totm nov 2023 SpaceX Discussion Thread

[tater]

Cross range probably resulted in the larger wings, not the basic shape.

It did have the added benefit of creating more abort from orbit options, but the once around from VAFB was probably the driver. The circumference of the Earth at VAFB latitude is ~24.8k miles, which means in an hour the Earth moves ~1000 mi East, and in a Shuttle orbital period, more like 1500 mi. The idea is that the USAF wanted the ability to fly basically not even 1 orbit, drop a payload, or use a camera in the bay, etc, then immediately land—except that Shuttle would then be over the Pacific. Least that's what I remember from back in the day, there are people here that know way more about Shuttle than me, though.

[RCgothic]

Well in any case, the reason re-entry bodies are blunt - to move the shockwave away from the hull - is also accomplished by being larger. 

1) Starship is larger.

2) There's no evidence cylinders don't work - in fact spheres do (Vostok and Voskhod) which are worse.

3) SpaceX aren't amateurs. They know what they're doing.

[tater]

[CatastrophicFailure]

SpaceX’s modeling of the whole belly-flop thing thus far appears to be spot-on: the SNs have been rock solid and 100% under control the whole way down, it’s reasonable to assume their models for the re-entry process, having lots of data from others to glean from, will be equally accurate. 

It’s the whole “never done before” flip thing at the very end that’s proving a bit... squirrelly...

[RealKerbal3x]

49 minutes ago, CatastrophicFailure said:

It’s the whole “never done before” flip thing at the very end that’s proving a bit... squirrelly...

The flip isn't really a problem. We've seen that as long as at least two engines are able to light, they're able to smoothly transition to vertical powered descent.

That's been the issue though, getting those engines restarted, running healthily, and correctly drawing propellant from the header tanks - all the way down to the ground. Hopefully the changes made on SN15 and onwards get those problems licked.

Edited April 20, 2021 by RealKerbal3x

[CatastrophicFailure]

1 hour ago, RealKerbal3x said:

That's been the issue though, getting those engines restarted, running healthily, and correctly drawing propellant from the header tanks - all the way down to the ground.

That being the problem. While orbital re-entry is a pretty well understood thing, this kind of flippy-sloshy slightly insane aerobatics is not, and that makes it very difficult to model. SpaceX is having to figure it out as they go along because there isn’t data to go on, or not very much that’s relevant. 

[RCgothic]

And it's the fluid dynamics of the fuel system that's tricky.

Not the aerodynamics.

[tater]

 

[JoeSchmuckatelli]

8 hours ago, kerbiloid said:

Anything known which is gliding (i.e. not falling like a bomb) and not flat below?

From Mike's thread, the flat bottom is to take advantage of the lifting body design, and land aerodynamically - right?  Shuttle lands on wheels and all that?  (presumes the graphic you provided lands on the skids, too) - and in that profile you need lift 

 

My understanding is that SX wants to slow down to effectively terminal velocity with broadside entry and then use the rockets for a powered descent for the last leg - where aerodynamic surfaces would not be able to assist. 

My question is a bit different - having looked at wind loads and turbulence on rectangular vs cylindrical buildings - the cylinders have the least turbulence (and, hence drag??) of the two.  If the concept extends to a reentry vehicle - Won't that make it more difficult to slow? 

[tater]

25 minutes ago, JoeSchmuckatelli said:

My question is a bit different - having looked at wind loads and turbulence on rectangular vs cylindrical buildings - the cylinders have the least turbulence (and, hence drag??) of the two.  If the concept extends to a reentry vehicle - Won't that make it more difficult to slow? 

We've all seen SS terminal velocity already, that's part of why it went to the alt it did in the tests so far.

On entry from space? <shrug> They have pretty good models of that at this point, they have modeled and flown 2 different capsules that reenter from LEO (3 if you count Cargo Dragon), they have flown many second stages that all have telemetry (and reenter destructively), and they have been designing SS (or versions of it under other names) for years. I would wager they have a decent understanding of what reentry looks like.

