totm nov 2023 SpaceX Discussion Thread

[Meecrob]

2 minutes ago, mikegarrison said:

Wasn't he already supposed to be landing ships on Mars by now?

How "on-time and under budget" was the Tesla Model 3?

There is a vast gulf between "out of touch crazy person" and "perfect record for being on-time and under-budget", and I argue that Musk is somewhere in that vast gulf rather than being either of the extremes.

Ok, you got me...I was a bit hasty in my comment. Let me say "compared to the regular schedule of the military-industrial complex". As soon as I posted, I knew someone was gonna bust my balls haha.

[CatastrophicFailure]

Just now, Meecrob said:

Ok, you got me...I was a bit hasty in my comment. Let me say "compared to the regular schedule of the military-industrial complex". As soon as I posted, I knew someone was gonna bust my balls haha.

[kerbiloid]

1 hour ago, Deddly said:

why does that matter for Starship? If I understand you correctly, you feel that changing Starship into a Shuttle-style spaceplane would be much superior and that their current design will not work well (or at all). As far as I understand things, all they need is some attitude control to keep Starship side-on to the airstream on re-entry in order to maximise drag and give them some level of control over the landing location.

1.
The faster a craft is accelerated down, the faster grows its total speed, the greater are the overloads and heating. 

So, a spherical Voskhod suffers from 10+ g.

The L/D (lift-to-drag) ratio keeps the craft flying forwards rather than falling down.
It makes the total accelerattion slower, allowing the craft to loose altitude and accelerate slower, reduces the mechanical acceleration and allows to dissipate the same total heat (the original potential + kinetic energy of the craft) more slowly, at lower equilibrium temperature, and do that in a thinner air due to its hrizontal speed and lift force.

So, an Apollo-like capsule withstands ~4 g overloads.

The same with cylinders.

A primitive cylinder, round in cross-section, like the Starship is like a sphere. Its L/D is very low, it tends to descent ballistically like Vostok.

So, all known viable cylindric designs are based on a rounded-triangle cross-section.
Like Apollo is a rounded triangle, so Shuttle, Buran, Bor, and other succesfully tested cylindric crafts are.
They are cylinders with rounded-triangle cross-section and flat bottom.
This significantly increases their L/D and allows to make the descent softer and cooler.

But even then their L/D appeared to be insufficient, so all of them had to increase it by adding the wings or side caissons (Bor/Spiral, Dreamchaser) and winglets.
They are not just rounded triangles in cross-section, but flattened rounded triangles with extended sides. And the only purpose of that is to raise the L/D.

If compare the Starship winglets area to the area of Shuttle/Buran/Bor/Dreamchaser/etc bottom side, they are pathetic.
Even if the drag force won't bend them, they look absolutely smaller than what the Shuttle designers were forced to place under Shuttle.

So, the cylindric shape of Starship is aerodynamically like an overturned Shuttle without wings.
Could it fly?
If it could, they wouldn't add the wings.

2.
Both hulls, a Starship round cylinder and the spaceplane flatbottom cylinder, have to dissipate the same energy (the initial mechanical energy of the ship in orbit).
They do it by interacting with air. The air gets hot exactly because it receives the mechanical energy of the ship, it doesn't have its own speed or heat.

But the flatbottom dissipates the heat uniformly, like the air meets with its bottom.
While the rounded cylinder will be heated not uniformly, but mostly along the narrow lowest part of the cylinder, because the other air will slide aside its rounded sides without taking its energy.

So, the starship-like hull will be overheating along the lowest part of the fuselage, and dissipate the heat mostly there.
As at the same time it will be falling and accelerating faster, the energy per time will be higheer, so the lowest band along the fuselage bottom will be overheated much greater than a flatbottom would.

3.
As the cylinder faces the air with only one side, there is no need in heat protection on the upper side.
So, the total heat protection is absolutely weird. The heat protection should be placed at te bottom side of the cylinder and thus make its cross-section asymmetric.

This should in turn turn the Starship rounded cylinder into an asymmetric cross-section figure, so keeping it round makes no sense again, but makes a flat-bottom design the natural shape for any cylindric spacecraft which has to aerobrake.

