Rigid Foam on Roof with Added Overhangs
Hey all,
I apologize in advance for the drawing I’m going to attach to the bottom of the email. But as my initial house plans (zone 3 California’s High Desert) near completion, we’ve set our attention on the details regarding air sealing and insulation.
I remember watching Matt Risinger build a roof with a layer of external insulation on top of the WRB and there was a system of rafter tails outside the sealed sheathing that were notched to fit over what he called “monopoly framing.” That meant the house was built without overhangs…or rafters piercing the outside wall.
I have a feeling there are details for this somewhere on the site, but I can’t find them. So in trying to explain the concept to the structural engineers, I made this rough drawing. They said that forces of uplift on the false rafters would stress the narrower tails that were fastened through the roof deck, so this wasn’t actually a viable plan.
I know I’m not breaking new ground here…so that means that I’ve done something wrong in my drawing. It’s taken me 60 years, but I finally “know” when I “don’t know” something. Can anyone look at this drawing and see if there’s some fundamental aspect of the detail that I’ve left out?
thanks so much…I really appreciate the time you guys spend helping inform amateurs like myself.
Mark
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Couple of things:
1. Is it possible you have a wind load that makes this more difficult in your area?
2. If I recall, Matt's house used LVLs for this application, which would make the narrower portion somewhat stronger. No idea if it would be strong enough, in your engineers' view.
Hi Mark,
I can think of a few things that are worth discussing. The first is that your structural engineer probably doesn't want to bother with things outside the norm. Doing anything beyond osb/paper/tile in your area is already going to be an uphill battle. The next is that they're mostly right. A 2x4 with a section removed, lvl or not, really just isn't that strong. I'd wager it's even worse with a bolt compressing the center plies apart in a lvl. However, that's ok. You may need to explain further that the structural sheathing is below the insulation, and everything above can be considered sacrificial/cosmetic when it comes right down to it. Sure, a high wind even might take the metal roof right off, but it wont effect the strength of the structure in the same way it would in a roof built with trusses. If you have to, point them directly to the youtube video so that they can get the idea for themselves. There's plenty of ways to approach fake rafter tails without making them this way.
Kyle and Patrick...
these are great comments, thanks. And Kyle's point about the structure vs. the cosmetic aspects of the roof are dead on. I'll push a bit and we'll see what happens. So far we're trying to use pin foundations, lime plaster and wood fiber insulation on the outside...so it would be a shame to have to compromise on this. At 4500 feet we have some of the best/biggest diurnal shifts around, so a well sealed house can be flushed at night and stay cool throughout the day.
Thanks again. GBA is an incredible resource. I appreciate everyone who takes the time to help me and all of us navigate the next generation of building.
"They said that forces of uplift on the false rafters would stress the narrower tails that were fastened through the roof deck."
I first read this as wind uplift on the rafter entire, but it may be worth clarifying that they don't mean the uplift force placed on the narrow portion of the tails when the overhang is under downward gravity load.
I assume you are going for exposed rafter look? And this is a flat roof? How long of an overhang are we looking at? How much exterior roof insulation is that, and are you sure it meets the requirements in your climate zone?
The 90 degree cut in the false rafter can create horizontal shearing forces which can propagate as a horizontal split (see 'dap' for one way old timers dealt with this with joists and the like, though not applicable to your use here). I would be somewhat surprised if such forces would be in play here, but maybe. I would think some positive fastening of the rafter overhang to the structure near the bottom of the false rafter would alleviate some uplift concern.
Could you link to the Risinger video you are referencing?
Adding false rafters to the exterior of a fully-sealed house is characteristic of the PERSIST system. If you do some google searching, you might find some details you like.
Your engineers are correct, but not very imaginative. Some areas of the high desert can see 100 mph winds, and you do have to design for local conditions. Uplift forces on a wide overhang can be pretty large. So design in a support/hold-down system. Diagonal brackets have served this purpose for centuries, bracing an overhang from both gravity and uplift forces.
But wide overhangs are getting somewhat out of favor in wildfire zones. I suspect this may be an issue in the high desert as well, though it might be possible to design a sacrifical overhang system. Not sure I'd want to put my name on that without some serious consideration, though....
We’ve done a number exterior roof insulation projects using foam held on by 2x4s on the flat, extending them for the overhangs. This way, the full dimension is used in the overhangs. Also helps to cantilever the sheathing, meaning it should extend up at least the distance of the overhang. We typically run a return 2x back to the wall to provide support for the soffit. You can adjust for different facia & shadow sizes. This works fine for overhangs up to about 18-24”.
I find it interesting that after doing it, this builder concludes that he would go back to conventional eaves.
https://www.youtube.com/watch?v=JR7sCKlChgY
I noticed too, that the cut outs in Matt Risinger's house are cut out of the top and not the underside of the rafter tail. So the rafter tail provides a support for the external insulation. I've followed Corbett Lunsford a ton...he and Kimchi and Kraut are great sites.
I'm going to watch this and reassess the design. I think we can still pull it off ... and diagonals would be a great solution.
thanks to everyone!
