Fuel 115 (2014) 295305

Contents lists available at SciVerse ScienceDirect

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

Experimental studies on the effect of injection timing in a SI engine

using dual injection of n-butanol and gasoline in the intake port
T. Venugopal, A. Ramesh
Mechanical Engineering Department, Indian Institute of Technology Madras, Chennai, India

h i g h l i g h t s
 System of varying blend ratio of alcohol and gasoline was developed.
 n-Butanol improves the torque and efciency at higher throttle position.
 Use of n-butanol blend (50% by mass) or neat gasoline is good at lower throttle than neat n-butanol.
 Completing fuel injection before the inlet valve opens reduced HC emission.
 Injection phasing or sequence inuences the HC and CO emissions little.

a r t i c l e

i n f o

Article history:
Received 3 September 2012
Received in revised form 1 July 2013
Accepted 2 July 2013
Available online 19 July 2013
Keywords:
n-Butanol
Simultaneous injection
But50S
Injection timing
Injection phasing

a b s t r a c t
Alcohols can be used in spark ignition (SI) engines along with gasoline in the blended form. However,
phase separation which occurs in the presence of moisture restricts the amount of alcohol that can be
blended. On the other hand, for effective engine operation, the ratio of alcohol to gasoline has to be varied
based on the engine operating condition. n-Butanol has properties close to gasoline and has not been
widely investigated as an engine fuel. In this work, two injectors were mounted in the intake port of
an automotive SI engine (bore = 62 mm, stroke = 66 mm, compression ratio = 9.4) to inject gasoline and
n-butanol separately so that the fuels hit the back of the intake valve. The engine was fully instrumented
for the measurement ofperformance, emissions and combustion parameters. Initially experiments were
conducted with simultaneous injection of n-butanol and gasoline (1:1 mass ratio = But 50S) using the two
injectors with different injection timings at 25% and 60% throttle positions at 3000 rpm. Subsequently different injection timings for the two fuels were tried to study the inuence of sequencing. The results were
compared with gasoline and n-butanol (But100) using a single injector. Around 26% reduction in hydrocarbon (HC) emission with simultaneous injection of gasoline and n-butanol (But50S) at 25% and 60%
throttle positions was observed with an injection timing of 64 CA before in let valve opening as compared to open valve injection. B100 was superior at high throttle positions and gasoline or B50S was suitable at 25% throttle as regards performance and emissions. At 60% throttle, injecting n-butanol just before
the start of injection of gasoline is benecial for reducing HC andcarbon monoxide (CO) emissions. In the
case of lean operation (equivalence ratio of 0.82) there is no signicant inuence of injection phasing
except for a small improvement in thermal efciency. On the whole this method of operating the engine
can lead to good engine performance over wide operating conditions since the ratio of the fuels can be
varied.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
Several alternative fuels like ethanol, butanol, methanol, biodiesel, natural gas, liqueed petroleum gas and hydrogen are being
investigated for their use in internal combustion engines. Bio-fuels
like bio-alcohol have the potential to control net global CO2 emissions and also to reduce consumption of fossil fuels. Alcohols can
Corresponding author.
E-mail address: aramesh@iitm.ac.in (A. Ramesh).
0016-2361/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.fuel.2013.07.013

be used as fuels for spark ignition (SI) engines due to their good
antiknock quality, high ame velocity and because of the presence
of oxygen in their molecule that helps combustion. Amongst alcohols, n-butanol is an emerging renewable fuel which can be produced from biological sources. A comparison between the
properties of gasoline, ethanol and 1-butanol which is also known
as n-butanol is shown in Table 1. Butanols caloric value and stoichiometric air fuel ratio are close to those of gasoline. Its latent
heat of evaporation is higher than gasoline. It is also less corrosive
than other alcohols [13]. The wide ammability limits and high

