totm nov 2023 SpaceX Discussion Thread

[mikegarrison]

1 hour ago, sevenperforce said:

If reusing a stage costs $1M and saves them the ~$18M it costs to build a new stage, and they can reuse stages ten times each, then it would take 66 launches at full commercial price to recover the $1B it cost to develop Falcon 9 first-stage reuse. That's a long time.

Instead they're reinvesting that capability by selling themselves Starlink launches.

Bingo. More than half of their reflights are going to Starlink (since the F9B5 debut, there have been 26 operational Starlink launches and 22 commercial launches on reflown boosters). So that's more like 80-90 commercial launches to recoup the dev costs. And even if they make that, their investment in upper stages and launch costs for Starlink still has to be recovered. 

Yeah, that's what I'm saying.

It costs a lot of money to develop a new launch vehicle, and even more money to develop a whole new technology like booster reuse. So you need a big enough market for it to close the business case.

In the case of the Falcon 9, they are providing their own market, which funds closing their own business case quite nicely. But that only works if Starlink actually makes money. Otherwise they just move the Falcon 9 dev costs from SpaceX to Starlink.

[Codraroll]

3 hours ago, SpaceFace545 said:

Please, in what timeline will rocket engines get down to the price of a cheap house.

Given how the housing prices seem to evolve nowadays, it wouldn't surprise me if this would be true in a few years even if the cost of rocket engines goes up.

[mikegarrison]

3 hours ago, RealKerbal3x said:

When they're no longer bespoke, single-use machines that have to be built, transported and installed using highly specialised tooling and vehicles.

Raptor is already eschewing a lot of that - it's the first full-flow staged combustion cycle engine to leave the test stand and yet it's transported on a flatbed and installed with a scissor lift.

Just so it's clear -- a 737/A320 commercial airplane jet engine costs between $5M-$10M, as far as I know. These are not "bespoke, single-use machines" -- they come off of a factory line and are installed/replaced interchangeably.

Being "single use" does not make an engine more expensive. It does make it more expensive PER USE, but making it last thousands of hours instead of 10 minutes actually makes the engine much more expensive to design and build. So Raptors should not be expected to be cheaper because they are reusable. Rather, that should be expected to make their initial costs either a little or a lot HIGHER, rather than lower.

Cruise missile engines are about the cheapest jet engines made, because they only have to last one flight (although they have to be super reliable for working the first try after sitting around unused for years, which has its own issues).

Elon is claiming that they are going to be making Raptor engines for 5% of the cost of an A320 engine. I don't know if this is really possible or not, but it sure sounds ... questionable.

[SOXBLOX]

5 minutes ago, mikegarrison said:

Raptor engines for 5% of the cost of an A320 engine. 

Just out of curiosity, which engine do you think is more difficult to manufacture? Raptor or the jet engine?

[JoeSchmuckatelli]

1 hour ago, sevenperforce said:

And even if they make that, their investment in upper stages and launch costs for Starlink still has to be recovered. 

Given how much I pay for internet... 

...times a whole bunch of people around the world...

...

Three weeks?

[sevenperforce]

1 hour ago, tater said:

I think I have seen that number before, is it from Musk or Shotwell?

Do we think that number is accurate? Seems like the actual cost of hardware for reuse was fairly low, and the landings were tested on flights that were completely paid for at retail.

That was what Musk said at the presser immediately after SES-10, the first successful reflight. A lot of his numbers are off-the-cuff but if he said a billion I doubt it is less.

Hardware costs for individual boosters are minimal, yes, and most of the test landings were incidental to commercial flights. But it took a long time to get there. SpaceX had to build facilities at McGregor to allow flight tests there rather than just stand tests. They had to dev and build and operate Grasshopper for eight flights. They had to dev and build and operate F9RDev1 for five flights. They build a Dev2 vehicle and leased space at Las Cruces and built a pad there, but canceled it. They had to do all the modeling and write all the code for hypersonic retropropulsion. They had to buy the barges and retrofit them as droneships and design all the software for autonomous station-keeping. They had to pay the boat people and rent the chase planes. They had to research the legs and design the legs and build the legs and destructively test the legs. They had to come up with the grid fin control hardware and write the software for it. All of that work by people who were getting paid by SpaceX and who weren't contributing to commercial activities.

I'm sure it's more than a billion.

39 minutes ago, zolotiyeruki said:

Even if the Merlin engine wasn't originally designed for relight, I'd be shocked if its design hasn't evolved to better facilitate relight.

