USAF Certification a possible hindrance to future F9 Development?

[Wigles]

Noting the significant amount of effort that goes into military certification (having been involved in military aircraft certification), this could this lead to either a stagnation of F9 development, or a fork in design's.

Aircraft modifications need to be assessed for their impact on the certification basis, and similarly I expect that modifications which SpaceX may wish to put on F9 in the future could require a costly and lengthy re-certification process. Now minor modifications will probably only require notification to the USAF and an internal assessment of the impact, but for majors there could be a multi-month delay while the modification is individually certified.

This could lead to either a stangation in overall F9 development (acknowledging that maintaining USAF certification would decrease risk & insurance costs for commercial customers) or a fork where the certified rocket doesn't get as rapidly updated as the commercial launcher.

Thoughts?

edit: made the subject less definitive, added word 'possible'.

[Llian Rhydderch]

Noting the significant amount of effort that goes into military certification (having been involved in military aircraft certification), this could this lead to either a stagnation of F9 development, or a fork in design's.

Aircraft modifications need to be assessed for their impact on the certification basis, and similarly I expect that modifications which SpaceX may wish to put on F9 in the future could require a costly and lengthy re-certification process. Now minor modifications will probably only require notification to the USAF and an internal assessment of the impact, but for majors there could be a multi-month delay while the modification is individually certified.

This could lead to either a stangation in overall F9 development (acknowledging that maintaining USAF certification would decrease risk & insurance costs for commercial customers) or a fork where the certified rocket doesn't get as rapidly updated as the commercial launcher.

Thoughts?

It has been argued by many that SpaceX needs to go after every segment of the mid-size to heavy launch market in order to have the sort of volume and launch rate needed to support their low-cost objectives in general and their reusable launch vehicle objectives in particular.  I tend to agree with this.

But your point is a good one.  All actions have a cost, and the cost of SpaceX pursuing US government military "certification" will certainly preclude a number of other courses of action, such as making iterative incremental changes to the design between successive launches for the US military without the Air Force certification bureaucracy getting in their shorts on design, and process, and sufficient verification and validation testing, and sometimes, resolving political issues with important generals and Congressfolk or the Administration.

My sense is that with the largely monopsonistic market that has existed in space launch for decades (for many reasons—reasons that are probably off topic here) with the small exception of the commercial geosynchronous commsat market, SpaceX has to endeavor to sell to that monopsony buyer:  the US government.  And it needs to do so in both the civilian and military sides of the US government space launch purchasing programs, NASA and USAF.

I look at this cost a couple of ways: 

1) it may be advisable^[1] to halt the innovation on the F9 launch system once the USG certifies the F9 for its use.  Move the innovation to other SpaceX launch systems that are not yet tied down by the government process:  FH initially and the MCT launch vehicle.  (It can be argued that by SpaceX modifying the F9 v1.0 design to the F9 v1.1 design, they have got 95%+ of what they want in the rocket, and that little additional innovation is needed in that particular design—so perhaps the cost is not high.  But I can see arguments on the other side of that too—the cost of impeding innovation is larger than is easily seen.)

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.  This is contrary to SpaceX design philosophy to this point, and would add significant costs to manufacturing, procurement, launch ops, etc.  But there is no getting around that this alternative is one way of dealing with the cost of the government additional regulation for military launches.

But yes, either way, selling to the military segment of the US launch market does incur costs.

Cheers,
  Llian


[1]—i.e., economically advisable from the point of view of the company.  What potentially deleterious effect this has on F9 technology innovation for the "progress of space launch" to the rest of us would remain, however, an unintended consequence of the regulatory aspects of the USG/USAF processes.  While unfortunate, it won't be the first time a government regulatory process has hampered innovation.

[rst]

This could lead to either a stangation in overall F9 development (acknowledging that maintaining USAF certification would decrease risk & insurance costs for commercial customers) or a fork where the certified rocket doesn't get as rapidly updated as the commercial launcher.

Thoughts?

Question:  we know the certification requirements for the original EELV program were rather less stringent than what's being required of SpaceX to compete for contracts -- in particular, the original EELVs were not required to have three successful launches in the same configuration before a DOD mission.  (The second launch of Delta IV Heavy was an NRO payload, after a first launch with premature engine shutdowns.)

So, is it possible that the DOD will loosen up on the requirements for a commercial launcher once one variant has achieved certification by the full three-launch-and-design-review criteria?

[hrissan]

Elon referred to certification as a "paper excersize".

Is it magic? If you fill lots of forms, the flying machine will not break?

Nowadays in the foreign trade "certification" means "trade war, but without violating WTO rules". Producer from the country A is requested to get certification in country B to sell things there. Producer spends lots of money and year or two, and close to the end the certification requirements change or another different certification is devised.

