International Journal of Mechanical and Production

Engineering Research and Development (IJMPERD)

ISSN(P): 2249-6890; ISSN(E): 2249-8001
Vol. 8, Issue 6, Dec 2018, 809-818
© TJPRC Pvt. Ltd.

ANALYSIS OF COOLANT TEMPERATURE AND SWIRL MOTION EFFECTS ON

CYCLE TO CYCLE COMBUSTION VARIATIONS IN SI ENGINE USING

REGRESSION AND CORRELATION TECHNIQUES

V. KRISHNAMOORTHY
Professor, Department of Mechanical Engineering, Saveetha Engineering College, Chennai, Tamil Nadu, India
ABSTRACT

The improvement of exhaust emission and engine performance in terms of fuel economy depends on the
understanding of combustion process in IC Engines on cycle to cycle basis. It also helps to improve engine stability. Two
port injected SI engines were analyzed for cycle to cycle variation in combustion to know the effect of initial combustion
processes on the following combustion. For different equivalent ratios the correlation between IMEP and pressures
almost showed same trends but with different mixture preparations indicated different tendency. With two engines for the
given referenced angles the dependency of IMEP on pressure increases when mixture becomes leaner. By deactivating
one intake valve and by varying the coolant temperature, mixture distribution in the combustion chamber was varied due

Original Article
to air motions and evaporation of fuel. With the Coolant temperature and air motion, correlation between pressures
related parameters were done.

KEY WORDS: IMEP, Correlation, SI Engine, Cold Start, Swirl, Cycle to Cycle Variations & Coolant Temperature

Received: Oct 03, 2018; Accepted: Oct 23, 2018; Published: Nov 17, 2018; Paper Id.: IJMPERDDEC201882

INTRODUCTION

There are various factors such as type of fuel, Air fuel mixture motion, residual gases and mixing of air
and fuel, causes cycle to cycle variation in combustion process though the IC engine may be under steady state
condition of operation with respect to load and speed (Ozdor, N, 1994). Other factors such as location of spark
plug, spark strength and heat dissipation in the electrode also contribute greatly to the cycle to cycle variation of
combustion parameters in Engines (Bates, 1989). These cyclic variations are excessively demonstrated by the
engine when cold starting, idling and lean mixture operations.

Combustion parameters such as flame propagation and pressures can be analyzed to know the cycle to
cycle variations in IC Engines. The variations in Combustion process can greatly be studied by visualizing the
flame propagation with the help of high speed vision equipment and laser diagnostics. (Keck and Heyward, 1987)
But with the help of measurement of Combustion pressure, the variation in combustion process can be easily
analyzed. Combustion pressure measurement also helps in detecting misfire, knocking and also to achieve efficient
operation of IC engines by spark timing optimization. (Spicher and Backer, 1990). Cycle to cycle variations in
Combustion process can be confirmed by measuring combustion pressure thereby evaluating the combustion state
(Young. M 1981). It is imperative to understand how the initial combustion state affects the later combustion
process and there by engine performance, so as to improve the engine stability. In this Research the effects of
previous cycle’s combustion on successive cycle’s combustion were analyzed for different air-fuel mixture and the

www.tjprc.org SCOPUS Indexed Journal [email protected]

810 V. Krishnamoorthy

coolant temperature with the pressure data. Two different spark engines with the port injected fuel system were compared
to have consistency in study.

EXPERIMENTAL PROCEDURES

Figure 1 shows the experimental setup in schematic sketch. Piezoelectric pressure transducer Kistler 6053CC was
mounted in the cylinder head to measure the combustion pressure and analyze the combustion process. The pressure data
were sampled at a rate of 1o CA (Crank angle) intervals for 100 cycles in the data acquisition system and processed to
obtain the pressure at referenced Crank angles and IMEP (Indicated Mean Effective Pressure). To measure air-fuel ratio,
laminar flow meter and exhaust gas analyzer were used. The specifications of two different four stroke SI (Spark Ignition)
engines are given in Table 1. The tests were conducted with fueling and spark timing controller, Dynamometer controller,
encoder, Throttle actuator, ECU (Electronic control unit) and computer. For consistency two different SI Engines were
used for experiments. To study the effect of fuel vaporization and mixing at low temperature conditions, the engine coolant
temperature of engine A were changed. In engine B, one intake valve was deactivated to induce higher swirl in the
combustion chamber and change the mixture distribution in the combustion chamber.

