Document 10549811

advertisement
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Secondary atomization of drop impactions onto heated surfaces
António L. Moreira1, Ana S. Moita2, Elvio Cossali3, Marco Marengo4, Maurizio
Santini5
1: Dept. of Mechanical Engineering, Instituto Superior Técnico, Lisbon, Portugal, moreira@dem.ist.utl.pt.
2: Dept. of Mechanical Engineering, Instituto Superior Técnico, Lisbon, Portugal, anamoita@dem.ist.utl.pt.
3: Dept. Industrial Engineering, Università degli Studi di Bergamo, Dalmine (BG), Italy, gianpietro.cossali@unibg.it.
4: Dept. of Industrial Engineering, Università degli Studi di Bergamo, Dalmine (BG), Italy, marco.marengo@unibg.it.
5: Dept. of Industrial Engineering, Università degli Studi di Bergamo, Dalmine (BG), Italy, maurizio.santini@unibg.it.
Abstract The work presented here addresses an experimental study aimed at characterizing the effect
of the impact angle on secondary atomization, at different boiling regimes. The study considers impacts of
water and isooctane droplets, onto a heated stainless steel surface, with known roughness. Sizing of the
larger droplets, within the range of 40µm up to a few millimeters is carried out by post processing of the
images recorded with two CCD cameras (Image Analysis Technique – IAT), while the diameter of small
droplets, within the range of 5.5µm up to 250µm, is evaluated based on phase Doppler anemometry
measurements. The PDA measurements are integrated in time up to the impact instant t0, before comparison
with the IAT probability distribution. Finally, a scaling of the IAT p.d.f. and of the integrated PDA p.d.f. is
performed by equating the count values where the two size ranges overlap, obtaining an extended p.d.f.,
which is used to evaluate secondary droplet size diameters, over a wide range of droplet diameters, from
5.5µm up to a few millimeters. The results show that the measured secondary droplet diameter is faintly
influenced by large impact angles (α>45º) but becomes considerably smaller for small inclination angles
(α<15º), as the larger droplets are swept away due to gravitational forces. The effect of the gravitational
forces in decreasing the measured mean droplet diameter is less significant for isooctane drop impacts, as the
mean secondary diameters are globally smaller, when compared to those obtained from water drop impacts.
Differences in secondary droplet size diameters between isooctane and water droplets are attributed to the
much smaller surface tension of isooctane.
1. Introduction
The impact of sprays on hot walls is a topic of interest in several technological applications,
ranging from those geared to cool high power electronic devices or to protect the epidermis during
cutaneous laser therapies, to fuel injection systems in internal combustion engines. In all cases,
technological developments rely on fundamental researches aimed at improving knowledge of the
interaction processes, such as those reported by Mundo et al.(1997, 1998), Lee and Ryou (2000),
Rusche (1997), and Tropea and Roisman (2000). Applications like fuel injection are probably those
demanding the most from the ability to accurately describe the physics of interaction, as discussed
in Tropea and Roisman (2005) and a large number of physical models have been developed for
gasoline and Diesel sprays, as in Thedorakakos et al. (1996), Lindgren and Denbratt (2000) and Bai
et al. (2002).
A renewable interest is now arising in this field due to the development of new lean internal
combustion engines, the Homogeneous Charge Compression Ignition. This is an attractive
advanced combustion process that results from spontaneous-ignition at multiple points throughout
the volume of the charge gas, rather than relying on maintaining a flame front as in conventional
engines (either spark ignited or Diesel). Although there is ample evidence that HCCI combustion in
four-stroke engines is dominated by chemical kinetics, experiments indicate that mixture
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
inhomogeneity plays an important role in HCCI combustion and is a key enabler for controlling
HCCI combustion, e. g. Girard et al (2002). However, and despite the fuel injection strategy chosen
(port injection or direct injection inside the cylinder) the quality of atomization and, therefore,
mixture preparation and combustion depends on the thermal and fluid-dynamic interactions
between fuel droplets and the valve surface.
In this context, two basic issues need to be addressed. The first is to be able to predict which
regime occurs under given conditions; the second is to be able to predict quantitatively the
characteristics of secondary breakup occurring at the impact, namely the magnitude and direction of
secondary droplets, as well as their size distribution and fraction of the total mass of fuel supplied.
