Wood Roof--> are rafter ties necessary?

[Case580b]

are rafter ties necessary?

assume:

-- a plywood-roof-diaphragm is designed to carry the largest possible in-roof-plane loads.

-- the top-plate of the wall is designed to be an adequate chord for the diaphragm.

-- the gable-end-walls can transfer the the diaphragm-shear at the ends-of-the-diaphragm to the top-plate below then to shear walls below.

 

[KootK]

By rafter ties, do you mean collar ties connecting the rafters across the span or hurricane clips at the ends of the rafters? Both are really more connected to gravity load design. If you don't have a load bearing ridge beam at the top of the rafters, then you probably need a collar tie. If you have net uplift forces at the ends of your rafters, then you probably need hurricane clips.

Neither collar ties nor hurricane clips are normally required as part of the lateral design. Where eve blocking is omitted, sometimes roof deck shear will be transmitted to the shear walls through something akin to the hurricane clips.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[BAretired]

Rafter ties are not necessary if the roof works as a folded plate, but that would require special attention to the nailing of the deck to the rafters, nailing to the top wall plate and nailing to the ridge board. In most circumstances, workmanship does not warrant that assumption but with special inspection from the EOR, it may be okay.

BA

 

[KootK]

Upon re-read, I'm pretty sure that you're talking about the issue that BAretired has responded to. Please ignore my earlier response.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Case580b]

Thank you for your very quick answers. The folded plate is what I was thinking about. A bit scary... but as BAretired suggests it is possible in theory.

Thanks, again.... 3 weeks of thinking, it is very helpful to have the ideas of others.

 

[KootK]

If you pursue it, I'd be very interested to hear how it works out and what the details wind up looking like. For lower pitched roofs, I imagine that the diaphragm forces and demand for stiffness would skyrocket.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Case580b]

Hi KootK , I will be happy to post or send drawings. Is there a limit on allowed size at this site?

Probably there will be 2-cross-ties for a 50 foot long wall, so it will not be all diaphragm transfer. Also, it may end up with partially supported ridge. Also, it is trickier as the 'gable ends' are really 4: 12 sloped roof planes. Rectangular pyramid, with 4:12 pitched roof planes, with partial ridge along long axis.



As far as I can tell the tricks are:

--adequate chord nailing at splices... looks to be about 50- 10d common at any splice, might make sense to add LSTA12 or LSTA18 at splices, exterior side of top plate

--adequate nailing at hips--- LVL and PSL can't be used (unless blocked, see below) as nail spacing is a minimum 3" OC for PSL and LVL top edges... so DF and probably minimum 3x lumber or blocking, parallel and nailed to hip. I like the idea of shaped 2x blocking, as the top edge can be ripped so sheathing fits flat to block top.

-- The hips act as collectors. It is probably reasonable to assume that collector-force from the long-wall-diaphragms can be transferred to the end-walls thru to the sloped end-wall-roof diaphragms, if it is designed for required capacity; i.e. collector force transfer and shear capacity needs to be adequate.

-- it might make sense to add wall-corner-straps as the calculated force without-transfer-of-diaphragm-shear-to-'Gable"-end-walls (6000lbs tension, no diaphragm transfer) (hips are pushing corner apart). Wrapping strap around corner is a bit questionable as wood might compress under straps, so there might be steel plates at the corners, under the straps, or angle iron at the corners, under the straps. Corners would have to be slightly rounded so as not to over stress (cut strap).

Also (without complete diaphragm transfer) the hips will be pushing out at the corners (6000lbs no diaphragm transfer) , so the hip has to have enough bearing at the corner so as not to exceed compressive strength of hip or top plates.

The above (other than top-chord splicing) seems like it becomes safety factor, or redundancy, if one is confident of transfer of shear thru diaphragms. So... I would be working with 4000 lbs at corners is probably adequate and safe.

Diaphragms are designed Dead + Live for 1.15/1.6 (ratio of Cd.s) times table values for allowable shear for diaphragms.

I would be happy to hear any other ideas you might have.

Cheers, Case580b

 

[XR250]

Unless there is a compelling reason to do so, stick with a structural ridge or use rafter ties (if possible). The amount and cost of detailing that would be required will make your head (and the contractor's) head explode. It will also likely not get built properly and may sag over time due to nail slip.

