Fuel 106 (2013) 98–105

Contents lists available at SciVerse ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Numerical study of combustion and emission characteristics of dual-fuel

engines using 3D-CFD models coupled with chemical kinetics
Amin Maghbouli a, Rahim Khoshbakhti Saray b,⇑, Sina Shafee c, Jafar Ghafouri d
a
Department of Mechanical Engineering, Faculty of Engineering, National University of Singapore, Singapore
b
Mechanical Engineering Faculty, Sahand University of Technology, Tabriz, Iran
c
Department of Mechanical Engineering, Faculty of Natural and Applied Sciences, Middle East Technical University, Ankara, Turkey
d
Azad University of Tabriz, Azad University Departments, Tabriz, Iran

h i g h l i g h t s

" Zheng and Yao’s mechanism can capture spray combustion of fuel rich regions.
" Cylinder peak pressure is delayed by shifting diesel to dual-fuel combustion mode.
" IMEP and ITE are increased by applying higher amount of pilot fuel.
" Higher NOx and CO levels are resulted by increasing pilot fuel quantity.

a r t i c l e i n f o a b s t r a c t

Article history: Dual-fuel combustion provides a relatively easy and inexpensive alternative to conventional diesel
Received 23 February 2011 engine combustion by drastically reducing fuel consumption with comparable performance characteris-
Received in revised form 29 September 2012 tics. Accurate simulation of the dual-fuel combustion requires utilization of a detailed chemistry com-
Accepted 23 October 2012
bined with a flow simulation code. In the present study, the combustion process within the diesel and
Available online 20 November 2012
diesel/gas dual-fuel engine is investigated by use of a coupled 3D-CFD/chemical kinetics framework. In
this study, methane and n-heptane are used as representatives of the natural gas and diesel fuels. The
Keywords:
multi-dimensional KIVA-3V code, with modified combustion and heat transfer models, incorporates a
Dual-fuel
KIVA
chemical kinetics mechanism for n-heptane and methane oxidation chemistry. The source terms in
Combustion energy and species conservation equations due to chemical reactions are calculated by integrating the
Chemical kinetics CHEMKIN chemistry solver into the KIVA code. The model is applied to simulation of a medium duty
Pilot fuel dual-fuel converted diesel engine. A chemical kinetics mechanism which consists of 42 species and 57
reactions is used for prediction of n-heptane oxidation chemistry. Simulation of dual-fuel combustion
is performed using the same mechanism with addition of a series of major methane oxidation pathways.
The results show that Zheng and Yao’s n-heptane mechanism which had been previously validated in
their work, can model the diesel and dual-fuel combustion, where fuel-rich zones are present. The pre-
dictive model of this study is validated using available published experimental data. Results show that
pressure and ignition delay predictions are in good agreement with experiments. Based on constant total
mixture input energy in dual-fuel combustion, increasing pilot fuel amount leads to shorter ignition delay
and peak pressure increment. It is found that concentrations of NOx and CO emissions tend to increase at
higher pilot fuel injection quantities.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction seem to completely take the role of internal combustion engines in

Although cutting edge studies on alternative power generation
units and technologies such as fuel-cells and hybrid-automobiles
are carrying significant amount of research projects, they do not
⇑ Corresponding author. Tel./fax: +98 412 3459065.
E-mail addresses: mpeaminm@nus.edu.sg (A. Maghbouli), khoshbakhti@sut.ac.ir
(R.K. Saray), sina.shafee@metu.edu.tr (S. Shafee), gh_jafar@yahoo.com (J. Ghafouri).
the near future. Therefore a considerable deal of research is concen-
trated on optimization of conventional engines. Based on ongoing
numerical calculations and recent achievements in practical appli-
cations, dual-fuel compression ignition engines are nominated to
meet requirements of the strict emission regulations. In practice,
the dual-fuel engines are conversion of conventional diesel engines
while NG provides majority of chemical energy. Injection of diesel as
pilot fuel is used here to ignite NG and air mixture [1,2].

