The Influence of Hydrogen Addition on a SI Engine—Theoretical and Experimental Investigations

Abstract

In a world with severe pollution regulations and restrictions imposed to internal combustion engines, improving efficiency and reducing pollutant emissions and greenhouse gases are important goals for researchers. A highly effective method to achieve the premises written above is to use alternative fuels, which may have a strong influence on combustion processes in spark ignition engines. In order to increase the heat release rate during combustion, the brake thermal efficiency, and to decrease the levels of pollutant emissions and greenhouse gases, the use of sustainable alternative fuels, in parallel with conventional fuels is a great choice. Among alternative fuels, hydrogen is an excellent fuel in terms of its physical-chemical properties, making it an attractive replacement for classic fuels in the combustion process. This article demonstrates AMESim 13.0.0/Rev13 theoretical and experimental investigations conducted on a supercharged spark ignition engine at 55% engine load and 2500 rpm speed, analyzes the effect of 2.15% hydrogen that substitutes gasoline on combustion, implicitly investigates energy and fuel efficiency of the engine and investigates pollutant and greenhouse gas emission levels. These experimental investigations confirm the theoretical study of thermo-gas-dynamic processes of a SI engine fueled with gasoline and hydrogen, and it shows the importance of engine tunings and hydrogen quantity on engine operation. The obtained results indicate the advantages of fueling the engine with both gasoline and hydrogen: the increase of the heat release rate which leads to the increase of maximum pressure and maximum pressure rise rate during combustion, the increase of the brake thermal efficiency, the decrease of the combustion duration, the decrease of the brake specific energetic consumption by 4.8%, the decrease of the levels of pollutant emissions by 11.11% for unburned hydrocarbons HC, by 12.5% for monoxide carbon CO, by 63.23% for nitrogen oxides NO_x, and by 33.7% for carbon dioxide CO₂ as a greenhouse gas. Further research directions can be developed from this research for other operating regimes and other hydrogen quantities.

1. Introduction

Studies

As a part of the community that is concerned about the health of city inhabitants and a clean environment, we have to find adequate solutions in order to make spark ignition engines more friendly to the environment. According to the International Energy Agency, the transport sector generated 254 Mt of CO₂ more in 2022 than in 2021 [1]. Many studies are focused on reducing pollutant emissions, especially in urban areas. Fully electric powertrain systems for automotive use do not seem to be the solution for achieving the European Union’s target of zero greenhouse gas emissions by 2050 [2] because of the cost of batteries, lack of infrastructure, and relatively low autonomy. Furthermore, it is very expensive to recycle the batteries required, and the autonomy decreases significantly over the years and in cold weather. Using low carbon fuels such as methanol (CH₃OH), methane (CH₄), or fuels that do not contain carbon molecules, such as hydrogen (H₂), ammonia (NH₃), or synthetic fuels, may offer a viable means of reducing CO₂ levels without imposing constrains on the utilization of internal combustion engines. Among these fuels, hydrogen stands out due to its physical-chemical properties that make it an attractive alternative to replace classic fuels in the combustion process.

In

Table 1. Hydrogen, gasoline, and diesel fuel properties [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25].

Characteristic	Hydrogen	Gasoline	Diesel Fuel
Chemical formula	H2	C8H18	C16H34
Mass Composition c [%]/h [%]/o [%]	0/100/0	85.4/14.2/0.4	85.7/13.3/1
Minimum theoretical amount of oxygen necessary for stoichiometric combustion [kmol/kg fuel]	0.2500	0.1065	0.1043
Minimum theoretical amount of air necessary for stoichiometric combustion [kmol/kg fuel]	1.1900	0.5073	0.4966
Molar mass [kg/kmol]	114	2.016	224
Stoichiometric air fuel ratio [kg air/kg fuel]	34.4	14.7	14.6
Diffusion coefficient [cm2/s]	0.63	0.05	-
Density [kg/m3] at 0 °C, 760 mmHg	0.09	740–760	830–870
Lower heating value [kJ/kg]	119,617	42,690	41,800
Minimum ignition energy [mJ]	0.02	0.2–0.3	-
Autoignition temperature [K]	845	740–810	473–493
Laminar flame speed [cm/s] at λ = 1	258	35	125
Flammability limits λs/λi [-] at p0 = 1 × 105 Pa; T0 = 293 K	0.394/10	0.7/1.15	0.34/1.68
Flammability limits inferior/superior [%vol.]	4.10/75.6	1.5/2.4	-
Stoichiometric dosage [%volume]	29.58	1.73	-
Adiabatic flame temperature [K], at λ = 1	2300	2411	2327
Research Octane Number	>130	90–98	-

, the properties of hydrogen are presented. For ease of understanding, the physical-chemical properties of hydrogen are presented in parallel with the properties of classic fuels used mainly in internal combustion engines.

According to studies by other researchers, hydrogen has the potential to change the world in terms of pollution. Hydrogen’s properties show that it can be used as a sustainable green fuel due to possibility of a reduction in pollution levels during combustion processes. A study by Ling-Chin [4] presents an overview of the technologies and challenges faced when using hydrogen in different transportation modes, emphasizing the differences and resemblances in storage technologies and refueling, supply chain, use in internal combustion engines and fuel cells. In all transportation modes, using hydrogen demonstrated great potential, nonetheless each sector encounters different obstacles and will need enhancements in performance, cost, infrastructure regulations, and standards and supportive policies to stimulate hydrogen utilization. A significant amount of studies are focused on using hydrogen as a fuel to produce electricity in fuel cells, but because of the high cost of fuel cells, the price of hydrogen (H₂ used in fuel cells has to have a 99.97% purity or higher), and the lack of infrastructure, this solution has not gained popularity. Another use of hydrogen is as fuel in internal combustion engines, which is a relatively simpler and cheaper solution and it supports the automotive industry that produces powertrain parts. Hydrogen properties like high research octane number, wider flammability limits, higher lower heating value, and high combustion speed defines hydrogen as a viable alternative fuel for spark ignition engines. Because of its low ignition energy of 0.02 mJ, the internal combustion engine can operate at ultra-lean mixtures with stable combustion and reliable performance that leads to fuel economy [5,6,7]. Due to its high laminar burning speed, hydrogen use can lead to the reduction of the combustion duration and cyclic variability. An important issue presented with hydrogen use in spark ignition engines is the possibility of the occurrence of abnormal combustion phenomena such as pre-ignition and backfire. Depending on the chosen hydrogen fueling method, manifold injection, port injection, or direct injection, different methods of forming the air-fuel mixture and different controls for the combustion process can be achieved. A more efficient combustion control can be ensured by applying the hydrogen direct injection (HDI) method. This method was used by Park [8] on a 2-L spark ignition engine to eliminate backfire and to adjust the H₂ dose and timing. With hydrogen use in the engine, performance and pollutant emission levels were improved. With hydrogen direct injection use, the engine torque increased by 35.9% [8]. The brake thermal efficiency (BTE) increased from 34% to 40% when the hydrogen flow rate increased from 1.55 kg/h to 1.95 kg/h, and the fuel injection timing was reduced from 195° BTDC to 60° BTDC. In lean mixtures with the coefficient of excess air at λ = 2.05, the NO_x emission level was 3000 ppm and CO₂ emission level was 500 ppm or higher.