296

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

Nomenclature
BSFC
BTE
But100
But50S
CA
CO
CO2
COV
FPGA
HC
HRR
IMEP
IVO

brake specic fuel consumption (g/kW h)

brake thermal efciency (%)
n-butanol
n-butanol and gasoline by 1:1 mass ratio simultaneously
crank angle ()
carbon monoxide (% vol)
carbon dioxide (% vol)
co-efcient of variance (%)
Field Programmable Gate Array
hydrocarbons (ppmv)
heat release rate (J/ CA)
indicated mean effective pressure (bar)
inlet valve opens

ame speed of n-butanol will allow the engine to operate with lean
mixtures, high thermal efciency and low cycle by cycle variations
[36]. Only limited studies have been conducted on n-butanol in SI
engines. Hence, in this work the performance of n-butanol in different proportions along with gasoline using a dual injection system was evaluated in a SI engine.
2. Background and objective
In experiments conducted on n-butanol in SI engines, blends up
to But40 (40% n-butanol and 60% gasoline by volume) performed
similar to gasoline. However, Hydrocarbon (HC) and carbon monoxide (CO) emissions were higher for blends like But60 and
But80. Emission levels of Nitrogen oxides (NOx) were lower due
to the high latent heat of evaporation of n-butanol [6]. Lower combustion duration was also recorded with the blends as compared to
operation on neat gasoline [3,7]. No differences in HC, CO and NOx
emissions were observed between But10 and gasoline [1]. Though
lower CO levels were observed with blends of n-butanol as compared to gasoline, HC emission levels were higher at low loads
[7]. Tests with ethanol gasoline blends showed that at 20% throttle
and low speeds (3000 rpm) E05 (5% Ethanol and 95% gasoline by
volume) gave the highest torque and at high speeds E30 was the
best. At 40% and 60% throttles, E20 and E30 were observed to be
the best. At high throttle conditions no clear trend was observed
[8].
Alcohol gasoline blends can separate in the presence of water.
However, the stability of blends of n-butanol and gasoline is good
even in the presence of moisture. This is because of the low

MBD
MBT
NDIR
NI
PP
NOx
SI
TDC
UEC
/

mass burn duration ( CA)

minimum advance for best torque ( CA before TDC)
non-dispersive infrared
National Instruments
peak pressure (bar)
oxides of nitrogen (ppmv)
spark ignition
top dead centre
universal engine controller
equivalence ratio (actual fuel air ratio by stoichiometric
fuel air ratio)

solubility of n-butanol in water, unlike ethanol which is fully miscible with water. Earlier studies have reported that n-butanol can
be used as a co-solvent to avoid separation of the phases when ethanolgasoline and methanolgasoline blends are prepared [1,9
12]. M20 (Methanol 20% by volume in gasoline), E20 and But20 reduced CO emissions even at fuel rich conditions as compared to
gasoline. HC emissions were observed to be higher than with neat
gasoline when alcohol gasoline blends were used instead [9]. Butanolgasoline blends (But35) improved the torque at low and medium speeds in a carbureted engine; However, the values of torque
at high speeds were similar to that with gasoline [13]. It has also
been reported that operating the engine with 100% iso-butanol at
part throttle, reduced the brake thermal efciency by 12%. The reason for this drop in efciency is poor vaporization of iso-butanol
due to its high latent heat [14]. Emission characteristics have been
reported to be different with different isomers of butanol; however
no major change in performance has been reported with different
isomers of butanol [15]. The amount of alcohol in the blend has to
be varied based on operating conditions to ensure good performance and low emissions [6,8,11,12,14]. In addition, the high latent heat of evaporation of alcohols leads to problems during
cold starting. Cold startability can be improved by heating the fuel,
heating the air and by starting the engine using only gasoline
[1,2,11,12]. These are however not possible with pre-blended fuels.
In an attempt to vary the alcohol to gasoline ratio on line blending
methods have been tried [12].
Simultaneous injection of alcohol and gasoline into the intake
manifold through separate injectors will also avoid the problem
of phase separation and will also allow the use of varying

Table 1
Properties of gasoline, ethanol and n-butanol [1,2,4].
Property

Gasoline

Ethanol

n-butanol

Chemical formula
Composition (C, H, O) (mass%)
Lower heating value (MJ/kg)
Density (kg/m3)
Octane number ((R + M)/2)
Boiling temperature (C)
Latent heat of vaporization (25 C) (kJ/kg)
Self-ignition temperature (C)
Stoichiometric air/fuel ratio
Laminar ame speed @1 bar, 393 K, = 1.1 (cm/s)
Mixture caloric value (MJ/m3)
Ignition limits in air (vol%)
Lower limit
Upper limit
Solubility in water at 20 C (ml/100 ml H2O)

