Hip Roof

[XR250]

Hi There:

I am sizing some hip rafters for a house.
The roof is a rectangle 34 ft. x 38 ft. with four symmetric hips and a 4 ft. long ridge. It has a 12:12 pitch.
The attic floor is fully sheathed with 3/4" plywood and the rafters sit on a plate on top of the plywood.
The house location is central NC, so snow loads are not huge.
I typically design my hips as beams and support them with a post down to bearing walls. However, in this case, since the attic floor will prevent rafter spread in all directions and the roof is pretty steep, would you guys thing there is enough folded plate action going on to preclude me from having to support the hips on posts or design them as beams?

Thanks in advance.

 

[RFreund]

I've wondered this myself, and I have seen it framed w/out posts many times. The hip seems to start to act like a compression member (and some flexure). I would need to look into this a little bit. The problem seems to be that over time there is some spreading unless the ties are carefully considered. Having said that the outward thrust from the hip at the wall would be resisted not only by ceiling joist ties but also the outside walls acting like a shear wall. The force at top is transferred into the diaphragm and really countered by the other hip.
The jack rafters will be tied at the bottom (although you will need to detail something as only ties in one direction can be continuous) then at the top the force would be transferred into the diaphragm (you may need some blocking here to transfer this force into the diaphragm??) I think ilevel (or weyerhaeuser? the guys who manufacture LVL's) might be able to help some with this.

EIT

http://www.HowToEngineer.com

 

[XR250]

Yup.
FWIW, in my 20 years looking at houses, I have only seen problems with hip roofs if they were low sloped. The high sloped ones, even if they are really cut up and have no floor diaphragm seem to work just fine.

 

[msquared48]

I've done this before and it works quite well for structures that are square or almost square where you can use the outside wall top chord as a tension ring. Then you do not need any columns. You just have to have a special connection at the corners to take the lateral thrust.

If it is a rectangle, then to get no columns, I either use a glulam or collar tie truss at the hip intersection or close to it. That works too.



Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[XR250]

Mike,

I feel like the thrust is resolved at the base of all the rafters (due to the floor diaphragm) and not just at the hip corner. As such there should not be a thrust load at the hip. Thoughts?
FWIW, I have done the tension ring thing too on the past but on much smaller roofs with no floor deck.

 

[msquared48]

Can't agree about the corners, but yes for the rafters connecting to the wall top plate, at least for the most part.

In the ones I have done, the Architect wanted an open vaulted ceiling, so there were no deiling joists to act as collar ties. With the hip sitting on top of a corner column, the only way to verify through simple calculations to take the lateral thrust of the hip wats to provide a ring beam of sorts with a substantial corner connection. If the roof diaphragm is deep enough, you may be able to justify the roof diaphragm acting as a deep beam, but I think that in order to accomplish this, you may get more spreading in the walls and cracking of the sheetrock at certain locations. I just prefer the corner connection as it is a more direct and verifiable solution.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[woodman88]

“...thing (I think you meant “think” here) there is enough folded plate action going on to preclude me from having to support the hips on posts...“
I hope not. The only way you will have any folded plate action is if the floor diaphragm or the connection of the rafters to the floor diaphragm fails. With a properly designed floor diaphragm the roof will be three hinge arches and modification of one. It should work fine. Be sure to take a careful look at your connections (to floor diaphragm and ridge) and the tension forces in the floor diaphragm.

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[nac521]

We typically turn the ceiling joist 4 to 8 feet from the hip end for a direct connection of the roof rafters to ceiling joists at the hip end. Add some strap ties and block a few bays back at the intersection of the turned ceiling joists.

 

[XR250]

@Mike;

I agree with you about the vaulted ceiling, but with a floor diaphragm, it is another story.

@woodman88 -thanks. With a 12:12 roof and our small snow loads, the loads in the floor diaphragm are pretty small

@NAC521 - Do you do this without a floor diaphragm and just the sheetrock ceiling? W/o the floor diaphragm, you are introducing the the rafter thrust from the ladder framing into the ceiling diaphragm as a tension force.

 

[msquared48]

@ ExcelEngineering:

I guess I don't understand the layout here. Can you poast a sketch?

