International Journal of Heat and Fluid Flow 33 (2012) 5969

Contents lists available at SciVerse ScienceDirect

International Journal of Heat and Fluid Flow

journal homepage: www.elsevier.com/locate/ijhff

Analysis the effect of advanced injection strategies on engine performance

and pollutant emissions in a heavy duty DI-diesel engine by CFD modeling
Raouf Mobasheri a,, Zhijun Peng a, Seyed Mostafa Mirsalim b
a
b

School of Engineering and Informatics, University of Sussex, Brighton, East Sussex BN1 9QT, United Kingdom
Engine Research Center (IPCO), Amirkabir University of Technology, Tehran, Iran

a r t i c l e

i n f o

Article history:
Received 15 June 2011
Received in revised form 5 October 2011
Accepted 7 October 2011
Available online 8 November 2011
Keywords:
Multiple injections
CFD simulation
DI diesel engine
EGR
Pollutant emissions

a b s t r a c t
An Advanced CFD simulation has been carried out in order to explore the combined effects of pilot-, postand multiple-fuel injection strategies and EGR on engine performance and emission formation in a heavy
duty DI-diesel engine. An improved version of the ECFM-3Z combustion model has been applied coupled
with advanced models for NOx and soot formation. The model was validated with experimental data
achieved from a Caterpillar 3401 DI diesel engine and good agreement between predicted and measured
in-cylinder pressure, heat release rate, NOx and soot emissions was obtained. The optimizations were
conducted separately for different split injection cases without pilot injection and then, for various multiple injection cases. Totally, three factors were considered for the injection optimization, which included
EGR rate, the separation between main injection and post-injection and the amount of injected fuel in
each pulse. For the multiple injection cases, two more factors (including double and triple injections
during main injection) were also added. Results show that using pilot injection accompanied with an
optimized main injection has a signicant benecial effect on combustion process so that it could form
a separate 2nd stage of heat release which could reduce the maximum combustion temperature, which
leads to the reduction of the NOx formation. In addition, it has found that injecting adequate fuel in postinjection at an appropriate EGR allows signicant soot reduction without a NOx penalty rate.
2011 Elsevier Inc. All rights reserved.

1. Introduction
The improvement of DI diesel engines to comply with the stringent exhaust emissions standards is closely linked to continued
development of the injection systems. Traditional injector design
is often suitable for injection timings close to TDC and cannot satisfactorily meet the requirements for very early or late injection
timings. A growing trend in the diesel engine industry is towards
wider use of electronically controlled high pressure injection systems which can inject fuel at any point in the cycle without the
injection rate changing owing to injection timing or engine speed.
Multiple injections have shown to be an effective means for
reduction of pollutants emissions in diesel engines (Li et al.,
Abbreviations: ATDC, after top dead center; BTDC, before top dead center; BSFC,
brake specic fuel consumption; CFD, Computational Fluid Dynamics; DI, direct
injection; EGR, exhaust gas recirculation; EVC, exhaust valve closing; EVO, Exhaust
Valve Opening; HRR, heat release rate; IMAP, intake manifold air pressure; IMAT,
intake manifold air temperature; IVO, inlet valve opening; IVC, inlet valve closing;
NOx, oxides of nitrogen; RPM, revolutions per minute; SOI, start of injection.
Corresponding author. Address: Shawcross 2B12, School of Engineering and
Informatics, University of Sussex, Brighton BN1 9QT, United Kingdom. Tel.: +44 (0)
1273 872562.
E-mail addresses: R.Mobasheri@sussex.ac.uk (R. Mobasheri), Z.Peng@sussex.
ac.uk (Z. Peng), Mo_Mirsalim@aut.ac.ir (S.M. Mirsalim).
0142-727X/$ - see front matter 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.ijheatuidow.2011.10.004

2004; Mendez et al., 2008; Husberg et al., 2008; Badami et al.,

2002; Mobasheri et al., 2011; Shundoh et al., 1992). Multiple injections divide the total quantity of fuel into two or more injections
per combustion event. A pilot injection is also usually dened as
an injection where 15% or less of the total mass of fuel is injected
in the rst injection. Many researchers are now investigating pilot
and split injection as an effective means to simultaneously reduce
NOx and soot emissions.
The benets of multiple injections have been found to be highly
dependent on the specication of the quantity of fuel in each injection and the dwell between injections. Shundoh et al. (1992)
reported that NOx could be reduced by 35%, and smoke by 60 to
80%, without a penalty in fuel economy if pilot injection was uses
in conjunction with high pressure injection. Nehmer et al. (1994)
studied the effect of split injection in a heavy-duty diesel engine
by varying the amount of fuel in the rst injection from 10% to
75% of the total amount of fuel. They found that split injection
better utilized the air charge and allowed combustion to continue
later into the power stroke than for a single injection case, without
increased levels of soot production. Tow et al. (1994) found that
using a double injection with a relatively long dwell on a heavy
duty engine resulted in a reduction of particulate emissions by a
factor of three with no increase in NOx and only a slight increase

