So here’s the deal, Got the design from the SIPs folks and with all the 2x in the splines I might as well stick frame the roof. I was thinking SIPs for the fact that thermal transfer is minimized but with double 2x every 4 feet whats the point.
So I seek out wisdom and experience. What I am thinking is to stick frame the roof, cover that with 5/8 ply then a layer of blue/pink foam and then another layer of plywood. No sleepers just ply foam ply. Securing the upper layer of ply to the rafters with 3.5″ nails. Roofing on top of the ply will be comp shingles.
We are building a “hot” roof, no venting. Straping the interior is not an option. The main reason for the sandwich is thermal transfer.
I am building on the east slope of the Cascades in Washingon. Last few days temps have been over 100. This past winter we had over 5′ of snow on the ground with temps as low as 20 below 0.F
So the question is will this work or will the upper sheet of ply fail or will the roof end up looking like a lake on a windy day with ridges and humps from the foam and ply combo. Anyone do this??
ps. the archy and eng. are both useless on this.
Thanks for the thoughts
Rob
Edited 7/26/2006 11:19 pm by level
Replies
You may wat to look into nailbase. Many of the sip companys make it. Basically a nonstructural version of the sip. I beleive it only has osb on one side.
I am curious as to why you need double 2x splines. I just built an sip house and only needed osb splines in the roof with the exception of the hip sections which needed the double 2x. Maybe a different roof design would help. What type of spans do you have?
Tom
Douglasville, GA
Thanks Tom,
double 2x are by the manufactor of the SIP's. Roof design is set in stone. How I build it is not. Spans of 7-14 feet. 85 lb snow load 2/12 pitch.
Non structrual sips would involve more exposed beams, and thats not an option. stick framing the roof and then putting sips on would be xpensive and make for a tall roof assembly.
thanks for the ideas.
Rob
Whose SIP's are you using. All of my roof sections have greater than 7' spans and I also have a 85# snow load, although at 4/12 and 10/12 pitch. The only areas that have double 2x splines are the hips sections. Everything else is osb spline.
I used R-control panels. Quite happy with the project so far.Tom
Douglasville, GA
thnaks Tom
I'm using priemer building located in Fife washington, thought they seem to have offices elsewhere. I'm not sure why so much wood. I speced 12" panels but the drawings came for 10', with a phone conversation I was told I could go foam splines up to 8 feet but after that I needed to go with wood. They did mention that they build with a fail factor of 3, so I can spec what I want, i.e all foam splines, but I'll assume the liabiltiy if any thing goes wrong.
I assume that the snow load you build is for your NH place, haven't heard of that much snow in GA.
thanks again
Rob
The snow load is for the place in nh.Tom
Douglasville, GA
Rob - a couple of thoughts - the first is that what you are proposing sounds very complex and labor intensive, and your concern about integrity, ridges and humps is well founded. I did a similar roof once and it was a headache, especially getting the spikes (I used 6") to hit the rafters. Edge details were another issue.
2nd, one of sips primary advantages, next to insulation is air infiltration. 2x's thermal transfer is probably negligible next to the advantage you gain with sips over typical construction. Anything over 5" and I would guess (just a guess) diminishing returns begins to be a factor.
3rd, if your design allows, could you simply do a visible frame (you are in large timber country after all) roof structure and sheath it with nonstructural (cheaper) sips? The savings might pay for the wood for the rafters and ridge. You don't need to cut mortises etc to "timber frame" - even doubled 2x done well can be very attractive. BTW - what is your wall design?
Finally - let us know what you decide and how it works out.
Hey Brian,
thanks for responding. Complexitty doesn't concern me, but final product does. this house, du to design and location has great thermal gain and loss potetnial. I want to minimize it. the house has big timbers, 16 and 18' glulams on 8, & 12 foot centers with two funky valleys and a ridge that will increase spans to over 14 feet. How 's the roof you did working out? any long term issues?
I'll keep you updated on which way I go,
Thanks Rob
The roof... its OK, but we used the white foam back then (tight budget) - if you go this way, use the blue or pink stuff, and long screws sound like a better idea than spikes.
I agree with the post that recommended the 5/8 and screwing to wherever - I'd want a lot of screws in this! Staggered joints - should be very tight. What R value are you pursuing?
I would do a thorough cost/time comparison - the sips are still quite efficient, even though its hard to bear the idea of a thermal transfer path with the lumber.
Treat every person you meet like you will know them the rest of your life - you just might!
