I/EC
Safety
Detonation Tests Evaluate High Pressure Cells Full scale prototype onstrates
by J. P. Weber/
performance
of a
new cell
design
Jacob Savitt,1 John Krc, Jr./' Armour Research Foundation;
Safety column in I / E C (October 1961, p . 52 A) featured background a n d design of a new type concrete high pressure cell developed at M o n s a n t o Chemical Co. T o determine how well the design objectives had been achieved, M o n s a n t o asked the A r m o u r Research F o u n d a t i o n of Illinois Institute of Technology to evaluate a prototype cell. M o n s a n t o built the cell at the A R F explosive testing facility at Coal City, 111., and explosive charges were detonated within the cell to simulate various conditions of sudden energy release. L A S T MONTH'S
The program was intended to determine the effects of 8 pounds of T N T detonated within the cell. For convenience in charge preparation, handling, and firing, Du Pont 50% ditching dynamite was used in place of TNT. The heats of detonation of T N T (p = 1.59 grams per cc.) and the ditching dynamite are 2010 and 1880 B.t.u. per pound, respectively. The heats of combustion are 6520 for T N T and 2500 B.t.u. per pound for ditching dynamite. The large heat of com1 2 3
Present address, Sandia Corp. Present address, Explosiform, Inc. Present address, Parke, Davis & Co.
128 A
dem-
and H. G Browne, Monsanto Chemical Co.
bustion of T N T results from an oxygen deficiency. The afterburning of the detonation products does not contribute to the air-blast pressure, but is of importance in creating rise of static pressure in a completely closed volume. Since the test cell has a large vent area per unit of volume, pressures within the cell are not sustained, and the increases due to afterburning will probably not be of sufficient duration to cause much increase in structural damage. For this reason the heat of detonation more truly represents the sudden energy release and is the better criterion for predicting air-blast damage to the cell and its surroundings. Thus, a given weight of ditching dynamite is only slightly less effective in causing damage than an equal weight of T N T . Electronic and photographic instrumentation was employed to obtain air-blast peak pressure measurements in the immediate neighborhood of the cell, shock velocity measurements in the immediate neighborhood of the cell (for pressure and shock position calculations), and air-blast peak pressure measurements at greater distances. In order to determine directly whether windows in nearby buildings might be broken by air blast, three glass windows were mounted in the test area. These windows
INDUSTRIAL AND ENGINEERING CHEMISTRY
were located at distances of 280, 500, and 800 feet. Two of the windows were scaled down to about one half the actual dimensions. For some of the shots, a 16-mm. Fastax camera was used to observe shock propagation from the cell, and the response of the cell and the safe wall to the blast. Framing rates were between 725 and 4660 frames per second, depending on exposure times required for the existing light conditions. Instrumentation
The gages used for obtaining pressure measurements were of the piezoelectric type employing barium titanatc. They were 1/t inch in diameter and were flush mounted in the centers of rectangular steel baffle plates having dimensions of Ve X 4 X 7 inches. The voltage output of the gages is directly proportional to the pressure, and is given bv
*-¥
where Ku is the gage constant, coulombs per pound per square inch : and C, total capacitance of the gage and the circuit in which it is used. Gage constants for the close-in gages ranged from 13.9 to 26.5 micro-microcoulombs per p o u n d per square inch, a n d for the distant gages they ranged from 76.5 to 196. Calibration of the gages was ac-
complished by subjecting them to several known pressures between 0.8 and 9.2 p.s.i. in the Armour Re search Foundation's 4-inch shock tube. The low pressure gages were also checked against a calibrated Altec-Lansing microphone system at pressures as low as 0.0012 p.s.i. generated by a speaker to check the gage linearity in the very low pressure range. The relation between the velocity of propagation and the pressure in a shock wave in air is given by
P. -1.16 Λ ( g - l ) where Ps is excess pressure of the shock front; P0, atmospheric pres sure ahead of the shock; U, shock front propagation velocity; and C, velocity of sound in still air ahead of the shock front. The measurement of the shock front propagation velocity, therefore, con stitutes a convenient method of de termining shock pressure, if the shock pressures are not too low. A foil-break probe system was used to detect the shock front. Holes 2 inches in diameter were drilled about 3 / \ inch deep at 2-foot intervals along a wooden two-byfour. One-half-inch wide aluminum foils, notched down to Vie inch in the center, were taped across the
Experimental
data led to these
openings. Pieces of Saran film were then taped over the foils. Breaking the foils unshorted resistors, resulting in voltage changes on the output cable. The circuit was balanced so that an up-and-down step func tion resulted as various foils were broken by the shock front. Because only one Tektronix 551 dual-beam scope was available for the first few experiments, the D, E, and F pressure gage signals were amplified and recorded on an ad ditional 4-channel oscillographic unit. A sequence timer was used to synchronize the recording camera on the 4-channel unit, the Fastax camera, and the firing of the charge. Shock arrival probes were used to trigger the sweep of the two Tek tronix scopes.
