totm nov 2023 SpaceX Discussion Thread

[RCgothic]

 

 

[tater]

 

[tater]

 

 

[magnemoe]

8 hours ago, tater said:

Pretty sure a "tank test" != launch vehicle/spacecraft test.

Agree, sounds much more like test of thinner hulls, This might has the impact that they has to use internal stiffing to keep SS structural sound then not pressurized. 
The usual with pressure and cryo testing. 
Now the weird one was lots of suborbital flights, probably to land SS accurate enough to get caught by the arms. 
Only 5 orbital flight a year however sounds very low with their construction tempo. 

[Beccab]

1 minute ago, magnemoe said:

Agree, sounds much more like test of thinner hulls, This might has the impact that they has to use internal stiffing to keep SS structural sound then not pressurized. 
The usual with pressure and cryo testing. 
Now the weird one was lots of suborbital flights, probably to land SS accurate enough to get caught by the arms. 
Only 5 orbital flight a year however sounds very low with their construction tempo. 

Thanks to how it is structured they can actually get 8 flights for a year when they pass from development phase to the operational phase, plus I'd be surprised if that number doesn't increase by the end of 2022

[wumpus]

17 hours ago, mikegarrison said:

I can't understand anybody arguing this.

I can tell you that we pressure test every single airplane before we let it go up for a first flight. And if 10% failed, not only would heads be rolling, but we probably wouldn't be in business anymore.

I'd assume that they test tanks  before putting them in in rockets.  Blowing up 10% of their rockets would be odd, but once they committed to mass producing rockets this became an option.  Since nobody else is mass producing rockets (including the Falcon 9 line), this would be a disaster anywhere else.

Anyone who can get Intel's 10nm process (allegedly in production since 2018) to 10% failure will  immediately be made a senior Intel fellow (it was likely 90% in 2018).  I'd be impressed if the old workhorse 14nm++++++++ process fails less often than 10%, certainly TSMC's competing 7nm process isn't 10%.  And those are shipped as Intel's main product.

Not that I could imagine a company surviving building 11 aircraft for every 10 it shipped, but in rocketry I'd have to expect that most of the costs are NRE.  This certainly helps Soyuz (all development costs are long since paid).  Blowing up 10% of rockets is hardly unknown, although typically that happens in flight.  Spacex lost the first three Falcon [1] flights, than had 21 successful flights (for the primary payload, if not the secondary).  Then they blew up another spacecraft, making sure that they wouldn't get below 10% failure rate before Falcon 9 launch 40 or so.  If 10% of your aircraft failed, not only would heads roll over the costs, I'd assume that it would be a safety issue worse than the 737MAX fiasco.  With all the flights everyday, sooner or later you'd see the same thing with crewed flights.  With new rockets, failure is to be expected and if you *aren't* losing a few you probably have far too much margin.  Crewed flight is different, and also some markets have unbelievably expensive cargo.  Hint, hint ESA, don't blow up Weber.

[tater]

Rockets ARE tanks. Testing "tanks" to me sounds like test articles, vs testing spacecraft/boosters.

There is no way they plan on losing 10% of the vehicles they cryo-proof, static fire, etc. They have lost exactly none in that way since what SN?

[Beccab]

11 minutes ago, tater said:

Rockets ARE tanks. Testing "tanks" to me sounds like test articles, vs testing spacecraft/boosters.

There is no way they plan on losing 10% of the vehicles they cryo-proof, static fire, etc. They have lost exactly none in that way since what SN?

SN7.1, but even that was tested to destruction proposefully

[tater]

 

[MKI]

I think the "10% Starships will fail" concept is another "Elon Engineering philosophy".

What SpaceX would like to achieve is "pushing the limits where 10% fail", since there are plenty of limits to push, this means producing 11 crafts for every 10 missions. Its somewhat an extension of the "re-add 10% back into the design, because you un-designed too much". Instead, SpaceX would like to meet current expectations, (whatever the current goals are), except force at least 10% to break/fail to know where the boundaries are with the current system. Iterate, and continue to "push boundaries".

 

Elon's background is in software, where there is a similar concept (at least in its idea) of Chaos Engineering, where you essentially try to break your system so you can test, and thus iterate against its weak points. Essentially you break your system expectantly, so its more prepared for the unexcepted. I'm sure Elon is familiar with the concept, and might be trying to apply it to Starship.

This sort of goal is actually a genius way to solve the problem of getting to Mars. The factory should be optimized to keep pumping out Starships, even once they are stable with the current system, they will continue to iterate in the direction they need, along with iterating into failures to continue to push boundaries. Again, this is only possible if you can iterate fast enough, which so far SpaceX seems to be doing just fine. Once you get that scale of production, you can use them to continue to learn more about the current system by essentially throwing them at the problem, and potentially throwing the problem at them to seek out failure points, and margins.

 

When doing something hard, making failure cheap, and being able to continue to iterate solutions can make iterating rather fast. SpaceX seems to be going this exact route, to tackle easily the hardest engineering problem available.  

 

 

 

 

[JoeSchmuckatelli]

6 minutes ago, MKI said:

I think the "10% Starships will fail" concept is another "Elon Engineering philosophy".