60

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

in BSFC compared to a single injection. Zhang (1999) used a single

cylinder HSDI diesel engine to investigate the effect of pilot injection with EGR on soot, NOx and combustion noise, and found that
pilot injection increased soot emissions. The author also showed
that reducing the amount of fuel in the pilot injection and increasing the interval between pilot and main injections could reduce the
pilot ame area when the main injection starts, resulting in lower
soot emissions.
It is well known that exhaust gas recirculation (EGR) is effective
for the reduction of NOx emissions (Ladommatos et al., 1997;
Hentschel and Richter, 1995; Ladommatos et al., 1998; Arcoumanis
et al., 1983). The application of EGR in diesel engines helps to
replace part of oxygen and nitrogen in the inlet air with carbon
dioxide and water vapor from the exhaust that have higher specic
heat capacities.
Ladommatos et al. (1997) conducted a detailed study of the effects of EGR in a 2.5 L, four-cylinder DI diesel engine. Their results
showed that the reduction in NOx emissions and the increase in
particulate emissions due to EGR could mainly be attributed to
the dilution function of residual gas to inlet charge oxygen. Hentschel and Richter (1995) investigated the formation of soot in a
1.9 L DI diesel engine and found that with increasing EGR rates,
the amount of soot formed was increase only slightly, but the
amount of soot oxidized during combustion decreased signicantly. Ladommatos et al. (1998) also observed that use of EGR
caused an increase in the ignition delay and shift in the location
of the whole combustion process further towards the expansion
stroke. This resulted in the combustion gases spending shorter
periods at high temperature, leading to lower thermal NOx formation as well as a reduced rate of soot oxidation.
Arcoumanis et al. (1983) reported that cold EGR resulted in lower NOx emissions at EGR rates below 30%, but at higher EGR rates
cold EGR seemed to offer marginally higher NOx emissions in comparison to hot EGR. As its inuence is so complicated, normally the
application of EGR must be considered to combine with other
optimizations, such as fuel injection strategy.
As mentioned earlier, multiple injections are considered as an
effective means to improve particulate emissions. Thus, it is of
interest to explore the possibility of simultaneous reduction in particulate and NOx emissions with the combined use of EGR and
multiple injections. Mikulic et al. (1993) investigated the effects
of pilot injection with EGR on engine emissions and fuel consumption and found that the lowest NOx emissions could only be
reached using a combination of EGR and pilot injection. They also
found that pilot injection in combination with EGR provided no
deterioration of fuel consumption and HC emissions. Uchida
et al. (1998) found that the combined use of pilot injection with
EGR results in little advantages for the NOx-BSFC trade-off since
smoke increased, especially under low load conditions. They argued that the smoke deterioration might be caused by the interference of the main injection sprays in a hot and higher equivalence
ratio zone near the injector nozzle. Pierpont et al. (1995) examined
the combined effects of EGR and multiple injections and achieved
signicant reductions in both NOx and soot emissions with only a
slight increase in BSFC when EGR was used in combination with
optimized double and triple injections.
Advanced injection strategies offer possible ways to improve
the mixing process which could lead to reduce both NOx and soot
emissions. In the current study, it was of interest to determine the
emission reduction capability of the combined effect of advanced
injection strategies and EGR in a DI diesel engine through Computational Fluid Dynamics (CFDs) simulation. For this purpose, three
factors have been considered for the injection optimization, which
included EGR rate, the separation between main injection and
post-injection and the amount of injected fuel in each pulse. Based
on those simulations, the optimum operating points for obtaining

the minimum amount of NOx and soot emissions have been

demonstrated.
2. Numerical procedure
2.1. CFD code and calculating meshes
The computational mesh was created using AVL ESE Diesel Tool
(ICE Physics and Chemistry, 2009). Because of the symmetrical
location of the injector at the center of the combustion chamber,
the CFD calculations are performed on 60sector meshes. Exhaust
and intake ports are not included in the computational mesh by
concentrating this simulation on in-cylinder ow and combustion
processes. Calculations begin at Intake Valve Closure (IVC) and end
at Exhaust Valve Opening (EVO). The same initial and boundary
conditions are used for all the computations. The time step used
for calculation is 0.2 deg of crank angle. The nal mesh consists
of a hexahedral dominated mesh. Exact number of cells in the
mesh was 34725 and 79311 at TDC and BDC, respectively. The
present resolution was found to give adequately grid independent
results.
2.2. The Spray Model
The standard WAVE model, described in Liu et al. (1993) was
used for the primary and secondary atomization modeling of the
resulting droplets. In this model the growth of an initial perturbation on a liquid surface is linked to its wave length and to other
physical and dynamic parameters of the injected fuel and the domain uid. Drop parcels are injected with characteristic size equal
to the nozzle exit diameter (blob injection). The Dukowicz model
was applied for treating the heat-up and evaporation of the droplets, which is described in Dukowicz (1979). This model assumes a
uniform droplet temperature. In addition, the rate of droplet temperature change is determined by the heat balance, which states
that the heat convection from the gas to the droplet either heats
up the droplet or supplies heat for vaporization. The spray wall
interaction model used in the simulations was based on the
spray-wall impingement model described in Naber et al. (1988).
This model assumes that a droplet, which hits the wall is affected
by rebound or reection based on the Weber number. The Shell
auto-ignition model was used for modeling of the auto-ignition
(Halstead et al., 1977). In this generic mechanism, 6 generic species
for hydrocarbon fuel, oxidizer, total radical pool, branching agent,
intermediate species and products were involved. In addition the
important stages of auto-ignition such as initiation, propagation,
branching and termination were presented by generalized reactions, described in Halstead et al. (1977).
2.3. The Turbulent Mixing Model
The k-e approach has been used to take account of turbulent
effects, while the complex oxidation process of diesel fuel has been
summarized by a single step irreversible reaction (ICE Physics and
Chemistry, 2009; Liu et al., 1993). The mean reaction rate has been
evaluated by means of the Coherent Flamelet Model (CFM) Colin
and Benkenida, 2004. For a diesel spray, the fuel droplets are very
close to each other and are located in a region essentially made of
fuel. After the evaporation of the fuel, an adequate time is needed
for the mixing from the nearly pure fuel region with the ambient
air. In this case, the mixing of fuel with air is modeled by initially
placing the fuel into the pure fuel zone of the ECFM-3Z model (Colin and Benkenida, 2004). A transport equation for the unmixed
fuel is solved where the source term for the transfer of fuel from
the unmixed to the mixed state can be described as follows:

61

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

sm

~FFu
y

q~ MM
q~ u MFu

~FFu is the mass fraction of unmixed fuel, MM is the mean moWhere y

lar mass of the gases in the mixed zone, MFu is the molar mass of
~ is mean density, q
~ u is the density of the unburned gases
Fuel, q
(the density of fresh gases that would be obtained if combustion
had not occurred), and sm is the mixing time.
2.4. The Combustion Model
The combustion model is based on the Coherent Flame Model
originally. The ECFM-3Z model (Colin and Benkenida, 2004; Hlie
and Trouv, 2000) distinguishes between all three main regimes
relevant in diesel combustion, namely auto-ignition, premixed
ame and non-premixed, i.e. diffusion combustion. The auto-ignition pre-reactions are calculated within the premixed charge of
fuel and air, with the ignition delay governed by the local temperature, pressure, fuel/air equivalence ratio and the amount of residual gas. Local auto-ignition is followed by premixed combustion in
the fuel/air/residual gas mixture formed during the time period
between start of injection and auto-ignition onset within the
ECFM-3Z modeled according to a ame propagation process. The
third regime is the one of diffusion combustion where the reaction
takes place in a thin zone which separates fuel and oxidizer. In the
ECFM-3Z it is assumed that the chemical time in the reaction zone
is much smaller than the time needed for the diffusion process.
Therefore the rate of reaction during diffusion combustion is determined entirely by the intermixing of fuel and oxidizer. This distinct
separation of the different ignition/combustion regimes makes the
ECFM-3Z model specically applicable to conventional as well as
alternative diesel combustion modes. In the conventional case
most part of the combustion can be assumed as diffusion type, in
the case of recently introduced alternative concepts a large amount
of fuel is consumed within premixed combustion.
2.5. The Pollutant Models
It is well known that the formation of NO depends mainly on
three different processes, the thermal NO, the prompt NO and
the fuel NO mechanism (ICE Physics and Chemistry, 2009). Usually
in automotive diesel engine applications the third one can be neglected, because there is no signicant amount of nitrogen in the
fuel. The two other mechanisms can contribute to the NO formation in engines, where mainly thermal NO is formed, but also some
amount of prompt NO can appear. The model used for this work,
covers these two contributions (ICE Physics and Chemistry, 2009).
The Hiroyasu model (ICE Physics and Chemistry, 2009; Hioyasu
et al., 1989) was also used to anticipate the soot formation. Generally, it is well accepted that the production of soot occurs in two
main phases, soot formation and soot oxidization. These processes
depend on the fuel composition, in-cylinder gas pressure, in-cylinder gas temperature, and local fuel and oxygen concentrations. The
soot formation model which has implemented in the current study
is based upon a combination of suitably extended and adapted
joint chemical/physical rate expressions for the representation of
the processes of particle nucleation, surface growth and oxidation.
3. Results and discussion
3.1. Model validation
The diesel engine used for the model validation is a single-cylinder version of a Caterpillar 3401 heavy-duty truck engine. The
engine specications are given in Table 1 (Wiedenhoefer and Reitz,
2000).

Table 1
Engine specications.
Engine type

Caterpillar 3401

Bore  stroke
Compression Ratio
Displacement
Connecting rod length
Squish clearance
IVO/IVC
EVO/EVC
IMAP
IMAT
Engine speed
Piston shape

13.719 cm  16.51 cm
15.1:1
2.44 l
26.162 cm
4.14 mm
32 ATDC/147 ATDC
134 ATDC/29 ATDC
184 kPa
310 K
1600 rpm
Mexican hat style

Table 2
Injector fuel system specications.
Injector type

Common rail

Injection pressure
Number of nozzle holes
Nozzle hole diameter
Start of injection
Injection duration
Fuel injected

Variable (up to 120 MPa)

6
0.26 mm
9ATDC
21.5 CA
0.1622 g/cycle

12

80
In-cylinder Pressure
(CFD Simulation)

10

70

In-cylinder Pressure
(Experiment)

60

Heat Release Rate

(CFD Simulation)

50

Heat Release Rate

(Experiment)

40
30

20
2

0
320

Heat Release Rate (J/deg)

In-cylindet Pressure (MPa)

_ F!M 
SFu

10
0
330

340

350

360

370

380

390

400

Crank Angle (degree)

Fig. 1. Comparison of calculated and measured in-cylinder pressure and heat
release rate.

The fuel delivery system was an electronically controlled, common rail fuel injection system (Wiedenhoefer and Reitz, 2000). In
all the injection cases studied, the same amount of fuel is injected
in each engine cycle. The main characteristics of the injection
system are listed in Table 2.
Fig. 1 shows comparisons between the predicted and measured
in-cylinder pressure and heat release rate. The result is based on
the assumption of uniform wall temperature 425 K for the cylinder
wall, 525 K for the cylinder head and 525 K for the piston top.
The trend predicted by the model is reasonably close to experimental results, although there are still some differences as can be
seen in Fig. 1. These discrepancies could be related to experimental
uncertainties in input parameters to the computations such as the
precise injection duration, start of injection timing and gas temperature at IVC.
Figs. 2 and 3 present comparisons between the predicted and
measured engine-out soot and NOx values for EGR levels of 0%
and 10%.

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

45

45

1.8
Single injection, EGR=0 %

40

40

1.6

35

35

1.4

30

1.2

25

20

0.8

30

NOx (g/kg-fuel)

NOx (g/kg-fuel)

Single injection,SOI=-9 ATDC

25
20
Experiment (EGR=0%)

15

Experiment (EGR=10%)

15

CFD Simulation (EGR=0%)

5
0
351

10

Soot (CFD Simulation)

5
391

411

431

451

471

491

Crank Angle (degree)

0.2

Soot (Experiments)

0
0

Fig. 2. Predicated NOx in comparison with measured data (Wiedenhoefer et al.,

2000).

