Starship heat shield

[spacexfanatic]

I counted NASA wireless scanner is 3 seconds per tiles for space shuttle I counted 20 hours for scanning all tiles but don't know how much time to process the data nor to change any tiles, for me it's more safe to get a well checked and maintained heat shield and this require time. to close the drame of the shuttle.

Anyway whatever less expensive and rapid methode to keep SS operable ans most important safe is welcome.

[sebk]

Please refere to the above response, it must be as easy as interchanging an airliner engine.

Do you know what it takes and how long to change out an airliner engine? 

https://blog.virginatlantic.com/change-aircraft-engine/

I saw a documentary on National Geographic it was too easy.

Do you realize this was multiple hour operation? Hint: the movie is not real time  See, that lighting is changed in the later phases of the operation? That's because hours have passed in the mean time.

Anyway, replacing dented tile is a fast operation. Replacing entire shield is not.

To replace a single tile:
1. remove the damaged one
2. place gap filler
3. click on the new tile
4. safety inspector checks the tile is properly placed

Done.

To replace entire shield structure:
1. Roll the vehicle to a building (50m tall structure is a huge sail, it has no rigidity, it will get damaged by wind when not firmly attached to either rocket body or a special stand; edit: 4m/s wind would exert a literal ton of force on such 500m^2 structure, if the structure weights 20t it would be pushed ~18° off vertical -- good luck doing any precise fastening work in such conditions; hence building is a must)
2. Undo few hundreds fasteners
3. Remove the structure
4. Put it on a special stand for detailed work
5. Hoist replacement structure
6. Carefully replace hundreds of fasteners
7. Safety inspector verifies hundreds of fasteners
8. Roll the vehicle back to the pad

Uffff. Hours have passed.

TL;DR:
The idea is completely wrong. Give it up.

[Oersted]

I saw a documentary on National Geographic it was too easy.

These forums are populated with engineers and industry-level experts. Please don't argue with them by referencing TV documentaries. There is a lot to be learnt just by reading the forums and not posting.

[livingjw]

- Heat shields have very little strength. They usually have a thin outer shell, or ablator, with insulation underneath. The insulation may be either rigid, like the Shuttle tile, or in the form of felt or batting. The whole has very little strength because you are trying to make it as light and as insulating as possible.

- There is also going to be thermal expansion mismatch between the heat shield and the underlying structure. Because of this tile or panels can only be so large before expansion joints have to be incorporated into the heat shield.

- A large heat shield would not be self supporting. It would rip apart of its own weight. Adding reinforcements to it only makes it heavier. Not good.

John

[john smith 19]

Rapid inspection of tiles and replacement of only the defective ones is a solvable problem in an era of computer vision.

It would have been a solved problem in the 90's if Marshall had let CMU install their shuttle inspection robot but apparently it was too risky as it might crash into the vehicle it was scanning (despite CMU's plan to use a very slow moving vehicle to carry the arm carrying the sensor package).

I'm guessing that would have made several hundred staff surplus to requirement as well.

[john smith 19]

- There is also going to be thermal expansion mismatch between the heat shield and the underlying structure. Because of this tile or panels can only be so large before expansion joints have to be incorporated into the heat shield.

True. In the case of the Shuttle tiles versus aluminum skin it was something like 3:1 in favor of the skin.

But

SS301 is substantially  lower in TCE.

As I also noted in another thread NASA solved this problem (and patented it, which has been expired for a decade) in the 80's. In principle large single pieces of TPS with substantial TCE (and anisotropic TCE) are viable.

Personally I  think it's a trade off. Replace a single piece the size of a whole half cylinder for any damage to that piece? That sounds excessive to me. OTOH too small and you end up with a huge jigsaw of pieces.

Ideally (and let's be clear this is a total fantasy) you want pattern to line up so that when a tile is damaged it only ever happens to a single tile and never crosses over and damages the edge of an adjacent tile as well.

Which is a total fantasy, given the surface area and number of tiles involved. 

IRL some times you'll get lucky, some times you won't. I think it's too early to say with any certainty what an optimal tile size (or placement) is

This is not a problem. It just means that know that SX have a scope of the problem it's time to start the field tests.  I strongly doubt what we are seeing now will be what finally goes to orbit. But it is a start.

[Barley]

Ideally you want a thermal protection system that never gets damaged, then you don't care how long it takes to repair.

That's a fantasy, but the toughness of the tiles and gentleness of operations is an important part of the engineering.

[rakaydos]

Ideally you want a thermal protection system that never gets damaged, then you don't care how long it takes to repair.

That's a fantasy, but the toughness of the tiles and gentleness of operations is an important part of the engineering.

That fantasy might apply to an individual flight, but it's better to plan for least expence for the lifetime of the vehical. Cheap repairs AND a durable shield is better than just a durable shield.

[Slarty1080]

Assuming that a single tile has been damaged and needs to be replaced, how would they do this? Do they have to use the narrow gap between tiles to get underneath or do they do something from the front of the damaged tile to remove it?

