Third time is not charm for the N1
After a two-year hiatus, Soviet engineers brought the third N1 Moon rocket to the pad for the launch on June 27, 1971. Numerous upgrades implemented throughout the vehicle gave project leaders some confidence that the latest attempt would go better than the previous two in 1969. But despite normal operation of the troubled first-stage engines at launch, some mysterious force led to the destruction of the rocket less than a minute in flight.
Third launch of the N1 rocket (No. 6L) at a glance (537):
When the smoke cleared over the devastated launch facility in July 1969, everyone involved in the second failure of the N1 rocket could guess that it would be a long time before the next launch attempt could be made. Though only six months had passed between the first two failed launches, the near-annihilation of the only available pad in the second attempt meant lengthy repairs.
No efforts would be spent to activate the second pad, informally known as Levy start (Left pad) which was largely completed at the end of 1968 and which mostly escaped damage after Vehicle 5L crashed at the near by Pravy (right) pad. There, things looked grim – just to dismantle the fallen and twisted structures, the military construction directorate in Tyuratam had to bring a team of explosive experts from Tashkent in the nearby republic of Uzbekistan. The repairs at the "right" pad started in August 1969 but would not be completed until 1972. (202) The rocket itself needed multiple modifications in the wake of the failure, even though there were initial hopes to complete them in months not years.
Much changed in the intervening time: NASA astronauts walked on the Moon just a few weeks after the 5L explosion, but the Apollo program was put on track to be curtailed in 1972, after launching seven lunar expeditions. The premature termination of the Apollo could actually give the USSR the political incentive to mount a more spectacular and sustained exploration of the Moon, even if it meant that the first Soviet landing would have to be pushed to 1975 or 1976. By 1971, an expert commission, which reviewed the situation, favored this approach.
Despite growing skepticism toward the ill-fated N1 rocket, the development of the L3 expeditionary complex which relied on it continued in the hope of getting it ready for landing the first cosmonaut on the Moon after as few as three successful flights of the N1. Using a dual launch scenario, the L3 system could be upgraded to deliver two or even three Soviet cosmonauts to the lunar surface with expeditions lasting up to a month. It could be followed by the construction of a permanent lunar base, though the time frame for all these activities was uncertain. Still, beyond the Moon, future space station plans and projects like a Mars sample return mission also relied on the successful introduction of the N1. (774)
Modifications of the N1 rocket for the third launch
The creators of the N1 rocket used the forced break between the second and third launches to learn lessons from the failures and prepare the best they could for the next attempt. First of all, the areas most vulnerable to fire inside the rocket's first-stage booster got additional blanketing. Also, protective filters were installed on the oxidizer and fuel lines leading to the engines to reduce the probability of foreign objects entering into the system and causing catastrophic explosions. (223) The latter upgrade was recommended practically immediately after the 5L disaster.
Additionally, fire suppression systems using freon and nitrogen were installed throughout the vehicle. The inter-stage and tail compartments of three stages were equipped with a thermal-control system.
Various instruments for telemetry and flight control were relocated farther away from the engines and were protected with additional thermal blankets.
On the propulsion side, engineers worked to eliminate the pressure-drop effect at the entrances into the pumps, which also had a catastrophic potential. They also introduced staged shutdown of peripheral engines on the first-stage booster for a gentler cutoff of the rocket's colossal thrust at the end of its burn in the stratosphere. For the first time, upgraded engines 11D51 and 11D52 for the second and third stages of the rocket were put through fully integrated firing tests, however the static firing of a fully assembled first stage remained out of reach, due to the lack of a test facility large enough to accommodate it.
In the meantime, flight control specialists changed the operational algorithm and the interaction between the KORD diagnostics system and the flight control system of the rocket. First, the emergency engine cutoff command, AVD, was blocked for the first 50 seconds of flight to give enough time even to a stricken rocket to clear the precious launch facility and avoid the kind of devastation seen in July 1969. The rocket was also specifically programmed to avoid the launch pad shortly after liftoff.
