Once any part of the pellets (?) or powder (?) or mix or crystals/clumps get "past" the melting temperatures, some part of each pellet or clump is going to start melting.
You cannot predict exactly how much nor where or what part of each clump or crystal or pellet is going to melt though. You just know that some part of each is going to start melting: probably at the "points" or highest part of each clump (if heated by gas passing above or by radiation from above) or the lowest part of each clump or pellet touching the tank floor or wall (if heated by contact from the wall of floor of the tank.)
Obviously, the least time above the melting point implies the least amount melted. More time, more melted parts on each clump until everything melts and you have control over your material again. But in the middle? You don't know.
Now, realistically, just like water "starts" boiling at 212 F, but the whole pot doesn't magically get to 212 all at the same time, you could "assume" that rushing a batch through a furnace at too hot a temperature might not melt too much - but that depends on what you want to happen. And what you don't want to happen.