 

[Case580b]

I agree with XR250.... but its about what the architect and owner want visually... vaulted ceiling, clear spaces and minimal visual structure. The project is part of a remodel with some of the structure to stay in place.

The things what gets one self into!

 

[XR250]

what is the ridge length and building width?

 

[BAretired]

I would try diligently to dissuade the owner from the folded plate idea. Scissor trusses would provide a vaulted ceiling and would correspond to standard construction practice.

Alternatively, an exposed glulam truss at the third points could support a ridge beam at 17' centers and would have an attractive appearance (the architect may not agree).

BA

 

[BS2]

I have designed many wood framed roofs that are vaulted but have never attempted or even seen an attempt at a folded plate design. I have seen enough sagging roofs without ridge beams or rafter ties to know that if you pursue this route, good luck, be careful, document your design here, and make sure your professional liability insurance is up to date. My typical design option order for vaulted construction is: ridge beam, scissor trusses, steel beam or frame, tension ring.

 

[XR250]

+1

 

[SteelPE]

I'm not sure specifically what the architect/owner is looking for that is pushing you in this direction. No matter what you do you are going to need some type of structure to support what ever loads you have up there. As others have said, I would run a ridge beam, if an LVL was "too much structure" then I would run a steel beam. I have built my fair share of houses (literally) and I would not have confidence in the type of structure you are proposing.

Is the structure to be exposed?

 

[Case580b]

Thank you all for your responses. I will talk to the architect and owner.

a few thoughts however..... some frivolous but...

-- i have attached an APA document I just got from APA... Folded Plates.... notice the water stamp 'Use at Your Own Discretion'

-- sags might be kind of cute, if calculated appropriately one would have a 'classical japanese style roof' ... a bit of a joke but i have wondered about the possibility for many years... does the style assume sagging or are a few curves just more elegant? probably the latter, but where did the idea come from... sagging?

-- post tensioned cables? just a wild idea... but next to the top plates, maybe one could add a 2 x4 on top of double 2 x6 top plate the building is rectangular... post-tensioning is not expensive,.... basically a perimeter band to keep everything locked together...


Thanks again for your experience. Cheers

 

[KootK]

You terrible tease! I must have that document but the link is bunk.

I was in Jamaica earlier this year and a noticed a number of outdoor patio structures built in a manner somewhat similar to what you've proposed. They would have collar ties but only every 2-4 rafters instead of at each one. It was fun to guzzle Redstripe and contemplate the implications/intent. They seemed to be performing fine. Of course their only real environmental load would be the occasional typhoon.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[KootK]

Okay, found it. It's pretty bad ass. If you do this I want to help. Anonymously.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[RFreund]

No way, it does exist!

So this is for gravity loads, right? So your top plates will act like a tension ring? Seems scary, but would be very interested in the design and details.

EIT

http://www.HowToEngineer.com

 

[KootK]

I will say that I like it much better as a conventional gable than as a hip roof.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Case580b]

i don't think i can use the folded plates for this project. It is a remodel... and there are just to many added uncertainties.

That been said one can still add a lot of extra strength, for not a lot of extra cost... a fancy sink and fixtures probably costs more.

adding xtra nails at top plate splices (as existing is dried fir it probably needs to be predrilled for new nails to avoid splitting)
adding stuff to increase strength of corners (strap around corners).

If this was new and i was trying to do folded plate... one could use Framing anchors like HDU crossed at each corner... For the maximum conceivable tension try to open the corner connection I calculated about 6000 lbs DL + LL. This would require an HDU-11. One board of a double 2x6 top plate (a 2x6) could carry the shear load if the the HDU is installed close as possible to the corner 1 3/4" bolt-CL from-corner. NDS 3.4.3 Shear Design allows for a substantial shear force reduction. on the exterior side allowable compression is 625 psi... so a plate with 10 square inches is probably appropriate. This is also the area of the face a 2x6 so if the top-plate boards have alternate overlaps i.e. one top-plate board extends all the way to corner in one direction, and the the other in the other direction than top plate has adequate compression.

The other tricky connection is the ridge

I have attached a PDF of the structure with a detail of the ridge connection..

Thanks all for your thinking about this....