FAILURE: SpaceX Falcon I Launch III - August 2

[Antares]

Vacuum engine test facilities are more limited by their ability to maintain a vacuum with hundreds of pounds of gas per second flowing into their chamber.  Total thrust can be a concern, but is usually not the limiter.

You could probably get a lot of information with a FOIA request for the trade studies that went into the selection of building A-3 at Stennis vs refurbishing Plum Brook vs one of the chambers at Arnold Engineering Development Center.

As for a complete list of everything that can go wrong with an engine, I'm sure the SSME FMEA would be about 99.99% of that.  Not sure if that's on L2 or if it's eye-tar.

[Comga]

This is *not* incorrect.  Read the thread.  The discussion was about the initial Falcon 5 (not Falcon 9) designs.  Aviation Week reported at the time (2003-4) that SpaceX had contemplated a future RL10-powered Falcon 5 upgrade.  Here's a link.   

http://www.aviationweek.com/aw/generic/story_generic.jsp?channel=awst&id=news/03294top.xml
 - Ed Kyle

I did read the thread and am casting absolutely no aspersions on your excellent summary of the Falcon 5/9 evolution.  However, A_M_Swallow was saying that the flight failure was due to changing between LOX/RP-1 and LOX/LH2, which was never done.  There was once talk of using RL-10's as you correctly pointed out, but this was never part of the hardware.

[A_M_Swallow]

This is *not* incorrect.  Read the thread.  The discussion was about the initial Falcon 5 (not Falcon 9) designs.  Aviation Week reported at the time (2003-4) that SpaceX had contemplated a future RL10-powered Falcon 5 upgrade.  Here's a link.   

http://www.aviationweek.com/aw/generic/story_generic.jsp?channel=awst&id=news/03294top.xml
 - Ed Kyle

I did read the thread and am casting absolutely no aspersions on your excellent summary of the Falcon 5/9 evolution.  However, A_M_Swallow was saying that the flight failure was due to changing between LOX/RP-1 and LOX/LH2, which was never done.  There was once talk of using RL-10's as you correctly pointed out, but this was never part of the hardware.

But was it part of the software?
The software is not the hardware.

When I admitted it was the upper stage requirements that changed rather than the first stage requirements I was accepting that is was probably a red herring.

[nobodyofconsequence]

whatever that means

Well, for a DOTCOM, that would be the guy responsible for product development and operations as opposed to general IT...
And there it WOULD include SI (as a responsibility)

A CTO is responsible for the representation of technology to the outside world, they are largely a "technology sales man". Engineering operations is typically led by a VP of engineering who had better be a damn good engineer, manager, and executive. In some firms, one reports to the other, but often they have nothing to do with each other.

Many times you don't have a CTO. You only have a COO - operations officer - if you have major manufacturing or other massive process function that requires tight execution of  stacks/silos/supply chains.
The whole "C" level is meant to split up and scale executive function when it outgrows the single CEO/president role.

Sometimes it is a mere corporate conceit.

[nobodyofconsequence]

Forgive me for asking, but if F1 first stage engine cutoff has a stutter, why not adaptively compensate by waiting for the trailing interruptions before staging, knowing that after the turbo runs dry you can only support combustion/diffusion for a limited period - then stage. E.g. no fixed interval. You may have to burn second stage a little longer digging into propellant reserves, but so what?

[Antares]

The shutdown impulse would be too unpredictable.  What data criterion would you use?  Pc?  Acceleration?  What if it didn't pulse?  No, it's best to do it on a timer.

[nobodyofconsequence]

1.The shutdown impulse would be too unpredictable. 

What data criterion would you use? 
2. Pc? 
3. Acceleration?
4. What if it didn't pulse? 

No, it's best to do it on a timer.

1. Really? There is a known amount of props when the turbo runs dry and stops (spin down time). So either it burns or diffuses, right? It's not just going to sit there.
3. If it burns, we get an impulse - the impulse size is prop consumption by known amounts (so we decrease worst case timer by consumption estimate)
2. If it doesn't, it's diffusion in a vacuum - exponential decay, lengthen  timer until drop off edge below combustion lower bound, then reset to safety pad.
4. What if the fixed timer length is always wrong? E.g. still don't wait long enough and have a later burp? Or too long and we coast too much to make it up with the second stage reserve?

Just intrigued by the issue. Have always been skeptical of "time outs", mostly because of the arbitrary choice made in setting them.

[Flometrics]

Based on what Elon said, the engine thrust taper off was longer than expected, and this is why the vehicle failed. If the first stage had been test launched with a dummy second stage, the nut failure and first stage/second stage contacts would have still happened but would have been much less expensive failures, and would have happened earlier. This path would have been safer, but not as sexy.

