I think part of my mistake was not thinking of the thermal energy packets (phonons?) as waves in boxes, with the box being the boundary of whatever thing has th

I think part of my mistake was not thinking of the thermal energy packets (phonons?) as waves in boxes, with the box being the boundary of whatever thing has the energy. Which is still weird for a gas expanding into a vacuum I guess, but works for a hot solid object or a confined gas.