C4C12
86, 14, 0
42.7
715765
90
25215
380500
300
14.7
52
3.75

C2H5OH
52, 13, 35
26.8
790
100
78
904
420
9.0
63
3.85

C4H9OH
65, 13.5, 21.5
33.1
810
87
118
716
343
11.2
57
3.82

0.6
8
<0.1

3.5
15
Fully miscible

1.4
11.2
7.7

297

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

alcoholgasoline ratios even under transient conditions. Such a

system will also allow the blend ratio to be changed in real time.
However, details of such operation with two injectors are not reported in literature. The objective of this study is to explore the
performance, emission and combustion characteristics of a SI engine with injection of gasoline and n-butanol through two injectors
into the intake port at varying injection timings and injection phasing or sequence. The results have been compared with neat gasoline and neat n-butanol which were injected through a single
injector. It was expected that injection phasing or sequencing of
the timing of start of injection of the two fuels could lead to charge
stratication when one fuel is injected at the closed intake valve
condition and the other fuel is injected when the intake valve is
open. Hence, the effect of injection phasing or sequencing the
injections of gasoline and n-butanol was also studied in this work.

3. Experimental set up and procedure

The schematic of the experimental setup used for this work is
shown in Fig. 1. The specications of the engine and the injectors
are given in Tables 2A and 2B respectively. The engine was coupled

Table 2A
Specications of the engine.
Engine
No. of cylinders/cycle
Ignition system
Bore  Stroke
Connecting rod length
Displacement volume
Compression ratio
Rated power
Cooling medium
Lubrication system
Oil sump capacity
Lubricating oil

3 wheeler auto engine (TVS King)

One/4 Stroke
Spark ignition
62 mm  66 mm
120 mm
200 cc
9.4: 1
6.5 kW @ 5000 rpm with gasoline
Air cooled
Pressurized lubrication
1.75 l
SAE 20W40

Number of valves
Intake valve opening
Intake valve closing
Exhaust valve opening
Exhaust valve closing

2
26
44
48
22

bTDC (694 CA)

aBDC (224 CA)
bBDC (492 CA)
aTDC (22 CA)

to an eddy current dynamometer equipped with a closed loop

speed controller. Air Flow rate was measured by using a roots type
airow meter (DRESSER Inc.USA, Model 2M175). Exhaust gas

Fig. 1. Schematic of the experimental set up.

298

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

Table 2B
Specications of the injector.
Number of holes
Spray cone angle
Flow rate (DP = 3 bar, 10 ms)
Injector location

4
15 with gasoline
13.5 with n-butanol
16.9 mg/injection (gasoline)
17.7 mg/injection (n-butanol)
94 mm from the inlet valve centre

Table 2C
Uncertainty of parameters.
Dynamometer torque and speed
Efciency
Equivalence ratio
Blend ratio of n-butanol
Exhaust gas temperature
HC
CO
NO
Peak pressure

0.13 N m and 3 rpm

0.6%
0.03
1.5%
4 C
12 ppmv
0.04 %vol
50 ppmv
1 bar

Fig. 2. Important crank angles and valve timings.

emissions were measured by calibrated emission instruments

(NO-based on a Rosemount USA make chemiluminescence analyzer, HC and CO based on a NDIR analyzers of Horiba Japan make).
Fuel ow rates (both n-butanol and gasoline) were measured using
precision weighing balances on the mass basis. A ush mounted
Piezo-electric engine pressure transducer (Kistler, Switzerland,
Type 6052C) and an angle encoder (Kubler, Germany) were used
with specially developed software. A National Instruments (NI,
USA) data acquisition system was used with this software to capture cylinder pressure data on the angle basis. An average of 100
cycles of cylinder pressure was used for the calculation of heat release parameters. Heat release rates were calculated based on the
rst law of thermodynamics as applied during the period where
the intake and exhaust valves are closed (closed valve period)
[16]. Heat transfer was calculated using the Hohenbergs correlation with an assumed wall temperature of 400 K [17]. Literature
indicates that the experimentally measured values of heat uxes
match closely with the Hohenberg correlation [1820] and hence
this correlation was used in this work for the calculation of wall
heat transfer. Initially, the mean cylinder charge temperature at
the crank angle where the inlet valve closes was calculated based
on adiabatic mixing of trapped residual exhaust gases at exhaust
gas temperature and inducted fresh charge of air and fuel at ambient temperature. The mean in cylinder gas temperature at different