It's a good thing chute recovery never really took flight (pun intended). Now, Merlin 1D *is* designed to be test-fired and re-test-fired and static fired before launch, like a car engine that can be restarted over and over and over again as long as you change the oil periodically. Not only does that make recovery by propulsive landing possible, but it makes rapid reflight possible. Nothing like the Shuttle program. The Orbiter never brought the RS-25s back; it merely brought the RS-25 parts back.

Most non-hypergolic liquid-fueled rocket engines designed through the history of spaceflight could not be readily relit. The J-2 and the RD-58 were some of the first, I believe. I can't remember whether the RL10 could originally be relit or not. They certainly weren't relit when they flew on the original Saturn I S-IV.

7 minutes ago, SOXBLOX said:

Just out of curiosity, which engine do you think is more difficult to manufacture? Raptor or the jet engine?

I know you weren't asking me, but because I often answer when I'm not asked...I think the Raptor engine was harder to design and I think the jet engine is harder to manufacture.

Finding a new way to do something that has never been done before is hard.

[Terwin]

6 minutes ago, mikegarrison said:

Just so it's clear -- a 737/A320 commercial airplane jet engine costs between $5M-$10M, as far as I know. These are not "bespoke, single-use machines" -- they come off of a factory line and are installed/replaced interchangeably.

Being "single use" does not make an engine more expensive. It does make it more expensive PER USE, but making it last thousands of hours instead of 10 minutes actually makes the engine much more expensive to design and build. So Raptors should not be expected to be cheaper because they are reusable. Rather, that should be expected to make their initial costs either a little or a lot HIGHER, rather than lower.

[...]

Elon is claiming that they are going to be making Raptor engines for 5% of the cost of an A320 engine. I don't know if this is really possible or not, but it sure sounds ... questionable.

Building engines in higher quantity will generally reduce the per-unit cost, be it a singe or multiple use engine.

A commercial jet engine is deigned for high reliability, low maintenance, and a high rate of use(possibly even more run-time than non-running time: running for multiple hours then refueled and re-started after a quick maintenance check). 

Orbital rocket engines burn for tens of minutes(if that) then coast for hours and may or may not burn for a few more minutes.

SH *might* get below a 1:5 burn:idle ratio (10 min burning up, 5 min burning down out of a 25 minute flight and < 50 minute time on the pad before the next launch might be a best-case scenario for re-use)

Also, for an airliner with 2 engines, 'engine out' capability is not something you want to be relying on, but with >25 engines on SH, even one failed engine per launch is not critical, just annoying.

(and long-run-time reliability is very expensive) 

[SpaceFace545]

1 hour ago, mikegarrison said:

Elon is claiming that they are going to be making Raptor engines for 5% of the cost of an A320 engine. I don't know if this is really possible or not, but it sure sounds ... questionable.

I think this is just his marketing, 1000 starships a year, 1000 starship Mars fleets, earth to earth, tunnels that will revolutionize traffic, brain implants to cure autism, etc.

[magnemoe]

52 minutes ago, mikegarrison said:

Just so it's clear -- a 737/A320 commercial airplane jet engine costs between $5M-$10M, as far as I know. These are not "bespoke, single-use machines" -- they come off of a factory line and are installed/replaced interchangeably.

Being "single use" does not make an engine more expensive. It does make it more expensive PER USE, but making it last thousands of hours instead of 10 minutes actually makes the engine much more expensive to design and build. So Raptors should not be expected to be cheaper because they are reusable. Rather, that should be expected to make their initial costs either a little or a lot HIGHER, rather than lower.

Cruise missile engines are about the cheapest jet engines made, because they only have to last one flight (although they have to be super reliable for working the first try after sitting around unused for years, which has its own issues).

Elon is claiming that they are going to be making Raptor engines for 5% of the cost of an A320 engine. I don't know if this is really possible or not, but it sure sounds ... questionable.

Cruise missile engines don't have to be super reliable either, storing an jet engine for decades is not an very hard problem and you assume some missiles will miss or get shot down anyway. 
Jet engines on passenger planes on the other hand has to be idiotic reliable. You save if you go from 3-4 engines to 2, however if one fail the other had to take the load. 
Not all are like the brave B-52 pilots facing 7 engine emergency landings. 


 

[CatastrophicFailure]

2 hours ago, sevenperforce said:

Wait, are they going to relaunch it after all??!!

Scuttlebutt I heard was that it’s just for inspections, for now. What they find will prolly determine if it flies again or not. Or if that’s just the easier way to get it back off the launch stand. 

2 hours ago, sevenperforce said:

Tory also wished my seven-year-old a happy birthday yesterday and did so with cool facts.

