A Novel Fuel-Flexible Combustor for Industrial Applications

Paper from the AFRC 2013 conference titled A Novel Fuel-Flexible Combustor for Industrial Applications by Yonas Niguse

In this study, a 60 kWth capacity swirl stabilized fuel flexible combustor utilizing flow-blurring injector is developed for industrial applications, by scaling-up a small scale system that has been produced and experimentally tested by our research group. A major advantage of the novel combustor is that it can efficiently and cleanly burn nontraditional fuels such as glycerol. The combustor development process was conducted in three stages. In the first stage appropriate criteria were selected to design and manufacture the liquid fuel injector, swriler and combustion enclosure. In the second stage, experiments were conducted to investigate combustion performance with diesel and vegetable-oil. Flame photographs and emissions measurements showed robust performance characterized by blue flames, and stable combustion with low emissions. In the third stage, experiments were conducted to investigate combustion of glycerol fuel. Results show a general trend of lower CO and NOx levels at higher ALRs, replicating the performance of the small scale injector. Most importantly, clean combustion of fuels with widely different physical and chemical properties has been demonstrated in the same system without the need for costly upstream fuel-processing steps.

[Type text] *Corresponding Author ([e-mail]) A Novel Fuel-Flexible Combustor for Industrial Applications Yonas G. Niguse, Ajay K. Agrawal, Robert P. Taylor, and William F. Cox Department of Mechanical Engineering The University of Alabama, Tuscaloosa, Alabama 35487, USA Abstract In this study, a 60 kWth capacity swirl stabilized fuel flexible combustor utilizing flow-blurring injector is developed for industrial applications, by scaling-up a small scale system that has been produced and experimentally tested by our research group. A major advantage of the novel combustor is that it can efficiently and cleanly burn nontraditional fuels such as glycerol. The combustor development process was conducted in three stages. In the first stage appropriate criteria were selected to design and manufacture the liquid fuel injector, swriler and combustion enclosure. In the second stage, experiments were conducted to investigate combustion performance with diesel and vegetable-oil. Flame photographs and emissions measurements showed robust performance characterized by blue flames, and stable combustion with low emissions. In the third stage, experiments were conducted to investigate combustion of glycerol fuel. Results show a general trend of lower CO and NOx levels at higher ALRs, replicating the performance of the small scale injector. Most importantly, clean combustion of fuels with widely different physical and chemical properties has been demonstrated in the same system without the need for costly upstream fuel-processing steps. 1. Introduction Fluctuating fuel prices, unabated energy sustainability concerns, and waste energy byproducts generated in industry have created the opportunity to develop fuel flexible combustion systems. A combustion system's capability to handle multiple liquid fuels depends on the fuel injector. Most combustion applications have limited fuel flexibility mainly because of the strong dependence of the injector performance on physical and chemical properties of the fuel. Thus, an ideal fuel injector would perform robustly with minimal dependence on fuel properties. The most common fuel injection techniques are: pressure driven as in direct injection systems, and kinetic energy driven as in twin-fluid atomizers. Less commonly used techniques include centrifugal energy driven atomization as in rotating discs, and effervescent, flashing, electrostatic, vibratory, and ultrasonic atomizers.1,2 Twin-fluid injectors utilize kinetic energy provided by a gas introduced in the injector system, mainly for the purpose of enhancing atomization of the liquid fuel. Air-blast (AB) injector is a typical example of twin fluid atomizers. In AB atomization, atomizing air and liquid are supplied separately to the injector. Air is delivered and swirled on the outer periphery of the injected liquid fuel at a relatively