Driver Calibration

The driver of the GALCIT Detonation Tube is intended to reliably and controllably initiate detonations in any mixture that is detonable in the tube [Akbar and Shepherd (1996)]. It consists of acetylene and oxygen cylinders, regulators, flash arrestors, and valves, an injection valve, a digital control circuit, and an exploding wire circuit. The control circuit actuates the electropneumatic valves to control the driver injection duration and ignition delay, and triggers the exploding wire. It is interlocked to various gas supply valves and hydraulic closure devices for safety purposes.

A manual fire signal starts the driver sequence. The control circuit opens the acetylene and oxygen valves and the injection valve for a programmed time period and then waits for a programmed delay period. The ratio of acetylene to oxygen is controlled by adjusting the cylinder pressure regulators. A fire signal is then sent to a trigger module that sends a high voltage trigger to the exploding wire spark gap. The spark gap switches the 2 $\mu$F capacitor bank (typically charged to 9 kV) through a small copper wire in the tube. The oxy-acetylene mixture is easily initiated by this discharge and transitions to a planar detonation wave which is transmitted to the test mixture.

Tests are periodically performed to verify the overall quantity of driver gas injected with each shot and to measure and adjust its equivalence ratio. To check the amount of gas injection, the driver is triggered several times without the exploding wire and the final pressure is measured after each injection. Measuring the detonation wave speed in this mixture and comparing it to equilibrium calculations (STANJAN) allow an estimate of the equivalence ratio. The driver is kept slightly lean to avoid formation of soot.

A number of tests with the driver transmitting blast waves into air have been performed to evaluate its equivalent energy. A summary of results from these shots is presented in Table .

 

Table: Driver Characterization Shot List

Shot	Press.	Flow Duration		Delay		D1-2	D2-3	$\Delta$P1	$\Delta$P2	$\Delta$P3
	(kPa)	(dial)	(s)	(dial)	(s)	(m/s)	(m/s)	(kPa)	(kPa)	(kPa)
48	100.0	560	4.442	200	1.025	712	658	430	330	270
581	100.08	560	4.442	200	1.025	-	-
59	25.2	560	4.442	200	1.025	-	-	361	210	150
60	25.2	560	4.442	200	1.025	999.5	846.7	305	206	163
61	50.2	560	4.442	200	1.025	823.6	739.1	340	224	185
62	75.2	560	4.442	200	1.025	-	-	434	312	253
63	100.0	560	4.442	200	1.025	711.8	652.1	350	270	230
642	98.9	69	0.440	200	1.025	-	-
65	99.69	260	1.997	200	1.025	453.5	450.7	95	90	80
662	99.47	137.7	1.000	200	1.025	-	-
67	100.0	751	6.000	200	1.025	744.8	683.2	500	330	310
68	100.0	628.5	5.000	200	1.025	716.8	663.3	450	340	290
69	100.0	505.8	4.000	200	1.025	689.3	635.3	430	280	240
70	100.0	383.1	3.000	200	1.025	655.2	599.5	350	230	200
71	100.0	560	4.442	200	1.025	705.7	644.4	420	300	240
72	100.0	560	4.442	50	0.500	706.6	651.9	500	330	270
73	100.0	560	4.442	336	1.501	707.9	645.3	400	280	250
74	100.0	560	4.442	479	2.002	696.1	646.6	510	380	280
75	100.0	560	4.442	765.1	3.003	701.1	646.4	350	270	240

¹Pressure signals too small on CAMAC data; oscilloscope failed to trigger
²Driver did not detonate

The injection and delay periods are programmed through dial potentiometers. The relationship between the numerical values on these potentiometers and the actual injection and delay periods has been measured, and for reference are given below.

$$\rm Injection~Period = 8.1513(Dial~Setting)-122.26$$

and

$$\rm Ignition~Delay = 3.5(Dial~Setting)+325$$

where the injection period and ignition delay are in units of milliseconds.

According to the approximate analysis by [Thibault et al. (1987)], the far field overpressure in a tube subjected to a blast wave at X=0 is a function of $\gamma$, X, P₀, and E_c, where E_c is the equivalent energy of the source:

$$\frac{\Delta P}{P_0} = \frac{4\gamma/(\gamma+1)}{\sqrt{1+4\gamma/ (\gamma^2-1)\cdot X/L_e}-1}$$

where

$$L_e = \frac{E_c}{P_0}$$

Far field is considered to be X/L_e > 0.3.

Solving these relations for E_c in terms of $\gamma$, $\Delta P/P_0$, and X allows the data in Table  to be used to plot equivalent energy vs injection time. These data, and a semilog curve fit are shown in Fig. .

  

Figure: Driver equivalent energy

Each point represents a shock pressure measured at one of the pressure transducers. The equivalent energy computed from the three transducers for each shot were averaged and used to calculate explosion lengths (L_e) and the nondimensional distance to each transducer. Since the nondimensional distance to each transducer was a function of the initial pressure and the injection time, the validity of the far field assumption was checked for each pressure trace and the data points were sorted and plotted accordingly. The ``far field'' data were found to lie along a linear curve in linear-log coordinates, so a semilog curve fit was made and plotted with the data. The applicability of this fit is limited to the range of injection times investigated. Below about 2 s injection time the driver slug itself will not initiate. Above 6 s, the equivalent energy begins to plateau and the length of the driver slug begins to become appreciable (especially at low initial pressures). However, the data are useful by demonstrating that the relationship between equivalent energy and injection time is not linear. The effect of varying the delay time has also been investigated but no trend with respect to equivalent energy has been found.

Note that the analysis illustrated here assumes that certain system variables are constant, namely oxygen and acetylene delivery pressures and flow rates. These can be affected by variations in cylinder pressure and in the detonation tube initial pressure. Acetylene delivery pressure depends strongly on the frequency of use since the gas is dissolved in acetone within the cylinder. Furthermore, the specific construction of each cylinder can affect its flow characteristics. Variations in the component flow rates can affect the equivalence ratio and the driver equivalent energy. Efforts have been made to compensate for these variations, and tests are performed periodically to correct them.