totm nov 2023 SpaceX Discussion Thread

[tater]

 

[RyanRising]

kinda funny how much the blankets underneath give it a look like the tiles are a peel-n-stick affair

[cubinator]

30 minutes ago, RyanRising said:

kinda funny how much the blankets underneath give it a look like the tiles are a peel-n-stick affair

Stainless steel was too expensive to weld; rocket is now built out of LEGO bricks. Delightfully counterintuitive!

[JoeSchmuckatelli]

32 minutes ago, RyanRising said:

he blankets underneath

Do we know what they're made of?  Rock Wool?

[sevenperforce]

9 minutes ago, JoeSchmuckatelli said:

Do we know what they're made of?  Rock Wool?

Probably a thermal curtain, in case of failure.

[tater]

15 minutes ago, JoeSchmuckatelli said:

Do we know what they're made of?  Rock Wool?

Not sure what the Flexible Insulation Blanket material is. Shuttle used a low density, fibrous silica batting, maybe this is the same?

Seems to be at least a cm thick, and then they have a mesh (carbon fiber?) on top, then the tiles.

Tiles take the brunt, and in case of any tile failure the mesh and remaining tile structure (the metal Y) hold down the mesh/batting, and the mesh is good to a decent temp itself.

 

 

 

[Beccab]

New evac notice, possibe static fire attempt today

Edit: closure cancelled

Edited July 15, 2021 by Beccab

[tater]

 

The black structure looks like a jig. Is this the first chomper?

[Rakaydos]

23 minutes ago, tater said:

 

The black structure looks like a jig. Is this the first chomper?

Chomper prototype. SpaceX never expects to get things right on the first try. (though sometimes they do anyway, and have to scrap the spares)

 

unless that's just a spar of the building with a nosecone-shaped curve.

[RealKerbal3x]

ASOG is arriving in Port Canaveral!

 

[magnemoe]

18 hours ago, tater said:

Not sure what the Flexible Insulation Blanket material is. Shuttle used a low density, fibrous silica batting, maybe this is the same?

Seems to be at least a cm thick, and then they have a mesh (carbon fiber?) on top, then the tiles.

Tiles take the brunt, and in case of any tile failure the mesh and remaining tile structure (the metal Y) hold down the mesh/batting, and the mesh is good to a decent temp itself.

Sounds likely, guess the blanket also add an pressure to the tiles so they can not vibrate 

[Dfthu]

4 hours ago, tater said:

-snip-

The black structure looks like a jig. Is this the first chomper?

Forgive me if I'm dumb, but is the chomper the thing they will use the grab the booster when it lands?

[RealKerbal3x]

17 minutes ago, Dfthu said:

Forgive me if I'm dumb, but is the chomper the thing they will use the grab the booster when it lands?

The chomper is the Starship payload bay.

Of course, this render is almost 2 years old, so the payload bay design may have changed by now.

[Rakaydos]

6 minutes ago, RealKerbal3x said:

The chomper is the Starship payload bay.

Of course, this render is almost 2 years old, so the payload bay design may have changed by now.

But chomper refers specifically to the aligator style reusable fairing, as shown above. Put shuttle doors or a cargo hatch on it, that's not a "real" chomper.

[RealKerbal3x]

Just now, Rakaydos said:

But chomper refers specifically to the aligator style reusable fairing, as shown above. Put shuttle doors or a cargo hatch on it, that's not a "real" chomper.

Yeah, the chomper is just a name for the payload bay in its last known configuration. Should they change it to something more hatch-like, it's not a chomper anymore.

[sevenperforce]

23 hours ago, tater said:

Not sure what the Flexible Insulation Blanket material is. Shuttle used a low density, fibrous silica batting, maybe this is the same?

Seems to be at least a cm thick, and then they have a mesh (carbon fiber?) on top, then the tiles.

Tiles take the brunt, and in case of any tile failure the mesh and remaining tile structure (the metal Y) hold down the mesh/batting, and the mesh is good to a decent temp itself.

Okay, so I think I might be onto something here.

We have only one decent image of studs on the nosecone, so I did a little bit of pixel counting and tried my best to map it onto a repeating pattern:

There's a single inverted row in the middle, spaced to match the low rows, and a series of upper rows that are all spaced to match each other but are more widely spaced than the lower rows. I extrapolated that pattern out and filled in all the studs with regular hexagon tiles in the places where they would fit without colliding:

^{(note: I'm not 100% sure that about having a 9-tile gap on the bottom and a 10-tile gap on the top. It may be as small as 7:8 or it may be as high as 12:13; I just have no real way of knowing for sure without better imagery)}

That's where I started thinking a little. We know these stud patterns are asymmetric; there is only one possible way to attach each tile to each stud pattern. You can't rotate a tile 120 degrees and just stick it on. Why do an inverted row? Is it just to remind the workers attaching the tiles that this is the "special" row? Maybe. But when I see asymmetric patterns laid out in a symmetric arrangement, I think about chirality and congruency.

I played around with it a little bit, and here's what I came up with:

I see two possibilities.

On the left, you have just three unique tile types: the mirror pentagon, the right-handed pentagon, and the left-handed pentagon. The tiles on the upper layer are congruent with the tiles on the lower layer. There is no straight seam and the gaps are quite small.

On the right, you have nine unique tile types: a mirror pentagon, four right-handed irregular hexagons, and four left-handed irregular hexagons. However, the right-handed and left-handed irregular hexagons are mirrored, and once again each tile on the upper layer is congruent with the tile directly below it, so you can simply rotate either matching tile 180 degrees to attach. This also has no straight seam, but it also has virtually no gaps at all.

