Int. J. Vehicle Design, Vol. x, No.

x, xxxx 1
Copyright 200x Inderscience Enterprises Ltd.
Application of Taguchis methods to investigate
factors affecting emissions of a diesel engine running
with tobacco oil seed methyl ester
Adnan Parlak*
Yildiz Technical University,
Naval Architech and Marine Engineering,
stanbul, Turkey
E-mail: parlak@sakarya.edu.tr
*Corresponding author
Hlya Karaba
Sakarya University,
Vocational School of Akyazi,
Sakarya, 54000, Turkey
brahim zsert, Vezir Ayhan and dris Cesur
Technical Educational Faculty,
Sakarya University,
Sakarya, 54187, Turkey
E-mail: ozsert@sakarya.edu.tr
E-mail: vayhan@sakarya.edu.tr
Abstract: Different kinds of vegetable oils and their methyl/ethyl esters have been
tested in diesel engines. However, studies reporting the effects of tobacco seed oil
and Tobacco Seed Oil Methyl Ester (TSOME) on emissions of diesel engines are
limited. One of the most important issues is to investigate parameters that affect
the yield of biodiesel and their interactions on emissions of a diesel engine. The
Taguchi method is a useful tool for this purpose. Two different catalysts (KOH
and NaOH), four different blends (B10, B20, B50 and B100) and four engine
speeds were used during full-load tests. Optimal catalyst type, engine speed and
TSOME blends on exhaust emissions were determined using Taguchis technique.
The Taguchi design method revealed that choosing right catalyst and the blend rate
are important two factors in view of minimisation of pollutant emissions.
Keywords: Taguchi method; catalyst; emissions; tobacco; methyl ester.
Reference to this paper should be made as follows: Parlak, A., Karaba, H.,
zsert, I., Ayhan, V. and Cesur, I. (xxxx) Application of Taguchis methods to
investigate factors affecting emissions of a diesel engines running with TSOME
blends, Int. J. Vehicle Design, Vol. x, No. y, pp.xxxxxx.
Biographical notes: AUTHOR PLEASE SUPPLY CAREER HISTORY OF NO
MORE 100 WORDS FOR EACH AUTHOR.
Author
please
supply
E-mail id for
remaining
authors.
IJVD 0(0) 07 Parlak et al.indd 1 1/18/2012 11:55:08 AM
2 A. Parlak et al.
1 Introduction
Vegetable oils are renewable, non-toxic, biodegradable, and have low-emission proles but
NOx (Hasimoglu et al., 2008; Altin et al., 2001; Karaosmanolu et al., 2000). Signicant
studies have been conducted on the effects of various raw vegetable oils, vegetables oil
methyl/ethyl esters on diesel engines. While some researchers focused on the effects of
vegetable oils and their esters on the performance (Nwafor and Rice, 1996; Sapaun et al.,
1996; Mc Donnel et al., 2000), some of the other studies have been focused on emissions
characteristics of engine fuelled with vegetable oil methyl esters (Murayama et al., 1984;
Scholl and Sorenson, 1993; Ali et al., 1995; Arregle et al., 1999). Among them, Murayama
et al. (1984) reported that vegetable oils and methyl ester of rapeseed oil offered lower
smoke and oxides of nitrogen (NO
X
) emissions. Scholl and Sorenson (1993) reported that
carbon monoxide, oxides of nitrogen (NO
X
) and smoke emissions were slightly lower for
soybean ester than that of diesel, whereas hydrocarbon (HC) emission showed 50% reduction
compared with diesel.
The effects of vegetable oil fuels and their methyl esters (raw sunower oil, raw cottonseed
oil, raw soybean oil and their methyl esters, rened corn oil, distilled opium poppy oil and
rened rapeseed oil) on a direct injected, four stroke, single-cylinder diesel engine exhaust
emissions was investigated by Altin et al. (2001). Niemi et al. (2002) reported that the carbon
monoxide emission was higher at all loads for different speeds with preheated-mustard oil as
fuel. The authors concluded that emissions decreased with preheating the oil.
Kalligeros et al. (2003) analysed the emission characteristics on a stationary diesel
engine fuelled with sunower oil methyl ester/diesel blends. They observed decreases in
particulate matter, carbon monoxide, hydrocarbon and nitrogen oxide emissions. Dorado
et al. (2003) tested the use of methyl ester of used olive oil as fuel in a direct-injection
diesel engine. They reported that carbon monoxide, carbon dioxide, oxides of nitrogen and
sulphur dioxide emissions decreased by 59%, 8.6%, 32% and 57%, respectively, and that
the smoke emission was low. They concluded that the methyl ester of olive oil could be
used as fuel.
Labeckas and Slavinskas (2006) analysed the emission characteristics of four stroke,
four-cylinder, direct injection, unmodied, naturally aspirated diesel engine when operating
on neat Rapeseed Methyl Ester (RPE) and its 5%, 10%, 20% and 35% blends with diesel
fuel. They found that carbon monoxide, hydrocarbon and visible emissions have decreased
while an oxide of nitrogen emissions increased for methyl ester compared with diesel.
As can be seen from the literature, there are various kinds of vegetable oil methyl/ethyl
esters, which were investigated their effects on performance and emission characteristics.
Among them, Tobacco Seed Oil (TSO) was also considered as a potential alternative fuel for
compression ignition engines (Parlak et al., 2009; Usta, 2005a, 2005b).
