Experimental Study of Steady Quasi-Cone-JetElectrostatic Sprays of Biobutanol
for Engine ApplicationsMaria S. Agathou1 and Dimitrios C. Kyritsis2
Abstract: Butanol produced from agricultural sources is emerging as a potentially renewable biofuel for use in engine applications. In thiswork, butanol electrostatic sprays were established within a narrow region of low flow rates. Spray phenomenology was investigated throughhigh-speed visualization for the low flow rate conditions in the vicinity of the lowest voltage for which electrosprays (e-sprays) could besustained. Spray structure was studied through droplet size and velocity measurements for a combination of conditions, performed by usingphase Doppler anemometry. Combined with high-speed spray visualization, these measurements revealed a stable e-spray operation withnarrow droplet size and velocity distributions; i.e., spray behavior was close to monodisperse. A similarity analysis was performed to developan empirical expression correlating appropriately dimensionless average diameter, flow rate, and applied voltage. DOI: 10.1061/(ASCE)EY.1943-7897.0000141. © 2014 American Society of Civil Engineers.
Author keywords: Biobutanol; Electrosprays; Internal combustion engines.
Introduction
To increase power output, improve fuel economy, and meet strin-gent emission regulations, scientists of the automotive, or moregenerally, the energy industry, focus their interest on the domainof techniques related to the design or operating conditions of en-gines (Pulkrabek 2004; Rakopoulos et al. 2004a, b, 2009; Abu-Jraiet al. 2009; Levendis et al. 1994); i.e., to diesel or spark ignitionengines (Giakoumis et al. 2012; Alkidas 2007; Rakopoulos et al.2008; Rakopoulos 2012) that are well established as dominatingpowertrain solutions in the world market. Particularly for the reduc-tion of pollutant emissions, the interest is concentrated on thedomain of fuel-related techniques, such as the use of alternativegaseous fuels that are friendly to the environment, such as methane,syngas, or hydrogen (Barlow et al. 2001, 2009; Rakopoulos andMichos 2008; Vancoillie et al. 2012), or oxygenated fuels thatpresent reduced particulate emissions (Rakopoulos et al. 2011,2012). In various countries, considerable attention has been paidto the development of alternative fuel sources, with emphasis onbiofuels that are oxygenated by nature, which possess the advan-tages of being renewable, biodegradable, and nontoxic, such asdiesel engines fueled with blends of diesel fuel with vegetable oilsand biodiesels, ethanol, n-butanol, or diethyl ether (Giakoumis et al.2013; Rakopoulos 2013; Rakopoulos et al. 2013).
Recent research has shown the possibility of efficientlyproducing butanol with the use of fermentation microorganisms(clostridia); namely, Ezeji et al. (2007) and Tashiro et al. (2007),who had earlier presented results on the coproduction of butanolwith ethanol and acetone (Tashiro et al. 2005), demonstrated bio-butanol production with the use of the microorganism chlostridiumbeijerinckii and highlighted the potential use of biobutanol as analternative fuel derived from renewable sources. Butanol has asignificantly increased energy density over to the currently widelyused bioethanol. Specifically, the energy densities of butanol andethanol are 36.4 and 24.8 MJ=kg, respectively, whereas the energycontent of gasoline is 44.9 MJ=kg. Notably, butanol mixes readilywith diesel fuel, unlike ethanol.
Butanol is particularly appropriate for the technology of electro-static atomization because of the −OH bond in its molecule, whichbecomes polarized under an electrostatic field of elevated intensity,converting butanol into a relatively conductive molecule. In thefield of electrostatic atomization, promising results have recentlybeen reported in terms of utilizing this technology for butanol(Agathou et al. 2007). Comparative studies of butanol with otherfuels of automotive interest have been reported (Agathou andKyritsis 2010, 2012a), and further results on butanol blends havebeen given (Agathou and Kyritsis 2012b). Early work was per-formed in the context of direct injection (Anderson et al. 2007a)and focused on the effect of the electrostatic field on the structureof electrostatically assisted sprays (Anderson et al. 2007b) and theperformance of sprays of gasoline-ethanol blends (Anderson andKyritsis 2007). The fundamental combustion properties of butanolhave been compared with those of well-established fuels such asmethane and heptane (Agathou and Kyritsis 2011, 2012c), andthe actual performance of biobutanol diesel blends was studiedin diesel engines (Rakopoulos et al. 2010b). The advantages ofbiobutanol blending in diesel are evident when it is compared tovegetable oil, biodiesel, or ethanol (Rakopoulos et al. 2007, 2010a).The behavior of butanol/gasoline blends in spark-ignition engineshas also been reported (Gu et al. 2012; Irimescu 2012).
The possibility of automotive application of the electrospray(e-spray) technology has been discussed for early injection in
1Senior Engineer, Hellenic Petrol, 5 Chimaras St., 15125 Athens,Greece; formerly, Dept. of Mechanical Science and Engineering, Univ.of Illinois at Urbana-Champaign, 1206 W. Green St., Urbana, IL 61801.
