Proton-M Failure Reaction and Discussion Thread - July 2, 2013

[ugordan]

From what I've read, and someone please correct me if I'm wrong, the engines gimbal in the radial direction, ie into/away from the vehicle centerline.

IMO Proton's first stage engines are vectored tangentially .

And if you think about it, if all the engines gimballed radially, you couldn't do roll control.

[Prober]

"According to a Khrunichev representative all Proton-M launches have been suspended until the commission [on Proton-M/DM03 failure from July 2] completes its work after which a new launch schedule will be drawn up."
"According to a source in the Russian space industry work at the Baikonur space center will probably be suspended for the next two or three months because of contamination."
"A scheduled launch of a Progress M-20M spacecraft from Baikonur on July 27 is likely to be delayed"


http://en.ria.ru/russia/20130702/182002715/Russian-Proton-M-Rocket-Falls-Shortly-After-Launch.html
 
reposted from plan thread above.

[muomega0]

Why no cut-engine command?

Please read the thread before replying - this has been mentioned several times. before 45 seconds, the thrust termination is not active to prevent damage to the launch pad.

Yes, sorry. No cut engine command before 45 seconds would be common sense...

No, inhibiting engine cutoff before 45 seconds is not common sense. It's a design decision, arguably a good one, but not one that will always achieve the desired result.

Pictures of gimbal angles here:

From what I've read, and someone please correct me if I'm wrong, the engines gimbal in the radial direction, ie into/away from the vehicle centerline.

IMO Proton's first stage engines are vectored tangentially .

" The thrust vector is controlled by gimbaling an engine with a hydraulic actuator within 7.5 degrees. To make this possible, the engine is mounted in the yoke bearings by means of special trunnions installed near the chamber throat. "

This is no Falcon 9....

Any all-liquid rocket, including Falcon 9, would fail in this fashion if it suffered an engine failure right off the pad.  SpaceX has unreleased video of one of its rockets doing something similar.

That said, I'm not yet convinced that this was a propulsion failure.  I see six engines burning on a rocket that is suffering some type of control problem.  It is, however, impossible to say for sure based only on the available video.
 - Ed Kyle

With regards to Falcon 9, did not Musk claim the Falcon 9 has sufficient thrust and gimbal margin to continue in-control flight even if it lost an engine at liftoff?

Anyways, after watching the Proton video a couple times, my suspicion of an engine failure has also waned substantially.

The main factor I previously focused on was the off-angle and apparently low velocity of the brown smoke, but knowing now that the brown exhaust is normal, on further review I believe the off angle is simply due to the extreme angle of attack of the rocket after pitching over.

Seems to me that roll alone should be easy to control given all the other engine pairs (heck, probably doesn't even need to be an *opposing* pair if you don't mind some pitch/yaw as well), but I guess I can see how single plane gimballing could go the other way - induce roll problems while controlling yaw/pitch when one engine goes wild.

Could what we're seeing here be the price you pay for such a single plane gimbal freedom system?

With all this talk about failed TVC, I am surprised that no one has discussed the unusual gimbal arrangement for the Proton first stage. Each engine has a single degree of freedom gimbal, meaning that each engine can only provide control along one angle. 

If one engine were to gimbal as far as possible and stay that way, I don't know how the remaining five engines could compensate; this leads me to believe that the motion control system normally imposes significant constraints on each engine's movements.

Based on these posts (thanks!) and the vid, my guess is that there is an override to the thrust termination constraint until 45 seconds and one thrust gimbal actuator failed in an adverse position.

Here is why i am guessing there is an override:
- gimbal fails in first few seconds of launch
- 7.5 degree thrust angle arm per engine (pictures here)
- brown smoke occurs when pitch is greater than 7.5 degrees
- "brown smoke" corrects the pitch angle, but not the pitch rate, and the roll rate is reduced   

A look at the video: http://rt.com/news/proton-m-rocket-takeoff-crash-514/
(vid=8s) limits reached (past +7.5 degrees )
(Vid=9s)  brown smoke  (engine throttling) begins, pitch correction begins, override kicks in, pitch introduces roll rate
(Vid=11s) orientation is ~ vertical, but initial smoke location has rolled 180 degrees, pitch rate is still present, roll rate is reduced
(Vid=12s)  pitch is exact opposite direction (-7.5 degrees), but brown smoke shuts off!  roll rate begins again (if the rocket had been kept throttled down it would have helped correct the orientation because of the 180 degree roll-if this make sense)
(vid=13s) brown smoke disappears, pitch angle continues to increase, roll is present (why did 'another' engine not throttle down? too much roll?)
(vid=17s) another brown plume, but rocket has a significant roll rate
              so one would have to be "cycling" the TVC gimbals and/or engine throttle to correct orientation "up"   roll is not good
(vid=18s)  rocket is horizontal..... well past point of compensation...