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[nac521]

@ Excel: No, I don't like to rely on the gyp ceiling to resist the tension. If there is no plywood, we sometimes add a flat strong back to the top or block and strap tie all the way across depending on the length and geometry. Even with plywood, I prefer to block and strap a few bays in to distribute the tension.

 

[RFreund]

I don't want to distract from the thread but....
For the case of a vaulted ceiling do you span the top plates horizontally between perpendicular walls? I would think that you wouldn't be able to get very far and may need to use an interior wall as a 'support' for the top plates, no? That does seam like a tough connection to make; steel angle?

EIT

http://www.HowToEngineer.com

 

[XR250]

@Mike - don't have a pic, but it is simply a 34' x 40' 12:12 hip with the rafter bearing directly on the attic floor sheathing (with a 2x6 sill plate)
As such, the thrust of each rafters goes directly into the floor diaphragm.

@nac521 - Nice - that is the way i have been doing it as well when the attic can be sheathed

@RFreund - that is a tough connection and you would likely need 2x6 exterior walls to get enough horizontal beam action and not have the interior walls too far apart and have them be continuous across the building. There are many ways to skin a cat on a vaulted hip roof, but most of them only work on smaller roofs. I have done a larger vaulted hip roof before, but it used steel bents made from W10x77 i-beams.

 

[RFreund]

So interestingly enough I was at my brother-in-law's home this weekend and they have a pretty big vaulted hip roof. Actually the ridged roof that frames into the hips is also vaulted. The span is approximately 50' with an 8:12 roof pitch. I know that the vaulted ridge portion is framed using scissor trusses but I'm not sure if the hip is made of trusses or stick framed. He has plans but no roof framing plan is included (he purchased the home). There is some minor drywall cracking a the ridge but you could almost contribute it to shrinkage cracks (really not too bad). There has to be some sort of 'plate action' here. There is only one return wall and it does not extend and follow the hip roof. My questions are:

1. How is this standing? Is there some way to use a 'scissor' type truss for the hip framing?
2. How do you ventilate the hip roof (space between the jack rafters)if it is stick framed?

I want to ask him if I can go up in the attic and even check the walls for plumb.

I will follow up with a sketch on a couple ideas.

EIT

http://www.HowToEngineer.com

 

[RFreund]

Attached is a sketch. I'm guessing you would need a girder truss for this application, but that would be one large girder truss.

Is there anyway that the roof sheathing on the hip side starts to act as a beam spanning from hip to hip? Meaning that as the wall (say West wall on the attached sketch) starts to deflect outward due to the thrust form the jack rafters, the roof sheathing actually starts to prevent the rafters from 'sliding down'?

EIT

http://www.HowToEngineer.com

 

[RFreund]

Forgot to attach

EIT
www.HowToEngineer.com

 

[woodman88]

Typically, with pre-fab wood trusses, the girder truss would be setback 8' from the end wall. With 2' oc jack trusses and a corner girder connected to it. Then behind it are the main span step trusses 2' oc.

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[msquared48]

RFruend:

The greater the slope of the roof diaphragm, the greater the contribution tro supporting the hip...some.

But look to the obvious for the solution here. The hips are more likely than not supported by a structural member, truss, collar tie, beam, whatever, at the intersection of the two hips, or very close to it.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[RFreund]

@Mike - Yeah good point, I see that now.

@Garth - See attachment, is this the layout you are referring to? It makes sense for most hips that I have seen, but these would be oddly shaped trusses as the ceiling is vaulted below. But maybe they are oddly shaped scissor truss with a flat top cord for a portion...

EIT
www.HowToEngineer.com

 

[woodman88]

Yes, that layout should work. You appear to have shown the truss spacing as 4' oc instead of the 2' oc that is probably needed. I think that the lower stresses in a oddly shaped scissor truss is easier to deal with than the high stresses in a full hip scissor truss at your span of 50'. Also at a full setback of 25' the connections and reactions will present possible structural problems for construction and in the future that would be so easy to avoid by using an 8' setback. Also at 8:12 top chord pitch you will probably need a cap truss so a 3:12 bottom pitch is likely. Unless your client wants to pay extra to field splice trusses.

But you should talk to a local truss manufacturer and/or their engineer about this.

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.