0.4

NOx (Experiments)

CFD Simulation (EGR=10%)

371

0.6

NOx (CFD Simulation)

10

Soot (g/kg-fuel)

62

SOI (CA, BTDC)

Fig. 4. The effect of injection timing on NOx and soot, single injection, EGR = 0%.

16

Soot (g/kg-fuel)

Table 3
Computational conditions for studied cases.

Single injection,SOI=-9 ATDC

14
12

Experiment (EGR=0%)

10

Experiment (EGR=10%)
CFD Simulation (EGR=0%)

CFD Simulation (EGR=10%)

Total fuel
Pilot (SOI)
Pilot duration
Separationa
Main (SOI)
Main duration

0.1622 g/cycle
30.075 ATDC
1.075 CA
30 CA
9 ATDC
21.5 CA

a
The period between end of pilot injection and start of main
injection.

6
4
2
0
340

355

370

385

400

415

430

445

460

475

490

Crank Angle (degree)

Fig. 3. Predicated soot in comparison with measured data (Wiedenhoefer et al.,
2000).

As illustrated in Fig. 2, increasing EGR, which causes dilution of

intake charge, and insufcient oxygen in intake charge, leads to
lower combustion temperature and therefore decreases NOx emission. In contrast, as it can be seen in Fig. 3, this variation has a
reverse effect on soot formation.
While NOx and soot formation processes can be predicted but
there is only one measured value for these two components. For
further assessment of predication capability of the model, the
trade-off between NOx and soot was simulated with several different injection timing. Results shown in Fig. 4 suggested that the
models used in this study can provide enough condence to the
following simulation results with regard to the combustion process
and emissions.
It is evident from Fig. 4 that the predicted trends are fairly similar to the experimental values. In particular, they capture the
trend of reduced NOx and increasing soot with fuel injection
retard.
3.2. Modeling methodology
As mentioned earlier, careful optimization of engine operating
conditions is required to get the full benet of combined effects
of multiple injection parameters. Based the above success of
validations with a single injection, simulation results for different

multiple injection cases have presented and discussed in the following sections. Totally, 24 different injection arrangements for
which split and multiple injection cases with variable fuel amount
for each pulse (up to 30% for the second pulse) and variable separation/dwell between pulses (up to 30 CA) have considered. The
optimization were conducted separately for split injection cases
without pilot injection and then for different multiple injection
strategies accompanied with an early pilot injection. In addition,
for multiple injection cases, two more cases (including double
and triple injections during main injection) were also evaluated
which will be discussed in next section.
Table 3 shows the parameters which were xed for all injection
cases.
The injection schemes used in this study are shown schematically in Figs. 5 and 6. The same amount of fuel is injected in all
the cases considered. Based on previous research which was done
by Mobasheri et al. (2011) at this operating points, the optimum
separation for simultaneous reduction of soot with low NOx emissions was obtained by using 20 CA dwell delay between the injection pulses for split injection cases without pilot injection.
3.2.1. Inuence of split injection strategies on fuel consumption and
exhaust emissions
In this section, the results obtained for different split injection
schemes based on strategies presented in Fig. 5 are considered.
Figs. 7 and 8 show the amount of soot and NOx emission for different split injection cases with 0% and 10% EGR rate. The labeling
scheme for the split injection cases gives the percent of the fuel
injected in the rst and last pulses, and the dwell between two
injections. For instance, 70(10)30 represents 70% fuel injected in
the rst pulse, 10 crank angle degree dwell between the two injection pulses and 30% fuel in the second pulse.

63

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

Injection Duration= 21.5 CA

Single Injection

100%
SOI
15.05 CA
70%

17.20 CA

30%

10-20-25-30 CA

80%

19.35 CA

4.30 CA
20%

10-20-25-30 CA

90%

9 CA BTDC

6.45 CA

10-20-25-30 CA

2.15 CA
10%

Crank Angle (degree)

Fig. 5. Injection proles for different split injection cases without pilot injection.

Injection Duration= 21.5 CA

Single Injection

100%
SOI
Pilot Injection

SOI
Main Injection
15.05 CA

10-20-25-30 CA

70%

5%

17.20 CA

25%

3.22 CA
10-20-25-30 CA

80%

5%

5.37 CA

2.15 CA

18.27 CA
85%

5%

30.075 CA BTDC

9 CA BTDC

15%

10-20-25-30 CA

10%

Crank Angle (degree)

Fig. 6. Injection proles for different multiple injection cases.

As illustrated in Figs. 7 and 8, some split injection schemes can

reduce NOx emissions signicantly, while some can increase NOx
emissions. It can be seen in Figs. 7 and 8 that the emission production histories of the split injections are changed signicantly from
the original single injection cases and the majority of split injection
schemes can simultaneously reduce soot and NOx emissions
compared to traditional single injection scheme, though some
cases have increased emissions. If considering all data in Figs. 7
and 8, it can be seen that the split injection strategy under 10%
EGR conditions can be more benecial for the substantial reduction
of NOx formation. As shown in Fig. 7, the minimum amount of NOx
formation was achieved with the case of 70(25)30, though it is just
a little lower than other several operating points. It may be due to
the fact that premixed combustion which is the main source of the
NOx formation is relatively low in comparison with other cases.
Higher amounts of the second injection pulse into the lean and

hot combustion zones cause the newly injected fuel to burn rapidly
and effectively at high temperature, resulting in high soot oxidation rates. The optimum engine performance for reduction of soot
and NOx emissions can be obtained with 20 CA delay between
injection pulses in the 80(20)20 and 90(20)10 cases, though the
lowest total soot is seen with the split injection ratio 90(25)10.
As can be seen in Figs. 7 and 8, the delay dwell does not affect
soot formation signicantly. The combustion of 30% fuel in the second injection pulse only causes a small effect of soot variations
compared to the other cases in this injection category i.e.
70(x)30. It is clearly seen in Fig. 9 that the 90(20)10 case shifts
the soot-NOx trade-off to the optimum level.
Figs. 9 and 10 show BSFC vs. NOx curves at 0% and 10% of EGR.
As shown in Figs. 9 and 10, for the 90(x)10 case the differences
between BSFC vs. NOx emission is lower than other cases. It can be
concluded that the split injection shows minimal effects on BSFC

64

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

0.3

0.6
70 (x) 30

0.55

0.45

Single Injection

(25)
(30)

(25)

Single inj.

