One-meter-diameter flying model: firing-system schematic. Note ballistic envelope labeled "hull."
"We got chased out of there," says Jerry Astl, describing the abandonment of Project Orion's first experimental test facility on the mesa above La Jolla at Torrey Pines. "Because every time we made a few shots the other employees got upset and began to squawk." Thanks to Carroll Walsh, the Orion group soon found another site. Walsh, the project's internal technical ombudsman, could find anything—from high explosives to transportation for Queen Juliana of the Netherlands, who accepted an invitation to visit General Atomic to view a TRIGA reactor but requested a four-engine plane be made available for her flight. "I called the Navy at the air base, and I said, 'The Queen needs a four-engine plane. You set it up, we need to have it; she won't fly in anything else.' And they said, 'Let me think about this.' I had to talk to an admiral. Damned if he didn't arrange it! And he had the plane sitting there and he had written orders for about fifty captains and so forth to justify taking this big plane."
Whatever needed doing, Walsh could get it done. "Freddy asked me to write three pages about the Saturn-boosted Orion vehicle, and get someone to fly it out to Gardner on the night plane," Ted noted in his journal at the end of October 1960, when Trevor Gardner, chairman of the Air Force committee preparing to advise the incoming administration in Washington on space, decided to include a small version of Orion in the options from which it was hoped the new president would choose something to announce. "Felt like a heel when Wild and I asked Carroll Walsh to take it out."[158]
Walsh, born in 1915, grew up in Oakland, California, just before the depression and was making his own gunpowder and launching homemade rockets by the age of twelve. He joined the Civilian Conservation Corps in 1933, building fire trails in Yosemite and learning to work with dynamite while earning one dollar a day—keeping five dollars a month and sending his mother and father the rest. After eighteen months in the CCC he enlisted in the Navy, starting as a second-class electrician's mate and advancing to become an officer during World War II. He had a knack for electronics and was comfortable around high explosives and nuclear bombs. "Guess who had charge of taking an atomic bomb and giving it a hammer test? Have these big weights hit it and see what happened? I did," he says. "I had Q clearance before I ever came to General Atomic. And remember the big experiment where they took all the ships out to sea out there in the Pacific, and they had the Big One and were going to see what happened to all those ships? Who was in charge of that stupid bomb? It was an enormous thing—as big as this room! I was." Walsh spent eight years working at the Navy Electronics Laboratory at Point Loma, developing an uncanny knowledge of who, among the network of San Diego military bases, had what. "Out at the test site, we needed an automobile. I could get one from the Navy for nothing. All of a sudden they needed a forklift. I knew right where to go get one. For nothing."
Point Loma is a steep, narrow, three-mile-long peninsula that guards the entrance to San Diego harbor and had been fortified with a series of gun batteries during the war. When the guns were decommissioned the area remained a naval reserve, adopted by the Navy Electronics Laboratory for radar experiments and other secret tests. One of the guards had been Lieutenant Commander Walsh. "There were all these big sixteen-inch gun emplacements—cement walls six-foot thick. I knew all about those places because I used to have to patrol those on my night duty. We had a lot of work going on that was secret and nobody was supposed to be there." Walsh persuaded his former colleagues to grant General Atomic the use of Battery Woodward and Battery Ashby, overlooking the Pacific Ocean on the west side of the peninsula—complete with ammunition bunkers and abandoned gun emplacements that were ideal for blowing things up. There was also a 75-foot-high steel tower that had been built for 1/4-scale static tests of Atlas missiles, with a blockhouse that had been used for monitoring the tests.
"When we got to Point Loma, the Navy treated us so well, it was just like home," explains Jerry Astl. "Those bunkers! They had held sixteen-inch guns. We got whatever we wanted. But the launch test tower had huge steel platforms on three sides, so that they could shoot off the missiles at different angles. We asked the Navy about taking them down, and they said sure, they also wanted them down—but it would take two or three months before they could bring in a big crane to do the job. But we didn't have two months, we needed them down yesterday! So I said I could do it. And Carroll Walsh gave me just enough C-4, and long lead detonators, to do the job. And we had it down in two hours! It was just like if I was back in World War II."
Astl insisted on long lead detonators, resistant to radar-induced currents that might make regular detonators explode. "Detonators with long leads shall be used whenever possible," it was stipulated in General Atomic's Safety Procedures at Point Loma Experimental Site, compiled by Brian Dunne and issued on July 17, 1959. "Detonators shall at no time be carried about or placed in pockets of clothing," and anyone attaching a detonator to an explosive charge, long lead or not, "shall short the firing line to check for sparks from possible stray current and shall then touch the ends of the line simultaneously to his tongue for small currents."[159] No experiments involving high explosives were to be conducted except under the supervision of a designated "physicist-in-charge." During the entire project, there were only a few minor accidents with high explosives and no serious injuries as a result.