The rocket's electric system was upgraded to reduce the probability of a stray engine cutoff command resulting from damage or short circuits in the cable network.
Finally, the rocket was equipped with many new and backup sensors, which required to add a new ground-based diagnostics system, 11T81, which would be recording some additional parameters of the rocket performance before its liftoff.
Flight program and payloads
The initial planning for the 6L mission of the N1 rocket started as early as January 1967, with mass calculations and other preliminary work. (774)
Early on, the 6L mission was apparently considered as a milestone in the transition from purely experimental N1 vehicles to a more nominal system. It appeared from notes made by the head of TsKBEM design bureau Vasily Mishin between January and March 1968, that Vehicle No. 6L was expected to be the first to carry the complete L3 lunar expeditionary complex with a functional though unpiloted LOK spacecraft, known as T1K, along with a dummy mockup of the LK lander. (As of March 1968, the first functional LK lander (T2K) was to be introduced with Vehicle No. 7L.)
During 1968, Group 8 at the TsKBEM design bureau was working on technical documentation for experimental units representing the 11F93, 11F94, T1K and T2K ships comprising the L3 complex. In parallel, Group 2 was working on the documentation for the actual flight vehicles. The T1K spacecraft launched on the 6L rocket was expected to introduce a total of seven new systems whose design documentation was scheduled to be ready in June 1968. For the ground testing of the L3 complex, a special full-scale simulator had to be completed ahead of the 6L launch.
At the time, however, the program had come under schedule pressure. By March 1969, Mishin had already discussed the option of simplifying the LOK vehicle on the 6L rocket, or using the L3S variant, which would derive from the much more simple, but readily available L1 spacecraft.
In early 1969, the development of the LOK vehicle for the 6L rocket fell significantly behind schedule. On April 2, Mishin's deputy Konstantin Bushuev reported that the experimental design mockup of the LOK spacecraft had not been completed and even certification for its early variant intended for the 6L mission had not been signed. A mockup for thermal tests was also behind schedule as was the fueling test mockup. Some work was underway on vibration tests of the Descent Module.
Another mockup of the LOK spacecraft was needed for static tests and some test articles of the vehicle would be used for separation tests. The full-scale simulator, KS, with all onboard systems, was to be built only for the piloted version of the LOK vehicle, to be introduced with the 7L rocket.
At the same meeting, Konstantin Vachnadze, the director TsKBEM's production plant, confirmed that the LOK's thermal mockup was behind schedule and that his factory was also lacking the Block I propulsion system for the LOK, the Rosa, DOK, DKP, Granula and the Efir systems, as well as windows. By far the most critical component of the LOK spacecraft was the Volna fuel cell power generator, EKhG, which was supposed to replace solar panels aboard the spacecraft.
Vachnadze urged Mishin to appeal to the government to accelerate supplies and make up the schedule.
Despite all the delays, in May 1969, Mishin still mulled an installation of a photo camera for taking pictures of the Moon aboard the LOK vehicle launched by the 6L rocket, indicating that the lunar flight scenario for the mission remained in place. (774)
It appears that soon after the N1 failures in 1969, the LOK was spacecraft slated to fly aboard the 6L vehicle was downgraded to a simplified prototype with only few operational systems. On July 24, 1969, Mishin discussed the LOK configuration for the 6L flight with his engineers.
However the rocket would still carry a fully functional Block G fourth stage and the Block D fifth stage.
The launch window ultimately selected for the 6L mission did not target the lunar orbit, but the Block G stage could still fire, putting the L3 mockup stack on a long elliptical orbit around the Earth, and Block D could also simulate braking maneuvers.
The night-time liftoff of the giant rocket was timed primarily to facilitate the evacuation of the personnel from the huge danger zone surrounding the N1 launch facility at Site 110.
Activities in 1969
By January 9, 1969, all the engines for the 5L and 6L rockets were reported to have been delivered and tested ahead of the first N1 launch at the design bureau led by Nikolai Kuznetsov. In the wake of the 5L crash on July 13, 1969, (blamed squarely on the propulsion system), Mishin considered removing all the engines from Vehicle 6L and putting them to their limits during firings at ground test stands.
