Quote from: Kabloona on 01/12/2016 12:42 amQuoteBecause apart from the engines, it's a similar structure, similar fuel, similar size... It can't be THAT different. I'd be more inclined to agree if you were talking about *mass* ratios, but you're not.You're talking about *cost* ratios of completely different components on a completely different rocket. How is the cost of one RD-180 related to the cost of nine M1D's? Not to mention different production processes for tanks, etc.Apples and oranges.Both are fruit... Some things are similar.If the engines are $15M, what costs $45M?The tanks? The Atlas cost breakdown doesn't apply directly, but it gives you an idea...
QuoteBecause apart from the engines, it's a similar structure, similar fuel, similar size... It can't be THAT different. I'd be more inclined to agree if you were talking about *mass* ratios, but you're not.You're talking about *cost* ratios of completely different components on a completely different rocket. How is the cost of one RD-180 related to the cost of nine M1D's? Not to mention different production processes for tanks, etc.Apples and oranges.
Because apart from the engines, it's a similar structure, similar fuel, similar size... It can't be THAT different.
Quote from: meekGee on 01/12/2016 05:59 amQuote from: Kabloona on 01/12/2016 12:42 amQuoteBecause apart from the engines, it's a similar structure, similar fuel, similar size... It can't be THAT different. I'd be more inclined to agree if you were talking about *mass* ratios, but you're not.You're talking about *cost* ratios of completely different components on a completely different rocket. How is the cost of one RD-180 related to the cost of nine M1D's? Not to mention different production processes for tanks, etc.Apples and oranges.Both are fruit... Some things are similar.If the engines are $15M, what costs $45M?The tanks? The Atlas cost breakdown doesn't apply directly, but it gives you an idea...The answer is probably along the lines of what JamesH wrote above. The $1.5M engine cost is probably an "unburdened" cost number, ie only *direct* labor and materials to build that one item, without any overhead.But the total $60M vehicle cost figure Elon refers to has to be a partially or fully burdened number including overhead of facilities, support and admin staff, etc, etc.So if you assume a 70/30 cost split between S1 and S2, say it's $42M partially or fully burdened cost for S1, a lot of which is overhead. Subtract maybe $12M overhead and you're left with $30M direct labor and materials costs, half of which is for 9 engines, and then you get splits closer to ULA's pie chart, with F9's making up 50% of S1 *direct* costs.Or, put another way, if you burdened that F9 engine proportionally with overhead, I'm betting the number would be more than $1.5M and the proportions would tilt towards ULA's numbers.
This means that no matter how many uses occur even if it is only 2, the result is a price reduction over that of a non-reusable version of an F9.
Quote from: Kabloona on 01/12/2016 12:27 amQuote from: meekGee on 01/11/2016 09:55 pmWell if the engines are ~$1.5M, and if the Atlas slide is correct, than the F9 does not cost $60M. So something's gotta give.It's an Atlas slide. Why would you assume F9 has identical cost breakdowns? It's a completely different vehicle.Because apart from the engines, it's a similar structure, similar fuel, similar size... It can't be THAT different. If the Merlins are 1.5 Million, then the engine stack is 15M. So the structures would have to be some 30M to get the entire rocket to 60M. That's very far from the Atlas breakdown... 2:1 instead of 1:2....
Quote from: meekGee on 01/11/2016 09:55 pmWell if the engines are ~$1.5M, and if the Atlas slide is correct, than the F9 does not cost $60M. So something's gotta give.It's an Atlas slide. Why would you assume F9 has identical cost breakdowns? It's a completely different vehicle.
Well if the engines are ~$1.5M, and if the Atlas slide is correct, than the F9 does not cost $60M. So something's gotta give.
Quote from: meekGee on 01/12/2016 12:31 amQuote from: Kabloona on 01/12/2016 12:27 amQuote from: meekGee on 01/11/2016 09:55 pmWell if the engines are ~$1.5M, and if the Atlas slide is correct, than the F9 does not cost $60M. So something's gotta give.It's an Atlas slide. Why would you assume F9 has identical cost breakdowns? It's a completely different vehicle.Because apart from the engines, it's a similar structure, similar fuel, similar size... It can't be THAT different. If the Merlins are 1.5 Million, then the engine stack is 15M. So the structures would have to be some 30M to get the entire rocket to 60M. That's very far from the Atlas breakdown... 2:1 instead of 1:2....Yes, it can be very different. For one simple example that jumps out from the chart, I expect the cost of avionics is very different, and seems to be a good chunk of ULA's costs, by the pie chart shown. ULA uses typical radiation resistant aerospace components and systems; SpaceX uses pretty COTS components for its systems and software, getting reliability by redundancy in the systems (many CPU's, etc), running much more modern software. Given the cost of those modern components (which are far faster than anything you can get "radiation hardened"), I don't understand how to spend the kind of money that ULA current has to spend, despite having to have that redundancy; modern semiconductor technology and Moore's law being what it has been. SpaceX gave a talk at a Linux conference a few years ago describing (to the extent that the export control laws allow) what they were doing in their systems.Sometimes, by revisiting systems from scratch, you can get to a very different (better) point that following the path of "incremental improvement". Clean sheets can be a real advantage; but you don't get them very often. SpaceX did a "clean sheet" on most everything.So don't presume the actual costs in the different rockets are "the same", unless you are comparing fundamental costs (e.g. propellants, or N kilograms of aluminum alloy)... And one of Musk's strengths is that he thinks terms of those fundamental costs, rather than what people have traditionally paid....
