A computational study to investigate the effects of insulation and EGR in a diesel engine

LINK DOWNLOAD MIỄN PHÍ TÀI LIỆU "A computational study to investigate the effects of insulation and EGR in a diesel engine": http://123doc.vn/document/560451-a-computational-study-to-investigate-the-effects-of-insulation-and-egr-in-a-diesel-engine.htm

I
NTERNATIONAL
J
OURNAL OF

E
NERGY AND
E
NVIRONMENT



Volume 3, Issue 2, 2012 pp.247-266

Journal homepage: www.IJEE.IEEFoundation.org


ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.
A computational study to investigate the effects of insulation
and EGR in a diesel engine


Syed Yousufuddin
1
, K.Venkateswarlu
2
, Naseeb Khan
3

1
Department of Mechanical Engineering, Jubail University College, Royal Commission - Jubail , P.O.
Box 10074, Jubail Industrial City- 31961, Kingdom of Saudi Arabia.
2
Department of Mechanical Engineering, K.L University Vaddeswaram, Guntur(Dist), Andhra Pradesh -
522502, India.
3
Shaaz College of Engg. & Tech. Himayatnagar, Moinabad Mandal, Ranga Reddy (Dist) - 500 075,
Andhra Pradesh, India.


Abstract
Higher heat losses and brake specific fuel consumption (BSFC) are major problems in an indirect
injection (IDI) diesel engine, which can be overcome by means of insulation. However, insulation
increases combustion temperature by about 200-250
0
compared to an identical standard IDI diesel
engine. Consequently, the NO
x
emission increases extremely due to this temperature increment. With the
proper adjustment of cold EGR mass fraction, it is possible to partially offset the adverse effect of
insulation on heat release rate and hence to obtain improved performance and lower NO
x
than the
baseline engine. At the first step of this work, the effects of insulation (without heat flux) on the
performance and emissions are studied at both part and full loads by a three dimensional model. The
results show that in the adiabatic case, BSFC is approximately 18% and 23% lower than baseline at the
full and part loads respectively. Also, soot emission shows 36% reduction at full load, while at the part
load, the value of which remains unchanged. At the second step of the present work, for reduction of
NO
x
production in the insulated engine, cold exhaust gas recirculation (EGR) method is utilized. Thus,
the model is studied at various cold EGR mass fractions, in which the EGR mass fraction increases from
0% to 30% at the same speed and operating loads. The optimum cold EGR mass fraction is obtained as
10% for part load operation. Results show that with adding this optimum EGR, the BSFC and NO
x

decrease by 15% and 6.5 % respectively at full load compared to the baseline and these reductions are
reached to 21% and 29% in the case of part load respectively, while it causes increment in soot emission
at full load operation and decreases slightly in the part load compared to the baseline engine. As an
engine is generally operated for a short time interval at full load condition, this increment can be omitted
when improvements in BSFC and decrease in NO
x
are considered together. The results of the model for
baseline engine are in good agreement with the corresponding experimental data. This agreement makes
the model a reliable tool that can be used for exploring new engine concepts.
Copyright © 2012 International Energy and Environment Foundation - All rights reserved.

Keywords: Adiabatic; Emission; Indirect injection; NO
x
; Soot.





