DOS-3: Skylab's challenger

On May 11, 1973, the USSR launched its fourth space station and the first major upgrade of the original Salyut design, just days ahead of the American Skylab. The 19-ton Soviet lab reached orbit as planned, but an immediate crisis put Soviet engineering and mission control teams to a severe test... Today, it is often only an asterisk in history books, but in 1973, the launch of DOS-3 produced a political earthquake across the Soviet space program and even got the KGB involved.

Previous chapter: DOS-7K No. 2


DOS-3 mission at a glance:

Spacecraft designation DOS-7K No. 3, 17K No. 12301, Kosmos-557
Spacecraft liftoff mass 19.4 tons
Launch date and time 1973 May 11, 03:20:00 Moscow Decree Time
Reentry date 1973 May 22
Launch vehicle UR-500K (Proton)
Launch site Tyuratam, Site 81

Development of the modified Salyut

The DOS-7K No. 3 was the first modified version of the Salyut space station developed at the TsKBEM design bureau in Podlipki near Moscow, after its two practically identical vehicles DOS-7K No. 1 and No. 2 had been launched in 1971 and 1972. The most significant drawback of the original Salyut dsigen was the fixed position of the four solar panels on both ends of the hull. In order to keep the panels exposed to the sun for an adequate power supply, the entire station would have to be oriented in the right direction and put into a spin around the sun-pointing axis. Due to the rather tenuous nature of this stabilization method, the station's thrusters had to fire often to restore the equilibrium, leading to a quick expense of the propellant. Not to mention that the spin stabilization would have to be interrupted for pointing onboard instruments at their targets. The problem could be solved by giving solar panels the capability to track the sun independently.

The newly proposed system included three custom-built solar panels attached to the middle section of the vehicle via specially designed rotation mechanisms, each with autonomous capability to track the sun. During the ascent to orbit, the panels would be protected by the fairing.

Another major improvement came with the more capable Kaskad attitude-control system developed under leadership of Boris Raushenbakh. It could operate either in high-efficiency mode saving propellant or switch to a fast mode by activating larger attitude control thrusters. The attitude control system could operate in conjunction with the new ion orientation sensor, which was able to monitor the position of the spacecraft in space along two axis simultaneously.

The station was also equipped with the experimental Delta navigation system developed at TsKBEM's Department 311 led by O.I. Babkov. (52) It allowed the crew onboard to independently forecast and determine flight parameters. (685)

Also, the station's thermal control system incorporated for the first time innovative thermal pipes developed at Department 506 led by A.V. Puchinin.

Additional improvements included the SRV-K closed-cycle water system for the crew which extracted condensate from the station's atmosphere and converted it into drinking water. The system was designed at Department 511 under management by I.V. Lavrov.

Thanks to all the improvements, the projected life span of the second-generation station was increased from 90 to 180 days.

The station's payload assortment also grew, including an X-ray telescope-spectrometer, the RT-4 X-ray mirror telescope, the ITS-K infra-red telescope spectrometer and other instruments. (52)

The preliminary design of the modified station, designated DOS-7K No. 3 or DOS-3, had been issued at the TsKBEM design bureau in 1970, or well before the original Salyut had a chance to fly.

New rotating solar panels, replacing the simple fixed arrays from Soyuz, and other upgrades aboard the second-generation Salyut pushed its mass beyond the capabilities of the UR-500K rocket. To bring it back into the limits, engineers eliminated some propellant tanks. In turn, the station's operational altitude was increased to 350 kilometers in order to reduce the amount of propellant for maintaining orbit. (52)

In the meantime, the initial parking orbit, where the rocket would release the heavier station was reduced accordingly. Counting possible errors during the orbital insertion process, the station was predicted to stay aloft from three to four days before reentering the atmosphere. Additional errors in prediction of atmospheric density at the planned altitude could cut the station's life in the initial orbit to just two days. Moreover, the lower the altitude of this fast-decaying orbit, the more propellant would be required to boost the station to a safe altitude. As a result, the mission planners prioritized the orbit-raising maneuver above everything else in the initial flight timeline, even at the expense of the in-orbit testing of the flight control system which would ultimately prove fateful. (685)

In April 1973, Vasily Mishin, Chief Designer at TsKBEM, documented the cost of the Salyut project. At the time, the entire program was estimated at 308 million rubles with 260 million already spent. The cost of one DOS station and its UR-500K launch vehicle were put at 17.5 and 6.5 million rubles respectively, totaling 24 million.

