FAILURE: SpaceX Falcon I Launch III - August 2

[dirkthefirst]

A significant investment as a precautionary measure ... now Elon will have to account to investors.  I get the feeling that the Falcon 1 will now just be a testbed for Falcon 9.  Maybe thats the way it has to be to make COTS a reality. Kind of an interesting statement.

SpaceX is invested almost entirely by Elon himself, not venture capitalists. For the most part, not entirely, he has to answer to himself.

This is no longer the case.

As a precautionary measure to guard against the possibility of flight 3 not reaching orbit, SpaceX recently accepted a significant investment.

Someone invested in SpaceX. They will now have to answer to whoever that person/company/organisation is.

[landofgrey]

yes it will

For CCAFS, the requirement for active destruct depends on the vehicle. for Falcon 9, yes (as of what I least heard) it does require active FTS, i.e. command destruct.

A ballistic sounding rocket would not.

[landofgrey]

This second-guessing is for amateurs.

After over 20 years in this business I've learned two things. 1) Never call someone an amatuer; and 2) we ALL started out as amatuers. I'll never be condescending and arrogant and call someone an amateur as a slight when I know full well that that person could be the next von Braun or Goddard. Neither I, you nor anyone else is at the top of the pyramid, so we all should keep our egos, and attitude, in check.

And as soon as you think you're not an amateur, and start acting like you aren't.... then you are one of them. And von Braun was always a "rocket kid" at heart... amateur and all.

[landofgrey]

Look up how many of the first Atlas or Thor missiles failed before a successful launch. It may have been 50 years ago, but physics, rocket science and basic engineering and manufacturing hasn't changed. People should remember that.

Go SpaceX

[dirkthefirst]

Look up how many of the first Atlas or Thor missiles failed before a successful launch. It may have been 50 years ago, but physics, rocket science and basic engineering and manufacturing hasn't changed. People should remember that.

One thing that has changed, however, is the amount of experience we now have. 50-60 years of launching rockets is a long time, and we've researched a great many failure modes in that half century.
It seems to me that on occasion SpaceX haven't fully examined what has gone before, and that some of their problems could've been avoided had they heeded warnings from other launch vehicles.

(the tank issues on Flight 2 being a prime example)

[Zachstar]

Look up how many of the first Atlas or Thor missiles failed before a successful launch. It may have been 50 years ago, but physics, rocket science and basic engineering and manufacturing hasn't changed. People should remember that.

One thing that has changed, however, is the amount of experience we now have. 50-60 years of launching rockets is a long time, and we've researched a great many failure modes in that half century.
It seems to me that on occasion SpaceX haven't fully examined what has gone before, and that some of their problems could've been avoided had they heeded warnings from other launch vehicles.

(the tank issues on Flight 2 being a prime example)

Great point!!

[ShuttleDiscovery]

Not again! Just read the news..

Hope the next one goes better than the last 3!

[sdsds]

Look up how many of the first Atlas or Thor missiles failed before a successful launch. It may have been 50 years ago, but physics, rocket science and basic engineering and manufacturing hasn't changed.

One thing that has changed, however, is the amount of experience we now have.

Another change is the availability of high fidelity computational simulations. SpaceX uses these techniques extensively. The lessons we are all learning by watching SpaceX, and for which we should be truly obliged to them, are about how even the best available computational models aren't the same as flight experience.

We should all be heartily congratulating SpaceX for having what it takes to collect real-world flight data that validates or disproves the models. They have now done that again, and next Elon says they will move forward with full vigor. More power to 'em!

[CentEur]

Elon's sent an e-mail:

On the plus side, the flight of our first stage, with the new Merlin 1C engine that will be used in Falcon 9, was picture perfect.

The problem I have with this statement is that I've seen both live and replayed that the first stage stubbornly tried to roll just to be corrected several times. It was unusual, and did not occur in Flight 2. Thus Elon assessing yesterday's first stage flight as "picture perfect" reminds me one of Woody Allen's characters asking "Who do you believe? Me or your lying eyes?" 