International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

248
1. Introduction
Combustion characteristics of an engine are very important parameters for interpreting engine
performance and exhaust emissions. They are also useful for engine design and optimization. Besides,
the combustion characteristics such as maximum cylinder gas pressure ,temperature and heat-release rate
can be used to explain the effects of engine-operating conditions on the performance or can be used as
fundamental parameters for comparing various alternative fuels under the same operating conditions [1-
3]. The combustion characteristics of the (IDI) engines are different from (DI) engines, because of high
turbulence intensity and greater heat-transfer losses in the swirl chamber [4]. This defect causes the
brake-specific fuel consumption (BSFC) of the IDI engine to increase and the total engine efficiency to
decrease compared to that of a DI diesel engine. Because of these disadvantages of the IDI diesel
engines, most engine research has focused on the DI diesel engines. However, because of higher air
velocity and rapidly occurring air-fuel mixture formation in both combustion chambers of the IDI diesel
engines, these engines have a simple fuel injection system and lower injection pressure level [5, 6]. In
addition, they do not depend upon the fuel quality [5, 6] and produce lower exhaust emissions [7] than
DI diesel engines. Especially, unburned hydrocarbon (HC) and carbon monoxide (CO) emissions were
significantly lower in these engines, which have homogeneous charge condition [8, 9]. The concept of
low heat rejection (LHR) engine aims to reduce the great heat loss transferred to cooling system in an
IDI engine. So, this energy can be converted to useful work [10]. Some of the major advantages of LHR
engines include better fuel economy, increased engine life, reduction in HC, CO and PM emissions, and
lower combustion noise due to reduced pressure increasing rate, increased exhaust gas exergy, and ability
of operating lower cetane fuels [11-17]. Also, the higher temperatures in the combustion chamber can
have a positive effect on diesel engines at adiabatic case, due to the self-ignition delay drop [18]. A lot of
experimental study has been done to utilize LHR engine concept to improve thermal efficiency by
reducing heat losses, and to improve mechanical efficiency by eliminating cooling systems. The
experimental investigations of Adnan Parlak et al. and Ekrem Buyukkaya et al. [19, 20] revealed that
with the proper adjustment of the injection timing, it is possible to partially offset the adverse effect of
insulation on heat release rate and hence to obtain improved performance and lower NO
x
. Dickey
performed experimental investigation [21] on insulated engines and showed that shorter ignition delay,
decreased premixed fraction and a corresponding increase in the amount of fuel burned during the
diffusion phase of combustion were observed in the case of LHR engine. The experiments of S.
Jaichandar et al. [22] show that high temperature operation of LHR causes considerable improvement in
fuel consumption and thermal efficiency, increased availability in the exhaust gas and NOx formation, a
reduction of soot formation. The main factors of influence on the NO
x
emissions produced by the
combustion process are stoichiometric equivalence ratio and flame temperature. However, Because of the
diffusive mixing of air and fuel occurring along the spray envelope, combustion is dominated by near-
stoichiometric burning, where production of nitric oxide is high [23]. According to Yiming Wang et al.
[24], the characteristics of high temperature combustion, the methods, such as decreasing hole diameter,
optimizing injection timing and employing a special type of impingement plate in the combustion
chamber which breaks the envelope of high temperature flame, can reduce the combustion duration ,
achieve high efficiency and low soot emission in LHR engines. Also an experimental investigation on
the effect of ceramic coatings on diesel engine performance and exhaust emissions which was performed
by Dennis Assanis et al. [25] shows that in the optimum thickness of ceramic coating, efficiency
increases 10%, exhaust CO levels were between 30% and 60% lower than baseline levels and unburned
HC levels were 35% to 40% lower for the insulated pistons. Also the NO
x
concentrations were also 10%
to 30% lower due to the changed nature of combustion in the insulated engines. Finally, smoke emissions
decreased slightly in the insulated engines.
Numerical studies about energy and exergy of exhaust gas stream, which is the most important source of
available energy and exergy in a LHRE, have been performed by many authors [26-29] at the optimum
injection timing and have shown that lower heat rejection from the combustion chamber through
thermally insulated components causes an increase in available energy that would increase the in-
cylinder work and the amount of energy carried by the exhaust gases, which could also be utilized. Many
fundamental aspects concerning of CFD simulation of IDI engines have been discussed earlier by
Pinchon [30]. Three dimensional modeling of combustion process and soot formation in an indirect
injection diesel engine using KIVA CFD code have been progressed by Zellat [31]. Strauss and
Schweimer [32], at Volkswagen AG, studied the combustion and pollutant formation processes in a 1.9 l
IDI diesel engine using SPEED CFD code for a part and full load. The global properties are presented
International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

