Numerical Analysis of Fuels Type Effect on Combustion and Emissions of Turbo-Charged Direct Injection Engine

: One of the most important sources of contaminant emissions, especially in urban areas, is the car. Certain types of fuels are considered very beneficial in reducing emissions. The objective of this work is to study the effects of different types of fuels on the combustion and flow characteristics through the modeling of a direct injection turbocharged diesel engine. In this study, the simulation of combustion and pollutant emission evolution of an engine alimented with five fuels (C 14 H 30 , C 16 H 34 , C 8 H 18 , C 5 H 12, and C 2 H 5 OH) was performed with CONVERGE CFD software. The results obtained confirm that the diesel engine is more powerful than an engine powered by light fuels such as gasoline. Injection of C 16 H 34 and C 14 H 30 resulted in higher pressure, temperature, and heat rate than C 8 H 18 (15.17% and 12.8% for pressure, 25.12% and 24.4% for temperature, and 54.54% and 31.81% for heat rate compared to C 8 H 18 ). For polluting gases, if the engine is powered by heavy fuel oil, there are fewer unburned hydrocarbons and carbon monoxide, on the other hand, more soot and NOx, compared to C 8 H 18 . For the gaseous pollutants, the injection of the C 14 H 30 generates more NOx and soot but less HC and CO (58.3%, 49.23%, 51.61%, and 2% respectively compared to C 8 H 18 ). On the other hand, injection of C 2 H 5 OH generates a lower NOx and soot emissions level if compared to diesel (reduced levels by 75% and 95% respectively compared to diesel).


Introduction
The majority of internal combustion (IC) engines run on petroleum fuels, which are finite and will be depleted in around 35 years (Saurabh Kumar, 2017).Due to limited energy supplies, there is a risk of future energy shortages.IC engines with less than 185 kW utilize over 1/3 of all petroleum fuels, and one of the important sources of pollution in the environment is the exhaust fume released by these engines.Many pieces of research including Heywood, 1988;Pranab Das, 2015;Bousbaa, 2021, have been conducted in recent years on IC engines, to reduce exhaust emissions by altering operational factors such as injection parameters, fuel type, EGR, and AdBlue.As a result, many technologies have been used to reduce pollutant emissions in diesel engines, such as injection timing, injection pressure, and multiple injections (Okude et al. 2007), retarded injection timing (S.Gnanasekaran et al. 2016), HCCI mode operation (Miyamoto et al. 1999), EGR (P.Das et al. 2015), and high swirl ratio.Because of better spray atomization and air-fuel mixing, the Common Rail (CR) fuel injection system has a very high injection pressure, which can minimize particle emissions (Flaig et al. 1999).Experimenting with an engine by changing numerous settings is time-consuming due to the intricacy of practical implementation.To better understand the complex phenomenon of combustion, currently, to study the parameters that influence the performance and evolution of pollutant emissions in automotive engines, numerical simulation has become one of the most powerful tools.Computational Fluid Dynamics (CFD) software CONVERGE, KIVA, STAR-CD, and others can be found in the market.They are used for the modeling and simulation of integrated circuit motors.It is possible to determine the temporal behavior of any variable of interest at any point in the computational domain by using numerical simulations.This results in a better and more in-depth understanding of the relevant processes necessary for their improvement.In addition, the numerical simulation may be used to explore phenomena that occur over long time-scales or in inaccessible locations that cannot be investigated using traditional experimental methods.Sayin and Canakci (2009) investigated the effect of varying injection times in a diesel engine.When injection time was