The structure of existing models tends to be similar in that all address both issues, e. g. Li Fan and
Reitz (2000), Bai et al (2002) and Lindgren and Denbratt (2004). However, it is known that heat
transfer alters the fluid dynamic behavior of impact and, therefore, universal functional relations
cannot be devised. The result is that the same models cannot be simultaneously used for cold start
and the warm up periods as observed by Fan and Reitz (2000), Tomoda et al., (2001) and Kuhnke
(2004). However, only a few comprehensive studies on drop/wall interactions, particularly on
disintegration mechanisms, have been reported, e.g. Karl and Frohn, (2000) and Cossali et al.,
(2005), which consider the effects of surface temperature. Furthermore, the fuel spray does not
impact perpendicularly to the surface and the fine tuning of injector position and inclination plays a
fundamental role on engine performance (e.g. Wang et al., 2004). The characterization of the
disintegration mechanisms of droplets impacting onto heated inclined surfaces is sparsely reported
in the literature (e.g. Karl and Frohn, 1996, Yao et al., (1988) and the studies are usually restricted
to the film boiling regime, i.e., when the surface is heated well above the Leidenfrost temperature.
In this context, the present work addresses an experimental study of the secondary atomization
of liquid droplets impacting onto heated, inclined stainless steel surfaces of known roughness, at
moderate absolute Weber numbers (245<We<600). The surface inclination angle, α is varied over a
wide range, in order to better understand its influence on secondary atomization. The range of
surface temperature considered here (100ºC<Tw<300ºC) is relevant in terms of droplet/valve
interaction in port fuel injection systems, since the hottest part of the valve may achieve
temperatures up to 260ºC - film boiling regime - while the temperature within the region where the
fuel is more likely to impinge is about 180ºC – nucleate boiling regime, (e.g. Panão and Moreira,
2005, Cowart and Cheng, 1999).
The effect of the inclination angle is analyzed in terms of morphology and secondary droplet
size distribution, for droplets of different liquids (water and isooctane). The morphological
characterization is carried out by visualization, while the quantitative analysis of secondary breakup
is performed by combining image analysis techniques and PDA measurements. The combination of
these two techniques allows to cover a wide range of droplet diameters, from 5.5µm up to few
millimeters, which is vital to have a full characterization of secondary droplet size distribution.
2. Experimental Methods
2.1 Experimental arrangement
Single water and isooctane droplets fall by action of gravity onto the target surface, with all
impacts occurring onto dry surfaces.
The impact conditions for water droplets (D0=2.75mm and U0=2.5ms-1) were kept constant
through all the experiments As for isooctane droplets, the impact conditions were varied
(245<Weber number<600), to infer on the conditions for which the droplet behaviour was closer to
that of water. This working condition was observed to be achieved for
D0=2.53mm and
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
U0=2.5ms-1. The target surface is an AISI316 stainless steel circular disc, (with an average
roughness, Ra=0.311µm and average peak-to-valley height, Rz=2.32µm), electrically heated from
bellow. The surface temperature, Tw is monitored by thermocouples (type K) and controlled by a
PID controller. In the experiments considered here, 100ºC<Tw<300ºC.
Homogeneity of the surface temperature is assured by numerical simulations and thermal
imaging analysis with an infrared thermal imaging camera (NEC San-ei Instruments, model
TH1102). Thermal images of the target plate are post-processed with the Remote Program TH51708 software and converted into spatial distribution grids of temperature values, with 238x254
points resolution.
The impact angle, α, defined as in Figure 1, is varied over a wide range, between 6º<α<90º.
Tables 1 and 2 summarize the working conditions and the physical properties of the liquids
used in the present work.
D0
U0
α
Fig. 1. Definition of the impact angle, α.
Table 1 Working conditions.
Parameters
Range
Impact angle
6º<α<90º
Initial droplet diameter (D0)
2.53mm (Isooctane)
2.75mm (Water)
Working
Temperatures
Ambient temperature
Droplets
Impact
surface
100ºC<TW < 300ºC
Pressure
Atmospheric pressure
Impact velocity (U0)
1.6<U0<2.6ms-1
(Isooctane)
2.5ms-1 (Water)
Working Fluids
Isooctane, water
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Table 2 Physical properties of the working fluids.
