Experimental Valuation Diagnostics of Hydrous Ethanol Sprays Formed by a Blurry Injector

ABSTRACT: Concerns about the rising fuel price and environmental changes have led to the search for alternative fuels and energy sources. The interest in improving the performance of power generation, with the aim of reducing costs, increasing operating efficiency, and reducing the emissions of pollutants, has driven the scientific community to work on new burning technologies. Flameless combustion is one of the best alternative new technologies for a clean and efficient one. The burning of liquid fuels in power generation and propulsion systems depends on the effective atomization to increase the surface area of the fuel and thus to achieve high rates of mixing and evaporation. This work described the spray characteristics of hydrous ethanol in a blurry injector for applications in a flameless compact combustion chamber. The experimental results are obtained over a range of relatively low flow rates with different air-toliquid mass flow ratios.


INTRODUCTION
Spray combustion is extensively used in power generation and liquid-fueled rocket engines.In general, before burning, liquid fuels need to be dispersed in small droplets that are rapidly vaporized and mixed with the oxidizer.The atomization process increases the surface area of the fuel, aiming at making the contact area between the fuel and oxidizer higher and, therefore, its rates of mixing and fuel evaporation and in the time available for complete combustion.Effective fuel atomization is essential to minimize emissions of particulate matter (PM), carbon monoxide (CO), unburned hydrocarbons (UHC), and nitric oxides (NO x ).
The increasing costs of fossil fuels, environmental concerns, and stringent regulations on fuel emissions have caused a significant interest for the use of biofuels.Ethanol has become an attractive alternative fuel for: it is a renewable energy source, easily available from common biomass sources, biodegradable, contributes to sustainability and is oxygenated, thereby providing the potential to reduce pollutants emissions.Due to its combustion characteristics, it also has been considered as a low polluting liquid propellant for the combustion rocket propulsion application (Gajdeczko et al., 2000).
The most typical mixing twin fluid atomization technique is the air-blast atomization.Air-blast injectors have been widely used and studied (Lefebvre, 1992a, b;Clack et al., 2004;Hoeg et al., 2008;Bolszo and McDonell, 2009;Batarseh et al., 2010).In this technique, air and liquid are supplied separately to the injector, and mixing takes place downstream of the Azevedo, C.G., Andrade, J.C. and Costa, F.S. nozzle orifi ce, externally.Th e liquid discharges through a circular orifi ce, while the air is supplied through an annular slot around the periphery, resulting in a conical discharge pattern.Th e main atomization technique is shear interaction caused by high relative velocities between the air and liquid.A liquid jet is exposed to a stream of air fl owing at high velocities, which impinge on the liquid jet outside the discharge orifi ce, producing threads and ligaments.According to Lefebvre (1989), their initial hydrodynamic instabilities are augmented by aerodynamics disturbances, so that they expand away from the nozzle and their thickness slenderizes.When the ligaments collapse, droplets are produced.According to Lorenzetto and Lefebvre (1977), the air-blast injector produces fi ner droplets as the supply pressure or mass fl ow rate of the atomizing air is higher, which also increases the power requirement of the atomizer.However, the air blast injector performs poorly with fuels of high kinematic viscosity, creating large droplets that burn in diff usion mode to result in high PM, CO, and NO x emissions.
Another typical atomization technique of internal mixing type is known as eff ervescent atomization (Lefebvre, 1988;Lefebvre et al., 1988).A pressurized gas is injected into the bulk liquid in a mixing chamber, upstream of the discharge orifi ce.Th e injected gas forms bubbles to produce a two-phase mixture that fl ows through the orifi ce.Th ey are expanded quickly when the mixture is exposed to a low-pressure zone at the injector exit, shattering the liquid into droplets.Th ere have been many studies reported in the literature involving eff ervescent injectors over a range of air-to-liquid mass and liquid fl ow rates (Lörcher et al., 2005;Konstantinov et al., 2010).According to Sovani et al. (2001), compared with an air blast injector, eff ervescent ones present advantages like the formation of a spray with fi ner droplets over a wide range of operating conditions, even for less refi ned fuels; the injector performance