There is less study TSO and TSOME on performance and emissions has not been tested in
diesel engines, yet. In studies conducted by Usta (2005a, 2005b), blends containing TSOME
in 10%, 17.5% and 25% proportions by volume were tested in a turbocharged-diesel engines
for the engine loads of 50%, 75% and 100%. The author reported that maximum power was
observed with 17.5% TSOME blend. The increase in brake power and thermal efciency
were about 3% and 2%, respectively. The results also showed that the addition of TSOME to
the diesel fuel reduced CO and SO
2
emissions while causing slightly higher NO
X
emissions.
In his other study, the author reported that TSOME addition (up to 25% in volume) did not
cause any signicant variation in the engine performance. On the contrary, the TSOME
IJVD 0(0) 07 Parlak et al.indd 2 1/18/2012 11:55:08 AM
Application of Taguchis methods 3
blends resulted in slightly higher torque and power than the diesel fuel at full load owing to
its slightly higher density and viscosity.
Investigation of the parameters, which affects the yield of biodiesel on engine
emissions of a diesel engine is important. The optimum operating parameters for a given
system can be determined using experimental techniques but experimental procedure
will be time consuming and expensive when the number of parameters is in the order of
20, 30, etc., like in the case of IC engines (Murugesan et al., 2009). The most common
optimisation techniques used for engine analysis are simplex method (Shroff and Hodgetts,
1974), response surface method (Satake et al., 2008), Articial Neural Network (ANN)
(Parlak et al., 2006), Genetic Algorithm (GA) (Alonso et al., 2007) and Taguchi method.
Taguchis technique has been popular for parameter optimisation in Design of Experiments
(DOE) for decades. Basic principle of Taguchi methods is to develop an understanding of
individual and combined effects of variety of design parameters from a minimum number
of experiments (Sava and Kayikci, 2007).
Nowadays, some researches begin to use Taguchi method for optimisation in
internal-combustion engines. Among them, Win et al. (2005) applied the Taguchi method
for optimising the diesel engine operation and injection-system parameters for low noise,
emission and fuel consumption. Anand and Karthikeyan (2005) applied Taguchi method
to optimise engine design and operating parameters for improving the engine efciency
without considering the combustion parameters. Murugesan et al. (2009) carried out an
optimisation analysis of direct-injection diesel engine run by Jatropha biodiesel using a
thermodynamic model in combination with Taguchi method.
However, there is no study using this technique in optimisation of some important factors
affecting exhaust emissions of a diesel engines running with TSOME. In the experimental
design, the parameters affecting the emissions are chosen as engine speed, catalyst type and
the amount of TSOME in the blend. The conditions, which maximise the brake torque and
the conditions, which minimise the brake-Specic Fuel Consumption (SFC) and exhaust
emissions were investigated.
2 Materials and methods
2.1 Production of TSOME
Transesterication method was used for producing TSOME. The weight of the oil, alcohol
and catalyst was measured by 0.0001 g sensitivity. Base catalyst was chosen because acid
value is found about around 1% in the oil analyses. Methyl alcohol in 99% purity was used
for transesterication process. Measured and premixed methyl alcohol and catalysts (KOH
and NaOH) mixtures were poured on glass beaker, then the catalyst is stirred until all catalyst
was completely resolved with alcohol. After TSO was heated up to desired temperature,
the prepared alcoholcatalyst mixture was added to TSO to start the transesterication
reaction by using the heating bath and glass balloon (1 lt) of Buchi rotary evaporator. The
temperature sensitivity of the heating bath was 0.1C. The mixture was stirred for 1 h and
then it was taken into the separatory funnel and waited until esterglycerin separation is
taken place.
After the separation process, glycerin, which was moved to bottom of the separatory
funnel was removed. It is necessary to clean the TSOME by sterile puried water to separate
IJVD 0(0) 07 Parlak et al.indd 3 1/18/2012 11:55:08 AM
4 A. Parlak et al.
remaining glycerin, mono- and di-glycerides. Hot-sterile puried water was added to
TSOME in the separatory funnel for the cleaning process, then the separatory funnel was
shaken upside down by repeating four times. After this process, TSOMEWater mixture
was left to settle for 12 h and then pure water and glycerin glimmers were removed from
the separatory funnel. After all this stages, remaining TSOME was separated from all of
the unwanted particles by using an centrifuge separatory device called NUVE NF400 in
4000 rpm and 3600 RCF for 1 h. Finally, TSOME sample was dried by heating up to 110 C
for 30 min. Table 1 shows the analysis of methyl ester.
Table 1 Analysis of methyl ester
Method KOH NaOH
Ester content, % (m/m) prEN 14103 96.5 97.0
Carbon residue, % (m/m) ENISO10370 0.17 0.17
Copper band corrosion (3 h at 50