Of course you know, you’re now obligated to legally change his name to “xxx, Son of @TheseJustWords.” 

2 hours ago, sevenperforce said:

Elon may send Doge to the moon with his tweets but Tory's Twitter game is way better.

That’s easier to do when every single tweet isn’t deluged with thousands of fanboy quips and fake spambots. 

[Starman4308]

In terms of manufacture costs, I suspect high-performance rocket engines may be more expensive to manufacture than jet engines for a few reasons:

1) Jet engines are giant high-performance turbines, whereas rocket engines are basically large high-performance turbines with some other bits attached. There is likely some additional expense from the larger size of a jet engine turbine, but both are very complicated pieces of high-performance machinery.

2) Jet engines have only the fuel to worry about, whereas rocket engines have to have plumbing for both fuel and oxidizer. This often also means either gearing or a separate turbine is required for the oxidizer pump, plus more complicated injector design.

3) Jet engines typically rely on air cooling, whereas rocket engines require either ablative or regenerative cooling, the latter of which involves a lot of extra plumbing.

4) There is simply more institutional expertise and tooling for making jet engines. That might change, but there's a lot more jet engines being built than orbital-class rocket engines.

5) The margins for rocket engines are even more extreme than jet engines, as the cost of each additional kilogram is steeper.

[tater]

2 hours ago, mikegarrison said:

Elon is claiming that they are going to be making Raptor engines for 5% of the cost of an A320 engine. I don't know if this is really possible or not, but it sure sounds ... questionable.

What's the total life in hours running (they can't use seconds, lol) of a commercial jet engine?

The engines on a SpaceX booster get used about 3 minutes per flight. Maybe 2-3X that for the upper stage?

Assuming jet engines must last for 10s of thousands of hours... that's 600,000 minutes per 10,000 hours. Assume just 10k hours for fun. 600k minutes is 200k booster flights, maybe 60k SS flights? If the Raptor is 20X cheaper, maybe it's only good for 100X fewer flights? That's still 6k booster flights, or 600 SS flights. 1000X less durability for 20X less cost? 600 booster vs 60 SS flights?

Made up numbers, obviously, but the total burn time is certainly orders of magnitude lower, so there must be a trade off there someplace (where I have no idea).

Edited May 11, 2021 by tater

[mikegarrison]

Really I responded to this whole issue to address the (IMO) nonsensical claim that the Raptor will be super cheap because it is reusable. That is only going to make it more expensive per engine, even if it makes it less expensive per flight. But Musk's claim was not that Raptors would cost $250K/flight, he was claiming they would cost $250K period.

I will point out, however, that we don't know what kind of cost he is talking about. The actual cost to produce an engine, if separated from the cost to develop it, is much less. When you sell an engine to a buyer, you have to price it according to your entire program costs. Musk may have been strictly referring to the marginal cost of making one engine, without including any other program costs. (And if so, it's not "apples-to-apples" with reported costs for other engines.)

[tater]

15 minutes ago, mikegarrison said:

Really I responded to this whole issue to address the (IMO) nonsensical claim that the Raptor will be super cheap because it is reusable. That is only going to make it more expensive per engine, even if it makes it less expensive per flight. But Musk's claim was not that Raptors would cost $250K/flight, he was claiming they would cost $250K period.

I will point out, however, that we don't know what kind of cost he is talking about. The actual cost to produce an engine, if separated from the cost to develop it, is much less. When you sell an engine to a buyer, you have to price it according to your entire program costs. Musk may have been strictly referring to the marginal cost of making one engine, without including any other program costs. (And if so, it's not "apples-to-apples" with reported costs for other engines.)

I assumed from the way he said that they would continue to reduce costs with mass production that he means the cost per engine, excluding dev. They are not selling the engines, so it's not like they have to recoup the dev costs there. He's talking in "crazy Mars colonizing" terms, I think—1000 Starships tanked by some thousands of tankers, etc.

 

You are certainly right that reuse should make them more, not less expensive, I suppose the issue is internal vs external use. RS-25s were sold for ~$40M at the time, I assume they made money on those. If Raptors were for sale we'd obviously have a better sense of their cost. Still, reuse on rockets is a very different time regime than jet engines that are expected to be pretty much flawless for thousands and thousands of hours of continuous use, vs a few minutes, followed by nothing for days, weeks, months, or years, then a few more minutes.

Edited May 12, 2021 by tater

[Starhelperdude]

6 hours ago, RealKerbal3x said:

 

maybe it's time for everyone else to start building engines that are less delicate and bespoke and more ruggedised. 

