet practical and can be coupled with detailed chemical kinetic mechanisms.

SAGE combustion model: This is a detailed chemical kinetics solver that can be coupled with any detailed chemical mechanism to accurately simulate the fuel combustion process and pollutant formation. For the analysis of nitrogen-based pollutant formation mechanisms, which is the focus of this study, the SAGE model is the optimal choice.

2.3. Boundary Condition Setting and Mesh Generation

In this study, the initial and boundary conditions of the simulation model were defined with reference to data from numerous scholarly and expert publications. The boundary conditions were set with reference to the experimental configuration of Yousefi et al. [21], and the wall temperatures were based on the results reported by Madihi et al. [23]. The boundary and initial conditions of the simulation model are listed in

Table 3. Boundary and initial conditions of the simulation model.

Boundary Zone	Boundary Type Setting	Value
Intake Port	Inflow	319 K/0.193 MPa
Exhaust Port	Outflow	700 K/0.101325 MPa
Intake Duct	Wall	350 K
Exhaust Duct	Wall	450 K
Intake Valve	Moving wall	450 K
Exhaust Valve	Moving wall	465 K
Cylinder Wall	Fixed wall	480 K
Piston	Moving wall	480 K
Cylinder Head	Fixed wall	480 K
Hydrogen Injector Inlet	Inflow	350 K
Hydrogen Injector	Fixed wall	350 K
Spark Plug	Fixed wall	550 K

, and the initial conditions of the computational domain are also presented in Table 3. The initial conditions of the computational domain are presented in

Table 4. Initial conditions of the computational domain.

Parameter	Intake Duct	Exhaust Duct	Cylinder	Hydrogen Injector
Temperature/K	350	650	1200	350
Pressure/MPa	0.101325	0.101325	0.1652	8.001325

.

To ensure simulation accuracy while maintaining computational efficiency, the base mesh size was set to 4 mm, with appropriate mesh refinement levels applied to other computational regions. This parameter setup strategy satisfies the mesh quality requirements in critical regions and achieves optimized allocation of computational resources by reasonably controlling refinement levels, which conforms to the best industrial practice for balancing accuracy and efficiency. The mesh refinement strategy is presented in

Table 5. Mesh refinement strategy.

Computational Domain	Refinement Level	Mesh Size (mm)
Base mesh	-	4
Local in-cylinder refinement	2	1
Intake boundary refinement	3	0.5
Spark plug spherical zone 1	4	0.25
Spark plug spherical zone 2	5	0.125
Hydrogen direct injection boundary refinement	5	0.125
Velocity-based AMR	3	0.5
Temperature-based AMR	3	0.5

.

The mesh refinement strategy was formulated based on the following principles: For regions with intense flow and combustion variations (e.g., near the spark plug and in the hydrogen injection zone), a finer mesh resolution was adopted to accurately capture flame propagation and spray development processes. For regions with relatively uniform flow (e.g., the main cylinder body), a coarser mesh was employed to improve computational efficiency.

Adaptive mesh refinement (AMR) technology was adopted to automatically adjust the mesh resolution based on velocity and temperature gradients, further reducing computational cost while ensuring accuracy. The highest level of refinement (0.125 mm) was applied in the spark plug region as the ignition process is highly sensitive to mesh resolution, and an overly coarse mesh may lead to ignition failure or inaccurate prediction of ignition delay. A mesh resolution of 0.125 mm was also adopted in the hydrogen injection zone to accurately simulate hydrogen injection and mixing processes. A base mesh of 4 mm was used for the main cylinder body, balancing computational accuracy with computational cost.

Figure 3 shows the influence of the base mesh size on in-cylinder pressure and heat release rate. It can be observed from the figure that the in-cylinder pressure and heat release rate curves are in close agreement at mesh sizes of 3 mm and 4 mm, with negligible differences in the calculated results. However, when the base mesh size is increased to 5 mm and 6 mm, it exerts a considerable influence on the simulation results. Considering both computational cost and accuracy comprehensively, a base mesh size of 4 mm is selected for the engine.

2.4. Validation of the Simulation Model

In this study, a simulation model of an ammonia–hydrogen dual-fuel engine was established based on the Caterpillar 3401 heavy-duty diesel engine. Since Amin Yousefi et al. [20] also built an experimental platform using the same engine type, the simulation model developed in the present work was validated against their experimental results.

For comparison, the three-dimensional engine simulation model was adjusted to be consistent with the experimental engine of Amin Yousefi et al. Numerical model validation was carried out under 50% load operating conditions using ammonia–diesel dual fuel by tuning the model parameters.