advanced, they discovered that NOX and CO2 emissions rose but unburned HC and CO emissions decreased.Multiple injections and split injection situations were analyzed quantitatively by Han et al (1996).They discovered that split injection greatly decreases soot without affecting NOX emissions, and numerous injections significantly reduce NOX emissions.Prasad et al. (2011) used extensive 3D CFD simulations to investigate the influence of varied piston bowl designs and injection timings on the combustion parameters of a CI engine.They discovered that at SOI of 8.60 BTDC, a strongly re-entrant piston bowl without a central projection was the optimum for swirl and intensification of TKE near TDC.Jayashankara et al. (2010) used commercial CFD code to conduct numerical research on diesel engine simulation with respect to injection time and air boost pressure.They compared the flow-field findings from CFD modeling to the work by Payri et al. (2004) and found that advancing the injection timing resulted in an increase in cylinder pressure, cylinder temperature, and NOx emissions.Simulation of supercharged and intercooled engines resulted in increased NOx emissions when compared to normally aspirated engines.Yu et al. (2017) studied methanol, ethanol, and butanol-gasoline blends' effect under various alcohol ratios on combustion, performance, and emissions of engine characteristics.They concluded that due to the oxygen in ethanol and butanol-gasoline blends which improves combustion quality, HC emissions decreased, while they were increased with the other fuels.However, alcohols-gasoline blends presented lower NOx emissions and BTE decrease.For the butanol-gasoline blends, the results indicate showed a lower BSFC for its higher LHV.Algayyim et al. (2018) studied the effect of injector hole diameter on the evolution spray behavior of butanol-diesel blends.In their conclusion, through the injection parameters, they can control the engine performance efficiency.Salman et al. (2019) used GT-Power Model to study the effect of injection timing on performance and pollutants emission in turbocharged diesel engines fuelled with Butanol-Diesel blends (5%, 15%, and 25% by volume).The results showed a lower amount of NOx and CO pollutants with the addition of butanol to diesel fuel, on the contrary, the HC and CO2 emissions are elevated.For the advance of the injection, a slight improvement is noticed in the thermal efficiency.On the other hand, the delay in the time of injection showed good results concerning the BTE.In addition, on the snapshot a rate of heat release and minimal pollution of HC, CO, and NOx.Samet et al. ( 2019) published an experimental study to improve performance and emissions in a spark ignition engine.They used Isoamyl alcohol/gasoline blends (30%, 20%, 10%, and 0%) at full load.Different CR (8.0:1, 8.5:1, and 9.0:1) and different values of speed (2600, 2800, 3000, and 3200 rpm) are used.The results show that the exhaust emissions decreased with isoamyl alcohol compared to gasoline at all CRs.In addition, the brake thermal efficiency increased by about 2.67% with a blend of 20% compared to gasoline on CR=9.0:1.The torque and effective power increased respectively by 2.03% and 2.51% with a blend of 20%.Zhiqing et al. ( 2022) used an AVL-Fire CFD code to study the effect of diesel/ethanol/n-butanol blends on combustion parameters, such as temperature, cylinder pressure, and pollutant emissions, such as NOx, CO, and soot.The results showed that diesel/ethanol/n-butanol blends reduced pressure and temperature, NOx, CO, and soot.
Hence, the present work aims to present a numerical simulation of combustion and emissions polluting (soot, NOX, HC, CO) in a direct injection engine fuelled by different fuels, utilizing code CONVERGE CFD (Richards et al. 2013).This research allows us to compare the combustion evolution and pollutants produced by various fuels in a quantifiable manner.

