SHOCK TUBE STUDIES ON T H E IGNITION OF HYDRAULIC FLUIDS GORDON B. SKINNER' Mansanto Research Corp., Dayton, Ohia Ignition of small drops of several hydraulic fluids in shock-heated air was studied, the specific quantity obtained being the temperature of the air (heated by the reflected shock) needed to cause ignition in 2 milliseconds. Correlation was established between these results and those obtained by the CFR compression ignition test. The shock tube method requires less than 1 ml. of fluid per evaluation, can be used to study fluids that require compression ratios above 50 for ignition, and simulates field conditions at least as well as the CFR test. Ignition appears to be governed more by chemical than by physical properties. One experimental fluid was considerably more fire-resistant than the best commercial one tested.
w
has been used extensively in the study of gas-phase reactions, relatively few studies have been made of gas-liquid reactions. The main reason for this has been the difficulty in interpreting the data obtained when two phases are initially present, because the additional steps of drop breakup, evaporation, and mixing need to be considered, as well as chemical kinetics. A study of the combustion process when n-hexadecane is injected as a fine spray into shock-heated air was made by Mullaney (5). By high-speed photography, he was able to observe injection of the spray, evaporation of the droplets, and spontaneous ignition. In most of his experiments, combustion started before evaporation was complete. Shock tubes have also been used by Morrell and coworkers (3, 4 ) and by Hanson, Domich, and Adams ( 2 ) to study the breakup of liquid jets by rapidly moving gas streams. These investigators used inert liquids to study atomization of the liquids without the complications introduced by combustion and were able to obtain equations relating atomization time to physical properties of the liquid and gas. This work develops a method of rating the ignition characteristics of fire-resistant hydraulic fluids that could be related to the single-cylinder engine test (7) currently used, yet require much less fluid for an evaluation The results also shed some light on the parameters governing liquid-gas reactions.
the quartz window. Since the wire curved downward slightly in the middle, the liquid remained together as a hanging drop, only a small fraction being in close contact with the wire. In all cases, the driver gas was helium at 85 p.s.i.a., while the initial air pressure in the low-pressure section was variable between 2 and 4 p.s.i.a. The reflected shock pressure was between 60 and 80 p.s.i.a. The high gas velocity (generally over 4000 feet per second) behind the incident shock wave led to rapid breakup of the drop of fluid. Within about 0.3 millisecond after first reaching the drop, the shock wave had reflected from the end of the tube and produced a region of hot, stagnant air in which the small droplets of fluid evaporated and ignited. The zone of reflected shock-heated air remained intact for about 10 milliseconds. Since the fluid was in the reflected shock region for most of the 1 to 5 milliseconds needed for ignition, the temperature behind the reflected shock seemed to be the most meaningful single number by which to characterize the experimental conditions. The ignition time of a fluid sample was taken as that between passage of the reflected shock past the pressure transducer and the moment when the light intensity reached 10% of its maximum value. These ignition times have been plotted against the calculated temperatures of the air behind the reflected shock wave.
Experimental
Results
The shock tube used for these experiments has been described in detail elsewhere (6). Briefly, it is 3 inches in diameter, with 12-foot low-pressure and 20-foot high-pressure sections. Incident shock speeds are measured by timing the passage of the shock wave between stations 55 and 7 inches from the closed end of the low-pressure section. Gas temperatures and flow velocities are calculated from the incident shock speeds. A piezoelectric pressure transducer is mounted in the top of the tube, 3 inches from the closed end, while a fused quartz window-covered except for a 2-mm. vertical slit-is in the side of the tube, also 3 inches from the end. Light emitted by combustion in the tube is detected by a multiplier phototube 8 inches away from the tube, which also is covered except for a 2-mm. vertical slit. Because of the slits, only light emitted from gas 3 inches from the end of the tube is detected. A 0.01-ml. drop of the liquid to be tested was placed on an 18-gage copper wire placed crosswise in the tube just below
Figure 1 shows the ignition delays found for xylene and three hydraulic fluids. MS-2110-H is essentially a light hydrocarbon lubricating oil, while MIL-H-19457 is a phosphate ester, one of the most fire-resistant fluids available, of which the general formula is
HILE THE SHOCK TUBE
Present address, Dayton Campus, The Ohio State University and Miami University, Dayton, Ohio.