[edkyle99]

I respectfully disagree.
IMHO, Proton retirement is long overdue, although its reliability record is not too bad. The call for retirement comes from two other facts.
1. In normal case - i.e., each and every time - Proton drops the first stage with some hundreds kg of unburned UDMH. Therefore, there must be crush site decontamination after each launch. Unfortunately, there are numerous examples of improper/ineffective crush site treatment.
2. In less-then-nominal case - Proton WILL always create a disaster site -- along with the crater. 200 tons of UDMH will always by a BIG problem, if they did not burn nominally.

UDMH soil and groundwater contamination is an issue, no doubt.  On the other hand, hundreds of tonnes of kerosene leached into the soil would also require cleanup.  The cost, if properly done, would still be a sizable percentage of the cost of a hydrazine spill cleanup. 

Angara, as I understand it, will use multiple boosters, which means multiple drop (and cleanup) sites compared to Proton, so the comparison is not straightforward.  Even Vostochny, as I understand it, will need drop zones on land.

 - Ed Kyle

[Kabloona]

As for the brown smoke, do a Google translate on the web page and read what he says in the "Stage 1" section.

http://www.bernd-leitenberger.de/proton.shtml

He seems to say that the brown smoke is normal on every launch and is a result of an MR change by temporarily throttling UDMH flow to one engine, resulting in reduced thrust, possibly for a crude method of pitch-over, though I don't know why one would do that instead of using TVC.

In any case, the brown smoke is apparently an artifact of normal engine throttling, ie unburned NTO when the UDMH flow rate is reduced.

Edit: And if the engines do indeed vector tangentially, then an engine stuck hard-over would explain the seemingly increasing roll rate, as well as the pitch instability.

[Lars_J]

User Artyom. posted this video in the live update thread:


I'm 100% speculating here, but based on that video, there appears to be no problem with an engine stuck at a maximum gimbal position (at least from THAT angle) - the whole rocket seems very nimble. To my eyes this looks ore like a software/flight control issue - the vehicle attempts to WILDY overcompensate almost immediately (and then overcompensates in the other direction) and finally loses control.

Again, this is my 100% amateur speculation.

[kevin-rf]

I'm 100% speculating here, but based on that video, there appears to be no problem with an engine stuck at a maximum gimbal position (at least from THAT angle) - the whole rocket seems very nimble. To my eyes this looks ore like a software/flight control issue - the vehicle attempts to WILDY overcompensate almost immediately (and then overcompensates in the other direction) and finally loses control.

Again, this is my 100% amateur speculation.

Other than you can clearly see by the plume, that one of the engines is not moving in sync with the others... 100% amateur here too.

[Danderman]

I should followup on my comment that a 45 second delay in the FTS could lead to a rocket landing in the city of Baikonur.

Any vehicle that flies from the launch pads to the city would have to be traveling at a 180 degree launch azimuth, and very few vehicles are sent into that orbit from Baikonuir.

For a Proton to hit the city of Baikonour within 45 seconds, it would have to be aimed there, or things would have to go very wrong very quickly.

The failed Dnepr launch, which did go along a 180 degree azimuth, ended up some 170 km downrange, with the abort occurring at the 73 second mark. The stack landed in another country south of Kazakhstan.

Anyone claiming a vehicle cannot reach the city of Baikonur within the 45 second FTS delay is forgetting that the vehicle continues flying after the engine thrust is terminated.

[Lars_J]

I'm 100% speculating here, but based on that video, there appears to be no problem with an engine stuck at a maximum gimbal position (at least from THAT angle) - the whole rocket seems very nimble. To my eyes this looks ore like a software/flight control issue - the vehicle attempts to WILDY overcompensate almost immediately (and then overcompensates in the other direction) and finally loses control.

Again, this is my 100% amateur speculation.

Other than you can clearly see by the plume, that one of the engines is not moving in sync with the others... 100% amateur here too.

No, I *think* that is caused by the 1-D/tangential gimbal capability of the engines. The engines are arranged in a circle, and can only gimbal towards their neighbors. This explains why there appears to be a lone engine not gimballing when the other engines appear to be. You can see that on each side when the rocket attempts to veer left and right.

[LegendCJS]

I'm 100% speculating here, but based on that video, there appears to be no problem with an engine stuck at a maximum gimbal position (at least from THAT angle) - the whole rocket seems very nimble. To my eyes this looks ore like a software/flight control issue - the vehicle attempts to WILDY overcompensate almost immediately (and then overcompensates in the other direction) and finally loses control.