Figure 4 presents a comparison between the simulation results and the experimental data. The experimental conditions used for validation were: 50% load, ammonia–diesel dual-fuel operation, diesel injection timing of −10° CA aTDC, ammonia energy fraction of 50%, and engine speed of 1600 r/min.

The quantitative comparison shows that the simulated peak in-cylinder pressure was 10.1 MPa, compared with the experimental value of 9.8 MPa, yielding a relative error of 3.1%. The crank angle at peak pressure was 8.2° CA aTDC in the simulation and 7.5° CA aTDC in the experiment, resulting in a phase error of 0.7° CA. The shapes of the heat release rate curves and the combustion durations were also in good agreement.

The minor discrepancies between the simulation and the experiment can be primarily attributed to the following factors: experimental measurement uncertainties, such as the measurement error of the in-cylinder pressure sensor (±0.1 MPa) and the crank angle positioning error (±0.1° CA); simplifying assumptions in the numerical model, such as the neglect of in-cylinder blow-by and local non-uniformity of heat transfer; and the difference in fuel properties as the experiment employed ammonia–diesel dual-fuel operation while the simulation used ammonia–hydrogen dual-fuel, which also contributes to a certain degree of deviation. Overall, the model is capable of accurately reproducing the combustion characteristics of the engine and is suitable for subsequent simulation studies.

3. Results and Discussion

3.1. Basic Condition Setting and Analysis of Mixture Evolution in Ammonia–Hydrogen Engines

In the basic setup, the hydrogen energy ratio was fixed at 20%, spark ignition timing at −15° CA aTDC, hydrogen injection timing at −90° CA aTDC, hydrogen injection pressure at 8 MPa, and EGR rate at 0. Under 100% engine load, the equivalence ratio was varied by adjusting the fuel quantity to simulate and investigate the combustion process of the spark ignition ammonia–hydrogen dual-fuel engine. The simulated operating conditions are listed in

Table 6. Simulated operating conditions under different equivalence ratios.

Item	Parameter
Engine Speed/(r·min−1)	1600
Hydrogen Fraction/%	20
Hydrogen Injection Timing/° CA aTDC	−90
Hydrogen Injection Pressure/MPa	8
EGR Rate/%	0
Ignition Timing/° CA aTDC	−15
Equivalence Ratio	0.5, 0.55, 0.62, 0.74, 0.89, 1.05, 1.48

.

An equivalence ratio Φ = 1 corresponds to the stoichiometric mixture at which fuel and air are completely combusted. Φ < 1 indicates excess air and a low ammonia–hydrogen proportion in the mixture, representing lean combustion. Φ > 1 indicates excess fuel and a high ammonia–hydrogen proportion, corresponding to fuel-rich combustion. Lean mixtures help to improve engine thermal efficiency and reduce combustion temperature, thereby reducing NO_x emissions to some extent. However, for ammonia–hydrogen dual-fuel engines, the variation in NO_x emissions with equivalence ratio differs from that of conventional fossil fuel engines: the NO_x emission peak in traditional gasoline engines typically occurs at a slightly lean equivalence ratio of 0.85–0.9, whereas, in this study, the NO emission peak appears at an equivalence ratio of 0.74. This difference is primarily attributed to two factors: The flame temperature of ammonia combustion is lower than that of gasoline, requiring a higher oxygen concentration to reach the critical temperature for thermal NO formation (1800 K). Ammonia fuel inherently contains nitrogen, resulting in a higher proportion of fuel–NO in total NO_x, and fuel–NO formation is more sensitive to oxygen concentration.

Lhuillier et al. [11] and Chen et al. [16] also observed a similar NO emission peak position in their experimental studies on ammonia–hydrogen engines, validating the rationality of the findings in this study. However, under lean-burn conditions, ammonia exhibits a slow burning rate and incomplete combustion; therefore, hydrogen is added to compensate for the low combustion efficiency of ammonia. In summary, this study primarily focuses on equivalence ratios below 1 while employing equivalence ratios of 1.05 and 1.48 as control cases to directly reveal the combustion and emission characteristics of ammonia–hydrogen dual-fuel engines.

The selection of the above baseline operating parameters in this study is primarily based on the following considerations:

A 20% hydrogen energy ratio: Based on a comprehensive review of the existing studies, this ratio ensures combustion stability while avoiding a sharp increase in NO_x emissions, making it a commonly used blending ratio for ammonia–hydrogen dual-fuel engines. An excessively low hydrogen ratio leads to combustion instability, while an excessively high ratio increases NO_x emissions and knock risk.

