by Neil Sheehan
53.
A FEW GRAINS OF SAND
Thor No. 101 arrived at Patrick Air Force Base through the wide front clamshell doors of a C-124 four-engine Globemaster transport from California on October 18, 1956, a record nine and a half months since the signing of the contract with Douglas. The Thor was numbered 101 in an attempt to fool Soviet intelligence into believing that the U.S. Air Force had a lot of these IRBMs near flight status, instead of just one. Mathison and the civilian contractors were approaching completion of one of the two launch emplacements for Thor and Jupiter and would have it ready in December. The second would not be finished until mid-1957. (The Atlas ICBM still held the highest priority as far as Schriever was concerned and so Moose had concentrated on completing the two launching stands for it by the end of the year, even though test firings would not begin for another six months.) One IRBM facility was enough for the moment and Mettler and Thiel had plenty of minor fixes to the missile and its various subsystems to keep them occupied until Mathison and the contractors were done. They had not had time to train a special launch crew for the blockhouse and so they formed a scratch one composed of themselves and other Ramo-Wooldridge engineers, a couple of Hall’s officers, and Douglas personnel.
By January, a month behind schedule, all last-minute adjustments were done and they had the missile mounted on the circular launch pad. It was time to conduct several “captive” or “hold down” firings, also called “flight readiness firings.” A rocket must be held upright on its launch pad by immensely strong steel latch mechanisms until the engine or engines swell to sufficient thrust to overcome the weight of the missile itself. At that point, the latches are thrown open and the missile lifts off into the air. If there were no latches to hold the missile in place, it would tip over before enough thrust had developed and blow up on the pad. The latch mechanisms permit the engineers, as part of the final flight readiness process, to further test the missile by igniting the engines, usually for short burns before they are shut down, without ever turning the rocket loose to fly. Mettler and Thiel had wanted to do this before coming to the Cape, but given the gallop to launch Schriever had imposed, there was no chance to construct concrete stands in California on which to lash down Thor for these captive firings. And so they were instead done using the latch mechanisms on the pad at Canaveral. All went well and January 25, 1957, was set as the launch date.
A countdown for a launch was (and still is) a complicated, tedious, and often exasperating procedure. Test missiles were instrumented with a myriad of sensors that served two purposes. Prior to the flight, they told the launch crew whether the missile’s systems—propulsion, hydraulic steering controls, guidance, and others—were correctly connected and functioning well. During the flight, the same sensors monitored these systems to determine whether they actually did function as they were supposed to do. If something went wrong, the sensor detected what had gone awry so that the problem could be remedied before the next launch. In addition to sensors monitoring the main systems, there were many others that measured such factors as speed and angle of flight, the temperature on the surface of the missile at various points, and whether the engines were shutting down instantly at just the right moment to hurl the warhead accurately toward its target. The information provided by the sensors was called telemetry, a word sometimes used to refer to the sensors themselves. Once the missile took flight, the telemetry was transmitted to radio receivers on the ground through small transponders. The electrical power to operate these transponders was provided by batteries fitted into the Thor. (Later a more sophisticated power supply was devised by building a generator into the missile body.)
The interior of the blockhouse was brightly illuminated by fluorescents hung from the ceiling. Rows of consoles, their faces covered with instruments that gave off readings from the sensors, were arrayed around a table in the center. The instruments on a console encompassed the telemetry for a particular system of the missile, for example propulsion or hydraulic controls, or monitored some other aspect such as temperature or speed. A crew member of this pickup launch team sat in front of each console and called off the instrument readings as Mettler, who was acting as launch director, orchestrated the countdown down from the table in the center. Thiel, as his deputy and the only man in the blockhouse with experience gained at White Sands and the Redstone Arsenal, moved from console to console, making certain they were getting correct readings and supervising anything else that required his expertise. If a malfunction is discovered during a countdown and it is a simple matter of something like a faulty electrical switch that needs only a few minutes to replace, then the countdown picks right up and marches on after it is repaired. But if the malfunction is substantial and requires longer to remedy, the countdown has to stop completely and resume again from the beginning, no matter how much time has been invested up to that point.