Again, this is my 100% amateur speculation.

Other than you can clearly see by the plume, that one of the engines is not moving in sync with the others... 100% amateur here too.

No, I *think* that is caused by the 1-D/tangential gimbal capability of the engines. The engines are arranged in a circle, and can only gimbal towards their neighbors. This explains why there appears to be a lone engine not gimballing when the other engines appear to be. You can see that on each side when the rocket attempts to veer left and right.

With the tangential gimbaling arrangement every engine should have equal  authority to effect roll rate, so one engine hard over and the rest healthy  shouldn't/ couldn't have led to the run-away roll increase we saw.

Unless, of course, the loss of gimbaling ability on one engine creates a different form of cross coupling between attitude control and roll control actuation than the control scheme is designed to handle.

A control authority and cross coupling analysis between roll and attitude actuation based on proton engine layout in both the healthy and one engine frozen configurations would be informative, and I leave it as an exercise for the first aerospace engineering student who reads this post.

[Stan Black]

http://88.210.62.157/content/numbers/258/02.shtml

An old N.K. article about using solids on Proton, that addresses issues about saving the launch site during an emergency.

[Danderman]

I checked with Glushko's encyclopedia, and the RD-253 engines indeed gimbal in a lateral direction.

[belegor]

"According to a source in the Russian space industry work at the Baikonur space center will probably be suspended for the next two or three months because of contamination."

http://en.ria.ru/russia/20130702/182002715/Russian-Proton-M-Rocket-Falls-Shortly-After-Launch.html

I only stumbled upon the bolded part during my second reading (I read "launches" first), so do I understand it correctly, that this would also include work on payload preparation, etc?

[Star One]

I'm 100% speculating here, but based on that video, there appears to be no problem with an engine stuck at a maximum gimbal position (at least from THAT angle) - the whole rocket seems very nimble. To my eyes this looks ore like a software/flight control issue - the vehicle attempts to WILDY overcompensate almost immediately (and then overcompensates in the other direction) and finally loses control.

Again, this is my 100% amateur speculation.

To put it very, very crudely it looks like a drunk desperately trying to keep his balance, as if its 'senses' were wildly off kilter.

If it is software issue how would you end up with it producing that kind of result in a launcher?

[edkyle99]

The failed Proton was the 13th orbital launch attempt from Baikonur this year.

That's just a coincidence of course. 

It was also a coincidence that last year's 13th orbital launch attempt, by a Proton M/Briz M, also failed.  That one was a Briz M failure in orbit.

The 13th launch the year before that (2011) was a completely successful Proton flight, and the year before that (2010) a completely successful Soyuz U/Progress launch, so, clearly, that number 13 thing is just a coincidence. 

 - Ed Kyle

[simonbp]

Another way to look at it is that there has been one Proton-M failure per year for the past four years...

[baldusi]

Another way to look at it is that there has been one Proton-M failure per year for the past four years...

The historical rate of failure of Proton has been 88.40%, or one in 8.77 launches. And the average launch rate from 2000 to 2012 has been 8.62. I don't think this is a coincidence. As I said before, you know the reliability of a Proton. Consistency is a good characteristic. High reliability is better, of course, but since Proton is consistent in its failures, the project that fit it take that into account and have some sort of redundancy or insurance.
The Russian government has been very unlucky, though, having been bitten twice in a three Glonass launch in a row. And they had decided that they wouldn't use insurance since they already could play the numbers.

[Kabloona]

For comparison, video below shows nominal Proton launch of SES-6. A persistent brownish exhaust is visible, especially around 1 minute into the video, then seems to slowly disappear. This is presumably due to the mixture ratio adjustment described in the German website I cited upthread. But clearly it's quite different from the intermittent brown plumes seen during the failed launch.



Now there's a report that a senior Russian source is saying that the brown plume may have been the result of an NTO leak. But the same source also says the engines were shut down at 17 seconds, which is evidently incorrect.

http://rbth.ru/news/2013/07/02/russian_rocket_crash_could_have_been_caused_by_oxidizer_pipeline_perfora_27701.html

Edit: After re-reading the quote, the source says "after the engines emergency shutoff, when the rocket began to fall..." there was brown smoke...but he's talking about 17+ seconds into the flight when it was clearly a goner...obviously the root cause happened sooner...maybe he's speculating that the brown smoke indicated a possible NTO leak that began just after liftoff and gradually grew more severe...

[Gorizont]

A bad day for my (beloved) Proton-rocket.

greetings...
Soeren

[input~2]

AFAICT, the 1st stage engines' plumes where already very different in length just 3 seconds after lift-off.
It seems that at least one engine had problems at this stage