Ignition timing of −15° CA aTDC: Determined through preliminary experiments, this timing achieves high thermal efficiency while avoiding knock. Overly advanced ignition timing leads to knock, while overly retarded timing reduces thermal efficiency and increases unburned ammonia emissions.

Hydrogen injection timing of −90° CA aTDC: Injection during the mid-compression stroke ensures sufficient mixing time between hydrogen and the ammonia–air mixture while avoiding wall impingement caused by premature hydrogen diffusion. Overly early injection causes hydrogen to adhere to the cylinder walls, while overly late injection leads to uneven mixing.

Hydrogen injection pressure of 8 MPa: This pressure ensures appropriate penetration depth and atomization of hydrogen within the cylinder. An excessively low pressure results in insufficient penetration and non-uniform mixing, while an excessively high pressure increases the cost and complexity of the injection system.

A 0% EGR rate: As the baseline condition, EGR effects are excluded to isolate the influence of the equivalence ratio. The effects of EGR will be investigated in subsequent studies.

The flow velocity and temperature vary at different locations in the engine. Therefore, different regions were refined, and sections from different orientations were selected for analysis. The sectional views in different directions are shown in Figure 5.

In this study, ammonia fuel is introduced into the cylinder through intake port injection. After the intake valve opens, ammonia enters the engine cylinder together with fresh air. To clearly illustrate the in-cylinder motion and mixing process of hydrogen, Figure 6 shows the in-cylinder velocity contours from the entry of ammonia fuel into the cylinder until hydrogen injection from the hydrogen injector.

It can be clearly observed that, when the intake valve opens, the in-cylinder gas velocity directs a small portion of gas to diffuse into the intake port through the intake valve gap. This occurs because, at −355° CA aTDC, the pressure inside the cylinder is higher than that in the intake port, causing a small amount of in-cylinder gas to diffuse back into the intake port through the valve gap. With the periodic and continuous injection of ammonia from the intake port, the direction of in-cylinder gas velocity shifts from the valve gap toward the cylinder interior after −315° CA aTDC. After −90° CA aTDC, high-pressure hydrogen is injected into the cylinder by the hydrogen injector. As the piston moves upward, an obvious upward flow can be seen on the lower-left side of the cylinder wall.

Figure 7 shows the ammonia mass distribution in the intake port cross-section. It can be observed that, at −365° CA aTDC after the intake valve opens, ammonia is mainly distributed on the right side of the cylinder. With continuous injection of ammonia, the in-cylinder ammonia–air mixture gradually becomes uniform, and ammonia has sufficient time to mix with air. The in-cylinder mixing process of ammonia is analyzed in detail below.

Figure 8 illustrates the in-cylinder mixing process of ammonia. Combined with Figure 7, it can be seen that, during the initial intake stage, from −365° CA aTDC to −345° CA aTDC, a larger proportion of ammonia enters the cylinder through the outer valve of the intake port, forming a local ammonia-rich region on the right side of the cylinder. As mentioned above, the intake port temperature is 350 K, and the intake pressure is 1 atm. Under this condition, the density of ammonia is approximately 0.60 kg/m³, while the density of air is about 1.0 kg/m³. Therefore, ammonia is lighter and has lower flow inertia in the mixture, making it more susceptible to flow paths and the geometric structure of the intake port.

Observing the intake port geometry, the intake path of the inner valve features a downward curved arc. Air, with higher density, tends to continue flowing downward along the curved arc under inertial effects during flow. In contrast, ammonia, with lower density and smaller inertia, is more easily deflected toward the upper part of the flow path in the curved region, thus reducing the proportion of ammonia entering the inner valve. Meanwhile, the geometric structure of the inner intake port increases flow resistance, which reduces the gas flow velocity through the inner valve. With decreased flow velocity, the inertia of denser air weakens, allowing it to more easily fill the entire inner valve path. On the contrary, the intake passage of the outer valve is relatively smooth with a more streamlined flow path, where ammonia and air undergo no obvious deflection or separation. Consequently, the mixture tends to retain its original distribution characteristics, resulting in a higher ammonia proportion being drawn in through the outer valve.

During the intake stroke as the piston moves downward, ammonia diffuses and mixes further inside the cylinder. By −215° CA aTDC, the in-cylinder ammonia distribution becomes relatively uniform. At −85° CA aTDC, high-pressure hydrogen injection causes a local reduction in ammonia mass fraction. During the compression stroke, as the piston continues to move upward, the mixture of ammonia, hydrogen, and air diffuses and mixes further within the cylinder. It can be seen from the figure that, by −15° CA aTDC, just before spark ignition, ammonia achieves a homogeneous state with uniform mixing throughout the cylinder.

