USAF Certification a possible hindrance to future F9 Development?

[Lar]

Yes, but in the process of describing how it works, with the level of detail that NASA required of them, it would have helped them think through the details better. I think I remember Gwynn Shotwell saying something to that effect.

^this

The best way to understand something is to try to explain it.
The best way to learn something is to try to teach it.

In the software world, explaining how your code works (in a formal code review before it gets promoted) often uncovers some very subtle bugs. So ya, I would expect there is some benefit from the process. As long as it's not adversarial and not overly bureaucratic.

[savuporo]

In the software world, explaining how your code works (in a formal code review before it gets promoted) often uncovers some very subtle bugs. So ya, I would expect there is some benefit from the process. As long as it's not adversarial and not overly bureaucratic.

There are almost always benefits, i have been on both sides of various hardware cert programs ( not aerospace ) including helping creating one - yes it always adds extra costs and can get bureaucratic, but it also does help both sides.
Certifications can take multiple forms - compliance, quality, safety etc and sometimes they are separate tracks, sometimes bundled programs. Mass volume product certification programs are usually simple pass/fail and whoever does the certification tends to be pretty rigid - you either pass or fail, and that's the end of it until you come back with fixes.
If you are talking "certification" where you have a very few products ever being certified, the programs tend to be much more flexible and cert criteria often evolves in lockstep with the products themselves - so both sides benefit.

[Llian Rhydderch]

2) alternatively, SpaceX could fork the design as you pointed out, and lock down a version/model of the F9 for the USG military launches, while allowing some innovation on the other model.  ....

I'd just like to point out that this will have very far reaching effects, including factory equipment, ground systems and potentially pads. Can get super expensive.

Agreed.  That was the "cost" of this option that I spoke about.  So while their are high costs of less innovation and certification costs to stay with the business-as-usual military procurement model, and slow down the movement of technology innovation into non-government market launches, there are, as you note, also high costs of maintaining two processes in production capital equipment and ground systems/pad and the people who run them if SpaceX were to fork the design and maintain an older technology (but ostensibly more stable) military model launcher separate from the an iteratively improving launcher SpaceX is innovating on for non-military launches.

[RocketmanUS]

I am sure that NASA reviews helped them a great deal to build a launcher with zero primary mission failures after losing 3 out of 4 F1s.

Not really.   NASA doesn't tell them how to design or fix their vehicles.  NASA just needs to know how the vehicle works.

What do you mean, how it works?
How the engine works or how it performs?
Rocket vibrations, sound, thermal, ect. ?
Does this include details of every aspect of the vehicle or just how the payload will ride to orbit?

I know the specs for a truck and how the driver and payload will ride on it but don't know how the truck works ( for example ).


Fork production-
Stop an evolution of the F9 and certify it for USAF and have a production line for it.
For other customers have a second line that continues to evolve.
At some point certify a new version of the F9 that has evolved for USAF ( such as full RLV, not just 1st stage  for example ). I would expect the evolution line to be the less production line ( experimental ) and the USAF version to be the work horse for most customer flights.

[AncientU]

Fork production-
Stop an evolution of the F9 and certify it for USAF and have a production line for it.
For other customers have a second line that continues to evolve.
At some point certify a new version of the F9 that has evolved for USAF ( such as full RLV, not just 1st stage  for example ). I would expect the evolution line to be the less production line ( experimental ) and the USAF version to be the work horse for most customer flights.

Don't think this is best path.
1) increases costs, probably lot more than keeping running certification on upgrades
2) upgrades, if/when needed, must be flown on USAF launches to keep reliability improving

If F9/FH are going to fly crew safely, they will be reliable enough for USAF.  Systems that are developed for USAF-specific payloads can be custom built and certified to USAF needs.

[oldAtlas_Eguy]

After thinking about it: the certification process for changes/alterations/upgrades to the LV is written into the Launch Services contract as directions/clauses and data deliverables.  So Some of the price increase for AF launches is the costing for these items over the duration between the contract signing and the actual launch.

This I saw as boilerplate clause and datadeliverables even back in 1980 for Atlas E/F contracts.

The certification process never stops. It is just much more low key than the original certifcation prior to the first contract award.

[R7]

Why wouldn't they [commercial customers] opt for certified, stable and proven version too?

They paid for insurance.

Surely insurance fees would be higher for first flights of any new mods than frozen design with longer flight history?

I'm assuming here that flight on frozen design would not cost substantially more (shouldn't a longer production run do the reverse) for commercial customers than flight on latest mod. Elon's "it's been a paper exercise" comment seems to suggest that they are already building the vehicles in ways compliant with certification.

[john smith 19]

Surely insurance fees would be higher for first flights of any new mods than frozen design with longer flight history?

Not necessarily. Keep in mind how those Spacex tests have been done. They are done after stage or payload separation.

I'm assuming here that flight on frozen design would not cost substantially more (shouldn't a longer production run do the reverse) for commercial customers than flight on latest mod. Elon's "it's been a paper exercise" comment seems to suggest that they are already building the vehicles in ways compliant with certification.

