Roof suction and tributary area 1

[YoungGunner]

I was running my own calcs for a long ridge beam (30', 10ft of trib) and found that my roof dead weight was enough to nearly offset all the suction for components and cladding. The roof dead load is about 15psf normally and then you have the self weight of such a massive beam. This has me concerned in comparison with truss manufacturers whose girders always seem to have 1,000# or more of uplift. But then I got realizing that those truss manufacturers aren't using the entire tributary area in determining their GCp factors - they seem to use only the area of each truss, and then add those truss uplifts to the girder.

So my question is - if I'm designing the connections of my ridge beam to the support, do I use the entire tributary area of that connection for my GCp, or do I add up the uplifts of every individual rafter along the beam per their respective GCp and base it on that? Each will yield very different results.

 

[driftLimiter]

I ususally would use the tributary to the connection itself. But you might also want to check MWFRS uplift pressures on this beam and see which is worse.

In the roof design guide by verco, they do this for girders (check mwfrs) and they have some commentary in there about why and when to use C&C vs MWFRS that you might find valuable.

Link

 

[ChorasDen]

My argument is that you should generally use MWFRS loads for this scenario (ridge beam supporting roof joists). C&C loading is really for one single plane of surface load. MWFRS loads are multiple surfaces or larger areas. With a ridge beam where you have joists coming into the side from each direction, you have (2) different loading planes (left and right side roof). C&C does not seem appropriate to me because how would you expect to achieve peak C&C loading on both the left and right plane simultaneously?

 

[phamENG]

The rafters should be sized based on the effective tributary width of the rafter, the connections at each end of the rafter should be sized based on the tributary area of the connection (half of the tributary area of the rafter, not half of the effective tributary area), and the ridge beam should be sized for the tributary area of the ridge beam.

It's all about the probability of simultaneous loading. Wind is erratic and non-uniform. If you look at real time surface pressure readings from a live wind test, you'll see that there are spikes all over the place, as well as valleys. The smaller the area, the higher the probability of the entire area being filled with one of those spikes. The larger the area, the lower the probability that it will all be in a spike and the average wind load ends up being lower.

 

[Harbringer]

I agree with phanENG. The tributary area of the beam is 300 ft^2 that's what the member and connections should be designed for. Simply follow the load path.

My argument is that you should generally use MWFRS loads for this scenario (ridge beam supporting roof joists). C&C loading is really for one single plane of surface load. MWFRS loads are multiple surfaces or larger areas. With a ridge beam where you have joists coming into the side from each direction, you have (2) different loading planes (left and right side roof). C&C does not seem appropriate to me because how would you expect to achieve peak C&C loading on both the left and right plane simultaneously?

Regarding using MWFRS or C&C for this ASCE 7-16 30.2.3 is pretty clear on when MWFRS can be used in lieu of C&C (areas > 700ft^2).

 

[ChorasDen]

ASCE 7-16 30.2.3 is pretty clear on when MWFRS can be used in lieu of C&C (areas > 700ft^2)

Thanks, I'm well aware of this section, but that wasn't what I was trying to convey. I was more focused on the concept that for a typical ridge beam, we might expect a windward and a leeward side of the roof, each side of the ridge is a different roof plane, and thus, might be expected to have different design loads upon it at any given discrete moment in time. While the design level pressure may be the same for either side, I'm trying to indicate that I would not expect to see the design level negative pressure on both the left and right plane simultaneously, as they are two separate planes of the roof, with one leeward and one windward. I glanced through ASCE, and I do not see this in the code section. Generally speaking, I see roof truss systems designed with MWFRS loads for the same reason, in that the truss has multiple planes, unless they are monoslope trusses.

 

[driftLimiter]

Actually the definitions of MWFRS and C&C clarify this multiple planes idea.

 

[ChorasDen]

Ah, thanks DriftLimiter, that's what I was looking for, in chapter 26.

 

[XR250]

I'm in a 115 mph wind zone, but I never worry about isolated ridge uplift for closed spaces.
In my opinion, for the short term loading, the roof diaphragm act as a rigid member and transfers all the uplift to the rafter connections at the top plates.

 

[YoungGunner]

Appreciate everyone's responses so far. Has definitely helped me do a deeper dive into the differences between C&C and MWFRS winds.

However, this actually makes my situation even more confusing. If I were to calculate MWFRS winds, my wind pressure is less than the dead weight of my roof right off the bat. So I would ultimately get net uplift, which feels wrong. I was already getting 0 uplift for my C&C cladding options, but that at least started above my dead load but the hefty weight of the ridge beam was enough to push it over the edge. Because I see 1,000# uplifts on girder trusses reported by truss manufacturer's all the time, this feels wrong. Has anyone done these calcs recently and can confidently say that uplift should be a thing? My roof dead load (trusses/framing weight included) is usually taken as 15 psf.

 

[YoungGunner]

I should add that I am considering:
1) the vertical component only relative to the plan (cos(angle)) which reduces the pressure a bit
2) we live in an area of higher elevation (>4,900ft) which is reducing the wind by 16%
3) My beam is a hefty glulam weighing about 27plf
4) I'm designing everything in ASD and am using the 0.6 factor to reduce my dead load (as well as my wind loads)

 

[RWW0002]

It is possible to get a 0 net uplift condition for a longspan ridge beam with pretty small tributary width.

That being said, a 2 psf net uplift over the 10' trib width seem pretty small, even for MWF wind (if 0.6 * 27 plf beam dead load is overcoming it). You might want to recheck your wind calcs. If 105 mph wind and the high elevation you mention you might be correct, but that sounds a bit low to me.

Also, I generally use 0.6 * actual DL for uploft calcs. Where I might use a conservative approximation of 15 psf or 20 psf for dead load on the truss, I run uplift calcs using 0.6*(less conservative) DL of 12 psf or so. Basically just recognizing that most truss systems are much lighter than we call for them to de designed for.