[rakaydos]

Assuming that a single tile has been damaged and needs to be replaced, how would they do this? Do they have to use the narrow gap between tiles to get underneath or do they do something from the front of the damaged tile to remove it?

Well, the tile is damaged already, so it's already written off. Shatter it with a hammer, pry off the remains, straighten the pins, and attach a new tile?

[Slarty1080]

Assuming that a single tile has been damaged and needs to be replaced, how would they do this? Do they have to use the narrow gap between tiles to get underneath or do they do something from the front of the damaged tile to remove it?

Well, the tile is damaged already, so it's already written off. Shatter it with a hammer, pry off the remains, straighten the pins, and attach a new tile?

Seems reasonable, but when attaching the new tile won't the face of the tile insulate the studs from their needed electrical connection for the weld?

[livingjw]

Assuming that a single tile has been damaged and needs to be replaced, how would they do this? Do they have to use the narrow gap between tiles to get underneath or do they do something from the front of the damaged tile to remove it?

Well, the tile is damaged already, so it's already written off. Shatter it with a hammer, pry off the remains, straighten the pins, and attach a new tile?

Seems reasonable, but when attaching the new tile won't the face of the tile insulate the studs from their needed electrical connection for the weld?

We are assuming that the tile are held on to the studs with clips. Short studs welded to the skin would probably not be damaged. If the studs are damaged, the skin in the area is probably also, so that is a separate problem.

John

[JamesH65]

I wonder what the optimal tile size is? There must be some cross over point as you get larger where it starts to become less easy to manufacture, install or handle, likewise getting smaller. Too small, too many and hence heavy, too large, need to be curved.

I suspect there will have to have some 'custom' tiles rather than hexagons for things like leading edgers and the nose.

[livingjw]

I wonder what the optimal tile size is? There must be some cross over point as you get larger where it starts to become less easy to manufacture, install or handle, likewise getting smaller. Too small, too many and hence heavy, too large, need to be curved.

I suspect there will have to have some 'custom' tiles rather than hexagons for things like leading edgers and the nose.

Well, since SpaceX has hired really smart engineers to work out the details of its TPS system, my guess would be, about the size of the tiles we have seen.  :^)

John

[Shrike DeCil]

For those who don't go out to an airport and look at a 777 or an airbus 380. the Shuttle was about DC9 sized. Bigger than any previous vehicle, but still quite small.

My quip not quite right (It won't be carrying a 747!) but:

NASA has planes that carry spacecraft, SpaceX will have spacecraft that could carry a plane.

[Jorge]

Rapid inspection of tiles and replacement of only the defective ones is a solvable problem in an era of computer vision.

It would have been a solved problem in the 90's if Marshall had let CMU install their shuttle inspection robot but apparently it was too risky as it might crash into the vehicle it was scanning (despite CMU's plan to use a very slow moving vehicle to carry the arm carrying the sensor package).

I'm guessing that would have made several hundred staff surplus to requirement as well.

RMS/OBSS won because it really was the best solution for reducing risk, despite being heavier than all the free-flyer solutions, including CMU's (and AERCam).

[GregTheGrumpy]

- Heat shields have very little strength. They usually have a thin outer shell, or ablator, with insulation underneath. The insulation may be either rigid, like the Shuttle tile, or in the form of felt or batting. The whole has very little strength because you are trying to make it as light and as insulating as possible.

- There is also going to be thermal expansion mismatch between the heat shield and the underlying structure. Because of this tile or panels can only be so large before expansion joints have to be incorporated into the heat shield.

- A large heat shield would not be self supporting. It would rip apart of its own weight. Adding reinforcements to it only makes it heavier. Not good.

John

John,

I like your answers & responses.  You provide information that I certainly don't have so I learn a lot.  You are generally patient and kind in your answers & I really appreciate that.

-g

[john smith 19]

RMS/OBSS won because it really was the best solution for reducing risk, despite being heavier than all the free-flyer solutions, including CMU's (and AERCam).

I suspect we're talking at cross purposes.  The CMU project I am referring to was for a robot to inspect shuttle TPS on the ground during refurbishment in the hangar. 

[Jorge]

RMS/OBSS won because it really was the best solution for reducing risk, despite being heavier than all the free-flyer solutions, including CMU's (and AERCam).

I suspect we're talking at cross purposes.  The CMU project I am referring to was for a robot to inspect shuttle TPS on the ground during refurbishment in the hangar. 

Fair enough. I mistook it for one of the in-flight robotic proposals.

[john smith 19]

Fair enough. I mistook it for one of the in-flight robotic proposals.

No. On the ground refurb.  The sensor package was designed to use multiple sensors to detect both surface visible damage and sub surface damage brought on by over temperature stressing of the tiles. In principle with every tile bar coded it could detect a damage one and issue the re-order immediately.

In the context of SS/SH you would probably have a second arm equipped to handle the actual replacement in near real time.

We'll see how SX actually do it.