Liquid Monopropellant Combustion

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

JOURNAL OF PROPULSION AND POWER

Vol. 25, No. 3, May–June 2009

Technical Notes
TECHNICAL NOTES are short manuscripts describing new developments or important results of a preliminary nature. These Notes should not exceed 2500
words (where a figure or table counts as 200 words). Following informal review by the Editors, they may be published within a few months of the date of receipt.
Style requirements are the same as for regular contributions (see inside back cover).

Liquid Monopropellant Combustion 10-mm-diam combustor has a film surface area comparable to a
droplet spray with a 10 m Sauter mean radius. The concept using
in Mesoscale Vortex Chamber wall film evaporation in small combustion chambers was demon-
strated in a 3:14-cm3 -volume combustor [6] with gaseous air used as
the oxidizer. Secondary air injection was recently introduced into the
Ming-Hsun Wu∗ concept to enhance mixing and contain the reaction zone completely
National Cheng Kung University, within the combustion chamber [7]. Furthermore, a portion of the
Tainan City 70101, Taiwan, Republic of China flow residence time in the chamber was consumed by the mixing
and between the fuel and oxidizer.
Yanxing Wang,† Richard A. Yetter,‡ and Vigor Yang§ To circumvent the problem of short flow residence time in a small-
scale combustion chamber and thereby avoid the difficulties
Pennsylvania State University, associated with the mixing of fuel and oxidizer, we studied the usage
University Park, Pennsylvania 16802 of nitromethane (CH3 NO2 ), a liquid monopropellant. A mesoscale
DOI: 10.2514/1.40662 vortex combustor developed in our previous work on gaseous com-
bustion [8] was used because of its potential to implement the
concept of wall film vaporization.

I. Introduction
II. Experimental Setup
A WIDE variety of meso- and microscale combustion devices
have been proposed and are under development to meet the
power and energy density requirements of propulsion and power-
The volume of the combustion chamber is 108 mm3 , with the
chamber diameter of 5.16 mm. The injection and the exhaust
generation applications [1–4]. Most of these devices employ gaseous ports both intersect the circumference of the cylindrical com-
fuels and, in particular, hydrogen, which has a fast reaction rate and bustion chamber tangentially, but at opposite ends. The tangential
high diffusivity, to demonstrate the feasibility of self-sustained positioning of the inlet creates a swirling flow that stabilizes the
combustion at millimeter scales or below. Liquid-phase storage and combustion process. Details of the combustor design, fabrication,
operations, however, are essential for taking full advantage of the and gaseous-fuel operation are found in [3,8]. For the present
high specific energy of a combustion-based power generator using operation using a liquid monopropellant, a high-pressure syringe
hydrocarbon fuels. pump (Harvard Instruments, PHD 4400) was used for liquid
At macroscales, liquid fuels are typically sprayed into a com- injection. A 2-mm-thick sapphire window capable of withstanding
bustor in droplet form to enhance the vaporization and ensuing pressure up to 2.758 MPa (gauge) (400 psig) was installed on
burning rates. If the fuel was injected as a film on the chamber wall, the combustor to provide optical access. The window was replaced
the liquid surface area would not be large enough to sustain the with an end cap featuring a thermocouple port when tempera-
needed vaporization rate. In combustors with dimensions in the ture measurements were conducted. The chamber pressure was
subcentimeter-size range, however, because the specific area of monitored by a transducer (Setra Systems, Inc., Model 207)
the wall film increases as the combustor volume decreases, a liquid connected to a port on the side wall of the combustor.
film can offer a surface area for vaporization that is as large as that The exhaust products were diluted with helium (99.999%) and
of a vaporizing spray. The film on the combustor surface also cools then analyzed by an FT-IR spectrometer (ThermoNicolet 670). The
the combustor wall to create a favorably distributed temperature gas cell of the FT-IR apparatus was equipped with ZnSe windows,
profile that prevents heat losses for both endothermic liquid decom- and the optical path was 10 cm. The cell volume was 100 cm3 , and
position reactions and gasification. Heat transferred to the wall, the temperature was kept at 130 C to avoid liquid condensation. The
which has been considered as heat loss in most miniaturized com- resolution of the spectra acquired was 1 cm1 .
bustor applications, can be used for fuel film vaporization. Atomizers
that are capable of producing submicron-scale droplets may,
therefore, not be needed for liquid-fueled combustion in meso-
III. Results and Discussion
and microscale combustors. Sirignano et al. [5,6] concluded that a To investigate film combustion using tangentially injected liquid
fuels in a vortex combustor, studies were initially conducted with
Received 29 August 2008; revision received 6 January 2009; accepted for the combustion chamber operating on liquid nitromethane and a
publication 12 January 2009. Copyright © 2009 by the American Institute of small amount of oxygen at atmospheric pressure (1 atm). Liquid
Aeronautics and Astronautics, Inc. All rights reserved. Copies of this paper nitromethane was injected from a port tangentially intersecting the
may be made for personal or internal use, on condition that the copier pay the chamber wall. A small amount of gaseous oxygen was also injected
$10.00 per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood tangentially in the same direction, but on the opposite side of
Drive, Danvers, MA 01923; include the code 0748-4658/09 $10.00 in the combustor. Ignition was achieved by first igniting a methane–
correspondence with the CCC.