C) ENISO2160 1.0 1.0
Cold lter plug point, C EN116 7.0 10.0
Water content, mg/kg ENISO12937 300.0 400.0
Total contamination, mg/kg EN12662 20.0 23.0
Idoine value, g iodine/100 g EN1411 122.0 118.0
Methanol content, % (m/m) EN1410 0.20 0.18
Triglyceride content, % (m/m) EN1410 0.11 0.11
Di-glycerites content, % (m/m) EN1410 0.20 <0.05
Total glycerol, % (m/m) EN1410 0.02 0.02
Phosphorus, mg/kg EN14107 4.0 <4.0
Monoglyceride content, % (m/m) EN14105 0.29 <0.29
Density, g/cm
3
, 15C EN ISO 3675 0.88 0.86
Cetan number ISO 3104 54.5 49.0
Lower heat value, MJ/kg DIN 51900-1 39.16 40.02
Sulphur, mg/kg EN ISO 20846 0 0.0
Viscosity, mm
2
/s EN ISO 3104 4.88 3.5
Pour point, C ASTM D97 6 12.0
Flash point,C ASTM D93 >100 152
2.2 Experimental setup
The engine was a single cylinder, naturally aspirated, four strokes, and water cooled, direct
injection. Diesel engine with a bowl in piston-combustion chamber. In line type high-pressure
injection pump and hole type spring-loaded injector were used in the fuel system. Test setup
is shown in Figure 1. Table 2 the specications of the engine are given in. A load cell with the
precision of 0.1 N was used for measuring brake load. Full-load tests were conducted with the
engine speeds of 1200, 1400, 1600, 1800, 2000, 2200 and 2400 rpm at constant static injection
timing of 35 Crank Angle (CA), Before Top Dead Centre (BTDC). The fuel consumption is
measured by an electronic controlled balance with the precision of 0.1 g. At each operating
condition, the dynamometer load, speed, fuel and airow rates were recorded after allowing
sufcient time for the engine to stabilise. All the total uncertainties of performance characteristics
are calculated. The accuracies and total uncertainties of characteristics calculated with respect
to measured values are shown in Table 3. After the load tests were conducted for the standard
engine with injection timing of 35 CA, the same test sequence was conducted for the engine
tests with TSOME blends of 10%, 20%, 50% and 100% on mass base.
AUTHOR
PLEASE
CHECK
THE
SPELLING
OF DI-
GLYCE-
RITES IS
CHANGED
TO DI-
GLYCE-
RIDES
CHANGE
IS OK
IJVD 0(0) 07 Parlak et al.indd 4 1/18/2012 11:55:08 AM
Application of Taguchis methods 5
Figure 1 Test setup
Table 2 Specication of the test engine
Engine type Super star water cooled
Bore [mm] 108
Stroke [mm] 100
Cylinder number 1
Stroke volume [l] 0.92
Injection pressure [MPa] 17.5
Injection advance [CA, BTDC] 35
Maximum speed [rpm] 2500
Cooling type Water
Injection type DI
Table 3 The errors in parameters and total uncertainties
Parameters Systematic errors,
Load, N 0.1
Speed, rpm 1.0
Time, s 0.1
Temperature, C 1.0
Fuel consumption, g 0.1
NOx, ppm 5% of measured value
CO, % 5% of measured value
HC, ppm 5% of measured value
Smoke, % 1
Total uncertainty, %
Specic fuel consumption 1.5
Brake torque, Nm 1.1
IJVD 0(0) 07 Parlak et al.indd 5 1/18/2012 11:55:08 AM
6 A. Parlak et al.
2.3 Design of experiment
The present experiments were designed to apply the Taguchis methods to establish the
effects of two catalysts, ve TSOME blending rates and four engine speeds for the purpose
of determining optimal conditions of the exhaust emissions. The three design parameters
(factors) and their levels are given in Table 4. Optimum experimental conditions, which
maximise the brake torque and minimise the emissions of SFC, NOx, CO, CO
2
, HC and
smoke were determined by Taguchis methods. In experimental design, as the effects of
engine speed on performance is well-known in comparison witht the catalyst types and
the amount of TSOME in blend, four engine speeds, which is important for the engine
performance are selected.
Table 4 Design factors and their levels
Symbols Factors 1 Level 2 Level 3 Level 4 Level
A Catalyst type KOH NaOH
B Engine speed, rpm 1000 1400 1800 2200
C Blends, % m/m 10 20 50 100
Basic principle of Taguchi methods is to develop an understanding of individual and
combined effects of variety of design parameters from a minimum number of experiments.
Taguchi method uses a generic Signal-to-Noise (S/N) ratio to quantify the present variation.
There are several S/N ratios available depending on the type of characteristics including
Lower is Better (LB), Nominal is Best (NB), and Higher is Better (HB). Since the
lower SFC and exhaust emissions are vital in the engine tests, the S/N ratio for the LB
characteristics is related to the present study, which is given by
S/N = 10log
2
1
1
n
i
i
y
n