***

So, the problem is absolutely not in the attitude control, it's in ability to hold the velocity vector close to horizontal as long as possible (and this requires a flat bottom), and in increasing and making uniform the part of surface which faces the air flow and dissipates the heat (the same need in flat bottom).

 

1 hour ago, Deddly said:

Now, all of that aside, Elon would be the first to say that he and anyone else could be wrong, and that includes me.

I honestly don't understand why Elon's engineering skills keep being mentioned at all.

He is not a space engineer, compared to the team of DC-X working for him, he is a manager.

His engineering skills are important to understand if the golden contacts are really needed, or silver ones are enough, and to understand what the engine, avionics, and hull developers suggest.
But this is definitely not a thing to take into account when they construct something. 

That's great when the engineers do not have to explain him every number, but the only numbers requiring his decisions are in dollars. Is the claimed piece of iron enough cheap to buy, or no.

1 hour ago, Deddly said:

what benefit is a spaceplane on the Moon and Mars?

On the Moon - absolutely no benefit.
On Mars - same, because Mars requires absolutely another aerodynamic scheme or only a rocket one.

That's in turn, another disadvantage of Starship. It's equally bad as both an aerobraking craft and a lunar lander.

1 hour ago, Deddly said:

I'm confident that the current Starship design is not the only option here. How could it be improved?

They should not use the same design for everything. This never had brought to anything good.

They should split the tasks into
a partially reusable superheavy of 500+ t payload, to make the Martian ships and the lunar bases reasonable;
an Orion-like ship for the Moon and the Mars (though, it's already claimed by Orion, so not viable)
a CST-100 (but already claimed by CST-100)
scrap the Dragon before it killed somebody (and anyway it can't get lunar), as CST-100 is enough for ISS and Bigelows
develop a standard set of lunar/Martian base modules
develop heavy orbital platforms for low-class customers

Concentrate on large scale cargo fargo, leaving the rare and political crewed flights for CST-100 and Orion which anyway unlikely will get gone.

Also the 29 engines is not so low-part design.

It's mostly the same design as N1 and for same reason: make a cheaper engine but many.

It could be understandable for 16 000 t heavy projects like UR-700M (just the biggest possible engines, and it required tens of them).

But as SSSH looks like a Shuttle equivalent, it needs normal 5..10 MN engines like F1, a sixpack of them, rather this fur-tree garland of NK-15 Raptors..

[Meecrob]

2 minutes ago, CatastrophicFailure said:

This is doubly true for anyone who has worked in say flight test or any development program where each week of work is like a millimetre of progress.

8 minutes ago, kerbiloid said:

Like Apollo is a rounded triangle, so Shuttle, Buran, Bor, and other succesfully tested cylindric crafts are.

You lost me...maybe I'm blind but Shuttle and Buran don't look like cones to me.

[kerbiloid]

9 minutes ago, Meecrob said:

Shuttle and Buran don't look like cones to me.

A cross-section is a rounded triangle with extended aside flat bottom.

Like the cross-section of Apollo, too.

Spoiler

 

Edited August 9, 2021 by kerbiloid

[tater]

12 minutes ago, kerbiloid said:

They should not use the same design for everything. This never had brought to anything good.

Start your own company and make something better.

They are building a Mars craft that also works on Earth.

 

14 minutes ago, kerbiloid said:

He is not a space engineer, compared to the team of DC-X working for him, he is a manager.

Management matters.

[mikegarrison]

7 minutes ago, kerbiloid said:

They should not use the same design for everything. This never had brought to anything good.

Actually, there is strong value in flexibility. For instance, most people don't own 20 cars (one to go to the grocery store, one to drive to work, one to drive on the weekends, one to carry the kids to the doctor's office, etc.). They often own only one, which they use for all those. And car manufacturers design parts that can be used on many of their car models, rather than each model having its own unique set of parts.

However, this does have to be balanced. A spoon and a fork work better than a "spork", and it's not such a burden to own both spoons and forks.

[kerbiloid]

2 minutes ago, tater said:

They are building a Mars craft that also works on Earth.