In-cylinder numerical investigation 2.1. Governing equations
In this part, we are using the CFD code CONVERGE (Richards et al. 2013) to perform a numerical investigation.In our numerical study, we worked on compression, atomization, spraying, combustion, and expansion.The following governing differential equations are solved by the CFD code.
Continuity equations for species m: (1) Where; ρ is the total mass density, ρm is the mass density of species m, u is the velocity, D is the mass diffusion coefficient, Ym is the mass fraction of species m, and Sm is the source term due to the chemistry and the spray.With summation of the previous equation over all species, the total mass conservation equation becomes; (2) Keeping mass constant in chemical reactions, the fluid momentum equation for the fluid mixture becomes: (3) where P is the fluid pressure, σij is the stress tensor and Si is the source term.
Finally, the internal energy equation becomes: (4) where e is the specific internal energy, K is the conductivity, hm is the species enthalpy, and T is temperature.

Engine geometry and computational details
CONVERGE is used to simulate a one-cylinder DI diesel engine using the specs listed in Table 1.A periodic engine sector example is adapted for the engine simulation to save computing time.CONVERGE multiplies relevant physical variables in the output files by a suitable factor in the event of a periodic engine sector. Figure 1 shows the numerical mesh that was used to approximate 1/6 of the engine combustion chamber.The Converge pre-processor was used to make it.The use of symmetry lowered calculation time and memory requirements dramatically.For the initial conditions, the temperature of the piston, cylinder head, and cylinder wall are based on experimental data.Note that the mesh is a dynamic mesh making the piston a mobile wall (see Table 2).The injection fuels and their specifications are shown in Table 3 with the important specifications of injection.The software solves Navier-Stokes equations using the approach of Finite Volume (Richards et al., 2013).
-Spay model: A modified KH-RT model is utilized to predict the spray breakup, with the assumption that aerodynamic instabilities (KH) are responsible for the primary breakdown of the injected liquid blobs.Examining the competing effects of the RT processes is used to model the subsequent breakup of these dips.The KH-RT model has been proven to be faster and more accurate than the other CONVERGE models, such as O'Rourke's model (Richards et al., 2013).
-Ignition and Combustion Models: For the combustion simulation, the Time Combustion (CTC) model will be used.The variation of species density m is modeled as follows: (6) Where ρ* is the species density's local and instantaneous thermodynamic equilibrium value, and τc is the typical time to reach equilibrium.The characteristic time is computed using the following formula (7) where τchem, τturb, and f represent the chemical kinetics time, the turbulent mixing time, and the delay coefficient that replicates turbulence's increasing influence on combustion respectively.Further, the chemical time is given by the following expression (Richards et al., 2013): where Echem, Achem, Ru, and Tg are the activation energy given, the constant set in the input files, the universal gas constant, and the gas temperature respectively.Further, the turbulent time is given by the following expression: (9) C2 is a constant in this equation.The turbulent timeframe functions as a sub-grid model for species non-uniformity in a cell.The combustion process may be slowed by the subgrid non-uniformity of species, which cannot be properly accounted for.As a result, the turbulent timeframe is used to slow down the combustion process.
-Model for turbulence: (Richards et al., 2013) The RNG k-ε model with fast distortion is engaged, and the normal k-constants are employed.This type is perfectly suited for this situation because it is built for quick compression or rapid expansion.
To properly model the temperatures near the wall or the resolution of a turbulent boundary layer is not sufficient; the law-of-the-wall must be used.In this study, the Han and Reitz (1997) model is used, this model accounts for compressible effects.
-Models of NOX and Soot Formation: (Richards et al., 2013) The expanded Zeldovich mechanism is used to describe the reaction process of NOX production. ( Chemical species that exist in these global reactions are employed in the single-step fuel conversion equation as follows: (11) The Hiroyasu formation model was used to model soot emission in this investigation. ( where m, sf, and so denote the mass of soot, soot formed and soot oxidized respectively.For more information on this model, readers are referred to (Richards et al., 2013).

