SpaceX COTS Demo 1 Updates

[Antares]

It's processes and systems engineering that guarantee success. Read The Secret of Apollo. Elon has given speeches on his belief that systems engineering is an artifact of the DoD. Until they adopt that or tell me they have some other "secret" I won't risk my DNA or my irreplaceable payload with them.

I'm enjoying this debate, though.  As they say, where you stand depends on where you sit.

[Jim]

The Dragon capsule will have a propellant load capable of 300 Delta-V at a minimum.  The flight control of the Dragon capsule's trajectory after engine-out (last minute of powered flight about 900 nm downrange) should adjust the reentry angle such that Dragon will finish in the Atlantic. 

What says the Dragon is active during ascent, most spacecraft aren't.

[Jim]

1.   If the Demo's go as well as us "enthused SpaceX fans" their manifest is expected to explode.

2.  The costs will likely go up a bit but they will not be "similar" to other launch providers.

You have no proof of the #1 and you are wrong about #2, their costs are getting close

[Jim]

Jim, I'm pretty sure that Dragon, whatever is actually inside, counts as a real spacecraft.

No, it is not the paying customer

[Robotbeat]

It's processes and systems engineering that guarantee success. Read The Secret of Apollo. Elon has given speeches on his belief that systems engineering is an artifact of the DoD. Until they adopt that or tell me they have some other "secret" I won't risk my DNA or my irreplaceable payload with them.

I'm enjoying this debate, though.  As they say, where you stand depends on where you sit.

Very interesting topic to me, actually. Would you bother explaining more your thoughts on this, perhaps on another thread?

[jongoff]

It's processes and systems engineering that guarantee success. Read The Secret of Apollo. Elon has given speeches on his belief that systems engineering is an artifact of the DoD. Until they adopt that or tell me they have some other "secret" I won't risk my DNA or my irreplaceable payload with them.

I'm enjoying this debate, though.  As they say, where you stand depends on where you sit.

While I actually agree with you that appropriately applied System Engineering can make a huge difference...I think that far too often Systems Engineering is just buzzword engineering in many aerospace circles.  I've seen it used well, but in spite of probably having more systems engineers per square inch at NASA than anywhere else in the planet, what else could you call the CxP debacles (and HEFT, and ESAS, etc) than total System Engineering snafus of the first degree.

I think that's part of the unfortunate aversion to appropriate systems engineering in the commercial space world.  When you see wunderkinden like Mike Griffin giving lectures on "How to Fix Systems Engineering" at JSC, with no sense of the irony involved, it tends to initiate baby-bathwater-tossing-reflexes on many peoples' parts.

Once again, not opposed to Systems Engineering, just tend to cringe whenever I hear that some NASA or DoD contractor lists "Systems Engineering" as their key area of core competence...

~Jon

[jongoff]

The Dragon capsule will have a propellant load capable of 300 Delta-V at a minimum.  The flight control of the Dragon capsule's trajectory after engine-out (last minute of powered flight about 900 nm downrange) should adjust the reentry angle such that Dragon will finish in the Atlantic. 

What says the Dragon is active during ascent, most spacecraft aren't.

Yeah, while it could be possible to do something like that, doing it 100% autonomously, and being able to handle all necessary aborts would take a *lot* of mission specific engineering...I'm with Jim, I just don't think it's realistic.

Now, if you have a pilot on board, it's quite a bit easier to have some real-time retargetting for safety purposes.  But trying to predict every single possible abort mode for every flight trajectory you use in advance?  Very non-trivial.  And in some ways likely to make things less reliable--the more complex your GN&C system gets, the more likely you'll end up needing that abort capability.

~Jon

[Jim]

What's going here? Not all of the pictures (different days, different times of day?) show this umbilical like this, but what gives?

Nobody is running with this?  Something is disconnected.

http://forum.nasaspaceflight.com/index.php?topic=22041.msg667684#msg667684

[Mapperuo]

What's going here? Not all of the pictures (different days, different times of day?) show this umbilical like this, but what gives?

Nobody is running with this?  Something is disconnected.

http://forum.nasaspaceflight.com/index.php?topic=22041.msg667684#msg667684

Had to go seaching through the NASA gallery, seems the photos when its disconnected are from the 3rd Dec (Abort day) but on the 4th Dec photos it's connected just fine. (Successful 2sec test day)

[mr. mark]

Jim, if I'm reading right you are saying that Dragon is not a spacecraft because it's not a paid for payload. That's crazy. Sorry

[Malderi]

Jim, I'm pretty sure that Dragon, whatever is actually inside, counts as a real spacecraft.

