DEVELOPMENT OF NON-TOXIC HYPERGOLIC MISCIBLE FUELS (NHMFs) FOR HOMOGENEOUS DECOMPOSITION OF ROCKET GRADE HYDROGEN PEROXIDE (RGHP)

DEVELOPMENT OF NON-TOXIC HYPERGOLIC MISCIBLE FUELS (NHMFs) FOR HOMOGENEOUS DECOMPOSITION OF ROCKET GRADE HYDROGEN PEROXIDE (RGHP)

B. M. Lormand and N. L. Purcell
Energy Conversion Office
Naval Air Warfare Center Weapons Division
China Lake, California

J. J. Rusek
School of Aeronautics & Astronautics
Purdue University

ABSTRACT

A newly discovered non-toxic hypergolic miscible fuel (NHMF), developed by the Naval Air Warfare Center Weapons Division (NAWCWD) at China Lake, California, was tested at the Mojave test site of the HMX Corporation in a Navy developed 300-lbf bipropellant engine using 98% rocket grade hydrogen peroxide (RGHP) as the oxidizer. The synthesis of the fuel, purification of the RGHP, and initial hypergolic testing were conducted at NAWCWD. These engine tests represent a continuation of research sponsored by the Ballistic Missile Defense Organization (BMDO), as part of the Navy Theater Ballistic Missile Defense (TBMD) initiative.

The novel fuels are eminently polar solutions in their initial state, and they subsequently transform to a stable colloid, without indication of agglomeration, after 18 months of storage. The fuels are relatively non-toxic, economical1, renewable, and, most importantly, appear to have ignition delays comparable to those of nitrogen tetroxide (NTO) and monomethyl hydrazine (MMH) systems. The density specific impulse values of new formulations exceed those of NTO/MMH. These candidate fuels, in concert with RGHP, will enable divert/attitude control systems, orbit transfer, and large launch vehicle applications.

BACKGROUND

The Naval Air Warfare Center has had a rich history of non-toxic propulsion over the past six years.1,2,3,4 The first experiments used kerosene-based fuels and lower strength hydrogen peroxide with mild ignition sources. Results were encouraging, and pointed the way to miscible fuels, that is, those that would mix in all proportions with hydrogen peroxide, if the catalyst was not present. This approach was chosen to reduce the potential of hard-starts and the dependency on non-renewable resources.

The next significant event was the discovery of miscible transition metal catalysts that could be effectively dissolved in the alcohol-based fuels. Ultimately, the transition metals convert to stable oxides, which disperse as stable colloids after a short aging time. It was found that hypergolic ignition delay improved with age. Typical values were 15 milliseconds initial delay to 12 milliseconds after asymptotic aging when measured in a velocity impact probe.

The last component was the discovery of synergistic sensitizers that had the effect of coupling the catalytic decomposition reaction to the combustion reaction. A beneficial side effect of the sensitizers is that they increase the local pH of the fuel, destabilizing the hydrogen peroxide even further, which then further reduces the ignition delay.

INTRODUCTION

Navy Theater Ballistic Missile Defense (TBMD) missions place extreme demands on kill vehicle propulsion systems. The engines must be fast acting, reliable, and precisely controllable; they also must accommodate location errors, thereby guiding the interceptor to a successful direct hit at the defined point on the target. Trade-off studies2 show that liquid propulsion systems provide the operating characteristics needed to accomplish the mission as described. Most importantly, they provide energy on demand, fast response, and fast action with high effective propulsion energy. Traditionally, conventional hypergolic bipropellants with these characteristics would have been chosen; however, Navy TBMD requirements also include the necessity to meet shipboard safety and environmental regulations. Conventional hypergoles are extremely toxic, cause irreversible physical damage, and are proven or suspected cancer agents. Because of this evidence, the logistic and operational impacts preclude their use.

The results of the past six years of propellant development at China Lake has led to a viable non-toxic hypergolic alternative. The development of an effective Cooperative Research and Development Agreement (CRADA) with HMX Corporation and NAWCWD to jointly test the propellant formulations has further confirmed that the NHMF will work for its intended purpose.

