also, I'm pretty sure that Shotwell said that the F9 figures (10.7 klb to GTO for example) include the fuel reserve for 1st stage boost back. Couple that with Musk's comments that the Merlin 1D can be pushed harder (> 10% I think), then it's probably not a conclusive exercise if only the published website figures are used.
Elon said today that certification required three flights of that distinct configuration.
Therefore, they'd need to fly three with those upgrades before they could propose those for any payload. FH might be available before then.
Depending when those upgrades become available, it even seems conceivable they might upgrade FH after the first couple of flights, then need to fly three more to get certified. Would seem to make sense for the initial FH flights to use the upgraded configs.
Cheers, Martin
This is specifically the kind of stuff that Musk is likely peeved about. ULA never had to qualify a rocket configuration on 3 flights before being able to use it for government launches.
"A Delta 4-Heavy, powered for the first time by higher-thrust RS-68A engines, successfully boosted the NROL-15 payload into orbit from Cape Canaveral on June 29, 2012."
http://www.spacelaunchreport.com/slr2012q2.htmlThere are Atlas configurations that have flown less than 3 times. 431 has flown twice, 521 twice and 541 twice. The first 541 launched a $2.5 billion NASA mission. Delta IV-M+(5,2) has only flown once with an NRO satellite as well.
Besides, pushing the Merlin-1D harder by 10%, if it was a hardware change, would likely warrant being termed a different variant(Merlin-1E). Likely, the engine is being restrained artificially for margin reasons(like some high performance cars are restrained for legal/safety reasons) and the higher thrust can be achieved through software once tested at that level on a test stand. Does SpaceX have to test any flight software patches orbitally 3 times before they can fly government missions again because technically the storage devices on the LV have microscopic switches in them that are in a different physical state?
Ya know, by any sufficiently strict definition, they haven't done that and couldn't without giving up continuous improvement.
Yup.
But any vehicle is going to evolve over its first three launches, while Musk seems to think their certification is in the bag. Implies they're not being unfair about it.
And v1.1 was a huge upgrade over 1.0, so fair to certificate that version.
Cheers, Martin
Besides, pushing the Merlin-1D harder by 10%, if it was a hardware change, would likely warrant being termed a different variant(Merlin-1E). Likely, the engine is being restrained artificially for margin reasons(like some high performance cars are restrained for legal/safety reasons) and the higher thrust can be achieved through software once tested at that level on a test stand. Does SpaceX have to test any flight software patches orbitally 3 times before they can fly government missions again because technically the storage devices on the LV have microscopic switches in them that are in a different physical state?
It's my suspicion that the 112% M1D will only fly with propellant densification, and this is sufficiently novel (no one has ever flown it?) that it might be considered a new unique config.
If nothing else, it could have the potential for more countdown aborts until they get the wrinkles ironed out.
Cheers, Martin
Depending when those upgrades become available, ...
I don't think any upgrades are necessary to increase the Merlin 1D performace the extra amount Musk mentioned. Extra qualification testing and internal analysis are probably all that is required, but I'm shooting from the hip here.
Verifying that extra performance capability sooner rather than later would probably be justifiable if they were competing for some of those flights in the block buy.
Besides, pushing the Merlin-1D harder by 10%, if it was a hardware change, would likely warrant being termed a different variant(Merlin-1E). Likely, the engine is being restrained artificially for margin reasons(like some high performance cars are restrained for legal/safety reasons) and the higher thrust can be achieved through software once tested at that level on a test stand. Does SpaceX have to test any flight software patches orbitally 3 times before they can fly government missions again because technically the storage devices on the LV have microscopic switches in them that are in a different physical state?
It's my suspicion that the 112% M1D will only fly with propellant densification, and this is sufficiently novel (no one has ever flown it?) that it might be considered a new unique config.
If nothing else, it could have the potential for more countdown aborts until they get the wrinkles ironed out.
Cheers, Martin
Engines are "flown" and "wrinkles ironed out" on the test stand.
Depending when those upgrades become available, ...
I don't think any upgrades are necessary to increase the Merlin 1D performace the extra amount Musk mentioned. Extra qualification testing and internal analysis are probably all that is required, but I'm shooting from the hip here.
Verifying that extra performance capability sooner rather than later would probably be justifiable if they were competing for some of those flights in the block buy.
Increasing thrust while not increasing vehicle mass will increase stresses at maxQ, while only delivering a relatively small payload increase.
Those increased stresses might be a concern. Agree it would be helpful to demonstrate early to avoid any doubt.
Cheers, Martin
This is specifically the kind of stuff that Musk is likely peeved about. ULA never had to qualify a rocket configuration on 3 flights before being able to use it for government launches.
Because the USAF was intimately involved with Atlas and Delta development. They know what processes went into the design and testing of the vehicles. And the USAF would design and testing standards as part of the procurement process. The USAF has never bought a COTS launch vehicle before. In the 90's, when NASA started buying launch services and start putting out procurement solicitations, they realized that anybody (like Joe's Rockets "R" Us) could respond. So that is where the concept of certification came about. That is where the 14, 3 and 1 flights come from with the different amounts of insight penetration.
I'm a bit skeptical as to whether a relative newcomer like SpaceX would be entrusted with something as precious as an X-37B (or a large spy satellite for that matter), at least not until they have many more successful flights behind them.