An experimental investigation developed by Niu [9] on a 1.8-L natural aspirated spark ignition engine fueled by hydrogen (direct injection) and gasoline (port fuel injection) showed a decrease of the initial phase of combustion and of the total duration of combustion at the increase of the hydrogen quantity. Niu [9] studied the regime of 1500 rpm and 30% throttle position at different coefficients of excess air (1, 1.2, 1.5, and 1.8), and observed higher in-cylinder pressure, heat release rate, and thermal efficiency with hydrogen use. The HC and CO emission levels were reduced by hydrogen fueling, but the NO_x emission level was increased with the increase of H₂. Shi [10] also specified an increase of NOx emission levels when the cyclic dose of hydrogen increased. Additionally, Ozyalcin [11] tried to reduce the NO_x emission level by using two types of selective catalytic reductions (SCR) made out of copper-zeolite or were vanadium based. In another study, Chitragar [12] used hydrogen additions in fractions from 0% to 10%, with steps of 2% by volume to fuel a 4-cylinder indirect injection SI engine fueled initially with gasoline. During experimental investigation, the same ignition timing was used (5° before TDC) for all speeds (2000 rpm, 2500 rpm, 3000 rpm, and 3500 rpm) with a wide open throttle. The addition of 8% hydrogen increased the brake power of the engine by 11.8%, the brake thermal efficiency (BTE) by 8%, and thermal efficiency by 5.78%. Higher hydrogen percentages led to the decrease of these engine performance parameters. Compared to gasoline, the CO emission level decreased by 16.33% and the HC emission level decreased by 13.48%, but NO_x emission level increased by 16.8% with a 10% addition of hydrogen at 3500 rpm speed. Others researchers obtained similar results [13,14,15,16,17]. Elsemary [18] studied the impact of ignition timing on a single cylinder carbureted engine fueled with gasoline and 24%, 28%, 29%, 31%, and 49% vol. hydrogen at a 3000 rpm speed regime. The ignition timing was set up at 30°, 35°, and 40° BTDC. Brake thermal efficiency (BTE) increased with the H₂ quantity when the spark timing was near to TDC (30° BTDC) except for when the percentage of hydrogen reached 49%. Brake specific fuel consumption (BSFC) decreased at the increase of hydrogen share, except in the 49% hydrogen case. At the same gasoline-hydrogen proportion, the HC and CO emissions levels were increased when ignition timing was increased (i.e., 40° BTDC). According to Molina [19] hydrogen direct injection (HDI) seemed to outperform port fuel injection (PFI) systems on a single cylinder SI engine for both fueling methods when the spark timing was optimized. Molina [19] showed the increase of the IMEP by 4% with hydrogen direct injection use. Molina [19] calculated a value of 41.2% for the thermal efficiency at HDI use versus a value of 40% for thermal efficiency at PFI use, at a similar coefficient of excess air λ = 3.2, and observed a lower tendency for abnormal combustion, a reduced cycle-to-cycle variability, and a lower NO_x emission level at high engine loads. Kim [20] conducted an experiment on a 4 cylinder engine fueled by hydrogen trying to increase the engine power by using two turbochargers types. The results showed that optimizing the compressor’s pressure ratio involved aligning the system for maximum efficiency and fine-tuning the exhaust gas flow rate when the engine is using hydrogen fueling and lean mixtures [20]. For the variable-geometry turbocharger, the engine effective power was increased by 1.65 times compared to the power performance of the naturally aspirated engine at 5000 rpm speed. Sometimes, depending on the technical condition of the engine, the use of hydrogen as a single fuel cannot entirely decrease the HC or CO₂ emission level if the oil reaches the combustion chamber where it produces polluting emissions [8,21]. Even in this case, the HC and CO₂ emissions levels were lower when using hydrogen than when using classic fuel [8,21]. In his study, Machacek [22] presented the possibility to decrease the NO_x emission level by integration of electric hybridization into an engine fueled by hydrogen which was mounted in a passenger car. In some operational cases, the NO_x emission level decreased by more than 90%, as well as the hydrogen consumption decreased by 16% compared to the standard engine fueled by hydrogen. Purayil [23] and Teoh [24] studied the possibilities of hydrogen use as fuel for internal combustion engines. Dual-fueling with gasoline and hydrogen offers flexibility and can increase engine performance, especially if are integrated the advantages of each type of fueling method: direct injection or port fuel injection [23,24]. Compared to hydrogen, valve port injection and the use of hydrogen direct injection can reduce abnormal combustion phenomena like knock, preignition, or backfire, but the NO_x emission level and cyclic variability are increased [23,24]. Thermal efficiency and engine power are increased when low percentages of H₂ at partial loads are used [23,24]. In lean mixtures and hydrogen use, the CO and HC emissions levels are decreased, but the NO_x emission level decreased only with the use of the exhaust gas recirculation (EGR) [23,24].

