With the selection of the ANFO and AN material decided, the plans for testing ANFO charges in sizes up to 4000 lbs continued. Realizing that both premixed and field mixed ANFO could be used but not having the experience or the foresight to know which method would be most adaptable to military testing requirements in the 500 ton range, both techniques were used. Sadwin, Pittman and a small field crew completed a 23 shot ANFO series in May 1968 in the Rattlesnake Flats area about 18 miles southwest of Hawthorne, Nevada.
Charges weighing 260 , 500 , 1000 , and 4000 lbs were fired. Again, with no guidance available, but to assure detonation of the ANFO, the charges were. boostered with cast Pentolite cylinders weighing 8 , 15, 24 , and 32 lbs respectively'. Tire boosters were placed at the center of the ground plane of the charges. All but one of the charges used loose, unbagged ANFO piled on the ground in roughly hemispherical shape. A thin corrugated paper fence was used to retain the lower portion of the pile.
One 1000 1b charge was made by stacking the 50 lb.bags of premixed ANFO in as close to a hemispherical shape as possible, (the 20 bags required made this stack resemble a cube more than a hemisphere). Five charges used the ready mixed ANFO; the other eighteen charges were made with field mixed ANFO with the mixing being done in a 4.5 cubic foot cement mixer (Figures 1-11 and 1 12). The emphasis on field mixing stemmed from the thought that in. the 500 ton size, field mixing would be less expensive than buying ready mixed ANFO.
Three 238 1b hemispherical TNT shots were fired as part of the test program so that a basis would be available to fudge the performance of the ANFO shots. The TNT shots also permitted a check on the operation of the airblast instrumentation. The instrumentation used in the program was minimal but adequate. Airblast was measured in the range from about 1.0 psi to 30.0 psi. A two point probe was used to indicate detonation velocity within the charge. High speed, (4,000 and 7,000 frames per second) cameras were used to observe the explosion. And after each shot, crater size measurements were made.
The results of the test program indicated that unconfined ANFO charges of about 260 1bs are required before stable detonation and blast conditions could be achieved. This was evidenced by the scalability of the pressure distance data for charges weighing from 260 lbs to 4,000 lbs and the leveling off of the TNT equivalence of ANFO for charges weighing more than 260 lbs. These results confirmed the suspicion (and hope), that the smaller charges used in the earlier bootleg tests were not sufficiently large in diameter or weight, to permit steady state conditions to be realized. The new data indicated that ANFO had an average detonation velocity of 4200 meters/second and an average equivalent weight, on a pressure basis, of 0.82 compared to TNT. The positive duration and impulse of the blast wave, although not stated in terms of equivalence, appeared to be somewhat less.
The detonation velocity data was heartening; the literature gave the detonation velocity of stoichiometric ANFO, in heavy, confining steel pipes, as 4200 meters/second. The 0.82 TNT equivalence of ANFO, however, was a bit disappointing a higher output would have been desirable. This relatively low output is attributable in large part to the fact that ANFO is in stoichiometric balance; it does not depend upon or utilize atmospheric oxygen in the explosion process as does TNT. TNT is extremely oxygen deficient; for maximum output, it depends an atmospheric oxygen to continue the combustion process. This, afterburning leads to longer duration blast waves and higher impulses. Some thoughts again were given to the use of additives in the ANFO to increase output, but again, they were dismissed as being too costly, tire consuming, and complicated.
The initial disappointment was soon mollified when it was considered that on one hand; there is nothing sacred about the size or yield or type of charges used on military tests; no matter which chemical explosive source is used and no matter how large a yield is realized, chemical explosions only simulate in a limited range some of the effects of nuclear weapons. Through knowledge of the explosion processes of both nuclear and HE sources, analysts can relate one to the other and utilize the simulated effects to advantage. And on the other hand, because of this analytical ability, whether the HE yield is 80, tons or 120 tons, test results can be interpreted as if the environment were produced by 100 tons. As a matter of fact, for charges with yields 18 7.0% different, at a given pressure level, the differences in distances at which this pressure occurs are less than 7%. The converse is almost true also; at a given distance, the pressure difference is about 7%. This is less than the scatter in measured data usually obtained in field operations. So, the chemistry of ANFO was accepted along with its reduced blast output, i.e., its 0.82 TNT equivalence.