j \
, (
( ,

(1)
Similarly, higher-brake torque are vital in the engine tests, the S/N ratio for the HB
characteristics is given as following, which is given by Sava Kaykci (2007)
S/N = 10log
2
1
1 1
n
i i
n y

j \
, (
( ,

(2)
where n is the number of repetition in a trial under the same design conditions, y
i
represents
the measured value, and subscript i indicates the number of design parameters in the
Orthogonal Array (OA). In the Taguchi method, a design parameter (factor) is considered
to be signicant if its inuence is large compared with the experimental error as estimated
by the Analysis of Variance (ANOVA) statistical method given by equations (3)(7) shown
below. If this is the case, the design parameter is a critical factor in determining the optimal
solution to the design problem:
SS
T
=
( )
2
2
1
/
N
i
T
S N i
N

, ]

, ]
]

(3)
IJVD 0(0) 07 Parlak et al.indd 6 1/18/2012 11:55:09 AM
Application of Taguchis methods 7
SS
A
=
2 2
1
A
K
i
i Ai
A T
n N

, ] j \

, ]
, (
( ,
]

(4)
v= N 1 (5)
V
factor
=
v
factor
SS
(6)
F
factor
=
factor
V
error
V
(7)
where, SS
T
is the sum of squares owing to total variation, N is the total number of experiments,
SS
A
represents the sum of squares owing to factor A, K
A
is number of levels for factor A. A
i

stands for the sum of the total ith level of the factor A, n
Ai
is the number of samples for ith
level of factor A. T is the sum of total (S/N) ratio of the experiments, v is the degrees of
freedom, V
factor
is the variance of the factor, SS
factor
represents the sum of squares of the factor
and F
factor
is the F ratio of the factor.
In Taguchi methods, the levels of factors given in the ANOVA are meaningful according
to 90% and 99% condence intervals. Design of experiments is composed considering
these condence limits. Required the minimum experimental layout according to Taguchi
methods are given in Table 5.
Table 5 Experimental layout
Exp. No
Factors
[A] [B] [C]
1 1 1 1
2 1 1 2
3 2 1 3
4 2 1 4
5 1 2 3
6 1 2 4
7 1 2 1
8 2 2 2
9 1 3 3
10 1 3 4
11 2 3 1
12 2 3 2
13 1 4 1
14 1 4 2
15 2 4 3
16 2 4 4
IJVD 0(0) 07 Parlak et al.indd 7 1/18/2012 11:55:09 AM
8 A. Parlak et al.
3 Results and discussion
3.1 Determination of optimal operation conditions by Taguchi method
In this study, the effects of using TSOME blends, which is yielded for KOH and NaOH
catalyst on exhaust emissions of a direct-injection diesel engines are studied. To investigate
the factors of catalyst type, engine speed and mass base TSOME percentage in the blend
(blending rate), Taguchis methods were used.
B10, B20, B50 and B100 blends of TSOME on exhaust emissions were tested for
comparison with data of standard-diesel engine. Two different catalysts (KOH and NaOH)
and four different blends were used during full-load tests. Optimal catalyst type, engine speed
and TSOME blend affecting emissions of diesel engine were determined using Taguchis
technique. Although the engine was tested for seven engine speeds for full-load testing, four
engine speeds was only taken for experimental design and in determining optimal conditions
for performance as the solution is rather simplied.
Experiments were designed considering the requirements of the increase in the brake
torques and the necessity of lowering the SFC and exhaust emissions. Table 6 shows ANOVA
for the parameters considered. According to the results of ANOVA, it is shown that engine
speed, blending rate and catalyst type affects the SFC within a 99% condence level. The
effect of blend rate on SFC, brake torque and brake power are found meaningful comparing
with the others factors.
Table 6 The Analysis of Variance (ANOVA)
Factors
Sum of squares,
SS
Degree of
freedom v
Variance,
VT
F
factor
TORQUE [A] Catalyst 0.70 1 0.70 44.21***
[B] Speed 1.15 3 0.38 24.11***
[C] Blends, % 0.41 3 0.14 8.65***
Total 2.27 7 0.32
Error 0.13 8 0.02
SFC [A] Catalyst 0.50 1 0.50 33.19***
[B] Speed 1.76 3 0.59 38.96***
[C] Blends, % 1.76 3 0.59 38.92***
Total 4.02 7 0.57
Error 0.12 8 0.02
NO
X
[A] Catalyst 0.22 1 0.22 10.47**
[B] Speed 18.99 3 6.33 295.30***
[C] Blends, % 0.70 3 0.23 10.89***
Total 19.91 7 2.84
Error 0.17 8 0.02
HC [A] Catalyst 0.14 1 0.14 0.03
[B] Speed 527.38 3 175.79 37.88***
[C] Blends, % 77.73 3 25.91 5.58**
Total 605.25 7 86.46
Error 37.12 8 4.64

IJVD 0(0) 07 Parlak et al.indd 8 1/18/2012 11:55:09 AM
Application of Taguchis methods 9
Factors
Sum of squares,
SS
Degree of
freedom v
Variance,
VT
F
factor
CO [A] Catalyst 4.05 1 4.05 2.164463
[B] Speed 98.99 3 33.00 17.63***
[C] Blends, % 20.20 3 6.73 3.59*
Total 123.24 7 17.61
Error 14.97 8 1.87
CO
2
[A] Catalyst 0.54 1 0.54 10.66**
[B] Speed 0.76 3 0.25 5.01**
[C] Blends, % 0.09 3 0.03 0.61
Total 1.39 7 0.20
Error 0.40 8 0.05
Smoke [A] Catalyst 0.00 1 0.00 0.000574
[B] Speed 11.01 3 3.67 1.803275
[C] Blends, % 19.77 3 6.59 3.23*
Total 30.78 7 4.40
Error 16.28 8 2.04