249
resolved for the swirl and main chamber and the swirl chamber throat separately. Also, the thermal NO
x

and soot formation are simulated and analyzed as well [32]. In fact, it is now commonly admitted that the
design of IDI combustion chambers has to rely increasingly more on fundamental knowledge of local
aspect requiring multidimensional simulation. As can be seen in the relevant literature, there are a few
attempts about using three dimensional modeling for finding the optimization state at adiabatic IDI diesel
engines up to now. In this present work, a CFD model has been used to predict flow field, combustion
process and emissions in the Lister 8.1 indirect injection diesel engine in baseline, adiabatic, and
adiabatic with EGR conditions. For optimization and simultaneously improvement of NO
x
emission and
performance at the adiabatic case, EGR applying method is used at both full and part loads. The EGR
percentage (EGR %) in this study was varied from 0% to 30%.

2. Problem formulation
2.1 Initial and boundary conditions
Calculations are carried out on the closed system from intake valve closure (IVC) at 165°CA BTDC to
exhaust valve open (EVO) at 125°CA ATDC. Figure 1 shows the numerical grid, which is designed to
model the geometry of the combustion chamber engine and contains a maximum of 42200 cells at
165°CA BTDC. It captures features like the glow plug channel, the piston and the swirl chamber throat.
The present resolution was found to give adequately grid independent results. A single hole injector is
mounted in pre-chamber as shown in Figure 2. Initial pressure in the combustion chamber is set to 86
kPa and initial temperature is calculated to be 384K. The Present work includes study of three operating
conditions of engine: base line, adiabatic and also adding EGR to the initial charge of adiabatic case at
two load operating conditions: a full load and a part load (50% load). At all the cases, engine speed is
730rev/min. All the wall boundaries temperatures were assumed to be constant throughout the simulation
in base mode, but allowed to vary with the combustion chamber surface regions. In adiabatic case, all
boundaries were assumed to be without heat flux. In EGR case 0% to30% of EGR mass fraction was
added to the inlet charge and it is assumed that EGR temperature is equal with temperature of inlet
charge (cold EGR).


(a) (b)

Figure 1. (a) Mesh of the Lister 8.1 indirect injection diesel engine, (b) Top view of the mesh

International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

250

Figure 2. Spray and Injector coordinate at pre-chamber

2.2 Model formulation
The numerical model is carried out for Lister 8.1 indirect injection diesel engine with the specifications
shown in Table 1. The governing equations for unsteady, compressible, turbulent reacting multi-
component gas mixtures flow and thermal fields were solved from IVC to EVO by the commercial AVL-
FIRE CFD code [33]. The turbulent flows within the combustion chamber are simulated using the RNG
ε
−k
turbulence model, modified for variable-density engine flows [34]. The standard WAVE model,
described in [35], is used for the primary and secondary break up modeling of the resulting droplets. At
this model, the growth of an initial perturbation on a liquid surface is linked to its wave length and other
physical and dynamical parameters of the injected fuel at the flow domain. Drop parcels are injected with
characteristic size equal to the Nozzle exit diameter (blob injection). The injection rate profiles at the full
and 50% loads are shown in Figure 3. The Dukowicz model is applied for treating the heat up and
evaporation of the droplets, which is described in [36]. This model assumes a uniform droplet
temperature. In addition, the droplet temperature change rate 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. A Stochastic dispersion model was employed to take the effect of interaction between the
particles and the turbulent eddies into account by adding a fluctuating velocity to the mean gas velocity
[37]. This model assumes that the fluctuating velocity has a randomly Gaussian distribution. The
spray/wall interaction model used in this simulation was based on the spray/wall impingement model
[38]. This model assumes that a droplet, which hits the wall is affected by rebound or reflection based on
the Weber number. The Shell auto-ignition model was used for modeling of the auto ignition [39]. 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
[33, 39]. The Eddy Break-up model (EBU) based on the turbulent mixing was used for modeling of the
combustion process in the combustion chamber [33] as follows:

⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
+
=
S
yC
S
y
y
C
r
pr
pr
ox
fu
R
fu
fu
1
.
,,min
ρ
τ
ρ