CHEMISTRY

Because of its high energy, relatively innocuous decomposition products, and ease of disposal, rocket grade hydrogen peroxide (RGHP) was a logical choice as an oxidizer. The idea for a non-toxic hypergolic miscible fuel (NHMF) came about by looking at the previously known instability of RGHP and transition metals, which involved RGHP decomposition rates varying from slow to very rapid. Combining a rapidly catalytic transition metal species with a combustible, low-toxicity organic solvent is the key to NHMF, while maintaining a locally high pH value.

Specifically, a NHMF is comprised of three parts: a colloidal transition metal oxide, a low-molecular weight alcohol, and a sensitizing agent. The metal oxide rapidly decomposes RGHP into hot molecular oxygen, steam, and heat. The heat and oxygen that are generated rapidly combust the alcohol while the sensitizer destabilizes the hydrogen peroxide and provides additional fuel combustion value. The sensitizer is added to modify the local pH and to increase the polarity of the alcohol, which makes larger portions of the metal catalyst soluble.

For the optimization of the NHMF/RGHP system, many chemical parameters have been taken into account. These include catalyst concentration, sensitizer species type and concentration, and alcohol type. These parameters were analyzed and compared by theoretical energy calculations, drop test hypergolicity experiments, and drop test ignition delays. The ignition delays were quantified by using a China Lake developed velocity impact probe. The varieties of NHMF formulations that were studied are shown in Table 1.

TABLE 1. NHMF Formulations Studied.

Formulation	Wt% Catalyst	Wt% Fuel	Wt% Sensitizing Agent
SS	30.0	70.0 MeOH	Inherent to Catalyst
Block 0	22.3	77.7 MeOH	Inherent to Catalyst
49B	17.8	82.2 MeOH	Inherent to Catalyst
49C	13.4	86.6 MeOH	Inherent to Catalyst
49D	8.9	91.1 MeOH	Inherent to Catalyst
61A	8.9	40.0 MeOH 40.0 1-BuOH	11.1
61C	8.9	40.0 MeOH 40.0 2-BuOH	11.1

LABORATORY TEST PROCEDURES

Theoretical Calculations
Theoretical calculations of all thermodynamic rocket performance parameters were performed using the Air Force Astronautics Laboratory (AFAL) specific impulse (Isp) program. Density Isp values take into account laboratory-measured fuel densities and literature oxidizer density values.

Hypergolicity and Ignition Delay Tests
The hypergolicity and ignition delays of NHMF formulations were obtained with as much consistency as possible. Parameters such as drop height, crucible size, and fuel and oxidizer portion were kept constant to minimize the dynamical variations of the test results. All RGHP solutions were made from distilled water and a supply of 98% RGHP with the inhibitor loading shown in Table 2.

TABLE 2. Inhibitor Loading.

Inhibitor	Concentration, ppm
PO₄^2-	0.95
Sn²⁺	2.95

Inhibitors were analyzed via inductively coupled plasma-atomic emission spectroscopy. The 98% RGHP was created from a stock 85% H2O2 solution by distillation at a vacuum of 8 torr and 500C and stored in amber glass bottles pickled in dilute nitric acid followed by distilled water rinses.

NHMF formulations were prepared by combining all components in small glass vials and were shaken to dissolve all components. All solutions were prepared, stored, and tested at ambient conditions.

Hypergolicity and ignition delay data were collected for all samples by adding one drop of RGHP to two drops of fuel, with drops being administered by standard plastic transfer pipette.

LABORATORY RESULTS

Figure 1 shows specific impulse as a function of oxidizer to fuel ratio (O/F) for the methanol-NHMF system. As can be seen, the peak value of Isp occurs for the 49D system, the lowest catalyst concentration, at an O/F of 3, on a mass basis. This and subsequent Isp analyses used chamber pressures of 1000 psia expanded to 13.2 psia. Lastly, the RGHP concentration used for these studies was 98% by weight.

FIGURE 1. Theoretical Isp Comparisons of Methanol-Based NHMF/98% RGHP
(1000 psi expanded to 13.2 psi).