is relatively insensitive to the liquid kinematic viscosity; the larger diameter of the orifi ce alleviates clogging problems and simplifi es fabrication.Gañan-Calvo (2005) describes the fl ow-blurring injector, or blurry injector, a novel twin fl uid atomization technique, which exploits the advantages of internal and external mixes.Th is injection method presents several advantages over other injectors, such as formation of a uniform spray, better atomization, high atomization effi ciency, robustness, excellent fuel vaporization and mixture with air, and potential for the application in compact combustion systems that can be used as portable power sources.Also, for a specifi ed liquid fl ow rate and total energy input, the fl ow-blurring injector creates about 5 to 50 times more droplet surface areas than any other pneumatic injector of the "plain-jet air blast" type.Figure 1 presents the scheme of the fl ow-blurring injector.
Th e fl ow-blurring injector consists of a fuel tube and an exit orifi ce both of diameter (d).Th e concept behind fl owblurring atomization is that the air is forced through a small gap between the fuel tube exit and a coaxial orifi ce located H distance downstream the fuel tube.As shown in Fig. 1, when H/d < 0.25, part of the air is forced a short distance into the fuel tube and the remaining produces shear layer as it leaves the injector orifi ce enhancing the atomization process.Th e back fl ow of air at the tip of the fuel tube results in a twophase turbulent fl ow passing through a positive pressure fi eld.Th is mixture undergoes sudden decrease in pressure, while exiting through the injector orifi ce.Due to the signifi cant pressure decrease, air bubbles in the two-phase fl ow expand and shatter the liquid into fi ne droplets.Th e fl ow-blurring injector is capable of producing internal and external mixes of the two phases simultaneously, providing then superiority over other injectors.injector, using kerosene and diesel burning in a swirl stabilized combustor operated at atmospheric conditions, and verifi ed that for such fuel and atomizing air fl ow rates, the fl ow-blurring injector produced three to fi ve times lower NOx and CO emissions as compared to the air blast injector.Reduction in emissions was attributed to improved fuel atomization that resulted in a decrease in the mean droplet size for the fl ow-blurring injector.
Sadasivuni and Agrawal (2009) used the fl ow-blurring injector in a compact combustion system with a counter fl ow heat exchanger.Th e volumetric energy density of the system was substantially higher than that of the concepts previously developed.Heat release rate of up to 460 W was achieved in a combustor volume of 2.0 cm 3 .Th e combustion system produced clean, compact, quiet, distributed, and attached fl at fl ame.No soot or coking problems were experienced during or aft er combustor operation on kerosene fuel.Simmons and Agrawal (2010) used laser sheet visualization and a phase Doppler particle analyzer to obtain the spray characteristics of a fl ow-blurring injector, operating with a confi guration where H/D=0.23 and using as working fl uids water and air.Th e authors also compared the performance of such injector with that of an air blast and from the results, they concluded that the fl ow-blurring injector can eff ectively atomize liquids at relatively low air-to-liquid mass ratio (ALR) compared to the air-blast injector, while reducing the pressure drop penalty in the atomizing air line.
Rapid fuel vaporization and mixing with oxidizer are key requirements for liquid-fueled small-scale combustion systems.Th us, the optimization of combustion systems is very attractive, since the use of non-renewable liquid hydrocarbon fuels is responsible for most of the energy production and pollutants emissions.Th erefore, improvements in the design and operation of this equipment are essential for current environmental and energy requirements.
Th e fl ow-blurring injector is eff ective in generating a fi ne spray for liquid fuels in mesoscale systems to promote vaporization.Th erefore, this work presents the characterization of hydrous ethanol sprays formed by a blurry injector with a divergent exit.Th e liquid and air mass fl ow rates were measured experimentally and, since lower fl ow rates and pressures were adopted, the injector will be considered for applications in a fl ameless compact combustion chamber.Flameless combustion is a homogeneous low temperature burning process leading to strongly reduced pollutant emissions and higher effi ciency compared to the traditional processes (Wünning and Wünning, 1997).Experiments are conducted for diff erent liquid and air mass fl ow rates at ambient conditions of temperature and pressure.