All these efforts to reduce the pollutant emissions and enhance engine performance with alternative fuels must be correlated with environmentally friendly ways to produce hydrogen [25]. Water electrolysis and photochemical water-splitting are sustainable methods to produce hydrogen. Many researchers proposed different technical solutions to achieve maximum electrolysis efficiency and best current density, such as using a 1–2 mm spacing between the electrodes, using low carbon steel electrodes, and high temperature alkaline electrolysis [26,27,28]. In the photochemical water-splitting method, the energy from the sun is used with special semiconductors constructed from photochemical materials to generate hydrogen and oxygen. This technology has the advantage of using seawater to produce hydrogen, but has the disadvantages of limited efficiency and high production cost [29,30,31,32]. Pana [33] developed a thermodynamic model to study the kinetics of hydrogen combustion in a spark ignition engine. The effects of hydrogen addition on combustion kinetics was explained through the results of the numerical model that showed the increase of the maximum pressure and rate of pressure rise. Knock or other abnormal combustion phenomena were avoided by hydrogen direct injection before inlet valve closure by tuning the H₂ quantity and H₂ injection timing. Negurescu [34] studied the qualitative load adjustment solution at a hydrogen direct injection engine to avoid abnormal combustion of air-hydrogen mixture. Theoretical research by numerical simulation were confirmed by the experimental results for different H₂ injection timings looking at combustion efficiency. Compared to gasoline fueling, hydrogen use in combustion kinetics is improved by hydrogen wide flammability limits, with an eight times higher flame speed and a ten times lower of ignition minimum energy. Negurescu [35] designed an direct injection SI engine and showed that the use of an ultra-lean mixture, λ > 3.2, and a load qualitative adjustment at a direct injection spark ignition engine leads to very low NO_x emission levels, close to zero. At stoichiometric excess, the air coefficient, λ = 1, the maximum pressure, the maximum pressure rise rate, IMEP, engine effective power, ISFC, and BSFC are superior with hydrogen use versus gasoline fueling. Other researchers studied hydrogen combustion on different thermal engines or thermal machines in order to evaluate the influence of hydrogen quantity on combustion performance. For example, Cernat [36] showed that hydrogen use influences the vaporization and combustion speed of fuel drops in a diesel engine. with hydrogen use, the mixture air-fuel is rapidly formed due to the increase of the vaporization speed of diesel fuel in the air-hydrogen mixture. The combustion kinetics showed that in air-hydrogen mixtures the diesel fuel drops have a lower combustion duration and develop a larger flame radius in flame structure due to the air-hydrogen mixture having wider flammability limits, higher LHV, and higher combustion speed [36]. Hosseini [37] studied hydrogen combustion in a diesel engine and demonstrated the reduction of the carbon based emissions, brutal combustion, and higher NO_x emission levels. Salzano [38] used a burner to study the influence of H₂ in a mixture with air at ambient temperature on flame speed. Compared to methane or propane, the influence of hydrogen was analyzed for percentages up to 20%. Eckart [39] showed that carbon dioxide addition reduces the hydrogen flame and the thermal diffusivity in a burner, the heat transfer reduced flame propagation, and lower Markstein numbers were achieved. Zitouni [40] studied combustion in a constant volume vessel and showed the influence of hydrogen and methane addition on combustion of ammonia blends. Hydrogen addition led to the increase of the combustion speed of the mixture and of the Markstein length [40]. Konnov [41] showed that the maximum laminar combustion speed of hydrogen appears at an equivalence ratio of 1.8 in rich mixture areas. Pio [42] studied different laminar combustion speeds determined by various researchers in different thermal machines and concluded that closed vessel microgravity offers the highest speeds, around 8 cm/s, depending on the equivalent ratio. Pio [42] used different mixtures of air-ammonia and additions of H₂ or CH₄ fuel in his study and showed that compared to methane, hydrogen use led to superior laminar combustion speeds that can exceed 240 cm/s versus 40 cm/s at maximum ammonia percent.

This paper aims to observe the effects of fueling a turbocharged spark ignition engine with hydrogen on energetic performance, combustion, and on pollutant emissions and greenhouse gas levels. The experimental study was completed by a theoretical study based on a numerical simulation in AMESim in similar operational conditions [43].

2. Experimental and Theoretical Investigations

2.1. Experimental Investigations

The research studies were conducted on a A15MF-1.5 DOHC engine that was used on DAEWOO Nubira, Cielo, and Espero automotive models. Figure 1 presents the scheme of the experimental test bench.

The engine was equipped with a turbocharger that provided the blower with a 1.5 bar supercharging pressure. Hydrogen was injected into the inlet valves ports by using hydrogen fueling injectors, which were actuated from the electronic control unit of the engine. By using hydrogen PFI, the available volume for the air in the inlet manifold was at its maximum. The electronic control unit also actuated the gasoline injectors as it is an open type of ECU. By using the Dastek Unichip software the opening duration of the fuel injectors could be modified. For hydrogen fueling, the opening duration of the gasoline injectors was reduced and the opening duration of the hydrogen injectors was increased until the value of the engine power or the value of the coefficient of excess air was restored to the value allocated for gasoline fueling. At a constant throttle position and a constant spark ignition timing, the hydrogen injector’s opening duration was set up so that at a specific H₂ cyclic dose the engine thermal efficiency would be maximized. At the same time, it was ensured that the in-cylinder maximum pressure did not increase excessively with the amount of hydrogen used. First of all, before the measurements, all equipment and measuring devices were calibrated. During experimental investigation, operational parameters like gasoline and hydrogen consumption, inlet air consumption, the level pollutant emissions and greenhouse gas CO₂, supercharging the pressure, engine tuning parameters, in-cylinder pressure diagrams and associated parameters like heat release rate were examined. The reference was determined by fueling the engine with gasoline only, next the gasoline was partially substituted with hydrogen with the purpose of maintaining the same effective power as gasoline fueling, at the same coefficient of excess air, λ = 1. During the experimental session, the percentage of hydrogen used to replace gasoline was set at the value at which the thermal efficiency is maximized for an unchanged spark ignition timing, in conditions of stoichiometric excess, an air coefficient, λ = 1, for the operating regime of 2500 rpm and a 55% engine load. Thus, a novelty aspect was establishing a value for the substitution percentage of gasoline with hydrogen (xc = 2.15%) which, at an unchanged spark timing, ensured a maximum effective yield, minimum brake specific energy consumption, for a stoichiometric air-hydrogen-gasoline mixture, λ = 1, at speed 2500 rpm and 55% engine load. The novelty of the paper is represented by the optimal tuning between engine speed, engine load, hydrogen cyclic dose, gasoline cyclic dose, supercharging pressure, exhaust gas temperature, and spark timing at a stoichiometric excess air coefficient with the operating regime. Also, the novelty is present in achieving the main objectives of the paper of using low quantities of hydrogen injected into the inlet port through an electronic hydrogen injection system connected back-to-back with the gasoline injection system, in a ready to use system that can be easily adapted to any electronic fuel injection (EFI) engine. The accomplishment of this specific objective lead to the main goal improving the engine’s performance, regarding the thermal efficiency and pollution level, by using small amounts of hydrogen injected into the inlet air through the intake valves ports.

