Resolving Hip Beam Thrust in Tension Ring

[trackman417]

Looking for some insight from you guys on this conundrum I'm having.

Currently have a hipped roof that takes the load from a gable roof. The ridge beam at the hip roof is transferred down to 4 hip beams(2 on each end). After it is all said and done, I have about 8000 pounds(pulling from memory here - on vacation, but can't stop thinking about this) of compression being pushed through the hip beams which then dump 6000lb(x and y) at the corners of the load bearing walls.

Original design of the roof assumed that the hip-beams took a couple hundred pounds of compression, not 6000lbs. Looking for some help as I am sure the builder would not like any detail I have to resolve these forces - except for the first option.

Ideas:
1).Assume collar ties will help support ridge board/beam(hip roof is 9.5/12)... Not sure if they would would work with this particular arrangement due to the uneven symmetry.

If they would work, would it be safe to assume the valley beams take the load from the whole jack rafter connecting to it with the collar tie arrangement?​

I like this idea because it is the SIMPLEST. Also seems like the rafters at the center of the hipped roof would have to do without the collar ties until the valley and ridge beam get farther apart. Would you guys recommend doubling up the collar ties for the first few rafters that can accept the collar ties?​

2). Add 15/32 shear walls on both sides of load bearing walls (3" edge nails, with 6" field nails in these areas)

Design connection with 1/8"(or similar)steel plate to connect both hip beams and drag the tension or compression into the top plates(only about 400psi in tension, less in compression).​

3). There are some 2X4 partition walls underneath here that I may be able to throw posts down into, but would not like the load path. Could ask the architect to make the room a little smaller to accommodate.

Would appreciate any recommendations/comments.

NOTE: The house is mirrored on the other side

 

[Ron247]

Per the building code, there should be a vertical brace from the roof down to a wall at every location where a hip, valley or ridge intersect each other. Does your house have those present?

What is your roof loading, the materials you have called out appear fairly large for a roof.

 

[trackman417]

Designing for:
Dead: 15 PSF
Roof Live: 30 PSF(Per Code in the area, with no allowance for reduction)
Snow: 20 PSF(not reducing due to the 30 psf Lr)

Members do seem large, but if I go with one less lamination(1 3/4") I exceed the capacity of the ridge beam... Keeping the members as deep as I can before they require multiple ply's just to get to desired/design depth.

Code 802.3 states that there must be a vertical brace designed to support each hip. If not, the hip beams should be designed to transfer that load.

I do have some walls that are in the vicinity.

 

[Ron247]

You probably will need to supply a section cut through your roof. It sounds like you do not have any ceiling joists to tie your rafters together at the top of wall and therefore you have to design your ridge board as a ridge beam. Problem is, your ridge beam does not go all the way to the walls due to the hip.

Is this a timber framed roof that does not have any ceiling joists at the top of wall?

 

[dhengr]

Trackman417:
Note that any unsymmetrical structuring or loadings makes using hip and valley beams all the more problematic when they are used to do too much of the structural work. On reasonably straight forward roof framing, good roof framing carpenters seem to have figured this all out, in part because they primarily worry about solid geometry, good cutting and fitting, and not too much about exact combined stress levels, unsupported length, buckling and deflections, etc. They just know what’s worked in the past from solid experience, assuming we don’t throw them too many curve balls. We engineers seem to have a hell of a time making these systems check out completely. Do you have ceiling joists at the wall top plates? Also note, that the bearing elevations of the various rafters and beams will vary if you want a constant overhang and facia elevation. So, you could have some trouble with your top plates as continuous tie beams, additionally so when several beams bear near a complicated wall plate intersection or wall corner. What connection detail do you provide at the lower bearing seat on that hip beam to transmit an 8k axial load and a 6k bearing load?

 

[bob33]

I assume the architect wants a sloping ceiling. Adding columns under the main ridge beam is one obvious solution, but you say there are load path problems. An expensive alternative is to redesign the sloping ridge beams (as struts) and the tops of the perimeter walls (as a continuous tie) to take the forces you mention. The main ridge beam would also have to be temporarily propped until all the rafter sheathing is in place, and secured to the walls, because the beams alone form a mechanism. My preferred alternative is wood scissor trusses. I would not rely on collar ties.

 

[MIStructE_IRE]

A section would be very helpful if you have one?

Have you seen the recent thread on the failed collar tied roof? Worth a read. Not sure how I link it here.

 

[trackman417]

I have attached three photos of the architecturals(I do not have autoCAD on my laptop of the work I performed yet). One of them is redundant, no insults intended. I just figured the more the merrier-if you guys work with visuals better than descriptions(like me).

MIStructE_IRE: I have just found that post and thank you for the reference

There will be ceiling joists. If memory serves me correctly, the ceiling joists will be about 2' below the bottom of the ridge beam.

Even if the ceiling joists did work to support the ridge board, I'm having trouble wrapping my head around if they would work due to the unsymmetrical roofing past the centerline of the ridge and the rafters being connected to the valley beam on the other side.

But I do understand that relying totally on the joists for support should be properly analyzed before the final/checked design is submitted. I will be pushing for vertical supports at both ends of these ridge beams. If the load path is wonky, so be it - it will be a "positive" path to the ground.