BLURRy INJeCToR
Figure 2 shows the injector developed that will be possibly used in a fl ameless compact combustor.Th e blurry injector consisted of a central liquid tube (d = 0.5 mm) and a coaxial atomizing air passage with 6 mm inner diameter.Th e two-phase mixture exits through the orifi ce of diameter (d=0.5 mm) in the discharge plate located, such that H=0.125 mm.As discussed, this geometry creates a turbulent mixing between the air and liquid phases at the tip of the liquid supply tube to produce a fi ne spray.

Test bench
Compressed air was used as the atomizing gas and was supplied from a high-pressure cylinder, controlled by a needle valve, and measured by a calibrated fl ow meter with an uncertainty of ± 1.5 standard liters per minute (slpm).Th e fl ow rates of hydrous ethanol were measured by rotameters, with the uncertainty in the measurements being ± 2%.Supply pressure in the fuel and atomizing air lines were measured using pressure transducers at locations depicted in Fig. 3.
Th e average droplet diameters and size distribution of the spray were measured using a laser diff raction system (Malvern Spraytec ® ) at atmospheric conditions.Th e operating principle of this system is the laser scattering produced by the droplets.Th e laser diff raction system can measure droplet diameters from 0.1 to 2,000 µm with accuracy of ± 1% of full scale (specifi ed by the manufacturer).It could measure the droplet size and distribution of sprays with obscurations up to 95% and calculates spray average properties along a sight line across the spray.
Th e laser measurements were taken 50 mm downstream of the injector exit, where the spray drop size was constant further downstream.Th e centre of the spray was positioned at the laser beam centre, so it could be fully covered by the laser beam.
Table 1 shows the properties of the hydrous ethanol.Density ρ, surface tension σ, and dynamic viscosity ν were determined by measurement in laboratory.

RESULTS AND DISCUSSION
Initially, the liquid flow rate was kept constant and the airflow rate was varied to obtain the variation in ALR in the injector.Then, the liquid flow rate was varied for different values of airflow rate.Air density was calculated considering the supply pressure and temperature of the atomizing air.

PReSSURe dATA
Figure 4 shows the pressure in the atomizing air line and the pressure in the hydrous ethanol one for diff erent air fl ow rates.Th e pressure measured was eff ectively that drop in the line because the injector was open to the room.
It can be seen in Fig. 4 that the air and liquid pressures in the injector increase with air fl ow rate being higher.Th e air pressure ranged from 1.02 and 2.88 bar for air fl ow rate from 0.082 to 0.24 g/s, and the liquid pressure varied between 0.94 to 2.34 bar for air fl ow rate from 0.082 to 0.24 g/s.Th e pressure is higher when there is an increase of both air and hydrous ethanol mass fl ow rates.

AIR-To-LIQUId MASS FLow RATIoS
Th e ALR for the operational conditions are depicted in Fig. 5.To obtain the plots in Fig. 5 the liquid fl ow rate was initially kept constant and the air fl ow rate was varied over a range to obtain the variation in ALR.Th e liquid fl ow rate was then varied and the entire procedure was repeated for diff erent values of air fl ow rate.
It is observed in Fig. 5 that for a given liquid fl ow rate an increase in the air one leads to an increase in ALR.Th e data in Fig. 5 also show an increase in ALR with a decrease in the liquid fl ow rate.Th e reason for the increase in ALR can be attributed to the fact that with the decrease in the area occupied by the liquid due to the decrease in its fl ow rate the area available for air fl ow increases, doing the same in the air fl ow rate.For the liquid fl ow rates analyzed, it was verifi ed that the air fl ow rate varied between 0.082 and 0.24 g/s and the ALR was seen changing from 0.21 to 2.88.