Hydrogen injection timing was fixed from −90° CA aTDC to −83° CA aTDC. Figure 9 shows the in-cylinder mixing process of hydrogen. During the compression stroke, the piston moves upward, and the in-cylinder gas turbulence intensity gradually increases. At −90° CA aTDC, the cylinder volume has decreased significantly, and hydrogen is injected into the cylinder at a high speed, generating turbulence and mixing rapidly with the ammonia–air mixture. Before spark ignition at −14° CA aTDC, two hydrogen-rich regions exist in the cylinder.

Since hydrogen has a high laminar burning velocity, a uniform hydrogen distribution would greatly shorten the combustion duration but lead to an excessively high in-cylinder peak pressure. Therefore, the presence of two hydrogen-rich regions can appropriately delay the combustion process to ensure that the in-cylinder pressure remains within a reasonable range.

The above analysis indicates that the in-cylinder mixing process of ammonia and hydrogen has a significant impact on mixture homogeneity. Due to its lower density, ammonia tends to accumulate on the outer valve side of the intake port, while hydrogen, owing to its high injection velocity, forms two localized hydrogen-rich zones within the cylinder. This non-uniform mixture distribution affects the subsequent combustion process and pollutant formation. The following section provides a detailed analysis of the effects of equivalence ratio on engine combustion characteristics.

3.2. Effect of Equivalence Ratio on Engine Combustion Characteristics and Performance

Figure 10a,b show the variations in in-cylinder pressure, heat release rate, and in-cylinder average temperature under different equivalence ratios, respectively. Figure 10c,d present the changes in pressure rise rate and peak pressure at various equivalence ratios.

As shown in Figure 10a, with an increase in the equivalence ratio, the in-cylinder pressure and heat release rate first increase and then decrease. Meanwhile, Figure 10b indicates that the in-cylinder average temperature exhibits the same trend. This is because a higher equivalence ratio corresponds to a larger amount of ammonia–hydrogen fuel, which theoretically releases more total energy. Under lean-burn conditions, sufficient air is available to support mixture combustion, so the in-cylinder pressure and heat release rate continuously increase, which also leads to a continuous rise in the in-cylinder average temperature.

When the equivalence ratio reaches 1.05, the fuel–air ratio is closest to the theoretical requirement for complete combustion. Under this condition, the ammonia–hydrogen mixture burns at the fastest rate, thus producing the highest heat release. When the equivalence ratio is further increased to 1.48, the in-cylinder ammonia–hydrogen mixture concentration is much higher than that of air. Although the fuel quantity increases, insufficient air causes partial fuel to fail in complete combustion, resulting in reduced combustion efficiency.

Notably, at an equivalence ratio of 1.05, the in-cylinder instantaneous heat release rate rises sharply at 8.3° CA aTDC, which further causes the in-cylinder pressure to reach 21.9 MPa at 9.8° CA aTDC. However, excessively high in-cylinder pressure is detrimental to the stable operation of the engine. Combined with Figure 9 and the following Figure 13, the cause can be concluded: at 8° CA aTDC, the hydrogen-rich region far from the spark plug reaches autoignition conditions due to high temperature and pressure, and the rapid combustion of the mixture leads to a sharp increase in the heat release rate.

It can be seen from Figure 10c,d that, when the equivalence ratio exceeds 0.62, the in-cylinder pressure rise rate and peak pressure become relatively high, which is unfavorable for the smooth operation of the engine. Therefore, equivalence ratio parameters corresponding to high pressure will not be adopted in subsequent studies.

In this study, spark ignition timing is denoted by SIT. CA10, CA50 and CA90 represent the crank angles at which the cumulative heat release reaches 10%, 50% and 90% of the total, respectively. The flame development period is defined as the duration from SIT to CA10, the flame propagation period as that from CA10 to CA50, the flame acceleration period as that from CA50 to CA90, and the total combustion duration as that from SIT to CA90.

Figure 11a,b show the combustion phases and the corresponding crank angles of CA10, CA50 and CA90 under different equivalence ratios, respectively.

As shown in Figure 11a, under lean-burn conditions, the combustion duration first increases and then decreases with the increase in the equivalence ratio. When the equivalence ratio reaches 1.05, the combustion duration is the shortest, only 25.5° CA. As the equivalence ratio further increases to 1.48, the combustion duration slightly prolongs to 34.5° CA. This is because the fuel quantity in the cylinder is higher than the air supply, resulting in incomplete combustion and a longer combustion duration.