The cratering performance of the 260 1b ANFO shots were compared to that of the TNT control shots: The ANFO crater radius was 7% smaller than that for TNT, its depth 9% greater, and its volume about 20% less. In light of the caliche material located about 1.5 ft below the surface, these comparisons were considered satisfactory.
NSWC was ready to move on to the follow on tests suggested in its March 1967 proposal to demonstrate the merits of still larger charges and to explore ways to utilize the charges for tests at sea as part of the Navy's hardening program. A Phase II ANFO proposal was submitted to NAVSEC in July 1968.
It was proposed that the Phase II work consist of two 20 ton and one 100 ton ANFO shots with pressure time and high speed photographic instrumentation providing the main measurements coverage. A one year program with six weeks in the field was contemplated. The estimated total cost for the program was to be $115,000. It was pointedly noted that the cost of TNT alone for a 100 ton charge would be about $200,000. The proposal suggested that a Phase III program with charges weighing up to 500 tons would be necessary (after the successful completion of Phase II) to adapt the ANFO simulation technique to sea trials for the Navy's ship hardening program and to other DNA uses.
NAVSEC heartily endorsed the proposal in October 1968 with the statement, "One important aspect of airblast hardening is testing and evaluation of equipment to determine if design specifications have been met and to locate areas of weakness. This may be accomplished for components by testing them in the conical shock tube and for complete systems and sub systems by exposing them to simulated nuclear airblast from chemical energy explosion sources. Presently used energy sources (TNT, detonable gases) are expensive, over sensitive, inconvenient, and impractical for specialized purposes such as sea operations. The Navy, however, has recently experienced a breakthrough in this area with the proposed utilization of ANFO an inexpensive, insensitive, and versatile explosive for this purpose". They forwarded the proposal to DNA for direct funding. Kelso, who, of course, had been following the program of the ANFO endeavor,, provided DNA funds for the task in January 1969.
A three shot program was undertaken with two 20 ton and one 100 ton ANFO charges. The field program, under the direction of Sadwin, was conducted at DRES, the site of the early TNT and detonable gas large scale experiments and military equipment tests. The initial objectives of the DRES ANFO program were to establish the scalability of ANFO charges from the 260 4000 lb range to the 20 100 ton range, and to gain experience in handling, mixing, and preparing large charges with the ultimate goal of fielding still larger charges, up to the 500 ton size considered useful for military tests.
By this time, i.e, early 1969., the demonstrated and potential merits and applications of ANFO were being recognized by an ever increasing number of military scientists. More information on the explosion effects of ANFO was being asked for than could be readily provided by NSWC itself. Four other U.S. agencies and DRES participated in the Phase II ANFO test program to help get this additional information.
DRES provided field, support and made blastwave time of arrival, crater size, and photographic measurements on each of the shots. BRL (Ballistics Research Laboratories) made side on and total head airblast measurements in the high, i.e., up to about 1000 psi, and moderate pressure regions. USGS (U.S. Geological Survey) made cratering studies and NWC (Naval Weapons Center) and NCEL (Naval Civil Engineering laboratory) made measurements on above ground and underground structures respectively: NSWC, in addition to directing the over all operation, was in charge of charge design and preparation, charge monitoring, e.g., determining the internal temperature, oil content and prill size of the ANFO charge, and low pressure blast measurements.
The ANFO used on all three events was supplied by a Canadian source located in Calgary, Canada. All mixing and bagging was performed on site using techniques and equipment developed by and for the mining industry. For the first two events, the 20 ton shots, AN was delivered to a Suffield railroad siding in a 70 ton capacity hopper car. The AN was augered into a mixing truck where the FO was introduced in correct and metered proportions. The 7 ton capacity mixing truck was driven. to the GZ (ground zero) area where, for Event I, the ANFO was augered into a bagging unit, and for Event II, the ANFO was augered directly into a fiberglass hemispherical container. For Event III, the 100 ton shot, it was found rare efficient to have the AN brought to the GZ area in 22 ton capacity tanker trucks; auger the AN into the mixing truck, and the ANFO directly into the fiberglass container. In the mixing operation the diesel fuel oil was colored with a red dye so that a continuous visual check could be made of the fuel oil content of the ANFO; a change in color tone of the ANFO would indicate a change in the FO proportion. The fuel oil content was periodically checked also by chemical analysis. For these three charges, the percentage of FO was found to vary from 5.85 to 5.95, acceptable limits to provide a stoichiometric mixture of ANFO.