large velocity to break up the fuel jet and to disperse the resulting spray in the combustion zone.3 The primary driving force of liquid break up and droplet formation is by the shear layer formed because of the high relative velocities between the two phases. However, a major shortcoming of this technique is that in highly viscous liquids such as glycerol or straight VO, shear layer instabilities are suppressed, giving rise to less effective droplet break up or larger droplet diameters. Another twin fluid injector is the effervescent atomizer (EA)1. In EA, a pressurized gas is injected into the bulk liquid inside the atomizer body, upstream of the nozzle orifice. Bubbles formed by the injected gas are then expanded rapidly when the two-phase mixture is exposed to low pressure zone at the orifice exit shattering the liquid into droplets. EA is reported to produce a spray with very fine droplets. However, this method has known drawbacks in that the spray angle is usually narrow and atomizing air must be pressurized to the fuel supply pressure.3 Besides, the spray produced can exhibit unsteadiness related to two-phase mixing and flow processes.3 A novel twin-fluid atomization technique which is known as Flow Blurring (FB) atomization was recently proposed by Gañán-Calvo.4 This technique is reported to produce finer droplets with up to fifty times surface area to volume ratio and atomization efficiency of tenfold when compared to AB atomization.5-7 Figure 1 shows schematic illustration of FB atomization's working principle. Atomizing air is forced through a small gap between the exit of fuel tube and a coaxial orifice located at distance ‘H' downstream of the fuel tube. Gañán-Calvo et al4,8 reported that for H/D of 0.25 or less (here D is the orifice diameter), the atomizing air flow turns radially as it enters the gap H and a stagnation point develops somewhere between the exit of fuel tube and the orifice. Thus, the atomizing air flow is bifurcated about the stagnation point, with part of the air being directed upstream into the fuel tube and the rest flowing out through the orifice. The back flow part of the air, that enters into the fuel tube, results in turbulent two-phase mixing with the incoming liquid, characterized by "turbulent inertial cascade mechanics".6 Eddy scales developed in the turbulent mixing play major role in determining spray droplet sizes. Figure 1. Illustration of Flow-Blurring Atomization Figure 2. Schematic of Swirl Stabilized Dual Fuel Burner Recently, our research group has produced and tested a swirl stabilized dual fuel burner of 7 kWth capacity which utilizes the FB injector. The burner can utilize gaseous fuels and liquid separately or both gaseous and liquid fuels at the same time. Figure 2 shows schematic diagram of the burner assembly with components labeled. Details on working mechanism of the system and components are covered in experimental set-up and scaling approach sections of this paper. Initial tests, with diesel and kerosene as fuels, resulted in much lower than normal CO and NOx emissions levels.10-11 Later Benjamin Simmons et al3,6-7,9 showed that FB injector has superior atomization capability in terms of producing smaller droplets when compared to AB, and further demonstrated the injector's capability of burning fuels of very high viscosity, including straight vegetable oil (VO) and glycerol with low emissions. Table 1 compares thermo-physical properties of these very different fuels. Results of spray experiments on the injector conducted by Simmons under different Air to Liquid Ratio (ALR) by mass showed that droplet saunter mean diameter (SMD) decreases with increase in ALR. Fuel flexible, clean combustion in the laboratory scale (7-kWth heat release rate or HRR) has distinct importance for solving some of the environmental and economic concerns associated with waste fuels. For example, currently crude glycerol is generated as byproduct of biodiesel production and it is considered as waste because, despite its significant energy content of 16 MJ/kg, it is very difficult to atomize and burn with traditional injectors and in its crude form, it is of limited use.12 For these findings to be considered a full success, large scale application combustor for industrial applications is very