So is Air Force certification the form of "trade war"? Why the insurance does not work? They have unique birds, so what? They may lose any of them, so should have Plan B anyway.

[Jim]

Question:  we know the certification requirements for the original EELV program were rather less stringent than what's being required of SpaceX to compete for contracts -- in particular, the original EELVs were not required to have three successful launches in the same configuration before a DOD mission.  (The second launch of Delta IV Heavy was an NRO payload, after a first launch with premature engine shutdowns.)

So, is it possible that the DOD will loosen up on the requirements for a commercial launcher once one variant has achieved certification by the full three-launch-and-design-review criteria?

Incorrect, the original EELV's were develop to USAF specs and paid for by the USAF.  The USAF then participated in the development.   

Certification is for vehicles that were not developed by the USAF.
NASA does its a cert for vehicles that were not developed by the it.  Atlas V went through the NASA cert process.

[Kabloona]

Elon referred to certification as a "paper excersize".

Is it magic? If you fill lots of forms, the flying machine will not break?

Nowadays in the foreign trade "certification" means "trade war, but without violating WTO rules". Producer from the country A is requested to get certification in country B to sell things there. Producer spends lots of money and year or two, and close to the end the certification requirements change or another different certification is devised.

So is Air Force certification the form of "trade war"? Why the insurance does not work? They have unique birds, so what? They may lose any of them, so should have Plan B anyway.

No, it's not a form of "trade war." The DoD payloads are extremely expensive and the Air Force is paranoid about losing a $1 billion payload. So they are doing everything they can think of to ensure the launch vehicles they purchase for those missions are reliable. The "certification" process is simply the Air Force going over the design with a fine toothed comb to decide whether it will meet their stringent requirements that they expect will result in high reliability.

As for insurance, the US Government does not buy insurance. They self-insure, which means they just swallow the cost of a lost payload and move on. So if they do lose one, they do have a "Plan B," which is to build another.

[AncientU]

This could lead to either a stangation in overall F9 development (acknowledging that maintaining USAF certification would decrease risk & insurance costs for commercial customers) or a fork where the certified rocket doesn't get as rapidly updated as the commercial launcher.

Thoughts?

Working from the assumption that DoD certification/USG launches are a means to an end, not the end in itself (i.e., a source of revenue, not a Corporate goal):
1) the revenue stream is obviously the largest in the space launch world, so setting up a separate unit to deal with the certification process paperwork and an ongoing stream of change (continuous improvement) documentation after certification, should be relatively easy to fund from this source;
2) there could be significant launch reliability lessons learned by SpaceX as a result of an end-to-end scrub of current design and operations (I would hope) -- this could help long term with hardware and operations reliability;
3) a downside risk could be letting full documentation of the  F9 v1.1 recipe out of Hawthorne -- no patents is SpaceX policy to avoid this -- but we can hope the certification process isn't 'leaky.'

Overall, I don't think stagnation need be too high of risk. The clash of cultures will be very much in the forefront, though.

[Jim]

But your point is a good one.  All actions have a cost, and the cost of SpaceX pursuing US government military "certification" will certainly preclude a number of other courses of action, such as making iterative incremental changes to the design between successive launches for the US military without the Air Force certification bureaucracy getting in their shorts on design, and process, and sufficient verification and validation testing, and sometimes, resolving political issues with important generals and Congressfolk or the Administration.

This is already happening with NASA. F9 is being certified by NASA and any changes have to be vetted by NASA

1) it may be advisable^[1] to halt the innovation on the F9 launch system once the USG certifies the F9 for its use.  Move the innovation to other SpaceX launch systems that are not yet tied down by the government process:  FH initially and the MCT launch vehicle. 

FH is the main vehicle to go after the military market since the F9 captures very few of the missions

While unfortunate, it won't be the first time a government regulatory process has hampered innovation.

Certification is not part of a regulatory process.  It is a procurement process.

[Jim]

3) a downside risk could be letting full documentation of the  F9 v1.1 recipe out of Hawthorne -- no patents is SpaceX policy to avoid this -- but we can hope the certification process isn't 'leaky.'

There is no risk.
a. there isn't anything really advanced
b.  It is ITAR information, and use USG can guard it
c.  It is propriety information, and  USG guards just like the other vehicles' information.

[Lar]

3) a downside risk could be letting full documentation of the  F9 v1.1 recipe out of Hawthorne -- no patents is SpaceX policy to avoid this -- but we can hope the certification process isn't 'leaky.'

If there are 100 people working on something, chances are there will be some minor leakage (at least on the level of people telling their wives little tidbits) unless it's been classified.  That is in no way intended as a slight against the USAF (especially not on Memorial Day!!!), it's just human nature.

While unfortunate, it won't be the first time a government regulatory process has hampered innovation.

Certification is not part of a regulatory process.  It is a procurement process.

