"Elon wants to self-certify"

[MP99]

SpaceX supporters have suggested that the Air Force is dragging its feet on the certification. It is “very difficult to pick up the pace,” Shelton says. “It just takes time. It takes money. It takes people. I think SpaceX would have a hard time going faster than we are going right now.”

Ultimately, however, the industry source says SpaceX’s interest is not simply in getting Falcon 9v1.1 certified to compete, as it is only eligible to lift a small portion of the projected national security payloads on the manifest. The holy grail is to raise enough questions about the certification process so as to short circuit it and, eventually, eliminate it in its current form before SpaceX fields the heavy vehicle.

Certification is thorny for SpaceX because the company does not fly standard configurations, as it often includes risk-reduction work on missions. “Elon wants to self-certify,” the industry source says, referring to SpaceX founder, CEO and Chief Designer Elon Musk. “The Air Force is not going to be able to certify [because] each time Elon flies a mission, it is a new rocket.”

[My highlighting]


Well, that is quite a contention.

I find it hard to argue that SpaceX will continue to evolve F9, eg Elon has talked re propellant densification.

My thought is "good luck getting the AF to go for self-certification". But, it's easy to see how there could be a clash between a Silicon Valley / forever-moving-target culture, and AF wanting to fly the config that it spent $60m+ certifying. (Or, if you want a humorous interjection, re-certifying the changes for every flight could cut a lot from SpaceX's hoped-for cost advantage.)

With the history behind F9 v1.1 (F1 > F9 v1.0 > v1.1 single-cores), it's easy to see how this is relatively mature. With the article quoting 2018 for FH certification, I'd think this will undergo a lot more evolution in it's early flights, what with being their first three-core, introducing cross-feed, etc...

So, if "Elon wants to self-certify" is a true representation of what they're going to fight for - is there any chance that the AF would go for that?

If I was in the AF's position, my first riposte would be "this payload cost us $xM - you insure it on your dime". But, of course, that still doesn't cover a long delay to any capability that would be lost by a failure.

Would such a change have to be authorised by Congress? Is there any chance that Congress might impose this as a cost-cutting measure?

I note that NASA has an alternative route to certification:-

14 consecutive successful flights (95% demonstrated reliability) of a common launch vehicle configuration, instrumented to provide design verification and flight performance data

This sounds more in line with what the AvWeek article suggests - could the AF go for such a process down the line? (With obvious issues if the design is continuously evolving.)

cheers, Martin

[Darkseraph]

Doubt it will happen. The Falcon 9 has less configurations than the EELV's and I would think AF would pay good money for him not to do resusability testing during missions and just use the expendable versions so that they're not changing much.

[docmordrid]

There should be some changes, otherwise it's a major disruption to make running improvements and all you're doing is keep $$$$ "independent" consultants (*cough* AC *cough*) in new Beemers.

[Darkseraph]

There should be some changes, otherwise it's a major disruption to make running improvements and all you're doing is keep $$$$ "independent" consultants (*cough* AC *cough*) in new Beemers.

Probably but I think the Air Force is more interested in the thing just working and launching on time rather than "improvements".

[docmordrid]

If the changes can improve both without involving the expensive rented help both sides can be satisfied.

The problem seems to be a lack of in-house expertise in the USAF. Could be time to change that, or for USAF to get over its NASAphobia. The status quo seems to be a huge price and (improvements) delay  driver and we can no longer afford to ignore those.

[yg1968]

It's a very poor article. Most of the content is based on the opinion of an unnamed source. If you are going to provide opinions and make accusations, you shouldn't remain anonymous.

[Lee Jay]

Remember when Elon said no FTS is needed on SpaceX rockets?  How'd that work out?

[laika_fr]

Spacex quickly improves the design until it flies humans, then the polemic will be over.

[OnWithTheShow]

But did ULA even complete "certification" flights of Delta and Atlas? And then wouldn't each Atlas variant then be a "different rocket" that would have to complete it's own 3 certification flights? Some variants haven't even completed 3 flights. 

[Jim]

But did ULA even complete "certification" flights of Delta and Atlas? And then wouldn't each Atlas variant then be a "different rocket" that would have to complete it's own 3 certification flights? Some variants haven't even completed 3 flights. 

not required, the USAF paid for and participated in the development of the Atlas V and Delta IV, hence no need for certification.   

When it comes to certification, only the common core matters.  The number of strap on SRM's, third stages or PLF sizes does not play into it.

So only two variants of Atlas V were needed to be certified by NASA, the 4XX and 5XX series.  This is due to the encapsulation of the Centaur by the 5m fairing which induces different load paths.

If NASA were to cert the Delta IV, there would be 3 configurations, 4m & 5m Mediums and the Heavy

[OnWithTheShow]

If only the core matters then FH should not require certification.

[Jim]

If only the core matters then FH should not require certification.

No,

a.  Like I listed, the Delta Heavy would require cert and so would be the FH.  The triple core configuration is not the same as a single core especially when it comes to load paths and hence the need for cert

b.  The FH side cores are different than the center core.

[savuporo]

"Each time Elon flies a mission, it's a new rocket"

This. A completely valid concern - they will not stop tinkering with the vehicle configuration, and have fundamentally bundled their R&D efforts with operational launches.
This must make other customers as concerned.

[Jason1701]

"Each time Elon flies a mission, it's a new rocket"

Not if he makes them reusable.

[savuporo]

"Each time Elon flies a mission, it's a new rocket"

Not if he makes them reusable.

