About this Journal Submit a Manuscript Table of Contents
Advances in Mechanical Engineering

Volume 2013 (2013), Article ID 567159, 19 pages

Research Article

Numerical Simulation of PAHs Formation and Effect of Operating Conditions in DI-Diesel Engines Based on a Comprehensive Chemical Mechanism

School of Aerospace, Tsinghua University, Beijing 100084, China

Received 22 November 2012; Revised 25 April 2013; Accepted 13 May 2013

Academic Editor: Moran Wang

Copyright © 2013 Bei-Jing Zhong and Jun Xi. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


Three-dimensional numerical simulations of polycyclic aromatic hydrocarbon (PAH) formation in a Chaochai 6102bzl direct injection diesel engine are performed. n-Heptane is chosen as the fuel. A detailed mechanism, which includes 108 species and 572 elementary reactions that describe n-heptane oxidation and PAH formation, is proposed. A reduced kinetic mechanism, with only 86 reactions and 57 species, is developed and incorporated into computational fluid dynamics (CFD) software for the numerical simulations. Results show that PAHs, which were mostly deposited at the bottom of the diesel combustion chamber wall, first increased and then decreased with the increase in diesel crank angle. Furthermore, the diesel engine operating conditions (intake vortex intensity, intake air pressure, fuel injection advance angle, diesel load, and engine speed) had a significant effect on PAH formation.

1. Introduction

The diesel engine is widely used in various kinds of power devices and has gradually become one of the main power sources in different types of vehicles because of its low oil consumption, high thermal efficiency, good compatibility, and energy savings, as well as low HC and CO emissions [1]. However, the diesel engine can emit a large number of soot particles during operation [25]. Research indicates that diesel engines may need 0.2–0.5% fuel oil to convert soot particles into extra fine particles (~0.1 μm diameter) [6] emitted from the exhaust pipe. These particles consist of hydrocarbons (including aromatic hydrocarbon materials) adsorbed on carbon black. Therefore, controlling soot emission is a key issue in the development of diesel engines. To improve engine soot emission and engine performance, understanding the soot structure and formation mechanism of diesel engines is necessary.

In general, nucleation and growth are two necessary processes considered in the modelling and simulation of soot formation. Nucleation considers the dehydrogenation of hydrocarbons and recombination of small molecules, which result in the formation of polycyclic aromatic hydrocarbons (PAHs) as the soot core. Growth should consider a series of processes, including the transition of PAHs to small particles and soot particle change by the collision, condensation, absorption, and chemical reactions of gaseous compositions on the surface of the particles [7].

The numerical simulation of turbulent diffusion combustion in direct injection (DI) diesel engines is highly complicated because of the strong nonlinear interaction between turbulent flow and complex chemical reactions. Some researchers [8, 9] directly used the EBU model in three-dimensional simulations to describe diesel engine combustion. Cordiner et al. [10] used a mixed 1D-3D numerical procedure, together with the Shell and characteristic-time combustion models, to analyse combustion and exhaust emission in a dual-fuel diesel/natural gas engine. In the procedure, the 1D code provides the pressure boundary conditions as input information for the 3D simulations. Chen et al. [11] and Lim et al. [12] performed three-dimensional simulations of diesel engine combustion and emission using the eddy-dissipation-conception (EDC) model with simplified reaction mechanisms. Hu et al. [13] developed a mixed-mode combustion model, which takes advantage of the mixing details provided by large-eddy simulation, to cover the major regimes pertaining to diesel engine combustion. The model uses kinetically controlled, quasisteady homogeneous, quasisteady flamelet, and partially premixed combustion modes. The combustion models used for the simulation of turbulent diffusion flames also include the probability-density-function (PDF) transport equation model and laminar flamelet model [14, 15]. Given detailed chemical kinetics, the first two models require substantial calculation because the transport equations for all species have to be solved. Unlike the first two models, the laminar flamelet model, derived separately by Peters [16] and Kuznetsov [17], treats the mixture fraction as an independent variable. This model uses the scalar dissipation rate for the mixing process and views the turbulent flame as an ensemble of thin, laminar, local one-dimensional flamelet structures embedded within the turbulent flow field [18, 19]. The model is used to solve the local nonequilibrium chemical process caused by aerodynamic strain and can decouple chemistry and turbulent flow in considering the detailed reaction mechanism and molecular transport process. This decoupling considerably reduces the need for additional calculations.

This study aims to perform numerical simulations of the actual operating process of DI diesel engines (Chaochai 6102bzl engine) and formation of PAHs, including benzene (A1), naphthalene (A2), phenanthrene (A3), and pyrene (A4), using the FLUENT CFD software. To this end, a simplified mechanism that is based on a detailed mechanism is proposed. The effect of diesel operating conditions on PAH formation is also analysed. Commercial diesel does not have a chemical formula and its components are complex, making calculations difficult. A single component of alternative fuels is therefore needed. Cetane number (CN) is a measure of fuel ignition quality. It determines the length of the ignition delay period and has a significant effect on soot formation and the combustion process [20]. Thus, the CN of alternative fuels must first be considered. The CN of commercial diesel is about 50 and that of -heptane—an ideal, widely used alternative fuel [2123]—is about 56. In the calculations, -heptane is used as the simulated fuel, and the unsteady-state laminar flamelet concept based on the -heptane combustion reaction mechanism is adopted as the turbulent flame model of the diesel engine.