The Soyuz 7K-T crew vehicle and its launcher, for delivering expeditions, were put at 6 and 1.6 million respectively, totaling at 7.6 million. As a result, the 24-milion DOS-3 project, including three 7.6-million Soyuz vehicles, was estimated at around 47 or 50 million rubles. (774)

Building DOS-3 space station and planning its mission

In the Cold War competition, the DOS-3 was rushed into production in order to reach orbit ahead of the planned launch of the US Skylab, which was promised to host at least three expeditions lasting from one to three months. To beat that Soviet specialists apparently considered doubling the resources aboard DOS-3 to sustain longer expeditions.

Inside the USSR, the Salyut project competed for time and resources with the military Almaz stations. There was an initial plan to alternate Salyut and Almaz launches and allocate Soyuz crew ships with even production numbers to one station, with the ships with odd numbers going to the counterpart program. There were also proposals for expeditions to different stations to overlap each other in orbit from a week to 10 days for various joint experiments and public activities like inter-station video sessions. However, numerous delays and failures in both programs quickly derailed these plans.

As of September 1972, crew vehicle allocations between the programs looked as follows:

  • Almaz (1) station: Soyuz 7K-T No. 61, 62, 63 (to be refurbished for Almaz), 64, 65;
  • DOS-3 station: Soyuz 7K-T No. 36, 37, 35 (to be refurbished for DOS-3);
  • DOS-4 station: Soyuz 7K-T No. 39, 40, 41. (774)

The dual space station program presented some challenges for mission control, which at the time was centered near Yevpatoria on the Crimean Peninsula. The facility could track two stations simultaneously, but it had telemetry processing capacity to support only one crew in orbit at a time. For each Soyuz 7K-T crew vehicle, ground control had to handle 256 one-time commands and monitor from 500 to 600 telemetry parameters. These numbers grew significantly with the introduction of Salyut and Almaz stations, overwhelming ground facility that had to process and display all that information. During the docked flight of the Salyut and Soyuz spacecraft, mission controllers could discontinue monitoring some of the parameters aboard Soyuz, but the transport ship would still require some attention despite the main data workload coming from the station.

DOS-3 crews

A total of four pairs of cosmonauts, that were officially formed on Oct. 10, 1971, for expeditions to the DOS No. 2 space station, had been switched to the DOS-3 project, after DOS-2 was lost in a launch mishap in 1972. The crews included Aleksei Leonov and Valery Kubasov, V. Lazarev and Oleg Makarov, A. Gubarev and Georgy Grechko, Petr Klimuk and Vitaly Sevastyanov.

The first three of these crews also underwent training in August and September 1972 for autonomous test flight aboard the Soyuz 7K-T No. 34, but after the mission had been cancelled, all four crews focused on preparations for expeditions to DOS-3, then planned for 1973.

Final development

On Dec. 14, 1972, the collegium of the Ministry of General Machine-building, MOM, gathered to review the DOS-3 project. The high-profile meeting was chaired by Minister Sergei Afanasiev, Head of MOM. According to the plans, the launch of the 7K-T No. 36 crew vehicle, carrying the first two cosmonauts to DOS-3, was to take place between two and 10 days after the launch of the station.

The general assembly of the DOS-3 station was largely completed at the Khrunichev plant in Moscow in the Fall of 1972. According to one source, the spacecraft was shipped to the Tyuratam launch site in December of that year (52), however, Mishin's notes indicate that it could be referring to a delivery to the checkout facility in Podlipki for tests and resolution of various technical issues, including problems with the newly developed driving mechanisms of solar arrays. A total of 311 replacements had to be performed, including the removal of 39 failed components.

According to Mishin's account, DOS-3 was shipped to Tyuratam during the night from Feb. 18 to 19, 1973, and it was expected to arrive at the launch site in Kazakhstan on Feb. 23, 1973. At the time, the completion of the work on the station propulsion and life-support system was slightly lagging behind schedule. (774)

In parallel, TsKBEM was completing two Soyuz crew ships for the project. Vehicle 7K-T no. 36 was shipped to Tyuratam on Jan. 27, 1973, and its integrated testing at the processing building at Site 2 was scheduled to be completed on February 21. However, the spacecraft was still waiting the delivery of its parachute system that was modified after the loss of the Soyuz-11 crew in 1971.