0016-2361/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.fuel.2012.10.055
 A. Maghbouli et al. / Fuel 106 (2013) 98–105 99

Nomenclature

A0 fluid flow constant ATDC After Top Dead Center

D diffusion coefficient CAD Crank Angle Degree
J heat flux CFD Computational Fluid Dynamics
M number of reactions in chemical kinetics mechanism CI Compression Ignition
P pressure CO Carbon Monoxide
Q_ c source term due to chemistry CO2 Carbon Dioxide
Q_ s source term owing to spray CTC Characteristics Time Combustion
t time D.I Direct Injection
u velocity vector DVODE Double-precision Variable-coefficient Ordinary Differ-
W molecular weight ential Equation solver
EDC Eddy Dissipation Concept
Subscripts EGR Exhaust Gas Recirculation
m number of species in chemical kinetics mechanism EVO Exhaust Valve Open
HCCI Homogeneous Charge Compression Ignition
Greek symbols HRR Heat Release Rate
dm‘ dirac delta function IMEP Indicated Mean Effective Pressure
Dh0f heat of formation of species m at absolute zero ISEC Indicated Specific Energy Consumption
q density ITE Indicated Thermal Efficiency
q_ cm density change rate due to chemistry IVC Inlet Valve Closure
q_ s density change rate owing to Spray NG Natural Gas
e dissipation rate of turbulence NOx Nitrogen Oxides
x_ m molar production rate SI Spark Ignition
SOI Start of Injection
Abbreviations
AHR Accumulative Heat Release

Nowadays, with the advent of powerful computers with high
computational power, numerical modeling has proved to be a
strong tool having flexibility and low cost. Generally speaking,
the mathematical framework for engine modeling can be classified
into two main groups: thermodynamic and dimensional models.
Thermodynamic models assume energy equation as the main con-
servation equation and are subdivided into single and multi-zone
models [3–5]. The physical processes such as turbulent flow field
effect are not modeled directly in thermodynamic models and in-
stead, empirical correlations with numerous constants are used.
In dimensional models, temporal and spatial variations of the flow
filed, pressure, composition, temperature and turbulence inside
combustion chamber are accounted [6]. Variety of sub-models
are applied in the dimensional models to simulate different aspects
of in-cylinder processes. Among these, the combustion model has
been found to be the most important one since it is directly related
to in-cylinder pressure and HRR. Coupling 3D-CFD flow simula-
tions with chemical kinetics mechanisms, as advanced combustion
models, are needed in establishing flexible simulation workstation
of existing CI, SI, dual-fuel and HCCI engines.
By reference to extensive research on modeling of diesel
engines, there are quantitatively limited publications on 3D-CFD
simulation of dual-fuel engines. Therefore, further numerical sim-
ulations of dual-fuel engines are needed. Most of the stud ies in en-
gine simulation are conducted to model the combustion process
because of its complex nature, such as work by Kong et al. which
developed a multi-dimensional model for simulation of diesel en-
gine with SHELL and CTC models used as the ignition and combus-
tion models, respectively [7]. Agarwal and Assanis developed a
multi-dimensional model by the using of a detailed chemical kinet-
ics mechanism of methane [8]. They also studied the effect of eth-
ane and hydrogen peroxide on fuel consumption rate of a D.I
natural gas engine. Results were only validated for temperatures
below 1300 K since the mixing effect on combustion was not mod-
eled. Kusaka and his colleagues developed a multi-dimensional
model to study combustion and emission characteristics of a
dual-fueled diesel engine and improved performance and
emissions characteristics by applying EGR and intake heating
simultaneously [9]. Hong et al. implemented EDC as a turbulent
combustion model, taking into account the effects of mixing-con-
trolled combustion in multi-dimensional calculations of a D.I nat-
ural gas engine [10]. Significant radical growth was used as
transition parameter between the ignition and combustion. Liu
and Karim developed Partially Stirred Reactor (PaSR) combustion
model which considered both chemistry and mixing effects in
modeling of diesel and dual-fuel engines with swirl chamber
[11]. The results revealed that swirl chamber enhances injected
fuel evaporation, mixing and combustion in both diesel and dual-
fuel operating modes. It should be noted that all multi-dimensional
researches mentioned above were accomplished via KIVA code
[12].
In order to study the combustion process within diesel and
dual-fuel engines, a coupled 3D-CFD/Chemistry model is first pre-
sented in this paper and then model predictions are validated
using measured in-cylinder pressure histories. The developed
model is used to study the combustion and emission characteris-
tics of the engine under study at different pilot fuel quantities.
2. Model description
In the combustion systems with presence of fuel spray, a num-
ber of processes such as fuel droplet break up, oscillation, collision,
evaporation and mixing with oxidizer are needed for initiation of
combustion. This causes formation of number of regions with dif-
ferent air/fuel ratio around the liquid fuel spray, where liquid fuel
is surrounded by fuel rich regions. Such a distribution of fuel and
oxidizer mixtures forms stratified charge inside the combustion
chamber. In combustion simulations of D.I systems, the effects of
stratified fuel and oxidizer distribution and its preparation should
be taken into account. This can be done by introducing a delay
parameter in preparation rate of homogenous fuel and oxidizer
mixture in the species production rate formulation by the use of
eddy turn-over time [7] and/or use of sub-grid scale mixing effects
100 A. Maghbouli et al. / Fuel 106 (2013) 98–105