2Associate Professor, Dept. of Mechanical Science and Engineering,Univ. of Illinois at Urbana-Champaign, 1206 W. Green St., Urbana, IL61801; and Professor, Dept. of Mechanical Engineering, Khalifa Univ.of Science, Technology and Research, P.O. Box 127788, AbuDhabi, UnitedArab Emirates (corresponding author). E-mail: [email protected];[email protected]
Note. This manuscript was submitted on April 15, 2013; approved onJune 6, 2013; published online on June 8, 2013. Discussion period openuntil July 20, 2014; separate discussions must be submitted forindividual papers. This paper is part of the Journal of Energy Engineering,© ASCE, ISSN 0733-9402/A4014008(10)/$25.00.
© ASCE A4014008-1 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.
Spark-ignited (SI) engines (Shrimpton 2003) and for transientatomization (Kourmatzis and Shrimpton 2011). More studies inthe automotive field include charge injection atomizers (Shrimptonand Kourmatzis 2010) and electrostatic atomization insertion intoCompression-ignition (CI) engines (Thomas et al. 2002). Togetherwith the work of Anderson and Kyritsis (2007), these works seemto indicate a possibility for the practical application of automotive,electrostatically manipulated sprays if appropriate microfabricatedinjectors are made available.
For low conductivity liquids, namely hydrocarbons, there is aregion of stable e-spray operation termed the cone-jet mode, inwhich the e-spray consists of three major sections: a conical me-niscus, which is formed at the outer surface of the nozzle and is also
known as a Taylor cone; a liquid ligament, which breaks up furtherdownstream because of Rayleigh instability; and a spray of nearlyuniform sized droplets (Tang and Gomez 1994, 1996; Gomez andTang 1994). Gomez proposed a mesoscale burner based on kero-sene e-sprays (Kyritsis et al. 2004a, b), in which two features of thee-spray were highlighted: First, when operated in cone-jet mode,the e-spray can produce practically monodisperse sprays; second,the droplet size of the monodisperse e-sprays can be controlled bythe mass flow rate per spray. The morphology of the e-sprays wasextensively studied by Cloupeau and Prunet-Foch, both in the cone-jet mode regime (1989) and in other functioning modes (1994).
The objective of this paper is to experimentally investigatewhether these particularly attractive features of the e-spray can
(a) (b)
Fig. 1. Electrospray configuration: (a) typical e-spray; (b) system creating the e-spray
Fig. 2. Electrospray imaging results for small flow rates of butanol
© ASCE A4014008-2 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.
be achieved for butanol. Previous studies on butanol e-sprays(Agathou et al. 2007; Agathou and Kyritsis 2010, 2012a, b) sug-gested an e-spray operation away from the stable cone-jet regimeand practically produced polydisperse sprays. In these distribu-tions, the 10% ratio of the SD over the average diameter, whichis widely used as a criterion for monodispersity, was not achieved;rather, this ratio had a value on the order of 35%. The purpose ofthis study was to investigate whether this could be cured by utiliz-ing smaller flow rates to achieve monodisperse sprays.
Experimental Methodology
E-Spray Experimental Setup
Fig. 1(a) shows a typical e-spray and Fig. 1(b) shows a schematicof the system realizing the e-spray. The apparatus consisted of asyringe pump that fed the injector with butanol and controlledthe flow rate. Commercially available glass capillaries were usedas injectors, after applying a layer of silver on their outer surfacesthat acted as a conductor. The capillaries had outer and inner
0
(a) (b)
(c) (d)
(e) (f)
5
10
15
0 5 10 15 20 25 30 35 40
1ml/hr1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Diameter [µm]
Per
cent
age
[%]
0
5
10
15
0 5 10 15 20 25 30 35 40
1ml/hr1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Diameter [µm]
0
5
10
15
0 5 10 15 20 25 30 35 40
1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Diameter [µm]
0
5
10
15
0 5 10 15 20 25 30 35 40
1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Diameter [µm]
0
5
10
15
0 5 10 15 20 25 30 35 40
2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Diameter [µm]
0
5
10
15
0 5 10 15 20 25 30 35 40
2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Diameter [µm]
Per
cent
age
[%]
Fig. 3. Distribution of the droplet diameters of low flow rates of butanol for various voltages: (a) 3.4 kV; (b) 3.5 kV; (c) 3.6 kV; (d) 3.7 kV; (e) 3.8 kV;(f) 3.9 kV; (g) 4.0 kV; (h) 4.1 kV; (i) 4.2 kV; (j) 4.3 kV; (k) 4.4 kV; (l) 4.5 kV
© ASCE A4014008-3 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.
diameters of 1 mm and 30 μm, respectively, and were charged ata potential on the order of several kV provided by a high voltagepower supply. The charged spray was collected at an electricground in the form of a steel mesh placed 11 mm from the capillaryexit. The liquid was drained in a liquid beaker placed beneath theelectric ground. The liquid under consideration was pure 1-butanol(n-butanol). Butanol has a relatively high electrical conductivitycompared to commercially available hydrocarbons. For this reason,e-spraying was possible without the addition of conductivityenhancers.