Mark
Hi,
If you are in a Wildland - Urban Interface zone - the rafter tails would need to be 4x6 or larger to count as heavy timber so they could remain exposed. If that is the case it seems like enough lumber is there. The wind up lift on an eave is less than 10 lbs per sqft of eave. - a Simpson SDWS screw can take that uplift. A SWDS screw driven just near the notch can keep the timber from splitting (with upward tension) and would be good form just to keep the outlooker from splitting when drying in the desert.
Tim!
Thanks so much. It's a great call...and we're in a wildfire zone...so I'm sure you're right on with the sizing of the rafters. The add'l thickness and most likely reversing the notch (if necessary) should do the trick. Again, I want to thank everyone!
Also...watching Corbett Lunsford was instructive. And I agree with his logic completely. It's just a matter of figuring out how to stack a bunch of external insulation on a 1:12 roof, exposing the beams and not making the whole thing look a bit cartoonish...which is really the only reason one is notching beams.
thanks again
Mark
In San Diego County they allow 4x6 rafter tails - check with San Bernardino county - it may need to be 4x8. You will also need wildland approved vents for under eave. If you can make a small soffit so it can send air to several rafter bays with only one vent - that will save $ on the vents
Do you plan to have the metal roof exposed on the underside of the eave? I used a upside-down steel angle as a fasia on a project. like that.
Tim,
I'm putting together a list of code stuff for my code guy in SBC DBS...I have to say, during covid they've been dynamite about getting back to me via email. I'll check on the rafter tails...and I'm not sure about the soffit detail. The roof will have an air gap beneath it to keep the heat off the roof...and allow for circulation. So as long as I leave a vent for that, I could just let the steel extend out over the rafter tails. With Cor-a-vent something to keep embers from getting under the roof.
Mike Guertin over at Fine Homebuilding illustrates two approaches to false rafters in this article: Adding Roof Overhangs. Although he doesn’t get into insulating, there are two drawings that may be of help to you.
The metal roofs I have specified have been placed directly on the roof sheathing. In order to get a Class A fire rating - you need a rated underlayment (typically fiberglass based roll) or board type densglass underlayment. The roof mfgs have a UL listed assembly to follow for the Class A rating. You may be off the books with the air gap directly under the metal roof.
The roof vents would need to be California WUI rated. I think some plastic ridge vent are approved. https://www.brandguardvents.com/ is a metal vent supplier that is approved.
I worried about the air gap...the ventilation is positive....but embers are the killer (literally) up there. So it's going to have to be well and truly sealed off from embers a lot better than the houses we grew up in.
I'm going to scale back my wish list a bit. The consensus on the site seems to be...first from Corbett Lunsford....that the extra work of the tails isn't necessarily worth it. And the external insulation on the roof deck isn't going to be Passive House/Northern Canada worthy...so it's probably one of those steps that falls very close to the "Not worth the trouble" line. I do know the ventilation from rainscreens is noticeable in the hot weather. It make a real difference not to have the surface that's getting sun is direct contact with the surface that's being insulated. If there's a way a can do that and not burn the place down next time fire burns through there, I think I'm just about where I want to be.
I'll check the details on Roof manufacturers and check with the county...I would think that the underlayment would be there to protect the house, I might put the air gap between the metal roof and the underlayment...which might be as simple as a roll on rainscreen, if it's fireproof.
Hi, To take a step back and look a the assembly - you could add cross strapping on the ceiling to get a thermal break an get the desired insulation system. Or go to a truss that would give you a greater depth on the roof member to allow for insulation and a vented attic space.
The roll underlayment for metal roof is GAF versashield. It is used with metal roofs to get the class A rating.
Tim,
I think the cross strapping (if we're talking about the same thing) on the underside of the rafters gives me support for the insulation and the service cavity I want for services. I think the Versashield is a must. I've got a call scheduled with the structural engineer to see what we can iron out. The truth is also if I can draw some braces into the design on the facade, a lot of the structural issues are less problematic. I've found some nice options in "Passive House Details" (by Corner, Fillinger, Kwok - Published by Routledge) that I'm going to talk through with the engineers. And ultimately, fire safety is going to be the tail that wags this design dog.
And for the designer in me...the width and weight of the fascia board is part of this discussion as well. The challenge is to insulate the roof sufficiently and not end up with a cartoonish layer of icing on the cake.
Any advances on this? Especially looking for a detail for applied exposed rafter tails over wood fiber board insulation (a “chainsaw” retrofit, wood fiber board above air-sealed sheathing).
Risinger did it with enclosed soffits by using the ripped down LVL studs.
Not what you're looking for exactly, but I have recently used a plywood, 1x4 lattice, plywood sandwich on top of 5" of foam to build ~12" overhangs. I've also used plywood with a 2x4 lattice on top, followed by a metal roof to retrofit overhangs on a vented roof. No building inspections round here to tell me I'm doing it wrong on my own house. This assembly keep the overhangs thin, like 3-3.5 inches. I've done a few hand calcs to ball park the strength and the overhang deflection under full snow load seems reasonable.