Liquid
Surface
Tension
[Nm-1]x103
Density
[kgm-3]
Kinematic
viscosity
[m2s-1]x 106
(20ºC)
(20ºC)
(20ºC)
Water
73.75
996.6
1.0
Isooctane
18.1
690
7.0
The morphological characterization of the impacting droplets is carried out by visualization of
droplet impact, making use of two CCD cameras (a Colour PCO SensiCam 1280x1024pixels, as in
Cossali et al., 2005 and a Kodak Motion Corder Analyser, Series SR, Model PS-120high-speed
camera, with maximum frame rate of 10kfps, as in Moita and Moreira, 2005), which are triggered
by the passage of the droplet through a laser beam hitting onto a photodiode. Each camera was used
at different parts of the work, to provide complementary information.
Local measurements of secondary droplet size were obtained using a phase Doppler DANTEC
system. Details on optical configuration are given in Table 3. The PDA system is triggered by the
same mechanism used for the CCD cameras.
Table 3 Optical configuration of the Phase Doppler Anemometer.
Transmitting optics
Value
Processor parameters
Value
Laser Power [mW]
300
U signal bandwidth [MHz]
1.2
Wavelengths [nm]
514 and 488
V signal bandwidth [MHz]
0.4
Beam spacing [mm]
60
S/N Validation [dB]
0
Transmitter focal length
[mm]
310
Spherical Validation [%]
10
Frequency shift [MHz]
40
Receiving optics
Value
Scattering angle [[Deg]
70
Receiver focal length [mm]
310
2.2 Measurement techniques
Qualitative characterization of morphology is based on extensive image analysis. Primary
droplet diameter, Do and impact velocity, U0 are measured directly from the frames recorded
immediately before impact. Accuracy in the determination of Do is ±1.4% of droplet diameter.
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Determination of impacting velocity, from multiple exposed images is better than 3%.
Post-processing of the recorded images is used to determine the size of the larger secondary
droplets generated at drop impact. A commercial image analysis code together with home built
routines is used to measure the size of the larger secondary droplets, having diameters within the
range from 40µm up to a few millimeters. The maximum (dmax) and minimum (dmin) diameters of
each detected secondary droplet were evaluated and all droplets for which dmax/dmin>2 were
rejected, as in Cossali et al., 2005. More details on the image post-processing procedure are
presented in Cossali et al., (2005) and Moita and Moreira, (2005).
An average of 50 images per τ =t/(D0/U0) is used, in order to obtain measurements statistically
representative to the studied phenomena. This Image Analysis Technique-IAT has an absolute
accuracy better than 25µm.
Figure 2 illustrates a secondary drop size distribution obtained with the IAT technique, using the
Colour PCO SensiCam. When the image acquisition system is set-up with the high-speed camera,
Kodak Motion Corder Analyser, Series SR, Model PS-120), better temporal resolution is achieved
(up to 0.1ms), although at the expense of spatial resolution (accuracy up to 30µm). Hence, using
both cameras allows obtain complementary information.
110
D10 [µm]
100
90
80
70
60
50
0
20
40
60
80
100
τ= t/(D0/U0)
Fig. 2. Temporal evolution of the mean secondary droplet diameter, obtained with the IAT, generated from a water
droplet (We=245) impacting onto a stainless steel surface (Ra=0.311µm, Rz=2.32µm, Tw=161ºC, α=90º), within the
nucleate boiling regime.
PDA measurements were performed at z=2.5 and z=5mm, above the surface and at r=0mm,
r=3mm and r=6mm (where r=0mm corresponds to the center of the impact region). The location of
the measurement points was evaluated based on the space distribution of the large secondary
droplets, obtained with the IAT.
The PDA size range, with the chosen configuration, is between 5.5µm and 250µm, which
partially overlap the IAT size range, in order to allow comparison and scaling of the values obtained
by the two techniques.
The PDA measurements are integrated in time up to the impact instant t0, before comparison
with the IAT probability distribution. This procedure is required as the PDA and the IAT techniques
are statistically different.
Finally, a scaling of the IAT p.d.f. and of the integrated PDA p.d.f. is performed by equating the
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
count values where the two size ranges overlap, obtaining an extended p.d.f., (by means of a least
square method), which is used to evaluate the secondary droplet size diameters. A detailed
description of this procedure may be found in Cossali et al., (2005). Figures 3 a) and b) depict the
temporal evolution of the mean secondary droplet diameters obtained from PDA measurements and
extended p.d.f., respectively, for a water droplet impact within the nucleate boiling regime.
