FAILURE: Sea Launch - Intelsat 27 - February 1, 2013 (0656UTC)

[sdsds]

exceedance of a pre-programmed roll limit

It would be great to know exactly what this means. I'm dubious about "roll" in the sense I understand it to have in technical circles. Was this really a "pitch" or "pitch rate" limit that was exceeded?

[ugordan]

exceedance of a pre-programmed roll limit

It would be great to know exactly what this means. I'm dubious about "roll" in the sense I understand it to have in technical circles. Was this really a "pitch" or "pitch rate" limit that was exceeded?

Since this was posted on the SeaLaunch website, it's unlikely to be a mistranslation and is likely meant to really mean roll, either roll angle or roll rate (I'd bet on the latter). Loss of nozzle gimbal control might have first manifested itself as a roll buildup and only later as yaw/pitch rates. Roll would be virtually unnoticeable from the low quality webcast of a night launch.

[edkyle99]

I notice that the first message of this thread mentions use of "Zenit 2S #SL36" on this flight.  Meanwhile, the NASA Mission Set Database website lists the flight as "Zenit 3SL-48".  http://msdb.gsfc.nasa.gov/launches.php  Does anyone know what the numbers mean, and if one or both are correct?

 - Ed Kyle

[LouScheffer]

  A long thin object, pushed from the bottom, starting from an (almost) balanced state, will diverge slowly and then faster as gravity accelerates the divergence.  See any youtube video on smokestack demolition for an example. 

Gravity can't do that to a rocket.  Look up the pendulum rocket fallacy.

The pendulum rocket fallacy says a rocket with engines at the top will still not be stable, despite appearances.  It does not apply directly here for two reasons:  the engines are not at the top, and no-one is claiming it should be stable.  But maybe your point is that gravity always acts on the center of mass, and causes no torques, and so (in the case where it does not interact with the guidance system) it does not accelerate the divergence.  This is a good point, which I had wrong.

However, I think this does not affect the conclusion that the behavior was consistent with little or no guidance applied:

   - Pushing from the bottom (or anywhere, as you point out) is inherently unstable, and must be actively controlled, and

   - The time scale from when control is lost to large-scale divergence, for an item the size of a rocket, is a few seconds.

Both these statements still seem correct for this case.

[Garrettishere]

exceedance of a pre-programmed roll limit

It would be great to know exactly what this means. I'm dubious about "roll" in the sense I understand it to have in technical circles. Was this really a "pitch" or "pitch rate" limit that was exceeded?

Since this was posted on the SeaLaunch website, it's unlikely to be a mistranslation and is likely meant to really mean roll, either roll angle or roll rate (I'd bet on the latter). Loss of nozzle gimbal control might have first manifested itself as a roll buildup and only later as yaw/pitch rates. Roll would be virtually unnoticeable from the low quality webcast of a night launch.

I agree. Rockets can take a bit of a change in roll angle perfectly fine, but generating a high roll rate is a big problem.
Exhibit A:

[360-180]

This BIM pump - is it used on the Atlas 5?

No

[Syndroma]

It wasn't even used on Energia. Zenit-only device.

[just-nick]

exceedance of a pre-programmed roll limit

It would be great to know exactly what this means. I'm dubious about "roll" in the sense I understand it to have in technical circles. Was this really a "pitch" or "pitch rate" limit that was exceeded?

Since this was posted on the SeaLaunch website, it's unlikely to be a mistranslation and is likely meant to really mean roll, either roll angle or roll rate (I'd bet on the latter). Loss of nozzle gimbal control might have first manifested itself as a roll buildup and only later as yaw/pitch rates. Roll would be virtually unnoticeable from the low quality webcast of a night launch.

I agree. Rockets can take a bit of a change in roll angle perfectly fine, but generating a high roll rate is a big problem.
Exhibit A:

Great video! Wasn't a Titan II lost due to a roll issue too, during testing? Aerodynamically generated, saturated the thrust vector capability? I've seen a video but couldn't find it tonight.

[anik]

Does anyone know what the numbers mean, and if one or both are correct?

SL36 is a serial number of Zenit-2S two-stage rocket used by Ukranian producer within Sea Launch.

You can look at practically all those numbers here: http://www.kosmonavtika.com/lanceurs/zenit/liste/z2s.html

[jacqmans]

Are there any good photos of the first seconds of this launch

[Galactic Penguin SST]

Over at Novosti Kosmonavtiki a member has posted the following information on the launch:

- In normal conditions pitch over would begin T+11.3 s after rocket control going inertial (not sure if I got it right) and roll maneuver at  T+14 s. As in this launch the launch platform was aligned less than 2 degrees from the east-west axis, roll over is almost not required.

- On this launch the hydraulic pump (BIM) failed at T+4.5 s, afterward the rocket lost all controls, and pitch program was not executed.

- The rocket's maximum deviation of roll was reached at T+16 s (30 degrees), while the pitch and yaw did not exceed the limit of 15 degrees. Then the safety program ordered engine shutdown after the rocket cleared the pad. Shutdown occurred at T+24 s (20 s after liftoff).

- The rocket went even wilder than early reports indicated: the final azimuth was close to 270 degrees (heading west!). It can also be seen from a photo I posted before.