The answer is probably along the lines of what JamesH wrote above. The $1.5M engine cost is probably an "unburdened" cost number, ie only *direct* labor and materials to build that one item, without any overhead.But the total $60M vehicle cost figure Elon refers to has to be a partially or fully burdened number including overhead of facilities, support and admin staff, etc, etc.So if you assume a 70/30 cost split between S1 and S2, say it's $42M partially or fully burdened cost for S1, a lot of which is overhead. Subtract maybe $12M overhead and you're left with $30M direct labor and materials costs, half of which is for 9 engines, and then you get splits closer to ULA's pie chart, with F9's making up 50% of S1 *direct* costs.Or, put another way, if you burdened that F9 engine proportionally with overhead, I'm betting the number would be more than $1.5M and the proportions would tilt towards ULA's numbers.
ULA uses typical radiation resistant aerospace components and systems; SpaceX uses pretty COTS components for its systems and software, getting reliability by redundancy in the systems (many CPU's, etc), running much more modern software. Given the cost of those modern components (which are far faster than anything you can get "radiation hardened"), I don't understand how to spend the kind of money that ULA current has to spend, despite having to have that redundancy; modern semiconductor technology and Moore's law being what it has been. SpaceX gave a talk at a Linux conference a few years ago describing (to the extent that the export control laws allow) what they were doing in their systems.
This will all get much more serious for vehicles going to Mars, all of which will leave even the limited protection of the Earth's magnetic field. SX may have to go the full rad hard route for that, but that's probably 10 years ahead. [EDIT Which is not to say that someone within SX is not already working the issues to see what they could get away with.
They may have to, but I imagine the first thing they'll try is additional redundancy. But I agree that someone within SpaceX will be working this, though I suspect the FH will be throwing stuff to Mars well within 10 years.
Considering SpaceX just hot fired the recovered stage at SLC-40 this speaks to the cost of refurbishment for a stage to be almost nil on the order of less than $2M toward an amount of $.5M.What does that do to the reuse price of a F9/FH?
Presumably, whatever knowledge SpaceX does acquire in terms of what components erode fastest and require the most frequent monitoring and replacement, will be kept a trade secret by them, since this would be the key to maintaining their 'first mover advantage'. So it seems doubtful that they would reveal/publish/comment on this stuff much, when it becomes part of the bread-and-butter of their business model.But perhaps we'll be able to glean information from any upcoming design changes which may result from their knowledge gained from reusable flights.
NET February 6 - SES-9 - Falcon 9 FT - Canaveral SLC-40 (or late NET January 23)NET February 7 March 20 - Dragon SpX-8 (CRS8), BEAM (Bigelow Expandable Activity Module) - Falcon 9 FT - Canaveral SLC-40 - 04:20 21:01 1st quarter NET February March (TBD) - Eutelsat 117 West B (Satmex 9), ABS 2A - Falcon 9 FT - Canaveral SLC-40NET 1st quarter February March - JCSat-14 - Falcon 9 FT - Canaveral SLC-40NET February March - AMOS 6 - Falcon 9 FT - Canaveral SLC-40 (or midyear)March 21 NET April - Dragon SpX-9 (CRS9) - Falcon 9 FT - Canaveral SLC-40 ~04:00spring April-May - FORMOSAT 5, SHERPA SSO: Arkyd-6, CNUSail 1, KAUSAT 5, SIGMA, CANYVAL-X 1, CANYVAL-X 2, STEP Cube [/qoute]
Later Friday, SpaceX CEO Elon Musk reported on Twitter that one of the outer Merlin engines had not performed perfectly.“Conducted hold-down firing of returned Falcon rocket,” said Musk. “Data looks good overall, but engine 9 showed thrust fluctuations.”Musk said inspections would quickly investigate if debris had gotten into the engine.