Also in February 1973, TsKBEM was completing tests of Vehicle 7K-T No. 37 at its KIS facility in Podlipki. Tests revealed issues with 13 instruments, but the overall completion of the ship was also held by the delay of parachute systems.

On March 12, 1973, Mishin had important phone calls with Minister Afanasiev, Head of State Commission Kerim Kerimov, as well as key manager Vladimir Barmin, responsible for launch infrastructure. They agreed on the following timeline for the upcoming launches:

Military Almaz OPS-1 (to be announced to general public as Salyut-2):

  • March 27-29 (launch);

Soyuz 7K-T No. 61:

  • March 30 (Delivery to the fueling station);
  • April 7 (launch);
  • April 22-25 (landing);

DOS-3 (to be announced to general public as Salyut-3):

  • April 26-28 (launch);

Soyuz 7K-T No. 36

  • After May 2 (launch).

As of March 22, Mishin still discussed with Barmin the possibility of launching DOS-3 around April 26, but on April 16, he called Ustinov, when he proposed to launch DOS-3 on May 6 and Soyuz 7K-T No. 36 on May 8, 1973.

DOS-7K No. 3 reaches launch pad

On April 28, 1973, the meeting of the Military Industrial Commission in Moscow set the launch of the DOS-7K No. 3 space station for May 8, 1973. The launch would come only two months after the first attempt to launch the military Almaz space station, OPS-1, which ended in failure.

On May 3, 1973, Vasily Mishin flew to Tyuratam to oversee final preparations of the DOS-3 station.

The next day, the UR-500K rocket integrated with the lab was rolled out to the launch pad at Site 81. As the tests of the payload section on the pad proceeded smoothly from 16:00 until 20:00 (Moscow Time), Mishin called Dmitry Ustinov, who supervised the industry for the Soviet government, from a command post at a near-by Site 83 and reported on the progress of the work.

On May 6, at 14:30 Moscow Time, Mishin and other managers drove to the Krainy airfield to meet the four cosmonaut crews training for expeditions to DOS-3. Later in the day, Mishin also reviewed the processing of Soyuz 7K-T vehicles No. 36 and No. 37, which were assigned to deliver cosmonauts to the station. Finally, at 17:00, the State Commission met at the launch facility at Site 81 to discuss preparations for the launch of the UR-500K rocket with the DOS-3 station.

On May 7, from 10:00 to 12:00, Mishin toured the processing building at Site 2, where the Soyuz 7K-T No. 36 spacecraft was being prepared for integration with the payload fairing. (774)

First launch attempt

Around midnight on May 8, 1973, Mishin and other top officials arrived at Site 81, where the UR-500K rocket was in three-hour readiness for liftoff scheduled at 03:20 Moscow Time. The preparations went smoothly until the 20-minute mark before launch, when the propulsion system was declared ready for launch after purging of the tanks, but a brownish cloud of propellant vapors was detected near the aft section of the first stage. The leak was traced to an oxidizer valve on the first stage and, at the 10-minute readiness mark, preparations for launch were aborted.

Launch managers made the decision to drain propellant from the rocket and cancel the launch attempt until a full investigation into the problem. The search for a problem appeared to be still ongoing, when the technical management met at 10:00 on May 9. Only at another such gathering exactly 24 hours later, did the accident commission report that inadequately tightened bolts were the reason for the leak. At the May 10 meeting, specialists also proposed to replace some sealants in the affected valves and, after pressurization tests, go for another launch attempt on May 11. (774)

However, according to some sources (231), Mishin went into a rage and refused to certify the rocket for flight, demanding a replacement vehicle, despite assurances from Vladimir Chelomei, the Chief Designer of the UR-500K rocket, that the problem was fixable. Mishin's attitude, while often excessive, was not entirely unreasonable in this case, because the painful memory of losing the DOS-2 station to a launcher failure a year earlier was still very fresh. However, the replacement of the rocket would delay the DOS-3 launch by one or even two months, pushing the program well behind the American Skylab, (set to go on May 14), and it would also eat up the limited life span of the Soviet station.

In the end, Anatoly Kiselev, then Deputy Director of the Khrunichev plant, which manufactured UR-500 series, led a repair team that replaced the valve on the pad. However, even after the success of the repair, Mishin reportedly demanded a new rocket. Only after many appeals by managers at TsKBEM and members of the State Commission, did Mishin finally budge and give the green light to the fueling of the original launcher for a second launch attempt on May 11. (52)

Race for time


Propulsion section of the second generation Salyut.

