Strap Reinforcement of Existing Rafter 4

[SRO]

I need to reinforce existing rafters for the addition of solar panels. The simple solution of sistering another rafter along side the existing one isn’t an option due to confinement issues of the attic access (rafter is too long to fit).

My solution:
Attach a Simpson coil strap “CS14” (1/16”x1-1/4” steel) at each end and run it under blocking that would be between the strap and the rafter to create a deeper beam (see detail).

My Process:
•Find required section modulus by solving Mmax = Fb’ x Sreq.
•Transform the steel to the equivalent wood section.
•Determine the neutral axis & new moment of inertia for the transformed section.
•Compared the section modulus above & below the NA to the required S
•Used the allowable deflection to determine the required moment of inertia & compared it to the one for the new transformed section.

My Question:
•Does anyone see a problem w/ this solution and is there anything I overlooked?

 

[larsacious]

Interesting concept! Have you considered just attaching the Simpson strap to the bottom of the beam without the blocking?

I did a quick and dirty analysis of the geometric properties of a SPF #2, 2x12 with a 14GA(.0747")x1.25" plate attached to the bottom and determined the following:
-"I" increased by 38%
-"S" top for the transformed section increased by 23%.

In the past, I have found the weight of solar panels attached to a pitched roof to be relatively light (less than 5psf).

You should be able to justify the existing rafters by connecting the straps directly to the bottom of them. Be sure your analysis is two parts in that you should analyze the rafters for the existing loads based on the section modulus of the existing beam only. After that, you should analyze the rafters for the weight of the solar panels and live load with the transformed section.

Good Luck!

 

[msquared48]

Should work just fine as it is a concept used to support moving sprinklers in farm areas.

The only change I would make would be to twist the strap at each end , nailing the strap to the side of the rafter rather than the bottom and clinching the nails. Using this approach, you could use two straps side by side if needed.

You know though, sistering could still be an option here as you can sister on smaller members rather than the same size. The calculations are more involved, but it may be simpler. Similarly, the sistering does not necessarily have to be full length.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[BAretired]

Will you have enough nails to develop the force at each end of the strap? The strap has a component normal to the rafter which will tend to pull out the fasteners.

BA

 

[SRO]

Larsacious:
I agree usually the small increase in DL of the solar panel is offset by the decrease in SL from changing the roof surface to a slippery surface. If the panels were to be mounted to every rafter there would be no problem I would absolutely make the “net reduction in load” argument, but the problem arises when the mounts skip one or two rafters and concentrate the load on a single rafter.


msquared48
I know it seams minimal, but are you ever concerned that twisting the strap and attaching it to the side will introduce torsion in the rafter? Granted, it could be resisted by blocking, however I’m trying to make this as efficient as possible & more blocking, means more labor, which could make this option less efficient then another.


BAretired
I had intended use wood screws instead of nails to attach the strap. The first screws would resist the pull out force (via NDS) and the following nails would transfer the tension in the strap to the rafter.

 

[SLTA]

you might want to check with Simpson and make sure they're ok with using screws instead of nails. sometimes they can be rather particular. 1-800-999-5099 and ask for the technical department.

 

[BAretired]

Actually, the described method is not exact. The transformed properties of the combined section are not needed.

The proper analysis is to apply one sloping force at each end and one vertical force at each block. The compatibility condition is that the deflection of the joist at each block is equal to the vertical deformation in the steel strap under combined loading.

Another point to consider is temperature effects. A change in temperature will affect the steel strap more than the wood joist.

BA

 

[BAretired]

SRO,

•Does anyone see a problem w/ this solution and is there anything I overlooked?

There is a problem with your solution. It is not correct. While it may not make a great deal of difference in the particular situation which you describe, it is important to note that the method is fundamentally wrong.

BA

 

[SpecialEddie]

Unless the rafters are deteriorated or damaged in some way, I don't quite understand why you would need to beef them up to support solar panels that, along with the racking system and mounting hardware, weigh probably less than 5 psf. Based on your sketch, your concern is obviously not uplift. If the rafters were designed for a roof live load of 20 or 16 psf, the solar panels should impose no significant additional stresses on the existing framing. Solar panels are not designed to be walked on, so I think you can justify your analysis by reducing your roof live load to zero over the area occupied by the panels. Solar panels are considered live loads (like interior office partitions), not sustained dead loads. At least they are in the county where I do most of my work. I would imagine it would be the same anywhere else (though I would not be surprised if it wasn't). I don't know how you are doing your analysis, but if you are including the solar panels as sustained dead loading AND adding 20 psf of live load to your load combination, you are counting for the weight of the solar panels TWICE and unnecessarily penalizing yourself. So in an indirect sort of way, the solar panels have already been accounted for in the original design. Something just doesn't seem right...

 

[SRO]

BAretired
I agree that technically this in an indeterminate structure and should be analyzed as such. However if I were to try and invoice the client for that I doubt I’d ever get another job. As right or wrong as that may be it’s the nature of the beast. I’m just trying to provide my client with the most efficient overall process I can. If the cost for extra analysis will save my client money during construction then that’s what I’ll recommend. I just can’t see that as the case here.

I did look into the thermal expansion, and with a temperature change of 150 degrees for this span the differential change in length it only 0.11”



SpecialEddie
In this region (Massachusetts) roof live load rarely governs it’s usually snow load which still needs to be included, although it usually does decrease enough to offset the additional load form the panels. The problem arises when the panels aren’t supported by every rafter but every other one, or every third one. This is due to the contractor’s preference to reinforce the rafters instead of having additional roof penetrations that may leak over time.