*At Least 90% condence.
**At Least 95% condence.
***At Least 99% condence.
Table 7 shows S/N values of the factors according to experimental layout designed by
Taguchi methods. Observed values in the table are the real values, which were calculated
using the experimental data. According to Taguchi techniques, to determine the optimal
conditions and to compare the results with the expected conditions, it is necessary to perform
a conrmation experiment. If the generated design fails to meet the specied requirement,
the process must be reiterated using a new system until the required criteria are satised.
In the study, the required conrmation tests were done and the results were found completely
condence with the observed values.
3.1.1 Brake torque and specic fuel consumption
Optimal conditions were determined by using Taguchi method for brake power. As the
generated design has not been included in the main experimental layout, the process was
re-iterated until the required criteria are satised. After conrmations test carried out in
99% condence level, the optimum design parameter (factor) combination were found as
A
1
B3C
3
(KOH- 2200d/d-B50) for brake torque and as A
1
B
2
C
1
(NaOH-1400 rpm-B10) for
SFC. Figures 2 3 show S/N values of factor levels for the brake torque and SFC. Since
this combination of design parameter had already been included in the main experimental
layout (see Experiment 7 in Table 5) there was no need to carry out an extra conrmation
experiment. According to the ANOVA, catalyst type and blend rate of TSOME are
meaningful within 99% condential level. Higher torque was obtained with KOH catalyst
and B50 TSOME blend. Similarly, lower SFC was measured with KOH catalyst and B10
TSOME.
Table 6 The Analysis of Variance (ANOVA) (continued)
IJVD 0(0) 07 Parlak et al.indd 9 1/18/2012 11:55:09 AM
10 A. Parlak et al.
Figure 2 S/N values of factor levels for brake torque
Figure 3 S/N values of factor levels for brake SFC
3.1.2 Exhaust emissions
According to ANOVA in Table 6, catalyst type and blend rate are signicant on NOx
emissions within 95% and 90% condence level, respectively. The catalyst type and TSOME
blend rate strongly affect NOx, HC and CO above 90% condence level. The effect of
catalyst type on smoke emission is not observed above 90% condence level. The change
of NOx strongly depends on the engine speed. ANOVA conrm that engine speed affect the
NOx emissions within 99% condence limit. Same effects were observed with the pollutant
emissions of HC, CO within 99% condence level. However, the effects of TSOME blend
on emissions were meaningful within 90% condence level.
S/N values of factor levels of smoke, NOx, HC, CO and CO
2
emissions for B10, B20,
B50, B100 blends of TSOME are shown in Figures 48. After conrmations test carried out
in 99% condence level, the optimum design parameter (factor) combination were found as
IJVD 0(0) 07 Parlak et al.indd 10 1/18/2012 11:55:09 AM
E
x
p
.
n
o
.
T
O
R
Q
U
E

(
N
m
)
S
F
C

(
g
/
k
W
h
)
N
O
X

(
p
p
m
)
H
C

(
p
p
m
)
C
O

(
%
)
C
O
2

(
%
)
S
m
o
k
e
,

%
O
b
s
e
r
v
e
d
v
a
l
u
e
S
/
N
r
a
t
i
o
O
b
s
e
r
v
e
d
v
a
l
u
e
S
/
N
r
a
t
i
o
O
b
s
e
r
v
e
d
v
a
l
u
e
S
/
N
r
a
t
i
o
O
b
s
e
r
v
e
d
v
a
l
u
e
S
/
N
r
a
t
i
o
O
b
s
e
r
v
e
d
v
a
l
u
e
S
/
N
r
a
t
i
o
O
b
s
e
r
v
e
d
v
a
l
u
e
S
/
N
r
a
t
i
o
O
b
s
e
r
v
e
d
v
a
l
u
e
S
/
N
r
a
t
i
o