Despite all the prior testing and preparatory work by Mettler and Thiel and their colleagues, Thor 101 required several countdowns before they reached the critical and potentially perilous step of fueling the missile. The RP-1 was a stable element. The kerosene was pumped without much concern from the storage tank at the launching emplacement into its tank in the upper portion of the Thor. The liquid oxygen, or LOX as it was called, which burned the kerosene at the highest possible temperature to generate maximum thrust from the rocket engine, was an entirely different matter. A high-speed turbo-pump within the missile mixed the LOX and RP-1 before feeding them into the burn chamber of the rocket motor. LOX was highly volatile. To keep it from vaporizing, “boiling off,” it had to be kept under cryogenic conditions—297 degrees below zero Fahrenheit—in an insulated storage tank that resembled a giant thermos bottle. To prevent or at least slow down vaporization of LOX after it had been pumped into the oxidizer tank in the lower section of the Thor, the first step in the entire fueling process was to cool down, “cold soak,” the missile as much as possible by flushing both the RP-1 and LOX tanks with liquid nitrogen stored at the same cryogenic temperature as the LOX. Nitrogen is an inert gas that could be pumped into and out of the Thor’s tanks without danger. The nitrogen also provided a means of checking for leaks, which had to be avoided at all cost, particularly where the LOX was concerned. Great care was imperative at every stage of its handling. The slightest spark would set it to burning fiercely once it began to vaporize. If it was contaminated in any way during the fueling process, instead of burning the kerosene in the rocket motor, it would explode.
But on this day, the fueling process went off without incident. The final steps in the countdown were completed. Mettler and Thiel and everyone else in the blockhouse felt their year of toil under unremitting pressure was soon to be rewarded with a ballistic missile sailing downrange across the islands of the Caribbean. “Everything looked just perfect,” Mettler remembered. He turned loose the simple computer of the day that controlled the electrical firing sequence to ignite the engine. Roaring flames enveloped the launch pad, thrust built to the lifting point, the steel latch mechanisms holding down the missile were thrown open, and Thor 101 began to rise. It rose about eighteen inches and then suddenly fell back on the pad. With a deafening blast and a shock wave that was felt in the blockhouse, the missile blew up, not only tearing itself into pieces but also damaging the concrete launch pad seriously enough that Mathison’s civilian construction crews needed two months to restore the pad to usable condition.
At the birth of Thor in 1955, Mettler had thought it both generous and daring of Schriever and Ramo to put a thirty-one-year-old engineer with no experience in rocketry in charge of the Ramo-Wooldridge technical direction team on a project this important. (He knew the decision had not been Ramo’s alone because the two men conferred on everything that mattered.) When he later thought about that traumatic day in January 1957, he reflected that had they been different men they would have fired him right then, if only to satisfy the bureaucratic reflex to single out a scapegoat for such a catastrophe. They were waiting f
or him over at the motel at Cocoa Beach. They had flown to Patrick earlier in the day, checked into the motel, and had intended to watch the launch from there. Schriever was due in Washington the next day to brief the Joint Chiefs of Staff. Mettler drove over to the motel and told the two men he had no explanation for the failure. It was a mystery he and the rest of the Thor team would have to unravel. He could tell that Schriever and Ramo were disappointed, but neither man reproached him. “I expect things like this to happen,” Schriever said. Mettler came away grateful and admiring of the courage of Schriever and Ramo in their willingness to risk giving him a second chance. He did not get off entirely lightly. Schriever took him along to Washington to join in the briefing the next day for the Joint Chiefs on the entire missile program. He had Mettler speak to the august body on the portion dealing with the Air Force IRBM. The chiefs, Schriever knew, would hold him and not Mettler responsible for the failure, but he seems to have reasoned that it would be best for them to hear the details from a man with firsthand knowledge.
