totm nov 2023 SpaceX Discussion Thread

[Flavio hc16]

4 minutes ago, derega16 said:

to Elon:  read the last phrase on the 1st paragraph, again, SLOWLY   

 

[Ultimate Steve]

Drainage doors.

[Hannu2]

That rocket seem to have some kind of transparent nose. Maybe there is an elite compartment with maximum view and crazy high ticket prices. Those tourists must be brave if the rocket hits the tube at couple of hundreds meters per second.

[kerbiloid]

Musk is testing  the Transparent Pods mod on his note, so all crafts arre automatically updated.

Edited April 8, 2021 by kerbiloid

[Xd the great]

Anyone can give a reasonable estimate of how many times starship and superheavy can be re-used? I know Elon aims for 1000 times, but given how much thermal and other stress the airframe rocketframe undergoes, I wonder if they can re-use them 20, 50, 100 or 500 times. Or 5,000 times.

Edited April 8, 2021 by Xd the great

[RCgothic]

The first piece of SN20 has been spotted:

 

This being the first piece of the next major block of starship upgrades beyond the SN15-class of starships and potentially the first piece of a starship to orbit.

 

Edited April 8, 2021 by RCgothic

[tater]

21 minutes ago, Xd the great said:

Anyone can give a reasonable estimate of how many times starship and superheavy can be re-used? I know Elon aims for 1000 times, but given how much thermal and other stress the airframe rocketframe undergoes, I wonder if they can re-use them 20, 50, 100 or 500 times. Or 5,000 times.

A huge unknown.

The thing is that the math on reuse is very, very different if we are talking about the 2016 (was it really that long ago?) "ITS" BFR concepts, and the current, steel version (see the point above that they can apparently build a SS/SH tank structure literally in the same order of magnitude as the price of retail propane tanks).

That's the original SpaceX ITS graphic on cost.

$230M for ITS booster (now SH), $130M for cargo/tanker ITS, $200M for crew ITS. Note those are substantially larger than the current ones.

It's completely reasonable to guess current costs could be 10X cheaper than that, which is... insane. Crew vehicle  might be half the cost of the slide above currently (I'll assume the extra $70M is still there for fittings).

Note that 1000 SH launches means the per launch booster cost for ITS is $230,000, if SH is 10X cheaper, it gets to the same cost with 100 launches. Tanker/cargo the same with just 10, and a Starship would only need 5 trips to have the same per mission cost.

Edited April 8, 2021 by tater

[RealKerbal3x]

Yay heatshield!

 

[Flavio hc16]

19 minutes ago, RealKerbal3x said:

Yay heatshield!

 

it would be nice to  land it and inspect the heat tiles.....

[cubinator]

How long do we have to wait for a 100 km hop?

[tater]

1 hour ago, cubinator said:

How long do we have to wait for a 100 km hop?

I don't think that's a thing.

SS to 100km by itself might well be possible, but it needs more Raptors, and much more propellant that it takes for a hop. It would not really test anything that pertains to real EDL, not high enough velocity.

Musk has set the aspirational date of July to orbit. Even with some delays, that's pretty soon. Testing SH really requires load testing it. Like the hopper SS versions that actually flew with mass simulators, except SH needs a 1300t mass simulator... might as well be a SS.

Edited April 8, 2021 by tater

[JoeSchmuckatelli]

3 hours ago, Xd the great said:

Anyone can give a reasonable estimate of how many times starship... can be re-used

So far the answer is zero - unless they are recycling metal. 

 

Goal is likely to be on the order of what you see with Falcon (okay, their goal may be higher - but at this point I'd be more than happy to see even a second flight out of SN15...) 

[RealKerbal3x]

2 minutes ago, JoeSchmuckatelli said:

Goal is likely to be on the order of what you see with Falcon (okay, their goal may be higher - but at this point I'd be more than happy to see even a second flight out of SN15...) 

IMO we won't see any reflights until at least SN20 - supposedly the first orbital vehicle - and probably longer, given that for these early prototypes are pretty much obsolete as soon as they fly due to changes made to the next one.

[Meecrob]

16 hours ago, tater said:

You are missing the fundamental observation.

Again, if SpaceX could buy the tanks for MUCH less, they would, period. There is zero reason for SpaceX to build a $1,000,000 GSE tank if they can buy the same capability off the shelf for a small fraction of that. There's no "in house" reason to do it if it costs a lot more.

The take away is... it DOESN'T cost them a lot more. It might not cost them more, at all.