oxygen or hydrogen–oxygen pilot flame. Liquid nitromethane was
Assistant Professor, Department of Mechanical Engineering. Member
then injected, while reducing the gaseous oxygen to a prespecified
AIAA.

Research Associate, Department of Mechanical and Nuclear Engineering.
level. Three different mass loadings of nitromethane (80, 89, and

Professor, Department of Mechanical and Nuclear Engineering. Member 94% by volume) with a constant oxygen flow rate were investigated.
AIAA. The chemical energy imparted to the system was 70, 100, and 160 W
§
Professor, Department of Mechanical and Nuclear Engineering. Fellow with the three nitromethane loadings, respectively. The correspond-
AIAA. ing flame temperature and flow development were determined
829
830 J. PROPULSION, VOL. 25, NO. 3: TECHNICAL NOTES

in the reactant at atmospheric pressure. The ignition delay time was


defined as the time required to reach an increase of 400 K above the
initial temperature, which is set at 1000 K in the present study. It was
obtained from the SENKIN [9] calculation using a nitromethane
chemical-kinetics mechanism that consists of 35 species and 134
gas-phase reactions. The effect of pressure on the ignition delay time
is more significant than that of the addition of oxygen at atmospheric
pressure. At temperatures lower than 1000 K, the acceleration of
nitromethane kinetics by oxygen addition, or an increase in pressure,
is even greater.
This analysis indicates the necessity of operation at an elevated
pressure for the combustion of liquid nitromethane monopropellant
in a small chamber. Therefore, we adapted the mesoscale vortex
combustor by adding a needle valve at the exhaust port of the
combustor to regulate the chamber pressure. Ignition was achieved
by employing a hydrogen–air pilot flame. Liquid nitromethane was
then injected while the backpressure valve was closed to increase
the chamber pressure to over 0.689 MPa (gauge) (100 psig). The
hydrogen flow was shut off once the flame was stabilized. Air was
removed gradually as the flow rate of nitromethane and the chamber
pressure were increased.
The combustion of liquid nitromethane at 2.413 MPa (gauge)
Fig. 1 Combustion of liquid nitromethane with a small amount of
(350 psig) in the 107:8 mm3 vortex chamber is shown in Fig. 3.
oxygen. The emission of the nitromethane flame is a brownish–bright- Liquid nitromethane enters the combustor through the tube on the
white color as compared with the bright-blue emission observed from the right side of the figure at the back end of the combustor. Exhaust
gaseous hydrocarbons. No carbon deposits or soot formation are products exit the chamber tangentially at the front end through the
observed in either case. tube shown on the bottom left side of the figure. The mass flow rate of
nitromethane was 0:021 g=s, which corresponded to a linear liquid
feeding rate of 10:2 cm=s in the injection port. Interestingly, the
primarily by the amount of nitromethane. Figure 1 shows a typical continuous yellowish–bright-white flame that appeared when a small
flame structure, which is dominated by the combustion of nitro- amount of oxygen was added was no longer observed. A very dim,
