How one toggle switch "Electron" ruined
The accident investigation was completed at the first launch of a private ultra-light Electron launch vehicle. It turned out that the cause of the emergency termination of the flight was the improper setting of the contractor’s equipment - it was only because of one mistakenly switched off toggle switch that the communication with the rocket was interrupted, and according to the current rules we had to give a command to its emergency detonation.
Launch PH Electron, photo Rocket Lab
On May 25, after three transfers due to the weather, the first test launch of a private ultralight Electron launch vehicle took place. The rocket, distinguished by interesting technical solutions and intended for launching up to 225 kilograms to a low orbit, flew in this launch with an overall-weight payload mockup. According to the cyclogram of the flight, the first stage successfully worked for two and a half minutes and separated. The second stage launched the engine and continued the flight. At the beginning of the third minute, the doors of the head fairing were safely dropped. The rocket reached an altitude of 224 km, and on Twitter, Rocket Lab wrote about the successful achievement of space, but after four minutes of flight, communication with it through the main channel was lost. In spite of the fact that the rocket flew over the sea and could not fall, it was not the fact that in densely populated places, but generally on land, the rules were adopted, according to which, in case of loss of communication, the command was transmitted to undermine the rocket. As a result, "Electron" did not go into orbit.
The basic version of the accident appeared very quickly, but the investigation took two months to eliminate the possibility of an error. It turned out that the communication channel, with which the emergency termination operators worked, was provided by the contractor. In parallel, telemetry from a rocket via another channel was received by Rocket Lab specialists. And if telemetry began to deteriorate through the contractor's channel and eventually disappeared, then through the Rocket Lab channel a normal flight could be seen until the transfer of the command to the emergency detonation. It turns out that for some reason they blew up a normally flying rocket. And this reason was found in the equipment, more precisely, in the contractor's equipment settings — the interference-proof coding (Forward error correction, FEC) was turned off.
When the rocket was on the launch pad, it was close to the receiving devices, and the telemetry signal level was very high. After the start, it began to move away, and the signal power began to fall. More precisely, the signal / noise level began to fall - the signal from the rocket weakened, and the background radio noise remained at the same level. As a result, at some point the noise began to clog the useful signal, and the quality of telemetry fell below acceptable levels. They thought about the problem of transmitting information via a communication channel with interferences even in the middle of the last century, and even then they came up with a general principle - along with useful information, additional information is transmitted, according to which, if there are errors, you can restore what was transmitted. The simplest example from 1950 is Heming's code:
On the left - the original message. Behind the lines on the right and below are special control bits, in this case, showing the parity of the number of units of a row or column. An error has crept into the message on the right (marked in red). According to the value of the control bits, you can correct the error and restore the original message - in the second row the number of ones must be even, and in the second column also. Therefore, in the cell at the intersection of the second column and the second row there should be one, not zero.
We return to the "Electron". In addition to checking equipment settings, Rocket Lab conducted a full-scale experiment — the signal recorded from the rocket was reproduced again with a similar power level, and the contractor’s equipment received telemetry without any problems with the included noise-proof coding.
The result was a tragicomic and very instructive story - the wrong position of just one switch was enough to destroy a healthy rocket. For Rocket Lab, this is a lesson to a greater degree of work with contractors than designing missiles, an experience, albeit bitter, but useful. Solving such a problem is simple, and Rocket Lab writes that they have already changed their procedures. For astronautics, in general, such errors are not news, for example, I recall the story of how at the beginning of the broadcast of the exit to the surface of the Moon the operators of one of the communication stations, pressing the wrong buttons, first turned Armstrong upside down , and then, trying to correct the situation, also inverted colors.
The first flight of "Electron" revealed the real problem. In the video you can see that the rocket rotated slowly in the area of work of the first stage. The speed was insufficient to stabilize the rotation, and in general for a full-fledged control system, which stood on the rocket, such stabilization is not needed. But it didn’t look like a serious accident either - the rocket kept a controlled flight. It turns out that it was still a mistake in the design, and the Electron should not rotate in normal flight. Now we have learned about this error, and by the next launch it should be fixed.
The second flight of "Electron" is scheduled to take place in early October. It still will not be a real satellite. The name “Still testing” is already known. I am glad that Rocket Lab retained a sense of humor. Errors and failures, alas, are inevitable, but judging by the seriousness of the approach that is visible in the work of Rocket Lab, they will be able to draw useful lessons from them.