important. The objective of this study is to design and test a scaled-up version of the novel burner system to pave the way to its utilization in industry. A large fuel injector allowing HRR = 60 kW was designed, fabricated and investigated. Detailed experiments were conducted under different ALR and heat release rates. This paper reports a summary of the burner scaling, experimental approach and results, which include flame photographs and emission plots from combustion experiments using diesel, straight VO and glycerol fuels under. Scaling approach, experimental setup and methods, results and discussions are presented next. Table 1. Comparison of Fuel Properties (at 25 oC) Property Diesel Vegetable Oil Glycerol Density [kg/m3], 25°C 834 925 1216 Kinematic Viscosity [mm2/s], 25°C 3.88 53.74 741 Lower Heating Value [MJ/kg] 44.6 37 16 Auto-Ignition Temperature [°C] 260 406 370 Vaporization Temperature [°C] 160-370 327 290 Standard Heat of Vaporization [kJ/kg] 250 410 - 640 974 2. Scaling Approach The original small scale burner was of 7-kW capacity (HRR=7 kW). The scaled-up burner was designed with a rating of 60 kW with liquid fuels. With dual fuels (combined liquid fuel-gaseous fuel operation), the burner can deliver up to 100 kW. This increase in capacity is achieved because the gaseous fuel supply system is independent of the liquid fuel injector design. The burner scaling is based on constant velocity scaling criterion.13 This criterion ensures that the residence time inside the combustion chamber is independent of the HRR. Thus, to keep all flow velocities closer to that of small scale burner, cross sectional areas were increased by an average factor of around 9 from the original design. Thus, all circular diameters were increased by a factor of around 3. In parts, where there is a limit on maximum allowable dimension, care was taken to ensure that the flow velocity did not exceed more than 50% of the original, and that proportionate length was added to counter the effects of increase in velocity. These modifications were implemented on the injector, swirler, dump plane and combustion enclosure. The remaining burner components upstream of the dump plane were mostly unmodified; hence, the mixing tube section was not changed. Figure 3 shows diagram of the burner and injector components that have been affected by the scale up. Table 2 summarizes the dimensions of the small scale and scaled -up systems. The small scale FB injector has orifice and fuel tube diameters (D) of 1.5 mm, and fuel tube tip to orifice spacing (H) of 0.375mm. In the scaled-up version, D and H are increased to 4mm and 1mm, respectively. The outer injector body was left unchanged with some minor fitting changes to accommodate the high flow capacity. The dump-plane diameter of 3.8 cm for the small scale burner was increased to 11 cm for the scaled-up system. Two thirds of the dump plane tube was filled with marbles to diffuse the expanding flow from the smaller diameter mixing tube. Figure 3. Schematic of Scaled-up Components of the Combustor Table 2. Summary of dimensions of small scale and scaled up burner components and injector (the letters in bracket show designations of the parameters in Figure 3) Dimensional parameter Small scale burner (mm) Scaled up burner (mm) Injector Orifice diameter (D) 1.5 4 Fuel tube diameter (D) 1.5 4 Injector's fuel-tube to orifice spacing(H) 0.375 1 Dump plane internal diameter (A) 38 110 Swirler outer diameter(B) 38 110 Combustor internal diameter(C) 100 200 Meshed metal screens were used at both ends of the marble section. The atomizing-air line was elongated to account for the increase in the dump plane tube length. The swirler scaling-up was conducted based on criteria of keeping same vane angle of 30o and outer diameter to hub diameter ratio between 0.4 - 0.6, so that the swirl number will remain around 1.5. Hence, the scaled up swirler was designed with the same swirl angle but a double swirl design as the outer diameter to hub diameter ratio for a single swirl was too large with the scale up. The inner swirler was kept the same as the original swirler, and the outer diameter of external swirler was 11 cm. Figure 4 shows, photographs of the small scale and scaled up swirlers. For combustion enclosure, a cylindrical quartz