Vaulted Hip Roof

[cancmm]

A friend of mine recently bought a house that has a stick framed hip roof over his deck, which is about 14'x20'. The roof does not look to be original to the house and obviously not designed/constructed to code, considering it is sagging significantly. It was probably constructed about 10-15 years ago. Given that I deal primarily with bridges, I do not have the expertise to fully redesign his roof for him, but I would like to let him know what to expect when he hires a engineer/contractor.

A quick description... The hip starts immediately at the connection to the house, which is a vertical wall, 2 storeys high. In other words, there is no ridge beam typical of rafter-framed construction. Similarly, there are no collar ties so the ceiling vaults increasingly as you get closer to the house. If you can imagine, the peak of the hip sits a few feet above the second storey floor. They simply connected the two "hip beams" to a board nailed to the side of the wall. I highly doubt they reinforced the stud wall by adding a built-up column inside. The other three sides of the roof are supported by a 4x8 beam with 4x4 posts at ~4' o.c. and 2x4 cross-bracing beneath the hand rail. I hope this gives the idea of what we're looking at...

My friend (actually his wife) likes the look/feel of the roof so they'd like to replace it as is. My primary concerns are as follows:
1) Hip beams are undersized (I checked). Easy enough to fix during reconstruction though.
2) The vertical reactions of the hip beams are being supported by the exterior stud wall of the house. I feel they need to add a column to the exterior face to take this concentrated load.
3) I'm assuming the lateral thrust is being absorbed through the second floor diaphragm of the house, as well as the cross-bracing of the deck rail. I'm concerned that this isn't good practice and the structure should be laterally self-supporting.

My questions are as follows:
1) Are wood diaphragms effective over the ridges of the hip? I'd assume they would be but just figured I'd ask.
2) Is it common to absorb rafter thrust through diaphragm action to exterior walls, rather than more common methods such as ceiling joists or collar ties?
3) Am I correct in assuming that stud walls should not take concentrated loads of this magnitude?
4) Am I overlooking anything?

Thanks!

 

[MiketheEngineer]

1. Yes
2. No - use a ridge rafter/beam
3. Walls most likely OK
4. Probably - since I am not there.....

 

[cancmm]

Thanks Mike.
Regarding (2), there is no room for a ridge beam in the current configuration. What about roofs with only one hip? Imagine a simple rectangular building with a hip at one end and a gable at the other. The thrust must be absorbed through through the roof diaphragm wouldn't it?

 

[msquared48]

The thrust, if any, that you speak of would only be absorbed by the roof diaphragm if the diaphragm was plywood and not skip sheathing.

Hopefully you do not have skip sheathing...

Oh, and don't tell me, but there is probably a tile roof somewhere here when the original framing was designed for shake or comp... an extra 10 psf or so of dead load, and a probable caues of any sagging seen in the roof framing too if this is the case.

You do need a Mike there!

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[OldPaperMaker]

cancmm,
You might want to provide a sketch with a plan & elevation view of this roof system to make sure that we are all talking about the same thing.

If the hip beams and rafters don't have collar ties or horizontal ceiling joists (at the top of the wall) then the only thing keeping the walls or post from moving out at the top of the wall is the roof sheathing acting as a "folded plate".

 

[ToadJones]

Maybe you could get a truss designer involved and use mono hip scissors trusses with mono scissors truss infills.

 

[shobroco]

If this roof is as I picture it, a tie at the corners of the top plate of the wall might be all you need (with properly sized rafters) turning the top plate into a ring beam, or a half-ring beam that is tied to the original house. This also counts on sufficient rigidity to prevent horizontal deflection at the centre of the walls, but I've done it more than a few times.