Measurements of Droplet Size and Velocity: PhaseDoppler Anemometry
For spray droplet size and velocity measurements, a Dantec fiberphase Doppler anemometer (PDA) was used (Dantec Dynamics,Skovlunde, Denmark). It was powered by a 5 WAr-ion laser andwas run in one-dimensional (1D) PDA mode, utilizing only514.5 nm of the Ar-Ion laser and collecting at a forward scatteringangle of 71.4°. For each measurement, 10,000 data points were se-lected, which was sufficient for adequate statistics. Droplet size andvelocity measurements were performed at a location on the central
0
(g) (h)
(i) (j)
(k) (l)
5
10
15
0 5 10 15 20 25 30 35 40
2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Diameter [µm]
0
5
10
15
0 5 10 15 20 25 30 35 40
2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Diameter [µm]
0
5
10
15
0 5 10 15 20 25 30 35 40
2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Diameter [µm]
0
5
10
15
0 5 10 15 20 25 30 35 40
2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Diameter [µm]
0
5
10
15
0 5 10 15 20 25 30 35 40
2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Diameter [µm]
Per
cent
age
[%]
0
5
10
15
0 5 10 15 20 25 30 35 40
2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Diameter [µm]
Fig. 3. (Continued.)
© ASCE A4014008-4 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.
spray axis located 11 mm from the glass capillary exit. The collec-tion point location is shown in Fig. 1(b). Measurements wereobtained at only one location, because the goal was to establishthe possibility of monodisperse sprays, not the spatial distributionof droplet size.
High-Speed Visualization
Visualization of the butanol e-sprays was achieved with a VisionResearch Phantom v7.0 high-speed video camera (VisionResearch, Wayne, NJ). Three Nikon extension tubes totaling82.5 mm in length (Nikon, Tokyo, Japan) were used to increasemagnification, along with a Tiffen 52 mm zoom lens (Tiffen,Happauge, NY). Camera exposure time was set to 96 μs andthe zoom and f-number (defined as the ratio of focal length ofthe diameter of the aperture) settings were adjusted to provide aclear view of the entire spray. Laser light from an Oxford LaserSystems LS 20-50 copper-vapor laser (Oxford Lasers, Shirley,MA) operating at 510.6 nm with a frequency of 10 kHz was di-rected toward the doublet of the concave spherical-cylindrical lensthat was used to generate the laser sheet. Both the high speed cam-era and the laser were synchronized by a Berkeley Nucleonics500C042 pulse generator (Berkeley Nucleonics, San Rafael, CA).To acquire images of a two-dimensional (2D) cross section of thespray, the laser sheet was aligned so that its top would be just belowthe capillary tip and the bottom would be approximately 200 mmbelow the orifice; thus, the whole spray was captured. The thick-ness of the laser sheet was approximately 1 mm.
Results and Discussion
E-Spray Phenomenology
The establishment of cone-jet sprays requires a finer increment offlow rate and applied voltages than those used previously (Agathouet al. 2007; Agathou and Kyritsis 2010, 2012a, b) so that they canbe tracked, especially for a fuel that has not been investigated fromthis aspect before and has potentially novel modes of e-spray op-eration. During earlier experiments (Agathou et al. 2007; Agathouand Kyritsis 2010, 2012a, b), the applied flow rate was in the rangeof 5–30 mL=h, the voltage varied from 4–7 kV, and the e-sprayswere unsteady and polydisperse. The distance between the capillarytip and the ground electrode (liquid beaker) was 11 mm; it was kept
constant during the current experiments and equal to the one usedbefore (Agathou et al. 2007; Agathou and Kyritsis 2010, 2012a, b).In this work, smaller butanol flow rates were investigated:1–4.5 mL=h. The typical structure of the ensuing e-sprays is shownin Fig. 2. These data were collected in the range of the minimumvoltage that could support an e-spray, i.e., 3.3–3.5 kV. For a lowervoltage, atomization was not achieved and only the fuel liquidligament was observed. For those low flow rates, the e-sprays weresteady with a conical stable meniscus at the outside surface of thecapillary tip, exactly as expected from the cone-jet mode. In somecases, satellite droplets were present that deviated from the mainbody of the spray; in the rest of the cases, the fan of droplets wasnarrower than the sprays reported by Tang and Gomez (1996), andGomez and Tang (1994). Atomization led to an e-spray structuresimilar to the results of Tang and Gomez (1996) and Gomez andTang (1994), in which the established sprays operated in the cone-jet mode with a stable meniscus and a time independent spray.
E-Spray Structure: Droplet Size Measurements
Droplet size distribution results were obtained for flow rates from 1to 4.5 mL=h with an increment of 0.5 mL=h. The applied voltagesvaried from 3.4 to 4.5 kV, with an increment of 0.1 kV. The goalwas to investigate the possibility of monodisperse sprays, the exist-ence of which is conditioned on the stability of the conical menisciat the injector tip. Probability density function results of the average
0
5
10
15
20
25
3 3.5 4 4.5 5
1ml/hr1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Ave
rage
dia
met
er d
10 [µ
m]
Applied voltage [kV]
Fig. 4. Effects of applied voltage on the mean droplet diameter ofbutanol
0
(a)
(b)
5
10
15
20
25
0 1 2 3 4 5
3.4kV3.5kV3.6kV3.7kV3.8kV3.9kV
Ave
rage
dia
met
er d
10 [ µ
m]
Flow rate [ml/hr]
0
5
10
15
20
25
0 1 2 3 4 5
4.0kV4.1kV4.2kV4.3kV4.4kV4.5kV
Flow rate [ml/hr]
Ave
rage
dia
met
er d
10 [µ
m]
Fig. 5. Effects of flow rate on the mean droplet diameter of butanol forvarious voltages: (a) 3.4–3.9 kV; (b) 4.0–4.5 kV
© ASCE A4014008-5 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.
