Roof without flat ceiling - how to resist thrust?

[buleeek2]

Hello,

I am wondering if there is a way to utilize wood trusses like in the picture below. The owners do not want ceiling joists and for the sake of the discussion, let's say that a ridge beam is not an option (the span would be over 35 feet). I am worried about the thrust, as once I tie the trusses down to the walls it will create a horizontal pin at each wall. Does anyone have a suggestion on what to do in such a situation? Maybe a bracket that allows for lateral movement?

 

[jayrod12]

Usually we get the truss supplier to design around it being pin-roller, but keep the lateral deflection to 1/2". Then just assume the walls are flexible enough to accommodate 1/4" each at the top of the wall, which is generally tolerable.

 

[SinStrucEng]

We do the exact same procedure as @jayrod12 described above.

 

[buleeek2]

Thanks for the replies. So, the truss supplier will keep the deflection to 1/2" and the EOR assumes that the walls will not resist lateral loads at the top?

What would you recommend to do for the wall lateral deflection due to wind? The shear walls will be approx. 35ft apart. See below.

 

[jayrod12]

35 feet isn't that far for the roof plywood to act like a folded plate diaphragm. The detailing is a bit onerous if you truly want to get everything to calc out. But it's done fairly regularly.

 

[TRAK.Structural]

I think what @jayrod12 and @SinStrucEng are saying is to fasten the scissor trusses at each wall with a pinned connection, and checking the walls for additional stresses due to 1/4" of outward deflection from the thrust. This allows you to use the roof structure to brace the top of the walls and transfer wall pressures to the roof diaphragm.

 

[jayrod12]

I think what @jayrod12 and @SinStrucEng are saying is to fasten the scissor trusses at each wall with a pinned connection, and checking the walls for additional stresses due to 1/4" of outward deflection from the thrust. This allows you to use the roof structure to brace the top of the walls and transfer wall pressures to the roof diaphragm.

That's correct.

 

[buleeek2]

What would you recommend to pin the truss to the wall? I usually go with H2.5A ties plus blocking in between trusses, or a35 angles.

 

[jayrod12]

What would you recommend to pin the truss to the wall? I usually go with H2.5A ties plus blocking in between trusses, or a35 angles.

As long as that calcs out for your uplift, and lateral loads, I see no issue there.

 

[XR250]

fasten the scissor trusses at each wall with a pinned connection, and checking the walls for additional stresses due to 1/4" of outward deflection from the thrust. This allows you to use the roof structure to brace the top of the walls and transfer wall pressures to the roof diaphragm.

As been discussed in many previous threads, I don't think anyone in wood construction does this. What do you do when you get near the corner?

 

[TRAK.Structural]

As been discussed in many previous threads, I don't think anyone in wood construction does this. What do you do when you get near the corner?

I've read some of the old threads, seems like many are against the slip connectors that are intended to relieve the thrust, the other option is a horizontally restrained connection (i.e. pin).

What's your concern at the corner? Perpendicular wall stopping the bearing walls from deflecting outward? I haven't had issue with this but maybe you could use those slip connectors for a couple feet near corners to relieve the thrust. Or maybe try to enforce making the connection after the dead load is applied, although this seems very unlikely to happen.

What do you do?

 

[jayrod12]

I pretend the corner isn't that stiff to be completely honest. Often with scissor trusses like those shown, I frame a full height wall and the first truss doesn't start until 2 feet in from the corner. so there is already some flexibility in the wall system by the time you get to the first truss.

When it is just a truss at the outside, there's still often a wall framed tight to the underside, so now it's a gable truss that doesn't need the pin-roller support conditions and therefore again, the first truss needing some flexibility in the walls is 2 feet from the corner and likely there's enough movement there.

Although I'm generally not a fan of this next statement, and please everyone don't bother pointing out how "dangerous" the coming thought is because I've heard it enough, this type of detailing has been done forever without issues that I've been made aware of. I'm yet to see a bottom chord of a scissor truss buckle laterally because it is now in compression because the walls weren't flexible enough to produce a pin-roller support. If someone has evidence of that failure mode, I would love to see it so I can adjust my mentality going forward.

Regarding the slip connectors, they end up not working anyway. Between the roof plywood being fastened to blocking that is in turn fastened to the top of wall rigidly, and the interior finishes on the trusses, you won't ever see those things actually move. and if you did, you'd be constantly having to repair your drywall joint or the exterior finishes as the trusses move in-out. It just doesn't happen.

 

[XR250]

The bottom line is if you keep the spread reasonable, it just works. As you near the corner the roof diaphragm does a bunch of folded plate stuff to reduce the spread.

 

[jhnblgr]

Run a ridge beam with two cross beams like a king post. Make the structure something people look at instead hiding it.

 

[jayrod12]

Run a ridge beam with two cross beams like a king post. Make the structure something people look at instead hiding it.

That's all well and good, but money talks and trusses are the best bang for your buck. If I tried forcing ridge beams on every client, I'd be out of designing houses where I practice.