2. Governing Equations

The process occurring in an engine cylinder is a turbulent reaction flow that is dominated by physical conservation laws, including mass, momentum, energy, and species conservations. Therefore, the governing equations describing combustion in cylinder are mass continuity, momentum conservation, energy conservation, species conservation, and the ideal gas equation of state.

2.1. Mass Continuity

The species conservation equation was written as where is the mass density of species, ; kg/m3; is the density of mixtures, kg/m3; is the velocity of fluid, m/s; is the diffusion coefficient, m2/s; is the Dirac delta symbol; is the source term caused in reaction; is the source term caused in spray.

After adding all species, the equation of continuity for the mixture is obtained:

2.2. Momentum Conservation

The momentum conservation equation is written as where is the static pressure, Pa; is the turbulent kinetic energy, kJ/m3; is the gravitational body force; is the surface tension, m; is the momentum increment per unit volume caused in spray, kg/(m2s).

For Cartesian coordinate system, the -, -, -direction momentum conservation equation are given as follows:




2.3. Energy Conservation

The energy conservation equation is written as where is the specific internal energy, kJ/kg; is the vector of heat flux, kJ, equals the sum of heat conduction and thermal diffusion, and is given by where is the heat conductivity; is the gas temperature, ; is the species enthalpy; is the heat of combustion; is the thermal source term caused in spray.

2.4. The Ideal Gas Equation of State

One has where is the universal gas constant, kJ/(kg·K); is the molecular weight of species; is the specific internal energy of species; is the constant-pressure specific heat of species, J/(kg·K).

3. Models and Boundary Conditions

3.1. Turbulence Model

Numerous models can describe turbulent flow; these include the single equation model, double equation model (the standard , RNG , and realisable models), Reynolds stress model, and large eddy simulation. Given the geometric curvature of combustor configuration and high-speed piston movement, the in-cylinder fluid flow is complex and involves eddy and secondary flow; hence, the realisable turbulence model proposed by Shih et al. [24] is selected for the present study. The realisable model is a relatively recent development and differs from the standard model. The kinetic energy and transport equations of the former have the same form as the standard and renormalisation group models, but the dissipation rate equations are different. The produced items of dissipation rate do not contain the produced items of turbulent kinetic energy. The realisable model is adopted for fluid flow, including boundary eddy flow, strong adverse pressure gravity, separate flow, and secondary flow [25, 26].

The modelled transport equations for and in the realisable model are as follows: where In these equations, represents the generation of turbulent kinetic energy by mean velocity gradients; is the generation of turbulent kinetic energy by buoyancy; denotes the contribution of the fluctuating dilatation in compressible turbulence to the overall dissipation rate; , , , and are constants; and represent the turbulent Prandtl numbers for and , respectively; and and are user-defined source terms.

3.2. Turbulence-Chemistry Interaction Model

In diesel engines, fuel is sprayed into the cylinder. The fuel evaporates, mixes with the surrounding gases, and then autoignites as compression raises temperature and pressure. The diesel engine nonsteady-state laminar flamelet model that is based on the works of Pitsch et al. [27] and Barths et al. [28] can describe the chemistry in a single one-dimensional laminar flamelet and model ignition, as well as the formation of the product, intermediate, and pollutant species. It is chosen in this study to predict the combustion process in the diesel engine with compression ignition [25]. Computationally expensive chemical kinetics is reduced to a one-dimensional model, which is significantly faster than the laminar-finite-rate, EDC, or PDF transport models, and calculates kinetics in two or three dimensions. The simulation results confirm good accuracy.

In a laminar diffusion flame, the species mass fraction and temperature along the -axis can be mapped from the physical space to the mixture fraction space. Thus, they can be uniquely described by two parameters: mixture fraction , defined in (7), and strain rate (or equivalently, scalar dissipation , defined in (8)). The chemistry is therefore reduced and completely described by quantities and . Finally, a set of simplified laminar flamelet equations can be obtained in the mixture fraction space, including species mass fraction equations: and a temperature equation: In (13) and (14), , , , and are the th species mass fraction, temperature, density, and mixture fraction, respectively; and are the th species specific heat and mixture-averaged specific heat, respectively; denotes the th species reaction rate; and represents the specific enthalpy of the th species.

The mixture fraction can be written in terms of the atomic mass fraction as [29]

where is the elemental mass fraction for element ; the subscript ox is the value at the oxidiser stream inlet; and the subscript fuel denotes the value at the fuel stream inlet. If the diffusion coefficients for all the species are equal, then (15) is identical for all the elements and the mixture fraction definition is unique. The mixture fraction is therefore the elemental mass fraction that originates from the fuel stream.

Scalar dissipation varies along the axis of a flamelet and must be modeled across the flamelet. An expression for variable density is used [30]:

where is the density of the oxidizer stream. For a counterflow diffusion flamelet, flamelet train rate can be related to the scalar dissipation at :

where is the scalar dissipation at the stoichiometric ratio; is the characteristic strain rate; denotes the stoichiometric mixture fraction; and is the inverse complementary error function.