(25)

0.4
0.35
(10)

(30)

0.3

(20)

80 (x) 20

0.28

90 (x) 10

BSFC (kg/kW-hr)

(20)

(30)

80 (x) 20

0.5

Soot (g/kg-fuel)

70 (x) 30

0.29
(10)

(10)

90 (x) 10

0.27
(25)
(20)

(20)

0.25

(30)
(25)

0.24

0.25

Single Injection

(30)

0.26

(10)

(20)

(30)

0.2

Single Injection

0.23

(25)

EGR= 0 %

EGR= 0 %

0.22

0.15
30

32

34

36

38

40

42

44

46

48

50

52

30

54

32

34

36

38

40

42

44

46

48

50

52

54

NOx (g/kg-fuel)

NOx (g/kg-fuel)

Fig. 9. BSFC vs. NOx trade-off, split injection, EGR = 0%.

Fig. 7. Soot-NOx trade-off, split injection, EGR = 0%.

0.32

1.5

0.31
1.3

(20)
(10)

(30)

(20)

0.7

Single Inj.

(10)

BSFC (kg/kW-hr)

Single Injection

0.9

70 (x) 30
(30)

(20)

0.5

0.29

(20)

0.26

(20)

90 (x) 10

18

19

20

21

22

23

(10)

Single Injection

EGR= 10 %

0.22
14

17

(10)

(30)
(25)

(10)

0.25

0.3
16

Single Injection

(25)

0.27

0.23

(25)

15

90 (x) 10

(20)
(30)

0.28

0.24

80 (x) 20

EGR= 10 %

14

80 (x) 20

(25)

1.1
(25)

70 (x) 30

(30)

0.3

(10)
(25)

(30)

Soot (g/kg-fuel)

(10)

(10)

(20)

15

16

17

24

18

19

20

21

22

23

NOx (g/kg-fuel)

NOx (g/kg-fuel)

Fig. 10. BSFC vs. NOx trade-off, split injection, EGR = 10%.

Fig. 8. Soot-NOx trade-off, split injection, EGR = 10%.

0.55

when the secondary injection were relatively small compared to

the main injection. In addition, as can be seen the trade-off characteristics for the 90(x)10 case is relatively different for two level of
EGR rates.

Single Injection

0.45

Soot (g/kg-fuel)

3.2.2. Inuence of multiple injection strategies on fuel consumption

and exhaust emissions
Benets of split injection for emission reduction were discussed
in previous section. In order to fully investigate the potential of
multiple injection strategies, the effects of pilot injection followed
by various main and post-injection schemes based on strategies
presented in Fig. 6 are considered in this section.
Figs. 11 and 12 show the amount of soot and NOx emissions for
different multiple injection cases for EGR levels of 0% and 10%,
respectively.
As illustrated in Figs. 11, for multiple injection schemes both
NOx and soot emissions decreased compared to split injection
schemes in majority of cases. The pilot injection, which was set at
about 30.075 CA BTDC, reduces the ignition delay and therefore
the amount of premixed combustion, leading to lower temperatures and NOx emissions. The results of Fig. 12 conrm EGRs
effectiveness at reducing NOx. In addition, Fig. 12 shows the

0.5

(10)

0.4
(30)

0.35

(25)

Single inj.
(20)

(30)

0.3

(10)

(20)

(10)

(30)

5 (20) 75 (x) 20

(25)

0.2

5 (20) 80 (x) 15

EGR= 0 %

0.15

5 (20) 65 (x) 30

(20)

(25)

0.25

0.1
29

31

33

35

37

39

41

43

45

47

49

NOx (g/kg-fuel)
Fig. 11. Soot-NOx trade-off, multiple injection, EGR = 0%.

effectiveness of multiple injections at controlling soot emission under EGR conditions. It can be concluded that by using multiple
injections the soot formation is accrued in the multiple regions in

65

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

1.3

0.3

1.2

(30)

(10)
(20)

Single Injection

0.28
(30)

(10)

BSFC (gr/kW-hr)

Soot (g/kg-fuel)

(25)

1
0.9

(30)

0.8

Single Inj.

(25)

0.7

(10)

(20)

5 (20) 65 (x) 30

0.6
5 (20) 75 (x) 20

(20)

(30)

0.5

5 (20) 75 (x) 20

(30)
(20)

0.27
(25)

(10)

0.25

(25)
(20)

(10)

0.23

0.3
13

(25)

(30)

(20)

EGR= 10 %

12

5 (20) 80 (x) 15

5 (20) 80 (x) 15

EGR= 10 %

14

15

16

17

18

19

20

21

22

(10)

0.26

0.24

(25)

0.4

5 (20) 65 (x) 30

0.29

1.1

0.22

NOx (g/kg-fuel)

12

13

14

15

16

17

18

19

20

21

22

23

NOx (g/kg-fuel)

Fig. 12. Soot-NOx trade-off, multiple injection, EGR = 10%.

Fig. 14. BSFC vs. NOx trade-off, multiple injection, EGR = 10%.