Figure 2 depicts specific impulse as a function of O/F for the methanol-butanol system. The peak Isp values are similar, but it should be noted that the addition of the branched alcohol raises the O/F peak to about 4, while the straight-chain alcohol raises the O/F peak to about 4.75.

FIGURE 2. Theoretical Isp Comparisons of Methanol-Butanol
Based NHMF/98% RGHP (1000 psi expanded to 13.2 psi).

Figure 3 shows density Isp as a function of O/F for the methanol-NHMF systems. The peak value occurs again at the lowest catalyst loading and the optimum O/F is invariant of catalyst concentration.

FIGURE 3. Theoretical Density Isp Comparisons of Methanol-Based
NHMF/98% RGHP (1000 psi expanded to 13.2 psi).

Figure 4 depicts density Isp as a function of O/F for the methanol-butanol system. A gain of about 5 g-s/cm3 is realized by the use of the branched alcohol.

FIGURE 4. Theoretical Density Isp Comparisons of Methanol-Butanol-Based
NHMF/98%RGHP (1000 psi expanded to 13.2 psi).

Figures 5 and 6 show Isp and density Isp, respectively, as a function of O/F for the methanol Block 0 NHMF, branched butanol NHMF, as compared to conventional toxic hypergolic systems. These figures show the peak NHMF Isp to be about 93% of the peak of the conventional hypergolic system. The figures also show the peak NHMF density Isp to be about 102% of the peak of the conventional hypergolic system.

FIGURE 5. Isp Comparisons of Hypergolic Bipropellant Systems (1000 psi expanded to 13.2 psi).

FIGURE 6. Density Isp Comparisons of Hypergolic Bipropellant Systems (1000 psi expanded to 13.2 psi).

Hypergolicity and Ignition Delay Tests
Hypergolicity and ignition delay information for all tested NHMF formulations can be found in Table 3. Critical hypergolic concentration (CHC) is the minimum concentration of one drop of H2O2 that will cause a hypergolic ignition with two drops of fuel in the velocity impact probe. Ignition delay measurements are a rough measurement of system reactivity only. They have an as-yet unknown correlation to rocket motor ignition delays. Figure 7 depicts high-speed photography (1000 frames/s) taken of a hypergolicity test. The ignition delays of 10 milliseconds for the SS formulation and 12 milliseconds for Block 0 were confirmed by photography.

TABLE 1. Reference Values.

Formulation	CHC (%H₂O₂)	Ignition Delay w/ 98% H₂O₂
SS	95%	10 ms
Block 0	97%	12 ms
49B	97%	15 ms
49C	98%	17 ms
49D	98%	23 ms
61A	95%	21 ms
61C	95%	30 ms

FIGURE 7. High-Speed Photographs of Hypergolicity. THREE-HUNDRED POUND-FORCE THRUST SCALE-UP TESTS

The NHMF formulations were initially tested in a 10-lbf thrust test motor using 98% and 95% RGHP oxidizer. These tests were very successful; the results from these tests indicated fast ignition and smooth combustion with approximately 150-psia chamber pressure. Measured parameters for these initial tests included chamber pressure, oxidizer injector pressure, and fuel injector pressure. In the next step, the investigators scaled up the thrust level and more thoroughly instrumented the motor to obtain more accurate performance measurements.

Engine Hardware
The thruster design chosen for these tests was an existing heavyweight test motor used on a previous program.3,4 The motor consists of a thick-walled steel chamber and an aluminum injector that attaches with a custom Marmon clamp device. A picture of this device is shown as Figure 8.

FIGURE 8. Chamber and Aluminum Injector.

The chamber was modified to include a pressure tap and O-rings, and a nozzle and chamber liner were fabricated. The injector is of a splash plate design consisting of 15 oxidizer jets and 15 fuel jets arranged in a circle, which alternate along the circumference. The oxidizer and fuel streams are oriented axially and parallel to each other until they impinge on a 45-degree conical splash plate. At this point, they "fan" into each other as thin sheets and re-converge outside of the splash plate. Figure 9 shows the injector/splash plate hardware.

FIGURE 9. Injector/Splash Plate Hardware.

Test Site
The rocket engine firings were conducted at HMX's test facility located at the Mojave Airport, California. The testing was conducted under a CRADA between NAWCWD and HMX, Incorporated.