(1)

where this model assumes that in premixed turbulent flames, the reactants (fuel and oxygen) are
contained in the same eddies and are separated from eddies containing hot combustion products. The rate
of dissipation of these eddies determines the rate of combustion. In other words, chemical reaction occurs
fast and the combustion is mixing controlled .The first two terms of the “minimum value of” operator
determine whether fuel or oxygen is present in limiting quantity, and the third term is a reaction
probability which ensures that the flame is not spread in the absence of hot products. Above equation
includes three constant coefficients (
fu
C
,
R
τ
,
pr
C
) and
fu
C
varies from 3 to 25 in diesel engines. An optimum
International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

251
value was selected according to experimental data [33, 40]. NO
x
formation is modeled by the Zeldovich
mechanism and Soot formation is modeled by Kennedy, Hiroyasu and Magnussen mechanism [41].

Table 1. Specifications of Lister 8.1 IDI diesel engine

Cycle Type Four Stroke
Number of Cylinders 1
Injection Type IDI
Cylinder Bore 114.1 mm
Stroke 139.7 mm
L/R 4
Displacement Volume 1.43 lit.
Compression Ratio 17.5 : 1
V
pre-chamber
/V
TDC
0.7
Full Load Injected Mass
54336.6 −e
Kg per Cycle
Part Load Injected Mass
52009.3 −e
Kg per cycle
Max Power on 850 rpm 8 hp
Max Power on 650 rpm 6 hp
Injection Pressure 91.7 kg/cm
3

Start Injection Timing 20° BTDC
Nuzzle Diameter at Hole Center 0.003m
Number of Nuzzle Holes 1
Nuzzle Outer diameter 0.0003m
Spray Cone Angle 10°
IVO= 5° BTDC
IVC= 15° ABDC
EVO= 55° BBDC
Valve Timing
EVC= 15° ATDC

0
5
10
15
20
345 350 355 360 365 370 375 380 385 390
Crank Angle[deg]
Injection Rate
50percent load
full load

Figure 3. Part and full load injection histories

3. Results and discussion
The calculations are carried out for the single cylinder Lister 8.1 IDI Diesel engine and the operating
conditions are both full and 50% load at constant speed of 730 rev. /min. Figures 4 and 5 show the
comparison of computed and measured [40] mean in-cylinder pressure and Heat Release Rate (H.R.R)
respectively for both loads. The results presented in those figures are global (cylinder averaged)
quantities as a function of time (crank angle). The measured heat release rate curve is derived from the
first law analysis of procured in-cylinder pressure data as follows:

International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

252
θγθγ
γ
θ
d
dp
V
d
dV
p
d
dq
1
1
1 −
+
−
=
(2)

In the above equation, p and V are in-cylinder pressure and volume vs. crank angle
θ
, and
35.1=
γ
.
The good agreement between measured and predicted data especially during the compression, start of
combustion and expansion strokes verifies the model. Also Peak values for both premixed and diffusion
combustion phases are computed as well. Comparing these figures shows the effect of load on the heat
release rate, combustion duration and in-cylinder pressure. In order to study the part load condition, only
the fuel injection timing and the amount of injected mass were changed than those of full load and the
rest of the conditions were unaffected. As shown in Figure 4, the peak pressures discrepancy between
experiment and computation is less than 0.2%. Increasing load to full load causes in-cylinder peak
pressure to increase to 50.2 bar from 42.3 bar and ignition delay decreased to 7.9 CAD from 10.7 CAD
with respect to 50% load. It is clear from Figure 5 that at the full load operation, due to injection of fuel
in later parts of cycle and also longer injection duration, much of the fuel is burned in diffusion phase.
Also Combustion duration at full load operation is longer than that of part load operation. Figure 4 also
shows the quantity of computational and experimental results for Start of Combustion (SOC) and
Ignition Delay (I.D) crank angle degrees in both loads. The discrepancies of SOC or I.D between
computation and experiment at part load and full load operation are as small as 0.8° and 0.1° crank angle
respectively. Table 2 also shows a comparison of computational and experimental data for performance
parameters of base line engine. It is clear that there are good agreements between them. These
agreements between experimental measurements and numerical predictions in the case of baseline engine
makes the proposed model a reliable tool to be used for exploring new engine concepts.