2.2. Theoretical Investigations

The simulation of the engine operation was conducted using the model ENGCCSI10 from the AMESim library which was adapted to match the experimental parameters [43]. The numeric schema of the engine that was designed for modeling is presented in the Figure 2. The AMESim model uses a combustion chamber in which the pressure and volume is variable and takes into consideration the heat transfer to the wall and the losses associated. To simulate the fuel, three parameters were used (x, y, z, and the chemical formula C_xH_yO_z represented this) and the heating value was introduced [43]. The mixture of gases in the combustion chamber (air, fuel, and burnt gases) is considered a perfect gas.

The model used the general hypothesis that the in-cylinder charge is made of a chemically and thermally homogeneous mixture of perfect gases [25]. The calculations were made in iterations, starting with the onset of compression processes and ending with the onset of exhaust processes. Every calculation step was divided in small steps of 1 °CA.

The coefficient of excess air on a cycle is determined by the following formula of cycle mass fuel dose [25]:

$${\mathrm{m}}_{\mathrm{c}\mathrm{y}\mathrm{c}\mathrm{l}\mathrm{e}}={\displaystyle \frac{{\rho }_0·V_{\mathrm{s}}·{\eta }_{\mathrm{v}}}{1+\lambda ·L_{\mathrm{t},\mathrm{a}\mathrm{i}\mathrm{r}}}}$$

(1)

where:

• ρ_o—inlet air density
• V_s—engine cylinder displacement
• η_v—volumetric efficiency
• λ—coefficient of excess air
• L_{t, air}—minimum air quantity necessary for stoichiometric combustion
• m_cycle—fuel mass quantity per cycle determined in kg/cycle.

The equation of equilibrium is given by the following [25]:

$${\displaystyle \frac{\mathrm{d}\mathrm{U}}{\mathrm{d}\mathsf{\alpha }}}={\displaystyle \frac{\mathsf{\delta }\mathrm{Q}}{\mathrm{d}\mathsf{\alpha }}}-{\displaystyle \frac{\mathsf{\delta }\mathrm{L}}{\mathrm{d}\mathsf{\alpha }}}-{\displaystyle \frac{\mathsf{\delta }{\mathrm{Q}}_{\mathrm{w}}}{\mathrm{d}\mathsf{\alpha }}}$$

(2)

where:

• $\mathsf{\delta }\mathrm{Q}—$heat released during the burning process
• $\mathsf{\delta }\mathrm{L}—$mechanical work
• $\mathsf{\delta }{\mathrm{Q}}_{\mathrm{w}}$—heat loss though the cylinder walls
• dU—internal energy

Heat is determined with the burning characteristic equation as follows [25]:

$${\displaystyle \frac{\mathsf{\delta }\mathrm{Q}}{\mathrm{d}\mathsf{\alpha }}}={\mathrm{m}}_{\mathrm{c}\mathrm{y}\mathrm{c}\mathrm{l}\mathrm{e}}·{\mathrm{H}}_{\mathrm{i}}· {\displaystyle \frac{\mathsf{\delta }\mathsf{\xi }}{\mathrm{d}\mathsf{\alpha }}}$$

(3)

where:

• dU—internal energy of the in-cylinder mixture
• m_cycle—cyclic fuel mass
• H_i—lower heating value of the fuel
• ${\displaystyle \frac{\mathsf{\delta }\mathsf{\xi }}{\mathrm{d}\mathsf{\alpha }}}$—the heat release rate during combustion

The heat transferred to the walls is determined with the Woschni relation as follows [25]:

$${\displaystyle \frac{\mathrm{d}{Q}_w}{\mathrm{d}\mathsf{\alpha }}}=K_{conv}·A_{wall}·\left(\mathrm{T}-T_{wall}\right)·{\displaystyle \frac{1}{6·n}}$$

(4)

where:

• K_conv—the heat transfer coefficient by convection from the gas to the wall
• A_wall—the heat exchange surface of the walls
• T—in-cylinder gas temperature
• T_wall—the medium temperature of the wall
• n—engine speed

Temperature variation on each interval is calculated with the following formula [25]:

$${\displaystyle \frac{\mathrm{d}\mathrm{T}}{\mathrm{d}\mathsf{\alpha }}}={\displaystyle \frac{1}{{\mathrm{N}}_{\mathrm{c}}·C_{\mathrm{v}\mathrm{c}}}}·\left({\displaystyle \frac{\mathrm{d}\mathrm{Q}}{\mathrm{d}\mathsf{\alpha }}}-{\displaystyle \frac{\mathrm{d}\mathrm{L}}{\mathrm{d}\mathsf{\alpha }}}-{\displaystyle \frac{\mathrm{d}{\mathrm{Q}}_{\mathrm{w}}}{\mathrm{d}\mathsf{\alpha }}}\right)$$

(5)

where:

• ${\mathrm{N}}_{\mathrm{c}}$—quantity of combustion gases in the cylinder
• $C_{\mathrm{v}\mathrm{c}}$—molar specific heat
• $\mathsf{\xi }$—heat release law (combustion law)
• Q_w—heat transferred to the wall

With a Wiebe type relation, the heat release speed was calculated as [25]:

$$\mathsf{\xi }=1- e^{{-\mathsf{\alpha }}_1·x^{f_1+1}}$$

(6)

where:

• x—relative angle
• f—kinetic coefficient
• α₁—start of combustion moment

The relative heat release rate is calculated as follows:

$${\displaystyle \frac{\mathsf{\delta }\mathsf{\xi }}{\mathrm{d}\mathsf{\alpha }}}={\mathsf{\alpha }}_1·{\displaystyle \frac{{\mathrm{f}}_1+1}{{\mathsf{\alpha }}_a}}·x^{f_1}·e^{{-\mathsf{\alpha }}_1·x^{f_1+1}}$$