0.28
5 (20) 65 (x) 30

60

(30)

0.27

EGR= 0 %

5 (20) 75 (x) 20

Rate of Heat Release (J/deg)

BSFC (gr/kW-hr)

(30)

0.26
(20)

(25)

(10)

0.25

5 (20) 80 (x) 15

(25)
(20)
(30)

(10)

0.24

(25)
(10)

0.23

(20)

EGR= 0 %

90 (20) 10

50

80 (20) 20
5 (20) 75 (25) 20

40

5 (20) 80 (30) 15

30
20
10

0.22
29

31

33

35

37

39

41

43

45

47

49

51

0
340

NOx (g/kg-fuel)

360

380

400

420

440

Crank Angle (degree)

Fig. 13. BSFC vs. NOx trade-off, multiple injection, EGR = 0%.
Fig. 15. The HRR curve, optimum injection cases, EGR = 0%.

60
EGR= 10 %

Rate of Heat Release (J/deg)

the combustion chamber and thus has more area for oxidation.
Finally, the fuel that is pulsed into the combustion chamber after
main injection ignites rapidly and thus will not contribute signicantly to soot formation in high temperature rich regions. Even
though the EGR reduces some of the intake oxygen content, the heat
added to the intake air enhances the soot oxidation to some extent
which leads to reduction of soot emission. When the percentage of
the second pulse injected fuel is larger than 75% of the total fuel, the
NOx formation history of the multiple injection has a more impact
to simultaneous reduction of Soot and Nox emissions. This trend
has also observed when 10% EGR is used. It can be also concluded
that the NOx chemistry is sensitive to the early combustion details
because these combustion products stay at a high temperature for
the longest time, and the combustion region is not cooled by the
vaporization of the continuously injected fuel that occurs in the single injection case.
Figs. 13 and 14 show BSFC vs. NOx curves at 0% and 10% of EGR
for different multiple injection cases.
Approximately the same trend of overall reduction of NOx
emission and increase of BSFC could be observed in different cases,
as illustrated in Fig. 13 and 14, although this trend is different for
20 CA dwell in 5(20)85(x)10 cases. From these results, it can be
summarized that the optimum engine performance for reduction
of soot and NOx emissions can be obtained with 25 CA and 30
CA delay between main and post-injection pulses in the
5(20)75(25) and 5(20)80(30)15 cases, respectively.

90 (20) 10

50

80 (20) 20
5 (20) 75 (25) 20

40

5 (20) 80 (30) 15

30

20

10

0
340

360

380

400

420

440

Crank Angle (degree)

Fig. 16. The HRR curve, optimum injection cases, EGR = 10%.

Figs. 15 and 16 illustrate the heat release rates for optimum

split and multiple injection cases for EGR levels of 0% and 10%,
respectively.

66

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

1800
1700

EGR= 0 %

1600

Temperature (k)

1500
1400
1300
1200

90 (20) 10

1100
80 (20) 20

1000
900

5 (20) 75 (25) 20

800
5 (20) 80 (30) 15

700

600
330 340 350 360 370 380 390 400 410 420 430 440

Crank Angle (degree)

Fig. 17. In-cylinder temperature, optimum injection cases, EGR = 0%.

1800
1700

EGR= 10 %

1600

Temperature (k)

1500
1400
1300
1200

90 (20) 10

1100
80 (20) 20

1000
900

5 (20) 75 (25) 20

800
5 (20) 80 (30) 15

700

600
330 340 350 360 370 380 390 400 410 420 430 440

Crank Angle (degree)

Fig. 18. In-cylinder temperature, optimum injection cases, EGR = 10%.

Single Injection

(80) 20 (20)

As shown in Figs. 15 and 16, the second fuel injection, occurred

at the late combustion stage, affects the in-cylinder pressure and
temperature that causes second peak in HRR diagram. In addition,
the amount of injected fuel in each pulse and the delay between
injections strongly affect the timing and magnitude of the second
peak. As illustrated in Figs. 15 and 16, the main combustion event
usually has a short auto-ignition delay for multiple injection cases
due to the high in-cylinder temperature produced by pre-combustion resulted of pilot injection. It can be seen that the second peak
is signicantly moved toward the expansion stroke for the
5(20)80(30)15 case. On the other hand, multiple injections is found
to reduce NOx emission signicantly since it reduces the magnitude of the combustion peak.
Figs. 17 and 18 show the cylinder temperature for optimum
split and multiple injection cases for EGR levels of 0% and 10%,
respectively.
As can be seen in Figs. 17 and 18, for the 5(20)80(30)15 case, the
second peaks are lower than the other cases for both EGR rates.
Moreover, for the 70(20)30 case, the rst peak are lower than the
other cases. In addition, after the second peak, the cylinder temperature tends to increase more in comparison with the other cases.
Fig. 19 shows the velocity eld contours for single injection case
in comparison with the three optimum injection cases at 360,
385 and 410 CA.
As can be seen in Fig. 19, the velocity eld within the cylinder
increases dramatically for the three optimum injection cases in
comparison with the single injection case at 410 CA especially
for 5(20)80(30)15 case. It can be concluded that multiple injection
had a signicant effect on ow led and causes the subsequent effects on soot oxidization and NOx formation.
The NOx distribution for the three optimum injection cases
compared to single injection case are shown in Fig. 20 at 370,
385 and 400 CA. Fig. 21 shows the comparison of in-cylinder soot
formations at same operating points.
As can be seen in Figs. 20 and 21, the local soot-NOx trade-off is
evident in these contour plots, as the NOx formation and soot formation occur on opposite sides of the high temperature region. It is
widely reported that the combustion of single injection caused the
rapid premixed combustion phases, because most fuel is injected

5 (20) 80 (30) 15

5 (20) 75 (25) 20

Fig. 19. The velocity elds contours, Single injection case in comparison with three optimum injection cases at 360, 385 and 410 CA.