After a three-day delay, the UR-500K launch vehicle with the DOS-7K No. 3 station lifted off from Tyuratam on May 11, 1973, at 03:20:00 Moscow Decree Time. All three stages of the rocket worked well and the 19.4-ton station successfully entered a 206 by 225-kilometer initial orbit with an inclination 51.6 degrees toward the Equator.

In the first hours of the mission, ground stations, spread across the USSR along the station's ground track, began a complex series of actions designed to bring the station to its operational orbit in the shortest time.

Around nine minutes after launch, the NIP-3 facility in Sary Shagan in Kazakhstan confirmed to the State Commission in Tyuratam that the solar panels and other deployable elements of the station had opened as planned. Then, 12 minutes after launch, the NIP-15 station in Ussuriisk in the Soviet Far East transmitted commands to the spacecraft to activate its new ion orientation system. The uplinked sequence included commands to the station's flight control system to rely on high-thrust attitude control engines to more quickly dampen the tumbling of the vehicle after its separation from the rocket and put it in a correct orientation for the altitude-boosting maneuver.

The spacecraft followed the commands, but within 10 minutes, as it was passing within range of Ussuriisk, telemetry specialists noticed excessive use of propellant by the station's attitude control system. When the spacecraft re-appeared over the main nerve center of the mission at the NIP-16 station in Yevpatoria, on the Crimean Peninsula, at the beginning of the second orbit, ground controllers confirmed the severe propellant spending and urged turning off the misbehaving attitude control system, thus aborting the upcoming orbit-raising maneuver. However, while their alarm calls traveled up the chain of command to flight director Yakov Tregub and he assessed the implications of such a radical move, the station had gone beyond the horizon in Yevpatoria. Relaying the necessary abort commands to Ussuriisk also took precious time and came two minutes after the station flew out of range in the Far East. (685)

Nevertheless, at 08:00 Moscow Time on May 11, the technical management met in Tyuratam and gave the green light to the rollout of an 11A511 rocket with the Soyuz 7K-T No. 36 crew vehicle to the launch pad on a mission to deliver the first crew to the newly launched station.

At the time, Mishin and other top officials in Tyuratam apparently still thought that the problem on the station was manageable.

Available records give a confusing timeline of the mission management realizing the gravity of the situation, but what is clear is that some key people responsible for the mission were out of the loop at the most critical moments.

Mishin's diaries indicate that "everything nominal" messages, delivered during the first orbit, were followed by the news from NIP-15 in Ussuriisk, that the pressure in the tanks of the Attitude Control System, SIO, fell from 304 to 270 atmospheres (indicating propellant expenditure twice as normal) during the second orbit of the station.

By the third orbit, Mishin got the news from NIP-16 in Yevpatoria that the propellant pressure had fallen further, to 160 atmospheres, and that the ion orientation system experienced anomalies. He also wrote that during the third orbit the complete expenditure of the SIO propellant had been reported. (774)

Probably due to newly established formalities, all these reports were likely coming to Tyuratam with some delay and presented only part of the picture.

According to Chertok, who was much closer to the center of the action in Yevpatoria, when the DOS-3 re-connected with NIP-16, at the start of its third orbit, data showed no fuel left onboard! (685) In the 40 minutes, when the station was out of range between the second and third orbits, telemetry specialists in Yevpatoria had re-analyzed the downlinked data and concluded that the automated orientation mode has been sending the spacecraft into severe tumbling along the yaw and pitch axis. It was already clear that such a chaotic rotation of the spacecraft along two axis was induced by high-thrust engines on the commands from the station's own flight control system. In turn, these commands were generated based on data from the new two-axis ion orientation sensor which suffered from strong interference of the natural ion flow in the Earth's ionosphere from the exhaust of... high-thrust engines, preventing the sensor from finding reference points.

Had the specialists managed to turn off the attitude control system driven by a confused sensor during the second orbit, the station would likely have saved enough propellant for attitude control using the more traditional "IKV plus IO" mode which relied on infra-red sensors and low-thrust engines.