Pieces of a launch vehicle operated by an experienced space company that carried a precious spy satellite still litter a Vandenberg hillside as of yesterday. This same space company went on to lose at least three more tier 1 national security payloads on similar vehicles and still kept US government business. Number of successful past flights is not an all encompassing metric nor a predictor of future success.
Sure, but poor decisions made in the past do not justify continuing this practice. Now that there are options, a company that loses a few Tier-1 payloads should expect to lose their contract, and the reliability of new (and old) entrants should be taken into consideration.
Of course estimating the reliability of new entrants is controversial. I'd prefer some very simple and public rule, such as "we add to the cost of the rocket (1/N)*(cost of the payload), where N is the number of successful launches (if no failures) and (M/N)*(cost of payload) if there are M failures in N launches. Like the electoral college, this is not perfect statistically but at least everyone would know the rules in advance.
Can't find any Atlas 511 on the current manifest and only a 501 for the NRO
also, I'm pretty sure that Shotwell said that the F9 figures (10.7 klb to GTO for example) include the fuel reserve for 1st stage boost back. Couple that with Musk's comments that the Merlin 1D can be pushed harder (> 10% I think), then it's probably not a conclusive exercise if only the published website figures are used.
Okay, so testing....
These upgrades are not going to need all up flights to get certified. Only new vehicle changes, such as entirely new engines or a major change in total stage impulse, require new certifications. Increasing the engine thrust by 12% would necessitate a new engine qualification program at the engine level. That'd be 20-30 tests and 2000-3000 seconds of run time, testing throttle, mixture ratio and inlet condition extremes. Densified propellant would need a full or delta engine qual program. Increased vehicle loads would just be tested during standard structural testing.
Software almost never needs to be flight tested. It can be run by setting up integration labs using flight-representative hardware. Even if it's just a bench test, cables need to be flight length. Errors and out of spec flight conditions can be injected to see how the system reacts. And it can be done 100s or 1000s of times.
Separation tests are probably the hardest to execute on the ground.
It would be reasonable to require one flight, but the changes discussed here should not reset the clock.
Anyway, NASA and USAF have a definition of "common launch vehicle configuration" that generally guides when a change requires a new certification, though it's big enough for bureaucrats to drive a truck-sized interpretation through.
1 Atlas-V(501) Possible X-37, performance wise it is F9 compatible. Integration wise, unknown
Can speak on this one
In "Open bidding" Orbitals Antares would be a bit fit for this launch, and integration.
Jim, a general Security question regarding USAF payloads. What level of "security" do workers have to maintain with classified programs?
What's the security level on the X-37 Program?
k on this one
In "Open bidding" Orbitals Antares would be a bit fit for this launch, and integration.
No, it doesn't have a 5m fairing or a Cape launch pad.
Jim, a general Security question regarding USAF payloads. What level of "security" do workers have to maintain?
What's the security level on the X-37 Program?
X-37 is unknown. Most workers don't need a clearance since they don't have access to the payload. Also, most USAF payloads are visually unclassified and people without clearance can be escorted around them
the Delta IV (5,4) is likely a WGS, which would be outside the capability of the F9.
k on this one
In "Open bidding" Orbitals Antares would be a bit fit for this launch, and integration.
No, it doesn't have a 5m fairing or a Cape launch pad.
Agreed the location of the program @ the cape is a problem ....however with all the SpaceX launches set for the cape maybe the program will wish to move to wallops, never know.
The fairing is a standard one, its up to Orbital if they wish to adapt the launcher.
This is specifically the kind of stuff that Musk is likely peeved about. ULA never had to qualify a rocket configuration on 3 flights before being able to use it for government launches.
Because the USAF was intimately involved with Atlas and Delta development. They know what processes went into the design and testing of the vehicles. And the USAF would design and testing standards as part of the procurement process. The USAF has never bought a COTS launch vehicle before. In the 90's, when NASA started buying launch services and start putting out procurement solicitations, they realized that anybody (like Joe's Rockets "R" Us) could respond. So that is where the concept of certification came about. That is where the 14, 3 and 1 flights come from with the different amounts of insight penetration.
One of your best posts ever Jim.
ULA stated that the Air Force would be saving ~$4B by doing the block buy, but from a bulk procurement standpoint the cost curve for saving money could be on a far smaller number than 36 cores. For instance, for buying the aluminum for the rocket bodies and tanks the aluminum supplier may just have a price that says "if you buy over X amount, it's this price regardless of how much you buy".
It's wider than that. The ULA factory has a a certain "standard" level of production where
all costs are at a minimum. That's the number of cores/year where no one is idle, there are no left over alloy plates to store etc.
In theory if the USAF did block buys at
that level ULA's costs would be minimum and what they charged the USAF would (presumably) also be minimum. Staff would be fully used (without any over time) etc.
So the price the Air Force would pay/core for 18 cores might not be any different than for the 36 core buy, or within a few percentage points. That alone would make opening the contract for competition worthwhile.
True.
IRL
neither policy is fair. I believe there
is a block buy size where ULA makes out and the USAF gets a lower unit price. But 36 cores is
way too high unless the factory was designed for a
huge production volume (I mean 2 a month or more).