methane as evidenced by the yellowish–bright-white emission from nearly uniform white emission accompanied by a bright-red–white
the flame. No soot or carbon deposits were observed on the surfaces. emission from small particles (possibly soot), which formed and
Much of the liquid nitromethane appeared to be vaporized shortly rapidly disappeared, was seen. The existence of soot indicates that
after being injected into the chamber. At atmospheric pressure, the temperature in the combustor might be well above 1650 K [10].
combustion with nitromethane alone was found to be unachievable. Liquid was not visually observed within the combustor under
The need for a small amount of oxygen at 1 atm can be understood pressurized monopropellant operations, although it was in the
by comparing the nitromethane chemical kinetics to those of a atmospheric experiments with oxygen. The combustor body showed
hydrocarbon–oxygen system. Unlike hydrocarbon oxidation, which a reddish emission, indicating that the wall temperature may exceed
has significant chain branching due to the presence of molecular the vaporization and decomposition temperatures of nitromethane
oxygen (via H  O2 ! OH  O), no equivalent branching reac- at 374 and 520 K, respectively, at atmospheric pressure. Liquid
tions occur in the nitromethane system. The main source of radicals nitromethane is not likely to exist in the combustor under these
is through pressure-dependent dissociation reactions, which occur conditions. The tangentially injected nitromethane and its ensuing
rather slowly at low pressures. The flow residence time in the mini- vapor and gaseous decomposition product in the vortex chamber
aturized combustion chamber becomes insufficient, and the reactions generated a flowfield similar to that seen with gaseous reactants, in
quench as the heat release rate cannot keep up with the enhanced heat which reactions were sustained and stabilized by three recirculation
loss rate. Nitromethane reactions accelerate dramatically, however, zones [8].
with increasing pressure. The acceleration effect of pressurization is Different flame characteristics in the mesoscale vortex combustor
compared with that of oxygen addition in Fig. 2. The figure shows were observed when pressures were below 2.413 MPa (gauge).
the dependence of the ignition delay time on the pressure of pure Intermittent brownish-white flames (see Fig. 4) were observed more
nitromethane combustion, as well as on the concentration of oxygen

Fig. 3 Visual observation of the combustion of liquid nitromethane in


the 107:8 mm3 vortex combustor. The mass flow rate of the
Fig. 2 Effect of pressurization and oxygen addition at 1 atm on the nitromethane was 0:021 g=s, whereas the operating pressure was
predicted ignition delay time of nitromethane combustion at 1000 K. 2.413 MPa (gauge).
J. PROPULSION, VOL. 25, NO. 3: TECHNICAL NOTES 831

Fig. 4 Visual observation of nitromethane flame at operating pressure


of 2.275 MPa (gauge). Note that the flame was oscillatory and the bright
emission shown in the picture only appeared intermittently. The mass
flow rate of the nitromethane was 0:0127 g=s.