Mishin's notes indicate that even after the realization that all the SIO propellant had been expended, there was still hope to stabilize the station using the nitrogen gas of the pressurization system. The station's propellant tanks were equipped with inflatable membranes which pushed fuel and oxidizer into the engines. Apparently, because the fuel was used up first, while some oxidizer still remained in its tanks, the membranes separating liquid propellant components from the pressurization gas were kept intact inside all SIO tanks. The analysis showed that around 450 liters of nitrogen gas under 17 atmospheres of pressure was available in the system. By blowing up the membranes, mission control could let the gas into the system, venting it through the valves and nozzles thus creating momentum, like with the use of hypergolic propellant, just with much less energy. Calculations showed that the available gas was the equivalent of just around seven kilograms of propellant.

According to Mishin, a decision was made to vent the oxidizer remnants and then try using nitrogen gas for stabilizing DOS-3 and even for boosting its orbit. (774)

End of the DOS-3 mission

Chertok remembered that some eight hours after launch, key members of the State Commission, including authors of the troubled command sequence, arrived at NIP-16. While still on the plane from Tyuratam to Crimea, they got the word about excessive propellant usage on DOS-3, but only in Yevpatoria did they learn that the station was out of fuel, Chertok claimed. (685)

Mishin's notes indicate that he departed for Yevpatoria somewhat later, between 16:00 and 18:00 Moscow Time on May 11, and landed at the Saki airfield in Crimea at 19:30, before driving to NIP-16 in Yevpatoria.

From 22:00 to 01:00 on May 12, members of the State Commission had a long and arduous meeting with the flight control coordination group, GOGU.

Mishin also had the highest industry officials supervising the program, including Minister Afanasiev, B.A. Stroganov and B.A. Komissarov arriving at Yevpatoria at noon on May 12. From 15:00 to 21:00 Moscow Time on the same day, they all participated in the meeting of the State Commission, where Chertok, Yeliseev and Gorshkov presented reports from their respective groups.

In an attempt to hide the failure of the already orbiting space station, the DOS-3 was publicly announced as Kosmos-557 without any real details on its mission.

On the morning of May 13, most officials departed Crimea but, even that late, Mishin documented an extensive effort to save DOS-3 or at least to take the station under control using pressurization gas.

On the afternoon of May 13, (right after his return to Moscow), Mishin recorded a meeting at his TsKBEM bureau (in Podlipki) discussing scenarios for boosting the station's orbit using pressurization gas in the SIO system. However, the gathering apparently ended with the conclusion that "proposed options were not supported by available resources."

Yet another meeting 24 hours later (afternoon May 14) apparently narrowed down a feasible scenario and decided to dispatch top specialists back to NIP-16 in Yevpatoria to supervise further attempts at orbit correction. Mishin and his deputies Konstantin Bushuev and Boris Chertok were held back in Moscow by a meeting scheduled for next day at the Military Industrial Commission, VPK.

On May 15, at 09:00 Moscow Time, other top specialists from TsKBEM, Bashkin and Konstantin Feoktistov, flew to Crimea.

In the meantime, Mishin's departure to Crimea was re-scheduled for 22:00 for the VPK meeting at Ustinov's office, but the pilots had to wait much longer. The soul-searching event was held from 16:00 to 22:30 Moscow Time. As Mishin put it, everybody was lambasting the work organization at TsKBEM, as the role of mismanagement in the latest mishap was becoming increasingly apparent. Right after the unpleasant meeting, Mishin, Chertok, officials from the the Central Committee of the Communist Party Sheulov and Khodakov, as well as the Chairman of the State Commission Kerim Kerimov headed to the airport for a flight to Yevpatoria rescheduled for 23:30. They finally departed half an hour after midnight on May 16.

On May 16, 08:30 Moscow Time, Mishin and other specialists analyzed the latest measurements of the station's resources and had no good news.

It was already 02:30 Moscow Time on May 17, when TsKBEM management met with the members of the GOGU mission control group, to discuss possible actions during the station's passes within range of ground stations during five consecutive orbits during the day.

In the meantime, back in Moscow, another Mishin's deputy Viktor Legostaev supervised a series of simulations using the Kaskad attitude control system aiming to confirm the scenario for saving the station. On May 18, Legostaev called Mishin in Crimea and delivered bad news — the propulsion requirements for saving the station were beyond those available aboard DOS-3, because of the station's fast tumbling. Other teams who performed similar assessments at the TsNIIMash research institute delivered the same verdict.

Nevertheless, on May 19, mission control made several attempts to stabilize the station and perform an orbit correction using nominal ion orientation mode No. 5, but to no avail.

