It is said temperature inside a nuclear explosion is higher than inside the Sun, or around 100 million Celsius. It is said temperature at the surface right under the explosion in Hiroshima (explosion occurred at 1800 ft above) was about 7700 K or 7500 C or around 16000 F or hotter than Sun's surface. Steel melts at around 1400-1500 C (depending of type) and vaporizes at 2800 C.
Here they say temperature reached 3000-4000 C for maybe 3 seconds. That would have melted anything iron, that melts ad 1538 C.
I know it is not temperature but heat that melts metal, which means a heat transfer need to occur from environment to metal and that needs time. However, at 7500 Celsius, or 5 times iron melting point or 3 times iron vaporizing temperature, at a much higher than atmospheric pressure and hot fluid surrounding metal structure moving at high speed, like in a blow torch but much faster, heat transfer may have occurred much faster than near melting in a stationary pool of liquid iron.
It also depends how big the piece of iron is but blast waves near epicenter traveling much faster than speed of sound would have blown the roof almost instantly probably before it had time to melt.
Here is the image with the dome. One can see there is no signs of melting the iron works on top or signs of melting (vitrifying) of concrete and/or bricks like some say found on the ground after explosion. Event smaller (thinner) pieces survived.
Critical mass (in nuclear physics) is when there is enough mass or density of naturally radioactive materials which through spontaneous and random decay (atom split) with emission of neutrons split other atoms, emitting one more neutrons to exceed a certain number of emissions (or atom splits) and initiate a cascade known as chain reaction.
In some uranium devices critical mass is achieved "simply" by putting together two smaller pieces.
Amounts smaller than critical mass can be encountered in small reactors, like those used to power Voyager space probe. It is understood the masses not much smaller than critical also harbor sub critical chain reactions which put out heat and radiation. That energy needs to be evacuated to prevent the masses from melting down. (Never heard of any cooling devices on stored nuclear weapons).
In practical devices, there is a competition between heating and vaporizing and the chain reaction. The expansion of material due to heating and vaporizing may be faster than the chain reaction, or stopping it.
Increased thermal agitation should also oppose to an accelerated chain reaction because it may stop neutrons from spreading fast inside the material, pretty much like the case of moving electrons. In other words, the devices may heat up the material to the point it boils and vaporizes before chain reaction can consume enough of the fuel and heat up to 100 million degrees.
When temperature increases, electric resistance of most material increases because of thermal agitation (or agitation of atoms). Faster moving atoms could be more difficult targets for neutrons (i assume a neutron has to hit a fissile atom at a certain angle and speed and/or atom also has to be quite stationary to take the hit and get split).

Pretty much like an ordinary explosive which will not do much damage if not encased in a suited shell, having time to build temperature pressure, you will have to encase the material in a steel shell in order to reach more efficiency and for such big yields the casing could be so heavy it would be hard to transport and use as weapon.
The thicker the shell, the bigger the pressure and temperature that can be achieved. Without it, there will be no explosion but a fizzle. Because being so heavy they should be using giant planes, like Spruce Goose, An 224, etc. and most likely destroy the plane too in the process.
At this point i believe the source of energy, nuclear or conventional does not matter but only the size of the shell.
Also, other inconveniences, it think nuclear devices do not last long enough to be stored, because of continuous decay of materials, heating up, loosing structural integrity, etc..
At barely sub critical mass or density a device would lower its fissionable mass and/or density changing continuously the ratio of enriched isotopes inside; material it may also heat up, like in a reactor. The whole steel structure will soon get damaged because of neutron emissions.
Conclusion. I think, like in the case of Hiroshima and Nagasaki in WWII when they only had two devices which could have been of a much bigger size and weight than shown because of heavy cases they might have some, just a few, not the numbers and readiness and deliverability they claim, by at least two orders of magnitude. Enough to keep the whole planet in check though, while they do their job (depopulation, dumbification, zombification) with their actors in all important positions in all countries.