Launch PH Electron, photo Rocket Lab
Chronology of events
On May 25, after three transfers due to the weather, the first test launch of a private ultralight Electron launch vehicle took place. The rocket, distinguished by interesting technical solutions and intended for launching up to 225 kilograms to a low orbit, flew in this launch with an overall-weight payload mockup. According to the cyclogram of the flight, the first stage successfully worked for two and a half minutes and separated. The second stage launched the engine and continued the flight. At the beginning of the third minute, the doors of the head fairing were safely dropped. The rocket reached an altitude of 224 km, and on Twitter, Rocket Lab wrote about the successful achievement of space, but after four minutes of flight, communication with it through the main channel was lost. In spite of the fact that the rocket flew over the sea and could not fall, it was not the fact that in densely populated places, but generally on land, the rules were adopted, according to which, in case of loss of communication, the command was transmitted to undermine the rocket. As a result, "Electron" did not go into orbit.
The basic version of the accident appeared very quickly, but the investigation took two months to eliminate the possibility of an error. It turned out that the communication channel, with which the emergency termination operators worked, was provided by the contractor. In parallel, telemetry from a rocket via another channel was received by Rocket Lab specialists. And if telemetry began to deteriorate through the contractor's channel and eventually disappeared, then through the Rocket Lab channel a normal flight could be seen until the transfer of the command to the emergency detonation. It turns out that for some reason they blew up a normally flying rocket. And this reason was found in the equipment, more precisely, in the contractor's equipment settings — the interference-proof coding (Forward error correction, FEC) was turned off.
Interference and toggle switch
When the rocket was on the launch pad, it was close to the receiving devices, and the telemetry signal level was very high. After the start, it began to move away, and the signal power began to fall. More precisely, the signal / noise level began to fall - the signal from the rocket weakened, and the background radio noise remained at the same level. As a result, at some point the noise began to clog the useful signal, and the quality of telemetry fell below acceptable levels. They thought about the problem of transmitting information via a communication channel with interferences even in the middle of the last century, and even then they came up with a general principle - along with useful information, additional information is transmitted, according to which, if there are errors, you can restore what was transmitted. The simplest example from 1950 is Heming's code:
On the left - the original message. Behind the lines on the right and below are special control bits, in this case, showing the parity of the number of units of a row or column. An error has crept into the message on the right (marked in red). According to the value of the control bits, you can correct the error and restore the original message - in the second row the number of ones must be even, and in the second column also. Therefore, in the cell at the intersection of the second column and the second row there should be one, not zero.
We return to the "Electron". In addition to checking equipment settings, Rocket Lab conducted a full-scale experiment — the signal recorded from the rocket was reproduced again with a similar power level, and the contractor’s equipment received telemetry without any problems with the included noise-proof coding.
The result was a tragicomic and very instructive story - the wrong position of just one switch was enough to destroy a healthy rocket. For Rocket Lab, this is a lesson to a greater degree of work with contractors than designing missiles, an experience, albeit bitter, but useful. Solving such a problem is simple, and Rocket Lab writes that they have already changed their procedures. For astronautics, in general, such errors are not news, for example, I recall the story of how at the beginning of the broadcast of the exit to the surface of the Moon the operators of one of the communication stations, pressing the wrong buttons, first turned Armstrong upside down , and then, trying to correct the situation, also inverted colors.
Benefit and Future
The first flight of "Electron" revealed the real problem. In the video you can see that the rocket rotated slowly in the area of work of the first stage. The speed was insufficient to stabilize the rotation, and in general for a full-fledged control system, which stood on the rocket, such stabilization is not needed. But it didn’t look like a serious accident either - the rocket kept a controlled flight. It turns out that it was still a mistake in the design, and the Electron should not rotate in normal flight. Now we have learned about this error, and by the next launch it should be fixed.
The second flight of "Electron" is scheduled to take place in early October. It still will not be a real satellite. The name “Still testing” is already known. I am glad that Rocket Lab retained a sense of humor. Errors and failures, alas, are inevitable, but judging by the seriousness of the approach that is visible in the work of Rocket Lab, they will be able to draw useful lessons from them.
All Articles