On May 20, Mishin's notes from Yevpatoria documented the last-ditch discussions of a controlled reentry of the DOS-3 over the Indian Ocean and indicated that mission control had made such an attempt. He returned to Moscow the next day without any good news. (774)

According to most sources, DOS-7K No. 3 reentered the Earth's atmosphere on May 22, 1973, over Bolivia, thus surviving in its low orbit for around 11 days.

Investigation with severe consequences

The consequences from the loss of the DOS-3 space station were hugely exacerbated by the fact that it was the third loss of a very expensive 20-ton space station in a row and the second such failure in a month! Moreover, DOS-3 was lost just three days before what turned out to be the troubled but ultimately successful mission of the American Skylab.

The Kremlin appointed the rising star of the Soviet space industry Vechaslav Kovtunenko, who led the KB-3 design bureau in Dnepropetrovsk, Ukraine, to chair a governmental investigative commission into the DOS-3 failure. The group also included members of the academia inside and outside the rocket industry among them Nikolai Pilyugin, Boris Bunkin and Boris Petrov.

Boris Chertok, a leading specialist in flight control systems at TsKBEM, who found himself at the center of the investigation, later outlined a multitude of factors contributing to the latest fiasco, including inadequate testing of the newly developed ion orientation sensor and poor coordination of mission control teams.

From the Soyuz flight experience, it was known that thruster exhaust can disturb the flow of ions in the ionosphere which is used by ion sensors to provide critical data to the attitude control system on the orientation of the spacecraft. The new ion sensor aboard DOS-3 replaced not one but two previous-generation devices that provided data separately for pitch and yaw axis. The new sensor had a single aperture for capturing the ions and when engine exhaust was causing interference in the ion flow, the device could produce faulty signals for two axis instead of just one. It was also known that the level of interference in the ion flow varied depending on the geographical latitude and the orientation of the vehicle relative to the Earth's magnetic field, so its exact effect was hard to predict.

Prior to the DOS-3 launch, there were no tests of the flight control system with the real ion flow, let alone any tests simulating the influence of gas exhaust on the sensor.

In hindsight, it was obvious to the investigators that such an interference-prone instrument should have been activated in flight only within range of the main ground station in Yevpatoria and only after the initial stabilization of the spacecraft using the tried and tested IKV infra-red sensors and low-thrust engines. However, it would take from two to three orbits for that mode to be implemented because the process of stabilization along the roll and pitch axis using infra-red sensors would take almost an entire orbit. In the meantime, the lower parking orbit selected for the mission meant that controllers had little time to spare before the station would lose altitude, requiring too much propellant to recover.

Ultimately, the decision to skip testing of the flight control system was recognized as a mistake made by its developers. Given the risky and time-strapped nature of the DOS-3 early flight, program managers had to at least prepare fast-reaction measures in case of likely problems. Even with the relatively crude technology of the day, it would have been possible to address the situation with adequate preparations, Chertok later believed.

Another problem uncovered in the aftermath of the DOS-3 failure was the limited communication and poor coordination within the chain of command and across the facilities involved in mission control. As in the case with the new hardware, it was the latest operational and bureaucratic procedures introduced with the intention of formalizing and properly documenting flight operations that were found overburdening and slowing down the decision-making process, which used to be more informal.

A string of planning errors

The investigation discovered that the original instructions available at the NIP-15 station in Ussuriisk called for the low-thrust attitude control mode to be transmitted to the station. At the same time, a telemetry engineer from TsKBEM, who flew to Yevpatoria ahead of the launch, also found the instructions for the low-thrust attitude control. Right before his departure for Crimea, that engineer and his boss simulated the low-thrust and maximum-thrust attitude control modes in a lab. For that simulation, these specialists had received baseline data on the possible interference to ion sensors in low- and high-thrust modes. That simulation showed that both modes worked normally, but with the low-thrust mode, the time window for the maneuver during the second orbit could be missed. Both specialists never questioned the baseline data on the interference levels in the high-thrust mode, so the manager sent an official telegram to Yevpatoria with a proposal to switch to the high-thrust attitude control. The flight manager accepted the proposal and telegraphed the updated set of commands to Ussuriisk to be relayed to the spacecraft during its first communications window.

During the 10-minute window that the station had flown within view of Ussuriisk facility, another flight control expert there was able to get access to the real-time telemetry from the station only after the newly established process of deciphering, processing and security clearance. The expert immediately noticed that in the high-thrust mode, the station was firing three thrusters in instances where his documentation called for using only one. He also discovered that the angular rotation of the vehicle was 10 times higher than the forecasted rate! "It was like a dog chasing its own tail, except that the station was doing it along two axis," Chertok remembered.