In an adiabatic turbulent diffusion flame system, the species mass fraction and temperature in the laminar flamelets are completely parameterized by and , and the time-averaged characteristic scalars can be determined from the PDF of and as

where represents the species mass fraction and temperature and denotes a joint probability density function. For nonadiabatic laminar flamelets, considering the computational cost, heat transfer to the system is assumed to have a negligible effect on flamelet species mass fractions. So the flamelet profiles are convoluted with the assumed-shape PDFs as in (18).

The flamelet species and energy equations ((13) and (14)) are simultaneously solved with the flow. To account for the temperature rise during compression, the flamelet energy equation (14) contains an additional term on the right-hand side:

where is the specific heat and is the volume-averaged pressure in the cylinder. This rise in flamelet temperature caused by compression eventually ignites the flamelet.

The flamelet equations are advanced for a fractional step using properties from the flow, which is then advanced for the same fractional time step using properties from the flamelet. The initial flamelet condition at the beginning of the diesel simulation is a mixed-but-unburned distribution. For the flamelet fractional time step, the volume-averaged scalar dissipation and pressure, as well as the fuel and oxidiser temperatures, are passed from the flow solver to the flamelet solver. After the flamelet equations are advanced for the fractional time step, a PDF table is converted into a nonadiabatic steady flamelet table. Using the properties in this table, the CFD flow field is then advanced for the same fractional time step.

3.3. Kinetic Model

Turbulent combustion in an engine cylinder is highly complex. Hence, three-dimensional turbulent combustion simulation with a detailed reaction mechanism is difficult to complete using current computer resources. Detailed reaction mechanisms should be reduced. For this purpose, we first propose a detailed chemical mechanism describing -heptane as a surrogate fuel for kerosene, oxidation, and PAH formation. It is then simplified to obtain the reduced-detail mechanism that can be coupled with CFD.

The detailed -heptane reaction mechanism that describes PAH formation and oxidation consists of 108 species and 572 elementary reactions [31]. The mechanism was taken from Wang and Frenklach [32] (R31–R572, refer to [31]) and Curran et al. [33] (R1–R30, refer to [31]). As the first aromatic ring (benzene) is formed, the growth of larger aromatic species essentially follows the H-abstraction-C2H2-addition (HACA) mechanism. The HACA mechanism and aromatic combination, which describe the details of the growth of aromatic hydrocarbons, can be found in [15]. The next stage is to integrate reaction kinetics in the multidimensional CFD solver. The detailed mechanism is simplified prior to multidimensional CFD modelling. Net reaction rate analysis and sensitivity analysis are conducted to simplify the detailed mechanism. The simplified mechanism includes 57 components and 86 elementary reactions (see Table 1), and can be coupled with the CFD multidimensional model. A comparison of the simplified and detailed mechanisms shows very close calculation results, with a 15% relative error for intermediate profiles. These results confirm that the authenticity of the prediction of flame structure and PAH (A1, A2, A3, and A4) formation generated by the mechanism simplification procedure is indistinguishable from that yielded by the detailed mechanism over the specified fixed range of conditions. The details of the simplification of this mechanism and its testing can be found in [34].

Table 1: Skeleton reaction mechanism obtained by sensitivity analysis.
3.4. Spray Model

FLUENT provides different spray models for droplet collision and breakup, as well as a dynamically varying drag coefficient that accounts for variations in droplet shape. These models are the droplet collision model, applicable to low-Weber-number collisions, and droplet breakup models; the latter include the Taylor analogy breakup (TAB) model, applicable to low-Weber-number injections and low-speed sprays into a standard atmosphere, and the wave model that is applicable to Weber numbers greater than 100 and popular in high-speed fuel injection applications [14]. The dynamic drag model provided by FLUENT is adopted for the accurate determination of droplet drag coefficients, which are crucial for accurate spray modelling in this work. It is compatible with the TAB and wave models for droplet breakup. When the collision model is switched on, the collisions reset the distortion and distortion velocities of the colliding droplets. When sprayed into the combustion chamber through the nozzle, the droplets may merge or break up.

3.5. Initial and Boundary Conditions

Even as the precise geometric shapes of a physical model are an important precondition to accurate simulation, creating a computational module strictly in accordance with actual conditions is very difficult because the overall structure of the diesel engine is highly complex. Therefore, the treatment of the structures requires simplification. In this study, the approaches to simplification are as follows. The intake and exhaust strokes are disregarded, with initial and boundary conditions obtained through experiments. The calculation is completed only for one-sixth of the full geometry because the diesel cylinder that is adopted is a fuel injection nozzle with six uniform jet holes. Figure 1 shows the geometric model and calculation grid of the DI diesel engine at a 10° crack angle (CA) before top dead centre.

Figure 1: One-sixth of calculation grids at the location of 10°CA BTDC.

The parameters of the geometric model and initial boundary conditions are shown in Table 2. The calculations, including those for the compression, combustion, and expansion strokes, are completed using the FLUENT CFD software solver.

Table 2: Dimensional and operational conditions for diesel modeled.

