I want to install a hydronic radiant heated tile floor on my existing patio that is currently an unfinished concrete slab, smoothed out and leveled but left rough. My radiant contractor says we need to isolate the radiant slab from the existing slab that will serve as a heat sink otherwise, obviating the radiant. To isolate the radiant from the concrete sub floor slab, he suggests 2" rigid XPS (also insulating vertically at the edges), then lay the pex on the XPS, drop welded wire mesh on top of the pex tubing, and pour another 1.5" concrete slab, Id use a small aggregate concrete mix with fiber and possibly polymer modified additives to add crack resistance). Once this slab cures, I can apply tile with modified thinset.
Does this sound feasible? My mason does pavers and is a good friend of mine for years (NOT A TILE GUY), so I trust him. He feels putting tile top on top of XPS is crazy and it would be a question of time where areas would compress and tile would crack.
Bulding Science Corp has literature on the issue but I see the slab is usually in the garage or basement slab and thus is a min of 4" of slab with rebar thus it creates its own flex-resistant plane.
I do not own a copy of the TCNA handbook but have seen architects reference it in different articles in regards to the Hydronic Radiant Heat Pex Tubing needing a minimum of 3/4" of mortar from the top of the tubing. Thus by using 1/2" oxygen barrier pex with 5/8" outer diameter, the 1.5" slab is good but does NOT address the flexibility and longevity of this slab design.
Any constructive input?