[13]. In addition, mixing-controlled models are applied to consider q_ cm ¼ W m x_ m ð3Þ

the time delay between fuel and oxidizer mixture preparation and
combustion to reduce unrealistic high HRR predictions due to igni- X
M

tion of fully premixed fuel and oxidizer [14]. Q_ c ¼  x_ m ðDh0f Þm ð4Þ

The chemical kinetics mechanism of n-heptane which was pur- m¼1

posively reduced by Zheng and Yao [15] is used in this study. This
mechanism includes 42 chemical species and 58 elementary reac-
tions and was derived from Lawrence Livermore National Laboratory
(LLNL) [16] detailed n-heptane mechanism. It was successfully
used in modeling combustion characteristics of HCCI engines with
low D.I ratios [15]. The advantage of using this mechanism is its ex-
tended equivalence ratio validity range so that combustion in fuel
rich regions can be taken into account sufficiently. While generat-
ing a reduced reaction mechanism, if C7H15-3 was only selected as
the product of H-atom abstraction among four distinct heptyl radi-
cals, there would be only one reaction path from C7H15-3 to C7H14 -2
through C7H15-3 + O2 = C7H14-2 + HO2 reaction. However, this reac-
tion competes with the first O2 addition to heptyl radicals, which
inhabits and underestimates the low temperature branching when
the mixture is rich [15]. This will cause simulation failure espe-
cially in modeling diesel engine spray combustion, where fuel rich
local regions are available around spray cone. To remedy this issue,
series of reactions that can effectively expand the range of chemis-
try predictions for wide range of equivalence ratio were introduced.
Therefore, reactions: C7H13 = C3H5-A + C4H8-1, C7H14-2 + OH =
C7H13 + H2O and C4H8-1 = C3H5-A + CH3 were added to the new re-
duced mechanism [15]. These three reactions introduce C7H14-2
and C3H5-A to the chemical kinetics mechanism. Moreover, the
generation path of C7H14-2 and consumption path of C3H5-A was
included in the new reduced mechanism. In dual-fuel engine
simulation, pilot fuel is injected into premixed mixture of NG
and air. Moreover, NG is assumed to be 100% methane to simplify
the calculations. As methane oxidation chemistry was not included
in this mechanism, series of major methane oxidation pathways
are added from the reduced chemical kinetic mechanism devel-
oped at the University of Wisconsin Engine Research Center
(ERC) [17,18].
Accurate simulation of dual-fuel combustion is highly depen-
dent on prediction of pilot fuel ignition delay as well as methane
oxidation chemistry. In order to cover this, manipulation of the
integrated chemical kinetics mechanisms and flow simulations is
essential. The KIVA-3V code [12] is selected as 3D-CFD framework
for simulations of reactive fluid flow. KIVA uses finite-volume,
temporal-differencing scheme in solution procedures of three
dimensional conservation equations and turbulence at the same
time. This solution procedure, namely the Arbitrary Lagrangian
Eulerian (ALE) method, decouples calculations of the diffusion
and convection terms from chemical source terms. Hence, each
computational cell can be treated as a homogeneously mixed reac-
tor at each time-step.
Continuity equation for species m and energy equation in terms
of specific internal energy are formulated in KIVA as given in Eqs.
(1) and (2), respectively [12].
  