What SpaceX would like to achieve is "pushing the limits where 10% fail", since there are plenty of limits to push, this means producing 11 crafts for every 10 missions. Its somewhat an extension of the "re-add 10% back into the design, because you un-designed too much". Instead, SpaceX would like to meet current expectations, (whatever the current goals are), except force at least 10% to break/fail to know where the boundaries are with the current system. Iterate, and continue to "push boundaries".

 

Elon's background is in software, where there is a similar concept (at least in its idea) of Chaos Engineering, where you essentially try to break your system so you can test, and thus iterate against its weak points. Essentially you break your system expectantly, so its more prepared for the unexcepted. I'm sure Elon is familiar with the concept, and might be trying to apply it to Starship.

This sort of goal is actually a genius way to solve the problem of getting to Mars. The factory should be optimized to keep pumping out Starships, even once they are stable with the current system, they will continue to iterate in the direction they need, along with iterating into failures to continue to push boundaries. Again, this is only possible if you can iterate fast enough, which so far SpaceX seems to be doing just fine. Once you get that scale of production, you can use them to continue to learn more about the current system by essentially throwing them at the problem, and potentially throwing the problem at them to seek out failure points, and margins.

 

When doing something hard, making failure cheap, and being able to continue to iterate solutions can make iterating rather fast. SpaceX seems to be going this exact route, to tackle easily the hardest engineering problem available.  

 

 

 

 

I very often kind of nod and then don't say anything when Elon / SpaceX talks about getting people to Mars. 

You know what I mean?  Like when the person you haven't seen since highschool meets you at a party and within 30 seconds wants to tell you about this fabulous thing they heard from their Guru/therapist/priest?  (or a ten-year-old say's she wants to be President or an NFL Quarterback when she grows up... like - I support the aspiration, but don't want to burst the bubble of hope/optimism).

 

Then, here, I am reminded that they really are serious about Mars.

 

Which - is kind of crazy.

 

 

                                                           ...until you see how they are iterating Starship.

So - in line with what you write above, if Elon / SpaceX are capable of pushing boundaries as you say - in a SUSTAINED way; continuing past the success of Starship and a reusable BFR...   

 

 

 

 

 

 

 

 

 

...it's not crazy to think they might just pull it off

[tater]

live in 10 min

 

Obviously splashdown is not for a while.

[insert_name]

deobit burn finished, closing nosecone now

[CatastrophicFailure]

[Minmus Taster]

We have a visual!

Two drag chutes are out!

Main chutes are out!

SLPASHDOWN  

Welcome back Inspiration4!

[tater]

 

[tater]

 

 

They were at $60M, so they now have >$110M (not counting Isaacson's $100M).

[tater]

 

 

 

(the capsule is ~3.7m dia, so that's more like 1m, but good enough)

[JoeSchmuckatelli]

4 hours ago, tater said:

 

Has anyone decoded the chutes for sekret messages?

[tater]

Wonder if SpaceX could make an LSS crew module test article to leave in LEO... then use it as a station for commercial Dragon flights. Retire LSS risk by testing ECLSS/power/thermals, and have place where commercial Dragons could go. Maybe it has 2 ports, 1 has a vehicle with >4 seats (5-6), but they only  send 1-2 SpaceX crew with it. Commercial vehicles dock with 4 tourists/whatever they are called.

[Beccab]

1 hour ago, tater said:

Wonder if SpaceX could make an LSS crew module test article to leave in LEO... then use it as a station for commercial Dragon flights. Retire LSS risk by testing ECLSS/power/thermals, and have place where commercial Dragons could go. Maybe it has 2 ports, 1 has a vehicle with >4 seats (5-6), but they only  send 1-2 SpaceX crew with it. Commercial vehicles dock with 4 tourists/whatever they are called.

Following the Shuttle Enterprise station plans, they could remove the heat shield and add lots of stuff in the trunk

[magnemoe]

1 hour ago, tater said:

Wonder if SpaceX could make an LSS crew module test article to leave in LEO... then use it as a station for commercial Dragon flights. Retire LSS risk by testing ECLSS/power/thermals, and have place where commercial Dragons could go. Maybe it has 2 ports, 1 has a vehicle with >4 seats (5-6), but they only  send 1-2 SpaceX crew with it. Commercial vehicles dock with 4 tourists/whatever they are called.

Now this was an pretty genial idea. Make it the first version, interior designed for zero-g. Use the planned nose docking port and an side one as shown in the interior blueprints of manned starship. 
Already thought about using an second moonship as an lunar gateway space station. This could be used an an fuel depot and potential as an rescue ship if the one landed is not able to reach orbit again but has an working life support.
 

[tater]

They need to test ECLSS at some point on the larger SS scale, might as well chuck it into LEO.

Maybe the life support can intentionally be modular to try out different concepts.

[Beccab]

Going back to Starship development now that inspiration 4 is finished, the GSE4 derived mini tank has now an even smaller white tank attached to it:

Most of the catching arm as also been aligned:


Raptor RC 67 has also been lowered from Booster 4 yesterday, and a raptor without plumbing has been seen being transported behind the SpaceX dragon splashdown stream

Also interesting look inside the scrapped SN17 aft section

Edited September 19, 2021 by Beccab

[Serenity]

Super happy everything went well and they are back home.

I was wondering is the splashdown supposed to be this way? It felt a bit of a rough one but maybe its like that.