This gives us some sense of what a Starship tank section actually costs. Apparently it costs about the same as buying 4 off the shelf 100,000 gallon propane tanks retail, delivered.

That's incredibly inexpensive. Musk said they picked steel because it was cheap. He said in some quote or tweet that he thought it might be cheaper to build a Starship than it is to build a Falcon 9.

It's gonna have a minimum of what, 34 Raptors? Right now that might be 100M$(?), but he aims to get them to .25M$/ea. The steel part could literally end up costing almost nothing. Like a few million, even with flaps, etc. Launch costs not amortized over 1000 launches to get the price/flight down, but cheap even if expended (cargo here, crew vehicle... not even sure I believe it will be a thing).

You are missing the fundamental observation of logistics and integration. I get what you are saying. I am trying to say that everytime you outsource something it has errors. You are playing the game of "broken telephone". For rapid iteration as SpaceX is obviously doing, you need to be in control of..well everything. How much does it cost to delay your testing schedule because your supplier had to re-tool to meet your demands? It could very well be a NASA contract worth 8 figures.

You are talking millions of dollars, I'm talking hundreds of millions of dollars.

[tater]

6 minutes ago, Meecrob said:

You are missing the fundamental observation of logistics and integration. I get what you are saying. I am trying to say that everytime you outsource something it has errors. You are playing the game of "broken telephone". For rapid iteration as SpaceX is obviously doing, you need to be in control of..well everything. How much does it cost to delay your testing schedule because your supplier had to re-tool to meet your demands? It could very well be a NASA contract worth 8 figures.

You are talking millions of dollars, I'm talking hundreds of millions of dollars.

Storage tanks are storage tanks.

They launch to orbit with off the shelf storage tanks every few weeks.

Retooling is not a factor, the tanks required are available in stock.

[Meecrob]

Storage tanks are not storage tanks though. There are commercially available storage tanks for applications that exist right now. Starship is a different scale though. The evidence is they built a storage tank bigger than any tank they bought commercially.

What I am saying is really that this is a guy who went to Russia to buy rockets and they told him to get bent, so he said "ok, I'll build my own rocket." This is a strategic move, not a financial one.

[tater]

10 minutes ago, Meecrob said:

Storage tanks are not storage tanks though. There are commercially available storage tanks for applications that exist right now. Starship is a different scale though. The evidence is they built a storage tank bigger than any tank they bought commercially.

Every single single Starship test up until now has used off the shelf tanks in the tank farm.

If those tanks cost anything like what they look to cost—10s of thousands each—then buying more is a trivial expense.

How much do you think a Starship tank costs them to build? Do you think each costs $100M, and they will build a few $100M tanks instead of buying a dozen off the shelf with smaller capacity each because having 3 tanks is cooler than having over a dozen, even if it costs 1000X more?

 

10 minutes ago, Meecrob said:

What I am saying is really that this is a guy who went to Russia to buy rockets and they told him to get bent, so he said "ok, I'll build my own rocket." This is a strategic move, not a financial one.

You realize he went to buy the rocket, and would have—but the Russians kept increasing the price. He built his own rocket because he didn't like getting ripped off. It was a financial decision.

[mikegarrison]

5 hours ago, Xd the great said:

Anyone can give a reasonable estimate of how many times starship and superheavy can be re-used? I know Elon aims for 1000 times, but given how much thermal and other stress the airframe rocketframe undergoes, I wonder if they can re-use them 20, 50, 100 or 500 times. Or 5,000 times.

So far the record is 1 time.

5 hours ago, tater said:

It's completely reasonable to guess current costs could be 10X cheaper than that, which is... insane.

Well yes, it is insane, but not for the reason you suggest.

Costs never get cheaper between the PD concept and the final product. Usually what happens is that the weight goes up, the costs go up, and the mission performance gets squeezed.

Edited April 8, 2021 by mikegarrison

[tseitsei89]

5 minutes ago, mikegarrison said:

 

Costs never get cheaper between the PD concept and the final product.

It does often happen though, You will polish, streamline, automate and upscale the production so it gets cheaper as they perfect the process

36 minutes ago, Meecrob said:

Storage tanks are not storage tanks though. There are commercially available storage tanks for applications that exist right now. Starship is a different scale though. The evidence is they built a storage tank bigger than any tank they bought commercially.

What I am saying is really that this is a guy who went to Russia to buy rockets and they told him to get bent, so he said "ok, I'll build my own rocket." This is a strategic move, not a financial one.

