Scissor Truss Framing and How to Analyze Correctly?

[BA_zach]

I am designing a small wood structure (barn), and the roof structure is similar to that of a scissor truss concept with kickout forces and I am looking for advice on how to analyze the scissor type framing correctly. For this discussion the geometry of the framing and boundary conditions can't change, so no the collar tie can't be located at the bearing. I've provided dimensions only for reference but am looking for the correct way of analysis scissor truss type framing. In the imagines, you'll see the framing is rafter's with a ridge beam and a sort of collar tie across to the rafters. The rafters will be bearing on a wood stud wall each size of the truss, which is where I show the boundary conditions.

In the first imagine, I've showed the method of restraining the horizontal reactions, which means I am counting on the wall to provide that rigidity. You'll also see the member utilizations of the members and the deflected shape (it is shown as magnified for this discussion). Because the frame is restrained horizontally at the bearing and thus not deflecting by kickout the member utilizations are within allowables - though close to unity.

The second image is the horizontal restraints as essentially nothing (very light springs), resembling the wall to not provide the stiffness and restrain the horizontal kickout. Thus allowing for the deflection to happen within the wall and roof framing. In this method, allowing the deflection to happen, the member utilization is way over unity due to the collar tie happening along the length of the one rafter. This is creating a high bending stress in the member.

My question is which approach is the correct approach. Consider the wall as restraining the kickout? Or allowing the kickout and deflection (within allowable deflection ratios) and sizing the members appropriately for the forces.

My thought is, how could you consider the wood stud wall to provide that rigidity? This would result in a shear at the top of the wall, creating an overturning moment at the base - which would be impossible to resolve. I'm thinking you have to allow the deflection, but limit the deflection to be within allowable for the building material and then size the roof framing members for the subsequent forces. I look forward to the discussion here and thank you very much in advance!

 

[phamENG]

Unless you're using concrete or masonry walls with some nice buttresses (please use flying buttresses....they're due for an architectural comeback), then the walls will NOT restrain those reactions.

If you have a ridge beam, your model is incorrect. In fact, a ridge beam is probably the way to go. Though if the building is 25ft wide, how long is it? That ridge beam could get pretty big.

If you have a ridge board, then your model is accurate. Keep in mind that the deflected shape is exaggerated. Useful tool for understanding how your model is moving, not for how much. Be sure to look at the actual deflection values.

 

[Celt83]

Run a search through the forum this has been brought up a handful of times.

The "best" option I found for modeling this, using the sign convention from your screenshot, was to assign Y and Z reactions to your wood wall bearing points, so remove the x springs you added in. Then apply an X and Z reaction to the ridge this reaction will need to then be hand checked to resolve through the roof diaphragm and out to your lateral resisting elements. As an alternate assign only Y reactions at the wood walls while keeping the X and Z reaction at the ridge that way your model isn't relying on any out of plane reactions at the wood wall.

My Personal Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

Open Source Structural GitHub Group:

https://github.com/open-struct-engineer

 

[BA_zach]

@phamENG agreed on the walls and not considering restraint with wood stud walls. So you basically design for the deflection and size the members accordingly. You are correct, and I spoke incorrectly, I have a ridge board. Thus, not relying on a vertical support at the ridge. The rafters, if equal span and equal loading, would have opposite horizontal forces at the ridge board and cancel out. Yes, the deflected shape is exaggerated, and in reality isn't deflecting much. But the deflection is what is causing the excess bending in the right rafter. Thank you for the input.

 

[JoshPlumSE]

I wouldn't really call this a scissor truss. It's more like a rafter system for the roof with a tie beam.

Here's how I would model it. Pinned on one side and vertical support only on the other side. Check the deflections (vertical at the ridge and lateral at the support). If either is too large, then this probably won't work.

 

[BA_zach]

@Celt83 I will search the forum and see what else is out there, thanks for the tip. I do like that way of modeling and then checking diaphragm capacity for distributing for out to lateral system. I do make it aa habit to check reactions to ensure there aren't reactions which shouldn't be there. Thank you for the input, and I'll be sure to do some reading on the other threads discussing this.

 

[dik]

see what sort of numbers you get with a pin and roller support...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[MIStructE_IRE]

I’d call this a raised collar roof, not a scissor truss. I’d do this as a pin and roller.

In reality.. its a roller - roller as both walls will deflect outward. We tell ourselves we need a pin and roller for stability.. so whatever your roller outward movement is, split it between the two walls and make sure its less than wall height/300 or whatever your code permits.