Experimental Investigation of a Baffled-Tube Ram

Experimental Investigation of a Baffled-Tube Ram Accelerator C. Knowlen1, J.F. Glusman2, R. Grist2, A.P. Bruckner3 University of Washington, Seattle, WA, 98195 and A.J. Higgins4 McGill University, Montreal, Quebec Canada

The baffled-tube ram accelerator is an innovation in ram accelerator technology that allows the acceleration of axisymmetric projectiles in the velocity range of 500 to 3000 m/s. This device has the potential to triple the thrust performance of the conventional smooth-bore ram accelerator while reducing its minimum starting velocity. The baffled-tube ram accelerator utilizes a series of internal baffles to suppress the forward surging of a combustion driven shock wave, thus enabling operation in propellants having two to three times the energy release of those used with conventional smooth-bore ram accelerators. An experimental and theoretical investigation of this device is currently in progress at the University of Washington. Operation at velocities between 620 and 1220 m/s has been demonstrated to date. Theoretical modeling indicates that momentum loss due to baffle interactions is a key factor in the baffledtube ram accelerator, which reduces its performance. Nevertheless, baffled-tube experiments have demonstrated thrusts 30-100% greater than that of a smooth-bore ram accelerator operating at the same fill pressure. The design, modeling, and experimental results from a 38mm-bore, two-meter-long baffled-tube ram accelerator apparatus are presented.

Nomenclature A cd cp d F h I = F/PA M m L P Q = q/cpT T t V

cross-sectional area coefficient of drag constant pressure specific heat diameter thrust enthalpy non-dimensional thrust Mach number mass length pressure non-dimensional heat release temperature time volume

β q γ ρ BTRA SBRA CJ Subscripts 1 2 b c eff

volume ratio heat release per unit mass specific heat ratio density baffled-tube ram accelerator smooth-bore ram accelerator Chapman-Jouguet

entrance state exit state baffle parameter chamber parameter SBRA equivalent for BTRA

1

Research Associate Professor, William E. Boeing Dept. of Aeronautics & Astronautics, Box 352250, Associate Fellow, AIAA. 2 Graduate Student, William E. Boeing Dept. of Aeronautics & Astronautics, Box 352250, Student Member, AIAA. 3 Professor, William E. Boeing Dept. of Aeronautics & Astronautics, Box 352250, Fellow, AIAA. 4 Associate Professor, Mechanical Engineering, 817 rue Sherbrooke Ouest, Senior Member, AIAA. 1 American Institute of Aeronautics and Astronautics

I. Introduction

T

HE baffled-tube ram accelerator (BTRA) is one of the newest concepts for a hypervelocity mass driver utilizing chemical propellant.1,2 This device has the potential to triple the thrust of the conventional smooth-bore ram accelerator (SBRA) while operating at the same fill pressure, and can, in principle, reduce its minimum starting velocity (700 m/s) by at least 25%. This new launcher has the potential to accelerate axisymmetric projectiles with masses ranging from grams to metric tons to velocities up to approximately 3000 m/s. Its low starting velocity ability allows for a wide range of pre-launcher technologies to be considered for any application, which may greatly reduce overall cost and operational complexities of this mass driver system. Thus the BTRA offers a significant advancement in ram accelerator technology for hypervelocity applications where axisymmetric projectiles are preferred. Even though the SBRA has a demonstrated velocity range (700 to 2700 m/s) while operating at velocities below the Chapman-Jouguet (CJ) detonation speed of the propellant, it only realizes about 30-50% of its theoretical thrust potential, as indicated in Fig. 1. This reduced thrust arises because the propellants must be diluted with inert gas to keep the combustion-driven shock wave on the aftbody of the projectile from being pushed forward and ahead of the projectile throat (point of minimum flow area and usually the maximum projectile diameter). 3 This empirical finding was key to getting the ram accelerator to work in the first place; and the appropriate amount of dilution was found to be that which reduced the stoichiometric heat release by 40-70%. Attempts to enable operation in propellants (CH4/O2/N2) with less dilution by increasing the projectile throat diameter were unsuccessful.4 Thus, Figure 1. Experimental velocity-distance data even though the SBRA experimental thrust was only a from a 38-mm-bore SBRA compared with the fraction of that which was theoretically possible, scaling velocity-distance profile predicted when using studies for various hypervelocity applications using the optimal propellants with same fill pressure and empirical SBRA thrust have shown that practical systems projectile mass. were readily feasible for applications such as aeroballistic ranges5 and direct space launch.6,7 The pursuit of higher ram accelerator thrust at a given fill pressure led to the invention of the baffled-tube ram accelerator. This device has a series of baffles inside a straight bore tube that create a sequential series of chambers, as shown in Fig. 2. In the configuration presented here, rails are used to guide the projectile and stiffen the baffle chambers. The baffle spacing is such that at least one baffle is always completely blocked by the projectile as it passes through the tube. This requires the straight shoulder section of the projectile to have the length of at least one chamber and two baffle thicknesses. The flow field of the BTRA (shown in Fig. 3) illustrates how the projectile ingests a fresh charge of propellant as it enters a chamber and compresses the propellant until the projectile shoulder reaches the baffle. Thereafter the propellant enters combustion zone behind the projectile which sustains very high pressure on its