@ qm @ qm
þ O  ð@ qm uÞ ¼ O  qDO þ q_ cm þ q_ s dm‘
@t q Engine cycle
@ Injection system
ðqIÞ þ ðquIÞ ¼ PO  u þ ð1  A0 Þr
@t
: Ou  O  J þ A0 qe þ Q_ c þ Q_ s
where q_ cm in Eq. (1) and Q_ c in Eq. (2) are the source terms that need
to be calculated by combustion model. Mathematical descriptions
of these terms are as follows:
By refer to above equations, it can be declared that the ultimate
goal of a sample combustion model is to determine the chemical
species net production rates (x _ m ). To calculate the molar produc-
tion rate of chemical species participated in the chemical kinetics
mechanism, the gas phase kinetics library of CHEMKIN-II [19] is
integrated into KIVA code. In this procedure, the chemistry routine
in the KIVA has been modified to perform chemistry solutions by
iterative calling of DVODE [20]. This new unit acts as an interface
between KIVA and CHEMKIN and updates the combustion source
terms of Eqs. (1) and (2). The binary linking file including species
and chemical reactions data in CHEMKIN format is generated by
CHEMKIN interpreter prior to each simulation. The KIVA code pro-
vides the species concentrations and thermodynamic information
of each individual cell at every time step to pass to CHEMKIN sol-
ver. The CHEMKIN subroutines construct M-set of stiff ordinary
differential equations and DVODE subroutine is then successively
called to compute the species net production rates at the end of
each time step. A parametric study was conducted prior to main
calculations that indicated a time step of 1 ls for combustion
chemistry calculations is small enough to provide accurate predic-
tions by chemical kinetics [8]. It should be noted that time step
sub-cycling option of ODE solver was also activated to ensure
acceptable accuracy level while stiffness of ODEs were high. Once
the model is developed according to explanations above, the flow
and chemistry solutions are fully coupled. Heat transfer subroutine
is also modified in the current model to calculate wall heat flux
based on Han and Reitz formulation [21] and The RNG k-e model
is used as turbulence model.
3. Experimental setup and engine geometry
3.1. Engine specifications
The model is validated using published experimental data of a
medium duty OM-355 diesel engine converted to operate under
dual-fuel mode. The experiments were conducted by Pirouzpanah
et al. [22] in the Diesel-gas research facility [23]. Engine specifica-
tions can be found in Table 1. A schematic of the experimental set-
up can be seen in Fig. 1. Table 2 lists conditions and input
parameters of simulation cases that are used in the present numer-
ical study under engine full load operation.
3.2. Computational mesh
The full circle computational mesh generation was carried out
by the ANSYS ICEM CFD software. Three different mesh resolutions
Table 1
OM-355 engine specifications.
ð1Þ Engine aspiration Turbocharged
Diesel – 4 stroke
Number of cylinders 6 – inline
Common-rail
Injection pressure 170 (bar)
Injector nozzle number 4–Centric
ð2Þ Nozzle diameter 0.31 (mm)
Injection angle 75 (deg)
Piston diameter  stroke 150  128 (mm)
Cylinder volume 11.58 (lit)
Compression ratio 16.1:1
Max. output power 240 (hp) @ 1400 (rpm)
Max. outlet torque 824 (N m) @ 2200 (rpm)
 A. Maghbouli et al. / Fuel 106 (2013) 98–105 101

present (mesh types 2 and 3). A study was performed prior to main
calculations to investigate the computational time and validity
based on the three grid types. Fig. 2 also shows the calculation
results of in-cylinder pressure for case 1. Results show that mesh
type 1 shows a great discrepancy compared to mesh types 2 and
3. This is due to insufficient resolution of computational grids
located in liquid fuel spray path and inadequate accuracy in predic-
tions of the evaporated fuel. The average computational time var-
ied from 22 h for mesh type 1 to 40 h for mesh type 2 and 90 h for
mesh type 3 on a 5 GHz workstation. In view of the reasonable
trade-off between accuracy and computational time, mesh type 2
is selected to be used in the calculations.

