I assist a PV solar panel installer with residential roof evaluations. Residential client has a (c. 1960s) rafter framed hipped roof. The design roof live load from that era was 20 psf, but today's design roof snow load for this region is 35 psf, and the local BO requires an engineer evaluation. This topic has been discussed on this board a few times. I could design a support at mid-span, or call for sistering a full-length rafter to existing. The existing rafter is 16' long, so getting that length material up into the attic is not feasible. Has anyone ever just sistered an 8' length to reinforce the mid-span of the existing? That would solve the issue. Shear is OK, I'm only concerned about bending. Any input is appreciated.
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations MintJulep on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Strengthen existing 2x4 rafters for additional 3 psf solar loading 2
- Thread starter TroyD
- Start date
- Thread starter
- #4
Retired & oldest,
Thanks. The local BO does require that the roof member meet today's design loadings. (And just this week they have adopted the 2018 IRC and IBC requirements). I like the strap idea and recall that method discussed on this board and elsewhere. Maybe I'll give the client several options, in case he cannot get 8' lumber up into the attic access.
Thanks. The local BO does require that the roof member meet today's design loadings. (And just this week they have adopted the 2018 IRC and IBC requirements). I like the strap idea and recall that method discussed on this board and elsewhere. Maybe I'll give the client several options, in case he cannot get 8' lumber up into the attic access.
EdStainless
Materials
My Dad's house was similar. But there was a vent into the attic that we removed for access.
We passed up 14 footers (with angle cut ends) for full length sistering.
And then we drove a $hitload of deck screws.
= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy
We passed up 14 footers (with angle cut ends) for full length sistering.
And then we drove a $hitload of deck screws.
= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy
I remotely remember that in modification works that do not alter the original structural arrangement and load carrying mechanism, a full code update is not required, which I consider that is your case. Hope someone here remember the rule and provide reference.
-
1
- #8
TroyD:
Straps won’t really do much for you unless they are highly tensioned before the are nailed off. If you just nail the straps on the bot. edge of the 2x4, they won’t really pick up any of the load until the 2x4 has deflected another inch or so, to start to tighten the strap and bring the nails, in their loose holes, into play. The straps need to be properly located, and tightly spaced edge nailing on the 2x4 can cause splitting which doesn’t improve the strength of the member.
One of the other problems with the relatively small added load of the solar panels, which tends to lull us to sleep, is that that load is not uniform, although that’s the way they sell it to make it look good. You had 20 or 25 lbs./sq.ft. snow load, plus 5 or 10 lbs. DL, and things seems to work just fine. Now, they want you to use 35lbs./sq.ft., just because the code writers say so, even though the loads didn’t really change. We just started to realize we might have been under designing and not paying sufficient attention to drifting, etc., and we are now using weaker materials. So, the codes are catching up. But, the bigger issue is that these loads will no longer be uniformly distributed to all of the rafters. As likely as not, every forth rafter will get a few concentrated loads, from the panel support system, which are made up of 35 lbs. snow, 5+3(the panels) or 10+3 lbs. DL, all on some contributory area (4’x6’ ?) and then applied at the support point. So, now two of the rafters only have some DL on them, and every forth rafter has much larger concentrated loading. You really have to analyze this loading situation before you determine a reinforcing approach. That most installers don’t let this bother them, a they charge ahead to sell panels, dose not make what they do right, it just attests to the fact that our codes and FoS are conservative, and we often overdesign. What they are doing just eats into our FoS, and that may even be o.k. in some cases with proper analysis and engineering judgement for an 8 or10% load increase, 3lbs. v. (35 + 5 or 10).
Straps won’t really do much for you unless they are highly tensioned before the are nailed off. If you just nail the straps on the bot. edge of the 2x4, they won’t really pick up any of the load until the 2x4 has deflected another inch or so, to start to tighten the strap and bring the nails, in their loose holes, into play. The straps need to be properly located, and tightly spaced edge nailing on the 2x4 can cause splitting which doesn’t improve the strength of the member.
One of the other problems with the relatively small added load of the solar panels, which tends to lull us to sleep, is that that load is not uniform, although that’s the way they sell it to make it look good. You had 20 or 25 lbs./sq.ft. snow load, plus 5 or 10 lbs. DL, and things seems to work just fine. Now, they want you to use 35lbs./sq.ft., just because the code writers say so, even though the loads didn’t really change. We just started to realize we might have been under designing and not paying sufficient attention to drifting, etc., and we are now using weaker materials. So, the codes are catching up. But, the bigger issue is that these loads will no longer be uniformly distributed to all of the rafters. As likely as not, every forth rafter will get a few concentrated loads, from the panel support system, which are made up of 35 lbs. snow, 5+3(the panels) or 10+3 lbs. DL, all on some contributory area (4’x6’ ?) and then applied at the support point. So, now two of the rafters only have some DL on them, and every forth rafter has much larger concentrated loading. You really have to analyze this loading situation before you determine a reinforcing approach. That most installers don’t let this bother them, a they charge ahead to sell panels, dose not make what they do right, it just attests to the fact that our codes and FoS are conservative, and we often overdesign. What they are doing just eats into our FoS, and that may even be o.k. in some cases with proper analysis and engineering judgement for an 8 or10% load increase, 3lbs. v. (35 + 5 or 10).
Hi there TroyD,
when you sister 2x to an existing member, make sure you use the idea of springs in parallel to get a correct bending moment. Since the span is 16' I am positive that you are using a rough sawn element 2x4 and you can consider that as a select structural if it is an old building, maybe that will help crunching numbers
I agree with Dhengr on the use of straps which may cause splitting of the member itself due to excessive nailing because of shear flow criteria.
when you sister 2x to an existing member, make sure you use the idea of springs in parallel to get a correct bending moment. Since the span is 16' I am positive that you are using a rough sawn element 2x4 and you can consider that as a select structural if it is an old building, maybe that will help crunching numbers
I agree with Dhengr on the use of straps which may cause splitting of the member itself due to excessive nailing because of shear flow criteria.