The 20 ton hemisphere for Event I was formed using the bagged ANFO. A rather smooth hemispherical surface contour that was formed by the pliant bags. This was considered to be an advantage over the reentry cornered surface of TNT block built; reentry corners and planer surfaces were considered to be a possible cause for blast anomalies. Eight hundred 50 lb bags were used in this charge. One hundred fifty of these bags were opened and the loose AWO used to fill the interstices between the full bags. This was done to provide as homogeneous mass of explosive material as possible. It is remembered that one postulated source of anomalies in TNT block charge construction was the non-homogeneity of the charge, particularly at the interfaces between the blocks. The loose ANFO between the bags was aimed at reducing charge construction induced blast anomalies.
Event II used bulk ANFO contained in a thin fiberglass hemispherical envelope open at the top to permit filling. This construction was used to determine the merits of bulk loading, the effects of light containment, and the difference between field operations in terms of time, difficulty, and cost for bagged vs bulk ANFO charges. Event III, the 100 ton charge (Figures 1 20) was built similarly to the Event II charge; its primary objective was to establish the scalability of large charges of ANFO. Each charge was boosted by a 250 1b hemispherical TNT/pentolite charge placed at the bottom center of the main charge and initiated with 100 grain per foot primacord.
Starting on 14 August and continuing on a weekly basis through 28 August 1969, the three ANFO shots were fired. Prior to the first firing, there was some speculation even wagering among the test participants and observers as to whether the charge would, in fact, successfully detonate or would, succeed instead only in spreading fertilizer over the DRES plains. DRES, for instance, remembered that in the late 1950s in its general studies of explosive materials, it had investigated briefly the properties of AN based explosives; it could not reliably detonate the small charges used. Others were aware of the heavy and multi point boostering used by the mining industry for ANFO confined in bore holes.
The first charge and the subsequent charges detonated successfully. Analysis of the test data by NSWC showed the reproducibility and scalability of the explosion effects. The NSWC blast data averaged over the 1 to 200 psi range indicated that the average TNT equivalence for the 20 100 ton ANFO shots was 0.94 for both the bagged and bulk charges. This is considerably higher than the 0.82 equivalence reported for the earlier 200- 4000 1b shots.
There are several reasons for this apparent but not necessarily real discrepancy. The earlier 0.82 equivalence was established over a 1 to 30 psi range (using the only data available) and a linear weighting method was used. Because the linear method gives undue emphasis to the data at the higher pressures, and pressures up to 300 psi were recorded for the Phase II shots, a logarithmic weighting scheme was used for the Phase II data. Using a common system for both the Phase I and II shots, i.e., logarithmic averaging over the 1 to 30 psi range, the Phase I data gives an equivalent weight of 0.86, the Phase II data 0.87.
Equivalent weight determinations may be inappropriate not only for ANFO but for any explosive comparisons unless statistically significant number of shots can be fired. For one, by quoting a single number, the illusion is given that (in the ANFO example), the ANFO pressure distance curve is parallel to that of the reference TNT curve. This is not so over the whole pressure range of interest. Hence, depending on the pressure level of particular concern, different equivalencies can be calculated. For another reason why equivalent weight, numbers should be used with some trepidation, equivalent weights, as determined from pressure distance comparisons, are extremely sensitive measures of the merits or yields of one explosive compared to another. A 20% difference in yield or equivalency leads to about a 7% difference in pressure at a given range. On a single shot or a small number of shots, this 7% difference is hardly discernible because of the scatter in the data. The results of either calculations or measurements for a nominal 6 ton charge vs an 11,242 lb charge are hardly discernible and have little significance.
Detonation velocity, deduced from DRES photographic measurements and ionization probes, indicated as average velocity of 4470 meters per second, about 5% higher than that obtained in the earlier 260 to 4000 lb Phase I Program. Crater measurements showed reproducibility in crater dimensions produced by the two 20 ton shots, and close agreement with the crater produced by a 20 ton hemisphere of TNT fired in the same area. A comparison between the 100 ton ANFO shot and a 100 ton TNT hemispherical charge indicated marked differences in crater radii and depths but only a 15% difference in estimated crater volumes; the TNT crater was wider, shallower, and had a larger volume. A major reason for these variations was attributed to the geologic formation underlaying each shot; the TNT crater struck water while the ANFO one did not.