For weeks, the explanation eluded Mettler and Thiel and their colleagues. They examined and reexamined the wreckage, checked and rechecked their instruments, went back over the countdown procedure again and again, always in vain. “The pressure came up, all the instruments were okay, and the damn thing was ready to go, and yet it didn’t go,” Mettler recalled of their bafflement. “In desperation, day and night we tried to find out what the devil was wrong.” Finally, they turned to examining every photograph and every bit of film that had been taken of the launch. They were sitting in a hangar, “sort of bleary-eyed,” as Mettler put it, watching a public relations film the Douglas company had made when all of a sudden they spotted the trail. The answer turned out to be not a complicated technological treatise, but once again an explanation as simple as the hot-box metal cupboards on the SAC bombers that had been destroying the vacuum tubes in the navigation and bomb release systems.
The film showed two technicians in white jackets, with “Douglas” printed in large letters across the backs, pulling a hose that was to be used to fill the oxidizer tank of the Thor with LOX. They were dragging one end of the hose, the end that was to be connected to the valve on the oxidizer tank, through sand. “Even a grain of sand in liquid oxygen under impact will explode,” Mettler said. And here was the LOX hose picking up a good many grains. They searched through the wreckage again until they found and collected the remains of the valve on the tank to which the hose had been connected. Fitting the pieces together, they saw that the valve had been shattered “from the inside out,” Mettler said. The probability that contamination from sand had caused the LOX to explode, rather than to burn, seemed certain, but to make sure they performed an experiment in the isolation of the Rocketdyne engine test grounds in Southern California. They set up a LOX tank with a fill valve identical to the one on Thor, contaminated the valve with sand while pumping in LOX, then closed it and ignited the LOX with an electrical charge. The valve immediately blew up. In the future, all hoses, valves, and other connections were kept in a state of pristine cleanliness that became known in the liquid-fueled rocket business as “LOX-clean.”
54.
MEDARIS GOES FOR THE ICBM
On April 19, 1957, approximately two months and three weeks later, the launch pad had been repaired, several captive flight-readiness firings for the new missile had been completed, and the countdown was under way in the blockhouse for Thor 102. The atmosphere was particularly tense, not only because of the failure of Thor 101, but also because this was the second and possibly the third attempt (Mettler and Thiel could not remember precisely when they were interviewed many years later) to launch Thor 102. A twenty-four-hour limit had been established for launch attempts. It was feared that if countdowns had to be aborted and resumed over and over until the launch began to run more than twenty-four hours, the blockhouse crew would become so fatigued that someone would read an instrument incorrectly with disastrous results. Therefore, once twenty-four hours had been expended, the launch had to be canceled. This was precisely what had happened on the first, and the second, attempt to launch Thor 102. The countdowns had been repeatedly stymied by the discovery of mechanical problems. The countdown had to be stopped and restarted several times again on April 19, but matters went better on this day and Mettler at last pressed the button to start the electrical ignition sequence.
Thor 102 responded with fire and the percussive waves of its thunderous voice. When the hold-down latches on the launch pad were thrown open, the rocket rose, slowly at first, and then with quickening acceleration, beginning to fly up and out toward the sea. There was immense relief and exhilaration in the blockhouse until, just thirty-two seconds into the flight, the missile exploded. The range safety officer, housed separately in a central control bunker, announced that he had blown it up. Every missile launched at Cape Canaveral carried a packet of explosives hooked up to a radio-activated detonator. If a missile went awry and turned inland, the range safety officer pushed his destruct button to protect the lives and property of Florida’s civilians. Thiel was astonished. The instruments in the blockhouse had indicated that Thor 102 was doing fine, heading right out for the Caribbean. Some of his German friends and former colleagues from the Redstone Arsenal, who were at Canaveral for Jupiter launchings, had watched the launch from a building not far away. They told Thiel the missile had been flying perfectly and should not have been demolished. He and Mettler hurried to the control blockhouse. Unhinged by rage, Dolf Thiel dashed inside and confronted the range safety officer, an Air Force major. “You son of a bitch, why did you destroy the missile?” he shouted at the man. “You’re nuts,” he yelled, grabbing the major and cocking back his fist to punch the officer in the face. Mettler caught hold of Thiel’s arm just in time. “Take it easy, take it easy,” he said, calming his deputy.
