totm nov 2023 SpaceX Discussion Thread

[Listy]

7 hours ago, tater said:

67m/s

It's a grim comparison, but AF447 fell belly first into the Atlantic with a roughly similar vertical speed - about 55m/s...

[RCgothic]

3 hours ago, tater said:

Scrub was from current speed:

 

Hang on, it was JRTI on station for this mission. How much more upgraded can the thrusters get?!

[RealKerbal3x]

Just now, RCgothic said:

Hang on, it was JRTI on station for this mission. How much more upgraded can the thrusters get?!

Next upgrade will have new marine-optimised Raptors.

[StrandedonEarth]

4 hours ago, RCgothic said:

Hang on, it was JRTI on station for this mission. How much more upgraded can the thrusters get?!

Plaid Mode upgrade

[sevenperforce]

18 hours ago, tater said:

Wasn't one of the official animations showing like 60-something? Still pretty low.

42-43 m/s was with all flaps fully extended, which probably isn't even a stable configuration.

18 hours ago, cubinator said:

That's on Earth, right?

Because the flaps have a much higher drag coefficient than the cylindrical body, they represent 46% of the total drag even though they are only 27% of the total cross-sectional area. Since terminal velocity is inversely proportional to the square root of surface area, that means folding the flaps all the way back could increase terminal velocity to 62 m/s. Also, that animation starts at 3K feet, where the air density is 10% lower, further decreasing drag.

Finally, my mass estimate is based on the production Starship, which could be much lower than these prototypes.

[magnemoe]

9 hours ago, RCgothic said:

Hang on, it was JRTI on station for this mission. How much more upgraded can the thrusters get?!

You have thrusters who keep drilling ships at position at very rough sea states, granted they use propellers to counter the main direction but. 
Limit is power plant on board in practice. Or how large of an gas turbine they bother to install to power them. 
It also make the thing harder to tow but you could fold them up. 

[cubinator]

4 hours ago, sevenperforce said:

22 hours ago, cubinator said:

That's on Earth, right?

Because the flaps have a much higher drag coefficient than the cylindrical body, they represent 46% of the total drag even though they are only 27% of the total cross-sectional area. Since terminal velocity is inversely proportional to the square root of surface area, that means folding the flaps all the way back could increase terminal velocity to 62 m/s. Also, that animation starts at 3K feet, where the air density is 10% lower, further decreasing drag.

................on Earth?

[sevenperforce]

26 minutes ago, cubinator said:

................on Earth?

Yes.

[Meecrob]

Haha, an astute observation Elon! Seriously though, I can't wait to see what they come up with.

Edited September 22, 2020 by Meecrob

[mikegarrison]

Landing gear is a major component of most airplanes. It has to take all the landing force, so it must be very strong. Retractable gear also has to fold up and fit in as small of a volume as possible. And then there are doors that must open and close. Brakes and steering mechanisms (although Starship won't need those).

In most airplane PD cycles, landing gear are sized and located very early on in the process, because it is hard to fit it in later. Looks like SpaceX is learning this lesson now.

Edited September 22, 2020 by mikegarrison

[tater]

4 minutes ago, mikegarrison said:

In most airplane PD cycles, landing gear are sized and located very early on in the process, because it is hard to fit it in later. Looks like SpaceX is learning this lesson now.

Yeah, their initial design had the 3 large fins with gear at the tips, Tin-tin style, they obviously had to rethink everything after that change (always seemed like a lot of forces on those hinges, though, and the flaps not working on return from Mars is a LOC level problem).

[mikegarrison]

3 minutes ago, tater said:

Yeah, their initial design had the 3 large fins with gear at the tips, Tin-tin style, they obviously had to rethink everything after that change (always seemed like a lot of forces on those hinges, though, and the flaps not working on return from Mars is a LOC level problem).

I forgot about those tripod fins.

This is part of the danger of their build it, break it, redesign it strategy. In their rush to go fast, they may have jumped the gun too much.