Results and discussions
The validation of the code is based on a comparison of experimental and numerical simulation results; the comparable quantity is the cylinder pressure for the 1600 rpm regime, which is shown in Figure 2. When compared to C14H30, it can be shown that the results are in good agreement with those obtained experimentally, with errors of less than 3%.
The SAGE model was used to forecast the combustion of light fuels C8H18, C5H12, and C2H5OH, while the CTC model was used to predict the combustion of heavy fuels C14H30 and C16H34.Figure 3 shows the greatest average pressure at TDC.Before TDC, the air is compressed to 70 bars, but during the combustion phase, it rises to 112 pressure, which gas oils attain after a short time at TDC, and 97 bars for gasoline after a short time at TDC, because of their properties such as calorific power, cetane, and octane number.
For the various fuels, Figure 4 predicts the average temperature evolution of the gases in the cylinder as a function of the crank angle.We see a quick temperature rise, indicating combustion.Because of the greater auto-   ignition delay (which results in a longer combustion period), the maximum average temperature for C14H30 and C16H34 is 1730K and 1800K, respectively, while the maximum average temperature for C2H5OH, C8H18, and C5H12 is 1470K, 1325K, and 1390K, respectively.Figure 5 represents the development of the heat rate as a function of crank angle computed throughout an engine cycle for the five tested fuels.All of the tested fuels show a quick increase in heat rate, indicating burning.Because of the cetane number, the heat emitted by burning is more relevant in the case of C14H30 and C16H34 in contrast to the other tested fuels, according to the study of Figure 5 for gas oils.Furthermore, the C16H34 fuel has a longer autoignition delay owing to its low cetane number, but the light fuels C8H18, C5H12, and C2H5OH have a shorter auto-ignition delay due to their higher-octane rating (Heywood, 1988).
For the various fuels, Figure 6 depicts the Nitrogen Oxides development NOX which is a function of the crankshaft angle.The combustion of C2H5OH emits less NOX mass than others fuels, as can be shown.C14H30 and C16H34, on the other hand, produce a lot of NOX when compared to C5H12  and C8H18; their mass at the conclusion of the cycle is 6. 5× 10 -5 g, 6.37×10 -5 g, 4× 10 -5 g and 3.4× 10 -5 g, respectively.This results because of their properties such as density and viscosity which influences the quality of atomization and vaporization of spray.
For each fuel, Figure 7 depicts the development of soot as a function of crank angle.For light fuels, the amount of soot reaches a maximum value of around 10° after TDC and around 15° after TDC for heavy fuels and then falls towards the conclusion of the diffusion period.Due to the oxygen content that favors combustion, we found that soot emissions in the situation where the engine is fuelled with C2H5OH are much lower (approximately 5.5×10 -7 g) than in the other studied scenarios.
The development of unburned hydrocarbons (HC) as a function of the crankshaft angle is seen in Figure 8.The combustion process is fully responsible for the hydrocarbon emissions.When compared to other fuels such as C5H12 and C8H18, the combustion of C14H30 and C16H34 generates fewer unburned hydrocarbons.Despite having higher density and viscosity than other fuels, the rate of HC emission is lower when C2H5OH is used.This is likely owing to the oxygen component, which accelerates burning (Cheikh KEZRANE, 2016).Figure 9 depicts the development of CO as a function of the crankshaft angle for the investigated fuels.The mass fraction of CO remains practically constant during the self-ignition delay, and right before ignition, the reaction rate rises rapidly, resulting in a fast increase in the mass, which indicates combustion.At the end of the diffusion combustion phase, this mass tends to stabilize.After combustion, C14H30, C16H34, and C2H5OH have masses of roughly 0 g, whereas C5H12 and C8H18 have masses of 3×10 -5 g and 5×10 -5 g, respectively.For CO, between 10° and 15° after TDC, the quantity of gas increases until it reaches a maximum value, then falls as the combustion progresses.CO emissions are reduced when C2H5OH is burned.The same explanations as before may be used to explain this result: The hydrocarbons C5H12, C8H18, C14H30, and C16H34 lack local oxygen (Cheikh KEZRANE, 2016).
Figures 10 and 11 show the NOX and soot contours for just three fuels (at TDC and 40° after TDC).When compared to other fuels, we see that C14H30 emits a lot of NOX and soot.Furthermore, we discovered that when the engine is fuelled with C2H5OH, the NOX and soot concentrations are lower than when the engine is fueled with other fuels.The prior results (Figures 8 and 9) corroborated these findings.

CONCLUSION
To evaluate the effects of injection of five fuels on combustion and emissions pollution evolution.CONVERGE tool has been validated by the experimental data for incylinder pressure evolution.The main conclusions of this study are summarized as follows:  C16H34 and C14H30 have higher pressure, temperature, and heat rate than C8H18 (15.17% and 12.8% respectively for pressure, 25.12%, and 24.4% respectively for temperature, and 54.54% and 31.81%respectively for heat rate compared to C8H18).
 The diesel engine is more powerful than a motor powered by light fuels like gasoline.
 The heavy fuels (the most common being C14H30) generate more NOX and soot but less HC and CO when it comes to gaseous pollutants (58.3%, 49.23%, 51.61%, and 2% respectively compared to C8H18).
 These findings suggest that the significant quantity of NOX and soot created by gas oil combustion is an issue.It's the polar opposite for the species.
 C2H5OH can be considered an alternative fuel for engines (the reduced levels of NOX and soot emissions are 75% and 95%, respectively compared to diesel).
 This study adds to our understanding of the process of various fuels combusting in engines.

Fig. 1 .
Fig. 1.Computational domain of the engine CATERPILLAR 3401 multistep kinetics model based on the Shell model (Richard et al., 2008) has been implemented in CONVERGE to model diesel ignition delay.The Shell model was developed to predict knocking in gasoline engines.To simulate auto-ignition in diesel engines, a simplified reaction mechanism was used.Eight reactions are given in this model by, (5)

Fig. 3 .
Fig. 3. Measured and Predicted in-Cylinder Pressure for different fuels.

Table 1 .
Caterpillar 3401 Engine standard specification

Table 2 .
Initial conditions

Table 3 .
Fuel injection system CONVERGE is used for in-cylinder flow simulations, performance estimation, and combustion analysis in engines.It is a ground-breaking CFD software that overcomes the simulations generation bottleneck mesh.