Facts and hypotheses about the accident "Union MS-10"
Haste is evil. You reread the publication with the chronicle of the accident “Union MS-10”, which was carried out hot on the heels, and you find errors. But now, when the dust has settled, the hands are not shaking from the situation “the ship went to an emergency landing, and the crew lost contact,” and in open sources there appeared better materials, it’s time to understand what happened.

Track of the Soyuz MS-10 flight on long exposure, photo by NASA / Bill Ingals
There were two translations of the launch, from NASA-TV and Roskosmos. More complete is the first, there are heard the crew talks with the ground after the accident.
Both broadcasts successfully complement each other - Roskosmos shows the view from the ground, and NASA-TV data from the camera inside the ship.
Having superimposed the time from the video on the timeline of the flight, we obtain the following sequence of events.
114 seconds: according to the flight program, the emergency rescue boom is separated.
117 seconds: the process of separating the first stage begins, which consists of several stages. He is clearly abnormal. NASA photographer Bill Ingals takes a photo that clearly shows the immediate cause of the accident - one of the four first-stage blocks was somehow separated differently and did not have time to go the same distance as the rest of the staff. The white dot is lower - the SAC boom dropped, this is normal.

NASA / Bill Ingals Photos
In the area of 119 seconds from the ground, we see a cloud caused by damage and depressurization of a tank, most likely the second stage. The view inside the ship shows a noticeable decrease in overload (normal for separating the first stage) and then a serious jerk to the left (obviously abnormal). In the next photo of Bill Ingals, we still see the fourth block that has departed and some fragments that do not happen in normal separation.


Alas, instead of real telemetry, Roskosmos broadcasts animation, how a normal flight should go, so the picture transmitted to them on NASA-TV ceases to correspond to reality. Most likely, communication with the ship disappears, because in the broadcast of Roskosmos, the announcer continues to read the report on paper as during normal flight (in case of problems, the announcer used to switch to the countdown of seconds without comment, and he reported on the apparent accident on the air and stopped reporting).
In the area of 160 seconds, the audio of the Soyuz ship's shortwave transmitter appears, transmitting the “AN” - “carrier crash” to the Morse code and the report by cosmonaut Alexei Ovchinin: “Two minutes and 45 seconds. Carrier accident. In the broadcast Roskosmos promptly turn off the audio channel, then the speaker stops talking, and after a few minutes everything is turned off. The broadcast NASA-TV heard the dramatic negotiations of the crew in the ship, going to an emergency landing. As a result, the ship landed, the astronauts, alive and well, were evacuated.
So, from the photo and video it is obvious that the cause of the accident was an abnormal separation of the steps. Can you try to say more precisely?
How is the separation of the side blocks (first stage) from the central (second stage)? This process is described in the most detailed and clear way on the KIK website of the USSR . In a brief retelling, it consists of the following steps.

Starting from Kuru, ESA video
At first, the engines of the side blocks are switched to the low thrust mode, and the steering engines are turned off. Further, the lower links of the side blocks are broken, and since the direction of thrust of the engines of the side blocks is specially made not coinciding with their axis, the lower parts of the side blocks begin to diverge to the sides.

Orbiter Simulator Screenshot
Then their engines are turned off, and since the second stage engine is running, the side blocks lag behind, move backwards and exit the upper mounts.

Top mount, photo KIK USSR

The red circle in the center is the valve cap of the oxidizer tank, photo KIK USSR
When the side blocks move a sufficient distance, the valve of the oxidizer tank opens, the jet stream from which spins the blocks. Well, in the last turn, the valve of the fuel tank opens, a jet of pressurized gas from which it additionally twists and leads away the stage. It is important to note that the upper part of the blocks passes a rather complicated path - first along the step down, then to the side.

KIK scheme of the USSR
What could have happened to the emergency unit? There are a lot of options - the engine finished working earlier, the engine worked longer than expected, the lower mounts did not work and the lower mounts did not disperse, a problem arose in the upper mounting, the oxidizer valve did not open. Do we have any additional information? Yes - we saw on the video how the astronauts were jerked to the left. But which way is it "left" on a rocket?
Here is a photo of the launch of the Union MS-10 from the same Bill Ingals.

If you look closely, you can see three protrusions that allow you to determine the position of the ship under the fairing.

These protrusions are VSK fairings (“Special cosmonaut Vizir” - a device for manual docking and orientation, in fact, a periscope) and two infrared vertical sensors.

NASA Photos
The rocket did not turn over in flight, so these fairings were at the top and at the time of the accident. VSK is installed under the feet of the astronauts (hence, by the way, an interesting conclusion - the Soyuz, like, for example, the shuttle, goes into orbit upside down). And the jerk to the left, which is visible in the broadcast, was in the direction of the emergency unit. This makes it less likely that additional out-of-standard versions of it are used, and it is more likely that, on the contrary, it somehow pulled the missile along with it. So, the leaked version of the oxidizer tank valve in the media looks quite plausible - if the unit hit the second stage, the power cone on its upper part ripped its wall and dragged it along, the astronauts would just pull in its direction.
But It is worth noting that there is no trust of the “favorite sources in the industry”, and the causes of similar accidents were very diverse. On the forum of the magazine "Astronautics News" led the list:
According to the cyclogram, the emergency rescue system bar is reset to 114 seconds. But the CAC of the Soyuz ship consists of two parts - a rod and engines on the head fairing.