It can be clearly observed from Figure 11b that, with the increase in the equivalence ratio, the combustion duration, ignition delay period and heat release center all show a trend of rising first, falling, and then rising again. Under the equivalence ratio of 1.05, the crank angle corresponding to CA50 is the smallest, which is 5.51° CA.

OH radicals exist at the flame front and serve as a key indicator of fuel energy release. In ammonia–hydrogen dual-fuel engines, OH radicals dissociated from hydrogen combustion promote the decomposition of NH₃. During ammonia oxidation, the first step is typically the dehydrogenation of NH₃ by OH or H radicals to form NH₂ or NH. Most intermediates are then converted to N₂H₂, which further reacts with O₂ via NNH to produce the complete combustion product N₂. Meanwhile, Zhu et al. observed in their research that the distribution of OH radicals is highly consistent with the temperature distribution: high temperatures favor the formation of OH radicals and thus promote flame propagation.

The laminar flame speed of pure ammonia is only about 0.15 m/s, whereas that of hydrogen can reach up to 2.7 m/s. Under the 20% hydrogen energy share condition in this study, the laminar flame speed of the mixture increases to approximately 0.45 m/s, representing an improvement of over 200% compared to pure ammonia. This significantly shortens the combustion duration, shifts the combustion phasing closer to top dead center (TDC), and thereby improves thermal efficiency. As can be seen from the OH radical distribution in Figure 12, hydrogen blending results in a more distinct flame front with a faster propagation speed, enabling the flame to cover the entire combustion chamber in a shorter period of time.

Figure 12 shows the effect of different equivalence ratios on the in-cylinder OH radical distribution. It can be clearly seen that, after spark ignition, OH radicals are distributed spherically around the spark plug. As the flame front advances toward the unburned region of the combustion chamber, OH radicals begin to diffuse toward both sides of the chamber, and their concentration increases, reaching a maximum at 15° CA aTDC. OH radicals are more concentrated near the spark plug, with the highest density in the center of the cylinder and a gradual radial decrease.

With increasing equivalence ratio, the in-cylinder OH radical concentration rises gradually and reaches its maximum at an equivalence ratio of 0.89. Notably, at equivalence ratios of 0.5 and 1.48, the OH radical concentration is relatively low. As shown previously in Figure 4, both excessively low and excessively high equivalence ratios are unfavorable for flame propagation, resulting in slow laminar flame speeds.

OH radicals mainly exist in the flame front where the temperature exceeds 1500 K. At an equivalence ratio of 0.5, excess air dilutes the mixture and absorbs combustion heat, lowering the overall in-cylinder temperature and suppressing OH formation. At an equivalence ratio of 1.48, insufficient oxygen leads to incomplete combustion, creating a fuel-rich oxygen-deficient reducing atmosphere that inhibits OH generation.

Figure 13 shows the in-cylinder cross-sectional temperature distribution of the engine at crank angles ranging from −10° CA aTDC to 25° CA aTDC. The spark ignition timing is fixed at −15° CA aTDC. It can be clearly observed that, under different equivalence ratios, high-temperature regions basically appear near the cylinder head side close to the spark plug. This is because, after spark ignition, the fuel–air mixture around the spark plug is ignited rapidly, forming a high-temperature burned zone. As the piston moves upward, the high-temperature region gradually expands toward both sides and eventually spreads throughout the combustion chamber.

At 5° CA aTDC, the area of the high-temperature burned zone in the cylinder gradually increases with increasing equivalence ratio. As can be seen from Figure 11, a larger equivalence ratio leads to an earlier heat release center, shorter combustion duration, and faster combustion speed of the in-cylinder mixture, which explains the above phenomenon. Meanwhile, it can be intuitively observed from the slice contours that the combustion of the fuel–air mixture in the chamber is poor at an equivalence ratio of 0.5. When the equivalence ratio increases to 0.62, the in-cylinder temperature rises significantly, and the flame front approaches the cylinder wall at 25° CA aTDC. When the equivalence ratio rises to 1.05, the flame front reaches near the cylinder wall as early as 5° CA aTDC. As the equivalence ratio further increases to 1.48, the flame front approaches the cylinder wall at 15° CA aTDC.

This further indicates that, when the equivalence ratio is close to the stoichiometric ratio, the high-temperature combustion zone expands faster, combustion intensity is enhanced, and flame stability is improved. However, the rapid and high in-cylinder pressure rise tends to induce engine knock.