(7)

where:

• ${\mathsf{\alpha }}_a$—combustion duration

The gas quantity in the cylinder is made of residual gases, burning products, and fresh air, in kmol and is calculated as follows [25]:

$${\mathrm{N}}_{\mathrm{c}\mathrm{y}\mathrm{c}\mathrm{l}\mathrm{e}}={[\mathrm{N}}_{\mathrm{r}\mathrm{e}\mathrm{s}}+\mathsf{\xi }·{\mathrm{N}}_{\mathrm{b}\mathrm{u}\mathrm{r}\mathrm{n}\mathrm{e}\mathrm{d}}+(1-\mathsf{\xi })·{\mathrm{N}}_{\mathrm{f}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}}]·m_{\mathrm{c}\mathrm{y}\mathrm{c}\mathrm{l}\mathrm{e}}$$

(8)

where:

• N_res—quantity of residual gas
• N_burned—quantity of burned gas
• N_fresh—quantity of fresh inlet air

The engine used to implement the model described above is Daewoo A15MF-1.5-L DOHC spark ignition engine that was turbocharged. The specifications of the engine are presented in

Table 2. Technical specifications for the A15MF engine.

Specification	Value/Technical Solution
Displacement [cm3]	1498
Compression ratio [-]	9.5
Valves per cylinder	4
Bore [mm]	76.5
Stroke [mm]	81.5
Connecting rod length [mm]	123.5
Maximum power [kW]/speed [rpm]	63/4800
Maximum torque [Nm]/speed [rpm]	136/3400
Fuel system	indirect multi point injection
Engine cooling	liquid coolant

.

The engine was fueled with gasoline and hydrogen in dual fueling mode, the substitution ratio of gasoline with hydrogen being calculate with the formula:

$$x_c={\displaystyle \frac{m_{H_2}{H}_{i_{H_2}}}{m_{H_2}{H}_{i_{H_2}}+m_{gasoline}{H}_{i_{gasoline}}}}·100 [\%]$$

(9)

where:

• m_H2—mass fuel consumption of hydrogen
• m_gasoline—mass fuel consumption of gasoline
• H_{i H2}—lower heating value of the hydrogen
• H_{i gasoline}—lower heating value of the gasoline

The theoretical and experimental procedures were developed under reproducible conditions. The procedure adopted for the evaluation of the uncertainties associated with the experimental measurements assumes that during the experiment at each point a set of three measurements were recorded. For each operation point, three sets of measurements were recorded. For each domain of data, the variation ratio (VR) was calculated according to the according to relation (10) [5]:

$$VR={\displaystyle \frac{\sqrt{\frac{1}{n}·\sum _{j=1}^n{\left(a_j-\frac{\sum _{j=1}^n{a}_j}{n}\right)}^2}}{\frac{\sum _{j=1}^n{a}_j}{n}}}$$

(10)

where:

• n—total number of the parameters
• a—studied parameter
• j—current parameter

A set of three measurements were recorded for the investigated operation regime, for gasoline and hydrogen-gasoline fueling. The VR was used to evaluate the measurement error for the results. The VR values determined for all measured parameters such as engine brake torque, engine speed, brake consumption of gasoline and hydrogen, supercharging pressure, gas temperature, pollutant emission and CO₂ emission levels did not exceed 1%. Thus, the data spreading was narrow, the measurement samples were uniform, and the VC value was below the limit of convergence for each measured dataset. The methodology for achieving reproducibility of the numerical test is included in the definition of the boundary conditions. Boundary conditions define the inputs of the simulation model for velocity and volumetric flow rate and define how a fluid enters or leaves the model. Conditions for the film coefficient and heat flux are defined by the interchange of energy between the model and its surroundings. Thus, boundary conditions connect the simulation model with its surroundings in order to develop the simulation process. Boundary conditions are defined as either steady-state or transient, but steady-state boundary conditions persist throughout the simulation. Transient boundary conditions vary with time, in crank angle degrees (CAD or °CA) or speed and are used to simulate a process of heat release or pressure rise during combustion. The initial conditions, a different type of condition, are active only at the beginning of the simulation process.

For the pipe network used in the modelling process, the air path for both intake and exhaust sides was modeled with a series of pipes, 1D-1D junctions, and 0D-1D connections. The type of 1D pipes allowed for defining the geometry of the system (lengths, diameters, interior roughness). The type of 1D-1D junctions allowed for connecting a pipe to other pipes. The type of 0D-1D connections allowed for connecting a pipe to a finite 0D volume or to the ambient.

The mesh size parameter correspond to the discretization length of the 1D system and its very important in the CFD 1D simulation because the accuracy of the final results and the calculation duration are directly influenced by its value. We used a small mesh size which corresponded to accurate results but a long calculation duration. The initial conditions were defined from static pressure, which defines the initial gas static pressure in the pipe [bar], gas temperature, which defines the initial gas temperature in the pipe [K], and the mass fractions, which define the initial mass fraction of each species of the mixture. According to the associated gas properties parameter, a list of 3 (IFP-Engine) or 12 (IFP-Exhaust) species can be used [43]. The initial mass fractions are then automatically weighted by their initial sum. The pneumatic library only uses single species of gas so it does not need to define any mass fraction. This created sub-model allows for defining thermodynamic properties of fluids used in a spark ignition engine system. A fluid mixture is composed of three gases and a liquid. The first gas corresponds to the air, the second one to the fuel, gasoline, and hydrogen, in a gaseous state, and the third to the burned gases produced by the stoichiometric combustion of the air-gasoline mixture and air-hydrogen-gasoline mixture. The liquid corresponding to the C₈H₁₈ is a gasoline fuel defined in order to take in consideration the vaporization of liquid gasoline as a gas. The icon associated with this sub-model must be inserted in the sketch of an SIE system using 3 gas IFP-Engine components [43]. This sub-model, must specify the following aspects: the mixture index (1 for the use of only one mixture for the simulation), the liquid index, and the level of liquid properties to be used. Here, components of ENG as constant liquid properties defined in this sub-model, and THH as thermal hydraulic liquid properties defined in a TFFDX sub-model (thermal hydraulic library) are used [43]. The level of gas properties to be used, air-gasoline, and hydrogen are constant, linear, pneumatic polynomials, or Janaf-Yaws the designed model using the linear gas definition [43]. For the linear gas definition, an ASCII file is used which contains data keywords like: level (level of gas properties used = 3000 null), tin (first reference temperature in the intake circuit, in K), texh (second reference temperature in the exhaust circuit, in K), Cpin (Cp at the first reference temperature, in J/kg/K), Cpexh (Cp at the second reference temperature, in J/kg/K), r (perfect gas constant, in J/kg/K), μin (viscosity at the first reference temperature, in kg/m/s), μexh (viscosity at the second reference temperature, in kg/m/s), λbin (thermal conductivity at the first reference temperature, in W/m/K), λbexh (thermal conductivity at the second reference temperature, in W/m/K), which are necessary to define absolute viscosity [43]. For the intake valve port modelling, the CFD 1D valve component receives part of its parameterization by the ENGVDEF02 component of the spark ignition library; the input parameters include the valve effective area for both inflow and backflow directions as a function of the valve lift, which can range from 0–590 mm² for the intake valve effective area for inflow and 25–520 mm² for the intake valve effective area for backflow, the value depends on the intake valve lift in the range of 0–10 mm [43].

