totm nov 2023 SpaceX Discussion Thread

[RealKerbal3x]

Raptor SN50 being prepared for installation on Starship SN10:

[MinimumSky5]

I'm taking the fact that their putting more and more tiles on the Starships as a sign that they're not seeing significant shedding during flight, so they're getting more and more confident with the design. I do still want to see their solution to tiling the cone, though.

[cubinator]

2 minutes ago, MinimumSky5 said:

I'm taking the fact that their putting more and more tiles on the Starships as a sign that they're not seeing significant shedding during flight, so they're getting more and more confident with the design. I do still want to see their solution to tiling the cone, though.

I honestly wonder if they really need to use different tiles for the cone. Maybe for the very tip, sure, but the tiles are small enough that the curvature would be small at the scale of individual tiles, and I think you could keep using the same hexagonal tiles most of the way to the top.

[RCgothic]

The problem is you can't tile a curved surface with regular hexagons.

[sevenperforce]

11 minutes ago, cubinator said:

I honestly wonder if they really need to use different tiles for the cone. Maybe for the very tip, sure, but the tiles are small enough that the curvature would be small at the scale of individual tiles, and I think you could keep using the same hexagonal tiles most of the way to the top.

One possibility would be to go larger, not smaller.

Continue the same hex pattern as far up as possible, but have a large, single-piece ceramic shield at the top which fills in all the oddly-shaped gaps. Or, at most, a handful of them.

1 minute ago, RCgothic said:

The problem is you can't tile a curved surface with regular hexagons.

Right. However, you can tile a curved surface with regular hexagons up to a certain degree depending on gap tolerance. Use irregular, custom shapes to handle any gaps too large to tolerate.

[RealKerbal3x]

 

[Deddly]

About these ceramic tiles... we know that the shuttle used ceramic tiles and that they sometimes had problems with them coming loose. I'm not implying Starship's tiles are exactly the same thing, but there is obviously a possibility that some might come loose or get damaged during the flight. I was wondering about this...

Apparently, if you use a blow torch on a block of steel, it will form a molten pool where you are heating it. But if you do the same with a block of copper, it conducts the heat away so well that a molten pool does not form until the entire block is near melting point. So... if SpaceX had a sheet of copper underneath the heat shield tiles and one of the tiles cracked or came loose, would the relatively small area of copper that gets exposed to the shock heating be able to withstand the heat by conducting it away as a heat sink to the rest of the mass of copper? If so, how thick would it need to be, and would it be worth the extra weight and expense for added redundancy and safety on a crewed mission?

[kerbiloid]

Niobium.
True spacemen use niobium. All 1950s-60s are about it.

No niobium - no starshipping.

[Terwin]

36 minutes ago, Deddly said:

About these ceramic tiles... we know that the shuttle used ceramic tiles and that they sometimes had problems with them coming loose. I'm not implying Starship's tiles are exactly the same thing, but there is obviously a possibility that some might come loose or get damaged during the flight. I was wondering about this...

Shuttle used glue over aluminum.

One of the instances where a tile fell off they managed to land safely because there was steel under the aluminum that managed to survive and keep the shuttle intact

Starship uses bolts over steel.

It *should* survive reentry naked, but without heat tiles it could negatively affect temper and other properties of the steel, especially as they intend to aerobreak from both the Moon and Mars(the shuttle never went faster than LEO).

If the tiles are still good, they may just need a quick walk-around before they refuel and relaunch.

Without tiles, or with damaged tiles; they may need to perform more in-depth checks, especially for hotter reentries such as Mars and beyond.

 

[RealKerbal3x]

36 minutes ago, Deddly said:

About these ceramic tiles... we know that the shuttle used ceramic tiles and that they sometimes had problems with them coming loose. I'm not implying Starship's tiles are exactly the same thing, but there is obviously a possibility that some might come loose or get damaged during the flight. I was wondering about this...

Apparently, if you use a blow torch on a block of steel, it will form a molten pool where you are heating it. But if you do the same with a block of copper, it conducts the heat away so well that a molten pool does not form until the entire block is near melting point. So... if SpaceX had a sheet of copper underneath the heat shield tiles and one of the tiles cracked or came loose, would the relatively small area of copper that gets exposed to the shock heating be able to withstand the heat by conducting it away as a heat sink to the rest of the mass of copper? If so, how thick would it need to be, and would it be worth the extra weight and expense for added redundancy and safety on a crewed mission?

