This construction project is being done over 10 years as $ becomes available.
Doing RFH on the main floor, the basement has RFH completed, and working well enough to make the upstairs comforatble in the toughest of winters. So finishing the upstair HVAC has not been a priority.
Have a 5/8″ CDX sub floor (main level over trusses) that was supposed to get another layer of 5/8″ sub floor with seems overlapped. Did not do the second layer because of door height concerns if using gypcrete or similar product. Got advise that a second layer was not necessary if using gypcete as it (gypcrete) would add rigidity.
A year later and the 5/8″ sub floor is bowed everywhere you walk and it flexes at joints.
Don’t think that laying gypcrete over this is the long term solution because I’m concerned that pouring gypcrete over an unlevel and flexing surface is not a good thing. Still have height concerns, though.
Is there a product that is hard (like epoxy) and liquid (self leveling) that I could pour onto (original 5/8″ cdx) to get everything back to even, before installing 2nd 5/8″ cdx? (Back to original plan)
If yes, I’ll hold my breath with the gypcrete and final floor heights to stay under the door openings.
Replies
I would put 5/8 plywood glued and screwed over the subfloor and forget about the gypcrete. Run your RFH under the subfloor. We routinely use aluminum fins fastened to the tubing and insulate below the tubing. This works fine even under hardwood floors. Alternatively you could block under all the plywood joints and go with the original plan.
ultra-fin?
I am not sure of the name since our plumbers handle this. The fins come in two pieces, are riveted in place and are about 8"x8" when assembled. The length of tubing is reduced by at least half and runs through the middle of the joists, not stapled to the subfloor. The joist must be insulated under the tubing.
it's Ultra-Fin. How do you like it?
Very much but it would be nice if it could be designed to eliminate the riveting.
First, you might wonder why this happened. I'm assuming that it was nailed on solid lumber joists, not I-joists. Was it glued down? Is there blocking to keep the joists from racking? Why is the floor system flexing, is it undersized for spans? I can't see adding the weight of gypcrete if that is the case.
Schelling's advice about the radiant may work out well from a heating standpoint but I'd be worried that adding another 5/8" of plywood wouldn't take care of all these issues if something else is wrong to begin with.
This is a circular design with the trusses meeting at the center and spreading to 24" on center at the wall. The actual gap is 22" from inside of truss to inside of truss.
Because of the the truss / circular configuration underfloor with fins is nearly impossible.
Why the flex. Because near the outside (4' from the outside) is the first seem in the bottom layer of 5/8" cdx. At that point the span is probaby 20". Had I immediately glued and screwed another 5/8" cdx my effective subfloor would have been 1 1/4. The plans call for the second layer spacing to not match the first, keeping seems in different spots. Should have eliminated any potential for sagg. But alas, I didn't, so what do I do know?
The struture is adequate, because I had the trusses beefed up to take a 1" concrete floor. Load factor was something like 20#psi, don't quote me on the latter, memory has faded with age.
Interesting.
Just FYI, a load factor of 20psi (not quoting) would be no where near strong enough to support normal floor loads, any structure above, much less a layer of gypcrete!
But let's say you've got plenty of support. Your only problem is going to out near the perimeter. Can't you just address the problem in this area? If what you say about structure is true, the inside of the wheel must be dramatically oversupported, and WAY capable of handling anything.
Who designed this?
Edited 6/3/2003 9:43:06 AM ET by Mad Dog
I agree the center is way capable of handling anything. Don't brag about the design too much because the Structure is built in 8' panels by a company in North Carolina called Deltec.
Used ICF's for the basment then framed the structure when the panels etc. arrived on site. For you pro's this is probably not something you'd do, but for a guy like me who will be a general, laborer, electrician, plumber etc. for the one and final house of his life, I liked what I saw when I visited the plant.
The lumber looked better than anything I'd ever seen in the yards, the walls were all built in jigs and the tolerences were tiny.
With each sub, part of my deal was that they hired me back as a laborer. I've learned a lot and really come to appreciate fine home building, because it doesn't (didn't) happen every day.
As far as the load's all my paperwork is up north at the house and while I can't remember the terminology, I am comforable that it would hold a concrete floor.