For the exhaust valve port modelling, the input parameters include the valve effective area for both inflow and backflow directions as a function of the valve lift which can be in the range of 0–260 mm² for the exhaust valve effective area for inflow and 25–255 mm² for the exhaust valve effective area for backflow; the value depends on the intake valve lift in the range of 0–10 mm [43]. The thermodynamic model requires the engine geometry parameters like bore, stroke, compression ratio, and combustion chamber height. The initial temperature, the initial pressure, and the gases mass fraction define the thermodynamic state in a combustion chamber at initial conditions. For the heat transfer evaluation, the cylinder head and piston heat transfer surfaces are used. For the combustion process, the Wiebe’s parameter is used to define the evolution of the parameters with a constant type being set up for spark ignition isochoric combustion [43]. The routine “Eng_wiebe_myamoto_twin” calculates the evolution of the burned mass fraction of fuel, gasoline, and hydrogen during combustion [43]. This routine evaluates the burned mass fraction during the combustion process for the heat release calculations with Wiebe’s law and pre-supposed curve shape according to the above equations [43].

3. Results

3.1. Result of the Experimental Investigations Study

Figure 3 shows the variation of the pressure during combustion of the air-gasoline mixture and for the combustion of the hydrogen-air-gasoline mixture. The presented diagrams were obtained from averaging 200 consecutive cycles. Under conditions of a constant spark ignition timing, with gasoline-hydrogen fueling combustion, the pressure curves start to increase significantly and reach higher maximum pressure due to higher combustion speed and a higher lower heating value (LHV) of hydrogen compared to gasoline. Also the maximum pressure tends to appear sooner per cycle, to get closer to top dead center (TDC) of the main phase of combustion, which may correspond to an improvement in isochoric combustion at H₂ use.

It is important that maximum pressure (Figure 4) increase is at normal values that do not affect the engine reliability. For a 2.15% hydrogen substitution at stoichiometric dosage, a 19% increase of maximum pressure is noted, from 38.22 bar to 45.51 bar. With gasoline fueling, the maximum pressure occurs per cycle at 376 °CA, but with hydrogen-gasoline fueling, the maximum pressure occurs earlier per cycle, at 370 °CA, closer to TDC. Similar results were obtained by other researchers [9,33,34].

The maximum rate of pressure rise showed a significant increase, Figure 5, from 1.51 bar/°CA to 2.13 bar/°CA which is related to the reduction of the combustion process duration with hydrogen use due to the air-hydrogen mixture’s higher combustion speed compared to the air-gasoline mixture. The increase of the maximum pressure rise rate of 41% with hydrogen use may be a criterion of the hydrogen dose limitation in order to maintain engine reliability, normal operation, and to avoid the possibility for knock appearance. The increase of the maximum pressure rise rate is correlated with the increase of the maximum pressure and demonstrates the acceleration of the rapid phase of combustion that becomes more intense. However, if the maximum pressure is limited at 45.51 bar and its maximum rise rate is at 2.13 bar/°CA by limiting the hydrogen percent at xc = 2.15%, the engine will run normally and reliability is maintained without causing damages to the engine. Similar results were obtained by other researchers [9,33].

From the heat release rate (Figure 6) and heat release laws graphics (Figure 7), it can be observed that the air-gasoline-hydrogen mixture burns faster (Figure 8) compared to air-gasoline mixture and has a higher release rate than gasoline fueling case. With hydrogen use, the maximum heat release rate is increased by 11.28% and it achieved sooner at 5 CAD. Due to the fact that the hydrogen quantity used is relatively low, the same engine adjustments were maintained and it can be observed that the initial combustion phase starts earlier. The onset of combustion occurred at 3 CAD sooner per cycle with hydrogen use compared to gasoline. Also, with hydrogen use due to air-hydrogen’s higher combustion speed, the total duration of the combustion process is shortened by 10 CAD, as seen in Figure 8. Other researchers [9,34,36,40,41,42] obtained similar results of hydrogen combustion and shortened duration.

The increase and the acceleration of the heat release rate leads to the increase of the maximum pressure and its rising rate during combustion, and the variation tendency of the heat release rate is correlated with the variation of the in-cylinder pressure. These aspects are related to the reduction of the total combustion duration with 10 CAD, the angle of 90% of mass fraction burned decreasing from 47 CAD to 37 CAD, for averaged values. Similar results were obtained by other researchers [2,9,10,12,34].

Figure 9 presents the brake specific energy consumption (BSEC), which is reduced with hydrogen fueling due to the higher lower heating value (LHV) of H₂. Due to combustion improvement with hydrogen use, the engine thermal efficiency is improved and the brake specific energy consumption decreases by 4.8%. The decrease of the BSEC is assured under the condition of a limitation of the maximum pressure at 45.51 bar and of the maximum pressure rise rate at 2.13 bar/°CA for a normal engine operation. Similar result of brake specific fuel consumption with hydrogen use were confirmed by other studies [2,5,6,7,8,9,12,13,14,15,16,17,18].

