Alexandria Engineering Journal (2015) 54, 373–382

H O S T E D BY
Alexandria University

Alexandria Engineering Journal

www.elsevier.com/locate/aej
www.sciencedirect.com

ORIGINAL ARTICLE

Effective reduction of in-cylinder peak pressures

in Homogeneous Charge Compression Ignition
Engine – A computational study
T. Karthikeya Sharma *, G. Amba Prasad Rao, K. Madhu Murthy

Department of Mechanical Engineering, NIT Warangal, 506004 TS, India

Received 12 January 2015; revised 9 April 2015; accepted 15 April 2015

Available online 2 May 2015

KEYWORDS Abstract HCCI mode of combustion is known for simultaneous reduction of NOx and PM emis-
HCCI engine; sions besides yielding low specific fuel consumption. The nature of volumetric combustion of HCCI
ECFM-3Z; engine leads to the development of high peak pressures inside the combustion chamber. This high
Swirl ratio; peak pressures may damage the engine, limiting the HCCI engine life period and thus demands
Peak pressures; sturdy designs. In this study an attempt is made to analyze computationally the effect of induction
EGR; swirl in reducing the peak pressures of a HCCI engine under various operating parameters. For the
Compression ratio study, specifications of a single cylinder 1.6 L, reentrant piston bowl diesel engine are chosen. For
the computational analysis ECFM-3Z model of STARCD is considered. This model is suitable to
analyze the combustion processes in SI and CI engines. As HCCI engine is a hybrid version of SI
and CI engines, ECFM-3Z model with necessary modifications is used to analyze the peak pressures
inside the combustion chamber. The ECFM-3Z model for HCCI mode of combustion is validated
with the existing literature to make sure that the results obtaining are accurate. Numerical experi-
ments are performed to study the effect of compression ratio, equivalence ratio, exhaust gas recir-
culation and boost pressure under different swirl ratios in reducing the in-cylinder peak pressures.
The results showed that swirl ratio has a considerable impact in limiting the peak pressures of HCCI
engine. The analysis resulted in achieving about 21% reduction in peak pressures are achieved when
a swirl ratio of 4 with 30% EGR is adopted when compared to a swirl ratio of 1 with 0% EGR. The
study revealed that out of the four operating parameters selected, lower compression ratios, higher
EGR concentrations, lower equivalence ratios, lower boost pressures and higher swirl ratios are
favorable in reducing the peak pressures.
ª 2015 Faculty of Engineering, Alexandria University. Production and hosting by Elsevier B.V. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

The Homogeneous Charge Compression Ignition (HCCI) is

* Corresponding author. considered to be the principally promising future IC
Peer review under responsibility of Faculty of Engineering, Alexandria engine combustion concepts. HCCI is a concept of hybrid
University.
http://dx.doi.org/10.1016/j.aej.2015.04.006
1110-0168 ª 2015 Faculty of Engineering, Alexandria University. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
374 T. Karthikeya Sharma et al.