The specialist had to immediately inform the flight managers in Yevpatoria about this issue, but, instead of a simple phone call, the latest rules required an official telegram to go through the process of writing, ciphering, transmission, deciphering and registration. It took so much time that when the information reached Yevpatoria, the station had already made a loop around the Earth and was within range of NIP-16, which was preparing to send commands to the station for the orbit-raising maneuvers. Right before the maneuver, flight control specialists there received an estimate from the recently established automated telemetry analysis group, STI-90, that the spacecraft was overspending its propellant. Flight control veterans, who were used to the manual analysis of data, initially dismissed the warning as an error generated by the newly installed M-220 computer. However, one telemetry specialist got suspicions, and breaking the rules, ran to a next-door building to take a closer look at the raw telemetry tapes. Another expert from the flight control group also got very worried and did the same.

The minute the flight controller saw the actual telemetry tape, he rushed to the telephone in order to call the flight control team to immediately issue a command to deactivate the flight control system. However, the phone in his room did not work and he had to "sprint" back to the flight control building. By that time, the telemetry engineer had already gotten hold of the flight managers and urged them to immediately command the station stop maneuvering. Additional moments of hesitation by the flight director Yakov Tregub, who had to cancel mission-critical maneuvers on the insistence of two young rank-and-file engineers, probably added to the precious time lost during the passage of DOS-3 over Crimea. The situation was exacerbated by the fact that at the time, all top officials from the State Commission and key developers of the spacecraft were on the way to the airport in Tyuratam at the very start of a six-hour trip to Yevpatoria, which kept them out of the loop in the decision-making process during what turned out to be a life-and-death situation for DOS-3.

In the course of the investigation, specialists conducted a series of simulations attempting to recreate the behavior of the flight control system during the DOS-3 mission. It turned out that the level of interference producing the kind of reaction of the thrusters that had been seen in flight would have to be 10 times stronger than predicted during the planning of the high-thrust attitude control mode. (685)

Mishin generally echoed Chertok, when formulating three main reasons for the DOS-3 accident:

  • Poorly planned flight program (without testing of the Ion Orientation and selecting the reliable IKV+IO attitude-control mode among other things);
  • Not all recommendations of a (previous) accident commission had been implemented on DOS-3 (namely, one failure led to a total loss);
  • Inadequate organization of the flight control operations (technical manager was not receiving necessary information).

The political fallout

According to Chertok, in addition to top scientists outside the rocket industry, the investigation into the DOS-3 fiasco saw the involvement of the state security services. Indeed, Mishin's notes record an interaction with somebody named Samusev from the KGB on June 6, 1973.

Fortunately, the feared secret police found no proof of sabotage and filed no charges, but agreed with the engineering conclusions of the government probe about systemic failures within the industry.

Given the timing of the incident, right after two other costly failures, some heads had to roll. In an unusual move, the veteran Flight Director Yakov Tregub was dismissed from his post. There were no comparable measures even after the loss of the Soyuz-11 crew two years earlier. (52)

In addition, Boris Raushenbakh, who led the development of the flight control system, was relieved from all his administrative positions and made a consultant, while Boris Chertok, who was Deputy of Chief Designer for all control systems, received strong reprimands from the Minister of General Machine Building and through the Communist Party organization at TsKBEM. Down the chain of command, many specialists involved in flight control and ion system development, as well as flight controllers involved in the fateful mission, got their bonuses cancelled in addition to some disciplinary actions. Similar measures were taken against military personnel manning Soviet ground station network.

Nevertheless, despite some talk within the Soviet government about curtailing the space station program after three failures in a row, both the Salyut and Almaz projects continued full steam ahead. (685)

On the morning of July 12, 1973, Ustinov chaired a meeting on the DOS program which heard reports from Mishin and Afanasiev on the status of future space stations, followed by the review of the results into the DOS-3 failure investigation and proposals for improving reliability delivered by Kovtunenko.



The article by Anatoly Zak; Last update: May 11, 2023

Page editor: Alain Chabot; Last edit: May 11, 2023

All rights reserved


insider content



Cutaway view of the working compartment aboard the second generation Salyut space station. Copyright © 2001 Anatoly Zak