Edited April 21, 2021 by tater

[Starhelperdude]

lol what do they do with starhopper now? do they use it for static fire tests or as a fuel tank? xD

[OrdinaryKerman]

8 minutes ago, Starhelperdude said:

lol what do they do with starhopper now? do they use it for static fire tests or as a fuel tank? xD

It seems to be a monument and weather station now

[cubinator]

47 minutes ago, Starhelperdude said:

lol what do they do with starhopper now? do they use it for static fire tests or as a fuel tank? xD

Weather station, camera mount, protecting spirit...

[StrandedonEarth]

2 hours ago, JoeSchmuckatelli said:

My question is a bit different - having looked at wind loads and turbulence on rectangular vs cylindrical buildings - the cylinders have the least turbulence (and, hence drag??) of the two.  If the concept extends to a reentry vehicle - Won't that make it more difficult to slow? 

Any tweaking the shape (cylinder vs ????) for a slightly lower terminal velocity would not be worth losing the simplicity of cranking out cylindrical shapes. Anything else adds complexity/cost/mass/time

[Silavite]

3 hours ago, JoeSchmuckatelli said:

From Mike's thread, the flat bottom is to take advantage of the lifting body design, and land aerodynamically - right?  Shuttle lands on wheels and all that?  (presumes the graphic you provided lands on the skids, too) - and in that profile you need lift 

 

My understanding is that SX wants to slow down to effectively terminal velocity with broadside entry and then use the rockets for a powered descent for the last leg - where aerodynamic surfaces would not be able to assist. 

My question is a bit different - having looked at wind loads and turbulence on rectangular vs cylindrical buildings - the cylinders have the least turbulence (and, hence drag??) of the two.  If the concept extends to a reentry vehicle - Won't that make it more difficult to slow? 

I believe that the cylinder would have less drag, but Starship's mass/area at reentry is more or less comparable with the STS. Some quick and dirty drawing with MS Paint gives an area of 370 m² of the STS' underside, and a very crude estimate of its reentry mass (from subtracting LEO payload from MTOW on Wikipedia) would be 86 tonnes, which gives a mass/area of 232 kg/m². I think that the projected dry mass for Starship right now is 130(?) tonnes, let's say 150 tonnes at reentry. Starship is 9 m in diameter and 50 m in length, but the nose cone has taper, so lets say that the projected area is equivalent to a 9-by-43 m rectangle. This gives 387 m². The control surfaces will add to this, so a fudge factor of 1.1 give a final area of 425.7 m². Starship's (very crudely) estimated mass/area ratio is thus 388 kg/m². This is higher than the STS, but they're in the same ballpark. Besides, STS was designed with the ability to retrieve a satellite, so I believe that its actual mass/area could go up to something like ~280 kg/m².

Somewhat offsetting this disadvantage is the fact that Starship has the ability to maintain flight at a higher angle of attack, thus the component of area seen by the flow is larger. (That said, while we all know that the AoA tends to 90 deg for the terminal descent, I'm unsure what AoA Starship is supposed to reenter at.)

Edited April 21, 2021 by Silavite

[kerbiloid]

Lifting body is not just for terminal velocity (all of them have it ~150 m/s at last). 

Radial acceleration and heat distribution.

[mikegarrison]

To use aerodynamics for control you need to be moving fast. Planes achieve this by moving horizontally, but they need a runway (or arrestor cables). Helicopters achieve this by moving just their wings (the rotors) fast. Starship (or Falcon) basically has no aerodynamic control at touchdown, so it relies entirely on thrust vectoring.

[RyanRising]

2 hours ago, Silavite said:

Somewhat offsetting this disadvantage is the fact that Starship has the ability to maintain flight at a higher angle of attack, thus the component of area seen by the flow is larger. (That said, while we all know that the AoA tends to 90 deg for the terminal descent, I'm unsure what AoA Starship is supposed to reenter at.)

This was for a return from Mars, not sure if that changes anything about the angle you fly it at, but we’ve had an approximate number on that before. 
“... & hypersonic angle of attack is ~70 degrees” -Elon Musk

But then again, seeing as that’s a return from Mars your heatshield is going to need to be significantly better than the Shuttle’s, almost certainly beyond the point where increasing AoA gets you to.