On July 20, 1969, Kuznetsov proposed the following schedule for delivering 30 new engines, forming a six-engine cluster at the center of the first stage and 24 engines arranged in a circle on the periphery of the booster:
In parallel with the supply of 30 operational engines for the rocket, 12 others were to be sent for performance tests, known as KONRID. Kuznetsov also had to supply eight operational and four KONRID engines for the second stage and four operational and four KONRID engines for the third stage.
On July 28, 1969, the Chief Designers' council met to discuss the N1-L3 project. Kuznetsov reported on the reasons for the disintegration of Engine No. 8 on Vehicle 5L, on measures to improve the reliability of the engines and on the delivery of engines for the next rocket. At the same meeting, Vladimir Barmin, in charge of the launch infrastructure, made proposals for repairing the "Right" pad and completing the "Left" pad.
Events in 1970
By March 22, 1970, the first set of six 11D51 and 11D52 engines for the 6L vehicle was scheduled for delivery between April and June 15 of the same year. At the time, 6L and 7L were to be the last two rockets equipped with expendable engines, before the introduction of reusable engines on Vehicle 8L. It meant, that the same engine could be first used for a test firing on the ground and then installed on a flight-worthy rocket, greatly simplifying the whole process. But for the next two flights, engineers had to weed out all the problems on the untested hardware and address numerous other issues and assumptions.
As of August 6, 1970, Mishin wrote that the 6L launch depended on the resolution of the discrete oscillations of the aft bulkhead at launch. The delivery of avionics and of the Volna fuel cell unit was still listed as the determining factor for the 6L launch date, indicating the remaining hope for launching the LOK spacecraft during the 6L flight. The discrete oscillation issue was discussed again at the Chief Designers' Council on December 22, 1970.
Still, by November 23, 1970, the assembled first stage booster (Block A) for the 6L rocket was rolled into Hall 3 of the vehicle assembly building at Site 112 in Tyuratam. The second and third stages were already at Hall 4 of the same building. However, on November 27, Pilyugin called and said that the B-61 avionics units had to be replaced on all three stages of the rocket.
As of December 8, 1970, Mishin wrote down a calendar of critical launches planned for the first half of 1971:
Final preparations in 1971
On January 1, 1971, Mishin discussed the 6L mission with Afanasiev again. A major technical meeting on the 6L flight took place on January 17, 1971. Once more, the discrete oscillation issue was brought up but it appeared to be approaching resolution.
As of March 24, 1971, the transfer of the vehicle to the fueling station was planned between 20th and 23rd of April 1971, followed by the rollout of the rocket to the "Left" pad at Site 110 on May 20, 1971. (774)
After another delay, Vehicle 6L was finally rolled out from the assembly building at Site 112 to the "Left" launch pad on June 13, 1971. At the time, the launch was apparently planned for 10 days later, on June 23, 1971. (705)
The preparations on the pad were going relatively smoothly despite a severe heat wave reaching some +40 and 42 degrees Celsius even in the shadow and further complicated by dust-carrying wind. Then, the desert-like conditions gave way to a tropical downpour right around the time of the planned fueling operations, bringing havoc to Tyuratam with flash floods, short circuits and washed away roads, railways and communications lines. After the rain, the launch team found lowered resistance in the electrical lines inside the rocket, prompting the activation of the thermal conditioning system to "dry out" the vehicle, while postponing its further processing.
On June 18, Mishin called Head of cosmonaut training Nikolai Kamanin and said that the 6L launch would be postponed for two days from June 20. Around June 24, the launch was postponed again, this time to June 27. According to General Karas, who oversaw testing, the telemetry system aboard N1 failed and there were other problems that threatened the June 27 launch window. (142)
To add to the usual pressures in Tyuratam, on June 25, the prolific Soviet propulsion engineer Aleksei Isaev died in Moscow, leaving many of his fellow rocketeers, now at the top of the Soviet lunar program, scrambling for funeral arrangements right in the middle of the final countdown to the N1 launch.