frequently, finally developing into an oscillatorylike mode when the


chamber pressure decreased to 2.068 MPa (gauge) (300 psig). In
comparison, a low-pressure limit of approximately 3.447 MPa
(gauge) (500 psig) was observed for nitromethane burning-rate
measurements in a strand burner [11]. The current results show that
self-sustainable combustion can be achieved in the vortex combustor
at pressures as low as 1.793 MPa (gauge) (260 psig). The extended
regime for stable burning may be attributed to the preheat effects of
both the combustor body and the recirculation zone in the chamber
created by the swirling flow. It is worth noting that the combustion
of a hydroxyl-ammonium-nitrate-based liquid propellant with a
burning rate at least 1-order-of-magnitude larger than nitromethane
was found to be self-sustainable even at atmospheric pressure in the Fig. 5 FT-IR spectrum of liquid nitromethane combustion in the
millimeter-size chamber [3]. The heat release rate of the propellant 107:8 mm3 combustor: a) 1.793, b) 1.999, c) 2.206, and d) 2.413 MPa
(gauge).
is, therefore, an essential factor dictating the operation of small
combustors. It also determines the heat feedback for vaporizing and
decomposing the liquid fuels.
The heat generated from combustion must exceed the heat essential for flame stabilization in the small vortex combustor do
required to evaporate and decompose the liquid nitromethane to not form.
achieve steady combustion, even during the oscillatory combustion Absorption spectroscopy measurements of the combustion
of nitromethane at lower pressures. The enthalpy of vaporization of product were performed to characterize the high-pressure nitro-
liquid nitromethane is approximately 5% of the enthalpy of reaction. methane combustion. Figure 5 shows the FT-IR spectra acquired at
The amount of energy that is available to contribute to vaporization steady-state operation with chamber pressures of 1.793, 1.999,
and decomposition, however, is largely determined by the combustor 2.206, and 2.413 MPa (gauge). The mass flow rate was maintained at
configuration, as well as the temperature distributions in both the 0:02 g=s. Various hydrocarbon compounds, such as CH4 , C2 H2 , and
reaction zone and the combustor body. The fluctuation of the flame C2 H4 , are found in the exhaust gas at the lower pressures of 1.793 and
zone at low injection flow rates may be the result of low thermal 1.999 MPa (gauge) (Figs. 5a and 5b) along with the major products,
feedback, which is insufficient to generate a steady flow of decom- CO, CO2 , and water vapor. The existence of unburned hydrocarbons
posed gaseous nitromethane and, consequently, causes the reaction indicates incomplete combustion. Nitromethane was not found in the
to quench intermittently. Pressure fluctuations were recorded when exhaust gas. The unburned hydrocarbons were almost undetectable
the flame was in the oscillatory mode. At high input flow rates and as the pressure exceeded 2.413 MPa (gauge) (350 psig) (Fig. 5d).
low operating pressures, the residence time becomes too short for Carbon in nitromethane was converted more efficiently to CO and
complete combustion, thereby resulting in less energy generation CO2 at higher pressures. Nitric oxide was not found in the spectra,
and leading to effects similar to those already discussed. In the cases which indicates that most of the nitrogen in the nitromethane was
in which the propellant enters the combustion chamber in the liquid converted into molecular nitrogen, which is undetectable using
phase, flame stabilization may become challenging, as the induced infrared–red spectrometry due to the lack of a dipole in the diatomic
flow from the vaporization of the liquid propellant in the combustion molecule.
chamber can alter the flow patterns. Our numerical analysis [12]
revealed that, although the liquid flow travels circumferentially along
the wall, due to the tangential injection, the sudden expansion of the IV. Conclusions
fluid with vaporization causes the vapor to propagate normal to the Self-sustained liquid monopropellant combustion in a cylindrical
interface. Consequently, all the streamlines are aligned generally volume of 108 mm3 (5.16 mm in diameter) was successfully
along the radii due to the relatively larger radial velocity induced by demonstrated and investigated. Pressurization of the combustion
liquid vaporization compared with the tangential injection velocity chamber was used to enhance the decomposition rate of several
of the liquid monopropellant. The recirculation vortices known to be pressure-dependent reactions to yield the necessary radical
832 J. PROPULSION, VOL. 25, NO. 3: TECHNICAL NOTES

