On May 29, 2015, Roskosmos announced the results of the investigation into the MexSat-1 launch failure. According to the official statement, the steering engine of the third stage failed due to excessive vibration loads, which had been caused by an increasing imbalance of the rotor in the turbopump. The problem was linked to the degradation of the material of the rotor under the influence of high temperatures and to the poor balancing system. The failure was characterized as a design flaw.
According to the agency, the head of Roskosmos Igor Komarov directed GKNPTs Khrunichev and its branches to develop a plan of measures aimed to resolve the issue, including:
- To replace the material making up the shaft of the turbopump rotor;
- To upgrade procedures for balancing the turbopump rotor;
- To upgrade the attachment of the steering engine turbopump to the framework of the main engine.
The agency also announced that the investigation had revealed a number of problems in the management of the quality control issues within the wider industry and promised to develop a plan of measures to resolve them within a month. The launch date for the next Proton mission would be announced in June 2015, Roskosmos said.
During a briefing with reporters on the same day, Deputy Head of Roskosmos, Aleksandr Ivanov said that the quick identification of the culprit in the MexSat-1 failure had become possible thanks to measures taken
during a similar Proton accident exactly a year earlier. At the time, the route cause of the failure was mistakenly characterized as a production defect, triggering a massive inspection and re-certification of already manufactured hardware at the Voronezh Mechanical Plant, VMZ. However in addition, new vibration sensors were installed in the turbopump of the engine for future launches. The telemetry from those sensors complemented by ground tests, including a live firing of the engine, enabled to finally re-qualify the issue as an old design flaw rather than poor production. As it turned out, under certain border-line conditions, the shaft of the turbopump tends to fail, even though, it had actually happened in just three launches out of more than 400 Proton missions since 1965.
Aleksandr Medvedev, First Deputy Director at GKNPTs Khrunichev, confirmed during the briefing that the Proton failure on Jan. 18, 1988, had also stemmed from the same design flaw. However, at the time, the rocket completely lacked vibration load sensors and investigators had to work in the dark in search for a culprit, Ivanov added.
The failed rocket of 2014 did carry some vibration sensors on the frame of the engine, however that location turned out to be too far from the turbopump to correctly pinpoint the problem, Ivanov explained. Moreover, investigators into the 2014 failure were under the influence of preceding accidents, which prompted them to focus on the production defects and quality control issues. (In 2013, Proton failed seconds after liftoff due to wrong installation of flight control sensors. In 2010, Proton plunged into the ocean due to mishandling of the fueling procedures.)
Medvedev assured that despite very careful examination of the quality control procedures, no violations of the established process had been found this time. According to Medvedev, the telemetry from the mission provided a very clear picture of the accident, while production and testing of the rocket before the flight caused most controversy. Despite that, all 34 members of the investigative commission, working on seven different aspects of the accident, ultimately came to a consensus, Medvedev said.
In his turn, Ivanov assured that the newly available information allows to fully remedy the problem. According to Komarov, the new material for the failed rotor shaft had already been identified in the wake of the May 2014 failure.
According to Medvedev, officials also concluded that the turbopump attachment system had contributed to the failure and would have to be redesigned.
Unfortunately, investigators were not able to find surviving fragments of the actual turbopump, which caused the MexSat-1 failure, at the crash site in Eastern Russia, Medvedev said. He added, that the crash site was apparently affected by fire and the effort to recover the hardware would continue. The examining the hardware would draw a line under the accident.