4. Results and Discussion

4.1. Reliability Analysis of the Models

Comprehensively verifying the models is very difficult because limited test results are available for comparison. Usually, pressure and temperature are the standards of simulation. The examined comparisons of the detailed mechanism were completed in the laminar flames [31, 35]. The results show that the mechanism can describe flame characteristics. As shown in [34], the simplified mechanism that is based on the detailed mechanism can also reproduce the major features of the detailed mechanism. Figure 2 shows the variations in average pressure in the cylinder with CA; these variations were determined by numerical simulation and testing. The figure shows that the computed pressure rise and maximum pressure, as well as their corresponding crank positions are essentially consistent with experimental data [36]. The maximum pressure in the cylinder (about 8.5 MPa) occurred at 4°CA after top dead centre. The maximum pressure showed an error of 1.16% only in comparison with the experimental results for 8.6 MPa. The comparison of the pressure curves shows that the numerical simulation of the DI diesel engine reflects actual conditions.

Figure 2: The variation of volume-average pressure in engine cylinder with crank angles.

Figure 3 shows the computed volume-average temperature in the cylinder against variations in CAs. The figure shows three special points, which correspond to the ignition of the fuel at 360°CA, maximum pressure at 364°CA, and maximum temperature at 379°CA. The average temperature in the engine cylinder was about 1400 K at the ignition point (i.e., 360°CA). Below this temperature, the engine flamed out, which is generally consistent with the conclusion of Liu et al. [37]. Kinetic analysis shows that the following important reactions occurred at the ignition point:

Figure 3: The variation of the simulated average temperature in cylinder with crank angle.

It is found that H, O, and OH imposed important effects on ignition, and H2 and C2H4 were produced from the decomposition of C7H16. On the one hand, the small hydrocarbon gases can be conducive to burning. On the other hand, they can increase CH3 and C3H3, after which C3H3 can be converted to A1. This conversion may, in turn, generate A2, A3, and A4.

As the fuel is burned, the considerable heat produced by combustion caused the gas pressure and temperature in the cylinder to sharply rise until the average temperature reached a maximum value of 1935 K at 379°CA, where fuel injection was terminated and combustion was completed. Similarly, the kinetic analysis shows that in addition to reactions , , , , , , , , , , and (see Table 1), other important reactions occurred at the maximum temperature (i.e., 379°CA):

At this point, adequate H, OH, and O were produced but less C3H3 was generated. Figure 4 shows the corresponding spatial temperature distributions at 360°, 364°, and 379°CA.

Figure 4: Temperature spatial distributions in three crank angles (i.e., corresponding to ignition, maximum average pressure, and maximum average temperature).
4.2. Spatial Distributions of Temperature and Species Concentrations

Figure 5 shows the spatial distributions of temperature and concentrations of the major species (C7H16, O2, CO2, and H2O) and PAHs (A1, A2, A3, and A4) at 360°, 364°, and 379°CA. The figure illustrates that the eddy formed in the combustion chamber caused the high-temperature region to propagate through the bottom of the chamber. When the crank moved to 364°CA, a local high-temperature region at 2643 K was formed at the bottom centre of the chamber. Conversely, when the crank moved forward to 379°CA, a high-temperature flame fully propagated through the bottom of the chamber. At the starting point of combustion (360°CA), the fuel (i.e., C7H16) aggregated at the jet flow centre, thereby producing severely insufficient and excess oxygen inside and outside the fuel jet, respectively. When the C7H16 was consumed, only a little fuel remained near the wall at the bottom of the chamber at 364°CA because of the low combustion rate near this wall. This low combustion rate resulted in a lower wall temperature and lower O2 concentration. In the subsequent combustion (up to 379°CA), the remaining fuel continued to combust until it was completely consumed, the O2 concentration at the bottom of the chamber continued to decrease, and the CO2 and H2O concentrations gradually rose, thereby forming a high-temperature region with maximum temperature and severe oxygen deficiency. Nevertheless, the oxygen concentration remained high in the unburned region.

Figure 5: Spatial distribution of the temperature and the mass fractions of C7H16, O2, CO2, H2O, A1, A2, A3, and A4 at three different positions of the crank angle (temperature: 1077–2643 K; C7H16 mass fraction: 0–0.1184195, O2 mass fraction: –0.233; CO2 mass fraction: 0–0.1285799; H2O mass fraction: 0–0.1123324; A1 mass fraction: 0–0.02788063; A2 mass fraction: 0–3.926658 ; A3 mass fraction: 0–1.383634 ; A4 mass fraction: 0–1.92 ).

As shown in Figure 5, large quantities of A1, A2, A3, and A4 started forming near the wall of the combustion chamber bottom at the ignition point (1750 K), which corresponds to 360°CA. The species continued to form at 379°CA (i.e., the point at which the largest average temperature occurred). At 360°CA, the temperature further improved after ignition, but the fuel remained in the stage of diffusion combustion. It was poorly mixed with air, resulting in pyrolysis and incomplete fuel combustion. At 364°CA (i.e., the point at which the maximum average pressure occurred), the mixing of fuel with air continued to proceed poorly because the diffusion flame had only begun to propagate through the bottom of the chamber. Thus, the fuel jet achieved easy pyrolysis near the wall at the bottom of the combustion chamber, forming A1, A2, A3, and A4. Moving the crank to 379°CA caused large quantities of PAHs to continue forming near the wall at the bottom of the chamber until they were nearly completely oxidised at the high-temperature region because of adequate O2 concentration.