Storage tank is a storage tank in this case. It is just needed to store liquid oxygen/methane, nothing more fancy than that. They built a bigger tank because they just had existing production line for starship sized tanks and used that to build the tank. They could just as well use several of those smaller tanks like they have done until now.

[mikegarrison]

12 minutes ago, tseitsei89 said:

It does often happen though, You will polish, streamline, automate and upscale the production so it gets cheaper as they perfect the process

Yes, costs come back down after the initial prototype. Sometimes. But that's not what I said or what @tater was comparing against. The numbers he was showing were powerpoint slide PD numbers.

To put it into context, didn't those same ITS slide decks show SpaceX already landing people on Mars by now?

(OK, no, I guess they showed SpaceX sending out crews in the next Mars transfer window, so a few years from now.) 

 

Edited April 8, 2021 by mikegarrison

[tater]

1 minute ago, mikegarrison said:

Yes, costs come back down after the initial prototype. Sometimes. But that's not what I said or what @tater was comparing against. The numbers he was showing were powerpoint slide PR numbers.

Absolutely, it was also a carbon fiber composite cost estimate, and presumably in some notional "Mars fleet" level of production (which I will believe when I see, lol).

The cost of just the tanks in steel vs carbon could easily be far, far cheaper, however. Total vehicle cost? <shrug>

It's not impossible that it's substantially cheaper than their powerpoint estimate, however (particularly for the booster).

The number I used for SS/SH was based on a lower limit on cost, which seems to be "cheaper to build than off the shelf GSE tanks cost."

As I said, I think their home built GSE could in fact be more expensive than that, just not vastly more. Can you envision them building a tank for 10s of millions (more?) if they can buy the same capacity for a couple hundred grand?

If the SN# vehicles only cost a few hundred grand each (not counting engines!), that helps explain why they can be so cavalier about blowing them up, or even scrapping vehicles already largely built (at least sub assemblies built).

The build, test, blow up, repeat is hard to grasp if costs are like the costs for other rockets, but much easier to wrap one's head around if it's actually not that expensive.

 

[StrandedonEarth]

Pure conjecture here, but there may be another reason for the 9m GSE tanks: to reduce the physical footprint  

I don’t know how much prepared land they have, or what kind of foundation those 100 kilogallon tanks need, but a vertical 9m 400 kgallon would have a much smaller footprint than four skinny 100 kgallon tanks, not to mention plumbing them altogether would be simpler (not a big issue, I’m sure, but still). Sure, Starbase is surrounded by empty land, but it needs to be preloaded to support any significant weight, and that takes time (a year at least, IIRC)  

So if they’re running low on preloaded land, they may have gone this route to save space. Commercial suppliers would have charged through the nose for custom tanks of that size, and probably have to retool. Then there’s shipping (or barging) on top. So while it may have been cheaper to buy enough 100kgal tanks, it’s almost certainly cheaper to build their own 400kgal tanks, and they take up less space!

Sure, they had to buy stock tanks to get going while the factory was getting set up and proofed. But now that they can build their own, why not?

Again, pure conjecture about a possible reason, even if not directly cheaper then COTS. Definitely cheaper than preloading more land...

And remember, commercial suppliers expect to make a profit unless dumping excess stock. SpaceX does not need to profit on tanks. 

[tater]

11 minutes ago, StrandedonEarth said:

SpaceX does not need to profit on tanks. 

No, they build them at cost.

Still, I doubt they cost terribly much relative to the costs we're used to seeing for rocket tanks.

SLS spent $6.7B to make 2 flight article core stages, and one for destructive testing

 

 

[zolotiyeruki]

Is there a corrosion issue with storing LOx in a mild steel tank?

[magnemoe]

4 hours ago, tater said:

I don't think that's a thing.

SS to 100km by itself might well be possible, but it needs more Raptors, and much more propellant that it takes for a hop. It would not really test anything that pertains to real EDL, not high enough velocity.

Musk has set the aspirational date of July to orbit. Even with some delays, that's pretty soon. Testing SH really requires load testing it. Like the hopper SS versions that actually flew with mass simulators, except SH needs a 1300t mass simulator... might as well be a SS.

An real suborbital jump could give valuable data on heating, And they need to test superheavy under load. so just lob SS up have if fly for some time, the do an burn to reverse its trajectory and land. 
This could be important, an major issue is that starship will have to overfly southern Texas to land back, I find this harder than the jumps they do now and something who will have junk falling everywhere. Make me wonder if they don't want to land the first ones some other place.