"They were killing people at Los Alamos," says Dunne, referring to the mishaps with ordinary high explosives that far outnumbered mishaps with fissionable material in fabricating nuclear bombs. "I studied the accidents that had happened at Los Alamos, and you come away chastened from that." Dunne limited the amount of high explosives that could be stored in the General Atomic magazine at Point Loma to 400 pounds. "For charges involving shrapnel and/or charges of one-half pound or more, and for all tests in the tower," he ordered, "GA men shall be sent to block the access roads on each side of the experimental site."[160] One of the access roads was within view of Point Loma College. "The real attraction there was a college for girls, that's why everybody, especially the young technicians, wanted to work on Point Loma," Astl explains.
In formalizing Walsh's arrangement for use of the Point Loma site, General Atomic explained to the Navy that "the tests involve the effects of small explosions, of the order of 10 grams to perhaps a kilogram, on certain materials."[161] Besides building mock-ups of gas-bag shock absorbers and developing explosive-driven plasma jets, the initial tests included hanging pusher-plate test specimens from the tower, restrained by flexible cables to keep them from flying off into the distance with every shot. It did not take long for the Point Loma crew to start wondering what would happen to a pusher plate, attached to a rudimentary payload, when a small explosive charge propelled it up.
There are two surviving flying models of Orion. One of them, one meter in diameter, is in the Smithsonian Institution's National Air and Space Museum in Washington, D C. The other model, one foot in diameter, is in Jerry Astl's garage in Solana Beach. "We started out very slowly with a one-foot-diameter model, with pressure gauges all around," explains Dunne. "The charges were tiny—about golf ball size. We did Fastax film and got some idea of the shock wave scaling laws." It is this model—intermediate between the mixing-bowl experiment and the one-meter-diameter model now in the Smithsonian—that Jerry Astl has kept above the rafters in his garage for thirty-two years, rescuing it from destruction, thanks to security guard John Iles, when General Atomic, following its purchase by Gulf Oil, was cleaning house. Its fiberglass shell is a beat-up, faded Air Force blue, in the shape of a squat bullet (or "Anita Ekberg Maidenform bra," as Astl says it was known among the technicians) modeled at 1:130 scale after one of the original 4,000-ton designs. Its aluminum pusher plate is scarred from ed blasts, and the polyurethane foam shock absorber has crumbled almost entirely to dust.
"I made it in three or four days," says Astl. "I found a guy in National City who was an absolute artist in fiberglass. Old man, grouchy like hell, but you go with hand sketches, you tell him something that will stimulate his interest and he will drop everything, all his customers have to wait, because he is going to make this bullet for me, and for a few dollars." The fiberglass shell is perforated from top to bottom with a series of holes for pressure gauges to measure how the shock wave refracting around the edge of the pusher plate impacted the hull. "It's good to know about shock wave scaling in this business," says Dunne. Another series of holes was situated to measure stress within the shell. "We were trying to get some idea if the charge is slightly out of line, how much torque is generated," says Astl. "It looks stupid, but it generated a lot of good information."
By February of 1959, the one-foot-model had been followed by the first of a series of one-meter tethered models, which were subjected to single shots. "I was shown a number of interesting movies of the 1/40 scale model which they are testing," reported Don Mixson after a visit to General Atomic on February 25, 1959. "The model weighs about 160 lbs. and employs a steel pusher plate. The scaled explosive charge is 2 lbs. of Composition C-3 molded into a roughly spherical shape. The model has also been overtested with a three-pound charge." The purpose of the tethered models was to investigate structural design, not ablation, yet Mixson noted that "no evidence of any pusher ablation has been observed. This also applies to the 1/130 model which shows no ablation after more than 50 shots."[162] One month later, AFSWC project officer Ron Prater made another visit and reported somewhat less optimistically that "testing of the 3-foot model is continuing at the Point Loma facility. It is doubtful if further testing of this model will yield significant results."[163]
Under General Atomic's original contract with ARPA, the construction of a free-flying model was out of bounds. But by 1959 the Orion physicists had convinced themselves of Orion's theoretical feasibility and wanted to get the next stage of the project off the ground. According to Brian Dunne, it was Carroll Walsh, watching tests of the tethered models, who observed: "You ought to fly this machine, then we'll get this program going." Building a flying model was brought up with Don Mixson, who gave enthusiastic support. "Yup, he approved it," says Astl. "He said, doggone it, we have to somehow convince people that indeed this does have feasibility, because every time I approach some higher-ups I hear stereotype answer: we use bombs to blow things to pieces, not to make them fly. So that was why there was so much emphasis put on making the damn thing fly, so they can show them it is not necessary to blow everything to pieces with bombs, they can be used for good, practical purposes. And I believe we succeeded in that."