This study further investigated the effects of equivalence ratio on engine performance, including combustion efficiency (CE), indicated thermal efficiency (ITE), and indicated mean effective pressure (IMEP). The combustion efficiency, indicated thermal efficiency, and IMEP under different equivalence ratios are shown in Figure 14.

As shown in Figure 14a, with an increase in the equivalence ratio, both combustion efficiency and indicated thermal efficiency first increase and then decrease. The combustion efficiency reaches its maximum of 95.18% at an equivalence ratio of 0.74. Notably, when the equivalence ratio increases to 1.48, the combustion efficiency is only 58.96%, while the indicated thermal efficiency is as high as 49.1%. The low combustion efficiency is attributed to the overly rich fuel mixture and suppressed ammonia cracking, whereas the high indicated thermal efficiency is driven by the rapid combustion characteristics of hydrogen, reduced heat loss, and optimized thermodynamic cycle.

Figure 14b shows that the indicated mean effective pressure first rises and then falls with increasing equivalence ratio, reaching a maximum of 2.34 MPa at an equivalence ratio of 1.05.

It is worth noting that, at equivalence ratios of 0.5 and 0.55, the combustion efficiency, indicated thermal efficiency, and IMEP are all relatively low. When the equivalence ratio increases to 0.62, the combustion efficiency rises rapidly to 90.84%, indicating that the mixture in the combustion chamber burns more completely under this condition. However, as the equivalence ratio continues to increase to 0.89, the combustion efficiency begins to decrease. As shown previously in Figure 10a,b, the further increase in equivalence ratio leads to a higher in-cylinder average temperature and shorter combustion duration. The concentrated heat release increases heat transfer loss to the cylinder wall, so the heat cannot be effectively converted into mechanical work, resulting in a decline in both combustion efficiency and indicated thermal efficiency. In contrast, the IMEP continues to increase because, despite the lower energy conversion efficiency, the increased total fuel quantity still drives IMEP to a higher level.

3.3. Effects of Equivalence Ratio on Engine Emission Characteristics

In the research regarding the development and application of ammonia–hydrogen dual-fuel engines, nitrogen-based pollutants mainly consist of NO_x (including NO, NO₂, N₂O, etc.) and unburned ammonia. NO is the main component of NO_x emissions, which can cause acid rain and ozone formation, posing direct harm to the environment and human beings. Therefore, NO is the focus of research on nitrogen-based pollutants. Although N₂O has the smallest emission magnitude, it is a potent greenhouse gas that destroys the ozone layer and causes long-term harm to Earth’s ecosystem, requiring focused research in the field of clean energy. NO₂ is a highly oxidizing gas that is directly toxic to the human respiratory system. Meanwhile, as an important precursor of photochemical smog, ozone and PM2.5, it affects air quality and also requires key research.

Figure 15b shows the emission characteristics of NO, N₂O, and NO₂ under different equivalence ratios. It is noteworthy that, although unburned NH₃ emissions are nearly zero at Φ = 1.05, NO emissions remain at a relatively high level (approximately 15 g/kW·h). This is because the in-cylinder average temperature exceeds 2200 K under this condition, leading to a significant increase in thermal NO formation. This indicates that Φ = 1.05 represents a typical performance–emission trade-off operating condition: the highest IMEP (2.34 MPa) and the lowest unburned NH₃ emissions (0.053 g/kW·h) are achieved, with a combustion efficiency as high as 93.1%. NO emissions are relatively high, necessitating a selective catalytic reduction (SCR) aftertreatment system to meet emission regulation requirements.

In contrast, although the combustion efficiency at Φ = 0.62 is slightly lower (90.8%), the NO emissions are only 2.1 g/kW·h and N₂O emissions are 1.2 g/kW·h, resulting in superior overall emission performance. This indicates that, in practical applications, the appropriate equivalence ratio operating range should be selected based on specific emission regulations and performance requirements.

It can be seen from Figure 15b that NO emission is low in the equivalence ratio range of 0.5 to 0.62 because the combustion efficiency of the in-cylinder mixture is low and the average in-cylinder temperature is insufficient to support the formation of thermal NO and fuel NO. When the equivalence ratio increases to 0.74, NO emission reaches the maximum. As shown previously in Figure 10a,b, at an equivalence ratio of 0.74, the combustion efficiency is the highest (95.18%) and the average in-cylinder temperature generally exceeds 1800 K. According to the Zeldovich mechanism, a large amount of thermal NO is produced under such conditions, thus increasing NO emission [32]. When the equivalence ratio increases to 1.48, NO emission is nearly zero. This is because, under fuel-rich conditions, the combustion efficiency is only 55.2% and the average temperature decreases, which significantly reduces the formation of fuel NO and thermal NO.

