I'm installing a slab on grade radiant floor in our new home.  I've laid 6" of clean rock over compacted soil, then 6 mil visqueen, then 2" xps foam, then 3" AB.  On top of that will be the tubing and concrete. 

I have two concerns.  The first is any moisture that is between the vabor barrier and the foam (I know it's there from wetting the AB to compact it).  The other concern is the AB itself, I heard two opinions on this.  One - don't do it because the moisture can take years to get out through the slab and Two - if you don't want your slab to crack, this is a proven method.  I'll note that 80 percent of the flooing in the house is the finished slab.  The other 20 percent is either tile or floating hardwood.

I'm planning to place the tubing this week and pour the slab the following week.  But I'm going to live here a long time and if what I've done needs correction, no time like the present.

What is AB?

What is AB?

Hi Badger,
AB is compactable crushed rock. It's different than the 3/4 minus we used in oregon as it's not just rock, but has a sandy/clay like material in it that when wetted and compacted becomes very dense. They use it under slabs everywhere in this neck of the woods.

Crush limestone or number 7 I think around here. It will have a typical moisture level and would never be "wetted" in a normal application. This degree of compaction is not needed.

I am not sure where you got this spec?

Soil, crushed rock, vapor barrier, foam, tube and concrete to finish. This is a very old (20+ years) formula that works.

You will have cracks, I don't care what you do.

Where tile is applied a crack control barrier is used.

If the foam were properly placed between the vapor barrier and the slab you would have no worries.

The moisture between the visqueen & XPS will easiliy be able to dry toward the interior through 2" of XPS, AB, and wood or tile flooring. But it would be a good idea to let the AB & concrete dry before putting down the wood. (Heating the slab would accellerate the drying of the AB.)

Even if there are pockets of liquid water between the XPS & visqueen, the vapor retardency of the XPS would be enough to protect the wood from too-high moisture levels. But the AB has to be quite dry to not have initial issues with the wood.

The concrete should not be heated until cured. This time is dependent on the mix but generally is two weeks.

Though moisture does come out of concrete, the curing process is not drying per se, rather a chemical reaction that never stops (it gets harder with time). Generally the drier the mix and the longer the cure time (under ideal conditions) the stronger the concrete.

Never turn a radiant floor on a green slab to "dry" it.

You may be interested to know that, Gypcrete and similar cementious underlayments are just the opposite.

Good point on the concrete curing-time issues. The wet AB would be a constant source of moisture beyond what the concrete needed though, and "cooking" the AB dry before putting down the finish flooring would still be prudent. There is very likely to be a LOT of water in it (and the concrete) even months after the fact if left to dry on it's own.

This (moisture) is one of the reasons why I don't understand why people even put AB or sand above the vapor barrier. Here in CA, the practice is rampant, but not one person - not even two structural engineers - could tell me why there's sand. In the end, I went back and forth with my SE, and we eventually just deleted it.

YMMV.

Jeff

Posted By jbaron on 10 Dec 2010 09:04 PM
This (moisture) is one of the reasons why I don't understand why people even put AB or sand above the vapor barrier. Here in CA, the practice is rampant, but not one person - not even two structural engineers - could tell me why there's sand. In the end, I went back and forth with my SE, and we eventually just deleted it.

YMMV.

Jeff

And your house fell down, right?

I'm not sure whether AB would protect or aggravate slab damage in the event of an earthquake, but clearly slabs have survived a very long time without it, in all kinds of conditions.

Thanks everyone for your advice.  My biggest concern was water between the vapor barrier and the foam just rotting away and getting moldy.   But as long as the moisture can get out and I don't have any floor coverings to trap it (which I don't), I don't think a bit of moisture coming through the slab will hurt.  In Arizona, it's pretty dry.

Foam doesn't feed mold, neither does poly, but organic contaminants of the AB might.

In AZ things are pretty dry on average OUTDOORS, but in a very tight house you'll still have high relative humidity unless you're actively ventilating. But if you keep it 30-40% RH indoors the AB & concrete will release it's moisture, as will anything that managed to drip under the foam (eventually.)

Here is a good article on why NOT to put the AB material between the foam and concrete



Building Science Article

Posted By NH Homeowner on 14 Dec 2010 12:53 PM

Here is a good article on why NOT to put the AB material between the foam and concrete



Building Science Article

Aha! The "why AB" is finally explained!