SAS scheme of the Soyuz ship, an illustration of a corporation of tactical rocket armament
From -15 minutes to 114 seconds in the event of an accident, engines on the rod and head fairing are triggered (from 4:45, RDG work is clearly visible at 4:52).
From 114 seconds to 157 seconds in the event of an accident, the ship is diverted on the RDG engines in the head fairing.

The screenshot of the Orbiter
A simulator in the event of an accident after 157 seconds, when the head fairing is reset, is separated from the third stage by regular pyrobolts, as if it went normally into orbit.
Accordingly, once the accident occurred at 119 seconds, the head fairing has not yet been reset. And according to the information from the Cosmonautics News Forum, when the first stage is separated, the automatic control system of the SAS is turned off for 6 seconds so that the control system of the launch vehicle can cope with disturbances. When SAS switched on again at 123 seconds, it estimated the position of the rocket as abnormal and tore the ship from it. The connection was lost, because the power supply and equipment remained in the instrument-aggregate compartment, which the CAC does not save. And already in the region of 160 seconds, when the descent vehicle was dropped from under the head fairing, its autonomous communication system was turned on, which is used during landing during landing. This, by the way, explains the accident information that appeared in the media at 123 seconds.
According to the latest news, the emergency commission will work until October 25. Also, information appeared in the media that the video control system was on the rocket, a picture with which we might even someday be shown. A serious accident with the destruction of the rocket, after which the crew, through the efforts of automation, is alive and well, leaves something inappropriate sensation of victory, and it will be very interesting to see these shots.

Track of the Soyuz MS-10 flight on long exposure, photo by NASA / Bill Ingals
Chronology
There were two translations of the launch, from NASA-TV and Roskosmos. More complete is the first, there are heard the crew talks with the ground after the accident.
Both broadcasts successfully complement each other - Roskosmos shows the view from the ground, and NASA-TV data from the camera inside the ship.
Having superimposed the time from the video on the timeline of the flight, we obtain the following sequence of events.
114 seconds: according to the flight program, the emergency rescue boom is separated.
117 seconds: the process of separating the first stage begins, which consists of several stages. He is clearly abnormal. NASA photographer Bill Ingals takes a photo that clearly shows the immediate cause of the accident - one of the four first-stage blocks was somehow separated differently and did not have time to go the same distance as the rest of the staff. The white dot is lower - the SAC boom dropped, this is normal.

NASA / Bill Ingals Photos
In the area of 119 seconds from the ground, we see a cloud caused by damage and depressurization of a tank, most likely the second stage. The view inside the ship shows a noticeable decrease in overload (normal for separating the first stage) and then a serious jerk to the left (obviously abnormal). In the next photo of Bill Ingals, we still see the fourth block that has departed and some fragments that do not happen in normal separation.


Alas, instead of real telemetry, Roskosmos broadcasts animation, how a normal flight should go, so the picture transmitted to them on NASA-TV ceases to correspond to reality. Most likely, communication with the ship disappears, because in the broadcast of Roskosmos, the announcer continues to read the report on paper as during normal flight (in case of problems, the announcer used to switch to the countdown of seconds without comment, and he reported on the apparent accident on the air and stopped reporting).
In the area of 160 seconds, the audio of the Soyuz ship's shortwave transmitter appears, transmitting the “AN” - “carrier crash” to the Morse code and the report by cosmonaut Alexei Ovchinin: “Two minutes and 45 seconds. Carrier accident. In the broadcast Roskosmos promptly turn off the audio channel, then the speaker stops talking, and after a few minutes everything is turned off. The broadcast NASA-TV heard the dramatic negotiations of the crew in the ship, going to an emergency landing. As a result, the ship landed, the astronauts, alive and well, were evacuated.
So, from the photo and video it is obvious that the cause of the accident was an abnormal separation of the steps. Can you try to say more precisely?
Materiel
How is the separation of the side blocks (first stage) from the central (second stage)? This process is described in the most detailed and clear way on the KIK website of the USSR . In a brief retelling, it consists of the following steps.

Starting from Kuru, ESA video
At first, the engines of the side blocks are switched to the low thrust mode, and the steering engines are turned off. Further, the lower links of the side blocks are broken, and since the direction of thrust of the engines of the side blocks is specially made not coinciding with their axis, the lower parts of the side blocks begin to diverge to the sides.

Orbiter Simulator Screenshot
Then their engines are turned off, and since the second stage engine is running, the side blocks lag behind, move backwards and exit the upper mounts.

Top mount, photo KIK USSR

The red circle in the center is the valve cap of the oxidizer tank, photo KIK USSR
When the side blocks move a sufficient distance, the valve of the oxidizer tank opens, the jet stream from which spins the blocks. Well, in the last turn, the valve of the fuel tank opens, a jet of pressurized gas from which it additionally twists and leads away the stage. It is important to note that the upper part of the blocks passes a rather complicated path - first along the step down, then to the side.