combustor of 20-cm diameter and 45-cm length is used in contrast to the 10-cm diameter and 45-cm length as well for the small burner. Fig 4. Photograph of the Small (left) and Scaled-up (Right) Inlet Swirlers 3. Experimental Set-up Figure 5 shows schematic of the experimental setup. Air compressor, liquid fuel reservoir and natural gas tanks are the main air and fuel supplies of the system. Compressed air is passed through a dehumidifier to remove most of the moisture present, a pressure regulator to reduce supply pressure to the desired levels, and water traps to eliminate any remnant moisture upstream of the burner system. Then, the air is split into primary air supply and atomizing air supply lines. Natural gas is supplied from high pressure gas storage cylinders. A pressure regulator is used to reduce the gas supply pressure from as high as 20 MPa to around 700 kPa. Natural gas is then passed through the flow bench before entering into the burner mixing tube. A peristaltic pump, with an uncertainty of ± 0.25 % of the reading is used to supply the liquid fuel. Components for meters and controllers are mounted on a single centralized flow bench. Laminar flow elements (LFE) are used to measure the primary air and natural gas flow rates. The atomizing air flow rate is measured and controlled by a Smart Track 2 Series 100 mass flow controller with ± 1% accuracy. The liquid fuel is supplied by a peristaltic pump and passes through a pulse dampener to suppress the pulsating nature of the liquid supplied by the pump. The fuel then passes through a valve and fuel filter. Combustion products were sampled continuously by a quartz probe with its tip tapered to 1 mm ID, mounted on a manual traversing system. The probe was traversed in the radial direction at the combustor exit plane. The gas sample was passed through an ice bath and water traps to remove moisture before reaching the gas analyzer. The dry sample was sent through the Nova gas analysis system with uncertainty of ±2 ppm to measure the concentrations of CO and NOx. The analyzer also measured O2 and CO2 concentrations, which were used to verify the equivalence ratio computed from the measured fuel and air flow rates. The main component of the experimental set up is a dual fuel burner, shown as a subsystem in Figure 5. Primary air refers to air introduced into the system for the sole purpose of oxidizing fuel. Marble balls are filled in the box plenum at the bottom of the burner to ensure uniform flow of air into the combustion zone. Atomizing air is supplied through a dedicated line that is connected to the injector. Atomizing air, in addition to assisting the injection process of liquid fuels, also contributes for around 10 - 30% of combustion air depending on the ALR used. Gaseous fuel is introduced from the side of the burner into the primary air stream and is mixed with primary air in the mixing tube. A swirler with axial curved vane angel of 30 degrees and the swirler number of 1.5 is used to stabilize the flame. A FB injector is used to atomize a liquid fuel supplied by a peristaltic pump. The combustion zone is enclosed in by a heat resistant quartz tubing combustor for diesel and vegetable oil combustion. However, with glycerol, a steel combustion enclosure with a downstream nozzle was used to replicate industrial designs. Figure 5. Schematic of the Experimental Setup Diesel, VO, and glycerol combustion experiments were conducted at different ALRs and heat release rates (HRR). Operation of the dual-fuel burner is started by first burning natural gas, then slowly introducing liquid fuel, and gradually turning off the gaseous fuel, so that liquid fuel combustion sustains. The HRR was controlled by manipulating the fuel flow rate. The equivalence ratio was held uniform at around 0.75 by controlling the total air flow rate which is the summation of the atomizing air and primary air. ALR was varied by changing the proportion of primary and atomizing air while keeping the total air flow rate to be constant. A parametric study of combustion performance was carried out by analyzing visual light flame images, infrared images of combustor wall, and CO and NOx emissions profiles. 