crank positions was calculated by the ideal gas law. The composition of the burned products was calculated based on complete
combustion for the calculation of properties [21]. An universal engine controller (UEC) that was developed in the laboratory using
National Instruments FPGA (Field Programmable Gate Array) hardware along with in-house developed software written in Labview
was used to control the injection timing and injection quantity
(pulse width of the signal given to the injectors). All experiments
were conducted at 25% and 60% throttle positions at 3000 rpm. A
special intake manifold incorporating two fuel injectors, one for
gasoline and the other for n-butanol was developed and used. Care
was taken to orient the injectors such that the sprays do not hit the
walls of the manifold but only impinge on the back of the intake
valve for good vaporization. Typical values of uncertainties of different parameters with 95% condence level [22] are given in
Table 2C.
In the rst phase of experiments, the injection timing (start of
injection) was varied from 0 to 720 CA. Here the same injection
timing was kept for both the injectors. In this case 0 CA indicates
the angle at suction top dead centre (TDC) and 720 is one complete cycle (i.e. two revolutions). This is the notation that was followed for the rst phase of experiments. The inlet valve opens at
694 CA i.e. 26 CA bTDC (before TDC) and closes at 224 CA
(Fig. 2). Hence, when the injection timing was varied, fuel injection
occurred during the closed valve, open valve and combined closed
valve-open valve periods of engine operation. The ratio of the
masses of the two fuels was maintained at 1:1. Experiments were
also conducted with neat (100%) gasoline and neat n-butanol using
a single 0.82 injector. In the above experiments the equivalence ratio was maintained at 1. After ndingthe best injection timing, the
second phase of experiments was done. This was a study of the effect of injection sequencing in a narrow range of injection timings
which were near the inlet valve opening (IVO) point i.e. certain degrees of crank angle before and after IVO as given in Table 2D. This
was again done at a xed fuel mass ratio of 1:1 (n-butanol to gasoline) at an equivalence ratio of 1. Here the start of injection of
either gasoline or n-butanol was kept xed at 630 CA (64 CA before inlet valve opening (IVO)) as this was found to be optimal from
the experiments in Phase 1. The injection timing of the other fuel
was varied from about 150 CA before IVO to about 150 CA after
IVO at 60% throttle. This range was kept narrow at 25% throttle
(Table 2D), since the pulse widths used were lesser as compared
to those at 60% throttle.Subsequently the inuence of phasing
the injection of the two fuels was also studied with lean mixtures
at an equivalence ratio of 0.82 and 60% throttle opening. In all
cases the spark timing was set at the minimum advance for best
torque (MBT) using the FPGA controller.
4. Results and discussion
The results of experiments conducted in the rst phase i.e. with
simultaneous injection (same start of injection for both n-butanol
and gasoline injectors) are presented and discussed below.
4.1. Simultaneous injection at 60% throttle position (Phase 1)
Figs. 37 indicate the inuence of injection timing on performance, emissions and combustion when the engine was operated
with gasoline, n-butanol and n-butanol + gasoline (But50S i.e. nbutanol and gasoline injected through separate injectors but
simultaneously in the mass ratio 1:1). As seen in Fig. 3a, injection
timings with But50S do not have a signicant inuence on brake
thermal efciency. Similar trends were seen in the case of torque
also. At 60% throttle, neat butanol (But100) leads to higher torque
and efciency than gasoline and But50S because of faster combustion at all injection timings. This is because of the higher ame

299

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

Table 2D
Engine operating conditions.
Experiments

Range of injection timings for gasoline and n-butanol

Throttle positions

Equivalence ratio

Effect of injection timing (Fuel ratio = 0, 50 and 100)

Effect of injection sequence (Fuel ratio = 50)

0720
150 bIVO to 150 aIVO
94 bIVO to 26 aIVO
64 bIVO to 86 aIVO

25% and 60%

60%
25%
60%

1
1
1
0.82

Fig. 3. (a) Brake thermal efciency vs. injection timing and (b) HC and NO emissions vs. injection timing (60% throttle).