combustion, between conventional combustion concepts of Flame Combustion Model-3 Zones [11,12] and the
spark ignition engine and compression ignition engine. HCCI Equilibrium-Limited ECFM (ECFM-CLEH) [13,14]. Each
concept is however not a modern finding. Already in the early combustion model possesses limitations and advantages and
twentieth century hot bulb engines operated with an HCCI- is appropriate for a particular set of problems. In general
like combustion. They were superior in terms of brake thermal ECFM-CLEH and ECFM-3Z are suitable for almost all types
efficiency compared with the contemporary gasoline engines of combustion regimes, but ECFM-3Z is mainly applicable for
and at the same level as the diesel engines. Research has premixed homogeneous turbulent combustion with both SI
revealed that high engine efficiencies, ultra low NO emissions and CI. Table 1 shows the applicabilities of the various
and low particulates are the benefits of HCCI engines. combustion models. Due to its vast range of suitability,
Volumetric autoignited combustion of the compressed lean ECFM-3Z has been used in the present investigation to
air–fuel mixture is attributed to these benefits. Though examine the impact of piston bowl geometry on flow and com-
HCCI concept is attractive, it suffers from few limitations such bustion characteristics. Fig. 1 shows the schematic representa-
as low specific output, narrow operating range, lack of control tion of the three zones of the ECFM-3Z model.
over the ignition process, long start-up time and high levels of Induction induced swirl has a predominant effect on mix-
CO and UHC emissions. The CO and UHC emissions can be ture formation and rapid spreading of the flame front in the
after treated using catalytic converters [1–3]. Heywood et al. conventional combustion process of a CI engine. This has been
explained the combustion characteristics of HCCI combustion well documented in the literature. However, it is observed that
along with the many other fundamental concepts about the no work has been done on the effect of swirl in HCCI mode.
combustion in an IC engine. Heywood reported that the prob- The main objective of the present study was to analyze the
lems such as high particulate matter and soot emissions effect of induction induced swirl in reducing the peak pressures
(because of fuel rich and diffusion rich regions) associated with of the HCCI engine under varying operating parameters.
the conventional CI engine can be overcome by HCCI engines. Because of the volumetric combustion the development of
Volumetric combustion and low temperature combustion of peak pressures inside the combustion chamber is very high.
HCCI engine resulted in low particulate and soot formations. This is one of the limitations of the HCCI engine which
Also, low specific fuel consumption was the most attractive of demands the rigid body construction of the engine and reduces
HCCI engines when compared with the conventional CI engines the engine life. In this regard a computational attempt is made
[4]. Onishi et al. conducted experiments to visualize the process to control the peak pressures in terms of induction induced
of combustion on a conventional SI engine in both SI and swirl along with other parameters.
HCCI modes using schlieren photography method. From their
experimental results they reported that very wee defined propa- 2. Methodology
gation of flame was found in SI operation mode but no visible
propagation of flame was found in HCCI mode of combustion; A DI single cylinder CI engine with two different piston bowls
proving volumetric combustion of HCCI engines [5,6]. was considered for the analysis. The specifications of the
Chen et al. numerically investigated the effect of EGR in engine are tabulated in Table 2 have been considered for the
reducing the pressure rise rate in HCCI engine. They incorpo- analysis. Reentrant and spherical piston bowls are taken.
rated CHEMKIN and SENKIN code for the analysis. They Care was taken to have same clearance volume with all the
reported that with increase in EGR fraction retarded the start two piston bowls. To study the parameters such as flame
of combustion and decreased the peak pressure and tempera- distribution inside the combustion chamber, rates of heat
ture rise. Under the same conditions with increase in EGR release, temperatures, pressures and emissions of NO and
ratio, extension in preparation of thermal ignition and CO a multidimensional CFD package of STAR-CD; ECFM-
advancement in LTHR timing were observed [7]. 3Z is used. The model of the piston bowl shaper was prepared
Swirl helps in homogeneous mixture formation of the fuel and meshed as per the specification without having any varia-
and air [8]. It also helps in NOx emission reduction [9]. The tion in the compression ratio. The analysis was started with a
increase in swirl ratio reduces the peak temperatures by starting angle of 680 CA and ending angle of 800 CA.
increasing the heat transfer to the combustion chamber parts.
This leads to a low temperature combustion process resulting
3. CFD model set-up
in low NOx emissions [10].
Performing these explorations (under different operating
parameters with induction induced swirl) solely in the labora- A spline was created in the shape of the piston bowl and was
tory would be expensive, inefficient and impractical because of transformed into a 2D template. The 2D template was then
the complex interactions of the many variables. Because of this cut into 3D piston bowl representing 1/6th of the piston bowl.
reason, a CFD tool Star-CD is chosen for the analysis. Several The 1/6 piston bowl computational mesh has around 128,000
modifications were made to Star-CD es-ice module so that it
could be used for HCCI engine modeling. Many commercial
CFD packages such as Open FOAM, Ricardo Wave, GT
Table 1 Combustion model capabilities.
Power, AVL FIRE are available to simulate the combustion
process in IC engines. In the present work ‘‘es-ice’’ of Model Applicability
STAR-CD is used for the analysis. The various models G-Equation Partially premixed SI and CI
developed for presaging the engine combustion processes are DARS-TIF Compression ignition
Transient Interactive Flamelets (TIF) model, Digital ECFM Non-homogeneous premixed SI
Analysis of Reaction System-Transient Interactive Flamelets ECFM-3Z Premixed and nonpremixed SI and CI
model (DARS-TIF), G-equation model, Extended Coherent
A computational study on effective reduction of in-cylinder peak pressures 375

Figure 1 Schematic representation of three zones of ECFM-3Z

model.

