inject gypcrete?

scotter

I bought a home built in 1999 a couple of years ago with radiant heat installed as a wood sandwich. (3/4 OSB, 3/4 OSB strips (sleepers) with pex-al-pex between the strips, then another 3/4 OSB). Hardwood above the final OSB. Portions of the strips I can access appear to be 4" wide, leaving 12" of open air gap where the pex-al-pex runs. Thermal images allow me to clearly see the piping. I suppose this means that there is very little thermal conduction from the pex-al-pex to the surrounding wood given the air space. The floor doesn't produce enough BTUs to keep the house warm during the coldest days, so the scorched air system makes up the difference.
How crazy would it be to inject/pour gypcrete into the hollow space in order to increase thermal conduction? One zone is an attic zone that only has the 3 layers of OSB without a finished floor. This could be a test zone to see if something could be done. Would many carefully drilled holes allow pouring/injecting of gypcrete into that void? What would be the likely BTU increase?
Are there other ideas to increase thermal transfer?

Jamie Hall

I wouldn't bother. Wood itself is, at best, only so so as a heat conductor. This is why metal plates are used for radiant floors. How many BTUh per sqaure foot are you trying to get out of the floor?

scotter

Thanks Jamie, I don't know how to do BTU calculations. The simple answer is 'more' so that the floor can heat on its own for the coldest days, without the scorched air system making up the difference. I hoped that replacing the air around the pex with gypcrete would increase the thermal transfer to the wood quite dramatically.

hot_rod

Be an interesting experiment, and yes conduction is a much better conduit for heat than air. The tube to the fill and the fill against the subfloor would need to be connected some how.

What is below the flow as the gyp will leak thru any crack.

Also gyp does shrink a bit as it sets, so it may not be in contact with the subfloor.

A 1/2 motor with a mixer in a 5 gallon bucket is how I have seen gyp installers mix small amounts for repairs.

May try a bag of sand mix Sackrete with some plasticizer added, it would be a cheap trial. The concrete or sand mix bags do not shrink like gyp.

Best to fill those sleeper systems with a pea gravel mix when they go down to get some mass and tube connection.

scotter

I believe the OSB is tounge and groove, so I don't think the gyp would leak to the garage below. But, good to think about. To avoid the shrink of gyp, maybe floor self leveler would work. I think it is heavier, but the void isn't very big. I don't think that would shrink.

hot_rod

Gyp will leak thru a crack the thickness of a business card. They use regular sheetrock spackle to smear over any cracks before they pour to prevent leaking where you don't want it. It a pretty soupy mix so it can self-level.

It's a lot like spray foam, it finds cracks you don't even know you had.

Something like a giant caulk gun to squirt in self leveler?

I tried one job with gyp between the sleepers, came back the next day and it was 1/4" below the sleepers, so no contact to the sub floor or final flooring:(

They make expansion cement that swells up. Used for placing anchor bolts, maybe be that is worth a small sample test.

Jon_blaney

Have you considered increasing the water temperature? Maybe a problem with hardwood floors?

STEVEusaPA

Did they put foil tape between the sleepers (underneath the tubing)? Is it insulated tight under the floor/tubing?

scotter

no foil under the tubing, but insulation under the floor.

I took the setpoint temp up to 133 on the lochinvar NKC199 for a time. The heat stripes in the thermal images were very strong and the floor temp on the stripes was 85+, but the temps between stripes was lower. Overall it seemed too hot for wood floors. I have it at 120 now. I have to consider the zone that is embedded in concrete in the basement.

hot_rod

What you have is sort of a joist bay installation, where. The tube warms the space in those sleepers.

If the tube is stapled down, or somehow touches the upper layers, that is conduction heat transfer

If it is installed with aluminum transfer plated, the heat is transferring to the floor that they are connected to, and warming all the wood by conduction. With that method the better transfer is downward😀, so good insulations needed below.

Can you shoot the floor from below with IR, just for grins?

scotter

As best I can tell, the tube is stapled down, which would transfer heat down. There aren't any metal transfer plates. There is sheetrock below, so I cannot get a good view from below (other than the access I created to see the insulation levels. I think I will assemble some materials to funnel the mixed slurry this week to do a test pour of one bag.

PC7060

Interesting concept.  I’d set up a vacuum pump to pull the slurry into the space.   Depending on the viscosity of the material and how well sealed the cavities are you should be able to do large sections at a time 

hot_rod

here is the product I would try. It goes off in a matter of minutes and it doesn’t shrink

rick in Alaska

I would absolutely not want to inject gypcrete in between the layers as there is no way it will dry before it destroys the osb. Gypcrete has a lot of water in it and will make that osb grow something incredible.
Rick

hot_rod

What is the sub floor material? If it is Advantech or plytanium, any of the top line subfloor items should hold up until the cure. I’ve seen those subfloor materials covered in water for weeks in new construction jobs during rainy season

I used Advantech for a small utility trailer, lasted 10 years before it started to breakdown.

evaporation is a good point, and where that moisture ends up.  The hydraulic cement I mentioned cures in 1/2 hour or less.

Have you seen that expanding pour foam they put fence posts in with?  Grind up some copper, aluminum, or gold and mix it in😉

Made the mistake of pouring a gyp job on a new home that had fiberglass batt insulation! Thought we would never get the moisture out. Lots of water in gyp mixed that comes out the top

When gyp pours on OSB they spray a white latex sealer first. Looks and smells like Elmers glue

hot_rod

A good visual of what transfer plates add to the conduction transfer. Expect the same with tube encased in concrete, cement, etc.

scotter

Great thoughts! The moisture/evaporation/cure problem is certainly cause for pause.
It may be that the only proper solution is to remove the top layer of OSB and install aluminum plates, then put back new OSB.
OR flow gypcrete once the top layer is off where I can seal the bottom and take care of moisture.
I was hoping for a repair rather than a reinstall.

hot_rod

The plates really want to contact the upper layer of osb, the graphic shown us an under floor application. you have a sandwich installation. So either fill the space with cement up to the top so it touches the upper osb, or something like the ThermoFin U fins. They could fasten to the sleepers, then osb over then, same result as the graphic shown above, but built from top down.