World Missile and Space Launch Vehicle Failures

[Art LeBrun]

This topic is to discuss and illustrate rocket/missile/SLV mishaps, anomalies and failures over the years.

Feedback/corrections on topic title and subjects within the topic encouraged.

Now open for postings.

[R7]

Would Wan Hu Flying Chair qualify as first reported mishap?

[Archibald]

As promised I've dug into Ariane V36 failure. And I've found some elements in that story that are even more disturbing than the cloth itself.
Looks like the Murphy law had a pretty good day.
Suffice to say that the two Japanese satellites lost this day very VERY unlucky.
Superbird B had been delayed because the Loma Prieta earthquake of 1989 had damaged Loral facilities in Palo Alto.
The delay resulted in the satellite booking on V36.
In the end the satellite survived an earthquake only to be lost on a rocket failure. 
As for the other satellite (http://space.skyrocket.de/doc_sdat/bs-2x.htm)  - its twin and backup was lost a year later in an Atlas failure.

More on this later.

[ugordan]

http://space.skyrocket.de/doc_sdat/incorrect-2x.htm

Don't tell me the forum's Auto-Incorrect feature (TM) attacks even URLs?

Can you repost that link, just drop one of the empty formatting tags into the offending word?

[Archibald]

http://space.skyrocket.de/doc_sdat/incorrect-2x.htm

Don't tell me the forum's Auto-Incorrect feature (TM) attacks even URLs?

Can you repost that link, just drop one of the empty formatting tags into the offending word?

    
I don't know whether to laugh or bang my head against the nearest wall. Oh, well...
Superbird B entry
http://space.skyrocket.de/doc_sdat/superbird-a.htm
(the companion satellite is listed along it)

[AnalogMan]

http://space.skyrocket.de/doc_sdat/incorrect-2x.htm

Don't tell me the forum's Auto-Incorrect feature (TM) attacks even URLs?

Can you repost that link, just drop one of the empty formatting tags into the offending word?

main part of url should be

space.skyrocket.de/doc_sdat/bs-2x.htm

[Archibald]

As promised, the story of Ariane flight 36.

(translated from the highly recommended Capcom espace website, whose webmaster is a member of this forum, by the way)

http://www.capcomespace.net/dossiers/espace_europeen/ariane/index.htm

On February 22 1990 the third Ariane 44L lifted-off from Kourou at 20h17 local with Japanese satellites Superbird and B--S-2X.
110 seconds after the rocket exploded in flight destroying its 3750 kg payload estimated at over one billion Francs (a few hundred of million euros).
The explosion took place while the eight Viking engines operated at full thrust. It was Ariane 36th launch since 1979 and the 5th failure.
The alert was given to Kourou to prevent a possible fallout of toxic gases. Soon French legionaries were to the waist in the mud of the  mangroves to find the pieces.
Motor -D was found some days later. Analysis shows a loss of thrust in at least one engine caused the launcher to slowly veer off the nominal trajectory until aerodynamic forces led to a breakup.
According to preliminary data, thrust of first stage and boosters (PALs) was nominal until   T+ 6.2 seconds.
At this point, D engine pressure droped from 58 to 30 bar until T+110 seconds.  A fire broke out on the defective PAL.
Ariane control system tried to correct the issue by steering the other engines. From T+90 seconds  however the engines were pointed to their maximum of four degree - and the deviation continued.
At T+110 seconds loads exceeded Ariane strength leading to the launcher breaking up at a height of 9000 meters, 12.5 km away from the launch pad.

ELA-2 launch complex went on undamaged despite the launcher passing too close from the top - 3 meters instead of the usual 7 meters, with black marks found on top of the structure.

This is a comparison of Ariane V36 (left) and V48 (right), both Ariane 44L.

V36 lack of thrust at takeoff has the launcher rising slowly - it takes one more second for V36 to clear the tower than V48.
On the picture the cross represents the actual position of V36;  the square represents the nominal position.
Almost from the start V36 was outside of the curve - and Flight Director M.  Gaillard accordingly looks concerned.

And now the priceless moment - why did Ariane failed...