[mikegarrison]

5 hours ago, JoeSchmuckatelli said:

My question is a bit different - having looked at wind loads and turbulence on rectangular vs cylindrical buildings - the cylinders have the least turbulence (and, hence drag??) of the two.  If the concept extends to a reentry vehicle - Won't that make it more difficult to slow? 

It's not all that simple. One thing to understand is that sharp edges are very important in fluid dynamics. Sometimes you want them (trailing edges of wings) and sometimes you don't (leading edges of wings).

Re-entry is more about controlling heating and deceleration than anything else. You don't *want* to slow down too quickly, because that is too much g-load for your structure (and people, if crewed). But you don't want to slow down too slowly either, because you don't want to sit there for a long time in a superheated plasma oven.

[Hannu2]

20 hours ago, kerbiloid said:

Any known aerobraked cylinder?

No. Any known spacecraft comparable to Starship in any way? No. Any significant advances in manned spaceflight in last 40 years? Very little.

Sometimes you must invent new things if you want to go further.

 

20 hours ago, kerbiloid said:

And I believe that aerodynamics hasn't changed since 1960s.

Physics is the same but technology is much more developed. Abilities to simulate things, plan optimal structures, new materials and fabrication methods etc. Many severe restrictions in 60s are simple problems now.

But this is the thing we will very probably see in relatively short future. I think this summer is too aspirational schedule but in next summer we may see how cylindrical Starship handles hypersonic reentry.

If I remember correctly there was debate like this before first Falcon 9 landings. Some prediction from 60's was that rocket engine can not work if you try to brake at supersonic speed. I do not know was it theoretical prediction or based on some tests with actual engines.

[kerbiloid]

26 minutes ago, Hannu2 said:

Any known spacecraft comparable to Starship in any way?

The biggest known are Shuttle and Buran. Both are winged lifting bodies with flat bottoms.

26 minutes ago, Hannu2 said:

No. Any significant advances in manned spaceflight in last 40 years? Very little.

Any significant changes in aerodynamics or metallurgy in last 60 years?

26 minutes ago, Hannu2 said:

Sometimes you must invent new things if you want to go further.

Yes, and what new is invented in the idea, old like rockets, "let the rocket just reenter".

Why did they have to put thr 2nd stage engines of Shuttle into the orbiter, rather than deorbit them together with the whole tank as a full-featured stage?

26 minutes ago, Hannu2 said:

Physics is the same but technology is much more developed. Abilities to simulate things, plan optimal structures, new materials and fabrication methods etc. Many severe restrictions in 60s are simple problems now.

Yes. And where is a test deorbit of, say, Falcon 2nd stage?
Without landing, just to aerobrake.

26 minutes ago, Hannu2 said:

But this is the thing we will very probably see in relatively short future.

Or won't, if the Starship structure is not enough tough.

26 minutes ago, Hannu2 said:

If I remember correctly there was debate like this before first Falcon 9 landings.

Falcon 9 re-enters at almost plane speed, and the discussion was (and is) mostly about its actual (not the declared) price/cost ratio, which is still unknown.
Nothing even close to 8 km/s.

 

Edited April 21, 2021 by kerbiloid

[tater]

1 hour ago, kerbiloid said:

Yes. And where is a test deorbit of, say, Falcon 2nd stage?
Without landing, just to aerobrake.

I bet they have data on most reentries. Just because we don't see it doesn't mean they don't have it.

1 hour ago, kerbiloid said:

Falcon 9 re-enters at almost plane speed, and the discussion was (and is) mostly about its actual (not the declared) price/cost ratio, which is still unknown.
Nothing even close to 8 km/s.

Yeah, but they have decent models.

The Shuttle was designed with far less ability to model anything accurately, and worked the very first time.

[kerbiloid]

7 minutes ago, tater said:

I bet they have data on most reentries.

For sure. Until it burns.

[tater]

 

[cubinator]

Aw man, now it's even earlier. Am I gonna be able to watch this live?