Figure 2. Baffled-tube ram accelerator tube section with axisymmetric projectile.

Figure 3. Propellant is ram compressed in annular chamber around projectile shoulder and reacts behind the baffle.

2 American Institute of Aeronautics and Astronautics

base. In this manner the BTRA allows the projectile to act as a one-way valve that enables propellant to pass around the projectile but keeps the combustion-driven shock waves from being pushed ahead of it. This operating characteristic allows the use of propellants with minimal dilution, which, in principle, results in at least twice the thrust when using the same fill pressure as a SBRA.

II. Theoretical Performance Modeling The SBRA flow field when operating with subsonic combustion is illustrated in Fig. 4. At velocities less than the CJ speed, combustion heat release behind the projectile thermally chokes the flow which stabilizes a normal shock on its tapered aftbody.8,9 At a given in-tube Mach number, the location of the normal shock moves forward toward the projectile throat as the propellant heat release is increased, resulting in higher thrust. If the heat release is excessive, the normal shock is pushed through the throat and an unstart subsequently ensues.3 In this situation, the projectile pushes a normal shock ahead of it which results in significant drag that 2 1 causes rapid deceleration. As the projectile velocity Figure 4. Thermally choked ram accelerator approaches the CJ speed, the normal shock, in principle, falls flow field with normal shock on projectile off the projectile base resulting in a cessation of thrust. aftbody sustained by combustion heat release. Applying the quasi-steady one-dimensional conservation Control volume entrance at “1” and exit at “2”. equations to the control volume indicated by dashed lines in Fig. 4 (stations 1 and 2 are the entrance and exit planes, respectively) and assuming thermally choked flow at the exit of the control volume (M2 =1), the following closed form equation for SBRA thrust-Mach number performance in terms of propellant properties at stations 1 and 2 can be determined:

𝐼𝑆𝐵𝑅𝐴 = 𝑃

𝐹

1 𝐴𝑏

=

𝑀1 𝛾1 𝛾2

ℎ1

(1 + 𝛾2 )√(

𝛾2 −1 𝑐𝑝1 𝑇1 𝛾1 −1

)

2

𝑀 (𝛾 −1) + 1 1 +𝑄

2 (𝛾2 −1) ℎ2 + 𝑐𝑝2 𝑇2 2

− (1 + 𝛾1 𝑀12 )

(1)

where the non-dimensional thrust, ISBRA, is the ratio of thrust to bore cross sectional area, Ab, and propellant fill pressure P1. The influence of combustion heat release added to the flow behind the projectile is determined by applying a nondimensional heat parameter, Q = q/cp1T1. Detailed non-dimensional thrust predictions as function of M1 are determined by iteration on the exit flow to find the corresponding chemical equilibria (representative theoretical results are plotted in Fig. 1). Because the flow is thermally choked, the exit flow is an entropy extremum, thus the details of the flow processes within the control volume are not significant in the SBRA. 10 At the CJ Mach number, MCJ, the non-dimensional detonation speed heat release parameter, QCJ, can be determined by setting ISBRA = 0 in Eq. (1) as shown below. Note this expression is simplified by assuming a constant value for  and letting h/cpT = 1.