1
5
1
.
2
0
3
4
.
1
9
2
5
5
.
2
6

4
8
.
1
4
9
3
9
.
0
0

5
9
.
4
5
1
1
.
0
0

2
0
.
8
3
0
.
3
6
8
.
8
7
9
.
4
0

1
9
.
4
6
7
5
.
0
0

3
7
.
5
0

2
5
1
.
3
8
3
4
.
2
2
2
5
4
.
3
7

4
8
.
1
1
9
7
8
.
0
0

5
9
.
8
1
9
.
0
0

1
9
.
0
8
0
.
3
3
9
.
6
3
9
.
2
0

1
9
.
2
8
6
5
.
0
0

3
6
.
2
6

3
4
8
.
7
6
3
3
.
7
6
2
7
0
.
8
9

4
8
.
6
6
9
5
5
.
0
0

5
9
.
6
0
1
0
.
0
0

2
0
.
0
0
0
.
3
4
9
.
3
7
9
.
6
6

1
9
.
7
0
7
7
.
0
0

3
7
.
7
3

4
4
8
.
4
1
3
3
.
7
0
2
8
4
.
1
8

4
9
.
0
7
1
0
0
5
.
0
0

6
0
.
0
4
9
.
0
0

1
9
.
0
8
0
.
3
2
9
.
9
0
9
.
5
0

1
9
.
5
5
7
1
.
0
0

3
7
.
0
3

5
5
5
.
5
0
3
4
.
8
9
2
4
7
.
8
9

4
7
.
8
9
1
2
8
9
.
0
0

6
2
.
2
1
1
1
.
0
0

2
0
.
8
3
0
.
2
7
1
1
.
3
7
9
.
3
0

1
9
.
3
7
5
8
.
0
0

3
5
.
2
7

6
5
1
.
2
0
3
4
.
1
9
2
6
8
.
6
9

4
8
.
5
9
1
3
1
0
.
0
0

6
2
.
3
5
9
.
0
0

1
9
.
0
8
0
.
2
4
1
2
.
4
0
1
0
.
0
0

2
0
.
0
0
5
6
.
0
0

3
4
.
9
6

7
5
1
.
8
8
3
4
.
3
0
2
4
6
.
2
5

4
7
.
8
3
1
2
2
1
.
0
0

6
1
.
7
3
1
5
.
0
0

2
3
.
5
2
0
.
2
7
1
1
.
3
7
9
.
9
2

1
9
.
9
3
6
7
.
0
0

3
6
.
5
2

8
5
1
.
8
5
3
4
.
2
9
2
5
5
.
8
9

4
8
.
1
6
1
2
6
9
.
0
0

6
2
.
0
7
1
4
.
0
0

2
2
.
9
2
0
.
2
6
1
1
.
7
0
1
0
.
3
0

2
0
.
2
6
6
5
.
0
0

3
6
.
2
6

9
5
6
.
5
7
3
5
.
0
5
2
5
6
.
6
9

4
8
.
1
9
1
2
9
8
.
0
0

6
2
.
2
7
7
.
0
0

1
6
.
9
0
0
.
2
0
1
3
.
9
8
9
.
6
3

1
9
.
6
7
5
8
.
0
0

3
5
.
2
7
1
0
5
4
.
0
7
3
4
.
6
6
2
7
5
.
6
6

4
8
.
8
1
1
3
2
1
.
0
0

6
2
.
4
2
6
.
0
0

1
5
.
5
6
0
.
1
7
1
5
.
3
9
1
0
.
1
0

2
0
.
0
9
4
6
.
0
0

3
3
.
2
6
1
1
5
3
.
5
6
3
4
.
5
8
2
6
0
.
4
3

4
8
.
3
1
1
1
8
9
.
0
0

6
1
.
5
0
1
1
.
0
0

2
0
.
8
3
0
.
2
4
1
2
.
4
0
1
0
.
1
9

2
0
.
1
6
7
0
.
0
0

3
6
.
9
0
1
2
5
3
.
5
6
3
4
.
5
8
2
6
3
.
9
9

4
8
.
4
3
1
1
9
7
.
0
0

6
1
.
5
6
1
0
.
0
0

2
0
.
0
0
0
.
2
0
1
3
.
9
8
1
0
.
6
0

2
0
.
5
1
6
2
.
0
0

3
5
.
8
5
1
3
5
4
.
0
7
3
4
.
6
6
2
6
6
.
0
2

4
8
.
5
0
9
8
7
.
0
0

5
9
.
8
9
4
.
0
0

1
2
.
0
4
0
.
2
3
1
2
.
7
7
9
.
6
0

1
9
.
6
5
7
0
.
0
0

3
6
.
9
0
1
4
5
4
.
0
7
3
4
.
6
6
2
6
7
.
1
8

4
8
.
5
4
9
9
9
.
0
0

5
9
.
9
9
3
.
0
0

9
.
5
4
0
.
2
0
1
3
.
9
8
9
.
6
0

1
9
.
6
5
6
9
.
0
0

3
6
.
7
8
1
5
5
1
.
1
6
3
4
.
1
8
2
9
3
.
3
7

4
9
.
3
5
9
8
7
.
0
0

5
9
.
8
9
2
.
0
0

6
.
0
2
0
.
1
3
1
7
.
7
2
1
0
.
1
6

2
0
.
1
4
5
9
.
0
0

3
5
.
4
2
1
6
4
8
.
7
6
3
3
.
7
6
3
0
7
.
8
3

4
9
.
7
7
1
0
0
9
.
0
0

6
0
.
0
8
1
.
0
0
0
.
0
0
0
.
1
0
2
0
.
0
0
9
.
8
2

1
9
.
8
4
3
4
.
0
0

3
0
.
6
3
Table 7 Experimental layout and results with calculated S/N ratios for performance and
emission parameters
IJVD 0(0) 07 Parlak et al.indd 11 1/18/2012 11:55:09 AM
12 A. Parlak et al.
A
2
B
4
C
4
(NaOH- 2000d/d-B100) for smoke, HC and CO, A
2
B
1
C
1
(NaOH-1000 rpm-B10) for
NOx and A
1
B
1
C
3
(KOH-1000 rpm-B50) for CO
2
. While maximum performance was obtained
with KOH catalyst, the lowest emissions were found with NaOH catalyst except CO
2
.
Figure 4 S/N values of factor levels for smoke emission
Figure 5 S/N values of factor levels for NO
X
Figure 6 S/N values of factor levels for HC
IJVD 0(0) 07 Parlak et al.indd 12 1/18/2012 11:55:10 AM
Application of Taguchis methods 13
Figure 7 S/N values of factor levels for CO
Figure 8 S/N values of factor levels for CO
2
4 Conclusions
The present study has applied the Taguchi method to investigate the three well-known factors
on the performance parameters. The conditions, which maximise the brake torque and the
conditions, which minimise the brake-SFC and emissions were investigated. The conclusion
of this study can be summarised as follows:
The Taguchi design method revealed that choosing right catalyst type and blend rate
are important in view of maximisation of brake torque and minimisation of SFC
increase rate and exhaust emissions.
Although KOH is found the best catalyst for brake torque, NaOH is the better catalyst
for SFC and emissions.
