DI Diesel engines, having the evident benefit of a higher thermal efficiency
than all other engines, have served for both light- duty and heavy-duty vehicles.
However direct injection diesel engines emit more particulates and oxides of
nitrogen than their counterpart and hence a reduction of such emissions is most
urgent. As a result, many technologies such as high injection pressure, multiple-injection
(Okude et al., 2007) retarded injection timing,
EGR, HCCI mode operation (Miyamoto et al., 1999)
and high swirl ratio have been used in high efficiency DI diesel engines in
order to reduce the pollutant emissions. The Common Rail (CR) fuel injection
system offer very high injection pressure which can reduce the emission of particulate
matters due to improved spray atomization and air-fuel mixing (Flaig
et al., 1999). However, the high injection pressure will increase
the NOX emission due to high peak temperature and oxygen rich regions near the
beginning of the combustion process. The application of both Exhaust Gas Recirculation
(EGR) and retarded injection timing in a diesel engine can significantly reduce
NOx emission but increase the soot (Han et
al., 1996). The coupling of these technologies like multiple injection
and EGR (Mingfa et al., 2009), common rail injection
with EGR (Millo et al., 2009) can improve the
engine performance and reduce the emission but there is always tradeoff between
NOx and smoke level.
The interaction between injection timing and EGR needs to be considered when they are applied to a HSDI diesel engine. This combination of injection timing and EGR achieve simultaneously reductions in both NOx and smoke level.
Multi-dimensional engine models or Computational Fluid Dynamic (CFD) models
have progressed to the point where three-dimensional analyses of the cold flow
within the combustion chamber are becoming practical enough to be employed as
part of the actual engine design process. The modeling approach to be described
is relatively simple in nature, which offers advantages related to reduce computational
expense and ease of implementation within existing engine CFD methodologies
(Jennings, 1992). Multi-dimensional modeling has been
shown in several previous studies to be a useful tool for engine design and
optimization, as well as for gaining a better physical understanding of the
in cylinder combustion process (Reitz et al., 2001).
Multi-dimensional models provide detailed geometric information on the flow
field based on solution of the governing flow equations (Heywood,
1988). Multi-dimensional CFD codes solve the full set of differential equations
for species mass, momentum and energy conservation and also account for the
effects of turbulence. These models are best suited to analyze the various processes
of mixture formation and combustion in greater detail (Stiesch,
2003). In multi-dimensional models the time dependent, instantaneous conservation
equations are time averaged and the turbulence correlations are considered to
be proportional to the gradients of the mean flow. Multi-dimensional models
of diesel engine combustion, account for temporal and spatial variations of
the flow field, pressure, temperature, composition and turbulence within the
combustion chamber (Ramos, 1989; Kong
et al., 1995).
This study presents a effect of injection timing and EGR on DI diesel engine combustion performance and Pollutants formation using the Computational Fluid Dynamics code STAR-CD.
The model is based on the numerical simulation of time averaged conservation equations of the transport processes of heat, mass and momentum of the gas phase within the cylinder. A standard RNG k-ε model for the turbulence with wall function treatment for near wall region has been adopted for the solution of conservation equations. The process of combustion in the cylinder after injection of fuel is modelled by dispersed Lagrangian multiphase model. The chemical reaction rate is modelled using global turbulent controlled reaction kinetics (eddy break-up model). The formation of thermal NOx is modelled by the Zeldovich mechanism. The mass, momentum and energy conservation equations solved for general incompressible and compressible fluid flows in Cartesian tensor notation are expressed as:
The laminar-and-turbulent characteristic-time combustion model of Magnussen based on the eddy break-up concept is adopted. This model relates the rate of combustion to the rate of dissipation of eddies and expresses the rate of reaction by the mean concentration of a reacting species, the turbulent kinetic energy and the rate of dissipation of kinetic energy. As per this model, the combustion rate (Rf) is described as:
Aebu, Bebu are dimension less empirical coefficients
and k/ε is the turbulent time scale. The Shell auto-ignition model is employed
in the present analysis with the following assumptions RH hydrocarbon fuel of
nominal composition CnH2n+2
||Radical formed from the fuel (R* denotes radical)
||Products consisting of CO, CO2 and H2O
The chemical model consists of the following set of equations:
The non-reactive species (inert products) in the termination equations are assumed to be equivalent to nitrogen. The rate coefficients of the above reactions take the Arrhenius form.