speed of n-butanol. The best spark timing which is also knock limited, was more advanced with butanol because of the reduction in
charge temperature due to vaporization of the fuel. It may be noted
that n-butanol has a higher latent heat of vaporization than gasoline and also a lower caloric value. The ability to use higher spark

advances without knock in the case of n-butanol has led to better

thermal efciency and torque.
The brake specic fuel consumption (BSFC) was higher for
But100 due to its low energy density or heating value on the mass
basis as compared to gasoline. BSFC values of But50S were in

300

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

Fig. 4. Heat release rate and mean cylinder gas temperature vs. crank angle (a) 630 CA injection timing (60% throttle), (b) 0 CA injection timing (60% throttle), and (c) 630
CA injection timing (25% throttle).

Fig. 5. Spark timing and combustion parameters (60% throttle).

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

301

Fig. 6. (a) NO emissions vs. injection timing (25% throttle), (b) HC emissions vs. injection timing (25% throttle).

Fig. 7. COV of IMEP and COV of peak pressure vs. injection timing (25% throttle).

between gasoline and But100. Typical values of BSFC were 304,

337.5 and 381.5 g/kW h for gasoline, But50 and But100
respectively at 60% throttle and at the best injection timing of
630 CA (64 CA bIVO). n-Butanol has higher hydrogen to carbon
ratio (2.5:1) as compared to gasoline (2.25:1). This is the reason
for the low CO2 emission levels with n-butanol. The CO2 emission
levels at 60% throttle were 13.9%, 14% and 13.2% for gasoline,

But50S and But100 respectively at best injection timing of 630

CA (64 CA bIVO). Similar trends were also found at 25% throttle
position. The increase in CO2 emission with But50S was probably
due to better combustion with the twin injection system. It may
be notedthat the HC emission with But50S at this condition is lower than But100. Similar trends have also been reported with blends
of ethanol with gasoline [8,23].

302

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

There is a signicant inuence of injection timing on HC emissions with But50S (Fig. 3b). Injecting the fuel at 630 CA i.e. at
640CA before IVO leads to the lowest HC levels (134 ppmv). It
may be noted that at this condition the injection duration is 54
CA which means that the injection is just completed before the intake valve opens. Here, back ow of hot exhaust into the intake
manifold when the intake valve opens will aid vaporization of
the fuel [24]. The maximum HC level was observed with open valve
injection of 0 CA (180 ppmv). It may be noted that at this condition, some of the injected fuel will directly enter the cylinder without hitting the valve. Faster combustion was observed at this
condition leading to higher cylinder pressures which could result
in increased ow of fuel into the crevices in the combustion chamber and hence elevated HC levels [24,25]. Since stoichiometric
operation was maintained there was no signicant observation
with respect to CO and NO levels. HC, CO and NO emissions are
not greatly inuenced by injection timing in the case of neat gaso-

line and neat n-butanol operation. However, n-butanol exhibits

higher CO levels due to the inuence of lower charge temperatures
which can affect fuel vaporization. It was also noted that the exhaust gas temperatures for n-butanol were 5565 C lower than
But50S and gasoline which could reduce post oxidation of HC
and CO in the exhaust. The variation of NO is seen in Fig. 3b. NO
emissions of But50S and gasoline were higher than But100 due
to higher charge temperatures. NO emission with But50S where
two injectors were used was similar to gasoline due to improved
vaporization. This is probably because in the case of But50S the
spray covered a greater portion of the inlet valve as compared to
gasoline that was injected with one injector. Such results have
been observed by the authors earlier under different operating
conditions [26].
Heat release rates occurred earlier with But100 as compared to
gasoline and But50S due to more advanced spark timings (Fig. 4a)
at the injection timing of 630 CA (closed inlet valve injection).

Fig. 8. Brake thermal efciency and NO emission vs. injection timing (a) 60% throttle, (b) 25% throttle and HC and CO emissions vs. injection timing, (c) 60% throttle, and (d)
25% throttle.