QCJ



2

   

  1  M 2 2 1 CJ    1   2  1 M CJ 



2



2

M CJ  1  1   h2  2  1   h1 1  1 2    M CJ       2       2  1  c p 2T2 2 c T 2    p1 1  2    1 M CJ  1

(2)

In lieu of detailed calculation of the Mach number dependent end state equilibrium at the thermal choking point, the value of QCJ can be used as a reference to compare the amount of chemical heat release per mass of propellants. If one assumes the Q value of the SBRA is constant and equal to QCJ, the non-dimensional thrust vs. Mach number profile thus determined will under-predict the maximum thrust by 5-10% for typical propellants,9 which is not significant for the qualitative comparisons of interest presented here. Listed later are the detailed calculations of Q based on the 0 K heat of formation enthalpies, these will not contain the “CJ” subscript. When operating a ram accelerator in a tube having a complicated internal geometry, there is no preferential area reference for thrust normalization. Thus, for the BTRA a more general non-dimensional thrust definition is used in which net thrust is referenced to an equivalent tube area, Aeff, determined by the ratio of the net baffle chamber volume, Vb +VcVrail to the total length, Lb + Lc of one baffle section, as shown in Fig. 5, where Vrail is the net volume of any rails that may be incorporated in between the baffles (not shown in Fig. 5). The BTRA thrust can then be normalized and compared with non-dimensional thrust of a SBRA based on the effective area of the baffle bore as follows: 3 American Institute of Aeronautics and Astronautics

I BTRA 

( Lb  Lc ) 1  Lc Lb I F F F F     SBRA p1 Aeff p1 (Vb  Vc  Vrail ) p1 Ab 1  d c2 d b2 Lc Lb   Vrail Vb p1 Ab  

with  

1  d d Lc Lb  1  Lc Lb  Vrail Vb 2 c

(3)

2 b

where  accounts for the net baffle chamber volume as a function of chamber length, Lc, chamber diameter, dc, baffle thickness, Lb, and baffle inner diameter, db (Fig. 5). Since Vc   1 , the greater the value of , the larger the chamber volume and the more propellant per unit tube length is Vb available to accelerate the projectile. The non-dimensional db dc Vc + Vb deff thrust expression for a SBRA is recovered when  = 1. Hence, to calculate the ideal BTRA thrust as a function of M1, the non-dimensional thrust from Eq. (1) is multiplied by p1Ab. Experiments have shown that low pressure detonation Lc+Lb waves travel slower through tubes with baffles than in Lc Lb 11 smooth-bore-tubes. In a similar manner, the thermally choked region behind a BTRA projectile is expected to Figure 5. Effective area for BTRA chamber. experience form drag, shock wave total pressure losses, and viscous effects as the flow behind the projectile passes through the baffles at relatively high speed. To account for these effects without detailed flow modeling, a baffle drag coefficient, cd, is applied to the choke point, resulting in a net drag force which is subtracted from the thrust calculated for the BTRA.12 This drag force is assumed to be proportional to the square of the velocity of the choke point in the lab frame, namely v2,lab = v1  v2, and an empirically determined baffle drag coefficient, cd. The baffle drag force is thus estimated as follows:

1 1 Fdrag  cd 2v22,lab Aeff  cd 2v22,lab Ab  2 2

(4)

Normalizing this drag by p1Ab results in a non-dimensional baffle drag, Idrag, that can be subtracted from the nondimensional thrust of the BTRA to estimate its general thrust-Mach number characteristics. The maximum thrust of a SBRA occurs when the projectile velocity is exactly the same as the flow velocity at the thermal choking point, which corresponds to a flow velocity in the lab reference frame equal to zero. Because the baffle drag equals zero at this point, the BTRA and SBRA are predicted to have exactly the same non-dimensional thrust at this projectile velocity. At lower projectile velocities, the flow at the choke point moves backwards away from the projectile, resulting in a pressure build up due to partial flow restriction that enhances the BTRA thrust relative to that of SBRA. As the projectile increases velocity, the choke point accelerates in the direction of projectile motion, resulting in greater baffle drag and ultimately a lower maximum attainable Mach number. Plotted in Fig. 6 are non-dimensional thrust vs. Mach number profiles for the SBRA and BTRA with representative baffle drag coefficient parameters of cd = 0.3 and 0.6 to illustrate their impact on thrust. The specific heat ratio is constant at  = 1.2 and the non-dimensional heat release value is Q = 12 for all of these profiles; which are values representative of a slightly diluted methane-oxygen propellant that has more than twice the heat release that can typically be used in a conventional SBRA. All three plots coincide at M1 ≈ 3.6, which corresponds to the peak thrust for the SBRA under these conditions. At Mach numbers less than the SBRA maximum, peak BTRA thrust increases with Figure 6. Non-dimensional thrust vs. in-tube increasing drag coefficient, going through a relative maximum in Mach number for SBRA and BTRA for: the Mach number range of 3.0 to 3.5 for the cd values used here, Q = 12,  = 1.2, cd = 0.3 and 0.6 which are comparable to those of a bluff-body in subsonic flow; 4 American Institute of Aeronautics and Astronautics