This seems like the most straightforward way of doing it. With this approach, you probably wouldn't even need any extra insulation under the seam regions. Almost the entire nosecone is tiled in the same ordinary hexagons as the body; you just have these mirrored, congruent tiles creating a seam every few rows. You can taper your ogive more and more by having fewer and fewer "regular" rows between seams, up until you are very close to the nosecone and you can do your final few layers in individually tapered, curved tiles.

Just for fun, here's the final version showing both configurations with the stud marks removed and with several layers of tesselation up the ogive:

Spoiler

 

Edited July 15, 2021 by sevenperforce
Fixed error in one embedded image

[sevenperforce]

16 minutes ago, sevenperforce said:

Okay, so I think I might be onto something here.

We have only one decent image of studs on the nosecone, so I did a little bit of pixel counting and tried my best to map it onto a repeating pattern:

There's a single inverted row in the middle, spaced to match the low rows, and a series of upper rows that are all spaced to match each other but are more widely spaced than the lower rows. I extrapolated that pattern out and filled in all the studs with regular hexagon tiles in the places where they would fit without colliding:

^{(note: I'm not 100% sure that about having a 9-tile gap on the bottom and a 10-tile gap on the top. It may be as small as 7:8 or it may be as high as 12:13; I just have no real way of knowing for sure without better imagery)}

That's where I started thinking a little. We know these stud patterns are asymmetric; there is only one possible way to attach each tile to each stud pattern. You can't rotate a tile 120 degrees and just stick it on. Why do an inverted row? Is it just to remind the workers attaching the tiles that this is the "special" row? Maybe. But when I see asymmetric patterns laid out in a symmetric arrangement, I think about chirality and congruency.

I played around with it a little bit, and here's what I came up with:

I see two possibilities.

On the left, you have just three unique tile types: the mirror pentagon, the right-handed pentagon, and the left-handed pentagon. The tiles on the upper layer are congruent with the tiles on the lower layer. There is no straight seam and the gaps are quite small.

On the right, you have nine unique tile types: a mirror pentagon, four right-handed irregular hexagons, and four left-handed irregular hexagons. However, the right-handed and left-handed irregular hexagons are mirrored, and once again each tile on the upper layer is congruent with the tile directly below it, so you can simply rotate either matching tile 180 degrees to attach. This also has no straight seam, but it also has virtually no gaps at all.

This seems like the most straightforward way of doing it. With this approach, you probably wouldn't even need any extra insulation under the seam regions. Almost the entire nosecone is tiled in the same ordinary hexagons as the body; you just have these mirrored, congruent tiles creating a seam every few rows. You can taper your ogive more and more by having fewer and fewer "regular" rows between seams, up until you are very close to the nosecone and you can do your final few layers in individually tapered, curved tiles.

Just for fun, here's the final version showing both configurations with the stud marks removed and with several layers of tesselation up the ogive:

  Reveal hidden contents

 

Alternately, if you want to be boring, there's a solution that uses the existing hex tiles and the existing truncated pentagon tiles along with just three other tile types: an elongated pentagon and two mirrored irregular hexagons. This one has a straight channel but it's extremely short -- just a few tiles long.

[sevenperforce]

Looking back through some of the earlier update images and came across something interesting.

Full-resolution image here.

A couple of curious things. First of all, we don't know what this is part of; it isn't labeled. It may be part of SN20 or it may not be. 

The careful observer will note that the empty section over to the far right has the insulation cut away to match the tiles themselves and has bare metal exposed with no tile attachment points at all. It's a very curious shape, too.

More interesting, to me, are the four rectangular sections which are all aligned together on the near side. The top two sections have the words "DO NOT INSTALL INSULATION OR TILES" scrawled on them in sharpie; the lower two sections have the words "REMOVE" with arrows pointing to all the tile attachment studs. What is this?

If this is not part of SN20, could we be looking at external mounting points for landing legs? 

EDIT:

Also noted that the curved heat shield tiles are attached directly to the aerocover with no insulative underlayment:

I wonder if they will put insulation inside the aerocovers....

Edited July 15, 2021 by sevenperforce

[CatastrophicFailure]

1 hour ago, sevenperforce said:

Also noted that the curved heat shield tiles are attached directly to the aerocover with no insulative underlayment:

Your (over)analysis of these tiles is, as always, fascinating.  More likely what we’re looking at here is just another pathfinder with half-finished test sections, that’s my guess. 

[Flying dutchman]

I really hope the faa clears SpaceX for their launch pad and that they get the Artemis money. If Boca chica is finito SpaceX will be set back years..

[tater]

[JoeSchmuckatelli]

1 hour ago, tater said:

That looks like it will add a lot of weight to one side of the craft. 

I presume the engines are powerful enough and can gimbal enough to compensate for it - but with one landing of a craft with virtually no tiles... The SS20 landing attempt will be interesting to watch (presumes SX films and releases) 

[tater]

I assume this is a platform to install/maintain the Raptors around the edge of the boosters.

 

[RealKerbal3x]

1 hour ago, JoeSchmuckatelli said:

That looks like it will add a lot of weight to one side of the craft. 

I presume the engines are powerful enough and can gimbal enough to compensate for it - but with one landing of a craft with virtually no tiles... The SS20 landing attempt will be interesting to watch (presumes SX films and releases) 

Undoubtedly they've run flight simulations with the full heatshield and know how to tune the flight control software to compensate for the offset mass.

The heatshield probably isn't adding that much mass anyway - it's likely only a few tons. With its high gimbal range, a single Raptor should be able to deal with that. It'll be no problem for three.

[tater]