generation and to increase the flow residence time, so that pure [3] Wu, M.-H., “Development and Experimental Analyses of Meso and
nitromethane operation in the miniaturized combustor was achieved. Micro Scale Combustion Systems,” Ph.D. Dissertation, Dept. of
Combustion of liquid nitromethane monopropellant at atmo- Mechanical and Nuclear Engineering, Pennsylvania State Univ.,
University Park, PA, 2007.
spheric pressure was not achievable because the flow residence
[4] Wu, M. H., and Yetter, R. A., “A Novel Electrolytic Ignition
time was so short that the slow reaction was quenched. Oxygen was Monopropellant Microthruster Based on Low Temperature Co-Fired
required to enhance the chemical kinetic rates of nitromethane Ceramic Tape Technology,” Lab on a Chip, 2009.
reactions to achieve self-sustained combustion at atmospheric doi:10.1039/b812737a.
pressure. Although flow residence time can also be increased by [5] Sirignano, W. A., Stanchi, S., and Imaoka, R., “Linear Analysis of a
lowering the injection flow rate of nitromethane, the lower chemical Liquid-Film Combustor,” Journal of Propulsion and Power, Vol. 21,
input power might result in insufficient heat feedback to evaporate No. 6, 2005, pp. 1075–1091.
liquid nitromethane and decompose the vaporized reactant, resulting doi:10.2514/1.14156
in an oscillatory flame pattern. The study showed that stable nitro- [6] Sirignano, W. A., Pham, T. K., and Dunn-Rankin, D., “Miniature-Scale
methane combustion could be reached as the chamber pressure was Liquid-Fuel-Film Combustor,” Proceedings of the Combustion
Institute, Vol. 29, 2002, pp. 925–931.
increased above 1.793 MPa (gauge); however, the FT-IR spectro- doi:10.1016/S1540-7489(02)80117-1
scopic measurements revealed that estimated chemical efficiencies [7] Mattioli, R., Pham, T. K., and Dunn-Rankin, D., “Secondary Air
of over 99% were not obtained until pressurization exceeded Injection in Miniature Liquid Fuel Film Combustors,” Proceedings of
2.068 MPa (gauge). the Combustion Institute, Vol. 32, No. 2,2009, pp. 3091–3098.
doi:10.1016/j.proci.2008.06.174.
[8] Wu, M.-H., Wang, Y., Yang, V., and Yetter, R. A., “Combustion in
Acknowledgments Meso-Scale Vortex Chambers,” Proceedings of the Combustion
This work was supported by the U.S. Air Force Office of Scientific Institute, Vol. 31, 2007, pp. 3235–3242.
Research under contract AFOSR FA9550-06-1-0183. The authors doi:10.1016/j.proci.2006.08.114
gratefully acknowledge the support of Mitat Birkan, contract [9] Kee, R. J., Rupley, F. M., Meeks, E., and Miller, J. A., “Chemkin III: A
monitor for the program. Ming-Hsun Wu would also like to Fortran Chemical Kinetcs Package for the Analysis of Gas-Phase
acknowledge support from the National Science Council, Chemical and Plasma Kinetics,” Sandia National Laboratories,
UC-405SAND96-8216, 1996.
Taiwan, under grant NSC 96-2218-E-006-293 and NSC 97-2221- [10] Glassman, I., Combustion, 3rd ed., Academic Press, San Diego, CA,
E-006-058-MY2. 1996.
[11] Boyer, E., and Kuo, K. K., “Modeling of Nitromethane Flame Structure
References and Burning Behavior,” Proceedings of the Combustion Institute,
Vol. 31, 2007, pp. 2045–2053.
[1] Fernandez-Pello, A. C., “Micropower Generation Using Combustion: doi:10.1016/j.proci.2006.07.025
Issues and Approaches,” Proceedings of the Combustion Institute, [12] Wang, Y., “A Numerical Study of Combustion in Meso-Scale Vortex
Vol. 29, 2002, pp. 883–899. Chambers,” Ph.D. Dissertation, Dept. of Mechanical and Nuclear
doi:10.1016/S1540-7489(02)80113-4 Engineering, Pennsylvania State Univ., University Park, PA, 2006.
[2] Dunn-Rankin, D., Leal, E. M., and Walther, D. C., “Personal Power
Systems,” Progress in Energy and Combustion Science, Vol. 31,
No. 5–6, 2005, pp. 422–465. C. Segal
doi:10.1016/j.pecs.2005.04.001 Associate Editor

You might also like