0
z=O, 1, or 2
Fluid AV, a proprietary hydraulic fluid of unspecified composition, was studied because it had a CFR rating close to that of xylene, but quite different physical properties. Since, in the engine test (7), the actual measured quantity is the compression ratio required to cause ignition of a spray of hydraulic fluid in 2 milliseconds' time as the piston is approaching the top of its stroke, it is appropriate to compare the VOL. 4
NO. 2
JUNE 1 9 6 5
147
]
“
.
.
)
O
*60
-
C 0
c
P
c
f .-Ea2 E 0
W
J
1200 1300 1400 Heated Air Temperature for 2 msec. Ignition, O K .
I
I 1150
1250
1200
Heated Air Temperature,
Figure 1.
1300
Figure 2.
OK.
Shock tube ignition of hydraulic fluids H MS-21 lo-H
0
MIL-H-19457
A
+
+
Xylene Fluid AV
engine compression ignition rating (expressed as a compression ratio) with the reflected shock temperature required to give ignition of the fluid in 2 milliseconds. Engine compression ignition ratings of the above four fluids have been measured ( 7 ) ) and have been plotted against the 2-millisecond shock temperatures in Figure 2. The correlation between the two tests is good, but the linear relationship found over the range covered experimentally cannot be expected to hold for higher or lower ratings. Of a number of exploratory hydraulic fluids tested, one of formula
0
I
Calibration curve, shock tube vs. engine test rating MS-2110-H MIL-H-19457
0
Xylene
A
Fluid AV
These drop breakup times do not correlate with the ignition data, since the breakup times of the most and least flammable liquids are the same, while the calculated breakup times for fluid AV and xylene, which have similar ignition temperatures, are different. Moreover, the boiling points of these latter two compounds differ considerably, being 325’ and 140’ C., respectively. Therefore, under these conditions the physical properties of the fluids are less important than chemical reactivity in controlling the ignition delays. The shock tube evaluation procedure seems to be a good one. I t requires less than 1 ml. of fluid per evaluation, an important factor when experimental fluids are being tested. I t can be used to study fluids that require compression ratios above 50 (the current limit of the CFR test) for ignition. Moreover, it simulates conditions in a hydraulic system at least as well as the CFR test does. Acknowledgment
had the high ignition temperature of 1365’ K. If the linear extrapolation of Figure 2 holds to this temperature, an engine test rating of about 80 is indicated.
The author acknowledges the suggestion, made by Gordon H. Ringrose, that the shock tube could be used to test the reactivity of hydraulic fluids; and the assistance of Edward S. Blake and Ralph E. DeBrunner, who furnished the experimental hydraulic fluids.
Discussion
There is, and probably will be for some time, a question as to the relative importance in the ignition process of the physical factors of drop breakup, evaporation, convective and diffusive mixing on the one hand, and chemical reactivity on the other. Morrell and Povinelli (4)have developed an equation for the time for breakup of liquid cylinders by shock waves, which should also apply approximately to drops. The breakup times of the 0.01-ml. drops of “standard)’ liquids used in the above experiments have been calculated, as follows.
148
Liquid
Calculated Breakup Time, Millisecond
MS-2110-H Fluid AV Xylene MIL-H-19457
0.24 0.31 0.81 0.25
I&EC
PRODUCT RESEARCHAND DEVELOPMENT
literature Cited (1) Brown, C. L., NAVENGRXSTA Rept. 95 648 C, “Fluid
Structure Factors Versus Fire Resistance,” Naval Engineering Research Station, Annapolis, Md., 1962. (2) Hanson, A. R., Domich, E. G., Adams, H. S., Phys. Fluids 6, ‘ ’1070 (1963). . (3) Morrell, G., “Eighth Symposium (International) on Combustion,” p. 1059, Williams and Wilkms Co., Baltimore, 1963. (4) Morrell, G., Povinelli, F. P., NASA Lewis TP 3-63; Paper presented at Fifth Liquid Propulsion Symposium of the Chemical Propulsion Information Agency, Tampa, Fla., Nov. 12-14,
(1959).
Id.Eng. C h m . 50, 53 (1958). Ruehrwein, R. A., J. Phys. Chem. 63, 1736 RECEIVED for review January 6, 1965 ACCEPTED April 15, 1965
Division of Fuel Chemistry, 147th Meeting, ACS, Philadelphia, Pa., April 1964.