Translated from this document (via google, with all the according caveats) 
http://liris.cnrs.fr/amille/enseignements/Master_PRO/BIA/chap10.htm

1989 - Les Mureaux, near Paris - Ariane assembly plant

During assembly of a Viking motor, a boilermaker can not connect two pipes as shown on the plan. Thus he gives a little bit of polishing. But to follow the procedure, he must warn his superior that he has done something that was not planned. Now it is Friday evening, there is nobody in sight.
Then, as the tradition of Boilermakers in this case, he disconnect the pipes and puts his handkerchief in the pipe.
He reasons that on monday, the red, flashy color of the cloth will catch his eyes and he will be reminded.
Alas !
During the weekend, he fell ill.
The colleague who replaces him Monday takes the two pipes; they adjusts easily, and voila.

The pipe goes into the Viking cooling assembly, the Viking gets bolted onto Ariane 4 stage 1 and the stage is ferried to Kourou.
During flight the cloth blocked the flow of water cooling the Viking; the Viking lost power, ruining the rocket trajectory from the beginning.

http://www.forum-conquete-spatiale.fr/t10688-retour-sur-l-echec-d-ariane-4-v36

http://www.ina.fr/video/CAB06056640/ja2-20h-emission-du-23-fevrier-1990.fr.html
(it come damn close of the launch tower...)

Was Arianespace lucky or unlucky on that flight ? it's anyone guess... others were VERY unlucky.
On this fateful flight Ariane 44L carried two Japanese satellites - Superbird B and B--S-2x.
Stage 1 of that launcher was initially planned to fly on V35 - an Ariane 40 to launch Spot 2 into polar orbit.
It happened that Superbird B was delayed by the 1989 Loma Prieta earthquake, which resulted in damages to contractor Loral in Palo Alto.
As a result Ariane flights and stage 1 were swapped.
V35 stage 1 went to V36 and in the process the rocket become an Ariane 44L configuration.
V36 inherited from a defective stage 1, since the -D engine featured the infamous cloth blocking the water coolant tube.
On January 21, 1990 V35 was perfectly nominal - Ariane 40 boosted Spot 2 into the right orbit. A month later V36 was not so lucky.
Had Ariane launched in version 40, the less powerful launcher probably would have collided with the umbilical mast, causing more damage. As a matter of fact only the cryogenic arm had to be changed after V36 failure.
In conclusion
- It can be said that the unfortunate Superbird B satellite escaped an earthquake only to be destroyed in a rocket failure.
- Yet Japan bad luck was far from over. The other satellite destroyed in the flight (B--S-2x) had a backup... that was launched a year later, on April 18, 1991. Unfortunately the satellite was lost when Atlas-Centaur AC-70 Centaur failed to ignite !
- Much like Proton and Titan Ariane carried toxic propellants.
On April 2 1969 Baikonur suffered a major mishap when a Proton exploded close from the ground. A similar disaster happened in Vandenberg AFB on April 18 1986 with Titan 34D-9 failure. Would Ariane 44L V36 have resulted in similar damage to ELA-2 ?
- since that day every single cloth used by Arianespace has a number on it. No cloth can go away.

For the record, most of Ariane failures (1982, 1985, 1986 and the twin1994 failures) can be traced back to the H-10 third stage.
The lowest composite failed only two times - the cloth, and flight test 2 in May 1980. The second Ariane ever was lost due to first stage pogo.
Ariane 5 was another matter.
Proof that man-rating is not exact science: Ariane 1 - 4 had a reliability very similar to the shuttle (2 failures in 144 flights, against 2 in 135 for the shuttle). The Viking is damn reliable, ask ISRO.
Yet it was Ariane 5 that was build to be man-rated, only to suffer a very rocky start (four failures in the first six years). Go figure...

[Davd]

- Much like Proton and Titan Ariane carried toxic propellants

FTR, at least one Proton failure (Soyuz 7K-L1 launch attempt September 1967) was caused this way, only the offender was a rubber plug instead of a cloth. One of the first stage engines shut down at liftoff and the booster crashed downrange (but the LES pulled the capsule to safety).