Test
Procedure
Six 1-pound charges were det onated in the cell, followed by one 2-pound charge, one 4-pound charge, three 8-pound charges, one 16pound T N T charge, and a 50-pound charge. The charges were all lo cated near the center of the cell floor and were supported about 2 feet above the floor by a wooden fourby-four. Number 6 electric blasting caps were used to initiate the dynamite charges, and firing voltage
was supplied by two 6-volt "hot shot" batteries in series. T o permit synchronization with the instrumen tation, the firing circuit was closed automatically at the instrumentation trailer. When high speed pho tography was employed, the firing circuit was closed at the camera remote-control unit. Positions of the D, E, and F gages were varied for some of the shots, and the A gage was moved in to 100 feet for a few shots. The number and position of shock arrival time probes were also varied during the experiments. The Β and C gages were located in the same positions for each of the experiments. The wooden wall in front of the cell, simulating the concrete safe wall, did not withstand the effects of charges of 2 pounds or more. The wall was rebuilt twice, and after the first 8pound shot it was no longer used. Test Results
Except for damage to the side door mounts, no damage to the cell was detected until after the first 8-pound test. In this shot, some concrete fell from an area across the top of the vent opening or mouth of the cell. The discolorations in the concrete in this area indicated that the concrete had been improperly mixed or settled.
conclusions:
• The test cell will not be seriously d a m a g e d by a sudden e n e r g y release equivalent to the d e t o n a t i o n o f 8 pounds of TNT. The cell damage will p r o b a b l y consist o f damage to the side d o o r supports, interior vent, and curtain support assembly. Some c o n c r e t e may be loosened from a b o v e the cell mouth. H o w e v e r , it is n o t e d that this may have o c c u r r e d in the evaluation p r o g r a m only because o f the inadequately cured c o n crete in this a r e a . If fragments are p r o j e c t e d by the explosion, spoiling o f c o n c r e t e from the outside of the walls may occur, but the rear spall plate w i l l stop the spoiled c o n c r e t e .
• The pressures at w i n d o w s S a n d C may be in the neigh b o r h o o d o f 0.020 p.s.i., but environmental conditions such as terrain, cell geometry, atmospheric conditions, and reflection factors could increase these pressures several times. Pressures at w i n d o w A will be much less due t o the increased distance and the d i r e c t i o n a l effects of the blast. The results from the Β and C gages are p r o b a b l y conservative, h o w e v e r , since in the real situation inter vening obstacles, such as automobiles o r trees, w o u l d constitute many more reflecting surfaces than w e r e present in the evaluation a r e a .
• For an energy release equivalent to 8 pounds o f T N T , fragments and projectiles will be confined to the area o f the cell.
• The safe w a l l does effect a reduction in pressures occurring at w i n d o w s Β and C. This is p r o b a b l y due to absorption of blast energy by the w a l l , a g g r a v a t e d by d i r e c t i o n a l effects, and by the c r e a t i o n of multiple shocks.
• For an e n e r g y release equivalent to 8 pounds the peak pressure o f the incident blast w a v e a t w a l l is a b o u t 25 p.s.i., and at the Ο g a g e near o f the cell the pressure due to reflection from w a l l will be in the n e i g h b o r h o o d o f 7 p.s.i. The behind the safe w a l l will be v e r y much less than but the e x a c t value has not been determined.
of TNT, the safe the side the safe pressure 25 p.s.i.,
• An energy release equivalent to 16 pounds of T N T will cause cracking o f the structure, but no gross d e formation. Pressures near the cell and at the w i n d o w stations w i l l be a p p r o x i m a t e l y t w i c e the values r e p o r t e d f o r the 8-pound charge.