The N1 No. L6 vehicle lifted off on June 27, 1971, at 02:15:07.5 Moscow Time, turning night into day for miles around Site 110. The engines of the first stage ignited and lifted the colossal rocket off the pad seemingly without a problem. Mishin's deputy Boris Chertok and his colleagues, who were at TsKBEM in Podlipki and were listening to intermittent reports from the firing bunker, heard how five seconds after liftoff the telemetry officer reported: "Pitch, Yaw normal, roll is increasing!"
In Tyuratam, telemetry specialists V.D. Tikhomirov and G.A. Emanuilova, who were watching the launch from the IP-2 ground station a few miles southeast of the pad, saw the massive rocket rumbling over them with a noticeable shaking and flexing from side to side. It was clear that something had gone terribly wrong with the flight control of the rocket, even though its engines continued firing. (223)
That account was also echoed by General Valery Menshikov who watched the launch from Site 1, also southeast from the N1 pad. He remembered that the wayward rocket climbed to an altitude of around 500 meters and flew overhead while flexing "like a willow tree in the wind." (704)
Moments later came the inevitable disintegration of the rocket and the engine shutdown. The powerless giant began falling back to the ground in the pre-dawn sky, followed by multiple explosions in the steppe, including one not far from Site 31 on the Eastern flank of the test range.
The debris from the rocket and the L3 payload were strewn across the grassland from three to 15 kilometers downrange from the launch site, but, apparently, nobody was hurt, and the rocket's launch facility remained intact. (223)
In the immediate aftermath
The post-failure analysis showed that all 30 engines of the first stage had developed the required thrust and lifted rocket off the pad as planned. However from the beginning of the ascent, starting just 0.5 seconds after liftoff, the vehicle exhibited abnormal motion around its roll axis. To an observer looking up from below, the rocket would appear to be rotating clockwise around its main axis. The vehicle's roll control thrusters were activated and rotated to a maximum position of 45 degrees in an effort to counteract the fatal torque, but without success. At T+14.5 seconds in flight, the rocket was already 145 degrees off its nominal position, even though the main engines continued accelerating the vehicle through the air. According to Chertok, at that point, the rocket was off by just eight degrees on its roll axis, but it rotated to 60 degrees at T+50 seconds.
According to some sources, the stabilization of the rocket along the pitch axis was also off the required angle, but possibly, it was the result of the flight control system "stalling" after the rocket exceeded the allowable roll criteria. The stabilizing gyroscopic platform aboard the rocket locked itself at T+39 seconds, after reaching its limits.
Essentially, the rocket was uncontrollable from the start and was heading in a random direction, but its engines continued firing due to an embargo on the emergency shutdown command for the first 50 seconds of the flight, in order to avoid damage to the pad.
At T+47.8 seconds in flight (774), as the rocket was leaning heavily along its wrong trajectory and far exceeded the allowable angle of attack, the excessive aerodynamic loads began tearing the vehicle apart. The payload section, along with the third stage, apparently broke off from the rest of the vehicle and was the first to impact the ground around seven kilometers downrange. (690)
The emergency engine cutoff was finally activated at T+50.1 seconds, shutting down all 30 engines of the first stage and leaving the first two stages in a free fall. They crashed and exploded around 20 kilometers from the launch pad, forming a 15-meter-deep crater. According to Chertok, without the 50-second engine cutoff embargo, the rocket would have fallen around one kilometer from the pad, producing an explosion equivalent of 500 tons of TNT, enough to destroy the launch complex once again. (685)
Investigation into the failure
It was not immediately clear what kind of force had caused the rocket to rotate, but the immediate late-night hypothesis at TsKBEM that a stray flight control command put the rocket in the fatal spin had been disproven by telemetry data by 10 a.m. in the morning on June 28. Georgy Degtyarenko (a TsKBM specialist in the flight control system) called from Tyuratam and said that the flight control system had moved the roll control thrusters to their limit in the effort to counteract the spin but that it was not enough. (685)
There were some unsubstantiated claims that the changes in the pitch control sequence introduced with the 6L flight and designed to prevent the damage to the launch pad, might have played some role. (223, 690) This legend probably originated from the fact, that under pressure from officials to immediately explain the failure, investigators first scrutinized all the new aspects of the 6L mission that could produce previously unseen problems.