67

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

Single

(80) 20 (20)

5 (20) 80 (30) 15

5 (20) 75 (25) 20

Fig. 20. NOx mass fraction contours, Single injection case in comparison with three optimum injection cases at 370, 385 and 400 CA.

Single

(80) 20 (20)

5 (20) 80 (30) 15

5 (20) 75 (25) 20

Fig. 21. Soot mass fraction contours, Single injection case in comparison with three optimum injection cases at 370, 385 and 400 CA.

during the ignition delay period under high ambient pressure and
temperature conditions and, thus, is combusted immediately. For
this reason, undiluted airfuel mixtures and fuel-rich region exist
locally in the combustion chamber, which usually causes the formation of harmful exhaust emissions and combustion noises. In
the single injection case, the soot formed in the later combustion
phase is difcult to oxidize for two reasons. First, it is close to
the end of the combustion period, and second, the temperature
decreases rapidly in expansion stroke. In the same manner, the
soot produced during the main combustion phase will not be oxidized easily for the lower temperature in-cylinder. It can be seen
that for optimum injection cases, NOx and soot mass fractions
are lower in comparison with the single injection case. It can be
concluded that, for the split injection case, the second pulse
injected fuel enters into a relatively lean and high temperature region which is remained from the combustion of the rst pulse. Soot

formation is therefore signicantly reduced because the injected

fuel is rapidly consumed by combustion before a rich soot region
can accumulate.
The temperature distributions at two crank angle degrees for
three optimum injection cases compared to single injection case
are shown in Fig. 22.
As can be seen in Fig. 22, at 410 CA the maximum temperature
in optimum cases has a higher amount than single injection case. It
can be concluded that injecting adequate fuel in post-injection
leads to the increase of temperature in late stage of combustion
process that allows soot reduction without a NOx penalty rate.
3.2.3. Using double and triple injections for main injection
The previous section has shown the potential of different multiple injection cases to reduce NOx and soot emissions. In this section, the main injection has divided in two and three pulses to

68

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

Single

(80) 20 (20)

5 (20) 80 (30) 15

5 (20) 75 (25) 20

Fig. 22. Temperature contours, Single injection case in comparison with three optimum injection cases at 385, 400.

SOI
Main Injection

SOI
Pilot Injection

4 CA

8.6 CA

8.6 CA
25 CA
40%

40%

5%

4 CA

5%

30.075 CA BTDC

3.225 CA
15%

4 CA

5.375 CA

5.375 CA

25%

25%

9 CA BTDC

5.375 CA

20 CA

25%

4.3 CA
20%

CrankAngle (degree)

Fig. 23. Injection proles for two multiple injection cases with double and triple main injections.

50
EGR= 0 %

5 (20) 25 (4) 25 (4) 25 (20) 20

5 (20) 40 (4) 40 (25) 15

1500

40

1400

35

1300

30

1200

25

1100

20

1000

15

900

10

800

700

0
340

350

360

370

380

390

400

410

420

430

Temperature (k)

Rate of Heat Release (J/deg)

45

Table 4
Soot, NOx and BSFC for two multiple main injection cases.

1600

600
440

Crank Angle (degree)

Fig. 24. Heat release rate and temperature for two multiple main injection cases.

Case

Soot (g/kg-fuel)

NOx (g/kg-fuel)

BSFC (g/kw h)

5(20)25(4)25(4)25(20)20
5(20)40(4)40(25)15

0.242
0.2311

28.43
30.21

0.2574
0.2751

explore its effects for more reduction of pollutant emissions. For

this purpose, two more injection schemes, as shown in Fig. 23,
has been proposed and considered based on optimum cases which
were obtained in last section.
Figs. 24 shows the heat release rate and temperature curves
based on strategies illustrated in Fig. 23.
It can be seen that due to double and triple injection during
main injection, the peak of HRR and temperature diagram is lower
than multiple injection schemes which were previously
considered.
The amount of BSFC, NOx and Soot emission for these cases are
summarized in Table 4.

R. Mobasheri et al. / International Journal of Heat and Fluid Flow 33 (2012) 5969

0.03
70 (20) 30

EGR= 0 %

80 (20) 20

CO (Mass Fraction)

0.025

5 (20) 75 (25) 20
5 (20) 80 (30) 15
5 (20) 25 (4) (25) (4) (25) 20 (20)

0.02

69

 Employing a post-injection combined with a pilot injection

results in reduced soot formation from diffusion combustion
and enhances the soot oxidation process during the expansion
stroke, resulting in decreased soot emissions, while the NOx
concentration is maintained in low levels.

5 (20) 40 (4) 40 (25) 15

Single Injection

0.015

Acknowledgments
The authors gratefully acknowledge the AVL Company to provide computational resources for this research.

0.01
0.005
0
340

References
360

380

400

420

440

460

480

Crank Angle (degree)

Fig. 25. Effects of different injection strategies on CO emissions, EGR = 0%.

It can be concluded from Table 4 that two proposed injection

cases can be more benecial for the substantial reduction of NOx
and Soot emissions, however the amount of BSFC in these cases
should be considered as a main disadvantage.
The effects of optimum injection strategies on CO emissions are
shown in Fig. 25.
It can be seen that the concentration of CO emissions for two
optimum
injection
case
including
5(20)80(30)15
and
5(20)75(25)20 are generally lower than other cases. The reason
for lower CO emissions can be considered to be that spray has relatively better fuel atomization and airfuel mixing, as well as a
more complete combustion characteristic, which can be explained
by the faster evaporation and vaporization of fuel spray droplets.
4. Conclusions
The effect of different multiple injection strategies on the
improvement of fuel atomization and the reduction of exhaust
emission characteristics was analyzed on a DI diesel engine. These
results were compared with those obtained from the single injection case at same operating points. The conclusions are summarized as follows:
 The study conrms the benet of combining EGR and multiple
injections as a benecial tool to control both NOx and soot
emissions simultaneously.
 Although EGR is effective at reducing NOx by lowering peak incylinder temperatures, there is a substantial trade-off in
increased soot emissions due to increased high temperature
rich regions. By using multiple injections, the amount of soot
formed in these regions is reduced considerably. Multiple injection schemes improve fuelair mixing and lean out the in-cylinder mixture, thus reducing the high soot forming regions.
 Compared to the single injection, split injection was very effective for reducing NOx and Soot emissions. However, the split
injection must be optimized for best emission reducing effects
by varying the fuel distribution in each pulse and the separation
between pulses for each operating condition.
 Investigation on multiple injection strategies showed the soot
level can be dramatically reduced if an early pilot injection is
combined with a main injection.

