It also makes DALR expansions tricky, because they are one «parcel» of gas expanding against another, which gives me a headache — topologically — whenever I try to imagine it happening in a uniform way, at which point my brain wants to shift gears into Navier - Stokes altogether because simple pictures of a cubic meter of gas expanding into other gas surrounding it as it rises forces one to think of what the other gas has to do to make room, and then you've got gas expanding into gas that is contracting, some rising, some falling, with dynamic viscosity
along the parcel
boundaries and I just can't solve or even visualize the PDEs in my head particularly well except in simple e.g. convective
roll sorts of ways.