From the article:

"So where did the idea of putting sand between plastic and concrete come from?  Like most bad ideas, it arose from good intentions—it was an attempt to reduce the curl of concrete slabs due to uneven drying.  If a slab dries only from the top—the top becomes drier than the bottom—duh…The top shrinks relative to the bottom and the slab edges curl upwards. This is not a big movement and is not a big deal except where you need ultra flat floor systems—like in chip manufacturing plants. Slab curl is very easy to control, use a low water-to-cement ratio (less than 0.5) and wet cure the top (wetted burlap works—Photograph 5). Or just use a low water-to-cement ratio and forget about doing anything else."

(Emphasis mine.)

Somehow I doubt the average homeowner is going to be bent out of shape by a coupla millimeters of concavity to the floor, even if you DIDN'T cure it with excess top water. 

My father's construction company poured a LOT of big commercial & industrial building slabs in the '70s & '80s  in WA.  Keeping the top surface wet by repeated mist applications for the first 48hours or so seemed to be a standard protocol, as was applying a sealer after the last wetting.  The only slab I can remember with a compacted sand substrate was a foundation for a 50' tall steel water tank. The sand was dead-flat to within 1/2", but slab it self was convex by design, an inch or two out of true-flat. The steel sides of the tank rested on a highly reinforced very substantial ring footing, but the reinforced slab was carrying the weight of the water.  (I happened to drive by the location last summer- it's been taken down after 30+ years of service.)

Boy, that article sure brings back a lot of DIY construction memories. The other not-so-fond concrete memory that I have was the decision as to whether or not to use Sipex, which the SE put on the plans as "optional." Sipex is a chemical that somehow acts or combines (sorry about my chemistry here) with the water in the concrete, and makes the concrete waterproof. (It's probably one of those "foreign sounding names" mentioned in the article.)

The only problem with Sipex - well, the only problem if you aren't intending to finish the floor, in which case you have another problem - is that Sipex is expensive. We have a full, underground concrete basement, and used about 120 yards of concrete altogether (basement, slabs, foundation) and our Sipex bill was going to be something like $6,000. No one could tell me whether or not I needed it, only that, well, "we just use it here." Since I'm in California, the state with the sand issue, this is a familiar refrain.

We didn't use Sipex either. I asked my retired father - a retired civil engineer - and his only question was "is there a drain around the basement?" When the answer was yes, the Sipex was also deleted.

I, though, would care about cupping. We didn't want to use burlap because we colored the concrete, so we purchased a one-use curing blanket that was composed of a layer of white fabric covered by a sheet of clear plastic. We wet the slab, layed the blanket down, and then kept it covered for 2 weeks.

Those decisions were very early on in my house project, and they taught me to be very, very wary and to ask a lot of questions.

Jeff

PS. Yes, Dana, my house is still standing, and before you ask, no, I do not have any water leaking into my basement. (And nor do I have any water flowing out of the foundation drain.)

I live in washington state and just did a floor we used compacted sand with 6 mill poly on that then 2 inch foam.  The tubing we stapled to the foam the 2 inch dobe blocks for the rebar to rest on 2 by 2  squares also ran rebar on 45s on in side corners since that is where it usually cracks.  We poured 5 inches of concrete with lots of plastsizeres and water reducers to eliminate shrinkage.  Now for crack control we waited 48 hours then saw cut on point 40 by 40 inch tiles 1 1/4 inches deep two blades wide the we waited a few days and covered with 6mill plastic the 1/2 inch osb until we finished construction the striped covering acid stained and sealed then grouted cuts and resealed turned out beautiful people actually think we set 40/ 40 tiles hence they are all dead level to on another it makes for a great indestructable floor.  There are zero cracks it is important to have the right temps outside during the curing process so concrete doesnt dry to fast  The only thing I would do differently is to order concrete colored than acid stain because stain gets drawn to the cuts and makes your cuts a tad bit darker also some water based sealers will react to the acid. For a crack control decrotive slab it worked great I have also done slabs where the grooved the concrete in the pour but not as effective and not deep enough.  You need to make sure your sawcutter is very good you only have one shot

Excellent article NH.

California is a different land.