Figure 6 shows the spatial distributions of the concentrations of some intermediate species (H, C3H3, A3-4), C4H4, and C2H2) at 360°, 364°, and 379°CA. The figure indicates that H, C2H2, and C3H3 formed primarily at the high-temperature region because of the combustion and decomposition of fuel, which resulted in more A1 production through the reaction .

Figure 6: The spatial concentration distributions of the intermediate species (H, C3H3, A3-4, C4H4, and C2H2) at positions of 360°, 364°, and 379°CA (A3-4) mass fraction: 0–4.366237 ; C4H4 mass fraction: 0–8.300977 ; C3H3 mass fraction: 0–0.0003500804; C2H2 mass fraction: 0–0.02670222; H mass fraction: 0–3.321838 .
4.3. Species Concentration Profiles

Figures 7 and 8 show the variation curves of the concentrations of the major species (C7H16, O2, CO2, and H2O) and PAHs (A1, A2, A3, and A4) as a function of the CA corresponding to the conditions shown in Table 2. Figure 7 illustrates that the C7H16 concentration first increased with fuel injection during the delay period prior to ignition. C7H16 was then quickly consumed until it reached its maximum at 360°CA (the C7H16 ignition point). The C7H16 injected during the delay (360°–364°CA) and slow burning (364°–379°CA) periods continued to combust during the spray process and was quickly and almost fully consumed when the injection process was terminated (379°CA). O2 gradually decreased as combustion progressed, whereas CO2 and H2O gradually increased until they reached a fixed value at the end of the fuel combustion process.

Figure 7: The variation of the C7H16, O2, CO2, and H2O concentrations with crank angle.
Figure 8: The variation of the PAHs concentrations with crank angle.

As shown in Figure 8, the PAH concentrations first increased and then decreased with CA. During the delay period (~360°CA), A1, A2, A3, and A4 started to form because of C7H16 pyrolysis before fuel ignition. These species gradually increased in the range 360–364°CA until they reached their maximum at 364°CA. This behaviour implies that the pollutants formed in a dominant manner, corresponding to the stages of diffusion flames in the diesel engine. The fuel injected in the combustion chamber underwent pyrolysis under severe oxygen deficiency before the ignition and diffusion combustion of the formed rich mixture at the jet centre in the initial stage of combustion after ignition. The pyrolysis generated a number of A1, A2, A3, and A4 species. The kinetic analysis shows that the important reactions for PAH formation occurred at , , , and . After moving the crank to 364°CA, strong turbulent flow and fuel vapour in the cylinder accelerated the mixing of fuel with air and considerably improved combustion, thereby generating less A1, A2, A3, and A4. At the same time, large quantities of previously formed A1, A2, A3, and A4 were oxidised at high temperature by reactions , , , and . Thus, the A1, A2, A3, and A4 concentrations were quickly reduced beyond 364°CA because of their dominant oxidation. Beyond 379°CA, fuel injection was terminated, and combustion gradually ceased. Therefore, the pressure and temperature in the cylinder of the diesel engine quickly decreased, and PAH concentrations gradually reached constant levels.

4.4. Effect of Operating Conditions on PAHs
4.4.1. Effect of Intake Vortex Intensity

Vortex intensity is one of the important performance parameters that affect fuel combustion and emissions, especially particle emissions, in diesel engines. Eddy motion promotes mixing of fuel and air, improves the formation of combustible mixtures, and enhances combustion efficiency. To study the effect of intake vortex intensity (defined as the ratio of the intake swirl speed to the engine speed) on combustion and emission, we simulated the diesel engine combustion process and PAH formation under different vortex intensities (0.5, 2.5, and 4.5), with the other parameters maintained at relatively constant levels.

Figure 9 shows the mean-averaged mass fractions of A1, A2, A3, and A4 as a function of CA under vortex intensities of 0.5, 2.5, and 4.5. We can see from the figure that A1 gradually decreased, whereas A2, A3, and A4 only slightly changed with increasing intake vortex intensity. The sharp points of A2, A3, and A4 were 5 −10, 6 −13, and 1 −13, respectively. Increasing the intake vortex intensity is advantageous to increasing airflow velocity and turbulence intensity in the engine combustion chamber, accelerating mixture formation, expanding the high-temperature region, and controlling combustion in the cylinder. The end results are improved combustion conditions and reduced combustion time, increased combustion efficiency, and reduced PAH formation, as confirmed by the simulation results.

Figure 9: The effect of the intake vortex intensity on PAHs formation.
4.4.2. Effect of Intake Air Pressure

Figure 10 shows the mean-averaged mass fractions of A1, A2, A3, and A4 as a function of CA under intake air pressures of 3.5, 4, and 4.5 MPa. The figure illustrates that A1 gradually decreased, whereas A2, A3, and A4 only minimally changed with increasing intake air pressure. The variations in PAHs showed a similar trend (Figure 9).