KIK scheme of the USSR
What could have happened to the emergency unit? There are a lot of options - the engine finished working earlier, the engine worked longer than expected, the lower mounts did not work and the lower mounts did not disperse, a problem arose in the upper mounting, the oxidizer valve did not open. Do we have any additional information? Yes - we saw on the video how the astronauts were jerked to the left. But which way is it "left" on a rocket?
Here is a photo of the launch of the Union MS-10 from the same Bill Ingals.

If you look closely, you can see three protrusions that allow you to determine the position of the ship under the fairing.

These protrusions are VSK fairings (“Special cosmonaut Vizir” - a device for manual docking and orientation, in fact, a periscope) and two infrared vertical sensors.

NASA Photos
The rocket did not turn over in flight, so these fairings were at the top and at the time of the accident. VSK is installed under the feet of the astronauts (hence, by the way, an interesting conclusion - the Soyuz, like, for example, the shuttle, goes into orbit upside down). And the jerk to the left, which is visible in the broadcast, was in the direction of the emergency unit. This makes it less likely that additional out-of-standard versions of it are used, and it is more likely that, on the contrary, it somehow pulled the missile along with it. So, the leaked version of the oxidizer tank valve in the media looks quite plausible - if the unit hit the second stage, the power cone on its upper part ripped its wall and dragged it along, the astronauts would just pull in its direction.
But It is worth noting that there is no trust of the “favorite sources in the industry”, and the causes of similar accidents were very diverse. On the forum of the magazine "Astronautics News" led the list:
1. 06/20/1967 - Voskhod - Plesetsk - Emergency shutdown of the propulsion system of the central unit A of the second stage PH at 124.39 sec due to the opening of the fuel tank of the central unit A of the second stage due to collision of the side unit D of the first stage with the central unit A of the second stage at 122 sec at separation due to non-disconnection of the telemetry connector.Two more questions remain unclear - why did the alarm signal appear at 160 seconds, and did the emergency rescue system work?
2. 03/28/1981 - Soyuz-U - Baikonur - Emergency shutdown of the propulsion systems of the central unit A of the second stage and side blocks B, C, D and D of the first stage PH at 118.13 s due to a collision of the side unit B of the first stage and the central unit A of the second stage 117.7 sec due to non-detachment of side block B as a result of disconnection of the plug-in plug Ш144 of side block B at the estimated time in the process of regular separation of the side and central blocks due to the failure of the DP-2 relay type DP-12 in the 11L1126 device or the grinding of individual veins in the B11U cable in flight due to accidental damage to the metal hose of the cable.
3. 03/26/1986 - Soyuz-U - Baikonur - Emergency separation of the side block G of the first step from the central block A of the second step of the PH due to the non-retraction of the upper part of the side block G due to the failure of the command to receive the keypad БУ302-0М opening the cover Б4300-845М the jet nozzle of the oxidizer tank due to the non-exit of the stem of the B4411-0A contact sensor as a result of its jamming in the housing due to the deformation of the stem when assembling the “package” at the cosmodrome, caused by the joining of the side block G with a random deviation from the requirements of technical documentation and the mutual position of the blocks at the time of input connections and a thrust bearing spherical bearing
About the work of SED
According to the cyclogram, the emergency rescue system bar is reset to 114 seconds. But the CAC of the Soyuz ship consists of two parts - a rod and engines on the head fairing.

SAS scheme of the Soyuz ship, an illustration of a corporation of tactical rocket armament
From -15 minutes to 114 seconds in the event of an accident, engines on the rod and head fairing are triggered (from 4:45, RDG work is clearly visible at 4:52).
From 114 seconds to 157 seconds in the event of an accident, the ship is diverted on the RDG engines in the head fairing.

The screenshot of the Orbiter
A simulator in the event of an accident after 157 seconds, when the head fairing is reset, is separated from the third stage by regular pyrobolts, as if it went normally into orbit.
Accordingly, once the accident occurred at 119 seconds, the head fairing has not yet been reset. And according to the information from the Cosmonautics News Forum, when the first stage is separated, the automatic control system of the SAS is turned off for 6 seconds so that the control system of the launch vehicle can cope with disturbances. When SAS switched on again at 123 seconds, it estimated the position of the rocket as abnormal and tore the ship from it. The connection was lost, because the power supply and equipment remained in the instrument-aggregate compartment, which the CAC does not save. And already in the region of 160 seconds, when the descent vehicle was dropped from under the head fairing, its autonomous communication system was turned on, which is used during landing during landing. This, by the way, explains the accident information that appeared in the media at 123 seconds.
Conclusion
According to the latest news, the emergency commission will work until October 25. Also, information appeared in the media that the video control system was on the rocket, a picture with which we might even someday be shown. A serious accident with the destruction of the rocket, after which the crew, through the efforts of automation, is alive and well, leaves something inappropriate sensation of victory, and it will be very interesting to see these shots.