Screed Over Glavel Foamed Glass Slab
I’m looking for some help to determine the most suitable (and available) aggregate to use for a screed layer. These are the planned layers bottom to top:
*compacted stone
*geotextile fabric
*compacted Glavel (~12″ thick)
*geotextile fabric
*screed layer (1-2″ thick)
*Stego 15mil VB
*2 layers 3/4″ Advantech
*finish flooring
I have watched or read most of the concrete-free video and articles available online, however my assembly is a little unique in that there is no insulation board.
I’m leaning toward 1/4″ minus gravel or coarse sand. My concerns are:
1. The screed layer needs to be walkable without creating a mess from boot prints / knee prints while the Stego VB is added/taped.
2. From what I have read, compacting this thin layer would not be needed. I’m concerned about any stretching that might occur within the geotextile fabric below the screed layer over time… potentially leading to settling / sagging in the flooring.
3. If 1/4″ minus is the best option- does anyone know a source in Maine (preferably Southern Maine).
Currently the Glavel is compacted and my rough plumbing is awaiting sign-off, so screed layer is next up.
Bonus question: Any collated screw recommendations (Senco) for attaching 2 layers of Advantech together for a floating subfloor? I know they can strip out easy with such shallow engagement. Or even non-collated suggestions that you may have had good success with? It’ll have to be 1 1/4″ or 1 3/8″ screws.
Thanks!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Steve, I recently advised a builder with a similar situation. Glavel is great but difficult to impossible to get the surface perfectly level at the perfect elevation. What we settled on was using two layers of 2x4s on the flat over the vapor retarder, shimming between the 2x4s to create a level surface, infilling with mineral wool batts, then a single layer of Advantech. The mineral wool cuts into the carbon savings of the overall system.
If you want to stick with your assembly, I think dry sand would be fine. Are you north of Portland? Dugas Excavating in Yarmouth has been my go-to for 20 years.
Thanks Michael! You've given me a good option to consider. I believe the elevations will be the challenge (Glavel is already complete). Although, maybe rips of Advantech, shimmed between them, instead of 2x4s would work for the sleepers. Seems like Advantech rips would be less susceptible to movement as well (Cupping, bowing, etc.)
Generally I understand that unfilled air spaces are not desired within assemblies, but would there be any downside to not infilling between the sleepers? (I already have a R20 floor with 12" of Glavel.)
Yes, I've used Dugas, they have 1/4" with no fines. I've yet to find a source for 1/4" minus. I'm doing some testing with different aggregates now. My concern with sand is that I need to walk on it.
Steve, leaving the spaces unfilled would probably be ok in your case, as long as the perimeter has a continuous thermal layer. You don't want a thermal bypass between the Glavel and the sheathing.
Here's an out-of-the-box idea: get clean 1/4" stone and course sand and mix them together. It wouldn't be a perfectly graded mix but the sand should fill the voids in the 1/4" stone enough to keep it from moving around. Do you have access to a concrete mixer? That would make it easier.
Edit: I was just emailing Dugas about a project we're doing and he said they have 1/4" minus in stock again.
Thanks Michael. I'll give them a call, that may be the shortest path for me.
You've given me some good Plan-B ideas though if something doesn't pan out. I'll let you know what I end up doing.
Unfortunately the 1/4 minus that they have isn't really 1/4 minus from what I was told. It's pretty clean and doesn't have the fines. I'm going to the pit tomorrow to take a look. I may end up trying to mix aggregates.... along the lines of your suggestion.
Have you tried Pike.
Or Crooker out of Topsham.
Some places may have the fines (sump) that you can mix back in. It's probably more angular than sand which is good.
Also, have you considered just 'stone dust' which is kinda close to 1/4" minus.
I imagine 1" would be hard to keep from slipping around on the geotextile. 2" still, but better. You'll want the non woven stuff for sure.