VOL. 53, N O . 11
·
NOVEMBER 1961
129 A
Ε GAGE F GAGE. D GAGE
Schematic d i a g r a m of instrumentation system M a p shows layout of test a r e a
The second and third 8-pound charges caused only slight additional damage to the cell. The overhead curtain support rail was torn off, a small patch of concrete fell from the inside of the rear wall opposite the charge position, and some ad ditional concrete fell from above the mouth opening. The 16-pound T N T charge caused several cracks in the side walls and the interior wall but no significant deformation of the basic cell structure. The 50pound charge, which was consider ably in excess of the energy design limit of the cell, did cause extensive structural damage ; however, frag
ments from steel members placed on the charge were not projected from the cell. The motion picture records of the T N T test show afterburning of the product gases outside the mouth of the cell. It is also apparent that much unreacted carbon was present. This information supports the opinion that afterburning within the cell does not result in significant ad ditional damage to the cell structure and that the energy release criterion should be the heat of detonation rather than the heat of combustion. In the experiments involving charges of 4 pounds and larger, the
pressures recorded from the gages located near the cell seemed too low, and the shock propagation velocities indicated by the foil-break probes were only slightly above sonic. Fastax films of the experiment re vealed that the recorded pressures were lower than expected because two shocks were produced. A weak shock preceded the strong shock from the mouth of the cell. This weak shock apparently is caused by leakage around the edges of the baffle plate in the cell and through
Continued on page 132 A
CHARGE WEIGHT. LB
Figure 1. 130 A
Peak blast pressures
INDUSTRIAL AND ENGINEERING CHEMISTRY
Figure 2.
Peak blast pressures at windows Β and C
SAFETY
AND
SAFES in a HAWS Emergency Drench Shower. Burning, corrosive, caustic contamination can inflict injuries more dangerous than blazing clothing! Contamination by acids, chemicals, volatile fuels, radioactive elements, etc., must be instantly countered by first aid. Immediate drenching with clear water is the first precaution against permanent injury. HAWS leads in design and production of Emergency D r e n c h Showers! Ask for our complete catalog.
Model 8590 Multiple nozzle shower drenches victim from all angles.
HAWS
DRENCH SHOWERS Since 1909 a product of HAWS DRINKING FAUCET COMPANY
1443 Fourth St., Berkeley 10, Calif.
Circle No. 51 on Readers' Service Card
TYPICAL REMOTE CONTROL FLUIDOMETER SYSTEM For Use Where Meters Must Be Located In Hard-to-get-to Places This animated picture illustrates a typical remote control Fluidometer application. Re mote control of this general type is recom mended when plant conditions or arrangement make direct control impractical or impossible. With a system of this type the Fluidometer con trol head can be located away from the meter. Shown at the left is the automatic resetting control head, which controls the operation of the shut-off valve to give completely automatic batching. Available either jacketed or unjacketed—"tailor made" to fit your needs. New Bulletin Fl-56 will be sent on request. For in formation on jacketed pipe and fittings write for Bulletin J-56. H E T H E R I N G T O N & BERNER I N C . ·
ENGINEERS-MANUFACTURERS
710 KENTUCKY AVENUE INDIANAPOLIS 7, INDIANA A Wholly Owned Subsidiary of American Hoist & Derrick Company Circle No. 52 on Readers' Service Card
132 A
INDUSTRIAL AND ENGINEERING CHEMISTRY
the vent pipe near the bottom of the interior wall. Pressures of the strong shock were calculated in some cases from measurements of the shock velocity taken from the Fastax film. I n the shots for which pressure gages were located 7 feet a n d 11 feet from the cell mouth, the weak shock signals were recorded, but the strong shock drove the traces off the scope. T h e time of arrival of each shock a t each gage was known and, therefore, the propagation velocities a n d shock pressure could be calculated. T h e peak incident blast pressures 9 feet from the m o u t h are shown as a function of charge weight in Figure 1. Included in this plot are the pressures measured 1 foot from the south wall, 5ι/ΐ feet back from the m o u t h . T h e direct blast wave reach ing this point is less intense t h a n the blast reflection from the safe wall. T h e pressures behind the safe wall were not readable for the 1-pound charges, a n d for larger charges the meaning of the readings from this position is questionable because the safe wall was blown down. T h e m a x i m u m pressure readings obtained at the window positions Β and C are shown as a function of charge weight in Figure 2. No pressures were recorded at the A position 800 feet north of the cell, although in some tests pressures were recorded 100 feet north of the cell. Pressures arc lower for charges fired within the cell than for charges fired outside the cell. Also, pres sures a t Β and C were found to be lower when the safe wall was present. For a 2-pound charge of T N T fired in free air, the expected pressure at a distance of 280 feet is 0.190 p.s.i., a n d at 500 feet it is 0.105 (2). T h e measured values are lower than these figures. This is probably be cause blast waves of short duration are more readily attenuated by surface irregularities than those of longer duration. Also, the snow and brush may have been responsible for some reduction in pressure. F u r t h e r m o r e , when charges are fired in the cell, some energy absorption takes place, and shock reflection p h e n o m e n a are present. Multiple shocks were recorded at Β and C positions on m a n y tests. If reflec tions in and near the cell h a d not generated multiple shocks, the single blast wave might have had a higher peak pressure.