0
10
20
30
40
50
200 230 260 290 320 350 380 410 440 470
CAD
Pressure [bar]
Experimental
model
motoring
0
5
10
15
20
25
30
35
40
45
200 230 260 290 320 350 380 410 440 470
CAD
Pressure [bar]
Experimental
model
motoring
(a) (b)

Figure 4. Comparison of measured [40] and calculated pressure for baseline engine at 730 rev. /min for
(a) full and (b) part load

0
20
40
60
80
340 360 380 400 420 440
Crank Angle[deg]
H.R.R[J/deg]
model
Experimental

0
10
20
30
40
50
60
340 360 380 400 420 440
Crank Angle[deg]
H.R.R[J/deg]
Model
Experimental

(a) (b)

Figure 5. Comparison of experimental [40] and calculated heat release rate for baseline engine at 730 rev.
/min for (a) full load (b) part load
International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

253
At Figures 6 and 7 are shown respectively in-cylinder gas pressure and temperature for base line,
adiabatic and adiabatic with 10% EGR at both loads operations. These results show that pressure in the
adiabatic without EGR and with 10% EGR cases increase 9 % and 7 % in respect to baseline respectively
at full load operating condition. Also it can be observed that in the part load operating condition pressure
increases 10 % and 8 % at adiabatic without EGR and with 10% EGR cases respect to baseline
respectively. Maximum pressure for baseline, adiabatic without EGR and with 10% EGR cases at full
load state are 50.4 bar, 55 bar, and 54.5 bar and those of at pat load are 44, 47.5, 47.5 bar respectively.
Peak of pressure for three cases can be observed at 6 degree ATDC. Results for temperatures in cylinder
are shown in Figure 7. It is understood from this curve that for the adiabatic case values of peak
temperature increase by 230 k and consequently causes that the amount of NO
x
emission increases. In
part load case maximum temperature for adiabatic condition reaches 1570k and increases by 160k
respect to the baseline. It is interesting to see that ignition delay with EGR and without EGR at the
adiabatic engine is similar to that of baseline engine. This means that the effect of decreasing oxygen
concentration and lower value of γ on ignition delay is equal to the increased charge temperature when
the engine is insulated. Also these figures show the effect of load on the cylinder pressure and
temperature at constant speed and injection timing. During part load operation as load is increased, more
fuel is injected later in the cycle. It can be assumed that the oxygen concentration in the prechamber
decreases as more fuel is injected. Therefore, the increases in fuel injection with load may reduce the
extent of burning of last portions of fuel injected into the prechamber. The unburned fuel in the
prechamber may therefore increase with load, but its burning rate in the main chamber may likewise
increase [42, 43]. This is primarily due to higher temperature reached in the main chamber during the
expansion stroke, as can be seen in temperature and pressure traces of Figures 6 and 7. Figure 8 shows
the behavior of heat release rate for three states at both part and full loads operation. It is clear from this
figure that at full load operation combustion period shifts towards expansion stroke, where pressure is
lower and total combustion duration gets long. Therefore, at full load operations, combustion starts in the
prechamber near 9
0
crank angle after fuel injection and then continuous at the main chamber while at the
part load combustion takes place only in prechamber. Also Figure 8 shows a comparison of various heat
release rate for both loads. The diagrams clearly show that with proper adjustment of EGR mass fraction
at full load operation, it is possible to partially offset the adverse effect of insulation on heat release rate
and hence to obtain improved performance and lower NO
x
. While the insulation with and without EGR
at part load operation does not have considerable effect on the amount of heat release rate. As shown in
Figure 9, accumulated heat releases in the adiabatic case is larger than the other cases thus, the amount of
burnt fuel in this case is more than the others. Because of shortage of oxygen in the charge of cylinder in
the EGR case, amount of burnt fuel and consequently accumulated heat release are less than the others.
Figure 10 shows the effect of insulation without and with EGR(EGR =10% rate) on the soot and NO
x