cells BDC, and 52,000 cells at TDC with tetrahedral cells. By

comparing the peak pressure results with a finer grid having
12,400–259,000 cells a grid independence study was performed. Figure 2a Grid independent study.
The numerical results represented that the peak pressure values
are almost constant after 128,000 representing grid indepen-
where nR is taken as 2 is the no. of revolutions of the crank for
dent nature, but with higher computational processing time.
power stroke per cylinder and N is the speed of the engine
Hence, the optimal cell size of 128,000 is considered for the
(rpm). The indicated specific fuel consumption (ISFC) is
analysis. The 3D mesh consists of 40 radial cells, 160 axial
shown in Eq. (3):
cells, 5 top dead center layers and 40 axial block cells. The grid
independent analysis representing In-cylinder pressures vs 30mfuel N
ISFCðg=kW hÞ ¼ ð3Þ
Number of cells is represented in Fig. 2a. The piston bowl P
shape and 3D mesh of the piston bowl sector is shown in In Eqs. 1–3 the power and ISFC analyses are qualitative only.
Fig. 2b and c.
Gross indicated Work per cycle (W) is used to analyze the 4. Modeling strategy
engines energy efficiency which is calculated from the displace-
ment of the piston and pressure in the cylinder using Eq. (1):
" # STAR-CD is a combined package consisting of several sub-
Z
paB2 h2 a sinð2hÞ models to simulate all the SI and CI engine computations.
WðN mÞ ¼ pðhÞ 2 sinðhÞ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dh ð1Þ The sub-model such as atomization and fuel spray, turbulence
8 h1 l2  a2 sin2 ðhÞ
(I–L), boundary wall function, NOx, auto-ignition, combus-
where a = crank radius, l = length of the connecting, tion (ECFM-3Z), was included in the STAR-CD es-ice pack-
B = bore of the cylinder, and h1 and h2 are the beginning age. Ideal gas laws and Cp as a function of temperature are
and the end of the valve-closing period. chosen to deal with fluid properties.
The relation between indicated work per cycle and indi-
cated power per cylinder (P) is shown in Eq. (2): 4.1. Swirl creation
WN
PðkWÞ ¼ ð2Þ The swirl inside the combustion chamber is generated by
60; 000nR
varying the intake charge velocity components in U and V
directions. The velocity components of U and V are calculated
as follows.
Table 2 Engine specifications. U ¼ ðx  xCSYS ÞðSwirl  RPM  2pÞ=60 ð4Þ
Engine specifications
Displacement volume 1600 cm3
V ¼ ðy  yCSYS ÞðSwirl  RPM  2pÞ=60 ð5Þ
Bore 12.065 cm where U and V are the intake charge velocity components in X
Stroke 14 cm and Y directions. RPM is the engine speed. x, xCSYS y, yCSYS
Connecting rod length 26 cm are the x and y Global and local coordinates. The local and
Compression ratio 21:1
global coordinates Z axis is to be matched after meshing to
Fuel n-Dodecane
eliminate error.
Operating conditions
Engine speed 1000 rpm 4.2. Spray injection and atomization model
Equivalence ratio 0.26
Inlet temperature air (Tair) 353 K
Inlet air pressure (Pair) 0.1 MPa Huh’s [15,16] model is useful in modeling spray and atomiza-
Cylinder wall temperature (Twall) 450 K tion processes in conventional CI engines. Nozzle generated
EGR 0% the turbulence stresses, and gas inertia are the mechanisms
involved in Huh’s model to deal with spray and atomization.
376 T. Karthikeya Sharma et al.

(b) @ TDC (c) @ BDC

Figure 2b and c Schematic representation of 3D piston bowl shape at TDC and BDC.

The turbulence generated in the nozzle during the jet leaving combustion. Auto-ignition of the premixed fuel and air takes
the nozzle hole develops a surface wave growth direct toward place because of high temperatures generated at the end of
formation of droplets. the compression stroke in HCCI mode. The ECFM-3Z model
Secondary break-up of the droplets takes place because of is capable of dealing with propagation of flame, turbulence,
continuous phasing and non-uniform pressure around the dro- double delay auto-ignition, postflame models and emissions
plet. The secondary break-up is mainly depended on Weber of the modern HCCI engines. Angelberger et al. [19] wall func-
number (Wed) and the dimensionless droplet diameter d* as tion model was used to resolve the wall-bounded flows; turbu-
shown in Eq. (6): Reitz–Diwakar model [17,18] is used in com- lence generated near-wall region. The turbulent and molecular
puting the secondary break-up of the droplets. lever mixing of gases in the two zones called unmixed fuel zone
and unmixed (air + EGR zone) forms a mixing zone and
qd Dd V2d;n
Wed ¼ ð6Þ where the combustion takes place.
r The mass fractions of the unmixed fuel (Yfum ) can be
where ‘n’ = unit normal to the wall, ‘Vd,n’ = the velocity nor- obtained from Eq. (9):
mal component normal to the wall, ‘r’ = surface tension   
@qYfum l
coefficient. þ r  ðquYfum Þ  r  D þ t rYfum
Premixed mechanism is considered for HCCI mode to pre- @t SCt
 
