I am in the process of designing a house with a barrel roof similar to FHB #126 (September 2002) Pg 102. The top plate on the two sidewalls will be arced and have the joists run across them to form the arc. I have a 22’6″ clear span across the barrel. The roof joists will be perpendicular to the top plate in all locations with the possible exception of where they line up with the front and rear wall. My structural Engineer has come back saying that all his calculations require 5 x 10 glulam beams spaced 12″ on center to cross this span. The house in FHB #126 used 2 x 10 DF lumber and has a snow load that I will never see. The design has already been kicked back from the City of San Diego due to engineering errors on the original plan submission, so I need to get the second submission right. However, 5 x 10 glulam is too heavy and pricey for my application. I work as an engineer, but do not have a PE license. This looks like a simple statics 101 problem, but I need to find supporting data to prove my point the structural engineer. <!—-><!—-><!—->
Does anybody know where I can get some design guidelines for barrel roof systems?<!—-><!—->
Thank You, <!—-><!—->
James
Replies
Check out the Western Wood Products Assn. lumber technical guide:
http://www.wwpa.org/default2.htm
You ned to register, but it's free. I have an older version of their span calculator, and it says that you'd be OK with sel. str. 2x10, but of course there are plenty of other factors, especially in earthquake land...
I was searching their site earlier tonight without much success. My main problem is determining the loading factor on each beam. My structural engineer claims that he has to calculate the loading in the vertical direction for each beam. The actual force will run down through the angle of the beam and into the top plate. Any side forces will be taken up by the bracing and roofing plywood. Do they specify how to block and cross brace a beam, at the angles, as the barrel gets steeper?
James
The WWPA tools don't get into the details that would be needed to fully engineer this application (blocking, bracing, etc.). It's handy for sense-checking, and good for conventional structures.
Don't know if this is an option, but I'll throw it out anyway.
I once designed some half-round trusses for a barrel vault. The top chord was built in short segments of 2X12. Then it was cut to a radius after the trusses were built.
When you get into gray areas in design like this one, there's a lot of room for disagreement between engineers. Opinions vary widely. IT might be a good idea to look for another engineer.
Assuming you have to stick with stick framing - Another idea might be to go with deeper members. Like maybe single 14" LVLs could take the place of 5X10 glulams. They'd be cheaper and easier to handle.
How about some pics when this is done?
If I read you correctly, you've got two bearing walls with circle-segment-arc tops, and the clear span between them is 22'-6". Right?
Your roof rafters will all run horizontal, parallel to the centerline of the imagined tube shape, and will span this distance. Right?
If the roof were simply flat, you would have a joist-framed platform, and sizing the joists would be a simple matter of checking allowable loads for the spans, using the prescribed roof load per code, and looking at various spacings and member sizes.
The TrusJoist site has downloads for their I-joist systems, that can tell you what size and spacing of various joists might be for your situation.
There are online span calculators that can do the same for various species and grades of sawn lumber.
I am certain you won't get by with 2x10s. That example must have had a shorter span.
Edit: I just went here http://www.cwc.ca/design/tools/calcs/SpanCalc_2002/ and plugged in specs for a roof rafter system, 2x10s in northern douglas fir, No. 1 and 2, 16" on center, loading 20 psf live, 10 dead, which is for no snow, moderate roof coverings, drywall beneath.
The calc gave something like a 21-10 clear span, just under what you need. 2x12s should do you just fine.
To be perfectly legit, your rafters should stand plumb in section, and depending on location and arc radius, should have top and bottom edges beveled to track the curve, so that sheathing and inside finish bear flush.
Edited 4/5/2006 11:15 am ET by Gene_Davis
Never use a magazine as a reference. No one knows how good/poor the engineering is.
If you want a different opinion, you need to look for a different engineer.
I think part of your problem is your design. Simple changes may give you a better result.
Here is a quick sketch I did of rafters on a barrel-shaped roof. In plan view, the roof is 18 feet wide, and the rafters span 22-6.
The underside radius is 20 feet, the rafters are made from 2x12s, and they have beveled tops and bottoms, except for the one falling on dead center. Spacing is 16 inches.
I would build this roof with solid blocking between rafters at the quarter points, thus you will have three rows of blocking.