18
D10 [µm]
D10 [µm]
18
14
10
14
10
0
20
40
60
80
τ= t/(D0/U0)
a)
100
0
20
40
60
80
100
τ = t/(D 0/U 0)
b)
Fig. 3. Temporal evolution of the mean secondary droplet diameter, generated from a water droplet (D0=2.75mm,
U0=2.5ms-1) impacting onto a stainless steel surface (Ra=0.311µm, Rz=2.32µm, Tw=161ºC, α=90º), within the nucleate
boiling regime. a) Integrated PDA measurements; b) extended p.d.f.
3. Results and Discussion
The main purpose of the present study is to characterize the effect of the impact angle on the
secondary atomization of single droplets impacting onto hot metallic surfaces. This characterization
is carried out in terms of a qualitative description of droplet morphology and a quantitative analysis
of secondary droplet distribution.
3.1. Morphological characterization
The effect of the impact angle on droplet morphology was investigated using two different
liquids, water and isooctane. Observation of drop impacts within the nucleate boiling regime show
that water and isooctane droplets have similar behavior: the first stage of spreading may be
considered isothermal, as it occurs at a temporal scale which is much faster than that of the heat
transfer from the surface to the liquid. Secondary atomization occurs only at the later stages of
spreading, when the rate of heat removal is large enough to cause phase transition. The process of
bubble explosions is triggered at the edge of the spreading lamella, where the time of contact with
the hot surface is larger. Visual inspection of the images suggests that isooctane droplets break-up
into very small secondary droplets within the nucleate boiling regime, compared to water droplets.
Concerning the effect of the impact angle, any significant modifications are observed in the
morphology of water or isooctane droplets, as the impact angle varies between 90º and 45º, except
for an asymmetry of the spreading lamella and a slight delay in the beginning of the boiling process,
when α decreases. This is illustrated in Figure 4: at α=90º the boiling of the water droplet, begins at
2.5<τ<4, while at α=45º, the boiling process is triggered later (4<τ<5.5), starting at the thinner
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
region of the lamella. Differences in droplet morphology only become more evident at small impact
angles, α≤15º, for which the lamella is significantly asymmetric and the secondary atomization is
weaker, as the time for which the droplet stays in contact with the surface is significantly reduced.
Contrarily to observations from Sikalo et., (2005) on water-glycerine droplets impacting onto nonheated smooth glass droplet, rebound of the impacting droplets was never observed. According to
Sikalo et al., (2005), for the present impact conditions, rebound should occur for a critical impact
angle 3º<α<6º.
Water
Isooctane
Water
Isooctane
Water
Isooctane
a)
b)
c)
Fig.4. Effect of impact angle on droplet morphology during impact of a water droplet (D0=2.75mm, U0=2.5ms-1) and an
isooctane droplet (D0=2.53mm, U0=2.6ms-1) onto a hot stainless steel surface, (Ra=0.311µm, Rz=2.32µm, Tw=161ºC),
within the nucleate boiling regime. a) α=90º; b) α=45º; c) α=15º.
When the surface temperature is increased up to 260ºC, (Figure 5) which is near the
Leidenfrost temperature for water droplets and quite above the Leidenfrost temperature of the
isooctane droplets, water droplets do not suffer major morphological modifications, for 90º<α<45º.
However, a noteworthy distortion of the lamella is already observed at α=45º. On the other hand,
for the same range of impact angles, the isooctane droplets are observed to generate a crown,
according to the mechanism similar to that described in Moita and Moreira, (2005). For α<15º
water and isooctane droplets show similar behavior.
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Water
Isooctane
a)
Water
Isooctane
b)
Water
Isooctane
c)
Fig.5. Effect of impact angle on droplet morphology during impact of a water droplet (D0=2.75mm, U0=2.5ms-1) and an
isooctane droplet (D0=2.53mm, U0=2.6ms-1) onto a hot stainless steel surface, (Ra=0.311µm, Rz=2.32µm, Tw=260ºC),
within the transition and boiling regime. a) α=90º; b) α=45º; c) α=15º.
3.2. Secondary droplet size distribution
Evaluation of secondary droplet diameter is based on PDA measurements. These measurements
are then processed and combined with extensive analysis based measurements, according to the
procedure described in paragraph 2.2. Figure 6 depicts the temporal evolution of the secondary
mean drop diameter (D10) and Sauter Mean Diameter, (D32) obtained from the PDA measurements,
at z=2.5mm from the surface, within the nucleate and film boiling regime, for various impact
angles.