For the excess air coefficient where λ = 1, xc = 2.15% hydrogen, the combustion efficiency is maximum, the brake thermal efficiency is maximum and the brake specific energy consumption is minimum in condition of unchanged spark timing. The goal is to maintain the unchanged spark timing and to find out at which hydrogen percentage is the combustion efficiency optimum, xc = 2.15% in this case, ensuring engine design and tune modifications are minimal. Thus, the hydrogen fueling solution of using a low hydrogen dose in the inlet air can be easily adapted to spark ignition engines that are new or already in use in order to ameliorate their pollutant performance.

The level of unburnt hydrocarbons emission (HC) is decreased at dual fueling use compared to gasoline fueling, Figure 10.

With hydrogen use, the HC emission level decreased by 11.11% compared to gasoline fueling. Higher speed combustion of the homogeneous air-hydrogen-gasoline mixtures favors the reduction of HC emission levels. The wider limits of flammability and the higher burning speed of the air-hydrogen mixture reduced the possibility of incomplete combustion and extinguishing the flame in the gas mass or at the wall, which influences the reduction of the HC emission level. Similar results were obtained by other researchers [10,12,20,23,24,37].

The CO emission level decreased by 12.5% with hydrogen use versus gasoline fueling, Figure 11. When using hydrogen, the faster combustion of the air-fuel mixture and the reduction of the duration of the combustion process can favor the reduction of carbon monoxide formation in the gas mass through dissociation reactions, which, together with the reduction of the carbon content in the mixture upon substitution of hydrogen for gasoline can ensure the reduction of carbon monoxide emissions, Figure 11. Similar results were obtained by other researchers [10,12,23,24,37].

The reduction of the in-cylinder carbon content with hydrogen use also influenced the CO₂ emission levels, Figure 12.

The CO₂ emission level decreased by 33.7% with hydrogen use compared to gasoline, Figure 12. The CO₂ emission level reduction can be attributed to the reduction of the in-cylinder carbon content in the air-fuel mixture with hydrogen use. The mass composition of the gasoline was 85.4% carbon, 14.2% hydrogen, and 0.4% oxygen and since hydrogen does not contains carbon, the substitution of the gasoline with hydrogen decreases the carbon content in the mixture by 2.2% at xc = 2.15. The reduction of the carbon content can explain the reduction of the CO₂ emission level and the reduction of the CO and HC emission levels. Similar results were obtained by other researchers [2,10,20,37].

The NOx emission levels decreased by 63.23% with hydrogen-gasoline fueling compared to classic fueling as seen in Figure 13.

The emission levels of nitrogen oxides depended on the amount of oxygen available in the combustion chamber, the temperature of the gases, and the time available for the initiation and development of the chemical reactions specific to the formation of nitrogen oxides. When using hydrogen, the amount of oxygen available in the cylinder is reduced, with hydrogen occupying space at the expense of oxygen, which implies the possibility of reducing the emission level of nitrogen oxides. The acceleration of the combustion process in the presence of hydrogen, an aspect noted by the change in the release speed of the heat of the reaction and the law associated with it, Figure 6 and Figure 7, and by reducing the duration of the combustion process, Figure 8, allows the reduction of the time available for priming and carrying out the chemical reactions involved in the formation of nitrogen oxides. At the same time, there is also an influence of the global temperature in the combustion chamber. The adiabatic temperature of the flame when burning air-hydrogen mixtures is 2300 K at stoichiometric lambda and 2411 K when burning air-gasoline mixtures, at lambda λ = 1. When using hydrogen, the influence of reducing the global combustion temperature can influence a reduction of NO_x emission levels. Similar results were obtained by other researchers [2,19,22,23,24,34].

3.2. Results of the Theoretical Study

The in-cylinder pressure diagrams obtained with the AMESim model at the theoretical study are presented in a compared way with the experimental pressure diagrams in Figure 14.

It can be observed that hydrogen fueling can lead to the increase of the maximum pressure when compared to gasoline fueling in both of the experimental and theoretical studies, due to higher LHV of hydrogen versus gasoline. The increase in maximum pressure and of the maximum pressure rise rate was associated with the increase of the flame speed in the presence of hydrogen due to the higher LBS and LHV of hydrogen compared to gasoline. These aspects are related to the acceleration of the combustion process due to a higher heat release rate in the presence of hydrogen in the in-cylinder mixture. Similar influences on the combustion process with the addition of hydrogen are present at experimental and theoretical study.

The combustion process had a decreased duration in the presence of hydrogen as seen in Figure 15. The heat release rate analysis shows that the combustion starts earlier by 1 °CA in the presence of hydrogen, the initial phase of combustion accelerated due the increase of the inflammability limits and the lower minimum ignition energy of hydrogen compared to gasoline. From the experimental study it was revealed that the total duration of combustion decreased by 10 °CA, which is validated by the results of the theoretical study which indicated a decrease of 9 °CA. The reduction of the combustion duration lead to the increase of the maximum pressure, Figure 14, and of the maximum pressure rise rate.

The acceleration of the combustion process with the addition of hydrogen, which will result in a shorter duration, is reflected in the behavior of the combustion law, Figure 16. The initial and the main phase of combustion were reduced because of the H₂ addition in the final cylinder mixture. The main phase of combustion was accelerated by 11 °CA in the theoretical model due to estimation of the heat transferred to the wall, but only by 6 °CA, in experiment, for averaged diagrams from 200 consecutive cycles, but the trends are similar and validate each other.

4. Comparison between Theoretical and Experimental Results

The results obtained on the test bench were compared to the results obtained with the numeric model. At first sight, the simulated results were pretty close to the experimental ones, but not identical. This occurred because the test bench can have multiple sources of errors and unstable combustion, especially because the excess air coefficient (λ) can differ significantly in a short amount of time.

Regarding the maximum pressure (Figure 14), when fueling only with gasoline at stoichiometric dosage, the experimental value was 38.22 bar at 376 CAD while the theoretical value was 36.64 bar at 379 CAD. When substituting gasoline with 2.15% hydrogen, the maximum pressure increased to 45.51 bar at 370 CAD in the experimental case and 45.57 bar at 375 CAD with the numerical model. The difference between experimental and theoretical maximum pressures was 4.2% for gasoline and 0.79% for the gasoline-hydrogen fueling case. The difference between the experimental and theoretical angle of maximum pressures was 0.13% for gasoline and 1.34% for the gasoline-hydrogen fueling case. This comparison shows a good approximation between the experimental and theoretical pressure diagrams. The increase of in-cylinder pressure and heat release rate leads to better thermal efficiency.