Pic unrelated: debris recovered from Atlas-Centaur AC-1 (dunno what it is, but judging by the large amount of wiring, an instrument unit?)

[kevin-rf]

Why do you insist on putting unrelated rockets (and parts of rockets) in your posts?

Also, I vote for a test stand accident thread!

[kevin-rf]

At this point, D engine pressure droped from 58 to 30 bar until T+110 seconds.  A fire broke out on the defective PAL.
Ariane control system tried to correct the issue by steering the other engines. From T+90 seconds  however the engines were pointed to their maximum of four degree - and the deviation continued.
At T+110 seconds loads exceeded Ariane strength leading to the launcher breaking up at a height of 9000 meters, 12.5 km away from the launch pad.

So in other words, Ariane engine out capability failed...

[Jim]

Pic unrelated: debris recovered from Atlas-Centaur AC-1 (dunno what it is, but judging by the large amount of wiring, an instrument unit?)

Centaur did not/does not have an instrument unit, which by definition is a separate unit.  That would be the Centaur equipment module.

[Art LeBrun]

Pic unrelated: debris recovered from Atlas-Centaur AC-1 (dunno what it is, but judging by the large amount of wiring, an instrument unit?)

Centaur did not/does not have an instrument unit, which by definition is a separate unit.  That would be the Centaur equipment module.

Once enclosed perhaps equipment bay? Just looking for a good proper term - aerospace or manufacturer.

[Jim]

Pic unrelated: debris recovered from Atlas-Centaur AC-1 (dunno what it is, but judging by the large amount of wiring, an instrument unit?)

Centaur did not/does not have an instrument unit, which by definition is a separate unit.  That would be the Centaur equipment module.

Once enclosed perhaps equipment bay? Just looking for a good proper term - aerospace or manufacturer.

I meant the object in both photos is the Centaur equipment module.

I have to reread my posts from an outside POV and not what I think it says.

[Davd]

If you watch the video of AC-1, you notice that the initial Centaur breakup produces a water vapor cloud (the product of the LH2/LOX mixing) but nothing like the huge expanding deflagration in AC-5. I would guess because the airstream and speed would prevent that from happening.

[Art LeBrun]

If you watch the video of AC-1, you notice that the initial Centaur breakup produces a water vapor cloud (the product of the LH2/LOX mixing) but nothing like the huge expanding deflagration in AC-5. I would guess because the airstream and speed would prevent that from happening.

My take on that large white cloud was the Atlas lox tank opening up followed by the RP-1 tank and the fireball. If you look closely you will see one of the two liquid nitrogen lines was carried aloft and can be seen flapping against the boattail and in the airstream.

On AC-5 the Centaur would have dropped almost instantly into the buckling Atlas so the mixing was much faster.

[Davd]

My take on that large white cloud was the Atlas lox tank opening up followed by the RP-1 tank and the fireball.

In most rocket failures the detonation is caused by propellant touching hot engine parts, although RP-1 and LOX do form a highly explosive gel when mixed.

Interestingly if you see the video of Atlas 3B, there's little or no explosion there, just a smoke cloud (the LOX?) after the vehicle breaks up. The thrust section survives nearly intact and the camera tracks it falling to earth with a little bit of flame coming out. Likely because of the angle it was flying in, the propellants simply dispersed and didn't ignite from contact with the engine.

If you look closely you will see one of the two liquid nitrogen lines was carried aloft and can be seen flapping against the boattail and in the airstream.

That was presumably ripped off the launch structure during liftoff?

On AC-5 the Centaur would have dropped almost instantly into the buckling Atlas so the mixing was much faster.

This one is a testimony to the enormous power of LH2/LOX. Most of the blast is a giant yellow-white fireball and you can hardly see any of the orange-red of the RP-1 burning after the first few seconds. That's all considering the Atlas was twice the size and propellant load of the Centaur.

[Jim]

In most rocket failures the detonation is caused by propellant touching hot engine parts, although RP-1 and LOX do form a highly explosive gel when .

No, from the hot gases coming out of the engine or destruct charge.

[Jim]

If you watch the video of AC-1, you notice that the initial Centaur breakup produces a water vapor cloud (the product of the LH2/LOX mixing) but nothing like the huge expanding deflagration in AC-5. I would guess because the airstream and speed would prevent that from happening.