The test cart is shown as Figure 10. It consists of propellant tanks, lines, valves, and transducers. These items and the data acquisition system were supplied by HMX, while the rocket engine, propellants, and flow meters in Tests 3, 4, and 5 were supplied by NAWCWD.

FIGURE 10. Test Cart.

ROCKET ENGINE TESTING

The first two tests were "fire for function" tests to determine if the engine would ignite in a well-behaved manner before expensive load cells were installed on the test cart. For this reason, the only data taken on the first two tests were chamber pressure, oxidizer and fuel manifold pressure, and oxidizer and fuel tank pressure. Flow meters had not been installed; therefore, in order to measure flow rates, the injector was calibrated on a flow bench with an accurate turbine flow meter using water, and the coefficients of discharge for the oxidizer and fuel orifices were determined.

Test 1: 27 May 1998
Because the investigators involved in the study had little experience on the ignition characteristics of this fuel in a motor of the selected size and configuration, they thought it wise to start slow. The design chamber pressure for these tests was to be 600 psia at a total propellant flow rate of 1.3 lb/s and an oxidizer-to-fuel ratio of 2.3 to 1. However, the investigators decided to open the nozzle throat up and burn the fuel at a lower pressure initially (300-400 psia). In addition, for the first test, the nominal flow rate was reduced to approximately 1.0 lb/s by reducing tank pressures. The propellant flow control valves used were air-actuated ball valves with the flow restricted to the air actuators to cause the propellant valves to open slower than normal. The first propellant tested contained 98.1% RGHP oxidizer and a NHMF fuel designated as SS fuel. The oxidizer density was measured to be 0.052 lb/in3 and the fuel density was 0.034 lb/in3. The assumed delivered characteristic velocity, C*, (theoretical C* x 0.92) estimated from the AFAL thermochemical code for this propellant class, was 4832 ft/s (approximation for a starting point only). Unfortunately, the optimum O/F ratio was mistakenly thought to be 2.3 to 1. Later, more extensive thermochemical calculations indicated an optimum of 2.5 to 1. The O/F ratio the investigators targeted on this test, however, was 2.3 to 1; therefore, these firings were fuel rich. To obtain the desired oxidizer and fuel injector pressures, HMX personnel assumed values for feed line losses based on their experience with this system.

The ignition start-up was smooth, judging from sight and sound. Figure 11 shows the firing. Unfortunately, the data acquisition system used for this test was highly filtered and had a slow enough response so no ignition transients could be observed from the data, although steady-state information was there. The motor burned for 2 seconds and, during this time, no sensible instabilities could be detected. The nozzle throat diameter was 0.825 inch and the expansion ratio was 3.899 for this hardware. Also, the motor characteristic length (L*) with the larger nozzle throat was approximately 33.6 inches. The predicted performance for this test was

mdottot = 1.003 lb/s
O/F = 2.3
C* = 4832 ft/s
PC = 281.8 psia (assuming Pa = 13.2 psia)
F = 203.6 lbf (assuming thrust correction factor = 0.95)
Isp = 203 s (delivered)

The actual performance, either directly measured or calculated from measured data, was:

mdottot = 1.047 lb/s
O/F = 2.492
C* = 4737 ft/s
PC = 288.2 psia
F = 208.3 lbf
Isp = 199 s

FIGURE 11. Motor Firing.

Test 2: 27 May 1998
Test 2 was conducted within 15 minutes of Test 1; therefore, the temperature conditions were the same and the propellants were identical, simply the residual left in the tanks after the first firing. The purpose of this test was to increase the chamber pressure to between 300 and 400 psia while maintaining an O/F ratio of 2.3 to 1. The predicted performance values were

mdottot = 1.3 lb/s
O/F = 2.3
C* = 4832 ft/s
PC = 365.5 psia
F = 273.8 lbf
Isp = 210.6 s

The actual performance calculated from measured pressures was:

mdottot = 1.305 lb/s
O/F = 2.287
C* = 4866 ft/s
PC = 369.2 psia
F = 274.9 lbf
Isp = 210.7 s

Test 3: 23 July 1998
The purpose of this test was to measure the performance of the Block 0 fuel with 98% RGHP at the higher pressure of 600 psia. A full set of instrumentation was planned for this test. Tank pressures, oxidizer and fuel injector pressures, chamber pressure, oxidizer and fuel flow rates, and thrust would be measured on two different data acquisition systems. The slow (filtered) data acquisition system would record tank pressures and injector pressures, while the faster system would record chamber pressure, flow rates, and thrust.