 

[HouseBoy]

I don't see 3 photos, only one cross section.

For ceiling joists to work, they would ned to be positioned at the plate height/bottom of rafters.
I don't think you have that ( per cross section and "2 ft below the bottom of the ridge"). If you did have that, I don't think the "unsymmetrical roof" would be an issue since I would position one beam in the ceiling at each end of the ridge (where the hip ridges meet the ridge).

Considering the volume ceiling, I think the hip ridges (14") will extend below the rafters (11.25") and I agree that the members (3 1/2" x 14 hip ridges) do seem large. The long valley beams (5 1/4 x 14 might not be too large but to me, that does define the depth you have to work with.

One thing I have done is to utilize "flitch-frames". Basically like a flitch beam but running up and down with the rafters and having a miter joint/welded connection at the peak. Your case might be a little more convoluted due to all of the roof planes involved but I'd try to get a "dogleg/flitch/ridge" to work in the location of the side dormer ridge. Looks like you might have enough depth. For example, instead of the 2x10 ridge board in that dormer, use an LVL that spans about 20 ft (can't tell exact dimension) across the room and underneath the "short ridge". Yes you still have some thrust at the tails of the hips but not at the valleys). Depending on the ceiling height, you might even be able to get a pair of flitch frames in there, one at each side of the dormer. I have used flat steel plates for this or steel channels and even wide flange beams at times.

As far as the thrust at the hip ridges themselves, I have uses an "L" shaped plate (in plan) on top of the wall plates with a saddle on top to connect the hip ridges to the wall plates. Metal straps at the plates would augment the need for continuous plates. It looks like you MIGHT be able to tie the corners together. If not, shear wall type behavior at the walls ought to work.

 

[trackman417]

Sorry for the other two promised photos. Seems a space does not count as a post to put photos in threads. Will post this last one as the third photo is just the exterior face of the building.

We have 3' from top of ridge beam to ceiling elevation and 8' from finish floor to bottom of ceiling

Thanks for all the opinions. I will definitely try one/all of them. First thing is to try and get the architect to move walls around so I can minimize the hip beam sizes and have a more direct load path. As much as I love talking theory about what works, it's a totally different animal putting it on paper especially without seeing any plans/calcs to supplement the design for construction.

 

[MIStructE_IRE]

Can you triangulate something like the attached? Would really need to see some full sections to get a proper idea of this though

 

[trackman417]

Sorry for the late reply. Figured I would spend the rest of my vacation not worrying about work.

Before diving in. I found a way to limit the amount of compressive load being thrown at the hip beams by dragging a ridge beam across to a door with a beefed up header which posts down to some LVL's into the floor below. This cut the amount of load being drawn into the hip beams by a fair margin(Now about 1500lb being pushed through the hips and about 1100 pounds being transmitted to the top chord).
[li]This load and arrangement is something I am comfortable with.[/li]

Mistruct_IRE, see attached photo with some dimensions requested.

HouseBoy. I like the idea of the "L" shaped angle to hold the angle as the hip beam pushes into it. Took it a step further to weld it to a mitered Channel that is then bolted to the corners of the top plate for 3 to 4 feet to drag the force into the top plates. Channel also helps keep the corners of the top plates from spreading/deflecting.(Not shown in pictures)
[li]--For sake of theory... If this saddle-type connection is 4 feet long in both directions, that would translate to 1100lb/4feet=275plf. Structural I sheathing with 10d nails @ 6" provide almost 950lb shear transfer for wind. Would an allowable load for this more "permanent" loading be the 950lb*(1.15/1.60)=680plf--Where 1.15 is the load duration factor for Live Load, 1.60 for Wind Load? Does that seem reasonable or not conservative enough?[/li]

Also,curious to see if there is any literature on this type of scenario? I am coming up with nothing online. Also took some time thumbing through "Design of Wood Structures" By Donald Breyer(very comprehensive book I bought in college), but nothing on this type of scenario.

https://files.engineering.com/getfi...4-49f9-444a-bd65-64874fd8bce2&file=Detail.JPG

https://files.engineering.com/getfi...b6e-43e6-b1cb-b3561d1649b6&file=Capture-3.JPG

https://files.engineering.com/getfi...-0886-48cb-8eb9-b35b8852ed74&file=Capture.JPG

https://files.engineering.com/getfi...b4c-4fa5-adb2-581a259a9af2&file=Capture-2.JPG

Thanks for all your help on this matter.

 

[jdgengineer]

Personally, I'd likely overframe the valley rather than have the two large valley beams you have, although that isn't really the main part of your question.

I wouldn't typically rely on hip thrust. I'd prefer a support point at the hip / rather intersection. More often that not I would just put in a welded kinked steel beam. If the loads were small I would consider a flitch plate like mentioned above.

If you can fit in a dropped cross ridge to span to the Dormer to reduce the load I'd highly consider that as well. With the reduced reaction, maybe you could get a large wood truss at the hip / ridge instead of the steel beam.

You may need doublers or more around the dormers to head out the main roof framing to allow for the dormer construction.