No, it is not the paying customer

Ah, if that was your rationale, you should've said so. I agree that interfacing with an external organization and interfaces will be a real test. Almost all of their Falcon 1 launches had those, though, whether TacSATs or RazakSATs or whatever, so they've had at least some experience with doing that.

It's probably not the level of engineering/analysis that goes into billion-dollar NRO recon birds, but I don't think SpaceX is going for that market anyways.

[Pedantic Twit]

What's going here? Not all of the pictures (different days, different times of day?) show this umbilical like this, but what gives?

Nobody is running with this?  Something is disconnected.

http://forum.nasaspaceflight.com/index.php?topic=22041.msg667684#msg667684

Had to go seaching through the NASA gallery, seems the photos when its disconnected are from the 3rd Dec (Abort day) but on the 4th Dec photos it's connected just fine. (Successful 2sec test day)

Here's the links for the curious.

Disconnected (3rd Dec):
http://mediaarchive.ksc.nasa.gov/detail.cfm?mediaid=49712
http://mediaarchive.ksc.nasa.gov/detail.cfm?mediaid=49713
http://mediaarchive.ksc.nasa.gov/detail.cfm?mediaid=49715
Connected (4th Dec):
http://mediaarchive.ksc.nasa.gov/detail.cfm?mediaid=49707
http://mediaarchive.ksc.nasa.gov/detail.cfm?mediaid=49708
http://mediaarchive.ksc.nasa.gov/detail.cfm?mediaid=49709

[jimvela]

Once again, not opposed to Systems Engineering, just tend to cringe whenever I hear that some NASA or DoD contractor lists "Systems Engineering" as their key area of core competence...

I think, though, that you'd have to agree that the only thing worse than having too many systems engineers is... not having enough. ??

Lets hope that the launch goes well, although I'm not drinking any kool-aid, I am hoping to see SpaceX succeed.

[iamlucky13]

Going back to the nozzle...

Since thermal concerns were mentioned, remember that radiative heating was implicated in the failure of the roll control actuator on flight one. SpaceX says they've fixed the issue, but increased radiative heating could affect the fix.

However, the heat transfer analysis is not simple. The nozzle extension clearly (from past launch videos) gets very hot. It gains heat from the plume via both radiation and convection. It then radiates all of that heat with a high emissivity.

The unconstrained plume has only radiation to heat the rocket by, and while the angle of view effectively subtends infinity, it does not maintain its temperature that far. It also, as a diffuse gas rather than a sheet of metal, does not have the same emissivity.

So between the plume being exposed closer or there being less nozzle to radiate, I don't know where the total heating of the 2nd stage ends up.

And as for SpaceX's analysis - I suspect they analyzed resulting ISP and delta-V, but didn't go as far as heat transfer.


Anyways - shortened nozzle should be noticeable in the 2nd stage firing video.

[Kabloona]

And speaking of the nozzle again, this bit from the SpaceX Facebook page confirms that it was in fact the photos taken at "closeout" that showed the cracks, not photos from earlier build/inspection at the factory:

Quote: "@Muttley, Ben is correct, we take closeout photos as part of our final review process and the engineers want to take a look at the 2nd stage nozzle."

[butters]

Going back to the nozzle...

Since thermal concerns were mentioned, remember that radiative heating was implicated in the failure of the roll control actuator on flight one. SpaceX says they've fixed the issue, but increased radiative heating could affect the fix.

However, the heat transfer analysis is not simple. The nozzle extension clearly (from past launch videos) gets very hot. It gains heat from the plume via both radiation and convection. It then radiates all of that heat with a high emissivity.

The unconstrained plume has only radiation to heat the rocket by, and while the angle of view effectively subtends infinity, it does not maintain its temperature that far. It also, as a diffuse gas rather than a sheet of metal, does not have the same emissivity.

So between the plume being exposed closer or there being less nozzle to radiate, I don't know where the total heating of the 2nd stage ends up.

And as for SpaceX's analysis - I suspect they analyzed resulting ISP and delta-V, but didn't go as far as heat transfer.