Analysis of the temporal evolution of D10 clearly shows a dependence of secondary droplet
distribution with small impact angles (α<15º) both in the nucleate and in the film boiling regimes.
Hence, the temporal evolution of D10 corresponding to α=90º is globally the largest, both in the
nucleate and in the film boiling regime. As the impact angle gradually decreases, D10 becomes
globally smaller, showing a sharp decrease for α<45º, which may be considered as a critical angle,
bellow which the effect of gravitational forces becomes an important parameter influencing
secondary droplet distributions. Mean diameter determined at α=15º and α=6º is significantly lower
than that from the other impact angles, in accordance to the distinct behavior observed in the
morphological characterization.
Summarizing, decreasing the impact angle leads to a decrease in the mean droplet diameter.
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
This trend is particularly evident for 0<τ<20, which is the most relevant time interval for droplet
impact study.
22
20
90deg
60deg
45deg
30deg
14
15deg
18
D10 [µ m]
D10 [µ m]
18
06deg
14
12
10
10
0
20
40
60
τ = t/(D0/U0)
80
0
100
190
170
150
130
110
90
70
50
30
10
15deg
06deg
60
80
100
170
150
60deg
30deg
40
190
90deg
45deg
20
τ = t/(D0/U0)
a)
D 32 [µ m]
D 32 [µ m]
16
130
110
90
70
50
30
10
0
20
40
60
80
τ = t/(D0/U0)
0
100
20
40
60
80
100
τ = t/(D0/U0)
b)
Fig. 6. Temporal evolution of D10 and D32 obtained from integrated PDA measurements of secondary droplets generated
from a water droplet impacting onto a stainless steel surface (Ra=0.3µm, Rz=2.01µm). Absolute Weber number,
We=ρU02D0/σ =245. a)Nucleate boiling regime (Tw=161ºC), b) Film boiling regime (TW=260ºC).
Regarding the effect of the impact angle, the temporal evolution of the secondary droplet
Sauter Mean Diameter (D32), analysis based solely on the integrated PDA measurements does not
provide any evident trend. This is associated to the fact reported in Cossali et al., (2005) that while
the mean diameter solely evaluated from PDA measurements is close to that evaluated from the
extended p.d.f., the Sauter Mean Diameter is quite different, since the larger droplets detected by the
IAT play an important role in increasing D32. This is shown in Figure 7, which depicts the temporal
evolution of D10 and D32 evaluated from the extended p.d.f. Comparing Figures 6 and 7 one may
easily confirm that the temporal evolution of D10 is very similar to that presented in Figure 6,
specially within the nucleate boiling regime, but D32 evaluated from the extended p.d.f. is
considerably larger than that based solely on PDA measurements. Observing the plots in Figure 7b)
it is now possible to notice a clear trend to D32 decrease with the impact angle. This is also a good
example of the relevance of including the large sized secondary droplets detected by the IAT, in
order to have a complete secondary droplet distribution.
18
24
16
20
14
90deg
60deg
45deg
30detg
15deg
6deg
12
D32 [µm]
D10 [µm]
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
16
12
10
8
0
20
40
60
80
0
τ = t/(D0/U0)
20
40
60
80
τ = t/(D0/U0)
a)
120
600
80
90deg
60deg
45deg
30detg
15deg
6deg
60
40
a)
D32 [µm]
D10 [µm]
100
400
200
b)
0
20
0
0
20
40
60
τ = t/(D0/U0)
80
b)
20
40
60
80
τ = t/(D0/U0)
Fig. 7. Temporal evolution of D10 and D32 obtained from extended p.d.f., obtained for secondary droplet distribution
from a water droplet impacting onto a stainless steel surface (Ra=0.3µm, Rz=2.01µm). Absolute Weber number,
We=ρU02D0/σ =245. a)Nucleate boiling regime (Tw=161ºC), b) Film boiling regime (TW=260ºC).