- The rocket impacted the ocean at T+56 s 2.5 km from the launch platform. Telemetry was received all the way till impact.

- All rocket control parameters were nominal, and the guidance system reacted correctly to the attitude disturbances, but the engine gimballing commands were not executed properly.

This BIM pump - is it used on the Atlas 5?  Hoping there's no common failure point as I'll be in California for the LDCM launch.

Looks like it wasn't - its analogue to this unit on the RD-180 was made by NPO Energomash in house (as opposed to by a Ukrainian company in the RD-171), and the design have some differences.

Are there any good photos of the first seconds of this launch

Unfortunately I doubt Sea Launch or someone else would release them publicly right now. Perhaps request Chris to try to find them on L2? 

[Syndroma]

In above post liftoff occurs at T+4 s. At T+0 s guidance system assumes control.

[Ben the Space Brit]

The TVS hardware failure theory is beginning to look pretty solid from what I'm reading here.

[360-180]

mistake link.
 http://novosti-kosmonavtiki.ru/forum/messages/forum12/topic13320/message1028043/#message1028043
see http://novosti-kosmonavtiki.ru/forum/forum12/topic13320/?PAGEN_1=25 mess #494 and http://novosti-kosmonavtiki.ru/forum/forum12/topic13320/?PAGEN_1=28 mess #552
http://novosti-kosmonavtiki.ru/forum/forum12/topic13320/?PAGEN_1=29 mess #575

[input~2]

"Zenit was uncontrollable right after launch"
http://www.interfax.ru/russia/news.asp?id=288440

[LouScheffer]

The TVS hardware failure theory is beginning to look pretty solid from what I'm reading here.

Sure, but now the question is what caused the TVS to fail?  It just broke? Could not get fuel or oxygen?  Turned off by mistake? Etc...

Until this is known (assuming it is a TVS problem) then commonality wil be hard to assess.

[sdsds]

The rocket's maximum deviation of roll was reached at T+16 s (30 degrees), while the pitch and yaw did not exceed the limit of 15 degrees. Then the safety program ordered engine shutdown after the rocket cleared the pad.

So ugordan, you were right: the report was that a roll limit was exceeded and these details confirm exactly that. FWIW I read the above to mean that at T+16 s the roll deviation was 30 degrees, which exceeded the acceptable deviation limit and triggered the automatic mission abort. It further seems to say that, at least at the time the abort was triggered, none of the other deviation limits had been reached.

To the comment from Garrettishere that, "Rockets can take a bit of a change in roll angle perfectly fine," I agree since roll doesn't directly effect trajectory, but Zenit apparently doesn't have a range safety officer; no one is comparing the instantaneous impact point to impact limit lines. So apparently (wisely) they use roll angle as a measure of overall flight control.

Finally, as suggested by LouScheffer, I hope the FROB continues to seek a deeper root cause. Still given the proximate failure was in a component not shared with RD-180 I'm hoping this mishap has no effect on the LDCM!

[Notaris]

  A long thin object, pushed from the bottom, starting from an (almost) balanced state, will diverge slowly and then faster as gravity accelerates the divergence.  See any youtube video on smokestack demolition for an example. 

Gravity can't do that to a rocket.  Look up the pendulum rocket fallacy.

The pendulum rocket fallacy says a rocket with engines at the top will still not be stable, despite appearances.  It does not apply directly here for two reasons:  the engines are not at the top, and no-one is claiming it should be stable.  But maybe your point is that gravity always acts on the center of mass, and causes no torques, and so (in the case where it does not interact with the guidance system) it does not accelerate the divergence.  This is a good point, which I had wrong.

However, I think this does not affect the conclusion that the behavior was consistent with little or no guidance applied:

   - Pushing from the bottom (or anywhere, as you point out) is inherently unstable, and must be actively controlled, and

   - The time scale from when control is lost to large-scale divergence, for an item the size of a rocket, is a few seconds.

Both these statements still seem correct for this case.

It is a little off-topic, but rockets are not inherently unstable, at least while flying in atmosphere. It is a question of center of mass location (where thrust forces are "acting") and center of pressure location (where aerodynamic forces are "acting"). At sufficient speed and air density (thus reasonable big aerodynamic forces) and with center of pressure further back than center of mass, a rocket is flying stable! A big influence on center of pressure can be achieved by fins at the bottom of the rocket (Typical case for model rockets, but also sounding rockets - though even "real rockets" (e.g. some Ariane 4 variants) have/had fins).

[Galactic Penguin SST]

Apparently RKK Energia seems to have given up on Sea Launch and is offering to the Russian government to take over it. (!) But the report is very interesting, including claims by RKK Energia CEO Vitaly Lopota that companies in China, the US, Australia and Ukraine are interested in buying SL, and that Lockheed Martin's Atlas V is a "perfect fit" for the Sea Launch platform (!?)......

http://www.russianspaceweb.com/sealaunch.html#bankruptcy

Original article in Russian: http://lenta.ru/news/2013/02/04/seastart/

[pippin]

though even "real rockets" (e.g. some Ariane 4 variants) have/had fins).

Heck, even Saturn V had them...