A smaller nozzle throat that measured 0.6569 inch was installed. The expansion ratio (Ae/At) measured 6.698. The smaller nozzle throat raised the L* value to approximately 53 inches. The RGHP oxidizer was at a concentration of 98.6%. For Block 0 fuel and 98% RGHP, the O/F ratio for maximum Isp should be approximately 2.75 to 1. This was to be an O/F sweep starting at O/F = 2.3.

The motor fired well, but the instrumentation failed to record flow rates and thrust. In addition, the pressure data channel calibrations were accidentally changed without the knowledge of the instrumentation engineer. Some of the data were later retrieved (chamber pressure and injector pressures). With these data, it is possible to estimate the performance for this firing, although with lower confidence than with a complete set of data.

The predicted performance for this test was

mdottot = 1.35 lb/s
O/F = 2.3
C* = 4866 ft/s
PC = 602.7 psia
F = 300.0 lbf (assuming thrust correction factor = 0.95)
Isp = 222.2 s (delivered)

The measured fuel injection pressure appeared to be in line with the predicted value; however, the oxidizer injection pressure was low. This gives rise to performance numbers that are off. If one assumes that the predicted oxidizer injection pressure is correct and that the measured data were due to a calibration error, the calculated performance values are

mdottot = 1.358 lb/s
O/F = 2.26
C* = 4890 ft/s
PC = 622 psia (measured)
F = 309.6 lbf
Isp = 228 s

Because there is no way to determine exactly where the calibration error was, these data are speculative and of low confidence.

Test 4: 23 August 1998
This test was a re-attempt of Test 3, but with a redesigned load cell attachment and a better understanding of how to use the flow meters. The fuel was the same Block 0 fuel of the previous test, but the RGHP's concentration was lower (approximately 97.6%), probably due to dust contamination from use in the previous tests.

Upon firing, the motor immediately blew the oxidizer manifold inlet cap off at the weld. The pressure data showed that the oxidizer injector manifold had almost twice the pressure of the fuel manifold. The data also showed a low level fuel injector pressure several tenths of a second before the oxidizer pressure came up strongly. This fact indicates the possibility of fuel contamination in the oxidizer injector manifold. The problem will be overcome in the future with judicial use of a nitrogen purge just before the "fire" command.

Test 5: 27 October 1998
This was a re-attempt of Test 4 and was intended to measure the performance of Block 0 fuel with 98% RGHP oxidizer using full instrumentation. The fuel was left over from the previous test and was, therefore, identical to the previous Block 0 fuel. The RGHP's concentration, however, had degraded from 98+% to 95.4%. It was noticed that the RGHP container had not been shipped in the specified upright position and some had leaked out through the vent cap on the bottle. It is possible that contamination had also found its way into the bottle through the vent cap during shipping. The investigators had no choice but to test with the 95.4% RGHP. This test would also give a data point on ignition with lower concentration RGHP.