Table 1. Overall Exponents of the Power Law and the Corresponding R2
Voltage (kV) 3.4 3.5 3.6 3.7 3.8 3.9 4.0 4.1 4.2 4.3 4.4 4.5
β 0.43 0.22 0.21 0.13 0.18 0.15 0.21 0.80 0.35 0.15 0.11 0.53R2 0.95 0.80 0.93 0.82 0.83 0.79 0.93 0.72 0.37 0.87 0.94 0.63
Note: Power law shown in Eq. (1).
0
(a) (b)
(c) (d)
(e) (f)
5
10
15
20
25
30
35
0 2 4 6 8 10 12 14 16
1ml/hr1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Velocity [m/s]
0
5
10
15
20
25
30
35
0 2 4 6 8 10 12 14
1ml/hr1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Velocity [m/s]
0
5
10
15
20
25
30
35
0 2 4 6 8 10 12 14
1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Velocity [m/s]
0
5
10
15
20
25
30
35
0 2 4 6 8 10 12 14 16
2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Velocity [m/s]
0
5
10
15
20
25
30
0 5 10 15 20
2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Velocity [m/s]
0
5
10
15
20
25
30
0 5 10 15 20
2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Per
cent
age
[%]
Velocity [m/s]
Fig. 6. Velocity distribution for butanol at various voltages: (a) 3.4 kV; (b) 3.5 kV; (c) 3.6 kV; (d) 4.2 kV; (e) 4.3 kV; (f) 4.4 kV
© ASCE A4014008-6 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.
droplet diameter are presented in Fig. 3 for 12 values of appliedvoltage. In data processing, the average droplet diameter d10 ispresented.
Contrary to the insensitivity of droplet size on applied voltageobserved in previous works (Agathou et al. 2007; Agathou andKyritsis 2010, 2012a, b; Tang and Gomez 1996; Gomez and Tang1994), a dependence on voltage was clearly observed for low flowrates of butanol. To capture this dependence, small increments ofvoltage were selected. The investigation took place for the voltagesthat did generate a spray. For example, for the lowest flow rates(1–2 mL=h), an e-spray could not be established for a voltage lowerthan 4.5 kV. Clearly, the electrostatic repulsion in the liquid was notsufficient to break the liquid meniscus.
First, at the lowest voltages (3.4 and 3.5 kV), the droplet sizedistribution for flow rates ranging from 1 to 3 mL=h is narrow.Hence, near the voltage limit where the e-sprays start forming,the sprays demonstrate a behavior that resembles a cone-jet andis expected to produce monodisperse droplets. However, after adeparture from the narrow distributions (Cases 3, 4, and 5), thee-sprays return to an almost cone-jet behavior for voltages ofapproximately 4.3 kV; this occurs for all flow rates, not only forthe lower ones, as in the case of low voltages (Case 1). In thesecases, a typical ratio of the SD of the distributions over the meandroplet diameter is equal to 18%, which is fairly close to the 10%threshold typically used as a criterion for monodispersity. As acomparison, in the results of Tang and Gomez (1996) and Gomezand Tang (1994), this ratio is typically equal to 6–8%.
Fig. 4 shows the dependence of droplet size on voltage for aconstant flow rate. Again, contrary to previous literature findings(Agathou et al. 2007; Agathou and Kyritsis 2010, 2012a, b; Tangand Gomez 1996; Gomez and Tang 1994), this relation is not mon-otonic. An increase of droplet size is clear that peaks in the vicinityof 3.7 kV, abruptly decreases to a minimum value in the region of4.3 kV and then another increasing trend. The voltages for whichthe minima are shown in Fig. 4 correspond to the narrowest dropletsize distributions of Fig. 3. Results for heptane (Tang and Gomez1996; Gomez and Tang 1994) indicate an insensitivity of dropletsize to voltage for flow rates smaller than approximately 10 mL=h.
The effect of liquid flow rate on average droplet size under con-stant voltage is shown in Figs. 5(a and b). The results indicate anincrease in droplet size with increasing flow rate. These findingsagree with the already established finding that droplet size can bedecreased with diminishing flow rates (Tang and Gomez 1996;Gomez and Tang 1994). A power law dependence, as reportedby Tang and Gomez (1996) and Gomez and Tang (1994), wasobtained with a satisfactory regression coefficient for many of thecases investigated here. The equation used for the regression was ofthe form
d10 ¼ α ×mβ ð1Þwherem = flow rate; α = constant; and β = exponent. The values ofβ and the corresponding regression coefficients (R2) are presentedin Table 1. In most cases, β is very small (<0.3), which actuallyindicates a very mild dependence of droplet size on flow rate.