4. Results and Discussions 4.1 Combustion of Diesel and Straight Vegetable Oil Flame Images. Figures 6 show photographic flame images of diesel and straight VO combustion for HRR = 54 kW or 90% of the design HRR for ALR between 1.5 and 3.1. The dark zone surrounding the flame shows extent of the combustion enclosure. It is clearly seen in both cases that as ALR increased, the flames became bluer, which is an indication of excellent fuel air premixing and clean combustion. ALR 1.5 ALR 1.9 ALR 2.5 ALR 2.8 ALR 3.1 Figure 6. Photographs of Diesel (top row) and straight VO (bottom row) flames HRR = 54 kW (all dimensions are in cm) For ALR < 2.0, the straight VO tended to show some yellow color regions characteristics of soot when compared to diesel. This result is attributed to the higher heat of vaporization of straight VO compared to diesel, which would require longer residence time to pre-vaporize droplet of a given diameter for the former case. Interestingly, for ALR > 2.5, completely blue flame was achieved even with straight VO. This result agrees with previous findings on combustion with small scale FB injector.5, 10 Thus, for ALRs above a critical value, soot-free combustion is achieved with FB atomization regardless of the fuel type or properties. For lower ALRs, the straight VO flames show a narrow but longer structure compared to diesel flames. Thus, fuel properties do affect the combustion process, however, they don't adversely affect spray droplet sizes in the case of FB injector, which agrees with previous findings on small scale burner.4, 7 Increasing the ALR resulted in fainter blue flames with a slightly longer liftoff distance for both fuels. As a general trend, increasing the ALR has a positive effect on flame quality until negative effect on flame stability at ALR > 3.0. Higher ALR results in higher fuel droplet velocities, which reduces the residence time for pre-vaporization, and thus, combustion is more susceptible to turbulent flow fluctuations at the injector exit. CO and NOx Emissions. Figures 7 show radial profiles of CO emissions at the combustor exit plane for diesel and straight VO for HRR = 54 kW and ALR = 1.9 and 2.8. The CO emissions are quite low (less than 30 ppm), with much lower levels (less than 15 ppm) for the higher ALR case. For diesel, CO values are highest near the center region and lower near the combustor wall. For straight VO, an opposite trend of lower CO values at the center and higher values near is wall is observed. In this experiment, the combustor was not insulated, and thus, the heat loss reduced the product gas temperature near the wall as compared to the center region. For volatile diesel fuel with lower heat of vaporization, good fuel atomization, fuel pre-vaporization, and fuel-air mixing occur closer to the center region, and thus, near the combustor wall, where there is abundant air, low CO is produced. For straight VO, however, larger droplets migrate towards the wall where high temperature is very important for complete combustion with low emissions.14, 15 Nevertheless, regardless of the profile shapes, CO emissions are generally very low for both cases, and the injector performed comparably with the small scale injector.3 Figure 7. Radial Profiles of CO Concentration at ALR =1.9 (left) and ALR =2.8 (right) Radial Position (cm) CO concentration (ppm) -10 -8 -6 -4 -2 0 2 4 6 8 10 0 5 10 15 20 25 30 Diesel VO Radial Position (cm) CO concentration (ppm) -10 -8 -6 -4 -2 0 2 4 6 8 10 0 5 10 15 20 Diesel VO Figure 8 show the corresponding radial profiles of NOx emissions at the combustor exit. For both ALRs, diesel combustion results in lower NOx emissions compared to straight VO. The difference is more evident at the lower ALR. At ALR = 2.8, NOx emissions for diesel are less than 10 ppm while those for straight VO are around 30 ppm. Because of the high heat of vaporization of straight VO, some larger droplets are not fully pre-vaporized upstream of the reaction zone, which results in locally rich combustion with higher temperatures and thus, higher NOx emissions. Still, emissions for both fuels are low, further indicating successful performance of FB atomization technique at higher HRRs. Figure 8. Radial Profiles of NOx concentration at ALR =1.9 (left) and ALR =2.8 (right) 4.2 Combustion of Glycerol Following the successful tests on the scaled-up system with diesel and straight VO, the system