[clongton]

Elon's sent an e-mail:

On the plus side, the flight of our first stage, with the new Merlin 1C engine that will be used in Falcon 9, was picture perfect.

The problem I have with this statement is that I've seen both live and replayed that the first stage stubbornly tried to roll just to be corrected several times. It was unusual, and did not occur in Flight 2. Thus Elon assessing yesterday's first stage flight as "picture perfect" reminds me one of Woody Allen's characters asking "Who do you believe? Me or your lying eyes?" 

It did not occur on F2 because the Merlin engine was an ablative nozzle. F3 used the new Merlin Regen nozzle which introduces some roll momentum to the engine. The movements you saw were the TVC successfully  countering, as programed. This is a known phenomena on Regen nozzles. Based on Elon's statements, the engine performed exactly as predicted, nozzle-induced "roll" and all.

There is no problem here. Don't look for hidden problems that don't exist. He has already told us what the main problem was; a 1st/2nd stage separation event. Once the details of what caused that problem become available, he will tell us those as well.

[jimvela]

Elon's sent an e-mail:

On the plus side, the flight of our first stage, with the new Merlin 1C engine that will be used in Falcon 9, was picture perfect.

The problem I have with this statement is that I've seen both live and replayed that the first stage stubbornly tried to roll just to be corrected several times. It was unusual, and did not occur in Flight 2.

This flight used the new Merlin 1C, with a regeneratively cooled chamber.  That chamber has a spiral winding.  I wouldn't be at all surprised if that induces a tendancy to roll. 

[Namechange User]

Great, tonight Alt.Space died and with it all hope of making the space program viable in the long term.

No it didn't.  If it did, then it was never meant to be then in the first place if it was all based on everything working perfectly in the very short term. 

Reading through this thread everyone seems to be under the impression that launching rockets is easy and a new company like SpaceX should be on the moon by now. 

Well, it's not easy and there will be failures.  Everyone has to expect that.

[Jim]

There is no problem here. Don't look for hidden problems that don't exist. He has already told us what the main problem was; a 1st/2nd stage separation event. Once the details of what caused that problem become available, he will tell us those as well.

There is an "issue"  The oscillation is not normal.  There is a mode that the vehicle control system is having "trouble" with ala Delta-III

[Jim]

Reading through this thread everyone seems to be under the impression that launching rockets is easy and a new company like SpaceX should be on the moon by now. 

The issue is that nuspace thinks Spacex can.

[pippin]

Great, tonight Alt.Space died and with it all hope of making the space program viable in the long term.

No it didn't.  If it did, then it was never meant to be then in the first place if it was all based on everything working perfectly in the very short term. 

Reading through this thread everyone seems to be under the impression that launching rockets is easy and a new company like SpaceX should be on the moon by now. 

Well, it's not easy and there will be failures.  Everyone has to expect that.

A big part that was the credo of alt.space indeed did:
That if you enter the stage with an entrepreneurial approach and a management style derived from the new economy you can do everything "cheaper, faster, more reliable".
That was SpaceX original claim. It may have died already earlier but tonight's failure was the last nail in the coffin.

The other fundamental of alt.space, which is that there will be a transition from public to private funding for space transportation and maybe space flight has not died.
That one stems from political changes, a more mature market and the increasing reluctance from the public to fund spaceflight through taxes. But it is about more than companies like SpaceX, that one includes old.space as well as the rest of the world. Let's see whether Orbital's more experienced guys will in the end be the ones to launch a "privately funded liquid fueled rocket" into orbit...

[Analyst]

1) It did not occur on F2 because the Merlin engine was an ablative nozzle. F3 used the new Merlin Regen nozzle which introduces some roll momentum to the engine. The movements you saw were the TVC successfully  countering, as programed. This is a known phenomena on Regen nozzles. Based on Elon's statements, the engine performed exactly as predicted, nozzle-induced "roll" and all.