Fig. 1. Schematic of the OM-355 engine setup. 1:Diesel fuel input. 2: Pump. 3: Air 4. Results and discussion
input. 4: Intake manifold. 5: Turbocharger. 6: Natural gas tank. 7: Valve. 8:
Regulator. 9: Natural gas and air mixer. 10: Coolant in. 11: Coolant out. 12: Exhaust Fig. 3 contrasts in-cylinder pressure predictions based on
outlet. 13: Shaft encoder. 14: Dynamometer.
incorporation of the ERC V2 n-heptane mechanism [17,18] and
the mechanism of Zheng and Yao [15] for case 1. The model pre-
Table 2 dicts identical pre-firing pressure traces using both mechanisms.
Operating conditions for different simulation cases. However, the ERC mechanism is unable to model the diffusion
combustion and only a part of premixed combustion is predicted.
Simulation case Case 1 Case 2 Case 3
Combustion mode Diesel Dual-fuel Dual-fuel The reason for such a discrepancy can be due to the difference in
SOI (CAD ATDC) 16 14 15 equivalence ratio validity range of the mechanisms. As it is also
IVC (CAD ATDC) 120 120 120 reported by Pang et al. [24] that the ERC mechanism is valid
EVO (CAD ATDC) 118 118 118
up to equivalence ratio of 1.0, whereas Zheng and Yao validated
Initial pressure (bar) 1.0 1.19 1.21
Initial temperature (K) 338 365 367
their mechanism up to equivalence ratio of 1.5 which is favorable
Engine speed (rpm) 1400 2200 1600 to this study. After initiation of injection, local fuel rich regions
Total air fuel ratio 19.4 23.8 24.9 with high values of equivalence ratio will be formed around
Diesel/gas ratio (% mass) 100:0 12:88 9:91 the spray cone. If the range of validity of the chemical kinetics
Diesel fuel flow rate (gr/cycle) 0.096 0.0093 0.009
mechanism is narrow, the chemistry of autoignition in prospec-
tive regions for combustion will be neglected. Authors had con-
ducted a number of simulations with ERC mechanism;
however, very high values of temperature at IVC (which was
are tested as illustrated in Fig. 2. Grid size is refined near injector far from real in-cylinder thermodynamic properties) were needed
nozzle and regions where fuel spray particles are likely to be to initiate combustion.

Mesh types

Mesh type 1
(2347 cells)
Z

X Y

Mesh type 2
(5752 cells)

Mesh type 3
(13408 cells)
Z

X Y

Fig. 2. View of applied test grids at TDC (left) and corresponding predicted in-cylinder pressure results (right) for case 1.
102 A. Maghbouli et al. / Fuel 106 (2013) 98–105
Fig. 3. Comparison of predicted pressure traces using different chemical kinetic
mechanisms for case 1.
The developed predictive model described in Section 2 is veri-
fied using the available experimental data [22] with operating con-
ditions as given in Table 2. Fig. 4 shows predicted cylinder pressure
histories compared to experiments. It can be seen that the model
predictions of ignition delay, peak pressure location and magni-
tude are in good agreement with experimental data. As it can be
seen on Fig. 4, there is a slight over-prediction in expansion stroke
pressure histories which is mostly observable in case 1, where
100% of chemical energy is provided by the diesel fuel injection.
This problem can be regarded to slightly higher model predictions
for diesel premixed heat release. By comparing the test cases, it can
be seen that case 1 corresponds to the conventional diesel combus-
tion, whereas in cases 2 and 3 only 12% and 9% of liquid fuel is in-
jected, respectively. This means that the rest of chemical energy in
cases 2 and 3 was provided by the premixed mixture of methane
and air. The current model slightly over predicts the heat release
of the diesel combustion, where 100% of the input energy is pro-
vided by liquid fuel. As the combustion approaches dual-fuel oper-