pare a fuel and air homogeneous mixture. b q wm
¼ min Yfum 1  Yfum þ x_ e Vap ð9Þ
p qu wf
4.3. Autoignition model
4.5. NOx model
Instead pre-ignition kinetics, ignition occurrence time is estab-
lished to compute time of the delay in Ignition.
The reaction of N2 of the air with O2 in the combustion leads
The auto-ignition delay sd is computed using Eq. (7):
to the formation of NOx emissions. Usually this reaction
sd ¼ 1:051  108 ½F0:05 ½O2 0:53 q0:13 e5914=Tl ½47=CN ð7Þ occurs at high temperatures generated because of combustion.
The emissions of NOx are highly sensitive to higher tempera-
where CN = cetane number (max = 60). To track the reac-
tures and are directly related. Prompt and thermal NOx are
tions development an ignition progress variable function is
the two sources for NOx emissions from an IC engine.
defined as shown in Eq. (8):
Thermal NOx emissions are in major portions from the diesel
ðdYigi Þ=ðdtÞ ¼ YTf Fðsd Þ ð8Þ engine NOx emissions. To model NOx emissions Extended
Zel’dovich mechanism [20] is considered. The Extended
As HCCI mode of combustion is controlled by cool flames a
Zel’dovich mechanism deals with the below equations by
double delay auto-ignition model is used. The reaction rates
Bowman [20]
slowed down because of low temperatures in cool flame
regime. Main-auto-ignition starts because of the increased O þ N2 $ NO þ N ð10Þ
rates of reaction after second delay.
The double delay auto-ignition model deals with two igni- N þ O2 $ NO þ O ð11Þ
tion progress variables and two nonempirical precomputed
delay times. It is possible to estimate the maximum fuel burnt N þ OH $ NO þ H ð12Þ
at each auto-ignition step from these ignition progress
With the partial equilibrium of Eq. (12) for the hydrogen
variables.
radicals,

4.4. Combustion model O þ OH $ O2 þ H ð13Þ

Heywood [21] developed a single NO rate equation to repre-
A Three Zone Extended Coherent Flame Combustion Model sent the Extended Zeldovich mechanism and is shown in Eq.
(EFCM-3Z) is used to simulate the HCCI mode of (14):
A computational study on effective reduction of in-cylinder peak pressures 377

Figure 3 Validation of the ECFM-3Z compression ignition model with the experimental results of external mixture formation of HCCI
engine.

( )
d f1  ½NO2 =K12 ½O2 Rd2 g To simulate the EGR model variable composition model is
½NO ¼ 2k1 f ½O½N2  used in the present analysis. Here the EGR is the ratio of the
dt 1 þ k1b ½NO=ðk2f ½O2  þ k3f ½OHÞ
re-circulating exhausts gas mass (megr) to the total mass of
ð14Þ charge that enters the cylinder (mI).
where K12 = (k1f/k1b)(k2f/k2b) and the subscripts 1, 2 and 3 Thus:
refer to Eqs. (10)–(12), respectively. megr
O2, N2, O, and OH are assumed to be in local thermody- egr ¼ ð15Þ
mI
namic equilibrium.
where

4.6. EGR modeling mI ¼ mair þ megr þ mf ð16Þ

For the Individual species
Exhaust Gas Recirculation technique improves the auto-
ignition capabilities as it retains some hot combustion gases megrðfÞ ¼ egr  mEðfÞ ð17Þ
for the next combustion cycle and also it reduces the NOx
megrðO2Þ ¼ egr  mEðO2Þ ð18Þ
emissions.
There are two EGR models as stated below.
megrðCO2Þ ¼ egr  mEðCO2Þ ð19Þ
(a) Variable composition –– it considers and components of
megrðN2Þ ¼ egr  mEðN2Þ ð20Þ
EGR called six scalars such as EGR_CO2, EGR_O2,
EGR_H2O, EGR_CO, EGR_N2 and EGR_H2 which
megrðH2OÞ ¼ egr  mEðH2OÞ ð21Þ
can be solved by transport equations. Out of all the scalars
the EGR_O2 scalar does not take part in the reaction. where mf = mass of the fuel.
(b) Fixed composition –– in this the mass fraction of each The ratio of mass of individual gas to the total mass
scalar (total sum should be equal to 1) should be entered of all exhaust gases gives the mass fraction of that particular
to define the EGR mixture. gas.
378 T. Karthikeya Sharma et al.