emissions in an IDI diesel engine. It can be seen that the NO
x
decreased significantly with EGR adding,
whereas the soot slowly increased. The effect of insulation and EGR adding on the soot emission is
found negligible at part load operation, as show in Figure 10. The results show that production of soot in
the both operating loads at the adiabatic condition is lower respect to base line. Reason of this happening
is increased cylinder gas temperature due to insulation of engine. Therefore, the effect of insulation in the
production of soot is positive. But diagrams for production of NO
x
emission in both part and full loads,
illustrate that production of NO
x
is very high in the adiabatic condition. In the adiabatic case the amount
of NO
x
is 172 % and 272 % more than baseline respectively in part and full loads and this is the negative
effect of temperature increasing in adiabatic case. Adding EGR to the initial charge is suitable method
for solving this problem. With this EGR adding to the adiabatic case we obtain lower NO
x
even less than
baseline engine in both operating loads. But it reduces performance characteristics of engine as power
and increases brake specific fuel consumption respect to adiabatic case. At full load operation, Soot
emission increases than that of baseline engine while it is not varied at part load operation. As an engine
is generally operated for a short time intervals at full load condition, this increment can be omitted when
improvements in BSFC and decrease in NO
x
is considered together.
Figure 11 represents the evolution of temperature, NO
x
and soot emissions at 360, 380, 400 and 420
crank angles for three cases at both part and full load operating conditions. It is clear from Figure 11a
that ,at 360°CA , flame propagations (temperature contours) are similar for three cases and combustion
started at the upper edge of swirl chamber throat(stoichiometric zone) and then propagated to pre and
main chamber at 380°CA . Also the development of the temperature fields at three cases between 360
and 380°CA show that the axial and the radial penetrations of the flame front are almost equal. Thus, the
International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

254
flame front reaches the lateral cylinder wall and the cylinder wall opposite to the pre-chamber at
approximately the same time. It reveals that the flame started in pre-chamber and then invades a large
portion of the main-chamber very quickly. Flame distribution which is shown in this figure also indicates
that the swirl generated in swirl chamber during the compression stroke becomes gradually weaker due
to opposing flow in glow plug channel (don’t shown contours). Therefore, at the end of injection period
at 380°CA the flame reaches near to injector location without distortion. Also the similar trends are
observed for flame temperature distribution at 360°CA and 380°CA. This would result in similar the
mixture formation processes for three cases at these crank angles. In the front view at 400°CA, the hot
gas from the pre-chamber reaches the opposite side of main chamber. This leads to the formation of two
large eddies each occupying a half of the main chamber and staying centered with respect to the two half
of the bowl(don’t shown contours). At 420°CA, these eddies are larger than that of 400°CA and the eddy
formed in right half of bowl is larger than that of left half of bowl. At 400°CA, the shape of high
temperature regions for baseline and adiabatic with EGR cases are the same. Also at adiabatic case for
400°CA and 420°CA, these regions more spread out in the main chamber and the values of local peak
temperature in cylinder are higher than those of two other cases. After 400°CA, these regions transfer to
the main combustion chamber and therefore, at part load operation most of NO
x
emission form in the
main chamber or transfer to it from prechamber. At the full load operation, the momentum of fuel spray
increases because of the increase of injection pressure and the increase of the injected fuel mass due to
the increased effective flow area with increased nozzle needle lift. Therefore, as shown in Figure 11d,
combustion started earlier at the full load operation because of fast mixture formation. It is clear from
this figure that flame can develop in whole of glow plug channel and stagnation zone because of fast
mixture formation and high swirl ratio in the prechamber at 380CA and 400°CA for three cases. Also
similar trends are observed for flame propagation at other crank angles. At Figures 11b, 11c, 11e, 11f the
production of NO
x
and soot in the main and prechamber are discussed for three cases at both part and full
loads operation. It can be seen from Figure 11b that at 380CA the NO
x
is produced in the throat and main
chamber for three cases at part load operation. At 400CA and 420CA reducing NO
x
quantities in the
swirl chamber throat is due convection to the main chamber. Also similar trends are observed for
baseline and adiabatic with EGR cases while at the adiabatic cases increase in the global NO mass
fractions follow the higher global temperature in cylinder. The main chamber NO formation is much
higher than that of swirl chamber at three cases. On the other hand the lower global temperature and the
lower heat release, encountered in the swirl chamber of engine, lead to a much lower production of NO
x
.
As shown in Figure 11e, at 380°CA, the NO forms in the swirl chamber throat, pre and main chambers
and then transferred into the main chamber at 400°CA for baseline engine. While at adiabatic without
and with EGR cases because of the high temperature in both chambers, the NO emission exist in the pre
and main chambers. After the end of injection lean mixture is formed in the upper region of prechamber
at adiabatic cases, which with the high temperatures and the slow movement leads to a center of NO
formation. At 420°CA most of the NO
x
is found in the main chamber.