Figure 10: The effect of the intake air pressure on PAHs formation.

The intake air pressure was higher, and the gas density and oxygen concentration of the engine cylinder were greater. These conditions enabled complete diesel engine combustion under greater excess air coefficients, resulting in more complete fuel combustion and less PAH formation. Increasing the amount of air in the cylinder may reduce PAH production and simultaneously promote the oxidation of previously formed PAHs. Hence, PAH concentrations decreased with increasing intake air pressure.

4.4.3. Effect of Diesel Fuel Injection Advance Angle

Fuel injection advance angle is defined as the time at which fuel is injected into the cylinder and then to the piston only as the experience point of the crank angle. It is an important parameter that affects diesel fuel combustion. Previous studies have shown that the fuel injection advance angle in DI diesel engines has a more significant effect on fuel economy, power, and emission performance than do other parameters. Large or small, injection advance angles have a direct effect on diesel engine power output and fuel consumption, resulting in crude diesel engine and unstable running conditions.

Figure 11 shows the mean-averaged mass fractions of A1, A2, A3, and A4 as a function of CAs under fuel injection advance angles of 6°, 8°, and 10°CA, with the other parameters kept constant. The figure illustrates that A1 gradually decreased, whereas A2, A3, and A4 did not change much with increasing fuel injection advance angle.

Figure 11: The effect of the diesel fuel injection advance angle on PAHs formation.

Increasing the fuel injection advance angle decreased the pressure and temperature when the fuel was injected into the cylinder. Thus, the delay period of ignition was extended. Fuel can mix well with air at a level before the top dead centre (TDC), and then the mixture promptly ignites and burns at the nearby TDC. Therefore, appropriate increments in injection advance angle can improve fuel quality and air mixture. Consequently, combustion efficiency is enhanced. However, immoderate increments in fuel injection advance angle increase the amount of fuel injected into the cylinder before ignition. When the fuel is ignited, the cylinder pressure rapidly rises, resulting in crude diesel. At an excessively small fuel injection advance angle, a fuel mixture begins to form and burn at a level beyond the TDC, where the cylinder pressure and temperature are low, resulting in reduced fuel efficiency.

4.4.4. Effect of Diesel Load

Diesel load regulation is controlled with the quantity of diesel fuel injection. The fuel injected into the cylinder increases with rising diesel load. Accordingly, the heat released by combustion is greater and the engine produces more peak torque. We completed a numerical simulation of different fuel deliveries per cycle per cylinder, keeping the same other parameters constant, to study the effect of diesel load on PAH formation and oxidation in the DI diesel engine.

Figure 12 presents the mean-averaged mass fractions of A1, A2, A3, and A4 as a function of CA under fuel delivery per cycle per cylinder levels of 50, 70, and 90 mg. The figure shows that A1 gradually increased, whereas A2, A3, and A4 only slightly changed with increasing fuel delivery per cycle per cylinder.

Figure 12: The effect of the fuel delivery per cycle per cylinder (load) on PAHs formation.

The different engine loads correspond to different fuel-to-air equivalence ratios. Hence, the effect of engine load on harmful gases in the exhaust is essentially attributed to the equivalence ratio. The fuel and air mixture deteriorated with increasing equivalence ratio, leading to the increased formation of PAHs. During heavy load conditions, the equivalence ratio and temperature are high, and the excess concentrated fuel/air mixture produced in the combustion chamber causes fuel pyrolysis and incomplete combustion. By contrast, during light load conditions, the diesel engine operates with substantial excess air. Thus, combustion is more complete and fewer PAHs are formed.

4.4.5. Effect of Engine Speed

The effect of diesel engine speed on diesel engine combustion and PAH formation was also investigated under fixed parameters. Figure 13 shows the mean-averaged mass fractions of A1, A2, A3, and A4 as a function of CA under engine speeds of 1800, 2300, and 2800 rpm. As shown in the figure, A1 gradually increased, whereas A2, A3, and A4 only minimally changed with increasing engine speed.

Figure 13: The effect of the engine speed on PAHs formation.

When the engine speed increased, more fuel was injected into the engine chamber over a short period, which was insufficient for fuel evaporation and fuel-air mixing. In turn, these resulted in incomplete combustion and the formation of large quantities of PAHs.

5. Conclusion

Three-dimensional numerical simulations of an -heptane-fuelled Chaochai 6102bzl DI diesel engine were performed. The simulations were carried out under actual working conditions and PAH formation (A1, A2, A3, and A4). A combustion model, along with laminar flamelet, nonsteady-state, turbulent flow, and spray models, was used in the numerical simulation. The comparison of the simulation and experimental results for the delay period and variation trends of the cylinder pressure indicates that the models established in this study can accurately describe the actual working conditions of an in-cylinder diesel engine. The numerical results show that during the delay period, large quantities of PAHs were formed because of the high-temperature pyrolysis of fuel at the front of the fuel jet. PAHs were formed near the wall at the bottom of the combustion chamber during the ignition stage and were then completely oxidised at a temperature higher than 2000 K. The simulated results for the diesel engine operating conditions indicate that increasing the intake vortex intensity and intake air pressure promotes complete fuel combustion and reduces PAH formation. Appropriate increases in fuel injection advance angles can also reduce PAH generation. The effect of engine load on the combustion process is essentially attributed to the equivalence ratio. The load increase caused a rise in equivalence ratio, creating a rich fuel mixture and resulting in elevated PAH production. Finally, increasing engine speed shortened the time consumed in fuel and air mixing, resulting in combustion deterioration and the formation of large quantities of PAHs.