The ANOVA results indicated that TSOME contributed to increase combustion
efciency up to 50% blend in the mixture. This reason caused to increase the
performance in the blends of 10%, 20% and 50% blends. This effect also caused to
decrease smoke, HC and CO emissions.
IJVD 0(0) 07 Parlak et al.indd 13 1/18/2012 11:55:10 AM
14 A. Parlak et al.
As the oxygen content in fuel blend increased, the heat content of the blend decreased
considerably and these factor lead to decrease performance and decrease SFC for
100% TSOME.
Taguchi Method is a powerful tool for experimental design so as to minimising the
experiments number, which will be conducted. The method gives to a media to nd
interactions among the factors affecting the performance parameters and optimal
conditions for the performance.
Acknowledgement
This work has been supported by The Scientic and Technological Research Council of
Turkey and was performed within The Support Programme for Scientic and Technological
Research Projects (1001) with the Project Number of 105M259.
References
Ali, Y., Hanna Milford, A. and Borg Joseph, E. (1995) Optimization of diesel. Methyl tallowate and
ethanol blend for reducing emissions from diesel engine, Bioresource Technology, Vol. 52,
pp.237243.
Alonso, J.M., Alvarruiz, F., Desantes, J.M., Hernandez, L., Hernandez, V. and Molto, G. (2007)
Combining neural networks and genetic algorithms to predict and reduce diesel engine emission,
IEEE Trans. Evol. Comput., Vol. 11, pp.4655.
Altin, R., etinkaya, S. and Yucesu, H.S. (2001) The potential of using vegetable oil fuels as fuel for
diesel engines, Energy Conversion and Management, Vol. 42, No. 5, pp.529538.
Anand, G. and Karthikeyan, B. (2005) An investigation and engine parameters optimization of a spark
ignition engine with gaseous fuels, 4th Dessau Gas Engine Conference, Germany, 23 June
Arregle, J., Ruiz, S., Desantes, J.M. and Delage, A. (1999) Characterization of the Injection
combustion process in a D.I. Diesel Engine Running with Rape Oil Methyl Ester, SAE Paper
1999-01-1497.
Dorado, M.P., Ballesteros, E., Arnal, J.M., Gomez, J. and Lopez, F.J. (2003) Exhaust emissions from
a diesel engine fueled with transesteried waste olive oil, Fuel, Vol. 82, pp.13111315.
Hasimoglu, C., Ciniviz, M., Ozsert, I., Iingur, Y., Parlak, A. and Sahir Salman, M. (2008) Performance
characteristics of a low heat rejection diesel engine operating with biodiesel, Renewable Energy,
Vol. 33, pp.17091715.
Kalligeros, S., Zannikos, F., Stournas, S., Lios, E., Anastopoulos, G., Teas, Ch et al. (2003)
An investigation of using biodiesel/marine diesel blends on the performance of a stationary
diesel engine, Biomass and Bioenergy, Vol. 24, pp.141149. AUTHOR PLEASE SUPPLY
REMAINING AUTHOR NAMES
Karaosmanolu, F., Kurt, G. and zakta, T. (2000) Long term CI engine test of sunower oil,
Renewable Energy, Vol. 19, pp.219221.
Labeckas, G. and Slavinskas, S. (2006) The effect of rapeseed oil methyl ester on direct injection
diesel engine performance and exhaust emissions, Energy Conversion and Management,
Vol. 47, pp.19541967.
Mc Donnel, K.P., Ward, S.M., Mc Nully, P.B. and Howard Hildige, R. (2000) Results of engine and
vehicle testing of semi rened rapeseed oil, Transactions of the ASAE, Vol. 43, pp.13091316.
Murayama, T., Oh, Y-T., Miyamoto, N., Chikahisa, T., Takagi, N. and Itow, K. (1984) Low Carbon
Flower Build Up, Low Smoke and Efcient Diesel Operation with Vegetable Oils by Conversion
to Monoesters and Blending with Diesel oil or Alcohols, SAE Paper 841161.
IJVD 0(0) 07 Parlak et al.indd 14 1/18/2012 11:55:11 AM
Application of Taguchis methods 15
Murugesan, Ganapathy, T., Murugesan, K. and Gakkhar, R.P. (2009) Performance optimization of
Jatropha biodiesel engine model using Taguchi approach, Applied Energy, Vol. 86, pp.24762486.