full load
0
10
20
30
40
50
60
200 250 300 350 400 450
CAD
pressure(bar)
base line
adiabatic
0.1 EGR
part load
0
10
20
30
40
50
60
200 250 300 350 400 450
CAD
pressure(bar)
baseline
adiabatic
0.1 EGR
(a) (b)

Figure 6. Comparison of calculated pressure at 730 rev/min at three operating conditions for (a) full and
(b) part load

International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

255
full load
300
600
900
1200
1500
1800
2100
2400
200 250 300 350 400 450
CAD
temprature(k)
base line
adiabatic
0.1 EGR
part load
300
600
900
1200
1500
200 240 280 320 360 400 440 480
CAD
temprature(k)
base line
adiabatic
0.1 E.G.R
(a) (b)

Figure 7. Comparison of calculated temperature at 730 rev/min at three operating conditions for (a) full
and (b) part load


full load
0
20
40
60
80
330 360 390 420 450 480
CAD
H.R.rate(J/deg)
base line
adiabatic
0.1 EGR

part load
0
10
20
30
40
50
340 370 400 430 460 490
CAD
H.R.RATE(J/deg)
base line
adiabatic
0.1 EGR

(a) (b)

Figure 8. Comparison of calculated heat release rate at 730 rev/min at three operating conditions for (a)
full and (b) part load


full load
0
600
1200
1800
2400
3000
320 360 400 440 480
CAD
accumulated heat release(J)
base line
adiabatic
0.1 EGR

part load
0
300
600
900
1200
1500
320 350 380 410 440 470
CAD
accumulated heat release(J)
base line
adiabatic
0.1 EGR

(a) (b)

Figure 9. Comparison of calculated accumulated heat release rate at 730 rev/min at three operating
conditions for (a) full and (b) part load

International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

256
full load
0
0.75
1.5
2.25
3
3.75
340 380 420 460
CAD
soot (g/kw.h)
base line
adiabatic
0.1 EGR
part load
0
0.5
1
1.5
2
2.5
330 360 390 420 450 480
CAD
soot(g/kw.h)
base line
adiabatic
0.1 EGR
(a) Soot at full load (b) soot at part load
full load
0
1.5
3
4.5
6
7.5
9
330 360 390 420 450 480
CAD
NO (g/kw.h)
base line
adiabatic
0.1 EGR
part load
0
0.5
1
1.5
2
2.5
3
3.5
4
350 370 390 410 430 450 470 490
CAD
NO(g/kw.h)
base line
adiabatic
0.1 E.G.R
(c) NO at full load (d) NO at part load

Figure 10. (a-d) Comparison of calculated exhaust emissions at 730 rev/min at three operating conditions
at full and part loads

