I know there are 1000 of them 100x smarter than me at all of this, but I have my doubts that the specs for "cold rolled" stainless will hold up after tens of re-entries, as it is effectively being tempered over and over.
From the sounds of it steel structure can basically handle much greater temperatures routinely and also even more in rare situations that might mean retiring the ship afterward but not losing everything. In his presentation Elon says that aluminum alloy or carbon fiber structure is basically done for at 300 to 400 Celsius, well that's the temperature at which stainless steel can just begin to get conditioned by heat, it doesn't structurally 'melt' until over 1000 Celsius.
Although its a curious topic, I would have no doubt that whatever temperature cycling they design parts of the structure to go through in their "rapidly reusable" regime, the effects of "heat fatigue" or whatever to call it will have been rigorously assessed .
Steel won't melt at 300-400 Celsius but "handling" those temperatures for 10 minutes will change the properties of the material, possibly by a lot, so you don't need "melting" for structural collapse (Twin Towers style).
As the GP says, if you put cold rolled steel, which has an elastic yield strength of >1500 MPa, at 300-400 Celsius, then it is only a matter of time (30-120 min) till the elastic yield strength sinks to 500 MPa or less.
So either this is a single use device, or the steel isn't reaching temperatures over 250 Celsius, or the steel isn't cold rolled but is a low strength steel instead (although that would have other problems).
Re-entry will take something on the order of 10m or less. Even the space shuttle only took 30 minutes while flying like a plane. There will be no circumstances where the ship is going to be subjected to that amount of heat for 30-120 minutes.
> or the steel isn't reaching temperatures over 250 Celsius
That might be the case. The space shuttle had an aluminum structure that would fail at 175C, and now we have 40 years of technology advancements for the heat shield.
Also I notice Cr-Steel's elastic yield strength at 400C is about 87% of its strength at room temp and I don't think it does alter significantly over these timescales at that temperature.
It will be interesting to see how high it can be routinely cycled without losing its temper, but it seems to certainly provide much more overhead for safety margins as well as routine operation.
As of the last time I heard, SpaceX was planning to use an extremely aggressive active cooling system: cold fuel would be bled through many holes in the hull, in a process similar to sweating. The hull may never get above room temperature.
No, the "sweating methane" cooling is no longer happening (per the presentation/q&a last night). Instead they're using ceramic insulating heat shield tiles - similar to the shuttle, but different materials, very different attachment mechanism, and a regular shape (tiles will all be hexagonal)
As I understand it, 300-series stainless is not heat treatable, so this should not be a problem. (304 stainless is routinely used in exhaust systems where they are heated up to glowing red temperatures every time. If the starship tank gets anywhere near that hot on entry I think there are worse problems given that on the other side of the tank is cryogenic propellants.
Stronger is not a specific enough term with metal, but tempering is the opposite of what you described, which is hardening.
Steel becomes harder and more brittle if quenched (hardening), and softer and less brittle if cooled slowly (tempering).
Cold Rolling steel is done to avoid the heat cycle which would otherwise result in tempering, which is why its specific stats don't strike me as particularly relevant past the first launch.
Heat cycles also have the effect of warping steel.
Uh what? This is exactly the inverse of what happens. Quenching leads to stronger, more flexible steel. Slow cooling leads to harder, but more brittle steel. The reason is grain structure - slow cooling allows larger grains, which don't allow the movement of dislocations as freely, but the loss of flexibility costs in ultimate tensile strength.
Please cite your correction or delete it if you find you are in error. I have provided my 2 sources which are in addition to my peer and personal experience in blacksmithing.
I don't know about his general claim, but a detail within his comment is right.
Smaller grain size leading to stronger material -- this is a result of more grain boundaries as the scale of the grain goes down. Hall-Petch strengthening.[0] A secondary effect is that with smaller grains oriented in random directions the metal is more resistant in general from stresses in all directions; whereas with larger grains you tend to get weakness in a particular direction (along the slip planes.)
This is common knowledge, at least it's basic material physics that I learned in college.
But I think the word GP was looking for was annealing. Annealed metal bends quite easily (that's the analogy in 'simulated annealing' in optimization theory). For instance the wire bonsai practitioners use is traditionally made of annealed copper. Copper is already pretty ductile but once annealed it's positively floppy. Bending it work-hardens it, causing it to stiffen back up. Eventually one learns that if you bend it just so, it will stay where you wanted it to stay.
You do not want your space craft reconfiguring itself.
Tempering is the correct terminology. Annealing is a slow and controlled heating, gradually reducing the heat over time.
from Wikipedia:
"Tempering [...] is done by heating the metal to some temperature below the critical point for a certain period of time, then allowing it to cool in still air."
Although its a curious topic, I would have no doubt that whatever temperature cycling they design parts of the structure to go through in their "rapidly reusable" regime, the effects of "heat fatigue" or whatever to call it will have been rigorously assessed .