AIUI, the S2 roll actuator anomaly on F9 flight 1 was caused by the opposite problem: a cold GOX vent impinging on the actuator and chilling by convection the RP-1 hydraulic fluid in the actuator until it froze or gelled stuck. SpaceX has responded to the problem by relocating the vent away from the actuator. They may also have added some insulation around the actuator (not sure).

As for the thermal environment: if we're talking about it here, then SpaceX has surely also thought about it come up with some sort of answer. It's inconceivable that they would just shrug their shoulders and hope for the best. Note that MVac is ground-tested without the nozzle extension, so they must already have the thermal data on that configuration.

[joshcryer]

You have no proof of the #1 and you are wrong about #2, their costs are getting close

#1 is what they say.

#2 is fact for anyone capable of 3rd grade math.

Arianespace and ESA CRS is $62k / kg.

Orbital with Tarus II and Cygnus is $88k / kg.

SpaceX with Falcon 9 and Dragon is $22k / kg.

The costs are substantially lower and that is still including an ISS resupply premium. In reality SpaceX is pricing, on their site, $11k / kg for GTO, you can do the math on the other options, but no one comes close.

Name one company that is "priced similarly."

SpaceX already is the lowest cost supplier with 30 launches on their manifest. If this wasn't viable, if they wrongly priced 5 years worth of launches, then clearly they will go bankrupt. Arguing for their pricing being wrong is effectively arguing that.

[Kabloona]

Going back to the nozzle...

Since thermal concerns were mentioned, remember that radiative heating was implicated in the failure of the roll control actuator on flight one. SpaceX says they've fixed the issue, but increased radiative heating could affect the fix.

However, the heat transfer analysis is not simple. The nozzle extension clearly (from past launch videos) gets very hot. It gains heat from the plume via both radiation and convection. It then radiates all of that heat with a high emissivity.

The unconstrained plume has only radiation to heat the rocket by, and while the angle of view effectively subtends infinity, it does not maintain its temperature that far. It also, as a diffuse gas rather than a sheet of metal, does not have the same emissivity.

So between the plume being exposed closer or there being less nozzle to radiate, I don't know where the total heating of the 2nd stage ends up.

And as for SpaceX's analysis - I suspect they analyzed resulting ISP and delta-V, but didn't go as far as heat transfer.


Anyways - shortened nozzle should be noticeable in the 2nd stage firing video.

Yes, the thermal effect was my first concern as well, but I can't imagine that their analysis did NOT go as far as heat transfer, as you say. THAT would definitely be "systems engineering" failure, IMO. A previous post speculated that they've already done the analysis for vacuum without the extension (which they've said they don't "need"), which would represent a more severe heating case, and therefore they've already bounded the problem. That would seem to make sense, and would explain why they're apparently comfortable going this route.

[sdsds]

Going back to the nozzle...
[...]
I suspect they analyzed resulting ISP and delta-V, but didn't go as far as heat transfer.

It's processes and systems engineering that guarantee success. [...] Until they adopt that or tell me they have some other "secret" I won't risk my DNA or my irreplaceable payload with them.

I suspect SpaceX has computer models and sufficient computational resources to run -- as many times as they like with as many variations as they can imagine -- complete simulations of an entire flight.  For the first stage that includes atmospheric, mechanical, thermal, and vibro-acoustic dynamics.  For the second stage, it includes all of those that still apply outside the atmosphere, plus solar effects and anything else known to be of importance in that flight regime. 

If they had that, then they could try out design variations (in the simulator) whenever they thought of them and quickly determine if they're broken.  Do you believe heavy reliance on simulation like that could be the "secret" that lets them avoid endless up-front systems engineering cycles?  Would it make their current behavior as regards the "short extension" mvac appear rational?

[Kabloona]

I suspect SpaceX has computer models and sufficient computational resources to run -- as many times as they like with as many variations as they can imagine -- complete simulations of an entire flight.  For the first stage that includes atmospheric, mechanical, thermal, and vibro-acoustic dynamics.  For the second stage, it includes all of those that still apply outside the atmosphere, plus solar effects and anything else known to be of importance in that flight regime. 

If they had that, then they could try out design variations (in the simulator) whenever they thought of them and quickly determine if they're broken.  Do you believe heavy reliance on simulation like that could be the "secret" that lets them avoid endless up-front systems engineering cycles?  Would it make their current behavior as regards the "short extension" mvac appear rational?

No, I think butters has it correct:

Quote: "Note that MVac is ground-tested without the nozzle extension, so they must already have the thermal data on that configuration. "