The effect of the impact angle on the secondary atomization of isooctane droplets was inferred
based on averaged PDA measurements. The measurements were mainly obtained within nucleate
boiling regime, since the data rate was too low within the film boiling regime. The low data rate is
associated to the fact that within the film boiling regime, isooctane droplets disintegrate mainly into
large secondary droplets which are ejected very close to the wall, where it is not possible to measure
using the PDA system with the current optical configuration. Moreover, at α>45º, the crown formed
during the first instants after impact, ejects a few number of small droplets, but most part of these
droplets is not detected since the crown has a small height but quite large upper and lower diameters
(which may be larger than 10mm), so the secondary droplets are ejected very far from the impact
region. Figure 8 depicts the time averaged droplet size distribution obtained from isooctane drop
impacts within the nucleate boiling regime. The secondary droplet size for isooctane droplets is
globally small, so that the effect of the gravitational forces that sweep away the larger droplets thus
contributing to the decrease of the measured mean droplet diameter is expected to be less
significant. Hence, the slight decrease of the mean droplet diameter when the impact angle
decreases from 90º to 45 º, which is shown in Table 4 is not significant. Secondary droplets
generated from isooctane drop impacts are also smaller than those resulting from water drops
impact, which is in accordance to the qualitative analysis and is attributed to the smaller surface
tension of isooctane.
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
0.1
p.d.f.
0.08
0.06
90deg
Table 4 Secondary mean and Sauter mean diameters,
evaluated from PDA measurements for water
(D0=2.75mm, U0=2.5ms-1) and isooctane (D0=2.53mm,
U0=2.6ms-1) drop impacts, within the nucleate film
boiling (TW=161ºC).
45deg
0.04
D10 (Isooctane)
[µm]
D 32(Isooctane)
[µm]
D 32 (Water) [µm]
90deg
30
48
95.8
45deg
28.4
46.3
58.7
0.02
0
0
50
100
150
Diameter [µm]
Fig.8. Secondary droplet size distribution evaluated from PDA measurements for isooctane (D0=2.53mm, U0=2.6ms-1)
drop impacts, within the nucleate film boiling (TW=161ºC).
4. Conclusions
The present paper addresses an experimental study aimed at characterizing the effect of the
impact angle on secondary atomization of water and isooctane droplets, impacting onto a heated
stainless steel, at different boiling regimes. The characterization of the secondary droplet
distribution is carried out by combining Image Analysis Technique (IAT) with phase Doppler
anemometry measurements. This procedure allows obtaining an extended p.d.f. of secondary droplet
sizes over a wide range of diameters, from 5.5µm up to a few millimeters. Analysis of the
secondary mean and Sauter Mean diameters, evaluated from PDA measurements alone and from the
extended p.d.f. show that while the mean diameter solely evaluated from PDA measurements is
close to that evaluated from the extended p.d.f., the Sauter Mean Diameter is quite different, since
the larger droplets detected by the IAT play an important role in increasing D32. Globally the results
support that the measured secondary droplet diameters are faintly influenced by large inclination
angles but become considerably smaller for small inclination angles. This is associated to the fact
that for large inclination angles and particularly at 90º, large secondary droplets stay on the surface
for larger periods of time, enlarging D10 and D32. As the inclination angle becomes smaller (α<15º),
large droplets produced by secondary atomization are swept away due to gravitational forces. The
effect of the gravitational forces in decreasing the measured mean droplet size diameters is less
significant for isooctane drop impacts, as the mean secondary diameters are globally smaller, when
compared to those obtained from water drop impacts. Differences in secondary droplet size
diameters between isooctane and water droplets are attributed to the much smaller surface tension
of isooctane.
5. Acknowledgments
A. S. Moita received a training stage at the Facoltà d’Ingegneria – Università degli Studi di
Bergamo, which was financed by a Marie Curie Training Sites Fellowship n. ENK-CT-2002-50518.
The authors acknowledge the contribution of the National Foundation of Science and
Technology by supporting A. S. Moita with a Fellowship (Ref:SFRH/BD/18250/2004) and by
supporting the research (POCI/EME/57944/2004). The authors are also grateful to Mr. Tiago Moita
for his technical support in the electronic part of the experimental arrangement at Instituto Superior
Técnico.
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
References
Bai, C., Rusche, H. and Gosman, A.D., “Modeling of Gasoline spray impingement”, Atomization and
Sprays, Vol. 12, pp. 1 -27, 2002.
Cossali, G. E., Marengo, M., Santini, M., Watanabe, J. “Secondary droplet atomization from single drop
impact on heated surfaces”, Proc. 18th ILASS-Europe2002, Zaragoza, pp. 111-116, 2002.