Semantics. It's a government process. It would not be the first time a government process hampered ___X___  (where X is any number of desirable things)

[Jim]

Semantics. It's a government process. It would not be the first time a government process hampered ___X___  (where X is any number of desirable things)

Wrong, it is not semantics.  Nobody is forcing SX.   SX does not have to be certified and it doesn't have to fly USAF missions.

[Lar]

Semantics. It's a government process. It would not be the first time a government process hampered ___X___  (where X is any number of desirable things)

Wrong, it is not semantics.  Nobody is forcing SX.   SX does not have to be certified and it doesn't have to fly USAF missions.

True, but not relevant to my point which is simply " It would not be the first time a government process hampered ___X___  (where X is any number of desirable things)"  In this case, X is "save money for the taxpayer" and if the certification process adds needless cost ... QED

(I am sure that you are about to reply that not one cent of cost added is "needless"... save it ... that assertion will never fly)

[ncb1397]

3) a downside risk could be letting full documentation of the  F9 v1.1 recipe out of Hawthorne -- no patents is SpaceX policy to avoid this -- but we can hope the certification process isn't 'leaky.'

There is no risk.
a. there isn't anything really advanced
b.  It is ITAR information, and use USG can guard it
c.  It is propriety information, and  USG guards just like the other vehicles' information.

a.) If there was anything SpaceX wanted to keep secret, you wouldn't necessarily know about it. Their engine seems to be really inexpensive with modern manufacturing techniques(mainly CNC) given that you can do 10 of them on a 56 million dollar mission. Tom Mueller famously built a medium sized engine in his garage before working on Merlin. IMO the engine details would be potentially a place where a competitor would want to replicate technology in order to potentially realize any cost savings and gain competitive advantage. Merlin has also gone through multiple design iterations over many many years which, if you wanted to gain the ability, design information would allow for skipping intermediate steps and going directly to the end-point.

c.)USG doesn't gaurd against USAF personnnel going to work for AJR and taking their brain with lots of information with them. Turn-over at SpaceX is one avenue of SpaceX trade secrets leaking. Design information residing at the Air force as well is simply another avenue that runs in parallel.

[Jim]

a.) If there was anything SpaceX wanted to keep secret, you wouldn't necessarily know about it. Their engine seems to be really inexpensive with modern manufacturing techniques(mainly CNC) given that you can do 10 of them on a 56 million dollar mission. Tom Mueller famously built a medium sized engine in his garage before working on Merlin. IMO the engine details would be potentially a place where a competitor would want to replicate technology in order to potentially realize any cost savings and gain competitive advantage. Merlin has also gone through multiple design iterations over many many years which, if you wanted to gain the ability, design information would allow for skipping intermediate steps and going directly to the end-point.

c.)USG doesn't gaurd against USAF personnnel going to work for AJR and taking their brain with lots of information with them. Turn-over at SpaceX is one avenue of SpaceX trade secrets leaking. Design information residing at the Air force as well is simply another avenue that runs in parallel.

a.  Not true.  If SX wants to be certified, there will be no stone left unturned

c.  Yes, it does.  And it is more than USAF, it is Aerospace Corp and the many support contractors, who have to sign disclosure statements.

And the USAF isn't all the issue, Falcon 9 has been going under NASA cert since the Jason-3 contract award.  NASA and it's contractors have been looking at F9 a lot longer.

Edit/Lar: fixed misattribution

[AncientU]

2) there could be significant launch reliability lessons learned by SpaceX as a result of an end-to-end scrub of current design and operations (I would hope) -- this could help long term with hardware and operations reliability;

On this point, does anyone know if the certification process is designed as a co-operative effort -- maybe like the NASA-SpaceX relationship where (I believe) that SpaceX is gaining from NASA's expertise while the SpaceX systems are being 'vetted' -- or as an adversarial or neutral process?  Co-operative would certainly open possibilities for improved systems and launch flow, while adversarial/neutral would be opportunity lost.

[ncb1397]

a.) If there was anything SpaceX wanted to keep secret, you wouldn't necessarily know about it. Their engine seems to be really inexpensive with modern manufacturing techniques(mainly CNC) given that you can do 10 of them on a 56 million dollar mission. Tom Mueller famously built a medium sized engine in his garage before working on Merlin. IMO the engine details would be potentially a place where a competitor would want to replicate technology in order to potentially realize any cost savings and gain competitive advantage. Merlin has also gone through multiple design iterations over many many years which, if you wanted to gain the ability, design information would allow for skipping intermediate steps and going directly to the end-point.

c.)USG doesn't gaurd against USAF personnnel going to work for AJR and taking their brain with lots of information with them. Turn-over at SpaceX is one avenue of SpaceX trade secrets leaking. Design information residing at the Air force as well is simply another avenue that runs in parallel.

a.  Not true.  If SX wants to be certified, there will be no stone left unturned

c.  Yes, it does.  And it is more than USAF, it is Aerospace Corp and the many support contractors, who have to sign disclosure statements.