This non-monotonic evolution of nitrogen-based pollutants in our ammonia–hydrogen engine exhibits strong qualitative consistency with the multi-species emission behaviors reported in broader advanced alternative dual-fuel combustion systems [33]. As systematically demonstrated, the local equivalence ratio and corresponding flame temperature command the shifting boundaries between fuel–NO_x reactions and thermal Zeldovich mechanisms. Specifically, at lower combustion temperatures typical of lean-burn regions (below 1400 K), the interaction between amine-intermediate radicals and NO₂ serves as a major chemical contributor to N₂O accumulation, while the abundance of HO₂ radicals actively recycles NO into NO₂. When the equivalence ratio approaches closer to the stoichiometric point (Φ = 1.05), the drastically intensified radical pool (H, O, and OH) combined with elevated peak thermal fields accelerates the consumption and self-reduction pathways of both N₂O and NO₂, thereby forcing their final concentration profiles downwards. This cross-species kinetic balancing perfectly supports the validity of our 3D CFD simulation results.

Meanwhile, it can be observed that N₂O is the main pollutant with the highest emission at equivalence ratios of 0.5 to 0.62. With an increase in equivalence ratio, N₂O emission gradually decreases and becomes nearly zero when the equivalence ratio exceeds 0.62. This is because N₂O is mainly formed under high-pressure and medium–low-temperature conditions (between 800 and 1200 K). NO₂ emission also shows a gradual decreasing trend with an increase in equivalence ratio. It can be seen from the figure that NO₂ emission is the highest at an equivalence ratio of 0.5, indicating that low combustion efficiency under lean-burn conditions promotes the formation of NO₂.

Figure 16 shows the evolution of the in-cylinder cross-sectional NO mole fraction under different equivalence ratios. It can be seen from the figure that, after spark ignition, at −10° CA aTDC, the ammonia–hydrogen mixture on one side of the spark plug is ignited rapidly, and a large amount of thermal NO and fuel NO are generated quickly in the high-temperature burned zone. As combustion proceeds continuously, NO gradually diffuses to the low-temperature regions on both sides of the combustion chamber. This is because the flame front gradually spreads to the unburned zone, and important intermediates such as HNO, NH and N are mainly formed at the flame front, which favors the formation of fuel NO. Meanwhile, it can be observed that, with the continuous increase in the equivalence ratio, NO forms faster at the same crank angle. This is because the average in-cylinder temperature rises gradually and the mass of the ammonia–hydrogen mixture increases with the rising equivalence ratio, which further intensifies the formation of fuel NO and thermal NO.

Notably, at an equivalence ratio of 0.5, low combustion efficiency results in a low average in-cylinder temperature, and a large amount of ammonia–hydrogen mixture is not fully burned, thus inhibiting NO formation. At 40° CA aTDC, it can be clearly seen that NO appears mainly in stripes on both sides of the combustion chamber and tends to diffuse outward, where NO exists mainly at the flame front. When the equivalence ratio is increased to 1.05, the in-cylinder NO mass fraction shows a downward trend at 20° CA aTDC and decreases substantially at 40° CA aTDC. This is mainly because the in-cylinder temperature is the highest and the combustion duration is the shortest under this operating condition, so a large amount of ammonia–hydrogen mixture is burned out rapidly. As the piston moves downward, the in-cylinder temperature drops quickly, and a large amount of NO is reduced to N₂ and H₂O through reactions with N, O and H radicals.

It is worth noting that, at an equivalence ratio of 1.48, insufficient oxygen and a low average in-cylinder temperature lead to incomplete combustion of a large amount of mixture, which restricts the formation of thermal NO.

Figure 17 shows the evolution of the in-cylinder cross-sectional N₂O mole fraction under different equivalence ratios. It can be clearly observed from the figure that the in-cylinder N₂O mass shows a decreasing trend with the increase in the equivalence ratio. The N₂O mole fraction is the highest under the equivalence ratio range of 0.5–0.62. This is because the average in-cylinder temperature is relatively low under low equivalence ratios, and N₂O is mainly formed under high-pressure and medium–low-temperature conditions (800–1200 K). As shown in Equation 10 above, NH reacts with NO to form N₂O. This also explains why the distribution of N₂O appears in a ring shape diffusing to both sides. NO formed at the flame front diffuses toward the low-temperature unburned zone along with the flame front, and the temperature in the burned zone decreases to meet the formation conditions of N₂O. Meanwhile, a large amount of N₂O is produced during the low-temperature oxidation of NH₃ via the reaction NH₃ + O₂ → N₂O + H₂O.