No, the  cloud is the LH2 condensing water vapor in the air.  The same effect would happen if there was no LOX onboard. LH2 tank ruptured and LH2 spilt down the sides of the vehicle.

NASA SP-4230

[Jim]

This one is a testimony to the enormous power of LH2/LOX. Most of the blast is a giant yellow-white fireball and you can hardly see any of the orange-red of the RP-1 burning after the first few seconds. That's all considering the Atlas was twice the size and propellant load of the Centaur.

The bulk of the  blast is from the RP-1 and not the LH2.   The RP-1 is consumed after a few seconds.  I am looking for the accident report that I have.

[Art LeBrun]

"That was presumably ripped off the launch structure during liftoff?"

It does appear ripped away or broken above the ground connection. The same event happened on 117D (Ranger 2).

[R7]

In most rocket failures the detonation is caused by propellant touching hot engine parts, although RP-1 and LOX do form a highly explosive gel when mixed.

In most rocket failures there's hardly any detonations at all, except for the FTS explosive charges. Propellant detonation requires them ending up premixed at ignitable ratio and an energy spike (spark) or deflagration-to-detonation transition (which is easier said than done, just ask PDE-developers).

[Davd]

No, the  cloud is the LH2 condensing water vapor in the air.  The same effect would happen if there was no LOX onboard.

I know that; the hydrogen would still form water vapor from coming into contact with the oxygen in the surrounding atmosphere.

LH2 tank ruptured and LH2 spill down the sides of the vehicle.

Which is pretty obvious. Somehow I was a little bit doubtful of Art's claim that that was the Atlas's LOX spilling.

Also during AC-1's ascent, there's a visible trail of LOX streaming from the Centaur. Not sure if this was intentional or not, but it does remind one of Atlas 3D where the drain valves failed to close at liftoff.

in most rocket failures there's hardly any detonations at all, except for the FTS explosive charges

I did mean deflagration, that was just a poor choice of words. Pay it no heed. Deflagration is what happens if you threw a lit match into a paper cup filled with gasoline. Detonation is what happens inside a car engine (after the gas and air are mixed and ignited).

[Jim]

I know that; the hydrogen would still form water vapor from coming into contact with the oxygen in the surrounding atmosphere.

No, the visible vapor isn't from the combustion of the H2.  It is from the cold of the LH2 condensing atmospheric water into the white visible vapor.  It happens with any cryogenic fluid released into ambient air.

[R7]

Detonation is what happens inside a car engine (after the gas and air are mixed and ignited).

No. (Sorry  Detonation is not supposed to happen in car engines, things are tuned/timed to avoid that. Some times things go wrong and it can happen, it's called the engine knocking, and it destroys the engine quickly.

Please take no offense, it's just me being the pedantic wiserear  Detonation and deflagration are just so well defined that easy to keep them apart. Muddy waters start when using words like "explosion" where definitions become less scientific and more opinion-based. Then again whether things deflagrate rapidly or detonate when they aren't supposed to do either it's usually a bad day anyway...

[Art LeBrun]

No, the  cloud is the LH2 condensing water vapor in the air.  The same effect would happen if there was no LOX onboard.

I know that; the hydrogen would still form water vapor from coming into contact with the oxygen in the surrounding atmosphere.

LH2 tank ruptured and LH2 spill down the sides of the vehicle.

Also during AC-1's ascent, there's a visible trail of LOX streaming from the Centaur. Not sure if this was intentional or not, but it does remind one of Atlas 3D where the drain valves failed to close at liftoff.

I believe the temporary venting was hydrogen. It stopped about 8 seconds prior to the weathershield failure.

[Davd]

I believe the temporary venting was hydrogen. It stopped about 8 seconds prior to the weathershield failure.

Ok makes sense since it's coming from the top of the Centaur where the hydrogen tank is. I now just remembered that this same effect can be seen on Saturn launches.

But enough about AC-1 (at least until Jim posts the accident report he claims to have). Moving onto the next subject in this thread, Atlas 3D which had a nothing short of spectacular failure yet no accident reports have turned up.