Figure 1: Schematic Diagram of Experimental Setup

Table 1
Engine A Engine B
Bore (mm) 75.5 82.0
Stroke (mm) 83.5 85.5
Swept Volume (cc) 374 448
Compression Ratio 9.5:1 10:1
Valve Timing 5/35 43/5 9/43 50/6

Impact Factor (JCC): 7.6197 SCOPUS Indexed Journal NAAS Rating: 3.11
Analysis of Coolant Temperature and Swirl Motion Effects
E on 811
Cycle to Cycle Combustion Variations in SI Engine Using
Regression and Correlation Techniques
The two engines were operated at some specific part load conditions, that is 2000 rpm / 2.0 bar BMEP
(Brake Mean Effective Pressure) for
or engine A and 1500 rpm / 2.0 bar BMEP for engine B.
B To study the coolant
temperature effect on combustion variation, two different engine operating conditions were employed.
e One with idling
condition during warming up period in which the engine speed
ed and mixture ratio were varied as the operating time elapsed.
Another with the fixed engine speed,, intake pressure and fuelling was applied. The data was analyzed to understand the
effect of fuel evaporation on Cycle to cycle variations in combustion process. The conventional statistical method of
regression for correlating the two variables and covariance are applied in this research work. (Ref 6:
6 Dunn & Clark, 1997).

RESULT & DISCUSSIONS

Variation in Pressure at Referenced Crank
Cr Angle

With referenced Crank angle the pressure variations were compared to validate the effect of initial combustion
pressure as parameters for cycle to cycle variation in combustion.

The variation in cylinder pressure for entire cycle is represented by the standard deviation divided with mean
pressure at referenced crank angles. In Figure 1,, variations in cylinder pressure in terms of percentage at referenced crank
angles were shown for different equivalence
equivalen ratios. The variation in pressure at referenced crank angles when the engine is
not firing i.e at motoring condition is only 0.2 – 0.3 % but during firing conditions the variation in pressure is substantial.

CAppeak in the Figure 2 means crank angle at which the peak pressure occurs. At 150BTDC (Before Top Dead
Centre) the variation in pressure is around 1.4% for equivalence ratio of Z = 1.0, which is slightly
sligh higher than the
variations in pressures of motoring conditions. The probable reasons attributed for this small variation is the lack of time
for combustion after ignition initiation.

Figure 2: Variations of Pressures at Referenced Crank Angle for

Different Equivalence Rates

It is very evident from the Figure

igure 2 that as the combustion progresses, the variation in pressure also increases
rapidly and also the leaner mixture operation is prone
pro for higher variation in combustion pressure than the stoichiometric
st
condition, implying unstable combustion with leaner mixture.

www.tjprc.org SCOPUS Indexed Journal [email protected]

812 V. Krishnamoorthy

Correlation between Pressures at Referenced Crank Angles

The dependence of later stages of combustion on the early portion of combustion was investigated to assess the
validity of initial combustion
tion pressure as parameter for cycle to cycle variation in combustion. Figure 3 shows the effect of
pressures at 15o BTDC on latter pressures for different equivalence ratios. It is seen from the figure 3 that the pressures at
10o BTDC show a higher dependence
ence on pressures at 15o BTDC. On the contrary later pressures
pre at TDC (Top Dead
Centre), 10o ATDC (After Top Dead Centre) and 20o ATDC
C don’t show any correlation with the initial pressure. Hence it
can be concluded that combustion after TDC is not affected by the very early combustion state. However, the correlation
between the pressure at 10oBTDC and later pressures indicate relatively increased tendency as indicated in Figure 4. As the
mixture becomes leaner, the correlation
orrelation coefficient improves but when the combustion proceed further for all operating
conditions these correlation coefficient decrease linearly and becomes less significant after TDC. From these trends of
correlation coefficient it is seen that the combustion is accelerated as the combustion proceeds,
proceeds and cycle to cycle variation
in combustion diminishes to a lesser value.
val It is presumed that the inhomogeneous mixture distribution in combustion
chamber due to port injection is not affecting much the subsequent combustion.