Starship tiles are mechanically attached, not glued like Shuttle tiles. That should help with tiles falling off.

But stainless steel is pretty heat resistant on its own. I'm not sure a copper sheet under the tiles would be worth it when the steel should be able to just about handle it on its own.

[sevenperforce]

1 hour ago, RCgothic said:

The problem is you can't tile a curved surface with regular hexagons.

The tiles have gaps between them, so it's fairly straightforward to wrap the tile matrix around a cone as long as you mind the gap.

As you can see, the angle starts four tiles down but the gaps are large enough that the bent-over tile sections (unbent tiles on right) don't collide. Wherever there is a collision you take out that tile and you use a large, single, customized tile to fill that particular space. 

If you don't want to use a customized piece then you can just tile the space irregularly and allow larger gaps. 

 

[Geonovast]

1 hour ago, Deddly said:

we know that the shuttle used ceramic tiles and that they sometimes had problems with them coming loose.

The shuttle's tiles were glued.  Starship's are mechanically attached.

[mikegarrison]

I think concerns about the reusability the heat tiles are premature. Unless you think it matters whether they are still attached to little pieces of flaming debris....

Edited February 5, 2021 by mikegarrison

[RyanRising]

35 minutes ago, mikegarrison said:

I think concerns about the reusability the heat tiles are premature. Unless you think it matters whether they are still attached to little pieces of flaming debris....

We're all thinking about the future here, and that includes talking about the merits and challenges of certain technologies past the first few EDL tests, which I fully expect are going to end how you suggest. This discussion of the tiles has been based on the premise that they will eventually get that stuff figured out.

Also, Re: the tiles being mechanically attached, I think if the tiles are as fragile as the Shuttle's, switching from adhesive attachment to mechanical attachment is trading the tile popping off for the tile shattering & leaving behind only the part that the fastener's stuck into. Hopefully they can figure out how to make it sturdier, but it also definitely helps that the stakes are a lot lower for losing a tile from a steel structure than they are with an aluminium one. That being said, making this heat shield fully and cheaply reusable is probably going to be one of the biggest challenges SpaceX faces with Starship. Space is hard, but reentry seems like another level on that.

Edited February 5, 2021 by RyanRising
Added on-topic heat shield discussion.

[zolotiyeruki]

1 hour ago, sevenperforce said:

The tiles have gaps between them, so it's fairly straightforward to wrap the tile matrix around a cone as long as you mind the gap.

As you can see, the angle starts four tiles down but the gaps are large enough that the bent-over tile sections (unbent tiles on right) don't collide. Wherever there is a collision you take out that tile and you use a large, single, customized tile to fill that particular space. 

If you don't want to use a customized piece then you can just tile the space irregularly and allow larger gaps. 

 

Another option is to switch to slightly-trapezoidal tiles for the nose section, and arrange them in rows.  That way, you only have three types of tiles: hexagonal for the cylindrical portion, half-of-a-hexagons at the joint between the rings and the nose cone, and trapezoid.

[paul_c]

How about making the tiles half as thick each as they need to be; then attaching twice the number in 2 layers? Curved sections will have big(ger) gaps but with all the tiles overlapping it still leaves the 1 tile thickness. For a sufficiently small gap, it would be approachable to engineer the situation to accommodate this with heat transfer thru the structure?

[kerbiloid]

8 hours ago, Terwin said:

Shuttle used glue over aluminum.

Afair, originally it was going to use a titanium hull instead, but they considered it too expensive and replaced with aluminium.

 

[RCgothic]

13 hours ago, sevenperforce said:

The tiles have gaps between them, so it's fairly straightforward to wrap the tile matrix around a cone as long as you mind the gap.

As you can see, the angle starts four tiles down but the gaps are large enough that the bent-over tile sections (unbent tiles on right) don't collide. Wherever there is a collision you take out that tile and you use a large, single, customized tile to fill that particular space. 

If you don't want to use a customized piece then you can just tile the space irregularly and allow larger gaps. 

 

As always an impressive diagram, but I'm pretty sure this ends up with a large number of different custom tiles.