Oddly, the official Atlas launch summary lists seem to imply that 3D failed because of the LOX valve not closing at liftoff instead of the (extremely obvious) launcher arm hangup at liftoff.

[Art LeBrun]

Here is a GD/A report I received recently on 3D and 7D. Not scanned from a GD/A document but I assume the wording is accurate.

It seems the B-2 thrust decay accounted for the "lurching" of 3D and not a launcher arm failure. I understand it looks abrupt like a sudden release but this is what I have for the moment. There is a dark stream in the B-2 exhaust which I do not understand. Obviously you need to discount the turbine exhaust just to the outside of B-2.

[Davd]

It seems the B-2 thrust decay accounted for the "lurching" of 3D and not a launcher arm failure.

It would seem that the engine already had some sort of problem that was exacerbated by the LOX spilling and causing an explosion.

There is a dark stream in the B-2 exhaust which I do not understand. Obviously you need to discount the turbine exhaust just to the outside of B-2.

There are a number of things that might cause that. Maybe the fuel/oxidizer mix in the engine was off from its normal value. Also you can seen from the video that there's a fire building up in the thrust section during ascent prior to the explosion @ T+28 seconds. That could have been the result of the liftoff damage or due to whatever problem caused the initial B2 malfunction.

As for 7D, on rewatching the video on Critical Past closely, I see it now. The launcher arm doesn't retract fully at liftoff.

[Art LeBrun]

Atlas Incidents in Images

Image 1 shows Atlas 6D in flight from VAFB January 26, 1960. Note that the two nacelle doors have not closed. No word of effect on flight.

Image 2 shows Atlas-Agena carrying first SAMOS recon satellite aloft on October 11, 1960. Note one Atlas nacelle door is only partially closed. The notable part of the image is the escaping Agena stage attitude control gas from the left side of the interstage. The happened during an umbilical disconnect which tore off part of the plate on the vehicle. Without the attitude control gas the Agena burn was unsuccessful in achieving orbit. There are multiple launch images showing the gas loss.

[Davd]

Atlas Incidents in Images

Image 1 shows Atlas 6D in flight from VAFB January 26, 1960. Note that the two nacelle doors have not closed. No word of effect on flight.

Image 2 shows Atlas-Agena carrying first SAMOS recon satellite aloft on October 11, 1960. Note one Atlas nacelle door is only partially closed. The notable part of the image is the escaping Agena stage attitude control gas from the left side of the interstage. The happened during an umbilical disconnect which tore off part of the plate on the vehicle. Without the attitude control gas the Agena burn was unsuccessful in achieving orbit. There are multiple launch images showing the gas loss.

As 6D and 57D both performed normally during flight, it does not appear that the nacelle doors affected anything.

Also Atlas 131D (Midas 6's booster) suffered liftoff damage similar to 7D however instead of causing pressurization gas to escape, the hydraulic fluid was lost leading to eventual loss of control and RSO destruct.

[kevin-rf]

Nice article on the space review concerning some failures, with root causes.

Launch failures: what’s changed?
http://www.thespacereview.com/article/2256/1

[Art LeBrun]

Nice article on the space review concerning some failures, with root causes.

Launch failures: what’s changed?
http://www.thespacereview.com/article/2256/1

Usually Wayne Eleazer gives dates and missiles with his failure reports but the three Atlas Ds have no such reference. My limited records indicate no such episodes of 3 Atlas D exploding on Cape pads much less the same pad. However, I am intrigued by the separate turbine exhaust duct so I will research that and see if it is visible in photos.

[libra]

Bringing back this oldie... my own little contribution to that topic, now published at The space review.

https://www.thespacereview.com/article/4085/1

Ariane, earthquake, cloth, and the Murphy law. A monument of absurdness - one can only laugh or just shake his head in disbelief.  Such bad luck, for everybody involved.

Yet, it could have been worse

(Samuel L. Jackson memorable -and very unfortunate - quote from Jurassic Park)

[Vahe231991]

I found this video from Youtube showing the first launch of the CZ-3B in February 1996, which ended in failure.