Currently they are building a hopping mockup.

[tater]

Just now, mikegarrison said:

And car manufacturers design parts that can be used on many of their car models, rather than each model having its own unique set of parts.

This is the bit that really applies here the most I think.

Once they have manufacturing in place for engines (3+ variants of the same basic engine), and for tanks, etc, they can always specialize if the need arises. The ebenfit is just scale. They can bang out something, and try it in a way that is not possible without that underlying capability.

[kerbiloid]

2 minutes ago, tater said:

Management matters.

Management does.

If the manager can give technical advices to the engineers, either the engineers need them (DC-X from PayPal? Seriously?), and that's not the best engineers, or he just pokes them.

[tater]

5 minutes ago, kerbiloid said:

Currently they are building a hopping mockup.

Yes, you keep saying this for some entirely inexplicable reason.

1. Mockups don't function at all. (at least not as the word is used in idiomatic, North American English) The NT lander shown was a mockup.

2. SN20 is a flight article spacecraft.

3. B4 is also flight article.

4. Your claim suggests there is some gain in building a "not rocket." We've already established pretty clearly that the launch market is chump change that no one would be super concerned about.

3 minutes ago, kerbiloid said:

Management does.

If the manager can give technical advices to the engineers, either the engineers need them (DC-X from PayPal? Seriously?), and that's not the best engineers, or he just pokes them.

That's why no one uses SpaceX for launches, and the existing launch providers so easily buried SpaceX.

Edited August 9, 2021 by tater

[kerbiloid]

3 minutes ago, tater said:

2. SN20 is a flight article spacecraft.

How many of 29 engines had they ignnited at once?

How many of ignited engines exploded?

So, what's a chance of no engine exploded if 29 of them onboard?

(The upper stage is described above)

Why not launch an 1:3 scaled starship on Falcon?

To test the cylindric reentry with flaps.

Edited August 9, 2021 by kerbiloid

[cubinator]

3 minutes ago, kerbiloid said:

Currently they are building a hopping mockup.

It "hops" to orbit. That's more than I can say of any rocket I've built. And they could very well put something more useful than a cheese wheel in the nose of that thing if they wanted to. They are just skipping the payload because they prefer to test to destruction several times as quickly as possible in order to improve the already partially-functional design.

[kerbiloid]

  

Just now, cubinator said:

It "hops" to orbit.

It wants to.

Edited August 9, 2021 by kerbiloid

[tater]

3 minutes ago, kerbiloid said:

How many of 29 engines had they ignnited at once?

Maybe this is a language issue? Flight article means "capable of flight," nothing more. The current SLS core stage has been "flight article" hardware since manufacturing started on it. It is MEANT FOR FLIGHT. That's the definition of flight article.

3 minutes ago, kerbiloid said:

How many of ignited engines exploded?

So, what's a chance of no engine exploded if 29 of them onboard?

Well they've lit 27 engines 3 times with no issues, this seems little different.

Edited August 9, 2021 by tater

[Deddly]

31 minutes ago, kerbiloid said:

A primitive cylinder, round in cross-section, like the Starship is like a sphere. Its L/D is very low, it tends to descent ballistically like Vostok.

A cylinder is not as good as a flat-bottomed craft as you say, but I believe that it is a mistake to compare it to a sphere.  I know that KSP is not exactly the most realistic simulator in the world, but it does teach something: a cylinder at hypersonic velocities actually has quite a lot of lift - enough to give considerable control over the altitude on re-entry. As much as Shuttle? No. But certainly more than a sphere.

[kerbiloid]

1 minute ago, tater said:

Flight article means "capable of flight," nothing more. The current SLS core stage has been "flight article" hardware since manufacturing started on it. It is MEANT FOR FLIGHT. That's the definition of flight article.

Does the flight means a real flight, with payload mass equivalent to LEO?

2 minutes ago, tater said:

Well they've lit 27 engines 3 times with no issues, this seems little different.

So, if I get this right, 9 launches x 3 engines, 6 of them one engine exploded itself or burnt a neighbor?