Keep in mind that SpaceX has still not actually competed against other launch service vendors. Their launches have been funded by the DOD, which has an interest in reducing launch costs, but can't spend too much on failure. Now SpaceX will have to launch 3 times successfully to even begin to compete. After that will the competitors cut their rates temporarily to force Spacex into bankruptcy?
The assumption that the competition will stand still is a foolish one, but it seems to be in every rocket business plan.
I want SpaceX to succeed, but I think they need to focus on getting to orbit with the Falcon 1, not developing new engines, space capsules and getting more government contracts that cause them to lose focus. 
Steve
 

[Antares]

You're losing at most 10's of fps in the coast for staging.  Other liquid rockets use 6 sec or so.

[nobodyofconsequence]

...Keep in mind that SpaceX has still not actually competed against other launch service vendors. Their launches have been funded by the DOD, which has an interest in reducing launch costs, but can't spend too much on failure. Now SpaceX will have to launch 3 times successfully to even begin to compete. After that will the competitors cut their rates temporarily to force Spacex into bankruptcy?
The assumption that the competition will stand still is a foolish one, but it seems to be in every rocket business plan...

If this is a business, he'll have to take the accumulated dev costs and allocate a portion of them against each successful launch. The more duds, the less the profitability, and at some point there's no way to have a successful business AND a low cost launcher.

The principle advantage to surviving the competition is the profitability of the service.

[nobodyofconsequence]

You're losing at most 10's of fps in the coast for staging.  Other liquid rockets use 6 sec or so.

That's nothing. See your point.

[Comga]

If this is a business, he'll have to take the accumulated dev costs and allocate a portion of them against each successful launch. The more duds, the less the profitability, and at some point there's no way to have a successful business AND a low cost launcher.

You are making an assumption about how Musk wants to conduct his business in this case.  From his past statements, it is not certain that he intends to recoup his initial investment.  He has not said that he will see what the market will bear and will make a profit selling rocket launches at 75% of the going rate.   

One does not set out to cut costs by a factor of ten (whether or not he can do it) just to undersell the competition.  You can either take him at this "visionary" word or not.

[pippin]

If this is a business, he'll have to take the accumulated dev costs and allocate a portion of them against each successful launch. The more duds, the less the profitability, and at some point there's no way to have a successful business AND a low cost launcher.

No, that's a wrong calculation. Sunk costs don't count. For each launch he has to decide if the cost of THAT launch will be recouped by the coming business, the money you spent before is gone, no matter how you decide, only money you are ABOUT to spend counts for a decision.

[William Barton]

Based on what Elon said, the engine thrust taper off was longer than expected, and this is why the vehicle failed. If the first stage had been test launched with a dummy second stage, the nut failure and first stage/second stage contacts would have still happened but would have been much less expensive failures, and would have happened earlier. This path would have been safer, but not as sexy.

Keep in mind that SpaceX has still not actually competed against other launch service vendors. Their launches have been funded by the DOD, which has an interest in reducing launch costs, but can't spend too much on failure. Now SpaceX will have to launch 3 times successfully to even begin to compete. After that will the competitors cut their rates temporarily to force Spacex into bankruptcy?
The assumption that the competition will stand still is a foolish one, but it seems to be in every rocket business plan.
I want SpaceX to succeed, but I think they need to focus on getting to orbit with the Falcon 1, not developing new engines, space capsules and getting more government contracts that cause them to lose focus. 
Steve
 

US competitors who "cut their rates to force SpaceX into bankruptcy" could wind up facing action by the Justice Dept. Foreign competitors... that's 100% politics (instead of only 90%).

[pippin]

US competitors who "cut their rates to force SpaceX into bankruptcy" could wind up facing action by the Justice Dept. Foreign competitors... that's 100% politics (instead of only 90%).

I would think that depends on who the customer is. Which in turn would make it quite unlikely ULA or others would go dump out SpaceX, would probably hurt 'em more than SpaceX does.

[Nate_Trost]

One has to wonder, if they win a services contract under COTS II, if they'll even bother booking any Falcon 1e flights/finishing the vehicle.

[Danderman]

One has to wonder, if they win a services contract under COTS II, if they'll even bother booking any Falcon 1e flights/finishing the vehicle.

I cannot imagine Elon trying to fly Falcon 9 without Falcon 1 having at least one successful flight.

[aero313]

US competitors who "cut their rates to force SpaceX into bankruptcy" could wind up facing action by the Justice Dept. Foreign competitors... that's 100% politics (instead of only 90%).

So it's OK for Elon to sell launches below cost, but not for anyone else to do it?

[Jim]

US competitors who "cut their rates to force SpaceX into bankruptcy" could wind up facing action by the Justice Dept. Foreign competitors... that's 100% politics (instead of only 90%).

So it's OK for Elon to sell launches below cost, but not for anyone else to do it?

They could get sued by the US gov't for overcharging, if they were to advertise lower prices that are not available to the US gov't. 

[pippin]

So it's OK for Elon to sell launches below cost, but not for anyone else to do it?

Depending on the market position: yes!