Figure 4 In-cylinder pressure vs Crank angle at different compression ratios and swirl ratios.

species, thermal energy, shear stresses wall fluxes the y+ inde-

Table 3 Peak pressures (MPa) at various compression ratios
pendency was achieved along with providing all necessary
and swirl ratios.
boundary conditions for a wide range of near-wall mesh den-
Compression SW1 SW2 SW3 SW4 Percentage sities. The variable values on the wall and at the near wall cells
ratio reduction are calculated by using hybrid wall functions.
between SW1
and SW4
6. Validation of ECFM-3Z, compression ignition model
CR 18 11.78 11.70 11.64 11.55 1.98
CR 19 12.43 12.33 12.25 12.14 2.27
CR 20 13.06 12.96 12.88 12.75 2.37
Many well established automotive industries and researchers
CR 21 13.71 13.60 13.47 13.34 2.70 are adopting STAR-CD CFD package for 3 dimensional anal-
Percentage 16.352 16.20 15.73 15.491 yses on IC engines. Many researchers like Pasupathy
between CR Venkateswaran and Nagarajan [9], Zellat et al. [23],
18 and CR 21 Bakhshan et al. [24] validated the STAR-CD package with
the experimental results and found the good agreement
between experimental and model results. In the present analy-
sis the Combustion aspects of a CI engine in HCCI mode were
5. Initial and boundary conditions done by using ECFM-3Z model by considering Ganesh and
Nagarajan [25] engine specifications to validate the model
The initial boundary conditions considered for the analysis results. Ganesh et al. conducted experiments in both direct die-
are; an absolute pressure of 1.02 bar, 353 K initial tempera- sel injection mode and diesel vapor induction mode at varying
ture, 0% EGR, 0.26 equivalence ratio. Wall temperatures of EGR fractions. They developed a fuel vaporizer to evaporate
combustion dome region are fixed as 400 K, dome regions as the diesel fuel and to prepare a homogeneous mixture with
450 K and piston crown regions as 450 K. The Angleberger air before entering into the cylinder. He has compared the
wall function mode [22] is considered. By solving momentum, engine in-cylinder pressures and emissions in both the cases
turbulence and mass equations boundary layers with no-slip and found that the diesel vapor induction results are good
are computed by ‘two-layer’ and low Reynolds number when compared with the direct diesel injection. Validation of
approaches. In the analysis a combined ‘‘two layered’’ and ECFM-3Z model results with Ganesh and Nagarajan [25]
‘‘low Reynolds number’’ hybrid wall boundary condition is experimental results was performed by choosing the same
used. The need for using a small value for y+ was eliminated engine boundary conditions and specification.
by the use of hybrid wall boundary condition, which creates Graphs of performance and emissions with simulation and
very fine mesh near the walls. By using the expression of experimental values are drawn to examine the variation in
asymptotic valid for 0.1 < y+ < 100 or by combining the results. It was found that the results of ECFM-3Z are in good
low and high Reynolds number expression for chemical coherence with the experimental results of Ganesh and
A computational study on effective reduction of in-cylinder peak pressures 379

Figure 5 In-cylinder pressure vs Crank angle at different EGR concentrations and swirl ratios.

NOx emissions showed a variation of 4.2% at 25% load and

Table 4 Peak pressures (MPa) at various exhaust gas recir- 1.82% at 75% load. The error in hydrocarbon emissions was
culation and swirl ratios. obtained as 5.474% at 25% load and 4.31% at 75% load. In
case of CO emissions the error was observed to be 4.69% at
EGR (%) SW1 SW2 SW3 SW4 % 25% load and 8.69% at 75% load. EGR was observed to be
Reduction
a dominant parameter in reducing error in the values of
between
SW1 and NOx emissions where as the error has increased with HC emis-
SW4 sions as the model has over predicted it. No variation in error
is observed with EGR in case of CO emissions.
EGR 0% 13.71 13.60 13.47 13.34 2.70
EGR 10% 12.70 12.63 12.56 12.45 1.94
EGR 20% 12.01 11.92 11.82 11.69 2.58 7. Results and discussion
EGR 30% 11.23 11.06 10.94 10.81 3.72
% Reduction 18.096 18.644 18.783 18.953 The effect of induction induced swirl in reducing peak pres-
between EGR sures has been studied computationally by varying operating
0% and EGR
parameters such as compression ratio, equivalence ratio,
30%
exhaust gas recirculation, and boost pressure. The swirl ratios
ranging from 1 to 4 are considered for the analysis. The simu-
lation results of the ECFM-3Z model are discussed below.
Nagarajan [25] and are depicted in Fig. 3. In the figures EDVI
– Experimental Diesel Vapor Induction, SDVI – Simulated 7.1. Compression ratio
Diesel Vapor Induction, EDDI – Experimental Direct Diesel
Injection and SDDI – Simulated Direct Diesel Injection at The variation of peak pressures of the reentrant piston bowl
respective EGR concentrations. The clear variation between with compression ratio for swirl ratios 1–4 are plotted in
DVI and DDI can be seen in Fig. 3. Fig. 4. From Table 3, it can be observed that lower compres-
The error analysis of the validation of ECFM-3Z with the sion ratios and higher swirl ratios are favorable in limiting
experimental results was performed. The Error between simu- the peak pressures. The peak pressures obtained at different
lated and experimental results is high in case of direct diesel compression ratios and swirl ratios are summarized in Table 3.
injection when compared with the diesel vapor induction. In The peak in-cylinder pressure increases with compression
case of in-cylinder pressures a maximum deviation of 7.56% ratio irrespective of the swirl ratio, but the percentage increase
at 363 CA during combustion and a minimum of 1.47% at in peak pressures is less at higher swirl ratios [26]. As swirl
348 CA were observed. The maximum deviation during com- ratio increases; reduced peak pressures are obtained at all
bustion is attributed to assumption of fixed wall temperatures compression ratios, but the decrease in peak pressures is high
instead of mapping to save computational time and storage. at higher compression ratios [27]. The reason for this
380 T. Karthikeya Sharma et al.