E-Spray Structure: Velocity Measurements
Velocity distributions are provided in Fig. 6 for six of the casespresented in Fig. 3. The panels presented in the figure correspondto the regions where a behavior closer to cone-jet was observed inthe previous section; that is, the cases that have the lowest ratios ofSD over the mean droplet diameter. The sprays showing this quasi-monodisperse nature developed in the region of approximately 3.4and 4.3 kV. Narrow velocity distributions are shown in the cases of
the lowest flow rates for voltages of 3.4 and 3.5 kV, whereas withincreasing flow rate the distributions become equally broad, show-ing insensitivity to voltage. This insensitivity is also observed forall flow rates of voltages 3.6, 4.2, 4.3, and 4.4 kV.
The effects of voltage and flow rate on the average velocityare shown in Fig. 7. An increase in droplet velocity is shownwith increasing voltage, irrespective of flow rate. Contrary to thenegligible flow rate effect, the voltage effect is substantial andtranslates to higher velocities for higher applied voltages. Reynolds(R) and Weber (W) numbers for the droplet motion were calculatedand are presented in Figs. 8 and 9 for constant flow rate andvoltage, respectively. Fig. 8 shows that R and W increase withincreasing flow rate, which is observed for all voltages. However,in Fig. 9, the relation between R and W with voltage is not mon-otonic. Instead, two monotonic regions and a velocity decrease atapproximately 4 kVare shown for all flow rates, which may imply atransition to a different flow regime for all cases.
The values ofW allow prediction of the droplet breakup mecha-nism according to the analysis of Pilch and Erman (1987). In par-ticular, the flow regime corresponding toW between 100 and 300 isthat of sheet stripping, in which the droplet breakup mechanism isbased on the creation of an ellipse from the initial droplet, whichbreaks down into a series of satellite droplets around its two edges.
0
5
10
15
2 2.5 3 3.5 4 4.5 5
Butanol
1ml/hr1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Uav
e [m/s
]
Voltage [kV]
0
5
10
15
0 1 2 3 4 5 6
Butanol
3.4kV3.5kV3.6kV3.7kV3.8kV3.9kV4.0kV4.1kV4.2kV4.3kV4.4kV4.5kV4.6kV
Uav
e [m/s
]
Flow Rate [ml/hr]
Fig. 7. Effects of voltage and liquid flow rate on the mean velocity ofbutanol
© ASCE A4014008-7 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.
Regions withW ≤ 350 correspond to voltages below 3.6 kV, wherethe narrowest droplet size distributions are visible. This mechanismof the generation of satellite droplets through an unstable sheet cancontribute to the formation of almost uniform size droplets. In starkcontrast to this, for W ≥ 350, droplet breakup occurs in the cata-strophic regime (Pilch and Erman 1987). In this scenario, secon-dary droplets develop in a cloud around the surface of the initialdroplet as they reduce its size in an arbitrary manner. For the flowrates under investigation, although the droplet size distributionsrevealed an ∼18% uniformity in size distribution, which wassignificantly improved compared to higher butanol flow rates(Agathou et al. 2007; Agathou and Kyritsis 2010, 2012a, b), theywere still polydisperse. The source of polydispersion can be attrib-uted to the breakup mechanism, which is the catastrophic regimethat does not favor the creation of uniformly sized droplets.
Nondimensional Controlling Parameters
In this section, the possibility is investigated of producing empiricalcorrelations among the three variables under consideration: thedroplet diameter, the voltage, and the flow rate. According toFernandez de la Mora et al. (1990), the nondimensional dropletdiameter is a function of the nondimensional voltage, the flow rate,and the dielectric constant, ε, of the liquid:
d�fðQ�;V�; εÞ ð2ÞThus, for a specific liquid, d� is a function of V� and Q�. The
formulation of the nondimensional parameters is dependent on themode of operation of the e-spray; for a stable cone-jet regime, it isas follows (Fernandez de la Mora et al. 1990):
Q� ¼ ρ · λ · Qε · ε0 · σ
ð3Þ
V� ¼ V − V0
V0
ð4Þ
d� ¼�ρ · λ2 · d3
ðε · ε0Þ2 · σ�1=3
ð5Þ
where ρ = density of the liquid; ε0 = electric permittivity of air; andV0 = onset voltage, which was assumed to be 3 kV for this analysis.Eqs. (3) and (5) come from dimensional analysis. As indicated inthe previous sections, e-sprays of low butanol flow rates present astable behavior, although the 10% criterion used in previousstudies to establish monodispersity was not satisfied. Conductinga surface fit over the nondimensional data, an empirical relationwas generated that correlated d�, V�, and Q� by utilizing athird-order polynomial:
0
100
200
300
400
500
600
700
0 1 2 3 4 5 6
3.4kV3.5kV3.6kV3.7kV3.8kV3.9kV4.0kV4.1kV4.2kV4.3kV4.4kV4.5kV4.6kV
Re
Flow Rate [ml/hr]
Butanol
0
100
200
300
400
500
600
700
800
0 1 2 3 4 5 6
Butanol
3.4kV3.5kV3.6kV3.7kV3.8kV3.9kV4.0kV4.1kV4.2kV4.3kV4.4kV4.5kV4.6kV
We
Flow Rate [ml/hr]
Fig. 8. Effects of flow rate on R and W
0
100
200
300
400
500
600
700Butanol
1ml/hr1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
Re
Voltage [kV]
0
100
200
300
400
500
600
700
800
3 3.5 4 4.5 5
3 3.5 4 4.5 5
Butanol
1ml/hr1.5ml/hr2ml/hr2.5ml/hr3ml/hr3.5ml/hr4ml/hr4.5ml/hr
We
Voltage [kV]
Fig. 9. Effects of voltage on R and W
© ASCE A4014008-8 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.