And the USAF isn't all the issue, Falcon 9 has been going under NASA cert since the Jason-3 contract award.  NASA and it's contractors have been looking at F9 a lot longer.

a.) I meant you or I personally won't know because we aren't reviewing the design. Or if you do have access to the design, the fact that you state there is nothing "advanced"(like this is some sort of desireable goal) just illustrates how information leaks. I think a simple engine that works would be more desireable than a complex engine as they both do the same thing. Note that the chinese use a locally copied version of the AK-47 called the Type 56 Assault Rifle and not an M-16 derived design.

c.) NDAs don't stop former USAF employees working for SpaceX competitors from using information personally retained during certification in their internal calculus.

[Llian Rhydderch]

Semantics. It's a government process. It would not be the first time a government process hampered ___X___  (where X is any number of desirable things)

Wrong, it is not semantics.  Nobody is forcing SX.   SX does not have to be certified and it doesn't have to fly USAF missions.

Wrong.  The government has the monopsopy on the military market.  So it is a state regulated market.  Noone can enter the market unless the comply with the (wise or unwise, good or bad, ...) regulatory process to compete in that market. 

That was my point.

And those regulations will tend to both slow innovation, and increase cost to anyone who wants to sell into that market.

[oldAtlas_Eguy]

Yes, the process has a tendancy to freeze vehicle configuration if your only market was the government. But SpaceX has other non-government customers that new configurations can fly and be used to then certify them to the governemnt.  Eventually the certification process will be just a paper exercise without meaning as vehicle inovation increases basic reliabilty way beyond that being sought by the AF.

Other pradigram inovations in operations may also greatly change the landscape.  Use of an on-orbit final checkout and repair facility in LEO and possibly even in GEO will make the "shippment" of payloads in common containers to orbit the way its done.  No more launch services just shippment contracts.

An aside its not the AF personnel doing the certification engineering review but Aerospace Corporation.  I found that Aerospace Corporation was very resistive to anything not done before as a official stance but individuals inside AC were vere progressive but they work in a represive corporate culture environment.

The really sad part is that inovative engineering, processes, and manufacturing have only experts inside the entity that is doing the new.  AC is giving "expert" opinions to the AF about these for which it has only a novice level of understanding.

[Coastal Ron]

a.) If there was anything SpaceX wanted to keep secret, you wouldn't necessarily know about it. Their engine seems to be really inexpensive with modern manufacturing techniques(mainly CNC)...

Just wanted to point out that CNC has been around for decades (1st generation CNC used paper tape), so advocating that it is "modern" is a misnomer.  Even friction-stir welding (FSW) is no longer "new", just yet another manufacturing technique that is now proven and well understood.

For the Merlin I would say that it is a number of factors that contribute to the overall cost advantage it brings.  The first is the design (i.e. pintle type), including it's size and lack of overall complexity when compared to much larger engines.  That then requires parts that are smaller in size and easier to manufacture in-house with affordable machinery.  SpaceX is supposedly also using 3-D manufactured parts, which again would only be possible for smaller designs at this point.

So overall it's really because SpaceX decided to use a cluster of smaller engines, instead of one large engine, that drove their ability to reduce the $/lb of thrust per launcher.

And not much of that is a secret.  Merlin type pintle engines were perfected decades ago, and SpaceX has not likely created any new manufacturing techniques.

Likely the only way someone can duplicate what SpaceX is doing is to essentially start from scratch, which means building a new Falcon 9 class rocket - and everyone else in the world has already locked in their design choices for their current generation of launchers, so until someone decides to build a brand new rocket we are unlikely to see someone else try to duplicate what SpaceX has done.

My $0.02

[rcoppola]

The certification process may not halt further innovations or iterative improvements. Post certification, the AF will have an intimate understanding of every aspect from design, production and operations. And in so having, they will be in an ideal position to understand and approve further system improvements or ask that the improvements be flown on another non-AF mission first to validate.

Now, the real concern becomes, if SpaceX will need to lock down the current Falcon 9v1.1 while rolling in enhancements to another line used to support fully commercial launches.

I am fully aware and in agreement with many of the concerns posted, however, the AF is not a typical USG, civil service bureaucracy. Many of these people have state-of-the-art training and are wickedly smart. We may be very surprised how the introduction of SpaceX into this process will begin to alter both themselves and the AF into an amazing partnership where each learns from the other. It's going to take some time for this to happen.

It's all about risk tolerance, of which the AF has little wrt their payloads and I don't blame them one bit. Personally, I think a mash-up of SpaceX and the AF will be great. They can both learn something from the other and they will.