i.e., 0.3 and 0.6. Conversely, the zero thrust Mach number decreases with increasing drag coefficient, which is consistent with velocity deficit data from detonation experiments in tubes with baffles.11,12

III. Experimental Apparatus A. Ram Accelerator Facility The 38-mm-bore facility used for ram accelerator experimentation at the University of Washington consists of a light gas gun pre-launcher, 16-m-long test section having 102-mm-O.D., and muzzle blast tank with a projectile catcher tube, as shown in Fig. 7. Propellant gases are metered with mass flow controllers and mixed in-line through the tubing routing the mixture to the desired segment of the test section.13 Stages containing different propellants are separated by thin Mylar diaphragms which are readily penetrated by the projectile. Finned-projectiles with a 15-mmthick polycarbonate sabot can be launched to test section entrance velocities as high as 1200 m/s. Fill pressures up to 200 bar have been successfully used in SBRA experiments in this facility.14

Figure 7. Ram accelerator facility (38-mm-bore) with finned-projectile having a conical diffuser. B. Baffled-Tube Apparatus Key to realizing the performance benefits of the BTRA is understanding how the baffles interact with particular projectile geometries in highly energetic propellants. To this end, an experimental BTRA system was designed and fabricated with removable inserts that slip into a shell tube having instrument stations evenly spaced throughout its length, as shown in Fig. 8. The selected baffle spacing results in 37 chambers per meter, each 31.9-mm-long, the bore is sized for 38-mm-diameter projectiles (db = 38.1 mm), and the inner diameter of the shell tube is dc = 76.2 mm. The baffle chambers have four tapered rails with a 20° wedge angle (tapering from an O.D. arc length of ~16 mm to an I.D. arc length of 2.5 mm) that are staggered at 45° between adjacent chambers to enhance structural rigidity. The resulting volume-increase-factor is  = 3.001 when accounting for the volume displacement of the rails. This system was designed for operational flexibility in that chambers can be readily removed and replaced with variations as desired. Compression on the inserts is provided by twin end caps to hold them place and effect a pressure seal. The BTRA test section was inserted into the existing ram accelerator system at the UW and instrumented with piezoelectric pressure transducers (PCB 119) and electromagnetic (EM) sensors of an in-house design.15 C. Data Acquisition System The data acquisition system is based on a National Instruments PXIe-1071 configured to collect data from the piezoelectric pressure transducers and EM sensors at 1.25 MHz across up to 32 channels. EM sensor data consist of up to 40 separate signals that are multiplexed into 6 of the 32 available channels, allowing up to 26 pressure transducers to be monitored with the remaining data channels. In addition to measuring the tube wall pressure field, the arrival time data obtained by pressure transducers and EM probes at known axial distances along the test section are used to calculate velocity–distance profiles, from which the thrust–Mach number characteristics of the BTRA can be determined.


Figure 8. A two-meter-long BTRA with a 38-mm-bore and 75-mm-shell inner diameter is installed in existing ram accelerator test section. This configuration has 37 baffle chambers per meter.