2) There is no problem here. Don't look for hidden problems that don't exist. He has already told us what the main problem was; a 1st/2nd stage separation event. Once the details of what caused that problem become available, he will tell us those as well.

1) Why are this movements not seen with other vehicles using Regen nozzles?
2) Are you sure of this? TVC doing its work is one thing, but your flight does not get smother. You may lose performace too. At least some work at the control algorithm is needed.

Analyst

[clongton]

1) It did not occur on F2 because the Merlin engine was an ablative nozzle. F3 used the new Merlin Regen nozzle which introduces some roll momentum to the engine. The movements you saw were the TVC successfully  countering, as programed. This is a known phenomena on Regen nozzles. Based on Elon's statements, the engine performed exactly as predicted, nozzle-induced "roll" and all.

2) There is no problem here. Don't look for hidden problems that don't exist. He has already told us what the main problem was; a 1st/2nd stage separation event. Once the details of what caused that problem become available, he will tell us those as well.

1) Why are this movements not seen with other vehicles using Regen nozzles?
2) Are you sure of this? TVC doing its work is one thing, but your flight does not get smother. You may lose performance too. At least some work at the control algorithm is needed.

Analyst

I didn't say that it was perfect - it was not. But the roll inducement was anticipated, and the TVC compensated. I would agree that more work is needed on the control algorithm to smooth it out, but that wasn't what caused the LOM. It was a 1-2 stage separation issue, details forthcoming.

[pippin]

but that wasn't what caused the LOM. It was a 1-2 stage separation issue, details forthcoming.

How do you know? I can imagine a few scenarios where too much roll induces a separation problem. Flight 2 was already close on this and it did NOT have the roll before engine shutdown. We don't know how the vehicle performed prior to sep.

[Lee Jay]

It did not occur on F2 because the Merlin engine was an ablative nozzle. F3 used the new Merlin Regen nozzle which introduces some roll momentum to the engine. The movements you saw were the TVC successfully  countering, as programed. This is a known phenomena on Regen nozzles. Based on Elon's statements, the engine performed exactly as predicted, nozzle-induced "roll" and all.

There is no problem here. Don't look for hidden problems that don't exist. He has already told us what the main problem was; a 1st/2nd stage separation event. Once the details of what caused that problem become available, he will tell us those as well.

Come on Chuck!  The Shuttle has three of them, and when a 5 degree roll was imparted by the SRBs during the last flight, and quickly and smoothly corrected by those regen engines, an investigation was still launched into the cause!  This stage didn't start to roll back and forth until it went supersonic, so I can't believe that roll wiggle was normal and expected.  Just because it didn't go unstable like the second stage in flight 2 doesn't mean it's acceptable.

[just-nick]

I think that the roll issue is both interesting and fair game.  It is something different and therefore, for us armchair steely-eyed-rocket-men, one of the few things we have to go on for our back of the envelope postmortems.

My thoughts: 

(1) if the nozzle was inducing some roll I'd imagine that it would be relatively steady and that the system would therefore eventually reach some sort of equilibrium where it was applying the right amount of roll force to counter the nozzle-induced roll.  It didn't, so the loop wasn't closed as well as all that.  I'm no control systems expert, but I do know that tuning all of the various constants and gains in a system can be a bear -- even in this age of simulation.  If I was flying in an airplane with an autopilot that compensated for some sort of semi-steady state force (like prop torque) that roughly my stomach wouldn't feel that good...  Cause of the loss or totally unrelated issue, I couldn't say at all.

(2) What was the ultimate cause of flight 2 again?  Wasn't it recontact that started some oscillations that the system couldn't damp out?

That's all.  My wife and I had two very good friends over for dinner last night who were understanding enough to pause the meal and come watch the launch with me.  We then drank a lot.

  --Nick