When the equivalence ratio is increased to 0.74–1.05, the higher combustion efficiency promotes the rise of in-cylinder temperature. High temperature is unfavorable to the formation of N₂O, and N₂O is oxidized and reduced to NO at high temperatures, which further promotes the formation of NO. When the equivalence ratio is increased to 1.48, insufficient oxygen in the cylinder and an average temperature above 1500 K inhibit the formation of N₂O.

Figure 18 shows the evolution of the in-cylinder cross-sectional NO₂ mole fraction at different equivalence ratios. It can be clearly observed that NO₂ is mainly distributed in the low-temperature regions on both sides of the combustion chamber under low equivalence ratios. This is because NO₂ is mainly formed in low-temperature regions below 1000 K, and important precursors for NO₂ formation are NO and O atoms. Moreover, NO₂ tends to decompose into NO and O₂ at high temperatures, so the formation conditions of NO₂ are relatively stringent. Therefore, under low equivalence ratios, improving combustion efficiency can reduce NO₂ formation.

It is worth noting that, when the equivalence ratio exceeds 0.74, NO₂ emission is nearly zero due to the high in-cylinder temperature. The formation of NO₂ is the lowest under fuel-rich conditions.

4. Conclusions

Based on a validated three-dimensional numerical simulation model, this study systematically investigated the effects of equivalence ratio on the combustion characteristics and four major nitrogen-based pollutant emissions of an ammonia–hydrogen dual-fuel engine under a fixed 20% hydrogen energy ratio, revealing the pollutant formation mechanisms and their in-cylinder spatiotemporal evolution laws. The main conclusions are as follows:

(1) The equivalence ratio has a significant impact on the combustion characteristics of the ammonia–hydrogen dual-fuel engine. The in-cylinder pressure, heat release rate, and mean temperature all increase first and then decrease with increasing equivalence ratio, reaching their maximum values at Φ = 1.05. The combustion efficiency peaks at Φ = 0.74 (95.18%), the indicated thermal efficiency peaks at Φ = 1.48 (49.1%), and the IMEP reaches its maximum value at Φ = 1.05 (2.34 MPa).

(2) Unburned NH₃ emissions decrease first and then increase with increasing equivalence ratio, reaching a minimum at Φ = 1.05 (0.053 g/kW·h). At Φ < 0.74, the in-cylinder temperature is below 1800 K, resulting in a slow ammonia decomposition rate and a sharp rise in unburned ammonia emissions. At Φ = 1.48, insufficient oxygen leads to incomplete ammonia oxidation, causing unburned ammonia emissions to rise again to 2.13 g/kW·h.

(3) NO emissions increase first and then decrease with increasing equivalence ratio, peaking at Φ = 0.74 (approximately 20 g/kW·h). This is because the combustion efficiency is highest at this condition, with the in-cylinder mean temperature exceeding 1800 K, leading to substantial thermal NO formation. At Φ = 1.48, the low oxygen concentration and reduced temperature under fuel-rich conditions cause NO emissions to approach zero. NO is primarily generated in the high-temperature zone at the flame front. At Φ = 1.05, the rapid temperature drop during late combustion reduces part of the NO back to N₂.

(4) N₂O emissions are mainly concentrated in the low-equivalence-ratio range (Φ = 0.5–0.62), reaching a maximum at Φ = 0.5 (approximately 4.5 g/kW·h). N₂O is primarily formed through the reaction between NH and NO in the medium-to-low-temperature range of 800–1200 K. When the in-cylinder temperature exceeds 1500 K, N₂O decomposes rapidly. Therefore, at Φ > 0.62, N₂O emissions drop rapidly to near zero.

(5) NO₂ emissions decrease monotonically with increasing equivalence ratio, reaching a maximum at Φ = 0.5 (approximately 2.5 g/kW·h) and becoming nearly zero at Φ > 0.74. NO₂ is mainly formed by NO oxidation in the low-temperature zone near the combustion chamber walls (<1000 K), while high temperatures cause NO₂ decomposition.

This study reveals the synergistic regulation mechanism of equivalence ratio on nitrogen-based pollutant emissions in ammonia–hydrogen dual-fuel engines, providing an important theoretical basis for combustion system design and emission-control strategy development. Future research will further explore the effects of hydrogen blending ratio and ignition timing on engine performance, as well as multi-parameter collaborative optimization methods.