D. Projectile Configuration The use of axisymmetric projectiles is an advantage of the BTRA as they are easier to machine and provide greater volume-to-surface ratio for more payload capacity. Initially, two-piece polycarbonate projectiles were used with an aluminum stud threaded into each end so that the pieces sandwiched an annular Neodymium magnet (32-mm-O.D.). These projectiles were later replaced by one-piece polycarbonate projectiles of similar external geometry with a single thin Neodymium disk magnet (9.5-mm-dia) mounted internally. A side-by-side picture of the projectiles is shown in Fig. 9. All projectiles tested had a nose cone half-angle of 15°, cylindrical shoulder diameter of 37.2 mm, base diameter of 17.8 mm, and total length of 170 mm. The aft tail cone frustum half-angle was varied from 9° to 12° while keeping the base diameter and overall length constant, thus changing the cylindrical shoulder length from 111 to 126 mm. The average mass of the projectiles used in this test series was 141 g with a standard deviation of ~6 g. A machine drawing for the single-piece projectile with an aft tail cone frustum half-angle of 20° is shown in Fig. 10, these were all later machined down to the prescribed 9° or 12° angles for various experiments as shown in Fig. 9.

Figure 9. Comparison of the one-piece and twopiece polycarbonate projectiles.

Figure 10. Schematic of basic design for a one-piece polycarbonate projectile (inches).


IV. Results and Discussion Experiments have demonstrated accelerations in the range of 50,000 to 100,000 m/s2 with projectiles having a nominal mass of 141 g in 2CH4+4N2O propellant at fill pressures of 10 to 20 bar. This particular propellant has approximately twice the heat release per unit mass than can be used in the SBRA. Representative tube-wall pressure data from a 15 bar experiment are shown in Fig. 11, along with the corresponding velocity-distance data obtained from tube-wall pressure transducers and EM probes outside of the BTRA section. In this case the average acceleration was ~59,000 m/s2 at an average Mach number of ~3.85 for a velocity gain of ~100 m/s over 2 m. The EM probes in the BTRA were ~20 mm from the annular neodymium magnet carried aboard the projectile, but the interaction of the complicated combustion flow field with the EM signal in the baffle chambers made the arrival time measurements unreliable. Furthermore, the shock waves reflecting within the baffle chambers led to complicated tube-wall pressure data that could not in all cases be interpreted reliably enough to get an accurate projectile arrival time, thus there were no details of the velocity history within BTRA section for some of the experiments. Nevertheless, the pressure on the projectile base evident in Fig. 11 is nearly five times greater than that in the annular region around its shoulder, indicating that thrust is due to combustion behind projectile. The non-dimensional BTRA thrust predicted for this experiment is IBTRA = 4.75, whereas the experiment yielded IBTRA-exp = 1.80; resulting in a non-dimensional thrust deficit of 2.95.

Figure 11. Tube wall pressure and velocity-data from BTRA experiment with projectile mass = 140 g. In low entrance speed experiments, BTRA operation was demonstrated at 620 m/s (M1 ~ 2.1) as shown in Fig. 12, which is ~12% lower than the lowest ram accelerator starting velocity (700 m/s) reported to date.16 Again the detailed velocity-distance data were not discernible from the tube wall data within the BTRA section; however, a velocity gain of over 250 m/s was realized. Note that M1 ~ 2 is the theoretical minimum starting Mach number for this BTRA

Figure 12. Pressure and velocity-data from BTRA experiment with entrance velocity of 620 m/s. 7 American Institute of Aeronautics and Astronautics

geometry under ideal flow conditions. In this particular experiment the projectile accelerated at an average rate of 80,000 m/s2 for 2 m before leaving the BTRA section. The non-dimensional BTRA thrust predicted for this experiment was IBTRA = 4.13, whereas the experiment yielded IBTRA-exp = 2.57. Thus the non-dimensional thrust deficit was 1.56. From the series of experiments used to determine the starting envelope of the BTRA in methane-oxygen with carbon dioxide diluent, the experiment shown in Fig. 13-left yielded velocity-distance data indicating that the projectile entered the test section at too low an entrance velocity (710 m/s, M1 = 2.31) which resulted in deceleration due to an unstart. When the entrance velocity, and thus Mach number, was increased slightly to 800 m/s (M1 = 2.52), the projectile entered the baffle-tube section and accelerated (Fig. 13-right) for a velocity gain of 120 m/s. Of the carbon dioxide diluent tests carried out to date, this Q = 12.09 result was the highest Q value to show positive ram acceleration. For comparison purposes, the typical propellant used in the SBRA results in unstarts when the Q value exceeds 5 at similar entrance speeds.16