d� ¼ 3.216þ 36.57 · V� þ 0.02068 · Q� − 167.3 · V�20.1811
· Q� · V� þ 0.000565 · Q�2 þ 209.3 · V�3 þ 1.305 · Q�
· V�2 − 0.006392 · Q�2 · V� þ 7.538 · 10−6 · Q�3 ð6Þ
The regression coefficient of the correlation is R2 ¼ 93.14% andall coefficients of the polynomial are reported with 95% confidencebounds. This correlation basically presents all of the current resultsin one equation, without revealing any fundamental characteristicsof the underlying physics.
Conclusions
Biobutanol e-sprays can be achieved in the cone-jet mode in a narrowregime of voltages near the onset of electrostatic atomization forflow rates on the order of 1 to 4.5 mL=h. The ratio of SD to averagedroplet diameter ranges from 18 to 23%. This is substantially largerthan the 10% criterion typically used to establish monodispersion,but it is also significantly smaller than the corresponding values mea-sured for larger flow rates (Agathou et al. 2007; Agathou and Kyritsis2010, 2012a, b), which reached 35%. Thus, substantially narrowerdroplet distributions can be achieved by operating at low flow ratesand voltages near the onset of the e-spray. High speed visualizationshowed a stable e-spray behavior under these conditions withoutoscillations of the liquid meniscus from which the spray ensued.The dispersion in droplet size is attributed to the high Weber numberof the droplet motion, which can cause secondary atomization.
Acknowledgments
The authors would like to acknowledge the support of the U.S.DepartmentofEnergy through theGraduateAutomotiveTechnologyEducation (GATE) Center of Excellence at the University of Illinois.
Notation
The following symbols are used in this paper:d = average droplet diameter (m);d� = nondimensional droplet diameter;Q = volumetric flow rate (m3=s or mL=h);Q� = nondimensional flow rate;q = total electric charge on the droplet surface (C);R = Reynolds number (Ud=v);U = velocity of the droplet (m=s);
UAVE = average velocity of the droplet (m=s);V = applied voltage (V);V� = nondimensional voltage;Vo = onset voltage (V);W = Weber number (ρU2d=σ);ε = liquid dielectric constant;ε0 = electric permittivity of the surrounding medium (s=Ωm);λ = electrical conductivity (1=Ωm);ν = kinematic viscosity (m2=s);ρ = density (kg=m3); andσ = surface tension (N=m).
References
Abu-Jrai, A., et al. (2009). “Performance, combustion and emissions of adiesel engine operated with reformed EGR. Comparison of diesel andGTL fuelling.” Fuel, 88(6), 1031–1041.
Agathou, M. S., and Kyritsis, D. C. (2010). “A comparative experimentalstudy of butanol electrosprays through phase-Doppler anemometry.”2010 Technical Meeting of the Central States Section of the CombustionInstitute, Combustion Institute, Pittsburgh, PA.
Agathou, M. S., and Kyritsis, D. C. (2011). “An experimental comparisonof non-premixed bio-butanol flames with the corresponding flames ofethanol and methane.” Fuel, 90(1), 255–262.
Agathou, M. S., and Kyritsis, D. C. (2012a). “Electrostatic atomization ofhydrocarbon fuels and bio-alcohols for engine applications.” EnergyConvers. Manage., 60, 10–17.
Agathou, M. S., and Kyritsis, D. C. (2012b). “Fuel composition effect onthe electrostatically-driven atomization of bio-butanol containingengine fuel blends.” Energy Convers. Manage., 60, 28–35.
Agathou, M. S., and Kyritsis, D. C. (2012c). “Experimental investigationof bio-butanol laminar non-premixed flamelets.” Appl. Energy, 93,296–304.
Agathou, M. S., Powell, J. W., Lee, C. F., and Kyritsis, D. C. (2007).“Preliminary experimental study of butanol electrosprays for powergeneration.” SAE Paper No. 2007-24-0020, Society of AutomotiveEngineers, Warrendale, PA.
Alkidas, A. C. (2007). “Combustion advancements in gasoline engines.”Energy Convers. Manage., 48(11), 2751–2761.
Anderson, E. K., Carlucci, A. P., De Risi, A., and Kyritsis, D. C. (2007a).“Electrostatic effects on gasoline direct injection in atmosphericambiance.” Atomization Sprays, 17(4), 289–313.
Anderson, E. K., Carlucci, A. P., De Risi, A., and Kyritsis, D. C. (2007b).“Experimental investigation of the possibility of automotive gasolinespray manipulation through electrostatic field.” Int. J. Vehicle Des.,45(1–2), 61–79.