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

However, the peak heat release rates were comparable. Heat release was more advanced with But100 and But50S than gasoline
due to earlier spark timings. We see that the peak mean gas temperature for But100 is greater than But50S at the injection timing
of 630 CA (Fig. 4a). The lowest peak temperature was observed
with gasoline operation. However, the trend of NO was different
i.e. it was higher for gasoline as compared to But50S and But100.
This may be due to the differences in the physical and chemical
nature of the fuels used. The post combustion temperature with
But100 is low because of more advanced combustion due to the
higher spark advance that was needed. The low post combustion
temperature affects the oxidation of HC and CO emissions in the
case of n-butanol. As seen in Fig. 4b, with an injection timing of
0 CA (open inlet valve injection) the peak heat release rates of
But100 and But50S are lower than gasoline. Fig. 5 indicates that
the crank angle at which 50% burn duration occurs is earlier with
But100 and But50S. This is because of the more advanced spark
timings that were used with But100 and But50S. There is no significant difference between the combustion durations of But50S and

303

gasoline. However, But100 exhibits lower combustion durations

on account of faster combustion.
4.2. Simultaneous injection at 25% throttle position (Phase 1)
The efciency of the engine is slightly higher with gasoline and
But50S than But100 as shown in Fig. 3a earlier. The inuence of
injection timing on efciency and torque is not signicant except
in the early open valve injection timing of 0 CA where the efciency with But50S is higher than with gasoline and But100. It
may be noted the in the case of But50S both the injectors (dual
injection) were used and hence the injection pulse duration was
short (1.7 ms). However, in the case of gasoline and But100 the
injection was done with a single injector and was for much longer
durations (4.2 ms for gasoline and 6 ms for butanol). Thus in the
case of But100 and gasoline the injection of the fuel extends well
into the intake stroke whereas with But50S the injection pulse
stops at about 30 after it starts. A long injection duration well into
the intake stroke will mean that most of the injected fuel is carried

Fig. 9. (a) Brake thermal efciency and HC emission vs. injection timing, (b) heat release rate vs. crank angle, and (c) NO emission vs. injection timing.

304

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

away by the inducted air and hence is not effectively vaporized by

the hot intake valve and by the back ow of exhaust gases. Thus
vaporization is better with But50S and this results in better heat
release rate and efciency. Low NO emission was observed with
But100 as compared to gasoline and But50S at 25% throttle, similar
to the trends at 60% throttle (Fig. 6a). However, the values of NO
with gasoline and But50S were a little higher at 25% throttle as
compared to 60% throttle due to more advanced spark timings that
had to be used for best torque.
Injecting the fuel at 64 CA before IVO (630 CA) was better in
the case of gasoline and But50S as regards HC emissions
(Fig. 6b). Injecting the fuel at 64 CA before IVO resulted in about
25% reduction in HC emission as compared to open valve injection
in the case of gasoline and But50S. However, But100 resulted in
very high HC levels as compared to gasoline and But50S. With open
valve injection, it is likely that signicant amounts of n-butanol
reach the cylinder before vaporization. The exhaust temperature
was around 5070 C lower with But100 than gasoline and But50S.
The lower exhaust gas temperature adversely affects post oxidation of HC in the case of But100 as indicated earlier.
The heat release rates were not signicantly inuenced by
injection timing with But50S and gasoline; Lower heat release
rates were observed with open valve injection than closed valve
injection with n-butanol. This is because in the case of injection
when the valve is open, the owing air stream carries away the fuel
into the cylinder. However, in the case of closed valve injection
(injection when the intake valve is closed) the fuel hits the back
of the intake valve and resides there and the hot intake valve effectively vaporizes the fuel. This improved vaporization of fuel, results
in higher heat release rates with closed valve injection. Even
though higher heat release rates were observed with n-butanol
as compared to gasoline and But50S, longer combustion duration
and lower post combustion temperature due to charge cooling
(Fig. 4c) have led to higher HC emission levels (Fig. 6 b) and lower
efciency. Hence, from the point of view of HC emissions it is not
advisable to use But100 at low throttle conditions. Open valve
injection increased CO emissions with all the three fuels. The coefcient of variation (COV) of indicated mean effective pressure
(IMEP) and the COV of peak pressure (PP) are indicated in Fig. 7.
It is seen that there is no signicant variation in these parameters
with changes in the n-butanol ratio and injection timing because
the operation is at stoichiometric conditions with the best spark
timing.
4.3. Phased injection at 60% and 25% throttle positions (Phase 2)
In these experiments the injection timing of gasoline was xed
at 64 CA before IVO and the injection timing of n-butanol was
swept from 150 CA before to 150 CA after the IVO. The same
was repeated with the injection timing of n-butanol being xed
and that of gasoline being changed. While this range of injection
timings was used at 60% throttle position a smaller range 94 CA
before to 26 CACA after IVO) was used at 25% throttle opening.
No major change in brake thermal efciency was observed with
injection sequence at both the throttle positions. In-general, injecting one fuel at open valve condition and the other fuel at closed
valve condition reduced NO emission (Fig. 8a and Fig. 8b). It was
observed that injecting n-butanol before the start of injection of
gasoline and vice versa produced about 10% lower HC emissions
(Fig. 8c) at 60% throttle compared to simultaneous injection of both
the fuels (at 64 CA before IVO). No signicant change in HC and CO
emissions was observed with injection phasing at 25% throttle
(Fig. 8d). Fixing one of the injections at 64 CA before IVO and moving the other towards the valve open timing increased the CO
emission due to low post combustion temperatures as discussed
earlier (Fig. 8c and 8d). It can be concluded that no major benets

