Initiation Attempts Using a Collapsing Shock Wave

The initiator was also used to generate an imploding shock wave in an attempt to initi- ate the test-section mixture. Recent computational simulations by Li and Kailasanath (2003b) have suggested that it is possible to initiate JP10-air mixtures using impul-

Figure 4.19: A schematic demonstrating the reduction in overfill volume due to sym- metry. For pure symmetry, the end flange wall would be located at the dashed line. The volume of overfill gas would be reduced to the gas shaded gray, which is half of the volume required in the case where no end flange wall is present.

sively started jets of JP10 and air to create an annular shock wave.

A preliminary investigation of this notion was examined by conducting imploding shock experiments with the present setup. (More detailed experiments on this topic were performed with a different initiator and are discussed in the next chapter.) In order to generate an imploding shock wave, the initiator was partially filled (roughly 30%–40%) with initiator gas. Detonation of this gas propagated a shock wave followed by a deflagration through the channels of the device. This shock wave then implodes at the focus, creating an imploding shock wave in the fuel-air mixture.

This technique was unsuccessful at initiating stoichiometric ethylene-air mixtures. Pressure traces from an experiment where 41% of the initiator (and all of the plumb- ing) was filled with initiator mixture are shown in Fig. 4.20. The location of the pressure and ionization probe traces from Fig. 4.20 are shown in the schematic in Fig. 4.21. The test section mixture was ethylene-air.

Pressure transducer P2 shows a shock wave with an overpressure of 12 bar that is propagated into the test-section mixture from the initiator. As this wave implodes, the pressure measured near the focus is 100 bar. Farther down the tube, pressure transducers P4 and P5 show a shock with an overpressure of 4 bar. Measured wave speed in the test section is roughly 630 m/s while UCJ is 1825 m/s. Initiation of the test section mixture was not successful. The pressure traces are similar to those pre-

-6.00 -4.50 -3.00 -1.50 0.00 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 Ion (V) Time (s) Ion shot 382 0.00 0.20 0.40 0.60 0.80 1.00 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 P (MPa) P5 shot 382 0.00 0.20 0.40 0.60 0.80 1.00 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 P (MPa) P4 shot 382 0.00 2.75 5.50 8.25 11.00 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 P (MPa) P3 shot 382 0.00 0.25 0.50 0.75 1.00 1.25 1.50 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 P (MPa) P1 shot 382 0.00 0.25 0.50 0.75 1.00 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 P (MPa) shot 382

Figure 4.20: Pressure and ionization traces from shot 150, a typical shock initia- tion experiment. Test-section mixture was stoichiometric ethylene-air at 1 bar initial pressure. Traces are labeled and correspond to locations shown in Fig. 4.21.

viously presented for the “failed initiation” case where an imploding detonation wave (instead of a shock wave) was propagated into the test section. In these experiments, the imploding shock wave was not of sufficient Mach number and the post-shock flow

P2 P3

P4 P5

Ion Ion Ion Ion

Spark P1

Figure 4.21: A schematic of the experimental setup used for attempted initiation of HC-air mixtures using an imploding shock wave.

was not of sufficient duration to initiate the ethylene-air mixture. This concept is addressed more extensively in the next chapter.

4.4

Summary

A dynamic planar initiator has been developed that is capable of producing a large- aspect-ratio planar-detonation-wave in insensitive mixtures. The planar initiator uses a single weak spark and a small amount of fuel-oxygen mixture to produce the planar wave in a short distance and is capable of initiating detonations in mixtures such as C3H8+5O2+9N2 and C2H4+3O2+10.5N2 (Austin, 2003). The device is currently in use on Caltech’s Narrow Channel Facility.

A dynamic toroidal initiator has been developed that creates an imploding wave in an insensitive mixture using a small amount of hydrocarbon-oxygen gas and a weak spark. The imploding detonation wave initiates detonations in propane-air and ethylene-air mixtures with sufficient amounts of hydrocarbon-oxygen gas. The minimum volume of sensitive-initiator gas required for hydrocarbon-air detonation exceeded the volume of the initaitor channels, casuing some gas to spill into the test section. This “overfill volume” of initiator gas was found to increase as the mixture sensitivity decreased and increase as the implosion was moved away from the test-section end flange. It was thought that the proximity of the end flange to the wave focus reduced the critical overfill volume by creating symmetry and providing additional surfaces for wave reflection.

not able to initiate the fuel-air mixtures for the single case tested. More rigorous ex- periments testing the effectiveness of imploding shock waves at detonation initiation are presented in the following chapter.

Chapter 5

Imploding Shock Wave Initiator

5.1

Introduction

In Chapter 4, the toroidal initiator results were characterized by the amount of acetylene-oxygen gas injected into the initiator channels during each test. In or- der to perform any type of gas-dynamical analysis, it is desirable to translate the amount of gas into a wave strength for the implosion. However, this conversion is not straightforward due to several undetermined factors in the experiment.

For example, while pressure measurements taken at the end flange could be used to infer the implosion strength, they were not aligned with the main axis of the implosion and were shown to be affected by both diffraction and reflection from the end flange itself. Furthermore, a contact surface separating the initiator gas and the test-section gas was created in the test section during the dynamic gas injection. The location of this contact surface was not measured and it varied with the amount of initiator gas used. The imploding detonation wave would have been affected by the contact surface in two ways. First, the density gradient at the interface would have affected the transmitted wave strength and created a reflected wave. Second, the test-section mixture was much less sensitive than the initiator gas, which could have caused the transmitted detonation wave to fail or substantially weaken. In an effort to create an experimental situation that would be simpler to analyze, a facility was designed to create an imploding annular shock wave (as opposed to the imploding detonation waves in previous chapters) that was used to initiate detonations in combustible

mixtures.

The facility was also designed to experimentally test the numerical work of Li and Kailasanath (2003b), who proposed that imploding shock waves could be used to initiate insensitive mixtures. Their numerical simulations (Li and Kailasanath, 2003a) found that detonations could be initiated in a 14 cm (5.5 in) diameter tube filled with stoichiometric ethylene-air using an imploding shock wave created from the injection of a converging annular jet of fuel or air from the outer diameter of the tube. At the injection point, the jet had a Mach number of 1, a pressure of 2.0 bar, and a temperature of 250 K. For a perfect gas with γ = 1.4, a jet with these properties could be generated from a reservoir with a total pressure of 3.8 bar and a total temperature of 470 K.

The concept of detonation initiation via a converging air jet is extremely appealing to designers of PDEs since it would eliminate the need for a spark plug and associated power supply or any sensitizer fuel. In flight, stagnation of the atmosphere would supply the hot, pressurized air needed to create the imploding wave.

The facility discussed in this chapter used a shock tube to create a reservoir of hot, pressurized air to generate imploding annular shock waves that were propagated into a 7.6 cm (3.0 in) diameter test-section tube filled with either stoichiometric ethylene- oxygen or propane-oxygen mixtures diluted with varying amounts of nitrogen. The strength of the imploding shock wave and the sensitivity of the test gas were varied in an effort to find the minimum shock strength required to initiate a detonation in each mixture. The total pressure of the air jet that was used to create the implosion ranged from 3.2 bar to 16.8 bar and the total temperature ranged from 420 K to 790 K. These jet properties were comparable to those proposed by Li and Kailasanath (2003a).