Anderson, E. K., and Kyritsis, D. C. (2007). “Experimental investigation ofcombustion of electrostatically charged ethanol blended gasoline drop-lets.” 5th Joint Meeting of the U.S. Sections of the Combustion Institute,Combustion Institute, Pittsburgh, PA.
Barlow, R. S., Karpetis, A. N., Frank, J. H., and Chen, J.-Y. (2001). “Scalarprofiles and NO formation in laminar opposed-flow partially premixedmethane/air flames.” Combust. Flame, 127(3), 2102–2118.
Barlow, R. S., Ozarovsky, H. C., Karpetis, A. N., and Lindstedt, R. P.(2009). “Piloted jet flames of CH4/H2/air: Experiments on localizedextinction in the near field at high Reynolds numbers.” Combust.Flame, 156(11), 2117–2128.
Cloupeau, M., and Prunet-Foch, B. (1989). “Electrostatic spraying ofliquids in cone-jet mode.” J. Electrostatics, 22(2), 135–159.
Cloupeau, M., and Prunet-Foch, B. (1994). “Electrohydrodynamic sprayingfunctioningmodes: A critical review.” J. Aerosol Sci., 25(6), 1021–1036.
Ezeji, T., Qureshi, N., and Blaschek, H. P. (2007). “Production of acetone-butanol-ethanol in a continuous flow bioreactor using degermed cornand chlostridium beijernickii.” Process Biochem., 42(1), 34–39.
Fernandez de la Mora, J., Navascues, J., Fernandez, F., and Rossel-Liompart, J. (1990). “Generation of submicron monodisperse aerosolsin electrosprays.” J. Aerosol Sci., 21, S673–S676.
Giakoumis, E. G., Rakopoulos, C. D., Dimaratos, A. M., and Rakopoulos,D. C. (2012). “Exhaust emissions of diesel engines operating undertransient conditions with biodiesel fuel blends.” Prog. Energy Combust.Sci., 38(5), 691–715.
Giakoumis, E. G., Rakopoulos, C. D., Dimaratos, A. M., and Rakopoulos,D. C. (2013). “Exhaust emissions with ethanol or n-butanol diesel fuelblends during transient operation: A review.” Renew. Sust. Energy Rev.,17(1), 170–190.
Gomez, A., and Tang, K. Q. (1994). “Charge and fission of droplets inelectrostatic sprays.” Phys. Fluids, 6(1), 404–414.
Gu, X., Huang, Z., Cai, J., Gong, J., Wu, X., and Lee, C.-F. (2012).“Emission characteristics of a spark-ignition engine fuelled with gaso-line-n-butanol blends in combination with EGR.” Fuel, 93, 611–617.
Irimescu, A. (2012). “Performance and fuel conversion efficiency of a sparkignition engine fueled with iso-butanol.” Appl. Energy, 96, 477–483.
Kourmatzis, A., and Shrimpton, J. S. (2011). “Electrical and transientatomization characteristics of a pulsed charge injection atomizer usingelectrically insulating liquids.” J. Electrostatics, 69(3), 157–167.
Kyritsis, D. C., Coriton, B., Faure, F., Roychoudhury, S., and Gomez, A.(2004a). “Optimization of a catalytic combustor using electrosprayed
© ASCE A4014008-9 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.
liquid hydrocarbons for mesoscale power generation.” Combust. Flame,139(1–2), 77–89.
Kyritsis, D. C., Roychoudhury, S., McEnally, C., Pfefferle, C., and Gomez,A. (2004b). “Mesoscale combustion: A first step towards liquid fueledbatteries.” Exper. Fluid Thermal Sci., 28(7), 763–770.
Levendis, Y. A., Pavlotos, I., and Abrams, R. F. (1994). “Control of dieselsoot, hydrocarbon and NOx emissions with a particulate trap and EGR.”SAE Paper No. 940460, Society of Automotive Engineers, Warrendale,PA.
Pilch, M., and Erman, C. A. (1987). “Use of breakup time data and velocityhistory data to predict the maximum size of stable fragments foracceleration-induced break-up of a liquid drop.” Int. J. MultiphaseFlow, 13(6), 741–757.
Pulkrabek, W. W. (2004). Engineering fundamentals of internal combus-tion engines, Pearson Prentice-Hall, Piscataway, NJ.
Rakopoulos, C. D., Antonopoulos, K. A., Rakopoulos, D. C., Hountalas,D. T., and Andritsakis, E. C. (2007). “Study of the performance andemissions of a high-speed direct injection diesel engine operating onethanol-diesel fuel blends.” Int. J. Alternat. Propul., 1(2–3), 309–324.
Rakopoulos, C. D., Dimaratos, A. M., Giakoumis, E. G., and Rakopoulos,D. C. (2009). “Evaluation of the effect of engine, load and turbochargerparameters on transient emissions of diesel engine.” Energy Convers.Manage., 50(9), 2381–2393.
Rakopoulos, C. D., Giakoumis, E. G., and Rakopoulos, D. C. (2004a).“Cylinder wall temperature effects on the transient performanceof a turbocharged diesel engine.” Energy Convers. Manage., 45(17),2627–2638.