Figure 13. Velocity-distance data from BTRA experiments with entrance velocities of 725 to 800 m/s for 1CH4 + 2O2 + 1.86CO2 at fill pressure of P1 = 12 bar. The non-dimensional thrust data from BTRA experiments in both CH4-N2O and CH4-O2-CO2 propellants are shown in Fig. 14-left along with the theoretical non-dimensional thrust curves for 1CH4+2O2+2CO2 and 2CH4+4N2O propellants. The theoretical Mach numbers for peak thrust were 2.94 and 3.86 for the CH4-O2-CO2 and CH4-N2O propellants, respectively. These experimental data were normalized by the corresponding theoretical non-dimensional BTRA thrust (Eq. (3)) and plotted versus in-tube Mach number in Fig. 14-right. The data set from the experiments in CH4-N2O propellant (Q ~ 13) carried out at Mach numbers less than that for maximum theoretical thrust had an average ratio of experimental-to-theoretical non-dimensional thrust of ~0.5. Experiments in CH4-O2-CO2 propellant (Q ~ 12) were carried out near the peak thrust Mach number (2.94). These resulted in an average experimental-totheoretical non-dimensional thrust ratio of ~0.85. Thus it appears that the BTRA can operate quite close to theory in the CH4-O2-CO2 propellant when experiments were carried out near the peak thrust Mach number (2.94). Experiments are currently in progress to explore the thrust characteristics of the BTRA over a much broader range of Mach numbers. The aforementioned data were used to calculate a thrust deficit vs. Mach number. The thrust deficit normalized by theoretical non-dimensional thrust are plotted in Fig. 15. The low values of thrust deficit for experiments in CH4-O2-CO2 propellant at near theoretical maximum thrust Mach numbers are consistent with minimal baffle drag. The larger thrust deficits for experiments in CH4-N2O propellant at Mach numbers lower than that for maximum thrust are contrary to the predictions for the baffle drag model. The implications of these results are still under investigation. Even though the conventional SBRA theoretical non-dimensional thrust is greater than that measured in the experiment for the Mach number range shown in Fig. 14-left, the average experimental thrust per unit fill pressure (F/P1) for the BTRA experiments was greater than SBRA theory by ~100% for CH4-O2-CO2 and ~30% greater for CH4-N2O propellants, as shown in Fig. 16. These enhanced thrust percentages indicate the expected benefit from operating with a BTRA at the same pressure as a SBRA having the same bore diameter as the baffles. This implies that even though the BTRA experiments had lower than expected non-dimensional thrust, they still outperformed a conventional SBRA utilizing the most energetic propellant feasible for its operation.


Figure 14. Experimental non-dimensional thrust vs. in-tube Mach number (left). Ratio of experimental to theoretical BTRA non-dimensional thrust vs. in-tube Mach number (right).

Figure 15. Ratio of non-dimensional thrust deficit to theoretical BTRA non-dimensional thrust vs.in-tube Mach number.

Figure 16. Experimental and theoretical thrust per unit pressure (cm2) vs. M1.

Computational fluid dynamic modeling is currently in progress to gain a better understanding of the combustion process as the projectile passes through the baffle chambers in a BTRA. The challenges of modeling reactive flow in the laboratory reference frame for the passage of a supersonic projectile through complicated baffle geometries are being addressed using the FLUENT computer code. In this manner, alternative baffle geometries are being explored that may improve the BTRA thrust.

V. Conclusion Baffled tube operating characteristics have been investigated in a 2-m-long baffled-tube ram accelerator test section having 37 chambers. Ignition and thrust have been observed in CH4-N2O and CH4-O2-CO2 propellants having heat releases more than thrice as great as that which can be used in conventional smooth-bore ram accelerators. Operation with entrance velocities as low as 620 m/s and peak velocities of 1220 m/s were demonstrated in 2CH4+4N2O propellant. The measured thrusts were 15-50% lower than predicted. Nevertheless, even though this apparatus did not perform as well as theoretically predicted, it still obtained 100% greater thrust than the conventional smooth-bore ram accelerator on a force per unit fill pressure basis. Thus the potential to reduce the ram accelerator section length by a factor of two for a given velocity gain while reducing then minimum ram accelerator entrance velocity by 12% have been demonstrated.