The investigators fired the motor five times in an attempt to vary the O/F ratio from 2.0 to 3.0. The data recovered, however, again exhibited problems that made it impossible to determine with any degree of confidence what the absolute performance of the propellant was. However, some interesting observations concerning these firings can be made. First, based on previous experience with this system, the propellant tank pressures were set to give a motor operating pressure of approximately 600 psia and 300-lbf thrust when a C* value of 4866 ft/s (delivered) was assumed. While the investigators cannot determine exactly what the mass flow rates or the O/F ratios were, all firings measured over 600 psig and over 300 lbf of thrust. It seems likely that the specific impulse was somewhere in the range of 210-235 lbf-s/lbm. In addition, after the first 0.75-second firing, which had a rough start with a chamber pressure spike of 800 psia, subsequent firings had very fast, smooth starts without pressure spikes. The initial rough start was probably due to the fact that the lines had not been adequately bled. Because the propellant valves were made to open slowly, it is impossible to determine ignition response time accurately; however, there is no indication from the pressure data that any significant ignition delay occurred. It should be noted that these data was obtained with 95.4% RGHP, and that the propellants were at a temperature of 58 to 590F. It should also be pointed out that the phenolic liner in this motor started out at half the thickness of the previous motor tests, and that it was completely gone when the motor was disassembled. Therefore, the motor L* ranged from 65 to 68 inches during this test series. The motor hardware was observed to be much cleaner than in the previous lower pressure firings. There was essentially no residue build-up on the motor hardware; however, significant nozzle erosion of about 0.0059 in/s was observed on the radius. No combustion instabilities of any kind were noted within the range of the instrumentation. Based on these and previous firings, these propellants appear to be excellent candidates for reduced toxicity hypergolic applications.

SUMMARY AND CONCLUSION

Government and Industry investigators jointly conducted rocket engine firings of newly-developed non-toxic bipropellants and hardware. The following conclusions were made concerning the propellants and their testing.

Specific impulse is maximized for a given catalyst, sensitizer, and fuel combination, by minimizing the catalyst concentration; however, ignition delay is sacrificed, to some extent.
The O/F can be varied, within bounds, by varying alcohol isomers within the fuel.
For a given chemical system, the optimum O/F is invariant of catalyst concentration.
Density Isp can be increased by the inclusion of branched isomers.
Current NHMF/RGHP combinations have roughly 93% of the specific impulse of conventional hypergoles.
Current NHMF/RGHP combinations have roughly 102% of the density Isp of conventional hypergoles.
NHMF/RGHP bipropellants exhibit stable combustion and ignition in 300-lbf engine configurations.
NHMF/RGHP bipropellants burn cleanly; however, nozzle erosion due to refractory particles needs to be explored.
L* values were very high due to the constraints of the current engine hardware. Lower and limiting values of L* need to be explored with variable geometry hardware.
Tests conducted to date show exceptionally good ignition delay values, density Isp values, and stability of propellants. Work needs to be continued on this critical technology.

ACKNOWLEDGMENTS

The authors wish to gratefully acknowledge Major Buford Shipley and Major John Kusnierek of the U.S. Air Force USAF and Jim Kiessling of BMDO for having faith and for sponsoring this vital effort.

REFERENCES

J. J. Rusek, M. K. Minthorn, N. L. Purcell, T. C. Pavia, J. R. Grote, G. C. Hudson, and B. McKinney, "Non-toxic Homogeneous Miscible Fuel (NHMF) Development of Hypergolic Bipropellant Engines," Presented at the 6th Annual AIAA BMDO Interceptor Technology Conference, August 1997, San Diego, CA. Paper UNCLASSIFIED.
T. R. Walls, R. B. Dillinger, G. W. Burdette, N. L. Purcell, and R. F. Dettling, "Performance and Safety Aspects of Monopropellant and Non-Toxic Bipropellant Liquid Divert/ACS Propulsion for Navy TBMD Applications," Proceedings of the 2nd Annual AIAA SDIO Interceptor Technology Conference, 6-9 June 1993, Albuquerque, NM. AIAA 93-2634. Paper UNCLASSIFIED.
M. K. Minthorn, N. L. Purcell, J. B. Moore, G. W. Burdette, and R. F. Dettling, "Non-Toxic Liquid Bipropellant Divert/ACS Propulsion Technology Demonstration Efforts," Proceedings of the 1995 JANNAF Joint Propulsion and Subcommittees Meeting, 4-8 December 1995, Tampa, FL. Paper UNCLASSIFIED.
N. L. Purcell, M. K. Minthorn, J. J. Rusek, R. F. Dettling, T. Pavia, and J. Grote, "Non-Toxic Liquid Bipropellant Performance Evaluation," Proceedings of the 1996 JANNAF Propulsion and Subcommittees Joint Meetings, 9-13 December 1996, Albuquerque, NM. Paper UNCLASSIFIED.