Rakopoulos, C. D., and Michos, C. N. (2008). “Development and valida-tion of a multi-zone combustion model for performance and nitric oxideformation in syngas fueled spark ignition engine.” Energy Convers.Manage., 49(10), 2924–2938.
Rakopoulos, C. D., Michos, C. N., and Giakoumis, E. G. (2008). “Avail-ability analysis of a syngas fueled spark ignition engine using amulti-zone combustion model.” Energy, 33(9), 1378–1398.
Rakopoulos, C. D., Rakopoulos, D. C., Giakoumis, E. G., and Dimaratos,A. M. (2010a). “Investigation of the combustion of neat cottonseedoil or its neat bio-diesel in a HSDI diesel engine by experimental heatrelease and statistical analyses.” Fuel, 89(12), 3814–3826.
Rakopoulos, C. D., Rakopoulos, D. C., Mavropoulos, G. C., andGiakoumis, E. G. (2004b). “Experimental and theoretical study ofthe short term response temperature transients in the cylinder wallsof a diesel engine at various operating conditions.” Appl. ThermalEng., 24(5–6), 679–702.
Rakopoulos, D. C. (2012). “Heat release analysis of combustion in heavy-duty turbocharged diesel engine operating on blends of diesel fuel withcottonseed or sunflower oils and their bio-diesel.” Fuel, 96, 524–534.
Rakopoulos, D. C. (2013). “Combustion and emissions of cottonseed oiland its bio-diesel in blends with either n-butanol or diethyl ether inHSDI diesel engine.” Fuel, 105, 603–613.
Rakopoulos, D. C., Rakopoulos, C. D., Giakoumis, E. G., and Dimaratos,A. M. (2012). “Characteristics of performance and emissions in high-speed direct injection diesel engine fueled with diethyl ether/diesel fuelblends.” Energy, 43(1), 214–224.
Rakopoulos, D. C., Rakopoulos, C. D., Giakoumis, E. G., and Dimaratos,A. M. (2013). “Studying combustion and cyclic irregularity of diethylether as supplement fuel in diesel engine.” Fuel, 109, 325–335.
Rakopoulos, D. C., Rakopoulos, C. D., Giakoumis, E. G., Dimaratos,A. M., and Founti, M. A. (2011). “Comparative environmental behaviorof bus engine operating on blends of diesel fuel with four straight veg-etable oils of Greek origin: Sunflower, cottonseed, corn and olive.”Fuel, 90(11), 3439–3446.
Rakopoulos, D. C., Rakopoulos, C. D., Hountalas, D. T., Kakaras, E. C.,Giakoumis, E. G., and Papagiannakis, R. G. (2010b). “Investigation ofthe performance and emissions of a bus engine operating on butanol/diesel fuel blends.” Fuel, 89(10), 2781–2790.
Shrimpton, J. S. (2003). “Pulsed charged sprays: Application to DISIengines during early injection.” Int. J. Numer. Methods Eng., 58(3),513–536.
Shrimpton, J. S., and Kourmatzis, A. (2010). “Direct numerical simulationof forced flow dielectric EHD within charge injection atomizers.” IEEETrans. Dielectrics Electr. Insul., 17(6), 1838–1845.
Tang, K., and Gomez, A. (1994). “On the structure of an electrostatic sprayof monodisperse droplets.” Phys. Fluids, 6(7), 2317–2332.
Tang, K. Q., and Gomez, A. (1996). “Monodisperse electrosprays of lowelectric conductivity liquids in the cone-jet mode.” J. Colloid InterfaceSci., 184(2), 500–511.
Tashiro, Y., Shinto, H., Hayashi, M., Baba, S., Kobayashi, G., andSonomoto, K. J. (2007). “Novel high-efficient butanol production frombutyrate by non-growing clostridium saccharoperbutylacetonicumN1–4 (ATCC 13564) with methyl viologen.” J. Biosci. Bioeng.,104(3), 238–240.
Tashiro, Y., Takeda, K., Kobayashi, G., and Sonomoto, K. J. (2005). “Highproduction of acetone–butanol–ethanol with high cell density culture bycell-recycling and bleeding.” J. Biotechnol., 120(2), 197–206.
Thomas, M. E., DiSalvo, R., and Makar, P. (2002). “Electrostatic atomi-zation insertion into compression ignition engines.” SAE Paper No.2002-01-3053, Society of Automotive Engineers, Warrendale, PA.
Vancoillie, J., Demuynck, J., Sileghem, L., Van De Ginste, M., Verhelst, S.(2012). “Comparison of the renewable transportation fuels, hydrogenand methanol formed from hydrogen, with gasoline—Engine efficiencystudy.” Int. J. Hydrogen Energy, 37(12), 9914–9924.
© ASCE A4014008-10 J. Energy Eng.
J. Energy Eng. 2014.140.
Dow
nloa
ded
from
asc
elib
rary
.org
by
Uni
vers
ity o
f Il
linoi
s A
t Urb
ana
on 0
9/07
/14.
Cop
yrig
ht A
SCE
. For
per
sona
l use
onl
y; a
ll ri
ghts
res
erve
d.