It's the "opportunity cost" of not being able to conduct live flight tests and feeding those improvements (or the fact some changes don't work

AFAIK F9 was designed from day 1 to be EELV compliant. Certification if really for the USAF to confirm  that it has been designed to do that. 

[Jcc]

So what major changes does F9v1.1 need, now that it has legs? I suppose making the second stage reusable, if ever. They would probably need to recertify for that. They certainly will not convert to methane engines, change the core diameter or any other disruptive change. No point in it.

[TrevorMonty]

So what major changes does F9v1.1 need, now that it has legs? I suppose making the second stage reusable, if ever. They would probably need to recertify for that. They certainly will not convert to methane engines, change the core diameter or any other disruptive change. No point in it.

I think there will be two 2nd stage versions, an expendable ( current version) plus a new reusable. That being case they can use current ( certified)  expendable for DOD missions.

[beancounter]

I don't think that Elon wants two streams for the F9.  He's pushing for self-certification which is what he currently has for his commercial launches I think!?
Cheers.

[Lar]

I don't think that Elon wants two streams for the F9.  He's pushing for self-certification which is what he currently has for his commercial launches I think!?
Cheers.

Wait, is he *actually* pushing for self certification, or is that just what his critics are claiming he is wanting?

[Joffan]

I don't think that Elon wants two streams for the F9.  He's pushing for self-certification which is what he currently has for his commercial launches I think!?
Cheers.

Wait, is he *actually* pushing for self certification, or is that just what his critics are claiming he is wanting?

The latter. The phrase was used by an anonymous critic of SpaceX, in the Aviation Week article that Chris rightly decided he didn't wish to host debate on.

[john smith 19]

The latter. The phrase was used by an anonymous critic of SpaceX, in the Aviation Week article that Chris rightly decided he didn't wish to host debate on.

Good point.

AFAIK Spacex are well into the EELV certification. They seem to be happy to do it. If they were really serious about not doing it I don't think they would have got this far without a much louder level of protest.

Speaking to the thread's title. Yes I think it will. How big a hindrance will depend on the USAF's attitude and willingness to update their certification as Space identify improvements and put them into F9 or F9R and the USAF is left behind.

[Wigles]

The latter. The phrase was used by an anonymous critic of SpaceX, in the Aviation Week article that Chris rightly decided he didn't wish to host debate on.

Good point.

AFAIK Spacex are well into the EELV certification. They seem to be happy to do it. If they were really serious about not doing it I don't think they would have got this far without a much louder level of protest.

Of course they are "happy" to do it, not getting EELV certification excludes themselves from what could be potentially 70-80% of the launches which SpaceX may be in a position to compete for (depending on outcome of block buy challenge).

I don't think that Elon wants two streams for the F9.  He's pushing for self-certification which is what he currently has for his commercial launches I think!?
Cheers.

Wait, is he *actually* pushing for self certification, or is that just what his critics are claiming he is wanting?

I agree with beancounter, I don't believe that he would want to have 2 versions. By putting all the major design changes in place now for the certification flights (eg legs) even if he is not using them currently (eg legs) means that the delta-certification requirement is much lower as there will be less design evolution.

Also, I didn't mean to imply that the certification or requirement for re-certification for changes would imply a design freeze. What I mean is that the timeline to incorporate a change will be increased as the certification assessment will have to happen after the design is complete but prior to flight on a revenue launch. This increased cost to incorporate modifications could tip the scales on borderline cost-benefit analysis which on average would lead to less incremental development and larger "block" upgrades.

As a bonus. If the FH core and boosters remain largely similar to the certified FH 1st stage, it makes FH certification much easier.

[Jim]

As a bonus. If the FH core and boosters remain largely similar to the certified FH 1st stage, it makes FH certification much easier.

The boosters aren't

[Jim]

Certification is not just of the hardware but of the integration processes.
Also, there are changes to the vehicle going on all the time.

[Wigles]

As a bonus. If the FH core and boosters remain largely similar to the certified FH 1st stage, it makes FH certification much easier.

The boosters aren't

They arent 100% the same but subsystems would be very similar, eg the engines are the same, the octoweb engine assembly would be 99% the same (except for crossfeed) the construction and testing and QA processes are the same, the internal piping would be largely similar.

Major differences would be the length, load paths, vibration & fatigue assessments, etc. still a body of work to do but less than starting from scratch.

[Jim]

As a bonus. If the FH core and boosters remain largely similar to the certified FH 1st stage, it makes FH certification much easier.

The boosters aren't

They arent 100% the same but subsystems would be very similar, eg the engines are the same, the octoweb engine assembly would be 99% the same (except for crossfeed) the construction and testing and QA processes are the same, the internal piping would be largely similar.

Major differences would be the length, load paths, vibration & fatigue assessments, etc. still a body of work to do but less than starting from scratch.

which I would say excludes the term " largely similar ".  That is why NASA and USAF treat the vehicle as a separate one for cert purposes.

[Owlon]

which I would say excludes the term " largely similar ".  That is why NASA and USAF treat the vehicle as a separate one for cert purposes.

We're just arguing semantics here. Many people would acknowledge those differences and still call them largely similar, I'm sure, but we call all definitely agree they're not identical to the core and that FH will need a not insignificant amount of work to be certified.