The heat release rate (Figure 15) when fueling with gasoline and gasoline and hydrogen (2.15%) at stoichiometric dosage showed very strong results between the numerical simulation and the experimental results. When fueling only with gasoline, the maximum heat release rate was 39.97 J/CAD at 368 CAD obtained on the test bench and 39.98 J/CAD at 372 CAD with the numerical model. For gasoline fueling, the difference between experimental and theoretical maximum HRR was 0.025% and 1.08% for angle of maximum HRR. When adding hydrogen, the heat release rate ascended to 45.58 J/CAD at 362 CAD by experiment and 42.01 J/CAD at 362 CAD with the numerical model. For gasoline-hydrogen fueling, the difference between experimental and theoretical maximum HRR was 8.1% and 0% for angle of maximum HRR.

In Figure 16, the combustion laws indicate that the addition of hydrogen accelerates combustion and also shortens it, mainly the initial and the main phase of combustion. When fueling only with gasoline, the combustion duration was 45 CAD in the experimental measurements, and 46 CAD in the numerical model. The difference between experimental and theoretical combustion duration was 2.1% for gasoline fueling. When 2.15% hydrogen was added, the combustion was reduced to 39 CAD (by experiment) and 38 CAD with the numerical model. The difference between the experimental and theoretical combustion duration was 2.5% for gasoline-hydrogen fueling. The initial and the main phase of combustion were reduced with the increase of the H₂ percentage in the mixture. The duration of the combustion process was reduced by 10 CAD in experiment and by 9 CAD in the results of the numeric model, the difference of 1 CAD being acceptable. Because of the specific calculations of the numerical model and the estimation of heat transferred to the wall, the main phase of combustion was accelerated by 11 °CA in the theoretical model, resulting in an acceptable difference of 5 °CA versus the experiment (with an average 6 °CA advance in the main phase position versus TDC), which is acceptable given the conditions of using averaged pressure diagrams from 200 consecutive combustion cycle to calculate the HRR. This observation was confirmed by studies like [9,10,12]. This comparison shows a strong approximation between the experimental and theoretical results and validates the numerical model with experimental results. In future research, it is possible that in order to reduce the workload, this theoretical model can be used to estimate the combustion performance, in terms of pressure variation and heat release rate, for different operating regimes and other cyclic doses of hydrogen.

The numerical model can reproduce with an acceptable level of error the results obtained experimentally.

5. Conclusions

Hydrogen used as fuel added into the inlet air of a spark ignition engine had the following main influences:

-

An increase of the maximum pressure by 19% from 38.22 bar to 45.51 bar, for 2.15% hydrogen and a stoichiometric dosage. The pressure increase is acceptable as the engine reliability was affected and the hydrogen cyclic dose was limited at 2.15%. Also, the angle of the maximum pressure appeared sooner per cycle, changing from 376 CAD, recorded with gasoline fueling, to 370 °CA with hydrogen use. The maximum pressure was achieved faster during combustion and the maximum pressure rise rate increased by 41% with hydrogen use compared to gasoline, the higher increase percentage is a condition limited by the hydrogen share at xc = 2.15%. By limiting the hydrogen share at xc = 2.15% it can be assured that the limitation of the increase in maximum pressure at 45.51 bar and of the maximum pressure rise rate at 2.13 bar/°CA for normal operation, knock avoiding and ensuring engine reliability. Similar tendency of pressure increase is present in the theoretical results obtained by thermodynamic model;

-

The maximum heat release rate rose from 39.96 J/CAD with gasoline fueling to 44.47 J/CAD with hydrogen-gasoline fueling, being increased by 11.28% which correlates with the increase of the maximum pressure and maximum pressure rise rate. The maximum heat release rate was achieved at 5 CAD sooner per cycle aspect which is correlated with an earlier start of combustion, at 3 CAD sooner per cycle with hydrogen use compared to gasoline. Due to higher air-hydrogen combustion speeds, the total duration of the combustion process was shortened by 10 CAD. These aspects are related with the reduction of the total combustion duration of 10 CAD. Similar tendency of heat release rate increases was present in the theoretical results obtained by the thermodynamic model. Similar variation tendency of combustion laws was present in the theoretical results also obtained by the thermodynamic model;

-

The brake specific energetic consumption decreased by 4.8% with the conditions of limiting the maximum pressure at 45.51 bar and the maximum pressure rise rate at 2.13 bar/°CA for normal engine operation, which requires the limitation of hydrogen percent at 2.15%;

-

The level of HC emissions decreased by 11.11% compared to gasoline fueling due to a higher combustion speed with homogeneous air-hydrogen-gasoline mixtures which favors the reduction of the HC emission level;

-

The CO emission levels decreased by 12.5% with hydrogen use versus gasoline fueling due to the improvement of the combustion process by using hydrogen, a reduction of the carbon content with the substitution of gasoline with hydrogen, and faster combustion with the air-fuel mixture with the reduction of the duration of the combustion process favoring the reduction of the weight of the dissociation reactions with the formation of carbon monoxide in the gas mass;

-

CO₂ emission levels decreased by 33.7% with hydrogen use compared to gasoline; the reduction can be attributed to the reduction of the in-cylinder carbon content by 2.2% and the air-fuel mixture with hydrogen use for xc = 2.15%;

-

The reduction of carbon content can explain the reduction of CO₂ emission levels and its influence on the reduction of CO and HC emission levels;

-

The NOx emission level decreased by 63.23% with hydrogen-gasoline fueling when compared to classic fueling, due to the reduction of the amount of oxygen available in the cylinder, hydrogen occupying space at the expense of oxygen, and the reduction of the combustion duration which implies the reduction of the time available for priming and carrying out the chemical reactions assigned to the formation of oxides of nitrogen. An influence of a lower adiabatic flame temperature from air-hydrogen mixture combustion is also possible;

-

As a result of the low density and the difficulty of storing large quantities of hydrogen onboard, fueling an engine only with hydrogen is not currently a feasible method in the automotive industry;

-

Using larger quantities of hydrogen for different engine loads especially in lean mixtures, the attention focused on engine tuning can be a direction for future research.