The usual PD cycle is you start by making major decisions that size the craft (and include stuff like where the landing gear will be). Then you start getting more detailed, filling things in, assessing the performance, tweaking, etc. Usually you find out that something just can't be made to work, so you cycle back to a new clean sheet of paper and start again. Only after cycling through this over and over do you get comfortable enough to start the detailed design.

I don't know if SpaceX follows that model, though. They seem to be trying to mix some elements of detail design and flight testing in while still working issues that would normally be considered early PD. They run a risk of doing a lot of detail design that they have to just throw away later, but maybe they feel like they are learning enough from it to make it worthwhile anyway.

[magnemoe]

1 hour ago, mikegarrison said:

I forgot about those tripod fins.

This is part of the danger of their build it, break it, redesign it strategy. In their rush to go fast, they may have jumped the gun too much.

The usual PD cycle is you start by making major decisions that size the craft (and include stuff like where the landing gear will be). Then you start getting more detailed, filling things in, assessing the performance, tweaking, etc. Usually you find out that something just can't be made to work, so you cycle back to a new clean sheet of paper and start again. Only after cycling through this over and over do you get comfortable enough to start the detailed design.

I don't know if SpaceX follows that model, though. They seem to be trying to mix some elements of detail design and flight testing in while still working issues that would normally be considered early PD. They run a risk of doing a lot of detail design that they have to just throw away later, but maybe they feel like they are learning enough from it to make it worthwhile anyway.

Problem with the tintin design was in part the need for bracing the top for the fins inside the tank and the fins needed to be much stronger to handle landings. Also center of drag would be forward of center of mass. 
Love the design but think current one is much lighter even with larger legs as all forces are on the skirt and trust structure who need to handle an fully loaded starship anyway. 
And we has less experience with landing legs for spaceship than we had with retractable landing gear back in 1930. 
The legs they will use on tankers will probably not be the same used landing on Mars or the Moon. 

[kerbiloid]

Why does Space-X keep blundering around in the dark, instead of just referring to my post to clearly understand their future design evolution?

Is it so hard?

 

[mikegarrison]

2 hours ago, magnemoe said:

Problem with the tintin design was in part the need for bracing the top for the fins inside the tank and the fins needed to be much stronger to handle landings. Also center of drag would be forward of center of mass.

You may have missed my point, I'm not sure. I'm not saying they should have stuck with that design. I'm saying that they may have progressed their design work so far now that it will be really tough to fit in better landing gear. Perhaps if they hadn't been in a rush to build hardware, they might have noticed this problem earlier. What if, for instance, there is a way to fit all the gear they want into a 9.5M body. Are they too committed now to a 9M body to have that option on the table?

This kind of thing happens, and it's always a trade between never committing to go forward versus committing to go forward too soon and locking yourself into trouble. SpaceX has pushed very hard to go through the various design gates as quickly as possible, trading a risk there for accelerated testing. But with something like landing gear, that is usually the sort of thing sized very early in the design process, they could find themselves trapped between having to scrap a huge amount of the design work they have already done or facing a very, very difficult design problem and possibly less-than-ideal landing gear.

Then again, maybe it will all turn out fine. I'm not saying it won't. But I think this gear issue is more of a threat than many people realize.

Edited September 22, 2020 by mikegarrison

[tater]

Interesting. Looks slightly swept leeward as the default to me (assuming left is belly).

[sevenperforce]

11 hours ago, kerbiloid said:

Why does Space-X keep blundering around in the dark, instead of just referring to my post to clearly understand their future design evolution?

Is it so hard?

They won't do this.

9 hours ago, mikegarrison said:

Then again, maybe it will all turn out fine. I'm not saying it won't. But I think this gear issue is more of a threat than many people realize.

I agree. It's really nontrivial and the amount of back-and-forth they've done suggests it's still unsolved.

1 hour ago, tater said:

Interesting. Looks slightly swept leeward as the default to me (assuming left is belly).

Yep, I agree.