dISCHARGe CoeFFICIeNT
Th e discharge coeffi cient is the ratio between the experimental mass fl ow rate and the maximum theoretical mass fl ow rate of the liquid in the injector.It is given by Eq. 1 (Delmeé, 1983): where c d is the discharge coeffi cient of the liquid; m l the experimental liquid mass fl ow rate, kg/s; A is the total cross-sectional area of the discharge orifi ces, m 2 ; ∆P l is the pressure diff erence of the liquid fl ow across the nozzle, Pa; and ρ l is the density of the liquid, kg/m 3 .At each test condition, the discharge coeffi cient was determined by substituting into Eq. 1 the measured values of liquid fl ow rate and pressure drop across the injector, along with injector fl ow area and liquid density.Figure 6 shows the typical curve of the discharge coeffi cients versus ALR.It is seen in Fig. 6 that for a given liquid fl ow rate the discharge coeffi cient decreases with an increase in ALR.Lefebvre (1983) has defi ned the discharge coeffi cient to be a measure of the extent to which the liquid fl owing through the fi nal discharge orifi ce   makes full use of the available fl ow area.Th erefore, the discharge coeffi cient depends on the amount of fl ow area available for the liquid phase.As ALR increases, the fl ow area available for liquid decreases and C d is inferior.Th e rate of change in discharge coeffi cient decreases with an increase in ALR, which is responsible for a slower rate of decrease in the liquid fl ow rate at higher values of ALR as seen in Fig. 6.Th e values of discharge coeffi cient shown in Fig. 6 vary from 0.022 and 0.157 over the entire operating range.

dRoPLeT dIAMeTeR dATA
Diff erent characteristic diameters can be obtained to represent a spray.In this work, the Sauter mean diameter (SMD) and the mass median diameter (MMD) were obtained with aid of the laser system.Th e SMD is the droplet size that possesses a volume-to-surface-area ratio proportional to that of the entire spray, and MMD is the drop diameter such that 50% of the total mass of spray consists of droplets of smaller diameter.
Figure 7 illustrates the eff ect of ALR on the SMD and MMD at diff erent liquid mass fl ow rates for hydrous ethanol.
Th e results show that the droplet size is strongly infl uenced by the ALR.Th e data presented in Fig. 7 conclude that the droplet size decreases with an increase in ALR for a given liquid fl ow rate.It is verifi ed that a decrease in liquid mass fl ow rate causes a decreasing in the mean drop size.Th e higher the ALR is, the higher the air fl ux will be, and then a larger smashing energy can be provided for liquid atomization.It can be speculated that this decrease in the droplet diameter value is due to two eff ects.First, the increase in ALR increases the air fl ow rate and the eff ective area occupied by air, decreasing the eff ective area occupied by liquid and liquid fl ow rate through the injector orifi ce.Increase in air fl ow area is benefi cial to atomization, because it reduces the area available for the liquid fl ow, i.e. it squeezes the liquid into thinner fi lms and ligaments as it fl ows through the injector orifi ce.Secondly, the increase in ALR is accompanied by one in exit velocities and turbulence inside the injector, resulting in improved atomization.
Figure 8 illustrates the eff ects of atomizing air velocity on SMD and MMD at diff erent liquid mass fl ow rates for hydrous ethanol.Table 2 shows the ranges of ALR, air velocity, SMD, and MMD measured.Figure 9 depicts the effects of ALR on cumulative drop size distributions and on representative diameters, Dx10, Dx50 and Dx90, i.e. the drop diameters such that 10, 50 and 90% of total liquid volume are in drops of smaller diameter.
The particle size distribution at a low ALR depicts the presence of larger droplets compared to the case of higher ALR, where the percentage of smaller size droplets have increased significantly, reflecting an improved atomization at higher ALR.As expected, it was verified that an increase in ALR leads to a decrease in droplet diameters, since the increase in air flow results in better atomization.

CONCLUSIONS
A blurry injector has been developed for applications in a compact flameless combustion chamber, and the spray characteristics were obtained for injection of hydrous ethanol.The discharge coefficient is seen to decrease with an increase in ALR, which is attributed to the decrease in an available area for liquid flow with increasing air flow.The average droplet diameters decreased significantly with increasing ALR and air velocity.Air and liquid injection pressures were higher, approximately, linearly with increasing air flow rates.

Figure 2 .
Figure 2. Schematic representation of the blurry injector.

Figure 3 .
Figure 3. Schematic representation of the test bench.
-to-liquid mass ratio.

Figure 7 .
Figure 7. Infl uence of air-to-liquid mass ratio on Sauter mean and mass median diameters.

Figure 8 .
Figure 8. Infl uence of air velocity on Sauter mean and mass median diameters.
*measured at 299.15 K; **measured at 298.15 K.Experimental Valuation Diagnostics of Hydrous Ethanol Sprays Formed By a Blurry Injector

Table 2 .
Ranges of air-to-liquid mass ratio, air velocity and average diameters.