Cossali, G. E., Marengo, M., Santini, M., “Secondary atomization produced by single drop vertical
impacts onto heated surfaces”, Exp. Thermal Fluid Sci., published online, 2005.
Cowart, J., Cheng, W., “Intake Valve Thermal Behavior During Steady State and Transient Engine
Operations”, SAE Technical Papers, 1999-01-3643, 1999.
Fan, L., Reitz, R. D., “Spray and combustion modelling in gasoline direct-injection engines”,
Atomization and Sprays, Vol. 10, pp. 219 – 249, 2000.
Girard, J.W., Dibble, R. W., Flowers, D.L. and Aceves, S. M., “An Investigation of the Effect of FuelAir Mixedness on the Emissions from an HCCI Engine”, SAE Technical paper, No. 2002-01-1758, 2002.
Karl, A., Anders, K., Rieber, M, Frohn, A, “Deformation of liquid droplets during collisions with hot
walls: experimental and numerical results”, Part. Syst. Charact., Vol.13, pp. 186-191, 1996.
Karl, A., Frohn, A., “Experimental investigation of interaction processes between droplets and hot
walls”, Physics of Fluids, (12) 4, p.785-796, 2000.
Kuhnke, D., “Spray Wall Interaction Modeling by Dimensionless Data Analysis”, Ph.D Thesis accepted
by Fachbereich Mathematik, Technischen Universitat Darmstadt, 2004.
Lee, S.H., Ryou, H.S., “Development of a new spray/wall interaction model”, Int. J. Multiphase Flow,
vol. 26, pp.1209-1234, 2000.
Lindgren, R., Denbratt, I., “Modeling Gasoline Spray-Wall Interactions and Comparison to
Experimental Data”, SAE Technical paper, no. 2004-01-3003, 2004.
Lindgren, R., Denbratt, I., “Modeling Gasoline Spray-Wall Interaction – A Review of Current Models”,
SAE Paper 2000-01-2808, 2000.
Moita, A. S. Moreira, A. L., “The interaction of Fuel Droplets with Heated Interposed Surfaces in IC
Engines”, SAE_ICE2005, Italy, 2005.
Mundo, C., Sommerfeld, M. and Tropea, C., “On the modeling of liquid sprays impinging on surfaces”,
Atomization Sprays 8, 625, 1998.
Mundo, C., Tropea, C. , Sommerfeld, M., “Numerical and Experimental Investigation of Spray
Characteristics in the Vicinity of a Rigid Wall”, Experimental Thermal and Fluid Science, vol. 15, pp. 228237, 1997.
Panão, M. R. O., Moreira, A. L. N., “Thermo and Fluid dynamics characterization of spray cooling with
pulsed spray”, Experimental Thermal and Fluid Science, 30, p.79-96, 2005.
Rusche, H., “CFD Simulation of Spray Impingement Processes”, PhD Thesis, 1997.
Thedorakakos, A., Gavaises, M. and Bergeles, G., “Modelling wall impaction of Diesel Sprays”, Int. J. Heat
and Mass Transfer, Vol. 17, pp. 130-138, 1996.
Sikalo, S., Tropea, C., Ganiec, E. N., “Impact of droplets onto inclined surfaces”, Journal of
Colloid and Interface Science, 286, pp.661-669, 2005
Tomoda, T, Kubota, M, Shimizu, R., Nomura, Y., “Numerical Analysis of Mixture Formation of a
Diesel Injection Gasoline Engine”, The Fifth International Symposium on Diagnostics and Modeling of
Combustion in Internal Combustion Engines, Comodia2001, Nagoya, 2001.
Tropea, C., Roisman, I, “Spray/wall interaction: how much physics do we need in our models?”, 6th
World Conf. on Exp. Heat Transfer, Fluid Mechanics, and Thermodynamics, April 17 -21, Japan, 2005.
Tropea, C., Roisman, I.V., “Modelling of Spray Impact on Solid Surfaces”, Atomization and Sprays,
vol. 10, pp. 387-408, 2000.
Wang, Y. P., Wilkinson, G. B., Drallmeier, J. A., “Parametric study on the fuel film breakup of a cold
start PFI engine”, Exp. Fluids, 37, p.385-398’ 2004.
Yao, S. C., Cai, K. Y., “The dynamics and Leidenfrost temperature of drops impacting on a hot surface
at small angles”, Exp. Thermal Fluid Sci, 1, pp. 363-371, 1988.
Download