ation mode (cases 2 and 3), where more premixed fuel and oxidizer
charges are available, the model results are much closer to the
experiments. Moreover, in case 1, the higher autoignition heat re-
lease was followed by combustion which is the main source of the
over prediction seen in the expansion stroke in this case. Fig. 5
indicates predicted gross AHR and HRR traces. It can be seen that
Case 1
Fig. 4. Comparison of measured and predicted in-cylinder pressure histories for cases.
the premixed heat release peak is more pronounced in diesel com-
bustion (case 1). However, in dual-fuel cases, the second peak is
delayed due to higher temperature required for methane
autoignition.
Major combustion species concentration histories are illus-
trated in Fig. 6. It shows that different trends for species concentra-
tion variations agree well with data in the literature [2]. The model
is able to predict CO2 and NOx with a good accuracy compared to
experiments. It should be noted that the exhaust experimental val-
ues of CO2 and NOx are marked in this figure. The diagrams of
methane and n-heptane concentration variations show that major-
ity of methane consumption is occurred when diesel fuel is burn-
ing in high temperature, diffusion combustion phase. Considering
constant total mixture input energy, additional parametric studies
are conducted on case 2. Three pilot fuel values of 0.0116, 0.0140
and 0.0186 [gr/cycle] are considered, namely, case 2a, case 2b
and case 2c, respectively. The amount of pilot fuel values in the
cases of the parametric study are considered to be 25%, 50% and
100% more than base line case. It should be noted that other influ-
ential parameters such as fluid flow and thermodynamic initial
conditions are kept constant in parametric studies. Fig. 7 shows
the effect of increasing pilot fuel quantity on pressure and HRR.
It is found that an increase in the amount of pilot fuel advances
combustion timing and increases peak pressure. It is also seen that
the second HRR peak is delayed as the amount of pilot fuel de-
creases due to increment in the methane/n-heptane ratio. Varia-
tion of IMEP, ITE and ISEC values with respect to pilot fuel
quantities are shown in Fig. 8 as the effective engine performance
parameters. It is seen that IMEP is slightly increased as more pilot
fuel amount is injected. Similar increasing trend is observed for ITE.
However, ITE values for cases 2, 2a, 2b and 2c do not show large
difference. A minimum value of 39% and maximum of 45% are re-
sulted from parametric study for ITE corresponding to cases 2 and
2c, respectively. Also in Fig. 8, the variation of ISEC is represented.
It can be seen that the ISEC provided by premixed methane and in-
jected n-heptane tends to decrease by applying higher amount of
pilot fuel. As the assumption for parametric study, input chemical
energy is the same for all cases 2, 2a, 2b and 2c. However, produc-
tion of slightly higher output power by increasing the pilot fuel
quantity is lead to the reduction in ISEC values represented in
Fig. 8c. Fig. 9 illustrates temperature and methane concentration
contours for the cases under parametric study. It can be implied
that methane burning is initiated from high temperature regions
located in reactive spray path. Fig. 9 also shows that at the same
CAD, a higher concentration of methane is consumed as pilot fuel
quantity was increased. The concentration change of CO and NOx
emissions is shown in Fig. 10. The results indicate that, increasing
pilot fuel under full load operating condition of the dual-fuel
engine leads to higher emission levels. NOx concentrations tend
Case 2 Case 3
 A. Maghbouli et al. / Fuel 106 (2013) 98–105 103