Thanks for the update,
I too like the idea of the screws. As for r-value, in the roof it's self (rafters) with blown in insulation I'll get in the mid 40's. On the roof with foam I'd be happy with 5 as a thermal break. Four inches of foam would give me 20, and that with the 40 would be plenty. As for the cost with 1 inch of foam my quick calcs show I'd be ahead of the game, I'll get some more prices tomorrow.
thanks for the thoughts
Rob
Edited 7/26/2006 11:17 pm by level
This is what we do. Here's the drill. In your climate don't use the Sips or the nailbase. use 5/8" decking, two layers of 2" thick polyisocyanurate with the joints staggered. Use 6" screws with 3" washers to screw the 1/2" top layer of ply to the 5/8" first layer of ply. Don't bother trying to hit the rafters. The 5/8" is good enough. Edge the perimeter with 2x6 ripped to 4" to give a solid perimeter. This is a great system. We use an ice and water sheild on the 5/8" ply to protect the structure. We use something less on the top layer to keep out water but if the top layer rots it can be replaced without affecting the structure.
The problem with one layer of foam in your climate is that a convective loop will transfer interior moisture to the top layer of ply and cause deterioration. The problem with Sips is that the same as one layer of ply with two added problems. The sips will fail if one layer deteriorates. The other problem is that in winter, one half of your structure has a higher moisture content and in summer the opposite layer has a higher moisture content. This causes seasonal movement that racks the layers and allows for air leaks. In Alaska, the results were horrific. In your area they would be less so but would exist.
Gotta go.
Thanks for the info Ray,
I'm aware of the SSIPS falure in Alaska, the powers that be tell me it was an instal problem. I'm not sold either way. I like your technique, how many roofs have you done this way/ How long have they been in use? Whats the thickness of your screws and whats your schedual on their use. any issue with highwinds or shear?
sorry for the questions but you seem to have the info I am seeking and I need to be able to answer question from owners and inspectors, and my own curiosity.
thanks agin and I look forward to more info
Rob
Start by looking at the Atlas roofing website under their acfoam nailbase installation pages. After getting an idea of the requirements there, consult with your structural engineer. I have done three roofs with this method. All my other projects lately have been with spray foam to the bottom of the roof deck. I prefer the panels.
thanks for the info,Atlas was a new one for me, good info but rotten experience on the phone so far.
Rob
Ray can you elaborate on that a bit.
>a convective loop will transfer interior moisture to the top layer of ply
Where is the loop? In the foam?
>SIPS...one half of your structure has a higher moisture content...
In both of these you seem to be talking about water vapor movement within the foam. Do I have you right? I guess I hadn't paid close enough attention to the perm rating available in the nailbase/SIP panels.
splat
The gaps between the foam will allow a complex system of paths for air movement to be created. Warm moist interior air will move through small cracks into the system. It will then rise to the cold exterior surface where it is deposited on the cold surface of the exterior layer of plywood. It happens slowly and constantly during the winter months. If you can perfectly seal the joints between the plywood panels from underneath, then the convective loops will not deposit moisture and will only result in a very minor heat loss.
The same thing happens in a SIPS roof. It is virtually impossible to install a SIPS roof in an airtight manner such that air from below cannot enter the system. Once there is a path, air will enter the joints and deposit moisture on the cold surfaces. The convective loops occur within the gaps and joints at the panel connections.
In Alaska, it was found that the cold exterior surface became very dry compared to interior surface, except at the joints. This causes the exterior skin to shrink relative to the interior skin. In the summer, the exact opposite occurs. This results in a system that slightly heaves from season to season. When this happens, the joints loosen up further which allows more moist air to enter the system in winter. Eventually they had interior moisture forming ice between the panels which further drove them apart and before long, they had disaster. It is easy to blame the installers but what does that mean? Is the system too hard to install correctly or is perfection required? The fact is, that it is very difficult to apply the sealants that they call for underneath and on all surfaces without them being messed up when the panels are driven together.
Sips simply suffer from having two structural skins that are subject to expansion and contraction due to moisture content. With one skin seeing only the outside condition and one skin only seeing the inside condition, there is going to be substantial seasonal movement in any climate but especially in extreme climates.
Level, welcome to my world..
I did the same calculations you are going thru.. I found that by being creative I was able to have it both ways.. First.
You are right to avoid thermal transfer with sticks.. Sips do save!
But now you need to be creative because as you've discovered SIP's won't carry the load.. Not over long spans that modern houses like!
What I did was timber frame the inside of the house then the SIPs could be supported by the timbers underneath..
Now before you get all worried about costs it was cheaper when I got clever. Buy the logs or timbers direct from a sawmill. If you like the looks of logs then have the sawmill just flatten just one side. Timbers are easier to install and if you are clever much much cheaper..
You can do a hybred whereby the timbers are supported by stick built outer / inner walls. a Semi whereby the timbers sit on top of timbers that sit on floor joists or a full timber frame whereby timbers sit on timbers which sit on timbers.. (horizontal- vertical - horizontal)
Ceiling loads can easily be supported every ten feet so if you count inner walls and such the number of beams or timbers is actually pretty small..
Hey Frenchy,
Thanks for responding. A redesign is out of te question. We have all the exposed beams we are going to get . The spans are what they are. I'm just the builder who's going it alone after the archy was let go for a variety of reason, and an engineer who doesn't want to deal with issues such as thermal bridging. Job's like this make me dream of becominging a design build firm.
thanks
rob