*
I’m designing a house for myself that will be using hydronic heating in concrete. I’d like to cover the concrete with engineered hardwood, specifically bamboo, and all the citations I’ve seen suggest that any such surface should be floated (or nailed) over a subfloor of a couple of layers of screwed-together 1/2-inch plywood, which seems like a lot of insulation over that warm concrete. Alternatively, I could glue directly to the concrete, but that would be pretty hard and acoustically unsatisfactory.
So how about gluing 3/16-inch cork sheeting to the concrete and floating the hardwood on top of that? The bamboo (or whatever) would be unitized and presumably not swell and shrink apart, the cork would provide cushioning and acoustic isolation, and the insulation from the concrete would be minimized. Or is there something I’m missing?
Or perhaps I could just use a floating cork-panel floor like in my last kitchen, which is beautiful, comfortable, and a real hit with everyone?
What do you think?
Replies
*
It is best to float an engineered wood floor (Harris Tarkett, Award, Kahrs, etc.) over radiant heat so it expands and contracts as a unit (as opposed to glueing or nailing it down). Most of these floating floor products are installed over a thin foam pad for cushioning and to avoid a clapping sound as it is walked on. Nothing else is needed and, in fact, anything else would reduce heat transmission through the floor.
*Hi Mike,NO DO NOT put a layer of cork over your radiant heat. Cork has the highest thermal resistance of any flooring and is not recommended for this application.Gabe
*Perhaps I'm carried away by my affection for cork, even though it would never be seen in this application, and without some engineering R-data handy, I'm not sure, but . . . Surely 3/32" (my earlier query mis-quoted the thickness) of cork would have no more insulating value than a full 1" of plywood . . . Or would it?It just seems to me that 1/16 of plastic foam would eventually degrade and disappear, but plastics have come a long way, and maybe I'm all wet on this.
*Hi Mike,No 3/32" of cork does not have the same thermal resistance as 1" of plywood. However, the thermal resistance of cork tiles is 2.094 BTU.in.The reason for my objections was that you are defeating the effectiveness of the radiant heat by installing warm materials on top. Radiant heat shines under ceramic tiles because it is a cold product.At about $3.00 dollars a square foot for cork floating floor, you can't go wrong installing it on a cold concrete surface, but half of it's benefits would be loss over a heated surface.Gabe
*Mike & Gabe-The cork issue, if it is kept to the dimension you where talking about won't compromise the effectiveness of a radiant floor.When talking about a slab on grade procedure like this, I'm sure the required water temperature is around 95-105 degrees F. So to increase it a few degrees becuase of the addition of the cork is no big deal. But before you start anything, you should start with a design of a radiant system to get the proper heat load calculations and water temps figured out.Dave H.
*Any airspace or insulation (which is of course an airspace) or reflective covering over a radiant floor will obviously waste energy -- if you insulate it you will have to turn the furnace/water heater up to compensate. As long as your surface is less resistant to heat loss than the other five sides of your floor, your ok. If your other 5 sides are not isolated from surround, you will loose heat faster by undesirable conduction than you will by the desirable radiant transfer above. If your concrete is properly isolated from water and humidity, you can use a composite wood floor glued directly to surface. This will improve the conduction up through floor and retain a large part of your radiant energy benefits. If tiles can be glued to concrete then so can composite flooring. If there is a moisture problem in the concrete slab, the interference with the radiant heating will be much more problematic than the type of flooring you use.Heavy duty vapro difusion barrier and fully isolate slab from conduction to ground, stem walls, footers. Insulate slab edges as well as underneath to a high enough rating and your heat will go to point of least resistence. By the way, an air space created by using sleepers to lift the floor will have a convection pathway that will dry you flooring from below. Dimensional lumber will have shrinkage problems no matter what connection you choose. Composite wood and ceramic floors will not have this problem.tedd
*Hmmmm. This one got my attention, so I ran the numbers. SCIENCE !!!The ASHRAE Fundementals Handbook lists 1/8" cork as having an R value of .28, and 1/4" plywood (laminate floor ?) as .31 for a total of R = .28 + .31 = 0.59 for the cork construction. U value, thermal transmittance, is the reciprocal of the added R values, 1/0.59 = 1.71/2" plywood is listed at R = .62. With two layers of 1/2" ply, and the laminate, total R = .62 + .62 + .31 = 1.55 for the plywood construction. U value is 1/1.55 = 0.6If we assume a concrete floor temperature of 95°F, a room temperature of 70°F, two inches of rigid foam under the slab (R = 10, U = 0.1), and also assume the ground temperature under the foam stabilizes at 60°F, we can do the conductive heat loss using the formula Q = U x A x deltaT, where Q = Btu's lost per given area.U = Thermal transmittance of a material, 1/ added R valuesA = area in question (we'll use ten square feet)deltaT = temperature difference on either side of insulationIn both cases the downward loss is Q = U x A x deltaT; Q = (0.1)(10)(95-60)= 35 Btus downward loss, ten sq. ft. area, or 3.5 Btu's/sq.ft.For cork, the upward loss (gain to the living space), ten sq.ft. area, is Q = U x A x deltaT; (1.7)(10)(95-70)= 425 Btus, or 42.5 Btu's/sq.ft.For two layers of plywood underlayment, the upward loss (gain to the living space) isQ = U x A x deltaT; (0.6)(10)(95-70)= 150 Btus, or 15 Btu's/sq.ft.For every dollar spent on useable heat with the plywood system, 81 cents is spent to heat the room, 19 cents to heat the ground.For every dollar spent on useable heat with the cork system, 92 cents is spent to heat the room, 8 cents to heat the ground.This ignores air layers between materials, and is a rough quesstimate. Nothing that can't be tuned in by raising or lowering floor temperature a little. Had to see it to believe it, myself. Definitely put a vapor barrier between the concrete and the wood floor.Please excuse all the math.Duncan
*
I'm designing a house for myself that will be using hydronic heating in concrete. I'd like to cover the concrete with engineered hardwood, specifically bamboo, and all the citations I've seen suggest that any such surface should be floated (or nailed) over a subfloor of a couple of layers of screwed-together 1/2-inch plywood, which seems like a lot of insulation over that warm concrete. Alternatively, I could glue directly to the concrete, but that would be pretty hard and acoustically unsatisfactory.
So how about gluing 3/16-inch cork sheeting to the concrete and floating the hardwood on top of that? The bamboo (or whatever) would be unitized and presumably not swell and shrink apart, the cork would provide cushioning and acoustic isolation, and the insulation from the concrete would be minimized. Or is there something I'm missing?
Or perhaps I could just use a floating cork-panel floor like in my last kitchen, which is beautiful, comfortable, and a real hit with everyone?
What do you think?