According to Mixson, "It was ARPA that originally decreed no modeling. However, I strongly believed in it and risked a court-martial for direct disobedience of orders by telling de Hoffmann that I would authorize the model program personally provided he used General Atomic's first-year 'profit' to pay for it."[164] Official authorization slowly followed. "There was some controversy over whether money should be spent on a flying model," remembers Don Prickett. "But we stuck our neck out and gave General Atomic the green light to spend the money, some of which we didn't even have programmed yet. But we had a lot of friends up through ABDC and into the Pentagon, that we felt we could bank on to give us some protection."
On June 19, 1959, Amendment No. 1 was added to the original contract's statement of work: "The contractor shall use its best efforts to construct, static test, and free-flight test a three-foot-diameter model of the device which shall be capable of free-flight using a small number of high explosives for propulsion."[165] For the Point Loma experimentalists, led by Brian Dunne, it was now full speed ahead. "We did the whole thing in five months—from the glimmer of an idea to flying model," says Dunne. "To maintain the flow in the group, we had to increase the speed; it keeps people more focused. I was trying to get it finished before the theorists woke up and decided to get involved in the design." In addition to Dunne, Astl, and Taylor, the group responsible for the meter model included Ed Day, Michael Feeney, Budy Cesena, Perry Ritter, Michael Ames, Richard Morton, Reed Watson, Richard Goddard, Menley Young, Jim Morris, R. N. House, Leon Dial, W. B. McKinney, Charles Loomis, and Fred Ross.
The final flying model weighed 270 pounds with its ballistic envelope and 230 pounds (dubbed the "hot-rod") without. There was a narrow window between charges powerful enough to lift the model and charges so powerful they blew it apart. "Three-pound charges were found to be too damaging, and 2.7-lb charges also proved to be too potent. As a result, 2.3-lb charges were used in all the tests."[166] There were many spectacular failures in the course of finding this out. There was also some tension between the need for safety precautions and the pressure to get things done. Dunne appointed Perry Ritter as the lead man at the Point Loma site. "Jerry's a daredevil, not the kind of guy you want setting detonators," he explains. "I believe my safety procedures were good because I am still kicking around," counters Astl, citing his accident-free record. "Here you have to have procedures because you are not at war with anybody. During the war I was probably violating them left and right."
The main challenge in building a flying model—which did not have to withstand ablation by nuclear explosions or provide shock absorption for a human crew—was a problem that plagued the full-size design: how to eject the bombs. "All you would see is a picture of this pristine pusher plate," says Dunne of the early conceptual sketches of Orion, "but nobody would draw a picture of how you would get all these A-bombs into place." The Point Loma crew tried equipping the 1:130 model with a two-charge stack, but, recalls Astl, "it didn't work, the first charge went off but the second one jammed." Defusing the live charge convinced Dunne they had to design an ejection mechanism that, if not fail-safe, would at least revert the charges to safe mode if their ejection failed.
For the meter model, a series of high-explosive charges—grapefruit-sized balls of C-4, shaped by hand and cushioned by polystyrene foam inside coffee-can-sized canisters—were ejected through the middle of the pusher plate at quarter-second intervals from a central stack. "It was a very bad temperature environment and a very bad shock environment," says Dunne. Each canister had to withstand the shock of the previous explosion and then escape safely through the remnants of that explosion to detonate at a predetermined distance below the ship. "There were two possible routes to propelling the charge," says Dunne. "The first, to go with black powder, but the gas in the delivery tube would be heating up from previous explosions. The other approach was to use a gas—we used nitrogen—which cools as it expands, so the shock wave meets a cool gas that attenuates the shock pressure going up that tube." Each individual canister had its own miniature pusher plate, attached by shock-absorbing foam, to cushion the blast. This was blown into shrapnel by the exploding charge, making it hazardous to be out in the open anywhere near the flying model, although, if all went well, "penetrating fragments were localized to an angle of about 10° from vertical in a downward direction as anticipated from canister design."[167]
One-meter-diameter
flying model: firing-system schematic.
Note ballistic envelope labeled "hull."