The authors gratefully acknowledge the financial support of the China National Science Fund (Grant no. 51036004) and the National High Technology Research and Development Program of China (Grant no. 2009AA05Z219).


  1. S. Daido, Y. Kodama, T. Inohara, N. Ohyama, and T. Sugiyama, “Analysis of soot accumulation inside diesel engines,” JSAE Review, vol. 21, no. 3, pp. 303–308, 2000. View at Publisher · View at Google Scholar · View at Scopus
  2. A. D. H. Clague, J. B. Donnet, T. K. Wang, and J. C. M. Peng, “A comparison of diesel engine soot with carbon black,” Carbon, vol. 37, no. 10, pp. 1553–1565, 1999. View at Scopus
  3. L. M. Pickett and D. L. Siebers, “Soot in diesel fuel jets: effects of ambient temperature, ambient density, and injection pressure,” Combustion and Flame, vol. 138, no. 1-2, pp. 114–135, 2004. View at Publisher · View at Google Scholar · View at Scopus
  4. Z. John, M. J. Wen, S. H. Thomson, S. N. Park, and M. F. Rogak, “Study of soot growth in a plug flow reactor using a moving sectional model,” Proceedings of the Combustion Institute, vol. 30, no. 1, pp. 1477–1484, 2005. View at Publisher · View at Google Scholar
  5. Y. Xin and P. G. Jay, “Two-dimensional soot distributions in buoyant turbulent fires,” Proceedings of the Combustion Institute, vol. 30, no. 1, pp. 719–726, 2005. View at Publisher · View at Google Scholar
  6. J. B. Heywood, Internal Combustion Engine Fundamentals, McGraw-Hill International, 1988.
  7. J. Xi and B. J. Zhong, “Soot in diesel combustion systems,” Chemical Engineering and Technology, vol. 29, no. 6, pp. 665–673, 2006. View at Publisher · View at Google Scholar · View at Scopus
  8. Y. Han, W. Park, and K. Min, “Soot and temperature distribution in a diesel diffusion flame: 3-D CFD simulation and measurement with laser diagnostics,” International Journal of Automotive Technology, vol. 12, no. 1, pp. 21–28, 2011. View at Publisher · View at Google Scholar · View at Scopus
  9. Z. Zheng and M. Yao, “Mechanism of oxygen concentration effects on combustion process and emissions of diesel engine,” Energy and Fuels, vol. 23, no. 12, pp. 5835–5845, 2009. View at Publisher · View at Google Scholar · View at Scopus
  10. S. Cordiner, M. Gambino, S. Iannaccone, V. Rocco, and R. Scarcelli, “Numerical and experimental analysis of combustion and exhaust emissions in a dual-fuel diesel/natural gas engine,” Energy and Fuels, vol. 22, no. 3, pp. 1418–1424, 2008. View at Publisher · View at Google Scholar · View at Scopus
  11. W. Chen, S. Shuai, and J. Wang, “Effect of the cetane number on the combustion and emissions of diesel engines by chemical kinetics modeling,” Energy and Fuels, vol. 24, no. 2, pp. 856–862, 2010. View at Publisher · View at Google Scholar · View at Scopus
  12. J. Lim, Y. Kim, S. Lee, J. Chung, W. Kang, and K. Min, “3-D Simulation of the combustion process for di-methyl ether-fueled diesel engine,” Journal of Mechanical Science and Technology, vol. 24, no. 12, pp. 2597–2604, 2010. View at Publisher · View at Google Scholar · View at Scopus
  13. B. Hu, C. J. Rutland, and T. A. Shethaji, “A mixed-mode combustion model for large-eddy simulation of diesel engines,” Combustion Science and Technology, vol. 182, no. 9, pp. 1279–1320, 2010. View at Publisher · View at Google Scholar · View at Scopus
  14. I. Dhuchakallaya and A. P. Watkins, “Application of spray combustion simulation in DI diesel engine,” Applied Energy, vol. 87, no. 4, pp. 1427–1432, 2010. View at Publisher · View at Google Scholar · View at Scopus
  15. M. Frenklach and H. Wang, “Detailed mechanism and modeling of soot particle formation,” Springer Series in Chemical Physics, no. 59, pp. 165–192, 1994. View at Scopus
  16. N. Peters, “Local quenching due to flame stretch and non-premixed turbulent combustion,” in Western States Section of the Combustion Institute, Spring Meeting, Irvine, CA, 1980, Paper WSS 80-4.
  17. V. R. Kuznetsov, “Influence of turbulence on the formation of high nonequilibrium concentrations of atoms and free radicals in diffusion flames,” Fluid Dynamics, vol. 17, no. 6, pp. 815–820, 1982. View at Publisher · View at Google Scholar · View at Scopus
  18. N. Peters, “Laminar diffusion flamelet models in non-premixed turbulent combustion,” Progress in Energy and Combustion Science, vol. 10, no. 3, pp. 319–339, 1984. View at Scopus