Arcoumanis, C., Bicen, A.F., Whitelaw, J.H., 1983. Squish and Swril-Squish
interaction in motored model engines. ASME J. Fluids Eng. 105, 105.
Badami, M., Millo, F., Mallamo, F., Rossi, E.E., 2002. Inuence of Multiple Injection
Strategies on Emissions, Combustion Noise and BSFC of a DI Common Rail
Diesel Engine. SAE paper 2002-01-0503.
Colin, O., Benkenida, A., 2004. The 3-zones extended coherent ame model
(ECFM3Z) for computing premixed/diffusion combustion. Oil Gas Sci. Technol.
Rev. IFP 59 (6), 593609.
Dukowicz, J.K., 1979. Quasi-Steady Droplet Change in the Presence Of Convection.
Informal Report Los Alamos Scientic Laboratory, LA7997-MS.
Halstead, M., Kirsch, L., Quinn, C., 1977. The auto ignition of hydrocarbon fueled at
high temperatures and pressures-tting of a mathematical model. Combust.
Flame 30, 4560.
Hlie, J., Trouv, A., 2000. A modied coherent ame model to describe turbulent
ame propagation in mixtures with variable composition. Proc. Combust. Inst.
28, 193201.
Hentschel, W., Ritcher, J., 1995. Time-Resolved Analysis of Soot Formation and
Oxidation in a Directed-Injection Diesel Engine for Different EGR-Rates by an
Extinction Method. SAE Paper 952517.
Hioyasu, H., Nishida, K., 1989. Simplied Three Dimensional Modeling of Mixture
Formation and Combustion in a DI Diesel Engine. SAE Paper 890269.
Husberg, T., Denbratt, I., Karlsson, A., Analysis of Advanced Multiple Injection
Strategies in a Heavy-Duty Diesel Engine using Optical Measurements and CFDSimulations. SAE Paper 2008-01-1328.
ICE Physics & Chemistry, 2009. AVL FIRE user Manual v.2009. 1, 2009.
Ladommatos, N., Abdelhalim, S.M., Zhao, Hu, Z., 1997. The Dilution, Chamical, and
Thermal Effects of Exhaust Gas Recirculation on Diesel Engine Emission
Part4:Effects of Carbon Dioxide and Water Vapour. SAE Paper 971660.
Ladommatos, N., Abdelhamin, S.M., Zhao, H., Hu, Z., 1998. Effects of EGR on Heat
Release in Diesel Combustion. SAE Paper 980184.
Li, T., Nishida, K., Zhang, Y., Yamakawa, M., Hiroyasu, H., 2004. An Insight into Effect
of Split Injection on Mixture Formation and Combustion of DI Gasoline Engines.
SAE Paper 2004-01-1949.
Liu, A.B., Reitz, R.D., 1993. Modeling the Effects of Drop Drag and Break-up on Fuel
Sprays. SAE Paper 930072.
Mendez, S., Thirouard, B., 2008. Using Multiple Injection Strategies in Diesel
Combustion: Potential to Improve Emissions, Noise and Fuel Economy TradeOff in Low CR Engines. SAE Paper 2008-01-1329.
Mikulic, L., Kuhn, M., Schommers, J., Willing, E., 1993. Exhaust Emission
Optimization of DI Diesel Passenger Car Engine with High-Pressure Fuel
Injection and EGR. SAE Paper, 931035.
Mobasheri, R., Peng, Z., Mirsalim, S.M., 2011. CFD Evaluation of Effects of Split
Injection on Combustion and Emissions in a DI Diesel Engine. SAE Paper 201101-0822.
Naber, J.D., Reitz, R.D., 1988. Modeling Engine Spray/Wall Impingement. SAE Paper
880107.
Nehmer, D.A., Reitz, R.D., 1994. Measurement of the Effect of Injection Rate and Split
Injections on Diesel Engine Soot and NOx Emissions. SAE Paper 940668.
Pierpont, D.A., Reitz, R.D., 1995. Effects of Injection Pressure and Nozzle Geometry
on D.I. Diesel Emissions and Performance. SAE Paper 950604.
Shundoh, S., Komori, M., Tsujimura, K., Kobayashi, S., 1992. Nox Reduction from
Diesel Combustion Using Pilot Injection with High Pressure Fuel Injection. SAE
paper 920461.
Tow, T.C., Pierpont, A., Reitz, R.D., 1994. Reducing Particulate and NOx emissions by
Using Multiple Injections in a Heavy Duty D.I. Diesel Engine. SAE Paper 940897.
Uchida, N., Shimokawa, K., Kudo, Y., Shimoda, M., 1998. Combustion Optimization
by Means of Common Rail Injection System for Heavy-Duty Diesel Engines. SAE
Paper 982679.
Wiedenhoefer, J.F., Reitz, R.D., 2000. Modelling the effect of EGR and multiple
injection schemes on I.C. engine component temperatures. Numer. Heat
Transfer, Part A: Appl. 37 (7), 673694.
Zhang, Long, 1999. A Study of Pilot Injection in a DI Diesel Engine. SAE paper 199901-3493.