 

[PMR06]

This is an interesting idea, but there is something that just doesn't feel right to me. I've seen similar retrofits before, but they used turnbuckles of some sort to prestress the new tension member.

Without running numbers, it feels to me that the strap and screws (not being prestressed) would have to stretch a bit before taking any significant load. What happens to the system if say each end of the strap deflects/stretches even 1/16"? Does that mean the existing rafter would have to deflect (i.e. take a larger share of load) before the composite behavior is engaged?

 

[SpecialEddie]

Hmmm. Makes sense. I have never lived in or worked on projects in areas with snow -- the thought never crossed my mind. Still yet, seems like a lot more work for the contractor to retrofit the rafters than to than to just throw in more supports and anchor the system at each rafter. But I guess with all that snow and moisture on the roof for extended periods of time, makes sense that minimizing roof penetrations would be a high priority.

 

[SpecialEddie]

PMR06 brings up sum interesting points. Have you considered possibly using a heavy duty pre-deflected Simpson holdown horizontally at each end at the bottom side of the existing rafters and using a 1/2" or 5/8" rod to tie the two together. Or possibly clevis and turnbuckle at each end with a steel rod or cable? Seems like it would be a lot simpler to calc if you could make the numbers work. I don't know the span of the rafters or any other project specific data, so I'm just thinking out loud...

 

[BAretired]

I like the prestressing idea too.

BA

 

[SteelPE]

I’m not sure about what type of roofing system you have… but have you evaluated any alternatives such as providing a strut from the center point of the rafter to the interior bearing wall (if you have one)?

In a normal “stick” built house, the addition of a strut could be easily accomplished. This would cut down on your span which would greatly increase the strength of your rafter (although you would add more weight to other portions of the structure.

 

[SRO]

Good point about the pre-stressing of the steel (thanks PMR06). I think that the additional cost of hold downs, threaded rods, & connectors vs. only steel strap might make this option less efficient than another solution unless I can minimize the locations that need reinforcing, but that’s something I’ll need to run the numbers on. It the rafter is jacked up during the strap installation to relieve the stresses, then when the jack is removed the rafter would deflect and pre-stress the strap. I know it’s mathematically the equivalent of shooting from the hip, but what is everyone’s thought on that?


SteelPE
That would definitely resolve the rafter issue. However (as you mentioned) the new load would have to be accounted for in an existing house, and it seems people want to avoid that like the plague.

 

[MikeHalloran]

>>> The simple solution of sistering another rafter along side the existing one isn't an option due to confinement issues of the attic access (rafter is too long to fit). <<<

I've been wrong lots of times before, but I thought you only had to sister the middle half.



Mike Halloran
Pembroke Pines, FL, USA

 

[Prestressed Guy]

This strap can easily be prestressed by installing it flat to the bottom of the rafter with only the ends attached. Then using a wedge and series of blocks, deflect the strap down to its final position. As it is pulled away from the rafter, it will induce tension in the strap and you will prestress the rafter. You sill need to sharpen you pencil a bit to figure out how much deflection is required to get the tension you want, but it is doable.

 

[dhengr]

SRO:
The prestressing concept is the correct way to handle what you showed in your sketch, and I thought that was what you had in mind, so I didn’t go through your transformed section verbiage in detail. The difficult part is the end connections of the straps to really allow the strap tension to be induced. There is considerable relaxation at those connections which tend to defeat your tensioning. You’ve really got to think that out and over design that detail. And, Haydenwse has about the right way to induce the tension on this type of job. Otherwise, it’s a bit of seat-of-the-pants engineering: since the length change of the strap, due to the blocks implies the strap tension; but the upward deflection of the rafter and connection relaxation reduce that tension. A system with the blocks (stand-offs) and a turnbuckle allows you to watch the rafter deflect and connection movement during the tensioning process.

You have an interesting (dangerous to yourself) engineering philosophy, where you are getting your tit surprisingly close to the wringer to save the client money in terms of your fees. Nobody is suggesting that you way over engineer the problem and solution, but you’ve got to get it right. You have to really think it through, and have a real leg to stand on; that is calcs. and concept well thought out, etc. including some notes on the approx. nature of the entire process. Because, even if you save the client some fees, by doing a half assed engineering job, when your solution goes south, you can be sure the client’s attorney won’t remember that the client wanted a solution on the cheap. They need your help and should be willing to pay a reasonable fee for that help. These jobs are usually interesting, but they are kinda dogs for all of us; gosh it’s a simple problem, why does it cost so much; if it’s so damn simple why do they need you? One always wonders if a $500 job, that really should have been billed at $1500, is worth the $3000+ it could cost you to defend yourself, all for having saved the client money, no mention of solving his problem practically and economically.

 

[SRO]

dhengr;
I’m more conservative than I appear to be in this post although I appreciate the concern. I am guilty of trying to think out better ways to accomplish something, however I do know when to jump ship. After reading Haydenwse’s post I really don’t feel comfortable pre-stressing the strap w/ blocks & wedges. I have previously used turnbuckles and used torque to get some indication of the tension in the threaded rod so I might still recommend that as an option to my client. Unfortunately, due to the additional cost I think we’ll be right back to square one and open up the roof to get lumber in the attic and sister the existing rafters.

I love sending invoices like these out.

Thanks for everyone’s help.