  19. N. Peters, “Laminar Flamelet Concepts in Turbulent Combustion,” in Proceedings of the 21st Symposium (Int'l.) on Combustion, pp. 1231–1250, The Combustion Institute, 1986.
  20. H. Pitsch, H. Barths, and N. Peters, “Three-D in ensional modeling of NOx and soot formation in DI-diesel engines using detailed chemistry based on the interactive flamelet approach,” SAE Transactions, vol. 105, no. 4, pp. 2010–2024, 1996.
  21. H. Pitsch, Y. P. Wan, and N. Peters, “Numerical investigation of soot formation and oxidation under diesel engine conditions,” SAE Transactions, vol. 104, no. 3, pp. 938–949, 1995.
  22. F. Tao and J. Chomiak, “Numerical investigation of reaction zone structure and flame liftoff of DI diesel sprays with complex chemistry,” SAE Transactions, vol. 111, no. 3, pp. 1836–1854, 2002.
  23. Z. J. Peng, H. Zhao, and N. Ladommatos, Effects of Air/Fuel ratios and EGR Rates on HCCI Combustion of n-heptane, a Diesel Type Fuel, SAE Paper, 2003.
  24. T. H. Shih, W. W. Liou, A. Shabbir, Z. Yang, and J. Zhu, “New k-ε eddy-viscosity model for high reynolds number turbulent flows—model development and validation,” Computers and Fluids, vol. 24, no. 3, pp. 227–238, 1995. View at Scopus
  25. FLUENT User's Guide, FLUENT Inc., 2006.
  26. K. Van Maele and B. Merci, “Application of two buoyancy-modified k-ε turbulence models to different types of buoyant plumes,” Fire Safety Journal, vol. 41, no. 2, pp. 122–138, 2006. View at Publisher · View at Google Scholar · View at Scopus
  27. H. Pitsch, H. Barths, and N. Peters, “Three-dimensional modeling of NOx and soot formation in DI-diesel engines using detailed chemistry based on the interactive flamelet approach,” SAE Paper 962057, 1996.
  28. H. Barths, C. Antoni, and N. Peters, “Three-dimensional simulation of pollutant formation in a DI-diesel engine using multiple interactive flamelets,” SAE Paper, 1998.
  29. Y. R. Sivathanu and G. M. Faeth, “Generalized state relationships for scalar properties in nonpremixed hydrocarbon/air flames,” Combustion and Flame, vol. 82, no. 2, pp. 211–230, 1990. View at Publisher · View at Google Scholar · View at Scopus
  30. J. S. Kim and F. A. Williams, “Extinction of diffusion flames with non-unity lewis number,” Journal of Engineering Mathematics, vol. 31, no. 2-3, pp. 101–118, 1997. View at Publisher · View at Google Scholar
  31. B. J. Zhong, D. S. Dang, and Y. N. Song, “3-D simulation of soot formation in a direct-injection diesel engine based on a comprehensive chemical mechanism and method of moments,” Combustion Theory and Modeling, vol. 16, no. 1, pp. 143–171, 2012. View at Publisher · View at Google Scholar
  32. H. Wang and M. Frenklach, “A detailed kinetic modeling study of aromatics formation in laminar premixed acetylene and ethylene flames,” Combustion and Flame, vol. 110, no. 1-2, pp. 173–221, 1997. View at Publisher · View at Google Scholar · View at Scopus
  33. H. J. Curran, P. Gaffuri, W. J. Pitz, and C. K. Westbrook, “A comprehensive modeling study of n-heptane oxidation,” Combustion and Flame, vol. 114, no. 1-2, pp. 149–177, 1998. View at Publisher · View at Google Scholar · View at Scopus
  34. J. Xi and B. J. Zhong, “Reduced kinetic mechanism of n-heptane oxidation in modeling polycyclic aromatic hydrocarbon formation in diesel combustion,” Chemical Engineering and Technology, vol. 29, no. 12, pp. 1461–1468, 2006. View at Publisher · View at Google Scholar · View at Scopus
  35. Y. Zhang and B. Zhong, “Benzene formation mechanism in n-heptane/air partially premixed counterflow flame,” Journal of Tsinghua University, vol. 48, no. 5, pp. 904–908, 2008. View at Scopus
  36. T. C. Zhang, C. L. Song, Y. Q. Pei, S. R. Dong, G. D. Wei, and T. Z. Yang, “Experimental and CFD study on exhaust particulate from diesel engine with common rail injection system,” Journal of Combustion Science and Technology, vol. 13, no. 2, pp. 136–140, 2007 (Chinese). View at Scopus
  37. Y. Liu, Y. Zhang, J. Qin, and N. Peters, “Simulation and experiment for three-dimensional combustion temperation field in direct-injection diesel engine,” Chinese Journal of Mechanical Engineering, vol. 43, no. 2, pp. 196–201, 2007 (Chinese). View at Publisher · View at Google Scholar · View at Scopus