was further investigated for operation with glycerol fuel. Glycerol is a very high viscosity and low energy density liquid, which traditionally is not considered as a fuel and is difficult to atomize and burn without preheating or pre-processing using traditional fuel injectors. The combustion enclosure for glycerol is made of steel and thus, infrared imaging was used to measure combustor outer wall temperature to determine the location and extent of the flame. Infrared Images. Figure 9 shows infrared images of combustor wall surface temperature for different ALRs. The high temperature flame zone is located slightly downstream for low (0.93) and high ALR (2.01), while it is closer the injector exit for intermediate ALRs of 1.32 and 1.70. The larger droplets produced for lower ALR = 0.93 will require more time for evaporation prior to combustion, which increases the flame lift. However, for higher ALR = 3.01, the droplets are smaller, but the flame lift increases because of the higher flow velocities at the injector exit. These two opposite effects constrain the optimum range of ALRs for glycerol combustion to between 1.0 and 2.0. Again, this result is consistent with previous findings in the small-scale burner. Radial Position (cm) NOx concentration (ppm) -10 -8 -6 -4 -2 0 2 4 6 8 10 0 20 40 60 80 Diesel VO Radial Position (cm) NOx concentration (ppm) -10 -8 -6 -4 -2 0 2 4 6 8 10 0 20 40 60 80 Diesel VO ALR- 0.93 ALR- 1.32 ALR- 1.70 ALR- 2.01 Figure 9. Infrared images of Combustor Outer Wall for Glycerol Combustion, HRR = 41 kW (All Dimensions are in cm) CO and NOx Emissions. For a given ALR of 1.32, operation with HRR = 60 kW resulted in lower CO levels at the combustor exit, by as much as an order of magnitude, when compared HRR = 41 kW. This result shown in Figure 10 (left) clearly indicates that the burner performs well with low CO emissions when it is operated at full capacity. The reason is that glycerol has high vaporization temperature and high heat of vaporization, and thus, for a given volume of combustion chamber, high HRRs result in locally higher temperatures, assisting fuel evaporation to favor premixed combustion with lower CO levels. Besides, high flow rate of primary air results in stronger swirling and thus, improved fuel-air mixing. The NOx levels increase slightly with increase in HRR as can be shown in Figure 11(right), which is expected because of the local high temperature zones. However, NOx emissions for both cases are rather low; generally within 10 ppm. Figure 10. Radial Profiles of CO and NOx Emissions for Glycerol Combustion Position (cm) CO Concentration (PPM) -6 -4 -2 0 2 4 6 0 200 400 600 800 1000 HRR-60 kW HRR-41 kW Position (cm) NOx Concentration (PPM) -6 -4 -2 0 2 4 6 0 5 10 15 20 HRR-60 kW HRR-41 kW Effect of ALR on emissions was investigated by conducting experiments at ARL = 1.32 and 1.94, with equivalence ratio of 0.75. HRR = 41 kW was used instead of the maximum HRR of 60 kW to extend the limit of ALR with stable combustion. Figure 11 shows radial profile of CO and NOx concentration at the combustor exit. Increase in ALR produces spray with smaller droplets, which decreases the CO levels as expected. ALR has negligible effect on NOx levels which are very low (< 4 ppm) for both cases. Figure 11. Radial Profiles of CO and NOx for Glycerol Combustion at HRR = 41 kW. 5. Flame Stability Considerations Simmons 6,7 have shown that an increase in ALR produces spray with smaller fuel droplets. Without doubt, spray with smaller droplets is desirable to improve fuel evaporation, fuel-air mixing, and combustion. However, additional factors based on fuel properties arise that can limit the maximum ALR without adversely affecting combustion stability. In this aspect, the Lower Heating Value (LHV) of the fuel is an important factor. At any given HRR, the mass flow requirement of fuel and hence, atomizing air for a given ALR increases with decreasing LHV. Figure 12 shows plots of atomizing air flow rate Vs ALR for diesel, straight VO and glycerol. Glycerol, which has the lowest LHV of the three fuels requires much higher atomizing air flow rate at any given ALR and HRR. The injector exit velocity which is proportional to the total mass flow rate (fuel and atomizing air) through the injector has an inverse effect