The five separate charges, each powerful enough to bring down an airliner, had to be individually triggered to detonate in sequence, a problem similar to setting up a string of mousetraps to catch five mice in a row, without ever catching two mice at once. "It was decided, because of the potentially hazardous nature of the system, which would necessitate a large number of operations involving the handling of high explosive, Primacord and detonators, to give first priority to the requirement that the explosive system be as safe as possible," wrote Dunne.[168] The timing, switching, and safe/arm mechanism depended upon a combination of electronics, mechanics, and pneumatics to ensure that the charges exploded when ejected during flight, but not if the ejection mechanism jammed or the model crashed. Each pulse unit contained a carefully measured length of detonator cord that uncoiled as the charge was expelled. When the canister reached the end of this umbilical cord, a switch was triggered within the ship so that the charge's individual detonator, powered by its own capacitor, ignited the PETN high-explosive core within the cord. "You tie the Primacord in a double overhand knot," explains Dunne. "The detonation wave goes through at 6,000 m/sec, and when it hits that knot it goes off high-order for sure." If all went well, the detonation of the final charge triggered a shotgun shell that ejected a 14-foot-diameter parachute stored in the nose of the ship.
Richard
Goddard, Walt England, and W. B. McKinney preparing the meter-model for
flight.
Despite all these triggers, detonators, and lengths of high-explosive Primacord the whole contraption had to be safe enough for the technicians to work on before it was armed for flight, and safe enough to approach afterward, when, as often happened, it crash-landed with live charges still in the stack. "In case the model in flight fails when all charges have not been detonated," Dunne reported, "it is necessary to know that no charge remains on any of the storage capacitors. A bleeder resistor across the firing capacitor allows the charge to leak off slowly, so that after 15 minutes it is nearly certain that there is not enough electrical energy left to fire a detonator."[169] In the fourteen pages of Safety Procedures at Point Loma Experimental Site two sentences are underlined for emphasis: "Under no circumstances, shall personnel approach the area before the 'all clear' signal" and "Never cut Primacord with diagonals or steel scissors."[170]
The one-meter models had fiberglass ballistic envelopes—identified as "hull" in the engineering drawings—giving them an uncanny resemblance to the lunar spaceship of Jules Verne. Unfortunately they had trouble getting off the ground. "We'd have a big roar of high explosive, but the damn thing would not move, it would just sit there bathed in this low-density gas," says Dunne. Freeman Dyson came down to Point Loma to observe the launches on weekends. "I think we should suspend these tests until we can get at least 1G acceleration," he remarked on August 11, 1959. "I took Freeman's advice and changed all my assumptions," says Dunne. "The way to get this thing moving is we'll take a tub and put some gunpowder—it took quite a bit, about a pound—in a silk toroid. It was a loud thump and the thing just lifted up beautifully. I could sense the entrainment of cold air under the pusher plate."
Dunne's notes from that summer of 1959 chronicle the difficulties that led, eventually, to a series of successful flights. June 3: "Excessive damage to shock absorber through ejector tube perforations. Height-of-rise approximately 5 ft." August 5: "Ballistic envelope rupture apparently caused by excessive venting of gases from explosive into envelope through perforations in upper end of ejector tube. Faulty envelope layer bonding also added to failure. Charge canister No. 2 failed and jammed in ejector tube, failing to detonate, safe/arm mechanism operated successfully. Ejector tube and safe/arm system damaged by fall of model." August 27: "Charge 3 ejected but failed to detonate because disarm mechanism actuated prematurely." August 31: "All charges ejected and fired successfully. Shock absorber and liner again heavily damaged." September 10: "Camera record lost and camera damaged by high acceleration forces. Ring plate badly damaged. Nylon shock absorber liner moderately damaged. All photo records except Fastax partially lost because shrapnel cut control line at 1.5 sec." September 20: "Acceleration lost because of No. 2 misfire. Model lost altitude and parachute failed to open soon enough to break fall. Model heavily damaged. Damage to safe/arm mechanism by fall caused it to remain in armed position." October 17: "Canisters 1 through 5 ejected successfully. All detonated except 5. Since parachute squib was in parallel with No. 5 detonator, parachute was not deployed. Model heavily damaged by fall. Safe/arm mechanism damaged and jammed in armed position. Demolition charge used to part Primacord remotely to avoid hazardous disarming operation." November 14: "Flight successful in all respects. Height of rise about 185 feet. Parachute deployed at peak of trajectory. Model landed undamaged."[171]