Thanks maine_tyler. I appreciate the ideas. I looked at Pike's aggregate list previously and didn't see anything along the lines of 1/4"minus, though I have not called them.
I had forgot about Crooker. I just looked at their aggregate list online. They do have 1/4" screened sand. I'm assuming this is 1/4" minus... I'll give them a call tomorrow.
I’ll be interested to see how your project goes. I’ll be doing something similar for my foundation, although with an earthen clay floor over Glavel. This way it’s not as critical to get the Glavel (or screed layer) perfectly level.
I’m curious, what thickness of geo-textile fabric did you go with? I wasn’t sure if the 3” pieces of foamed-glass would tear the fabric.
Thanks, best of luck.
Adam
Hi Adam- I'll let you know how things go. I'm quite pleased so far with the Glavel for my application / goal (DIY friendly, no concrete contractor). I'm quite curious about your earthen clay floor as well. Has this been done before?
The geotextile I'm using is 8oz non-woven. In the past I've only ever used cheap landscaping fabric from local box stores for various landscaping projects. This 8oz non-woven fabric is impressive in terms of strength / quality.... I'm very confident that Glavel will not tear it, even with compaction.
As far as I know, it hasn’t been done before. The folks at Glavel, or Aero-aggregates haven’t seen it used under Earthen clay floors. And within the natural building community I haven’t heard of it so far. But it should hold up just fine, since clay floors are often installed directly over aggregate (with or without a vapor barrier between).
I think it’s actually a great fit for a clay floor, especially with embedded hydronic heat. Since there is no stone aggregate that would otherwise contribute to the thermal mass of the floor, then there’s only 2-3 inches of thermal mass to hold heat in the clay layer. So the interior temperature will hopefully respond relatively quickly.
I do kind of wonder if the 'no need to compact' is at odds with 'won't move.'
I've worked with a lot of fine surface aggregates for accessibile trail construction (such as stone dust, 1/4" minus, and custom mixes) and the types that compact and hold up the best are the angular crushed stone varieties-- but they really don't self compact.
Sand (or pea gravel) would be more 'self compacting' with the tradeoff being that it is more easily displaced with movement. I wonder if you could sequence construction (screed and lay plastic in sequenced rows) to avoid excessive disturbance.
Creative idea maine_tyler. I'll definitely give it some consideration.
Glavel suggests that their product be separated from other aggregate with geotextile fabric. I plan to reach out to them tomorrow to see if that's just a suggestion, or a requirement for some reason. If it's not a requirement, I'm thinking my best bet would be to add a layer: 1/4minus, 3/8 minus, etc. directly to the top of the Glavel, compact it flat, and maybe hand tamp any low spots. This would provide a walkable surface and would not settle. Then I could wrap the geotextile over it, which would cushion the Stego a bit (not needed but a perk). All speculative at this point... we'll see.
Thanks to everyone for your great ideas!
I've not used Glavel or any similar product, but I would think adding non-foamed aggregate on top of it that can work its way into voids could lower the R-value at the top.
I'd be curious to hear their feedback on your plan.
That would definitely lock everything together nicely, but it does seem like keeping fines from the screed layer out of the glavel layer would be important. Of course if it doesn't get rained on, there may not be a ton of fines migrating down, but its still probably not a by-the-book approach.
Thanks Patrick and maine_tyler.
The response from Glavel is that it's acceptable to add a finer aggregate screed layer directly to the top of the Glavel (without geotextile separation). Apparently some of their larger commercial customers have done this rather than adding thousands of square feet of geotextile.
Patrick- I hadn't really thought about any R-value implications to this approach. It's a good question. It may be flawed logic on my part, but I'm going to consider it as thermally comparable to a concrete slab over EPS assembly. In essence, I'll have ~R18-20 Glavel on the bottom, and then a top layer of less insulative material. Any top screed layer should only penetrate a few inches into the Glavel.