SAFETY Although the pressures measured in these experiments were lower than those theoretically predicted, other workers have reported even lower pressures under similar conditions. Data obtained by the University of Utah indicate that great variation in pressures may occur at long dis tances (3). Data on surface shots presented by Cook show that pres sures at considerable distances may be as small as one eighth of the values predicted on a still free air basis (7). In view of the behavior of air-blast waves at large distances and of the effect of environmental conditions on their propagation, it would seem reasonable to assume that the pres sures under some conditions might be several times as great as those measured during this program. T o the extent that the ground surface represents a dissipative mechanism, one would furthermore expect that initial pressures measured some dis tance above the ground would be somewhat higher than the initial pressures measured here. The windows were not broken by air blast from any of the firings up to and including the 50-pound charge. This is a good indication that the 8-pound T N T energy re lease is extremely unlikely to cause damage to windows as near as the minimum distance tested. T h e re sponse of large glass panes to mul tiple shock waves is quite compli cated, however, and unusual stress concentrations in the glass and im proper framing should be avoided. Literature Cited (1) Cook, Μ . Α., " T h e Science of High Explosives," p . 368, R e i n h o l d , New York, 1958. (2) Stoner, R. G., Bleakney, W., ./. Appl. Phys. 19, 670 (July 1948). (3) University of U t a h , Institute for the Study of R a t e Processes, Explosives R e search G r o u p , First A n n . R e p t . of F u n d a mental Investigation of Air Blast a n d G r o u n d Shock, T e c h . R e p t . I I , April 1. 1957. Division of Industrial and Engineering Chemistry, Chemical Safety S y m p o s i u m , 140th Meeting, A C S , Chicago, 111., S e p t e m ber, 1961.
Our authors like to hear from readers. If you have questions or comments, or both, send them via The Editor, l/EC, 1155 16th Street N.W., Washington 6, D.C. Letters will be forwarded and answered promptly.
Pulsafeeder meters "skunk essence" with never a whiff of leakage Metropolitan Utilities District of Omaha adds skunk-like odorants to natural gas to warn customers of leaks. A t first there were problems: odorization varied with gas flow rates; odorizing equipment often broke down under high loads; neighbors com plained about leakage of the odorant. T h e n , four years ago, Lapp Auto-Pneumatic Pulsafeeders took over the job. Since then, pump leaks have been non-exist ent. Adjustments that used to take hours are made in seconds. Maintenance costs 9 5 % less. Odoriza tion is absolutely uniform. And based New, Informative Film on present conditions, Metropolitan Utilities estimates their Pulsafeeder Pump life to be over 20 years. Tells the story of safe, ac curate, metered-pumping. If your material is hard to handle Now, know all the facts . . . if you're looking for accurate from origin—to develop m e n t — t o a p p l i c a t i o n of metering . . . if high maintenance costs Pulsafeeder Metering are eating into your profit. . . it's time Pumps. to put Pulsafeeder to work for you. 30 enjoyable minutes in
"HARD TO HANDLE"
sound and full color. Write for further details and ar rangements for showing at your plant or outside loca tion of your choice.
WRITE for Catalog 59. Lapp Insulator Co., Inc. Process Equipment Division, 1423 Poplar Street, LeRoy, N e w York.
Lapp Circle No. 20 on Readers' Service Card VOL. 53, NO. 11
·
NOVEMBER 1961
133 A