On November 16 Brian sent Freeman, who had already returned to Princeton, a more detailed report: "Wish you could have been with us to enjoy the Point Loma festivities last Saturday. The Hot Rod flew and flew and FLEW! We don't know how high yet. Ted, who was up on the side of the mountain, guessed about 100 diameters, by eyeball triangulation. Six charges went off with unprecedented roar and precision. We think we have it all recorded. By now you realize that V0 is fairly healthy (about 20 diameters per second by means of a one-pound toroid of an ancient Chinese compound). The chute popped exactly on the summit and it floated down unscathed right in front of the blockhouse. As you can imagine, this event let loose quite a bit of enthusiasm, and we are planning a champagne party for Wednesday down at the Battery. We are planning to put the model on exhibit in your room."[172]
Jerry Astl caught the historic flight on film—at 4,000 frames per second, using a Fastax camera, an extreme version of the spring-powered Robot gun camera he had salvaged during the war. When General Atomic started doing high-explosive tests, Astl had borrowed a Fastax camera from Convair. "The Fastax I borrowed," he explains, "was still the old Fastax, you put in 400 feet of film, you pressed the button, and a few seconds later 400 feet of film is gone. It didn't have stop and go." General Atomic later purchased their own Fastax that could expose as many as 8,000 frames per second, and Astl almost never missed a shot. "When I was making movies of these flights, it was usually action," he says. "Everybody was in the bunker, the only idiot who was out most likely was me, because I had to run the cameras." Astl also edited the footage, added titles and graphics, and produced a brief six-minute film that helped to keep the project alive at the end of 1959. "After we finished the testing of the flying models, we finished the film and the fellows took it for presentation to ARPA," he recalls. "Before anybody saw the film there was all kinds of debate and Ted said after the film was finished you had never heard such beautiful quiet. They were caught speechless. Bombs explode everything, and here the bombs propelled something and it came down in one piece."
"This seems a little excessive," says retired General Ed Giller, reviewing Astl's film forty years later, in 1999. "It's amazing that it survives that much high explosive, and it keeps going. I wouldn't believe that it would do that." That was the meter model's last flight. The neighbors at Point Loma were starting to complain about the noise and the Orion gang had to move farther out of town. Carroll Walsh found a new site, six miles inland from General Atomic on a decommissioned Marine Corps staging area known as Camp Elliott, in Sycamore Canyon. Formerly a dairy farm belonging to Mr. Green, the new test site, on 1,012 acres, had two small houses, a barn, a wild jackass, and large numbers of rattlesnakes, which "were systematically picked up by a couple of people working on the test program, with forked sticks, and then put into a chickenwire pen, and I remember seeing maybe twenty in there," says Ted.
"Hot
Rod" 1-meter model, with ballistic envelope
removed, clears the top of the 75-foot Atlas test tower, November 1959.
Green Farm had no neighbors, so the Orion crew could make noise without complaint. Brian Dunne and later Howard Kratz set up a facility for firing explosive-driven plasma jets at sample pusher-plate targets, but there were no more explosive-driven flights. "I had Mike Ames hide the meter model in the cow barn out at Green Farm, up on the second floor, boarded up," says Dunne. "Experimentalists are prone to cannibalism, and I knew that someday someone would need a piece of aluminum plate or something and they'd just go in and start cutting it up. It was stashed from 1959 to 1979." When Project Orion was partially declassified, Dunne and General Atomic s public relations officer Earl Zimmerman sent the Hot Rod to the Smithsonian. The museum tried to interest Gulf Oil, General Atomic's new owner, in underwriting an exhibit, but, says Dunne, "Gulf didn't want to be in any way associated with anything having to do with bombs."
The success of a 300-pound model driven by high explosives offered no conclusive evidence of the feasibility of a 4,000-ton ship driven by nuclear bombs. But the flights helped to convince Air Force (and later NASA) officials that the project deserved attention, and they inspired the Orion team in 1959. "The purpose of the flying models was to demonstrate that a vehicle possessing in rudimentary form the same engineering components as a full-scale ship, including pusher-plate and shock absorbers and ejector system, could be made to function correctly," Freeman later wrote. "The model flights were the most beautiful part of the whole project. We had a launch site on a hillside covered with flowering shrubs and cactus, overlooking the Pacific Ocean. We usually went out early on Saturday mornings to set up the model and were ready for the countdown about lunchtime. I often wondered what the Saturday afternoon sailors on the ocean thought of us, when some weird-looking object rose briefly from the test stand and blew itself into a thousand pieces. I still keep in my desk drawer a bag of aluminum splinters which I collected after one of our test flights, to prove to myself that all these happy memories are not just dreams."[173]