I believe this option will mitigate most of my concerns:
1. Concern over stretching geotextile, and resulting settling, under the screed layer is gone.
2. Concern over walkability due to a thin screed layer will be gone since it will be locked into and compacted into the top portion of the Glavel.
3. Concern over future settling of the screed layer due to having a thinner non compacted layer will be reduced.
Also, it should be an easy and cheap install. If there is a high or low spot it should be easy to add/remove and hand tamp as needed.
This is my plan for now, but feel free to shoot holes in it ;-)
Thanks again everyone.
A quick follow-up for future readers:
This is what I ended up doing for my concrete-free slab:
1. Base layer- bedrock / some existing soil / 3/4" stone. 3/4" stone added and compacted in lifts until it was ~14" below top of ICF stem walls.
2. Non-woven geotextile over stone and up ICF walls
3. Glavel. 12" in 2 lifts. Each lift compacted. surface of Glavel approx. 1 1/2-2" below ICF top surface.
4. 3/4" stone. A minimal layer spread by bucket / rake just to fill some of the voids in the Glavel, to better support the screed layer.
5. Trimmed geotextile to just below the top of the ICF
6. 1/4" minus. There was no 1/4" minus available in my area, but I needed something that was fine for screeding and would compact. I had the supplier mix 1/4" clean stone and stone dust. I added the 1/4"minus and compacted it very well until the final compacted surface was ~1/4" over the final elevation. It was basically walkable (with care) without leaving any significant footprints.
7. Trenched out 3 shallow spots for screed pipes. I leveled the pipes and checked them with a laser. I filled in the trenches around the pipes so they wouldn't move while screeding.
8. Screeded the floor flat and hand tamped a few spots. Added some sand and troweled a few spots as needed.
9. Pulled the screed pipes out and filled/compacted the trenches.
10. Hand tamped and touched up a few spots as needed and double checked flatness with a laser. The 15x15 floor was within an 1/8"
11. Added a thinner geotextile over the top.
12. Added Stego 15mil and taped / sealed to "flaps" over ICFs.
13. Added double layer Advantech subfloor, staggered and offset all seams.
It was a fair bit of work, but the floor is extremely solid, level, and flat. Time will tell, but I don't believe I will have any settling issues.
That sounds like success! Keep us posted as construction proceeds, and if you have any ideas to improve or simplify the system.
Steve41,
Nice work!
Thanks! and thanks to everyone for all of your ideas / help.
Michael- As far as ideas to improve or simplify (Ref item #'s from post 16):
#2 and #5: I'd leave the geotextile just short of the top of the ICF to start with. My original plan was to envelope the Glavel completely (recommended by Glavel), so I had a few feet extra that ran vertically and was taped to the studs temporarily. I ended up not placing a geotextile layer directly over the Glavel, so I had to trim the geotextile near the top of the ICFs. This was a pain due to the close proximity of the ICF and Glavel and the grit that had clung to the geotextile through the process. 8oz non woven geotextile is tough stuff.
#11: I added a 4 oz non woven geotextile over the screed layer for two reasons: To help protect the Stego VB and to reduce risk of settling / migration of the screed layer over the top of the ICF. There were a few spots where the screed layer was slightly above the ICF due to slight variation of the stem walls. I'm glad I did this (so far). It was also "cleaner" to work on the Stego with the geotextile under it. (taping / sealing).
#12: I hadn't expected the 15mil Stego VB to be so severely wrinkled. There were permanently set wrinkles which were quite severe. Likely fine under a slab, but more of a challenge to use with this assembly. Beyond vapor sealing, I wanted the Stego to be taped with no holes/gaps so that I didn't have to contend with high Radon. All of the wrinkles proved very challenging to seal (ref. pic) . I would investigate a less wrinkle susceptible product if I had to do it over again. (10 mil?). I used zip tape and Contega HF on the Stego. I would not use Stego's red tape, it's just not sticky or robust enough and will gap at any stressed area in time.