-
1
- #12
- Thread starter
- #13
Thanks dhengr, retired, Satya, etc.
Dhengr: I agree with your assessment of the additional solar loadings. The solar equip weight is approx. 3 psf, but the mounting attachment config result in point loads of ~29 lbs. The snow will sit on top of the PV panels, so that becomes a point load also. The mounting rails allow for flexibility in locating the attachments on the roof, and they figure that out in the field. The distribution of point loads onto the rafters is somewhat random. So you have to use good judgement and run a few different scenarios to figure out worst case.
Retired: I need to spend some time digesting these links for code compliance on existing structures. Thank-you.
Dhengr: I agree with your assessment of the additional solar loadings. The solar equip weight is approx. 3 psf, but the mounting attachment config result in point loads of ~29 lbs. The snow will sit on top of the PV panels, so that becomes a point load also. The mounting rails allow for flexibility in locating the attachments on the roof, and they figure that out in the field. The distribution of point loads onto the rafters is somewhat random. So you have to use good judgement and run a few different scenarios to figure out worst case.
Retired: I need to spend some time digesting these links for code compliance on existing structures. Thank-you.
Those links provide a perspective as to what the model codes say.TroyD said:
HOWEVER,
Note that those links are for model codes which mean nothing unless the local city/authority has adopted them and/or not changed them.
Many cities adopt the ICC codes outright, but many others adopt and significantly amend/revise these model codes to suit their local conditions.
Especially where existing buildings are involved. Some reduce the 50% rule to 20%.
I would suggest a call to your local AHJ or a visit to the local government website to verify what in fact is adopted and what in fact is altered locally.
I agree with radiocontrolhead.
In addition to the shortcomings of sistering and strapping (and whatever) indicated by others above, there is the nasty little issue of resisting the thrust at the bottom end of the rafter where it (presumably) is nailed to a ceiling joist with one or two nails.
1960's hip roofs were fine but if you had to design one today using real engineering principles I don't think you'd detail it the way most of them were built back in the day (even assuming material strengths from that era).
For a 1960 house with a stick frame roof, I would expect there are a couple of load bearing walls located somewhere inside the plan. Most likely you could slip in a few double or triple 2x10 beam, strategically placed within the ceiling framing and with props up to some boards that would support a group of rafters.
In addition to the shortcomings of sistering and strapping (and whatever) indicated by others above, there is the nasty little issue of resisting the thrust at the bottom end of the rafter where it (presumably) is nailed to a ceiling joist with one or two nails.
1960's hip roofs were fine but if you had to design one today using real engineering principles I don't think you'd detail it the way most of them were built back in the day (even assuming material strengths from that era).
For a 1960 house with a stick frame roof, I would expect there are a couple of load bearing walls located somewhere inside the plan. Most likely you could slip in a few double or triple 2x10 beam, strategically placed within the ceiling framing and with props up to some boards that would support a group of rafters.
Assuming the roof can take the "overall" load of the panels, here is what I do to deal with the point loads that dhengr mentioned..
Take a 4x6 "strongback" distribution member and install it tight up against the underside of the rafters perpendicular to their span. I hang it from each rafter with (2) hurricane ties or a Simpson LTS12.
This takes the point load from the rack leg and distributes it to the adjacent rafters evenly +/-.
Given your situation, I would also be concerned with the overall capacity of the roof as a 1960's hip roof defies analysis unless the hips are supported by a post and are sufficient in size (both unlikely) - similar to Houseboy's comment. Maybe try to keep the panels over the central area of the roof that is easily analyzed.
Good luck.
Take a 4x6 "strongback" distribution member and install it tight up against the underside of the rafters perpendicular to their span. I hang it from each rafter with (2) hurricane ties or a Simpson LTS12.
This takes the point load from the rack leg and distributes it to the adjacent rafters evenly +/-.
Given your situation, I would also be concerned with the overall capacity of the roof as a 1960's hip roof defies analysis unless the hips are supported by a post and are sufficient in size (both unlikely) - similar to Houseboy's comment. Maybe try to keep the panels over the central area of the roof that is easily analyzed.
Good luck.
radiocontrolhead
Structural
Have you been up in the attic?
Out of curiosity what is the framing size and spacing? These rafters For houses of this age are usually braced by kickers along their length at a pretty redundant fashion
Out of curiosity what is the framing size and spacing? These rafters For houses of this age are usually braced by kickers along their length at a pretty redundant fashion
- Thread starter
- #19
radio,
The rafters are 2x4 (possible actual 2x4, species unknown, likely old-growth D.Fir or similar) at 16" o.c. The rafters along the sidewalls are braced at mid-span with a continuous strongback and kickers down to the interior bearing wall. But the solar panels are being installed on the hipped roof plane at the south-facing endwall. No bracing there. 2x6 girders at hipped ridge support the endwall rafters and are framed into the peak.
The rafters are 2x4 (possible actual 2x4, species unknown, likely old-growth D.Fir or similar) at 16" o.c. The rafters along the sidewalls are braced at mid-span with a continuous strongback and kickers down to the interior bearing wall. But the solar panels are being installed on the hipped roof plane at the south-facing endwall. No bracing there. 2x6 girders at hipped ridge support the endwall rafters and are framed into the peak.
Another idea, if not mentioned before. Beef up the ceiling system with sufficient new ceiling "joists". Then use struts up to the rafters to give them adequate support mid span. A lot easier than attaching things to the underside of the roof.
- Status
Similar threads
- Question
- Locked
- Question
- Question