are there with injection phasing except a 10% reduction in HC

emissions at 60% throttle. All the above results are at an equivalence ratio of 1.
Since many small engines operate with lean mixtures for good
fuel economy, experiments were also conducted with leaner mixtures. A difference in efciency and torque was observed at an
equivalence ratio of 0.82 with injection phasing. Injecting one fuel
at the IVO point and another one at 64 CA before IVO produced
higher efciency and lower HC emissions as seen in Fig. 9a. Heat
release rates were observed to be higher with phased injection in
the regions where the efciency is higher as compared to simultaneous injection (Fig. 9b). Injection phasing could lead to charge
stratication which in turn can favorably inuence combustion.
Open valve injection can result in charge stratication as reported
in literature [27]. The NO emission level was increased by around
15% at the best efciency point as compared to simultaneous injection (Fig. 9c). CO emissions were less than 0.1% vol at all operating
points due to operation with lean mixtures (/ = 0.82).

5. Conclusions
Based on the experimental results on injecting n-butanol and
gasoline in a SI engine using two injectors the following conclusions are made.
 n-Butanol improves the torque and efciency at 60% throttle
position by around 1.5% at the best injection timing of 64 CA
before IVO. However, at the lower throttle position of 25% it is
inferior to gasoline and B50S operation. A 1.52% drop in efciency was observed with n-butanol as compared with gasoline
at 25% throttle position. It is better to use n-butanol at high
throttle positions and gasoline or B50S at 25% throttle to
improve performance and emission.
 Completing fuel injection before the inlet valve opens (i.e. a
start of injection of 64 CA before IVO) is best for reducing HC
emission. Around 26% reduction in HC emission with simultaneous injection of gasoline and n-butanol (B50S) at 25% and
60% throttle positions was observed. The effect of injection timing on HC emission with 100% n-butanol was found to be less at
both throttle positions.
 Injection phasing or sequence mainly inuences the HC and CO
emissions. Injecting n-butanol just before the start of injection
of gasoline is benecial for reducing HC and CO emissions than
simultaneous injection (at 60%throttle). However, the difference
is only about 10%.
 With lean operation (equivalence ratio of 0.82) there is no signicant inuence of injection phasing except for a small
improvement in thermal efciency.
On the whole with dual injection n-butanol has to be used at
higher throttle positions and gasoline or B50S is suitable at lower
throttle positions (25%) for good performance and low emissions.
Injection timing mainly inuences HC emissions and the best
injection timing was 64 CA before IVO. Injection phasing has a
small inuence on emissions. Injecting n-butanol just before the
start of injection of gasoline is desirable.

References
[1] Wallner T, Miers S, McConnell S. A comparison of ethanol and butanol as
oxygenates using a direct injection spark-ignition engine. J Eng Gas Turb Power
2009;131:19.
[2] Owen K, Coley T. Automotive Fuels Handbook. USA: Society of Automotive
Engineers; 1990.
[3] Szwaja S, Naber JD. Combustion of n-butanol in a spark-ignition IC engine. Fuel
2010;89:157382.