Figure 3:
3 Correlation for Combustion Pressures with
Pressure at 15o BTDC

Figure 4:
4 Correlation for Combustion Pressures with
Pressure at 10o BTDC

Correlation between Pressure and IMEP

IME

The Correlation between IMEP and pressures at referenced crank angles are analyzed as presented
presen in Figure 5 in
which R means the linear correlation coefficient, there by confirming the effect of combustion pressure on IMEP.

Impact Factor (JCC): 7.6197 SCOPUS Indexed Journal NAAS Rating: 3.11
Analysis of Coolant Temperature and Swirl Motion Effects
E on 813
Cycle to Cycle Combustion Variations in SI Engine Using
Regression and Correlation Techniques

Figure 5:
5 Correlation between IMEP and Pressures for
Equivalence Ratio φ = 1

The linear correlation between pressure and IMEP

IMEP for different mixture ratios at referenced Crank angles are
shown in Figure
igure 6. It is evident from the Figure 6 that the correlation coefficients are increased as the combustion proceeds
irrespective of mixture ratios. Hence IMEP is
is more closely correlated with the later pressures than the initial pressures.
As the pressure at the referenced crank
rank angle increases, the IMEP also increases. In leaner mixture operations, the
correlation coefficient increases,
s, implying IMEP is more dependent
depe nt on the later stage of combustion.

Figure 6:
6 Correlation between IMEP and Pressures for
Various Equivalence Ratios

Effect of Coolant Temperature on Correlation Coefficient

During the warming up period the level of variation in pressures at specified crank angle is changed with the
coolant temperature and it is also seen that the variations in pressure at specified crank angles increases as the combustion
progresses. Figure 7 shows the variations of pressure corresponding to the Crank angles
es and coolant temperatures. It is
clear from the figure 7 that variation in combustion is higher with the low coolant temperatures. It is difficult for the fuel to
evaporate at low coolant temperature and hence more liquid fuel is induced into the combustion
combust chamber causing mixture
in-homogeneity. The cycle to cycle variation in mixture distribution also seems to be increased. Hence variation in
combustion is due to variation in pressures caused by mixture ratio variation, in-homogeneity
homogeneity and cycle to cycle variation
in mixture distribution.

www.tjprc.org SCOPUS Indexed Journal [email protected]

814 V. Krishnamoorthy

Figure 7: Variations of Pressures at Idling with

Different Coolant Temperatures

The effect of initial combustion on later stage of combustion during warming up period is examined in Figure 8.
The trend of correlation shows almost same as normal operating condition for increased coolant temperature. However
with low temperature of 250C, the effect of initial combustion is relatively poor. This implies that the mixture distribution
at cold start is more inhomogeneous and the initial combustion is not sufficient due to difficulty in fuel evaporation.
At idling condition another peculiar phenomenon is that the crank angle at which maximum pressure occurs is very close to
the TDC.

Figure 8: Correlation between the Pressure at 5o ATDC and

Coolant Temperatures During Idling Condition

Due to the retarded spark timing, the crank angle at which the peak pressure occurred should be retarded, but this
angle appeared very close to TDC. The reason for this is that the slow combustion is dominant in the peak pressure
occurring at idling.

The effect of coolant temperature on the IMEP dependency on the pressure at specified crank angles as shown in
Figure 9. This trend is nearly same as the fully warmed up condition; however the case of 25oC indicates relatively less
correlation than the higher coolant temperature cases.

Impact Factor (JCC): 7.6197 SCOPUS Indexed Journal NAAS Rating: 3.11
Analysis of Coolant Temperature and Swirl Motion Effects on 815
Cycle to Cycle Combustion Variations in SI Engine Using
Regression and Correlation Techniques

Figure 9: IMEP Dependency on Pressures at Crank Angles with

Coolant Temperature During Idling Condition

By keeping the fixed engine speed, intake pressure and fuelling at the part load condition, the effect of fuel
vaporization on cycle to cycle variation was studied. In this case, with the lower coolant temperature the correlations
between initial pressure and later pressures are very poor and only the case of 85oC show relatively higher correlation as
the normal operating condition as shown in Figure 10. In Figure 11, the correlation does not exist appreciably except for
the case at 85oC as the combustion proceeds.