Photographic coverage of the explosions shoved the presence of anomalies essentially only in the bulk loaded, fiberglass contained charges. The TTCP working group studying anomalies had access to a Phase II ANFO results. In its report, it concluded "An ANFO barge built with stacked bags produced no anomalies attributable to the charge material; however, some Type 5 anomalies attributable to charge construction lion were evident. A Type 5 anomaly is one in which a fireball perturbation affects .the shock front. Some anomalies of all types were observed on the cased ANFO charges; these were considerably less in magnitude and extent than those observed on similar sized TNT charges."
The measurements of the internal temperature of the charges showed that there was no internally generated self heat; only small variations in temperature occurred and these were associated with diurnal air temperature changes. The ANFO was shown to be a stable mixture; there was no evidence of the fuel oil settling out of the mixture.
The field operations provided the sought after experience in handling large ANFO. The ease of charge preparation was demonstrated by the short span of time, fourteen days, required to prepare and fire three shots. The costs of bagged ANFO and fiberglass contained bulk ANFO charges were about the same, with the cost of the container equaling the cost of the additional manpower required for the bagging and stacking operations.
During the early 1970s, ANFO spheres were being investigated with DNA support for use as a direct counterpart for the TNT spheres. In October 1971, two 25 ton spherical ANFO charges were fired at MS. These charges, designated ANFO IV and ANFO V (and considered to be follow on to the 1969 three shot series at DRES) used bagged ANFO. One charge (ANFO IV) was constructed tangent to the ground, the other was half below half above the surface. Limited by a particularly austere budget, it was not possible to make as extensive a measurement effort as was possible for the three earlier large ANFO shots, but airblast, crater size, and photographic measurements were made. The test data were sufficient to provide judgment on the performance of spherical ANFO shots. Comparison could be made directly with similarly configured TNT charges fired at the same site.
It was found that ANFO IV and. V, with their 0.82 TNT equivalence, produced the same blast as 20 ton TNT shots. Some blast anomalies were observed by DRES on ANFO IV, none on ANFO V. It was conjectured that the anomalies stemmed from the somewhat asymmetrical construction of the charge, the rough outer surface created by the bag construction, and the possibility of air pockets entrapped in the ANFO bags. In the 1969 series, the bags were contoured into a relatively smooth outer curve by butting the ends of the bags together; on the 1971 tests the outer bags touched only at the inside corners. The crater obtained on ANFO V matched very closely the dimensions of the DISTANT PLAIN crater produced by a 20 ton TNT spherical charge half buried while the ANFO IV crater dimensions fell half way between those of DISTANT PLAIN 5A and 6A, 20 ton spherical TNT charges tangent to the surface; the ANFO IV crater had a volume of 16,540 ft3 , the DISTANT PLAIN, 5A crater 24,087 ft3, and the DISTANT PLAIN 6A crater 7,064 ft3. It is important to note that both DISTANT PLAIN shots were fired in the same area at DRES, the ANFO IV shot in another area. The difference in crater sizes for the similar TNT shots is an indication of the problems associated with crater (and ground motion) predictions: even small differences in geologic structure and materials can lead to large differences in actual test results.
These difficulties in ground motion and crater studies were stated by D. S. Randall, PI, after PI ran a four shot spherical ANFO test series in November 1971. Each charge consisted of 1200 lbs of ANFO contained in a hollowed out styrofoam cube resting on the surface.. Two shots were fired over a silty playa material; the other two over a 10 ft layer of clay above shale of unknown thickness. Each pair of ANFO shots was compared to a 1000 lb TNT shot of similar geometry fired at each site. Craters produced by the ANFO charges were neither consistently larger nor smaller than craters produced by TNT charges of the same yield. Essentially, this is the same result as obtained at DRES in ANFO IV and V. Hence, Randall's statement, in November 1972, "Craters produced by equal energy charges of different explosives are so strongly influenced by the characteristics of the test bed that no general predictive relationship between charge mass and crater size can be generated at this time."
Incidentally, PI was well aware of the airblast and fireball anomaly problem; significantly they note that on their 1200 1b ANFO sphere tests, "The fireballs expanded spherically without any evidence of anomalous behavior." The spherical ANFO tests conducted in October-November 1971 certainly did not provide data that could allay the concerns of the ground motion and cratering community. However, these data, in being compared with TNT results, did tend to bring into sharper focus than in the past, the whole ground effects problem i.e., the dependence of effects on ground geology and the difficulty of determining this ground geology with sufficient resolution to permit accurate predictions. Unfortunately, ground characteristics are not as easily defined as atmospheric characteristics.
|Join the GlobalSecurity.org mailing list|