God, I'm excited.

[tater]

A reasonable take on the flap geometry:

 

[sevenperforce]

27 minutes ago, tater said:

A reasonable take on the flap geometry:

 

It looks like it has a reasonable degree of passive stability.

My suspicion is that with the gimbal kick from those engines it can perform the tail flip even if the aft fins aren't fully folded. Just might take a little more thrust than normal.

[cubinator]

7 minutes ago, sevenperforce said:

My suspicion is that with the gimbal kick from those engines it can perform the tail flip even if the aft fins aren't fully folded. Just might take a little more thrust than normal.

If that's the engine configuration, you could even light just the bottom two there during the flip for a little extra torque.

[zolotiyeruki]

I have a question:  Assuming SS is passively stable in the roll axis, why would you need to have actuated rear fins at all?  You could control pitch and yaw with just the canards.  The best system being no system, and all that...

[sevenperforce]

47 minutes ago, cubinator said:

If that's the engine configuration, you could even light just the bottom two there during the flip for a little extra torque.

I don't think you'd want to mess with engine start transients during the flip recovery and terminal descent.

41 minutes ago, zolotiyeruki said:

I have a question:  Assuming SS is passively stable in the roll axis, why would you need to have actuated rear fins at all?  You could control pitch and yaw with just the canards.  The best system being no system, and all that...

I've thought about this a lot, actually, and I think it's a potential failure mode hedge. If you have only two actuated control surfaces and one of them locks up, you're screwed. If you have four, then you can lose one and compensate with the other three.

It also allows broader control range for different entries. Remember that the ship which goes to Mars needs to be able to return to Earth. The Earth return is also going to be the highest-energy entry, so you want the most data on it. Therefore you want the Mars return to use the same EDL mode as your basic ship so you get the most data. That also means your Mars EDL needs to be incorporated in the same system.

EDIT: Moreover, if you only have two forward canards, yaw is strongly coupled with roll and so you will have very little yaw authority. Honestly, yaw is challenging to control even with four actuated surfaces.

Edited September 22, 2020 by sevenperforce
Added thoughts

[RealKerbal3x]

56 minutes ago, sevenperforce said:

Mars EDL

The thing that I - and I'm sure many others - find the most exciting about this is the fact that this is actually being developed with Mars landing capability as an important design consideration. It's not some design study or technology demonstrator that may or may not lead to an operational vehicle. They genuinely want to land it on Mars, and as far as I can tell there's no other vehicle, in development or otherwise, that is intended to have this capability.

SpaceX is breaking new ground here. I remain convinced that when humans land on Mars (whenever that may be), it'll be aboard a Starship.

[sevenperforce]

A little look at how having all four flaps allow for multiplanar control authority:

Spoiler

NOTE: I actually have the yaw angles pointed in the wrong direction; I misidentified the image.

If you were to only have the forward fins actuated and have the aft fins fixed, then yaw and roll would be coupled: you couldn't yaw in one direction without rolling in that direction too. 

[zolotiyeruki]

1 hour ago, sevenperforce said:

I don't think you'd want to mess with engine start transients during the flip recovery and terminal descent.

I've thought about this a lot, actually, and I think it's a potential failure mode hedge. If you have only two actuated control surfaces and one of them locks up, you're screwed. If you have four, then you can lose one and compensate with the other three.

It also allows broader control range for different entries. Remember that the ship which goes to Mars needs to be able to return to Earth. The Earth return is also going to be the highest-energy entry, so you want the most data on it. Therefore you want the Mars return to use the same EDL mode as your basic ship so you get the most data. That also means your Mars EDL needs to be incorporated in the same system.

EDIT: Moreover, if you only have two forward canards, yaw is strongly coupled with roll and so you will have very little yaw authority. Honestly, yaw is challenging to control even with four actuated surfaces.

That's a fair point about the yaw/roll coupling, although I wonder how much roll you'd actually get, given the much larger surface area of the (hypothetically fixed) rear fins.