Figure 6 In-cylinder pressure vs Crank angle at different equivalence ratio and swirl ratios.

The peak pressure decreases with increase in EGR concen-

tration irrespective of the swirl ratio, but percentage reduction
in peak pressures is high at higher EGR levels with higher swirl
Table 5 Peak pressures (MPa) at various equivalence ratios
ratios. This can be attributed to release of reduced net energy
and swirl ratios.
with decrease in volumetric efficiency of the engine at higher
EQR SW1 SW2 SW3 SW4 % EGR levels and owing to oxygen availability for the combus-
Reduction tion process. As swirl ratio increases; lower peak pressures
between are obtained at any EGR concentration, but percentage reduc-
SW1 and
tion in peak pressures is high at higher swirl ratios and higher
SW4
EGR concentrations. The reason for this phenomenon is
EQR 0.26 13.71 13.60 13.47 13.34 2.70 increase in turbulence owing to increased wall heat transfer
EQR 0.46 20.89 20.89 20.79 20.63 1.21 because of increased swirl ratios [28,29]. This study also indi-
EQR 0.66 26.55 26.55 26.09 25.83 2.71
cated modest shift in the occurrence of maximum peak pres-
EQR 0.86 30.14 29.89 29.64 29.32 2.72
% Increase 119.708 119.112 120.89 120.300
sure towards TDC with increase in EGR concentrations [30].
between EQR
0% and EQR 7.3. Equivalence ratio (ø)
30%
The variation of peak pressures of the reentrant piston bowl
with equivalence ratio with swirl ratios 1–4 is plotted in
Fig. 6. From Table 5 it can be observed that lower equivalence
ratios and higher swirl ratios are favorable in reducing the
phenomenon is increase in turbulence owing to increased wall peak pressures. The peak pressure increases with increase in
heat transfer due to increased swirl intensity. equivalence ratio irrespective of the swirl ratio, but the per-
centage increase in peak pressures is marginal at all swirl ratios
7.2. Exhaust Gas Recirculation (EGR) [31,32]. The reason behind it is increased equivalence ratio
injects more fuel leading to rich mixture formation process,
The variation of peak pressures of the reentrant piston bowl combustion of more quantity of fuel leads to increase in peak
with EGR concentration with swirl ratios 1–4 are plotted in pressures. As swirl ratio increases; decrease in peak pressures is
Fig. 5. From Table 4, it can be observed that higher EGR obtained at all equivalence ratios, but percentage reduction in
concentrations and higher swirl ratios are favorable in peak pressures is high at high swirl ratios and lower equiva-
reducing the peak pressures. The peak pressures obtained at lence ratios. This may again be due to increased turbulence
different EGR concentrations and swirl ratios are plotted in owing to increased wall heat transfer losses because of
Table 4. increased swirl ratios.
A computational study on effective reduction of in-cylinder peak pressures 381

Figure 7 In-cylinder pressure vs Crank angle at different boost pressures and swirl ratios.

