In the eighth minute of the flight, a fire started in the engine compartment of the second stage. It ended in an explosion, and not from the self-detonation system, as it should have, but as a result of uncontrolled gorenje. This was not the case when testing the first version of the second stage of Starship V1. Almost certainly, the root of the problems was the change in parameters when creating the second version of the ship.
Today, at 2.30 Moscow time, SpaceX conducted the eighth test launch of its Starship system. As in the seventh flight, the second stage was new — Starship V2, which had not been launched before this year. The volume of its tanks has increased by a quarter, they hold 1,500 tons of fuel, not 1,200, as in Starship V1. In addition, it is 1.8 meters longer than the previous one. Due to this, the total length of both Starship stages increased to 123.1 meters. The special importance of Starship V2 is that it should become the basis for Starship HLS, which will return humanity to the moon in the next few years.
However, the eighth flight was largely according to the scenario of the seventh: the first stage worked fine, but the second did not. The first one, as in the previous two times, was landed on a Mechazilla (6 minutes 57 seconds after launch). There were a couple of engines that didn't restart to brake on time.
But since Starship, like the N-1 Korolev, has dozens of engines in the first stage, the failure of some of them still does not threaten normal operation. Unfortunately, the scenario of the seventh flight was repeated by the second stage: it died 20 seconds before the first scheduled engine shutdown.
The video broadcast from the ship's engine compartment shows the start of the fire about 7 minutes 44 seconds after takeoff. Apparently, the fire started in the area of one of the Raptor vacuum engines (the ship has three of them, and three more atmospheric versions of Raptor). After 21 seconds, the engines started shutting down. At 8.07 a.m., a moderate explosion occurred. After that, only two of the six engines were working, which caused the ship to spin around its own axis. At 9.09 a.m., another engine shut down, and at 9.35 a.m., communication with the stage was lost.
The beginning of a fire in the engine compartment
Image source: SpaceX
It looks like the remaining fuel detonated around that time. The debris scattered over a significant area, causing a temporary halt to flights from several airports. Since all engines were running at the time of the fire, the most likely cause was a rupture of pipelines and leakage of fuel components into the engine compartment.
This happened despite the pipeline evacuation system installed on the Starship V2 version of the ship and was absent on the Starship V1. At the same time, Starship V1 flew, and without such explosions. The cause of the leak may be different, but so far industry observers suspect two main ones: harmonic vibrations of the ship, and, less likely, fluctuations of the "Pogo" type.
The first often occur when the rocket is lengthened during modernization without increasing the rigidity of the structure. Aerodynamic forces act more strongly on a rocket with a changed length, and this is especially evident as fuel consumption increases, when the weight balance of the rocket changes dramatically, and the whole rocket becomes many times lighter.
Catching the first stage this time was tougher than ever before. In other words, the speed when touching the Mechazilla was higher, and the fuel consumption for braking before picking up, on the contrary. less than before. This is a deliberate policy of the company, trying to minimize fuel costs when fishing the first stage. In theory, the second one will have the same path, after it has been worked out sufficiently to start fishing.
Image source: SpaceX
Since the problems started shortly before the scheduled engine shutdown time, harmonic fluctuations are the most likely reason for the loss of Starship V2 in both the seventh and eighth launches. In the 1960s, similar fluctuations led to the death of both Korolev and von Braun rockets. Usually, the problem was solved with additional stiffeners welded from the inside. However, before the eighth Starship flight, additional reinforcements were installed inside the ship. Nevertheless, the scenario of the seventh flight was repeated.
A less likely cause of the ship's demise — org/wiki/%D0%9A%D0%BE%D0%BB%D0%B5%D0%B1%D0%B0%D0%BD%D0%B8%D1%8F_%D1%82%D0%B8%D0%BF%D0%B0_%C2%AB%D0%BF%D0%BE%D0%B3%D0%BE%C2%BB" target="_blank" rel="nofollow">Pogo type fluctuations . They are triggered by the instability of combustion in the engine chamber. Although this was a common cause of accidents on Soviet and American rockets in the 1960s, it is doubtful in the case of Starship. Its stages are undergoing bench tests on the Ground, during which the combustion in Raptor engines shows excellent stability. Gorenje Of course, such a test does not reproduce the conditions of combustion at high altitude, where the pressure of the outside air is low.Gorenje At the same time, it is not obvious why a vacuum Raptor could operate less stably than an atmospheric Raptor at sea level.
Despite the fact that it is too early to definitively diagnose the causes of the death of the ships in the seventh and eighth flights, the general conclusions from the incident still suggest themselves. Elon Musk, the technical director and chief engineer of the company, concentrates all the time on working out the heat shield, rightly considering this to be the most important problem of reusable ships. But it seems that in doing so, he did not pay enough attention to more traditional rocket problems, such as harmonic vibrations.
At the same time, if it's just them, then this problem should be relatively easy to solve by reinforcing the structure with stiffeners. If the root of the evil is in the operation of vacuum Raptors, then it takes a little longer to find a way out. But even this is quite realistic in a moderate period of time. In parallel, it is worth noting the flawless capture of the first stage of the Starship three launches in a row. This shows the impressive reliability of Mechazilla and the high prospects for reuse of the first stage of the system in the coming months.
Alexander Berezin