PLEASE SUPPLY INITIAL FOR THE AUTHOR MURUGESAN AUTHOR
Niemi, S.A., Murtonen, T.T., Lauren, M.J. and Vaino O.K.L. (2002) Exhaust Particulate Emissions of
a Mustard Seed Oil Driven Tractor Engine, SAE paper 2002-01-0866.
Nwafor, O.M.I. and Rice, G. (1996) Performance of rape seed oil blends in a diesel engine, Applied
Energy, Vol. 54, pp.345354.
Parlak, A., Islamoglu, Y., Yasar, H. and Egrisogut, A. (2006) Application of articial neural network to
predict specic fuel consumption and exhaust temperature for a diesel engine, Applied Thermal
Engineering, Vol. 26, Nos. 89, pp.824828.
Parlak, A., Karabas, H., Ayhan, V., Yasar, H., Soyhan, H.S. and Ozsert, I. (2009) Comparison of the
variables affecting the yield of tobacco seed oil methyl ester for KOH and NaOH catalysts,
Energy and Fuels, Vol. 23, No. 4, pp.18181824.
Pramanik, K. (2003) Properties and use of jatropha curcas oil and diesel fuel blends in compression
ignition engine, Renewable Energy, Vol. 28, pp.239248. AUTHOR PLEASE CITE THIS
REFERENCE IN TEXT.
Ramadhas, A.S., Jayaraj, S. and Muraleedharan, C. (2005) Characterization and effect of using rubber
seed oil a fuel in the compression ignition engines, Renewable Energy, Vol. 30, pp.795803.
AUTHOR PLEASE CITE THIS REFERENCE IN TEXT.
Sapaun, S.M., Masjuki, H.H. and Azlan, A. (1996) The use of palm oil as diesel fuel substitute,
Journal of Power and Energy Part A, Vol. 210, pp.4753.
Satake, K., Monaka, T., Yamada, S., Endo, H., Yamagisawa, M. and Abe, T. (2008) The rapid
development of diesel engine using an optimization of the fuel injection control, Mitsubishi
Heavy Industries Limited. Tech. Rev. Vol. 45, pp.610.
Savas, . and Kayikci, R. (2007) Application of Taguchis methods to investigate some factors
affecting microporosity formation in A360 aluminium alloy casting, Materials and Design,
Vol. 28, pp.22242228.
Scholl, K.W. and Sorenson, S.C. (1993) Combustion of Soybean Oil Methyl Ester in a Direct Injection
Diesel Engine, SAE Paper 930934.
Shroff, H.D. and Hodgetts, D. (1974) Simulation and Optimization of Thermodynamic Processes of
Diesel Engine, SAE Technical Paper, 740194.
Usta, N. (2005a) An experimental study on performance and exhaust emissions of a diesel engine
fuelled with tobacco seed oil methyl ester, Energy Conversion and Management, Vol. 46,
pp.23732386
Usta, N. (2005b) Use of tobacco seed oil methyl ester in a turbocharged indirect injection diesel
engine, Biomass and Bioenergy, Vol. 28, pp.7786.
Win, Z., Gakkhar, R.P., Jain, S.C. and Bhattacharya, M. (2005) Investigation of diesel engine operating
and injection system parameters for low noise, emissions and fuel consumption using Taguchi
methods, Proc. I Mech. E., Part D, J. Automob. Eng., Vol. 219, pp.12371251.
Nomenclature
A
i
Sum of the total ith level of the factor A
ANOVA Analysis of Variance
ANN Articial Neural Network
BTDC Before Top Dead Centre
CA Crank Angle
CO Carbon Monoxide
CO
2
Carbon Dioxide
IJVD 0(0) 07 Parlak et al.indd 15 1/18/2012 11:55:11 AM
16 A. Parlak et al.
F
factor
Ratio of the factor
GA Genetic Algorithm
HB Higher is Better
HC Hydrocarbon
K
A
Number of levels for factor A
KOH Potassium Hydroxide
LB Lower is Better
N Number of experiments
n
Ai
Number of samples for ith level of factor A
NB Nominal is Better
NaOH Sodium Hydroxide
NOx Nitrous Oxides,
RCF Relative Centrifuge Force
RPM Revolution per Minute
RPE Rapesed Methyl Ester
SFC Specic Fuel Consumption (g/kWh)
SO
2
Sulphur Dioxide, ppm
S/N Signal/Noise
SS
A
Sum of squares due to factors A
SS
factor
Sum of squares of the factor
SS
T
Sum of squares due to total variation
v
total
Degrees of freedom
v Variance of the factor
y
i
Measured value
T Sum of total (S/N) ratio of experiments
TSOME Tobacco Seed Oil Methyl Ester
IJVD 0(0) 07 Parlak et al.indd 16 1/18/2012 11:55:11 AM