8. Conclusions
Table 6 Peak pressures (MPa) at various boost pressures and
swirl ratios.
High in-cylinder peak pressures of HCCI engine due to volu-
Boost SW1 SW2 SW3 SW4 % metric combustion may damage the engine and thereby
pressures Reduction demands a rigid engine construction. A computational analysis
(bar) between
is undertaken to find the impact of induction induced swirl in
SW1 and
SW4
lowering peak pressures of HCCI engine. A study has been
attempted using Extended Coherent Flame combustion model
BP 1.0 13.71 13.60 13.47 13.34 2.70 for peak pressure reduction of a HCCI engine with induced
BTP 1.25 17.39 17.24 17.03 16.95 2.52
swirl motion under different operating conditions. The present
BP 1.5 21.37 20.91 20.61 20.32 4.90
BTP 1.75 25.01 24.69 24.21 23.85 4.63
investigation disclosed that ECFM-3Z of STAR-CD is well
BTP 2.0 28.87 28.39 27.86 27.38 5.16 suitable for HCCI mode of combustion with necessary modifi-
Percentage 110.545 108.777 106.731 105.235 cations, in coherence with the existing literature. It was found
increase that swirl ratio has a considerable impact in limiting the peak
between BP pressures of HCCI engine. The analysis resulted in achieving
1.0 and BP about 21% reduction in peak pressures when a swirl ratio of
2.0 4 with 30% EGR is adopted when compared to a swirl ratio
of 1 with 0% EGR. The effect of four operating parameters
viz; compression ratio, exhaust gas recirculation, equivalence
7.4. Boost Pressure (BP) ratio and boost pressure under varying swirl ratios suggested
lower compression ratios, higher EGR concentrations, lower
The variation of peak pressures of the reentrant piston bowl equivalence ratios, lower boost pressures and higher swirl
with boost pressures with swirl ratios 1–4 is plotted in Fig. 7. ratios are favorable in reducing the peak pressures. The study
From Table 6 it can be observed that lower boost pressures revealed that adoption of high swirl ratios associated with high
and higher swirl ratios are favorable in mitigating the peak EGR levels would lead to significant reduction in peak pres-
pressures. The peak pressures increase with increase in boost sures in HCCI mode.
pressures irrespective of the swirl ratio, but the percentage
increase in peak pressures is less at higher swirl ratios Acknowledgements
[33,34]. As swirl ratio increases; lower peak pressures are
obtained at any boost pressure, but the percentage decrease The authors thank Dr. Raja Banerjee, Associate Professor, IIT
in peak pressures is high at higher boost pressures. The reason Hyderabad for allowing to use computational facility, Mr. B.
for this phenomenon is increase in turbulence owing to Siva Nageswara Rao from CD-adapco, Bengaluru and Mr.
increased wall heat transfer losses because of increased swirl P. Madhu computer Laboratory supervisor IIT Hyderabad
ratios. for their support during the simulation work.
382 T. Karthikeya Sharma et al.