However, after the initial rush for a quick explanation, the investigation quickly shifted to very complex issues of aerodynamics and gas dynamics, which could produce an unexpected force rotating the giant rocket.
Initially, Vladimir Roshin, the leading aerodynamist at TsKBEM, expressed serious doubt that such a major factor could be overlooked in the exhaustive research into the rocket's shape but with one caveat: "Unless its shape had been disfigured during the upgrades of the aft bulkhead," Chertok remembered Roshin saying.
The subsequent investigation into the N1 failure was further compounded by the tragic loss of the Soyuz-11 crew just three days after the 6L crash. Yet, Mishin's notes indicate the sober determination of the engineers to overcome the double catastrophe and their confidence in the eventual rehabilitation of both systems.
The 6L probe turned into a months-long painstaking effort to explain the bizarre behavior of the rocket with mathematical calculations and with model simulations. As late as February 21, 1972, Mishin's work agenda listed the meeting on the aerodynamics of the 6L accident. Experimental work aimed at understanding the issue concluded only at the end of March 1972, even though the mystery of the failure would never be truly resolved.
Around October 15, 1971, the head of Soviet Academy of Sciences Mstislav Keldysh signed off on the results of the formal investigation into the 6L failure. (774)
The official conclusion was that the abnormal rotation of the vehicle had been caused by an unexpectedly powerful turbulent flow at its bottom, produced by main engines and further exacerbated by the unsymmetrical flow around the engine components protruding from the rocket's aft bulkhead. The resulting torque force far exceeded the maximum capabilities of the existing roll control system, which would have to be upgraded. (52) The absence of the same phenomenon during previous launches was explained by the fact that the premature shutdown of individual engines (experienced by previous rockets) would weaken the effect.
The theoretical explanations were reportedly confirmed by computer modeling, which used actual telemetry data from the failed launch to determine gas dynamics around the vehicle. Still, the analysis of the mysterious force that doomed the 6L vehicle continued for many years.
In the 1970s, the TsNIIMash research institute tested a 1-to-50 scale model of the N1 rocket in the U-11 wind tunnel. The facility used cold air compressed to 200 atmospheres and generating around 50 atmospheres of pressure around the model. A multi-stage compression system was then used to simulate temperature of around 4,000 K and 2,500 atmospheres matching gas dynamics of the rocket's ascent. The experiments, which lasted just tens of a second, reportedly produced a rotation effect matching the force of around 42 tons per meter that had been recorded by the telemetry during the ascent of Vehicle 6L. (247)
Soviet dynamics expert Boris Rabinovich wrote that even a 0.1 degree deviation of the exhaust flowing from the engines and from the turbopump drainage pipes would produce a turbine-like torque effect in the circular space between the two rows of engines of the N1, with force of around 50 tons per meter...
Vasily Mishin (left) and Nikolai Kuznetsov next to the NK-33 engine, which was designed for the upgraded version of the N1 rocket. Mishin is known to visit Kuibushev's NK production plant on August 18, and November 16, 1969. Credit: Progress
N1 rocket arrives at the launch pad at Site 110 in Tyuratam. Credit: Roskosmos
N1 rocket on the launch pad. Credit: Roskosmos
Photos often attributed to the launch of the 6L vehicle on June 27, 1971. Credit: Roskosmos
A camera installed at one of the lightning towers or an access gantry recorded the liftoff of the N1 rocket, possibly on June 27, 1971. Credit: Roskosmos
The impact crater left by the crash of N1 Vehicle No. 6L on June 27, 1971, as seen by a US reconnaissance satellite on Nov. 16, 1972. Credit: USGS
A wind tunnel model of the N1 rocket.