The modelling of fuel injection processes is an essential part of DI diesel engine simulation. The existing fully coupled stochastic Lagrangian-Eulerian approach used in STAR-CD has been enhanced to avoid the necessity to empirically tune coefficients or other inputs of the spray model. The velocity of the liquid fuel (injection velocity) as it exits the nozzle and enters the combustion chamber, is one of the most important parameters in a spray calculation. It strongly influences the atomisation (break-up processes), the spray penetration, the inter-phase transfer processes and the droplet-droplet interaction. In the present analysis, the effective nozzle model simulates the nozzle. This model determines the injection velocity based on the injector parameters like nozzle hole cross-sectional area, nozzle hole diameter, roughness and the discharge coefficient:
is the volumetric flow rate through the injector, ρd is the
fuel density and Δp is pressure drop across the nozzle.
Nox model: In the present analysis Zeldovich mechanism was used for NOx prediction.
Three different mechanisms have been identified for the formation of nitric oxide during the combustion of hydrocarbons, namely Thermal NOx, Prompt NOx and Fuel NOx. Among these, estimation of thermal NOx important in diesel engine combustion. Thermal NOx is strongly temperature dependent. It is produced by the reaction of atmospheric nitrogen with oxygen at elevated temperatures. For thermal nitric oxide, the principal reactions are generally recognized to be those proposed by the following three extended Zeldovich mechanisms.
PRE PROCESSING AND GRID GENERATION OF THE GEOMETRY
The preprocessing mainly involves in the creation of basic 3D model, grid generation
and fixing of the boundary conditions. The creation of the geometry is done
in GAMBIT, the mesh generation package of FLUENT. The partially generated grid
in GAMBIT is exported to STAR-CD for completing the mesh. In this analysis a
complete hexahedral structured mesh was created for the ports and cylinder.
For performing transient flow simulations it is necessary to use moving grids
that incorporate the piston and valves motion. The CFD package STAR-CD possesses
the above capabilities required for the simulation of transient flow cases.
Cell layer addition and removal are controlled by event and moving mesh commands.
Before starting the flow simulation, the correct valve and piston movement for
entire cycle at any crank angle position is verified from the mesh preview (Fig.
1). The entire computational work was done at IC engine Simulation laboratory.
Pentium IV processors computer was used for the computational work and it took
6 months to complete the entire work.
The engine studied in this work is a direct-injection Diesel engine. Table 1 shows the configuration of the engine . The inlet valve axis is offset from the cylinder axis by 18.5 mm in the x direction and 2.0 mm in the y direction.
|| Engine specification
|| Computational domain of the engine
RESULTS AND DISCUSSION
Effect of early injection on peak pressure and temperature: Simulations were carried out to study the effect of early injection on combustion performance and emission characteristic of the engine operating at a speed of 1000 rpm. A single step profile was considered for study. The start of injection (SOI) is 12°, 16° and 20° bTDC. Figure 2 below shows the effect of early injection on average cylinder pressure with respect to crank angle.
Figure 3 shows the variation of average cylinder temperature.
From the graph peak pressure and temperature shows an increase with advancing
the injection timing. The high peak pressure 13.8% (115 bar) and 5.9% (109 bar)
and Peak temperature 2.3% (1197 K) and 1.3% (1185K) obtained when the fuel is
injected at 20° bTDC and 16° bTDC respectively. Since the fuel is injected
at early stage of compression stroke, at this point in-cylinder temperature
and pressure is very less. It increases the ignition delay or more specifically
physical delay. Due to long ignition delay large parts of fuel made premixed
mixture and give more aggressive combustion (Zhu et al.,
2003). It results in increase in peak pressure and temperature. It is also
observed that the occurrence of peak pressure advances with early injection.
||Comparison of ignition delay, peak heat release rate and combustion
duration for injection timing
Effect of early injection on heat release rate: Figure
4 shows the variation in cylinder averaged instantaneous heat release rate.
The ignition delay, combustion duration and peak heat release rate obtained
for different injection timing are summarised in Table 2.