References [17] R.D. Reitz, R. Diwakar, Structure of High-Pressure Fuel Spray,

SAE Technical Paper Series, SAE Paper No. 870598, 1987.
[1] T. Karthikeya Sharma, G. Amba Prasad Rao, K. Madhu [18] R.D. Reitz, R. Diwakar, Effect of Drop Breakup on Fuel
Murthy, Performance of HCCI diesel engine under the influence Sprays, SAE Technical Paper Series, SAE Paper No. 860649,
of various working and geometrical parameters, Int. J. Theor. 1986.
Appl. Res. Mech. Eng. 1 (2012) (2012) 2319–3182. [19] C. Angelberger, T. Poinsot, B. Delhay, Improving Near-Wall
[2] K. Epping, S. Aceves, R. Bechtold, J.E. Dec, The potential of Combustion and Wall Heat Transfer Modeling in SI
HCCI Combustion for High Efficiency and Low Emissions (No. EngineComputations, SAE Technical Paper Series 972881,
2002-01-1923), SAE Technical Paper, 2002. 1997, pp. 113–130.
[3] B. Liu, W. Su, H. Wang, H. Huang, Characteristics and energy [20] M.A. Patterson, S.-C. Kong, G.J. Hampson, R.D. Reitz,
distribution of modulated multi-pulse injection modes based Modeling the Effects of Fuel Injection Characteristics on
diesel HCCI combustion and their effects on engine thermal Diesel Engine Soot and NOx Emissions, SAE Paper 940523,
efficiency and emissions, Front. Energy Power Eng. Chin. 1 (4) 1994.
(2007) 420–427. [21] J.B. Heywood, Internal Combustion Engine Fundamentals,
[4] J.B. Heywood, Internal Combustion Engine Fundamentals, McGraw-Hill Company, 1988.
McGraw-Hill Inc., New York, NY, 1988, ISBN 0-07-028637-X. [22] V. Moureau, G. Lartigue, Y. Sommerer, C. Angelberger, O.
[5] S. Onishi, S.H. Jo, K. Shoda, P.D. Jo, S. Kato, Active Thermo- Colin, T. Poinsot, Numerical methods for unsteady
Atmosphere Combustion (ATAC) – A New Combustion compressible multi-component reacting flows on fixed and
Process for Internal Combustion Engines, SAE Paper 790501, moving grids, J. Comput. Phys. 202 (2) (2005) 710–736.
1979. [23] M. Zellat, S. Duranti, Y. Liang, C. Kralj, G. Schmidt, J.M.
[6] T. Karthikeya Sharma, G. Amba Prasad Rao, K. Duclos, Toward a universal combustion model in STAR-CD for
Madhumurthy, Combustion analysis of ethanol in HCCI IC engines: from GDI to HCCI and application to DI Diesel
engine, Trends Mech. Eng. 3 (1) (2012). combustion optimization, in: Proceedings of 14th International
[7] Gen Chen, Norimasa Iida, Zuohua Huang, Numerical study of Multidimensional Engine User’s Meeting, SAE Cong, 2005.
EGR effects on reducing the pressure rise rate of HCCI engine [24] Y. Bakhshan et al, Multi-dimensional simulation of n-heptane
combustion, Front. Energy Power Eng. Chin. 4 (3) (2010) 376– combustion under HCCI engine condition using detailed
385. chemical kinetics, J. Engine Res. 22 (Spring) (2011).
[8] R. Manimaran, R. Thundil Karuppa Raj, Computational [25] D. Ganesh, G. Nagarajan, Homogeneous charge compression
studies of swirl ratio and injection timing on atomization in a ignition (HCCI) combustion of diesel fuel with external mixture
direct injection diesel engine, Front. Heat Mass Transfer formation, Energy 35 (1) (2010) 148–157.
(FHMT) 5 (1) (2014). [26] A. Viggiano, V. Magi, A comprehensive investigation on the
[9] S.P. Venkateswaran, G. Nagarajan, Effects of the re-entrant emissions of ethanol HCCI engines, Appl. Energy 93 (2012)
bowl geometry on a DI turbocharged diesel engine performance 277–287.
and emissions––a CFD approach, J. Eng. Gas Turbines Power [27] R. Song, T. Hu, L. Zhou, S. Liu, W. Li, Effects of compression
132 (12) (2010) 122803. ratio on the combustion characteristics of a homogeneous
[10] F. Maroteaux, L. Noel, A. Ahmed, Numerical investigations on charge compression ignition engine, Front. Energy Power Eng.
methods to control the rate of heat release of HCCI combustion Chin. 1 (4) (2007) 463–467.
using reduced mechanism of n-heptane with a multidimensional [28] Y. Gu, J.R. Mayor, W.J.A. Dahm, Turbulence-augmented
CFD code, Combust. Theor. Model. 11 (4) (2007) 501–525. minimization of combustion time in mesoscale internal
[11] J.P. Duclos, M. Zolver, T. Baritaud, 3D modeling of combustion engines, in: 44th AIAA Aerospace Sciences
combustion for DI–SI engines, Oil Gas Sci. Technol. 54 (2) Meeting and Exhibit, AIAA-2006-1350.
(1999) 259–264. [29] G. Chen, N. Iida, Z. Huang, Numerical study of EGR effects on
[12] O. Colin, A. Benkenida, The 3-Zone Extended Coherent Flame reducing the pressure rise rate of HCCI engine combustion,
Model (ECFM3Z) for computing premixed/diffusion Front. Energy Power Eng. Chin. 4 (3) (2010) 376–385.
combustion, Oil Gas Sci. Technol. – Rev IFP 59 (6) (2004) [30] B. Krishna, Murali, J.M. Mallikarjuna, Effect of engine speed
593–609. on peak tumble flows in a motored internal combustion engine –
[13] D. Abouri, M. Zellat, S. Duranti, F. Ravet, Advances in combustion an experimental investigation using particle image velocimetry,
modeling in STAR-CD: validation of ECFM CLE-H model to J. Appl. Fluid Mech. 4.1: 1-14-2011.
engine analysis, in: 18th International Multidimensional Engine [31] J. Martinez-Frias, S.M. Aceves, D. Flowers, J.R. Smith, R.
User’s Meeting at the SAE Congress, 2008. Dibble, Equivalence Ratio-EGR Control of HCCI Engine
[14] G. Subramanian, L. Vervisch, F. Ravet, New Developments in Operation and the Potential for Transition to Spark-ignited
Turbulent Combustion Modeling for Engine Design: ECFM- Operation (No. 2001-01-3613), SAE Technical Paper, 2001.
CLEH Combustion sub-model (No. 2007-01-0154), SAE [32] M. Christensen, B. Johansson, Influence of mixture quality on
Technical Paper, 2007. homogeneous charge compression ignition, J. Fuels Lubricants
[15] K.Y. Huh, A.D. Gosman, A phenomenological model of Diesel (1998) 107, SAE Transactions.
spray atomization, in: Proc. Int. Conf. on Multiphase Flows [33] J.O. Olsson, P. Tunestål, B. Johansson, Boosting for High Load
(ICMF ’91), Tsukuba, 24–27 September, 1991. HCCI (No. 2004-01-0940), SAE Technical Paper, 2004.
[16] C. Bai, A.D. Gosman, Development of Methodology for Spray [34] M. Christensen, B. Johansson, P. Amnéus, F. Mauss,
Impingement Simulation, SAE Technical Paper Series, SAE Supercharged Homogeneous Charge Compression Ignition
Paper No. 950283, 1995. (No. 980787), SAE Technical Paper, 1998.