If he wants to certify some future reusable rocket configuration for USAF payloads, i'm sure they'll be happy to do that, once it has flown the required demo flights, given payload specs etc.
That is not what is happening right now though

[RonM]

"Each time Elon flies a mission, it's a new rocket"

This. A completely valid concern - they will not stop tinkering with the vehicle configuration, and have fundamentally bundled their R&D efforts with operational launches.
This must make other customers as concerned.

That is a concern. Rockets are not software. You don't release a rocket and hope the customers will finish QA for you by risking their payload.

That said, the fly back and landing testing of the first stage after the payload is on it's way to orbit is okay, as long as the modifications don't jeopardize the launch.

[savuporo]

...as long as the modifications don't jeopardize the launch.

This discussion has happened here before - it is very hard to reasonably demonstrate that any hardware, software or process change did not affect the launch. That would be a small  "certification" process in itself.

[RocketGoBoom]

"Each time Elon flies a mission, it's a new rocket"

This. A completely valid concern - they will not stop tinkering with the vehicle configuration, and have fundamentally bundled their R&D efforts with operational launches.
This must make other customers as concerned.

As long as each rocket gets into orbit, despite the incremental improvements, then gradually SpaceX will earn the trust of even the doubters.

If SpaceX has a total failure during a launch, then all of the naysayers will have plenty of ammunition to aim at the tinkering process of v1.1, v1.2, v1.3.

But if SpaceX goes another two years with all rockets getting into orbit, landing stage 1 in Florida with full HD video, etc. Then it will be tough for the old school guys to keep denying SpaceX's way of doing things. In fact, all of the other players will likely have to start copying SpaceX if they want to remain relevant.

[su27k]

"Each time Elon flies a mission, it's a new rocket"

This. A completely valid concern - they will not stop tinkering with the vehicle configuration, and have fundamentally bundled their R&D efforts with operational launches.
This must make other customers as concerned.

Jim said multiple times that ULA improve their LV constantly, an Altas V this year is not exactly the same from an Atlas V last year. I was expecting him shooting down the self-certify as non-sense, but it didn't happen, strange...

[Jim]

Jim said multiple times that ULA improve their LV constantly, an Altas V this year is not exactly the same from an Atlas V last year. I was expecting him shooting down the self-certify as non-sense, but it didn't happen, strange...

And each improvement is vetted by the USAF and NASA.  ULA doesn't get to change the Atlas without approval.

As for the self-certifying, I thought that it was so ludicrous that it didn't need any help to point that outl

[Jim]

As long as each rocket gets into orbit, despite the incremental improvements, then gradually SpaceX will earn the trust of even the doubters.

If SpaceX has a total failure during a launch, then all of the naysayers will have plenty of ammunition to aim at the tinkering process of v1.1, v1.2, v1.3.

But if SpaceX goes another two years with all rockets getting into orbit, landing stage 1 in Florida with full HD video, etc. Then it will be tough for the old school guys to keep denying SpaceX's way of doing things. In fact, all of the other players will likely have to start copying SpaceX if they want to remain relevant.

Timeliness is almost as important.

[Prober]

"Each time Elon flies a mission, it's a new rocket"

Not if he makes them reusable.

doubt the AF is going for reusable to save a few $$.

[savuporo]

As long as each rocket gets into orbit, despite the incremental improvements, then gradually SpaceX will earn the trust of even the doubters.
If SpaceX has a total failure during a launch, then all of the naysayers will have plenty of ammunition to aim at the tinkering process of v1.1, v1.2, v1.3...

This is not about "naysaying" or "doubting" ( rockets dont fly on faith, by the way ). This is about the fundamental issue of certifying a moving target.

I dont know how USAF certifies their rockets, but if i ran my product through FCC cert testing tomorrow, and then went and changed the antenna designs or duty cycle triggered by the software, i would have to re-certify the whole thing if i wanted to ship the new config to a customer.

[RocketGoBoom]

"Each time Elon flies a mission, it's a new rocket"

Not if he makes them reusable.

doubt the AF is going for reusable to save a few $$.

I have no doubt at all that the AF will be using reusable rockets if SpaceX proves the technology works. The AF won't be the first customer. But if it becomes common and reliable for commercial launches, then the AF will follow a few years later.

[clongton]

As long as each rocket gets into orbit, despite the incremental improvements, then gradually SpaceX will earn the trust of even the doubters.

If SpaceX has a total failure during a launch, then all of the naysayers will have plenty of ammunition to aim at the tinkering process of v1.1, v1.2, v1.3.

But if SpaceX goes another two years with all rockets getting into orbit, landing stage 1 in Florida with full HD video, etc. Then it will be tough for the old school guys to keep denying SpaceX's way of doing things. In fact, all of the other players will likely have to start copying SpaceX if they want to remain relevant.

Timeliness is almost as important.

To be fair Jim most of their recent launch delays (not all) were not related to the launch vehicle. Things like weather, a fire in range radar outpost and ISS equipment failures come to mind.

Yes they need to get their *own* launch delays into better territory, but let's not be just saying "they delayed x times", when not all of those delays had anything to do with them.

[Chris Bergin]

Another somewhat pointless thread, based on a "mobile site" article on another site where there's one comment on their comment section and already going on 30 here.

I don't want this place turning info a free forum for other sites. Just throws all the costs on us and all the benefit in visitors to the external sites.

Let's give this place unique value via informed discussion and original content, as opposed to the above.

Besides, it's a boring thread. I didn't set this place up for boring conversations. Desperate to talk about it? This article has a comment section below it.

Locked.