Details for 4-foot-wide brow roof?
CZ6A, 45 degrees Lat., Ottawa ON
Finishing up the details of our new build. I’m going over my plans with my Engineer and he has concerns about the integrity of my brow roof design. Specifically, he’s worried about snow load and sag (creep) over time. This is for the south facing wall and its large windows.
Do any of the builders here have experience with design details for a 4ft deep brow roof on a flat two story wall?
My two story design has 9ft ceilings and 6ft tall window openings on the 1st floor centered on the height (18″ from the floor). At 45 degrees North I need a 4ft deep brow roof about even to ceiling height to effectively shade these south facing windows throughout the summer. Roof slope is 8:12
After fumbling through the code calculations for snow load, it seems my roof should be designed for 2.2kPa, or 46 lbs/ft2 + 5 lbs/ft2 (for solar panels) = 51 lbs/ft2. So a 4ft deep brow roof would support about 204 lbs/ft, or about 408 lbs/truss on 24″ centers.
My Engineer’s concern is the moment (bending load, 816 lbs/ft) that load would place on my exterior 2×4 wall. Any insight would be helpful!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Just to be clear, this is what I'm referring to as a "brow roof", the roof that shades the first floor windows:
Lance,
The reason that a designer or builder hires an engineer is to help with the details. If the engineer is working for you, ask the engineer to draw the necessary details for the roof design and attachment.
Lance, "brow roof" is a fine term, but if you're interested, the proper term is pent roof.
In case it helps you visualize things, bending moment is essentially the same as torque. (Torque is an applied force vs. bending moment is an internal force.) Rounding off your numbers, per linear foot of roof you have a 200 lb force applied 2' from the wall (at the center of the pent roof), or a 400 ft-lb force. The wall has to resist this force with a lever distance of 2.67 ft (from your 8:12 pitch), or 600 lbs. With studs are 24" o.c., that means a 1200 lb horizontal force on each stud. That's a pretty big force. An engineer should be able to design a solution without much trouble, but it will probably involve hardware, and something other than standard 2x4 studs.
Martin, I agree that should make sense, but I’m a “special” customer with specific requirements.
His original proposal included attachment points inside the wall, which would have made for an extremely difficult (if not impossible) shell to air seal, and chock full of thermal bridging. My proposal is to attach small triangular trusses to the outside of an already sealed envelope, sealing the fasteners as I go.
He would like to see support posts now, but I would still like to consider a post free design if that’s possible.
Lance, is there a reason you can't add brackets below the roof? We hit 0.25 ACH50 on this house, even building over an existing foundation and using some double-hung windows at the owner's insistence.
Lance,
I second Michael's recommendation of brackets. Perceptually, to make people comfortable, loads need to be seen to be taken-up. Thats why columns have almost always been designed by architects to be much larger than they need to be structurally. A pent roof without brackets or posts often just looks wrong.
Pent-roof it is then, thanks for the correction!
Also, I got my torque units wrong (thanks Michael), it should have been ft-lbs, not lbs/ft. Beyond that though, I'm not following your math?
Using rounded numbers, my 4' deep roof load is 200 lbs fer foot of length. 200 lbs force centered 2' from the wall is 400 ft-lbs, divided by 2.67' truss height is a force of 150 lbs per foot of length pulling outwards at the top of the truss. Multiply by 2 for trusses on 24" centers and we have a horizontal force of 300 lbs per truss.
I used an on-line beam deflection calculator to model that as a point load cantilever; 300 lb load at the end of a 32" cantilevered 2x4 comes up with a deflection of 0.028", or an angle of 0.06 degrees.
This certainly doesn't seem like a crazy situation, but I'm not well versed in creep in lumber structures. This is where I hoped to find someone who'd done a similar unsupported detail.
Michael, the detail on that house you did, how deep was that pent roof and what was the slope? Was that an Engineered detail, or just something you put together?
Malcolm, I have to respectfully disagree. Architecture often tries to visually defy the laws of physics, pushing the limits of unsupported spans and cantilevers. Many modern designs rely on steel for support as wood products just don't have the strength to handle these extremes. Architecture is subjective, like anything else, so what looks right and wrong depends on the audience. I think unsupported structures look really cool, so long as they stay put! :-)
Lance, you're right--I did the math without writing it down and did it backwards. The horizontal load on the stud is indeed 300 lbs.
Modeling it as a point load on a beam, which is what the situation would be if you turned it 90°, the deflection is more like 0.8", or L/120 on an 8' stud, assuming the load is applied 2.5' from the end on an 8' stud. (I wrote the math down this time, using standard engineering principles, as I always do when I do my own engineering, and can explain it if you're interested--calculating deflection from point loads is trickier than for distributed loads.)
The pent roof on my project is 3' deep and a 2:12 pitch. At least that's how I drew it; the builder may have grown it slightly. The top lands at the rim joist so the connection concerns were different than yours.
I agree with Malcolm that visually supporting the load often looks best, but I also understand that modern design revels in visually unsupported elements. I tend to find them disconcerting but it's your house. Good structural engineers are accustomed to working with architects' crazy ideas; this does not seem like it should be that difficult to solve. For example, changing to a 2x6 stud reduces deflection to L/480 (but connection details still need to be solved). The trussed roof itself should be a self-supporting rigid triangle, and won't deflect, given proper connection details. It's the horizontal loads at the wall (out at the top, in at the bottom) that are the concern.
Lance, do you need to place solar panels on the pent roof? If not, would it work to utilize metal roofing on the pent roof to minimize the snow buildup? Just thinking of ideas....