on residence time in the pre-vaporization and reaction zones of the combustor. Low flow velocities associated with high energy density fuels permit operation at high ALR values without adverse effects on flame stability. Thus, diesel and straight VO flames were stable even at the high ARL = 3.1. However, the higher flow rate requirement of fuels with low energy density limits the maximum ALR for stable combustion. For this reason with glycerol, flame stability issues were noticed at ALR > 2.0. Position (cm) CO Concentration (PPM) -6 -4 -2 0 2 4 6 0 200 400 600 800 1000 ALR-1.32 ALR-1.94 Position (cm) NOx Concentration (PPM) -6 -4 -2 0 2 4 6 0 2 4 6 8 10 ALR-1.32 ALR-1.94 Figure 12. Atomizing Air Flow Rate Vs ALR for Different fuels, HRR = 60 kW. 6. Concluding Remarks In this study, a recently developed novel swirl stabilized dual fuel burner of 7 kWth capacity utilizing a FB injector is scaled-up to industry capacity to deliver up to 60 kWth with liquid fuel(s) and 100 kWth with dual fuel(s) operation. The system is experimentally investigated for combustion performance with diesel, straight VO, and glycerol fuels. For ALRs above 2.5 for diesel, and above 2.8 for straight VO, the flames are completely blue and stable, replicating what has been observed in the small scale combustion. Both diesel and straight VO combustion produced very low CO emissions comparable to those in the small scale system. NOx emissions were higher for straight VO compared to those for diesel, possibly because of the high heat of vaporization for the former case, which made it difficult to complete pre-vaporize the fuel before combustion. Generally, the combustor performed comparably with the small scale combustor. Optimal condition for stable flame with lowest emission levels with glycerol was established at ALR of 1.32 and HRR of 60 kW. At this condition, CO levels were less than 70 ppm and NOx levels were less than 15 ppm indicating clean and complete combustion of glycerol, a fuel with extremely unfavorable physical properties including kinematic viscosity, LHV, evaporation temperature, and heat of vaporization. The burner was capable of utilizing three fuels with widely different properties, while producing complete combustion with very low or acceptable emission levels. Following the invention FB injector concept, and its application in small scale burner systems, the present study represents a significant step towards developing a fuel flexible combustor for industrial applications. Air to LiquidmassRatio (ALR) Atomizing Air Flow Rate (SLPM) 1 1.5 2 2.5 3 100 200 300 400 500 600 Diesel Veg Oil Glycerol Acknowledgments This research was supported by the US Department of Energy Award EE0001733. References 1. Lefebvre, A. H, Atomization and sprays, Hemisphere Publishing, NY (1989) 2. Kreith, F., Fluid Mechanics, CRC Press LLC, USA (2000) 3. Simmons, B. M.; Panchasara, H. V.; Agrawal, A. K. Proceedings of ASME Turbo Expo 2009: Power for Land, Sea and Air, GT2009-60239 4. Gañán-Calvo, A. M. Applied Physics Letters 2005, 86, (21), pp. 2141-2142. 5. Simmons, B. M.; Kolhe, P. S.; Taylor, R. P.; Agrawal, A. K. The 2010 Technical Meeting of the Central States Section of The Combustion Institute, Champaign, Illinoi, 2010 6. Simmons, B. M.; Agrawal, A. K. GT2011-46832, Proceedings of ASME Turbo Expo 2011, June 6- 10, 2011, Vancouver, CAN. 7. Simmons, B. M.; Agrawal, A. K. Atomization and Sprays, Vol. 20, No. 9, 2011, pp. 821-825. 8. Llompart, J. R.; Gañán-Calvo; A. M. Physical Review E 77, ( 2008) 036321 9. Simmons, B. M.; Agrawal, A. K. 50th AIAA Aerospace Sciences Meeting, Nashville, Tennessee, 2012-0523 10. (10) Panchasara, H.; Sequera, D. ; Schreiber, W. ; Agrawal, A. K. 5th US Combustion Meeting, 2007, San Diego, C A, Paper G-12-Spray 11. Panchasara, H.; Sequera, D.; Schreiber, W.; Agrawal, A. K., Journal of Propulsion and Power, Vol. 25, No. 4, 2009, pp. 984-987. 12. Bohon, M. D.; Metzger B. A.; Linak, W. P.; King, C. J., Roberts, W. L. Proceedings of the Combustion Institute, Volume 33, Issue 2, 2011, Pages 2717-2724Baukal, C. E. Industrial burners handbook , CRC Press (2004) 13. Terigar, B. G.; Balasubramanian, S.; Lima, M.; Boldor, D. Energy Fuels, 2010 24, 6609-6615 14. Sharma, B. K.; Perez, J. M.; Erhan, S. Z. Energy Fuels, 2007, 21 (4), pp 2408-2414