Figure 10: Correlation Between the Pressures at 15o BTDC and

Later Pressures for Part Load Condition for
Different Coolant Temperature

Figure 11: IMEP Dependency on Pressures at Crank Angles for

Various Coolant Temperatures During Part Load

www.tjprc.org SCOPUS Indexed Journal [email protected]

816 V. Krishnamoorthy

Effect of Air Motion

The engine B was

as operated with one deactivated intake valve to know the effect of intake air motion on cycle to
cycle variation in combustion. As one valve
valv is deactivated, the swirl motion in combustion chamber is increased there by
affecting the mixing of air and fuel and their distributions. From Figure 12, it is evident that as the mixture goes leaner, the
pressure variation increases. When both the valves are active, the swirl level is low and for equivalence ratio of 0.91 the
variation in combustion pressure is relatively higher than the rich mixture as shown in Figure
igure 12 (a). When one intake
valve deactivated, the swirl motion is high but trend of variation in pressure is same as in the case of low swirl motion as
shown in Figure
gure 12 (b). Comparing the Figures
igures 12 (a) & (b), it is evident that the pressure variation is higher for high swirl
motion before TDC than for lower swirl motion. Whereas after TDC the variation in pressures is lower due to the faster
combustion with increased
ased swirl motion.

Figure 12: Variation in Pressure for Engine B with Swirl Variation for Crank Angles

The Coefficient of variance (COV) in IMEP indicates that the cycle to cycle combustion variation is more stable
for higher swirl motion as shown in Figure
Figure 13 (a) & (b). When the mixture is lean, the effect of swirl on combustion
variation is predominant as COV in IMEP for Z =91 with one intake valve deactivation is 3.2%
3.2 whereas it is 6.1% for
other engine. The IMEP with rich mixture decreases in the early combustion as the initial pressure increases. This is
because the initial combustion is faster due to the
t increased turbulence.

Figure 13: Correlation between IMEP and Pressures for Engine B for Crank Angles

Impact Factor (JCC): 7.6197 SCOPUS Indexed Journal NAAS Rating: 3.11
Analysis of Coolant Temperature and Swirl Motion Effects on 817
Cycle to Cycle Combustion Variations in SI Engine Using
Regression and Correlation Techniques
CONCLUSIONS

The effect of engine coolant temperature and air motion on cycle to cycle combustion variation is very evident.
As the combustion proceeds, the variation in pressure increases at the referenced crank angle. This variation in pressures
increases with leaner mixture and with lower coolant temperatures. The pressures at specified referenced crank angle
generally showed good correlation to each other. In port injection, the subsequent combustion is less affected by the early
combustion due to less uniform mixture distribution in combustion chamber.

The correlation between IMEP and pressures at crank angles increases when the mixture is lean and on the
contrary this correlation is poor at low coolant temperature but increases as the engine warms up. However, the increased
swirl motion due to one valve deactivation makes combustion more stable and dependency of IMEP on pressures show
different trend in early combustion with rich mixture.

REFERENCES

1. Young. M (1981), Cyclic Dispersions in the Homogeneous-Charge Spark Ignition Engine – A Literature Survey, SAE Paper
810020.

2. Bates S (1989), Flame imaging studies of cycle by cycle combustion-variation in SI four stroke cycle engine, and SAE paper
892086.

3. Keck J and Heywood J (1987), Early flame development and burning rates in Spark Ignition Engines and their Cyclic
variability, SAE paper 870164

4. Spicher U and Backer H (1990) Correlation of Flame propagation and In-Cylinder Pressure in a Spark Ignition Engine SAE
Paper 902126.

5. Dunn, O.J. Clark, V.A (1997), Applied Statistics, Analysis of Variance na Regression, John Wiley & Sons

6. Ozdor, N., Duglar, M. and Sher, E (1994) Cyclic Variability in spark Ignition Engines – A Literature Survey, SAE Paper
940987

www.tjprc.org SCOPUS Indexed Journal [email protected]