T. Venugopal, A. Ramesh / Fuel 115 (2014) 295305

[4] Broustail G, Seers P, Halter F, Morac G, Mounaim-Rousselle C. Experimental
determination of laminar burning velocity for butanol and ethanol iso-octane
blends. Fuel 2011;90:16.
[5] Gu X, Huang Z, Wu S, Li Q. Laminar burning velocities and ame instabilities of
butanol isomers-air mixtures. Combust Flame 2010;157:231825.
[6] Dernotte J, Mounaim-Rousselle C, Halter F, Seers F. Evaluation of butanol
gasoline blends in a port fuel-injection spark-ignition engine. Oil Gas Sci
Technol 2010;65:34551.
[7] Yasar A. Effects of alcoholgasoline blends on exhaust and noise emissions in
small scaled generators. Metalurgija 2010;49:3358.
[8] Wu CW, Chen RH, Pu JY, Lin TH. The inuence of air-fuel ratio on engine
performance and pollutant emission of an SI engine using ethanolgasolineblended fuels. Atmos Environ 2004;38:7093100.
[9] Rice RW, Sanyal AK, Elrod AC, Clemson, Bata RM. Exhaust gas emissions of
butanol, ethanol and methanolgasoline blends. J Eng Gas Turb Power
2010;113:37781.
[10] Mukov Z, Posp M, Gustav S. Volatility and phase stability of petrol blends with
ethanol. Fuel 2009;88:13516.
[11] Abdulghani A Al-Farayedhi, Al-Dawood AM, Gandhidasan P. Effects of blending
crude ethanol with unleaded gasoline on exhaust emissions of SI engine. SAE
paper 2000-01-2857.
[12] Kumar A, Khatri DS, Babu MKG. An investigation of potential and challenges
with higher ethanolgasoline blend on a single cylinder spark ignition
research engine. SAE paper 2009-01-0137.
[13] Yang J, Wang Y, Feng R. the performance analysis of an engine fueled with
butanolgasoline blend. SAE paper 2011-01-1191.
[14] Irimescu A. Performance and fuel conversion efciency of a spark ignition
engine fueled with iso-butanol. Appl Energy 2012;96:47783.
[15] Regalbuto C, Pennisi M, Wigg B, Kyritsis D. Experimental investigation of
butanol isomer combustion in spark ignition engines. SAE paper 2012-011271.

305

[16] Hayes TK, Savage LD, Sorenson SC. Cylinder pressure data acquisition and heat
release analysis on a personal computer. SAE paper 860029.
[17] Hohenberg GF. Advanced approaches for heat transfer calculations. SAE paper
790825.
[18] Finol CA, Robinson K. Thermal modelling of modern engines: a review of
empirical correlations to estimate the in-cylinder heat transfer coefcient.
Proc IMechE J Automob Eng 2006;220:176587.
[19] Heinle M, Bargende M, Berner HJ. Some useful additions to calculate the wall
heat losses in real cycle simulations. SAE paper 2012-01-673.
[20] Lejsek D, Kulzer A, Hohenberg G, Bargende M. Novel transient wall heat
transfer approach for the start-up of SI engines with gasoline direct injection.
SAE paper 2010-01-1270.
[21] Ganesan
V.
Computer
simulation
of
spark-ignition
engine
processes. India: Universities Press (India) Ltd.; 1996.
[22] Holman JP. Experimental methods for engineers. 7th ed. India: The McGrawHill Companies; 2007. p. 4862.
[23] De Melo T, Machado G, De Oliveira E. Different hydrous ethanolgasoline
blends-FTIR emissions of a ex-fuel engine and chemical properties of the
fuels. SAE paper 2011-36-0080.
[24] Lang KR, Cheng WK. Effects of fuel injection strategy on HC emissions in a
port-fuel-injection engine during fast idle. SAE paper 2006-01-3400.
[25] Yang J, Kaiser EW, Siegl WO, Anderson RW. Effects of port-injection timing and
fuel droplet size on total and speciated exhaust hydrocarbon emissions. SAE
paper 930711.
[26] Venugopal T, Akash AJ, Ramesh A. Performance and emission characteristics
of simultaneous injection of n-butanol and gasoline in a four stroke SI
engine. In: Proc 22nd Nation Conf IC Engines Combust, NIT-Calicut India;
2011. p. 1427.
[27] Kiyota Y, Akishino K, Ando H. Concept of lean combustion by barrelstratication. SAE paper 920678.