The early injection timing 20° bTDc and 16° bTDC shows higher peak heat
release rate a compared to base case of 12°. bTDC. Early injection timing
leads to longer ignition delay which results in accumulation of large amount
of evaporated fuel before the start of combustion. Longer ignition delay is
due to lower values of pressure and temperature inside the cylinder during the
initial period of fuel injection at advanced injection timings. The longer ignition
delay leads to rapid burning rate and the pressure and temperature inside the
cylinder rises suddenly. Hence, most of the fuel burns in premixed mode causing
higher peak heat release rate and shorter combustion duration. Whereas the baseline
case 12° bTDC the ignition delay is short causing accumulation of relatively
less amount of evaporated fuel. Shorter ignition delay is due to pressure and
temperature inside the cylinder during the initial period of fuel injection
being high. The shorter ignition delay shortens the mixing time which leads
to slow burning rate and slow rise in pressure and temperature. Hence, most
of the fuel burns in diffusion mode rather than premixed mode resulting in lower
peak heat release rate, longer combustion duration.
Effect of early injection on NOX and soot emission: Figure 5 shows the cylinder averaged NOx emission at various crank angles. From the graph it is observed that the NOx emission increases with early injection timing .In the early injection more quantity of fuel burns instantly in premixed combustion period. It gives very high combustion temperature and hence NOx formation rate increases.
Figure 6 shows the cylinder averaged soot emission at various crank angle. From the graph it is observed that the soot emission decreases with early injection timing. In the early injection combustion temperature increases which increases the oxidation reaction and hence Soot formation decreases.
Effect of exhaust gas recirculation on peak pressure and temperature:
The Start of Injection (SOI) is 20° bTDC, 10% EGR, 20% EGR and a single
step profile is considered for study.
||Cylinder average pressure
||Cylinder average temperature
||Average heat release rate
||Cylinder average NOx emission
|| Cylinder average soot emission
|| Cylinder average pressure
|| Cylinder average temperature
Figure 7 shows the effect of EGR on cylinder average combustion
Figure 8 presents the variation of cylinder temperature with EGR.
From the Fig. 8 it is observe that peak pressure and temperature decreases with increasing the EGR fraction. The peak pressure is decreased by 5.5% (109 bar)and 13.86 % (101 bar)with 10% EGR and 20% EGR respectively. The peak temperature is decreased by 4.5%(1145 K) and 9.4% (1094 K) with 10% EGR and 20% EGR respectively. This is due to the fact that, exhaust gas has a higher specific heat than air and amount of oxygen available for combustion is comparatively less in case of 10 and 20% EGR. Also, observed that start of combustion was retarded with increasing EGR fraction, due to diluting the air.
Effect of exhaust gas recirculation on NOx and soot emission:
Figure 9 and 10 show the cylinder averaged
NOx and soot emission at various crank angle. From the graph it is
observed that the NOx emission decrease with increasing EGR fraction
|| Cylinder average NOx emission
|| Cylinder average soot emission
The presence of EGR will act as heat sink and also diluting the charge, there
by reducing the combustion temperature which will reduce the formation of NOx
From the graph it can be observed that the soot emission increases with increase in EGR fraction. The presence of EGR reduces combustion temperature which decreases the oxidation reaction and hence soot formation increases.
A computational investigation of the combustion process was performed using STAR-CD. The following conclusions drawn based on this analysis,
Peak pressure increases with early injection. Peak pressure is 109 bar and
115 bar when the fuel is injected at 16° bTDC and 20° bTDC respectively.
Occurrence of peak pressure also advances with early injection. Since the combustion
duration decreases with early injection, it is an indication of HCCI mode of
Nox emission increases with early injection (without EGR) the value is 22.5 and 36.8% higher than the conventional combustion at 16° bTDC and 20° bTDC respectively.
Soot emission decreases with early injection and the value obtained is 19.7
and 44.7% lesser than the conventional combustion when the fuel is injected
at 16° bTDC and 20° bTDC, respectively.
Due to supplementation of EGR peak pressure decreases compared with early injection without EGR. Cooled EGR works as heat sink and drops the temperature avail inside the combustion chamber. It results in drop in combustion pressure. EGR also retard the occurrence of peak pressure.
Due to the presence of EGR, NOx emission decreases. From the simulation it is also found that the 10% EGR fraction simultaneously reduce the NOx and soot emission. The engine peak pressure also improved by 7% compare with base line 12° bTDC injection timing.
From the above simulation, combination of 20° bTDC injection timing and
10% EGR fraction improve the engine performance and simultaneously reduce the
NOx and soot emission.