Acknowledgments The authors would like to thank the undergraduate and graduate students who have worked tirelessly in the ram accelerator laboratory over the past years as well as the sponsors of this research program: Joint Center for Aerospace Technology Innovation (JCATI), Washington Research Foundation (WRF), and EnergeticX, LLC. The authors declare the following competing financial interest: Carl Knowlen, Andrew Higgins, and Adam Bruckner have a financial interest in HyperSciences, Inc., which has licensed the described technology from the University of Washington in concert with EnergeticX, LLC.

References Higgins, A. J., Knowlen, C., and Kiyanda, C. B., “Gasdynamic Operation of Baffled Tube Ram Accelerator in Highly Energetic Mixtures,” 20th International Colloquium on the Dynamics of Explosions and Reactive Systems. McGill University, Montreal, Canada, July 31 – August 5, 2005. 2 Higgins, A. J., “Ram Accelerators: Outstanding Issues and New Directions,” J. of Propulsion and Power, Vol. 22, No. 6, 2006, pp. 1170-1187. 3 Higgins, A. J., Knowlen, C., and Bruckner, A. P., “Ram Accelerator Operating Limits, Part 1: Identification of Limits,” J. of Propulsion and Power, Vol. 14, No. 6, 1998, pp. 951-958. 4 Higgins, A. J., Knowlen, C., and Bruckner, A. P., “Ram Accelerator Operating Limits, Part 2: Nature of Observed Limits,” J. of Propulsion and Power, Vol. 14, No. 6, 1998, pp. 959-966. 5 Wieland, K., and Bruckner A. P., “Ram Accelerator Ballistic Range Concept for Softly Accelerating Hypersonic Free-Flying Models,” J. of Aircraft, Vol. 31, No. 6, 1994. 6 Knowlen, C., and Bruckner, A. P., “Direct Space Launch Using Ram Accelerator Technology,” Space Technology and Applications International Forum. American Institute of Physics. 2001. 7 Knowlen, C., Joseph, B., and Bruckner, A. P., “Ram Accelerator as an Impulsive Space Launcher: Assessment of Technical Risks,” International Space Development Conference. Dallas, TX, May 25-28, 2007. 8 Hertzberg, A., Bruckner, A. P., and Bogdanoff, D. W., “Ram Accelerator: A New Chemical Method for Accelerating Projectiles to Ultrahigh Velocities,” AIAA Journal, Vol. 26, No. 2, 1988, pp. 195-203. 9 Bruckner, A. P., Knowlen, C., Hertzberg, A., and Bogdanoff, D. W., “Operational Characteristics of the Thermally Choked Ram Accelerator,” J. of Propulsion and Power, Vol. 7, No. 5, 1991, pp. 828-836. 10 Knowlen, C. and Bruckner,* A.P., “A Hugoniot Analysis of the Ram Accelerator,” Shock Waves, Vol I, Proceedings of the 18th Int’l Symposium on Shock Waves, Sendai, Japan, Takayama, K., ed., Springer-Verlag, Berlin, pp. 617-622, 1992. 11 Gu, L. S., Knystautas, R., and Lee, J. H., “Influence of Obstacle Spacing on the Propagation of Quasidetonation,” Progress in Aeronautics and Astronautics, Vol. 114, 1988, pp. 232-247. 12 Tanguay, V., and Higgins, A., “On the Inclusion of Frictional Work in Non-Ideal Detonations,” 20th International Colloquium on the Dynamics of Explosions and Reactive Systems. McGill University, Montreal, Canada, July 31–August 5, 2005. 13 Knowlen, C., Bundy, C., Schwab, R., and Bruckner, A. P., “University of Washington High Pressure Ram Accelerator Facility,” Proceedings of the 50th Aeroballistic Range Association, Pleasanton, CA, November 8-12, 1999. 14 Bundy, C., Knowlen, C., and Bruckner, A. P., “Unsteady Effects on Ram Accelerator Operation at Elevated Fill Pressures,” Journal of Propulsion and Power, Vol. 20, No. 5, 2004, pp. 801-810. 15 Bogdanoff, D. W., Knowlen, C., Murakami, D., and Stonich, I., “Magnetic Detector for Projectiles in Tubes,” AIAA Journal, Vol. 28, No. 11, 1990, pp. 1942-1944. 16 Knowlen, C., Schultz, E., and Bruckner, A. P., “Investigation of Low Velocity Starting Techniques for the Ram Accelerator,” 33rd AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit, Seattle, WA, July 6-9, 1997. 1
