I assume the ultimate temperature of the melt would depend on the mass of the molten material and the equilibrium between internal heat generation and conduction/radiation of heat away from the melt. I presume it could get awfully hot - enough to easily melt steel and even concrete with which it is said to react chemically.
The melting temperature is determined by the particular material. How hot the liquid gets is determined by the heat source (volumetric heat rate) and heat transfer mechanism (conduction, convection, thermal radiation)
My own question: imagine a suspended fuel rod undergoing partial melt - the cladding melts on one place - does the lower end of the rod then drop to the bottom of the containment? It seems the cladding alone gives the rod structural integrity?
Reactor cooling is presently being achieved by 'injection' using fire pumps and others -considering the required volume of water one must assume that this is a closed, albeit leaky, loop. How does this mode of cooling differ from the one which failed after the tsumami and which we are told they are trying desperately to restore?
The heat source currently is the decay of fission products in the ceramic fuel pellets minus that which has been lost to the coolant (water or steam). Some fission products are gases (Xe, Kr), and some are volatiles (i.e., low melting point, e.g., Cs, I), some of which are soluble in water.
The Zircaloy-2 cladding surrounds the ceramic pellets, but it has certainly breached (cracked or corroded) and MAY have melting IF the cladding temperature reached ~1800°C.
The fuel rods sit between stainless steel (SS304) tie plates. Stainless steel melts at ~1400-1450°C. Only if cooling is insufficient, i.e., stagnant superheated steam could the steel or Zircaloy reaches those temperatures. If water is present - it boils, so those temperatures would not be realized. If the steam is 'wet' or 'moist', then those temperatures are not realized.
Nevertheless, before those temperatures are reached, the Zircaloy-2 would chemically react with the steam/water as in oxidation/corrosion. In that case, the Zircaloy-2 cladding may open up through cracks or ruptured hydride blisters, in which case the water/steam can communicate with the ceramic pellets. That's how the fuel particles and fission products get out.
If the bottom tie plate is not uncovered, i.e., if the water level covers the bottom tie plate, it won't melt. Any broken away cladding or fuel pellet may fall between the gaps between the fuel pellets. About every 20 inches, spacer grids are located, and they would tend to capture fuel pellet fragments and pieces of cladding. Wherever water is present, the fuel does not melt.
BWR fuel assemblies are surrounded by Zircaloy-2 channels (which facilitate the axial/vertical flow of coolant in the core). These channels (assuming they don't melt) would confined the fuel fragments and cladding to the box formed by the channel and bottom tie plate.
The bottom tie plate sits on a block of stainless steel. If covered by water, it does not melt. Then there is the structure underneath the core that contains the control rod drives. If there is water there, that does not melt.
All of the above sits inside a stainless steel lined pressure vessel of carbon steel. If water is in the bottom of the pressure vessel, it does not melt. Underneath the pressure vessel is several thicknesses of steel reinforced concrete. I expect that the bottom of containment is flooded with water. As long as there is water present - there is no melting.