Finally in Figures 11c and 11f, we present the evolution of the soot mass in the main and pre chambers
for three cases at part and full load operation respectively. As indicated by these figures, in the three
cases at part and full load conditions, the main cause of the exhaust smoke is spray-wall impingement
which leads to fuel adhesion on the wall and the stagnation of a rich fuel-air mixture. Under part load
operation, fuel spray impinges against the chamber wall from the beginning of injection due to the low
swirl ratio in swirl chamber. Therefore in Figure 11c, dense soot cloud shown by the white area appears
near the spray-wall impinging point from the early stage of the combustion period. Some part of this
dense soot flows out to the main combustion chamber just when the initial flame flows out (at 380°CA of
three cases, in Figure 11a). This simultaneous flow out is confirmed by the observation of the main
combustion chamber that the dense soot appears at almost the same time of the flame (at 380°ATDC of
Conventional, in Figures 11a and 11c ). As the formation of the dense soot continues throughout the
injection period due to the continuous fuel-spray impinging, the dense soot continuously flows out to the
main combustion chamber and cannot be fully oxidized. This incomplete oxidation is shown by the fact
that the dense soot does not disappear and spreads to the whole area of the main combustion chamber (at
400°CA in Figure 11c) while a part of soot emission remains in glow plug channel because of the
stagnation of a rich fuel-air mixture region. At 420°CA, dense soot in the main chamber disappears
because of complete oxidation. Also at full load operation, the harder spray impingement than the part
load operation causes more fuel adhesion on the wall near this point. This adherent fuel is not quickly
evaporated and formed fuel vapor is hardly carried out of this area, because the stagnation zone is formed
here due to the chamber shape. Thus, the rich fuel-air mixture stagnates in this zone under the condition
of high temperature and insufficient oxygen to form the dense soot cloud. At 360°CA, the soot is
produced in regions of high fuel concentrations, when cold fuel is injected into areas of hot gases at
upper edge of swirl chamber throat. It is obvious from flame distributions (Figures 11a and 11d) that in
the full load operation, the flow field more interact with the spray than in the part load operation.
Therefore, at 380°CA, The zones of the soot formation are occurred in the glow plug channel and in the
International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

257
center of swirl chamber. Similar trends are observed at this crank angle for three cases. At 400°CA, there
are two centers of soot production visible in the contour plots, one has attached itself to the glow plug
channel opposing wall and the other spreads itself to the whole area of the glow plug channel. At high
loads, we observe that the main part of soot formation takes place in the prechamber due to low swirl
ratio and the insufficient oxygen mass available. It can be seen from Figures 11a, 11b, 11d and 11e that
the NO
x
formation in the main chamber is intensified in areas with equivalence ratio close to 1 and the
temperature higher than 2000 K. In addition to this, as shown in these figures that the area which the
equivalence ratio is higher than 3 and the temperature is approximately between 1600 K and 2000 K
yields in greater soot concentration. A local soot-NO
x
trade-off is evident in these contour plots where
the NO
x
and Soot formation occur on opposite sides of the high temperature region.


360° 380° 400° 420°

baseline



adiabatic



Adiabatic
with EGR





Figure 11. (a) Contour plots of temperature at three cases for part load operation at 360, 380, 400,
420°CA (from left to right)










International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

258
360° 380° 400° 420°

baseline


adiabatic


EGR




Figure 11. (b) Contour plots of NOx mass fraction at three cases for part load operation at 360, 380, 400,
420°CA (from left to right)























International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

259
360° 380° 400° 420°

baseline


adiabatic


EGR




Figure 11. (c) Contour plots of Soot mass fraction at three cases for part load operation at 360, 380, 400,
420°CA (from left to right)
























International Journal of Energy and Environment (IJEE), Volume 3, Issue 2, 2012, pp.247-266
ISSN 2076-2895 (Print), ISSN 2076-2909 (Online) ©2012 International Energy & Environment Foundation. All rights reserved.

260
360° 380° 400° 420°

baseline


adiabatic


EGR




Figure 11. (d) Temperature contour plots at three cases for full load operation at 360, 380, 400, 420°CA
(from left to right)