6 chances of 27 of explode, 21 chance of not explode.

(1-21/27)²⁹ = looks like about a zero.

Say, they want a 0.98 probability of explosion.

p²⁹ = 0.98

p = 0.98^1/29 ~=0.9993

7 of 10 000 engines should explode, not more.

 

[cubinator]

2 minutes ago, kerbiloid said:

  It wants to.

And it will, in short order. 

If it blows up before it gets there, they will build another one in a couple weeks and fix whatever blew up the last one.

[kerbiloid]

6 minutes ago, Deddly said:

A cylinder is not as good as a flat-bottomed craft as you say, but I believe that it is a mistake to compare it to a sphere.  I know that KSP is not exactly the most realistic simulator in the world, but it does teach something: a cylinder at hypersonic velocities actually has quite a lot of lift

Conical capsules have L/D 0.3 .. 0.4

F-117  L/D = 4, and it's flat-bottomed and winged

Cylinder maybe ~1.

A plane (not F-117) 15..30

[insert_name]

SpaceX has purchased Swarm technologies. Somewhat strange considering spaceX has a much higher quality product already with starlink, albeit at a much higher price

https://spacenews.com/spacex-to-acquire-swarm-technologies/

[tater]

Just now, kerbiloid said:

Does the flight means a real flight, with payload mass equivalent to LEO?

It means it is capable of being launched (in this case orbitally, not a hop). Nothing more.

It has to accomplish NOTHING to be designed for flight. SLS could blow up on the pad when they launch it—still "flight article."

Come up with your own word, but that's what "flight article" means in English, sorry.

 

Just now, kerbiloid said:

So, if I get this right, 9 launches x 3 engines, 6 of them one engine exploded itself or burnt a neighbor?

6 chances of 27 of explode, 21 chance of not explode.

(1-21/27)²⁹ = looks like about a zero.

Say, they want a 0.98 probability of explosion.

p²⁹ = 0.98

p = 0.98^1/29 ~=0.9993

7 of 10 000 engines should explode, not more.

 

 

Version 0.X of Raptor. The current ones are like Raptor ~1.?, and Raptor 2 is OTW.

They obviously don't want it to blow up, but that's not relevant.

They fire Raptors daily at their other facility. They might have fired hundreds of times by now.

3 minutes ago, insert_name said:

SpaceX has purchased Swarm technologies. Somewhat strange considering spaceX has a much higher quality product already with starlink, albeit at a much higher price

https://spacenews.com/spacex-to-acquire-swarm-technologies/

"Swarm revealed the acquisition in a series of filings with the Federal Communications Commission Aug. 6 where the company sought approval to transfer its existing satellite and ground station licenses to SpaceX. The companies signed the merger deal July 16, according to the filings, under which Swarm would continue to operate as a wholly owned subsidiary of SpaceX."

The bit I think matters bolded.

[kerbiloid]

5 minutes ago, tater said:

They fire Raptors daily at their other facility. They might have fired hundreds of times by now.

But on launch they exploded.

[RCgothic]

It's flat out false that a cylinder doesn't have lift, and even if it weren't, Starship still has fins.

Every single Spacecraft with substantial wings that has been mentioned up thread as proof that Starship "can't work" is intended to glide to a runway. A L/D ratio sufficient to manage a runway approach is not a requirement of starship. Wings are heavy and useless in space. They're not needed 

Rule #1: Make your requirements less dumb.

[kerbiloid]

2 minutes ago, RCgothic said:

It's flat out false that a cylinder doesn't have lift, and even if it weren't, Starship still has fins.

Cylinder L/D?
Fins area + cylinder area compared to shuttle total lifting area? Those fins are negligible.

[tomf]

I think the point Kerbiloid is making is that a cylinder is a worse shape for reentry heating then the flat bottomed shapes of all(?) Successful manned spacecraft.

My guess as to the reason that the simulations that spaceX must have run are making them believe it might work might be simply sure to scale. SS is significantly larger than any capsule, and by the time it is re-entering is going to be pretty empty. I suspect it is going to be "fluffy" enough that the reentry heating is manageable with a suboptimal shape.