It turned out that the mishap was the result of incompetence by a Ramo-Wooldridge technician and panic by the range safety officer. The range safety officer had three instruments with which to monitor the flight of a missile. One was an optical device with powerful magnification and a reticle with two parallel lines that enabled him to determine whether the missile was flying straight up, as it was supposed to do, or veering off to one side. The second was a standard radar tracking scope, referred to as a “skin radar” because the impulses of its transmitter were reflected back from the outer surface of whatever object it was tracking. The third was a different radar that had just been added to these first two monitoring instruments in the control center blockhouse at the instigation of someone on Mettler’s staff. It was a Doppler Velocity and Position radar, otherwise known by its acronym, DOVAP. The Doppler radar tracked the Thor through a beacon installed in the missile and provided a more precise reading of the speed and position of the missile than the skin radar.
There was nothing wrong with this particular Doppler radar apparatus, but when the Ramo-Wooldridge technician installed it, he hooked it up backward. Instead of showing Thor 102 heading out to sea, the DOVAP showed it flying inland toward the city of Orlando. To make matters worse, someone from Ramo-Wooldridge had also advised the range safety officer to rely on the DOVAP as his principal monitoring instrument. The major had been understandably alarmed to see this IRBM streaking for inland Florida on the DOVAP scope. He should, however, have kept his head and followed commonsense procedure by looking at his two other instruments to be certain he was getting a correct reading. No electronic device is entirely dependable. A range safety officer is supposed to act with dispatch, but not to destroy lightly an expensive missile. Seconds counted, but he had seconds to spare. Instead, he had panicked and mashed the destruct button. He was transferred not long afterward to Alaska.
At least they knew now they had a missile that could fly, and on May 20, 1957, the countdown began in the blockhouse for Thor 103. It turned into a day of ill omens. Time after time they ran into problems that required them to stop the countdown and took long to remedy and they had to star
t the countdown over again. On through the night and into the next day the relentless process continued. By the morning of May 21 they had reached the twenty-four-hour limit, when the rules said they had to cancel the launch to avoid the chance of a grievous error from crew fatigue. But Thiel’s blood was up and he did not want to quit on this third attempt to demonstrate that Thor could take to the air with the best of missiles. He was convinced that with just a bit more time they would achieve liftoff. The blockhouse had a direct Teletype line to the headquarters in California. Thiel persuaded Mettler to send an urgent message to Lieutenant Colonel Richard Jacobson, Schriever’s overall chief of test facilities and operations, and ask for an exemption to continue. Jacobson’s response was swift: “Negative. Approval is not granted. Terminate the test in accordance with standard operating procedures.” Schriever and Ramo were late flying to Patrick and had just arrived at the motel at Cocoa Beach. Mettler and Thiel decided to go over Jacobson’s head. Mettler telephoned Schriever at the motel and said the missile was nearly ready to go and asked permission to exceed the twenty-four-hour limit and launch it. Bennie’s instinct was that Mettler and Thiel and the entire launch crew must be so exhausted by now from the interminable countdowns and from the stress they were all feeling that the wise thing to do was to stop and get a night’s sleep. “Look, let’s call this off and we’ll start again tomorrow,” he said to Mettler. But Mettler was insistent. They really wanted to go now, he said. “Well, I won’t second-guess your judgment,” Schriever said, relenting. “If you want to go, go ahead.” He and Ramo walked out of their rooms to watch the launch. All of a sudden, the horizon over at the Cape was illuminated by a huge flash. “It lit up the sky,” Schriever remembered.