I watched a video of russian engines beeing made, it looked like a normal factory (no offense, I like russian engines and the asthetic of the factory is cool)

--------------------

it's interesting how SpaceX never made a hydrolox engine and how Blue Origin never made a Kerolox engine

[SpaceFace545]

1 hour ago, Starhelperdude said:

I watched a video of russian engines beeing made, it looked like a normal factory (no offense, I like russian engines and the asthetic of the factory is cool)

--------------------

it's interesting how SpaceX never made a hydrolox engine and how Blue Origin never made a Kerolox engine

At least they take more care when transporting their engines then with goosenecks and scissor lifts.

[tater]

16 minutes ago, SpaceFace545 said:

At least they take more care when transporting their engines then with goosenecks and scissor lifts.

Has SpaceX dropped a lot of engines? You should have video, there are cameras pointed at the site 24/7.

They're just engines, they're supposed to be able to deal with the vibrations of 30 engines right next to them, after all.

16 minutes ago, SpaceFace545 said:

At least they take more care when transporting their engines then with goosenecks and scissor lifts.

Has SpaceX dropped a lot of engines? You should have video, there are cameras pointed at the site 24/7.

They're just engines, they're supposed to be able to deal with the vibrations of 30 engines right next to them, after all.

[.50calBMG]

Yeah, transporting engines on a truck out in the open is such a terrible idea. Clearly nobody would be crazy enough to do that. It would make those engines unreliable and bad.

Spoiler

Certainly a scissor lift would be the worst place for a rocket engine. Nobody would ever do that.

Spoiler

Seriously, why is this a bad thing? The engines are robust enough to be transported like a regular piece of equipment and still work. If you can treat something as complicated as a rocket engine the same way you would treat an engine going into a truck and it still works, it just proves that the engine can work outside of a lab test environment. If you are depending on an engine to land on another planet after multiple burns, vibrations from adjacent engines, and the stress of reentry, I would hope it can take a ding from safety chain or some dust.

 

Spoiler

Also, as someone who lives in a ~$90k house, since when is $250k cheap?

 

Edited May 12, 2021 by .50calBMG

[FleshJeb]

2 hours ago, .50calBMG said:

Also, as someone who lives in a ~$90k house, since when is $250k cheap?

Median home price in my area is $700k+

I could actually afford to buy a $250k rocket engine and live in the bell. The HVAC system might be a bit over-specced though.

[Xd the great]

2 hours ago, FleshJeb said:

I could actually afford to buy a $250k rocket engine and live in the bell. The HVAC system might be a bit over-specced though.

I wonder if they could land starship at areas affected by hurricanes and earthquakes, unload the supplies, then use the cargo space to provide temporary living quarters...

[GuessingEveryDay]

2 hours ago, Xd the great said:

I wonder if they could land starship at areas affected by hurricanes and earthquakes, unload the supplies, then use the cargo space to provide temporary living quarters...

Well, first they would actually require everyone to evacuate. Because no one wants to get cooked by a Starship. Then assuming the area affected by the natural disaster has a population of 30,000. And Elon plans for the Starship to have a capacity of 1,000 humans. That would mean 30 Starships for sleeping area. Those 30 starships would have to carry an assortment of beds, food, and water. Along with a few elevators from Lunar Starship. But sure, that would be a cool idea. Just have to get past all the regulations, and the town. It would be easier to just launch the town's population up into space until the hurricane passes. Not only do they get to avoid the hurricane, but they also get to go into space!

[DDE]

2 hours ago, Xd the great said:

I wonder if they could land starship at areas affected by hurricanes and earthquakes, unload the supplies, then use the cargo space to provide temporary living quarters...

Reminds me of the UR-100 maritime rescue ICBM proposal.

Hammers, nails, et cetera.

Edited May 12, 2021 by DDE

[kerbiloid]

A trailer park for Starship habitats.

[tater]

 

[Pastapower]

2 hours ago, GuessingEveryDay said:

Well, first they would actually require everyone to evacuate. Because no one wants to get cooked by a Starship. Then assuming the area affected by the natural disaster has a population of 30,000. And Elon plans for the Starship to have a capacity of 1,000 humans. That would mean 30 Starships for sleeping area. Those 30 starships would have to carry an assortment of beds, food, and water. Along with a few elevators from Lunar Starship. But sure, that would be a cool idea. Just have to get past all the regulations, and the town. It would be easier to just launch the town's population up into space until the hurricane passes. Not only do they get to avoid the hurricane, but they also get to go into space!

I find it hard to believe that 1000 people could fit in one starship!