Don't know how to only address the problem at the outsides, but you are correct, that is where the problems are. Supporting from underneath is a pain in the butt, because of the slight outward angles and the need to custom measure and cut each support piece. Life's too short.
That's why I thought about some kind of pourable epoxy type stuff to smooth the base before laying on another layer of 5/8"
Hi Bick,
I'm thinking that maybe you could solve most of your problems if you heavily glue down 3/4" tongue in groove subfloor, (I'd use firply in this case) and screw at least the outer two radial tiers, if that is how you lay it.
The extra eighth inch will help to straighten the floor and control the wobbliness you have now. You should belt sand the high spots where the moisture has raised your seams first. You might consider adding blocking to the floor joists wherever it seems most needed, and especially under the seams where the first layer did not fall too well across the joists, because I'm doubting from here that you sawed the edges in pie shapes to land on the joist centers? Therefore some more weakness built in to the outer rings of ply. If you can jam the blocking up into the first ply layer from the bottom by shooting toenails into the joists, it will be solid and may even help correct some of the sag. Then facenail as usual. If nothing else, it will keep the following fix from developing the same sag.
How low do the dips in your first layer seem to be? Put a level down and look at the gap. If it's pushing an eighth inch or more, you'll have to address this but it's a pretty easy indicator as you go. You could take a sharpie and circle the whole area, and mark how much to shim between the joists, rip a whole lot of shimstock from some 2x scraps, and glue that all down, then set your new plywood on and go from there.
Lots of construction adhesive, lots of screws.
Good luck
Thanks for the input. Agree that 3/4 is the way to go. Need to figure out if that will raise the finish floor above the door threshold. That woud be bad, eh.
Believe it or not, the package ships with the plywood precut into pie shapes. The whole thing fits like a puzzle.
Like the shim idea, but really don't have much in the way of dips. The dips only happen when standing on a seam out near the perimeter. Not to say there isn't some dip that stays around when not standing on it. Mostly in the kitchen and other high traffic areas. It's those areas I worried about when putting more plywood on top, whatever the thickness of the top layer.
Thanks again, you've given me some new things to think about.
Pretty bizarre, pie shaped plywood! Anyway, I'm thinking your troubles are soon to be over, and I would slap that plywood down, overlapping the offending seams, glue as previously recommended and just blast away with ring shanked nails, 3/4" ply will take care of this.
If an extra eighth of an inch takes the finished floor over your threshold, hmmm. That's also bizarre. But you could remove the door , add the ply and reinstall on top of the new stuff.
What was the maximum span between joists at the perimeter anyway?
24" on center at the perimeter, I think. The only measurements I remember doing were from inside to inside of joist and that was 22 1/2. But if my math is correct that ends up being 24 OC.
The framer was real concerned about how much I added to the floor whether in plywood, gypcrete, concrete or whatever. The stairs coming up from the Basement front door has consistent rise, because the top floor had a 2x surrounding it and he assumed (as did I) that the ???crete would come up to that level. That way when adding finished floor the rise from steps, to landing, to main floor would remain consistent.
That's another reason that I would like to stay withing my alloted 1 1/2 inches for next layer of sub floor, tubes, thermal mass etc.
Find a flooring supplies distributor, DAP, which was bought by BOSTIC, has a product called 80P it is self leveling portland cement base to which you add a latex milk emulsion for adhesion and flexability, Ardex also has a similar product. You can try http://www.goldingssupplies.com
Thanks for the suggestion. Have sent an email to DAP to see if they agree if the latex version that pours can be used in a Radiant Floor Heat application.
What does Deltec say? We almost put one up but family medical issues sent us in another direction. I went through plant, met their engineers, etc. I was very impressed. How many doors are you talking about for threshhold issues? I would consider rehanging doors and doing the 3/4 beef-up. I ended up build a 38x48 addition to existing house, used 12" ladder trusses with 3/4" t&g osb subfloor. My crew did not do very good job putting down subfloor, they tried to go fast, gluing, corner nailing, and coming back to nail off. Not a good practice, unless you get back before the set time for adhesive. Getting three inches of rain just after did not help. They also thought they were doing a good job by getting tight seams.
If I ever did it again I would use 16" trusses and plywood. I get a little bounce with my 19' span.