Edited February 6, 2021 by RCgothic

[Elthy]

If SpaceX uses modern manufacturing i dont see an isse with lots of differently shaped tiles. When a Starship comes back from orbit they will do some sort of inspection of the tiles, propably augmented using computer vision. This way it should take only a few minutes to know which tiles need replacement which would also output the exact shape since its tied to the location. If the only have a few types of tiles its easy to manufacture a bunch of them in advance. But even if everyone is shaped differently its not to hard:

Somewhere in the manufacturing process of a tile its shape is determined, maybe by milling of using a mold for sintering. So the cataloge numbers of the tiles that need to be replaced have to be send to that machine, if its milling the tiles it would just load the right toolpaths for each one. If its molding it should simply draw the right mold from a automated storage or print it when they are single use. The end result is always the same, you get exactly the tiles you need, propably within days or even hours if done right.

Edited February 6, 2021 by Elthy

[mikegarrison]

1 hour ago, Elthy said:

If SpaceX uses modern manufacturing i dont see an isse with lots of differently shaped tiles. When a Starship comes back from orbit they will do some sort of inspection of the tiles, propably augmented using computer vision. This way it should take only a few minutes to know which tiles need replacement which would also output the exact shape since its tied to the location. If the only have a few types of tiles its easy to manufacture a bunch of them in advance. But even if everyone is shaped differently its not to hard:

Somewhere in the manufacturing process of a tile its shape is determined, maybe by milling of using a mold for sintering. So the cataloge numbers of the tiles that need to be replaced have to be send to that machine, if its milling the tiles it would just load the right toolpaths for each one. If its molding it should simply draw the right mold from a automated storage or print it when they are single use. The end result is always the same, you get exactly the tiles you need, propably within days or even hours if done right.

This is pretty much exactly the Space Shuttle system. Every tile had a number and a specified location. And yet, this caused problems.

Let me point out some of the potential problems with different-shaped tiles:

• Inventory -- you can't store one big stack of tiles; instead you need lots of stacks of different-shaped tiles
• Manufacturing errors -- different-shaped tiles means a chance to put a tile in the wrong place
• Cost -- it's simply cheaper to make large numbers of standardized parts
• Cost again -- it's also cheaper to assemble things using standardized parts

None of these are absolute deal-breakers, but they are reasons why I'm sure SpaceX would vastly prefer to use uniform tiles rather than having to essentially custom-build and custom-install every tile.

[Elthy]

Obviosly it would be easier with one tile for everything, but that wont be possible. Dont forget that the Spaceshuttle was build in another era with conservative NASA design. I guess even on the last flights most stuff regarding the tiles was done by hand, including inventory and ordering new ones. But if you automate the process most of the (expensive) work is done by the computer. Modern manufacturing has no issue with that if you design the process properly in the first place.

[Scotius]

I hope it will not be too complicated. Spaceship's shape is much simpler than Shuttle's after all. Cylindrical-to-conical tapering pipe vs. delta-winged spaceplane with elements of lifting body mixed in

[magnemoe]

43 minutes ago, Scotius said:

I hope it will not be too complicated. Spaceship's shape is much simpler than Shuttle's after all. Cylindrical-to-conical tapering pipe vs. delta-winged spaceplane with elements of lifting body mixed in

Yes, I assume they will need specialized tiles tappering on the nose, either use special tiles preferable in an pattern who can be reused or filler tiles of various types kind of like the pentagon fillers on an soccer ball you want repeatable patterns. At the nose this changes again as the curving get stronger and you probably need smaller tiles. 

You want standards as you want an inventory so you can swap tiles fast without manufacturing spares. 

[paul_c]

There's obviously a sliding scale between number of distinct tile shapes and gap size. I imagine if they worked out a tolerable gap size, their inventory could be 10-12 variants, rather than thousands like the space shuttle. Of course, too big a gap negates (or renders inaccurate) a lot of the aerodynamics work if its done on a smooth stainless steel shell!

[JoeSchmuckatelli]

So - don't try to use the hexes for the cone.  Instead - use strips: https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.atlasroofing.com/download.php%3Fuid%3D1505%23:~:text%3DCones%20with%20three%20to%20eight,up%20evenly%20to%20a%20point.&ved=2ahUKEwi3mdmBzdXuAhVLGs0KHaX4DQUQFjABegQIEBAF&usg=AOvVaw3YfbJijR8HMKDfnAo3V7rM

Another idea :

https://www.designingbuildings.co.uk/wiki/Conical_roof_slating

 

 

 

Edited February 6, 2021 by JoeSchmuckatelli