Case 1 Case 2 Case 3

Fig. 5. Predicted gross HRR and AHR histories for simulation cases.

Casse 1 Case 2 Casee 3

Fig. 6. Predicted in-cylinder concentration histories of major combustion species for cases.

Fig. 7. Pressure (left) and HRR (right) traces for case 2 at different pilot fuel rates: case 2 = 0.0093 [gr/cycle], case 2a = 0.0116 [gr/cycle], case 2b = 0.0140 [gr/cycle] and case
2c = 0.0186 [gr/cycle].
104 A. Maghbouli et al. / Fuel 106 (2013) 98–105

(a) (b) (c)

Fig. 8. IMEP (a), ITE (b) and total ISEC (c) versus injected diesel mass for case 2 and three parametric study cases.

Caase 2 Case 2

Casse 2a Case 2a

Casse 2bb Caase 2b

2

Casse 2c Caase 22c

Fig. 9. Temperature [K] and methane concentration [gr] contours at -2 CAD ATDC for case 2 at different pilot fuel rates.

to increase slightly; however, there is a considerable increment in
exhaust CO levels.
5. Conclusions
After discussing model development and input conditions of
numerical calculations, the main conclusions are made below:
1. Zheng and Yao’s reduced n-heptane mechanism is capable of
simulating high D.I ratios needed in modeling mixing-con-
trolled combustion of diesel and dual-fuel operating modes,
where there is no need for tuning model constants in order to
simulate different operating cases.
2. Predicted in-cylinder pressure traces are in good agreement
with the experimental results. This includes peak pressure mag-
nitude and location.
 A. Maghbouli et al. / Fuel 106 (2013) 98–105
Fig. 10. CO (left) and NOx (right) concentration change histories for case 2 at different pilot fuel rates.
3. Simulations show that ignition tends to delay by shifting the
diesel combustion to the dual-fuel combustion mode.
4. Increment in the pilot fuel quantity, with the assumption of
equivalent input energy, decreases ignition delay and elevates
in-cylinder peak pressure. It also intensifies combustion cham-
ber mean temperature rapidly which result in quicker con-
sumption of methane fuel.
5. Parametric studies on dual-fuel operating mode show that IMEP
and ITE are slightly increased by applying higher amounts of
pilot fuel quantity. However NOx and CO concentrations are
increased.
Acknowledgment
Authors would like to gratefully acknowledge contribution of
Mr. Ali Y. Nobakht and Dr. Arash Rahimi.
References
[1] Kusaka J, Ito S, Mizushima N, Daisho Y, Saito T. A numerical study on
combustion and exhaust gas emissions characteristics of a dual-fuel natural
gas engine using a multi-dimensional model combined with detailed kinetics.
SAE paper 2003–01-1939; 2003.
[2] Karim GA, Liu Z, Jones W. Exhaust emissions from dual-fuel engines at light
loads. SAE paper 932822; 1993.
[3] Soylu S. Examination of combustion characteristics and phasing strategies of a
natural gas HCCI engine. Energy Convers Manage 2005. http://dx.doi.org/
10.1016/j.enconman.2004.02.013.
[4] Scott B, Dennis F. Development and validation of a quasi-dimensional model
for HCCI engine performance and emissions studies under turbocharged
conditions. SAE paper 2002–01-1757; 2002.
[5] Komninos NP, Hountalas DH, Kouremenos DA. Development of a new multi-
zone model for the description of physical processes in HCCI engines. SAE
paper 2004–01-0562; 2004.
[6] Ramos JI. Internal combustion engine modeling. Taylor & Francis; 1989.
105
[7] Kong SC, Han Z, Reitz RD. The development and application of a diesel ignition
and combustion model for multidimensional engine simulation. SAE paper
950278; 1995.
[8] Agarwal A, Assanis D. Multi-dimensional modeling of natural gas ignition
under compression ignition conditions using detailed chemistry. SAE paper
980136; 1998.
[9] Kusaka J, Ko-ichiro T, Daisho Y, Saito T. A numerical study on combustion and
exhaust gas emissions characteristics of a dual-fuel natural gas engine using a
multi-dimensional model combined with detailed kinetics. SAE paper 2002–
01-1750; 2002.
[10] Hong S, Assanis D, Wooldridge M. Multi-dimensional modeling of NO and soot
emissions with detailed chemistry and mixing in a direct injection natural gas
engine. SAE paper 2002–01-1112; 2002.
[11] Liu C, Karim GA. A Simulation of n-heptane fueled HCCI engines using a 3-D
model with detailed chemical kinetics. Proc ASME 2005;5:1220.
[12] Amsden AA. KIVA-3V: a block-structured KIVA program for engines with
vertical or canted valves. LA-13313-MS; 1997.
[13] Golovitchev IV, Atarashiya K, Tanaka K, Yamada S. Towards universal EDC-
based combustion model for compression-ignited engine simulations. SAE
paper 2003–01-1849; 2003.
[14] Kong SC, Marriott CD, Reitz RD, Christensen M. Modeling and Experiments of
HCCI Engine Combustion Using Detailed Chemical Kinetics with
Multidimensional CFD. SAE paper 2001–01-1026; 2001.
[15] Zheng Z, Yao M. Charge stratification to control HCCI: experiments and CFD
modeling with n-heptane as fuel. Fuel 2009. http://dx.doi.org/10.1016/
j.fuel.2008.09.002.
[16] http://www.llnl.gov.
[17] http://www.erc.wisc.edu/chemicalreaction.php.
[18] Patel A, Kong SC, Reitz RD. Development and validation of a reduced reaction
mechanism for HCCI engine simulations. SAE paper 2004–01-0558; 2004.
[19] Kee RJ, Rupley FM, Miller JA. CHEMKIN-II: A FORTRAN chemical kinetics
package for the analyses of gas phase chemical kinetics. Sandia report SAND
89–8009; 1989.
[20] Brown PN, Byrne GD, Hindmarsh AC. VODE: a variable coefficient ODE solver. J
Sci Stat Comput 1989;10:1038–51.
[21] Han Z, Reitz RD. A temperature wall function formulation for variable-density
turbulent flows with application to engine convective heat transfer modeling.
Int J Heat Mass Transfer 1997;40:613–25.
[22] Pirouzpanah V, Jeihouni Y, Afghani M. Reduction of pollutants emissions of
OM-355 diesel engine to Euro2 by converting to dual-fuel engine (diesel/gas).
In: Proc first conf auto fuel CNG; 2003. p. 84–94.
[23] http://www.dieselgas.co.nz.
[24] Pang KM, Ng HK, Gan S. Development of an integrated reduced fuel oxidation
and soot precursor. Fuel 2011;90:2902–14.