Roof Rafters Question 4

[DevinTheStudent]

Hi all,

have a few questions about collar ties and roof rafters in a vaulted ceiling. See attached photo for more information.

Q1) In this situation do the collar ties help support roof rafters and can this be considered a support allowing a span of 10'-6''?

Q2) Rafters should have been bird mouthed however can leaving them as is still distribute load accordingly?

 

[SteelPE]

1) I always thought collar ties were to help with roof uplift... which is why in the US they needed to be located within the upper 1/3 of the roof rafter. The help prevent the rafter from separating from the ridge board under an uplift scenario. If you have a structural ridge beam and can make a proper connection against uplift I am not even sure collar ties would be required.

2) I may be wrong with this, but I have always look at a birds mouth as providing providing a way for the roof rafter to transfer the load to the supporting member through a pure "vertical" load path. If the support is sloping (like you would have without a birds mouth) then there is a lateral component to the load that needs to be accounted for. This load may be able to come out through the roof sheathing, but it's something that would need to be resolved.

 

[1503-44]

The collar is a "moment" connection making the two rafters a continuous beam, but bent shaped. The stronger that upper triangle is, the less spreading from kick-out load you will have at the wall bearing points. It would help if you drew the collar tie to scale at the right place.

Just my opinion, but I think the 2x6 rafters are way too small for this application. My rule is 1" depth for each foot of span, more or less. You will effectively 10'-6" cos(pitch angle) cantilever beam. You also have a pretty good snow load creating a large moment at the 10-6" connection point. Will a 2x8 work? You should check the bending stress there. I think the 2x6 would have too much deflection there as well.

A bird mouth fits better to direct reactions down into the wall. You may experience more pronounced kick-out load on the wall without that feature.

 

[phamENG]

1) Your span is looking at the full length of the rafter, not the horizontal span length...have you adjusted your loads accordingly? If you're using span tables, they consider the horizontal projection of the joist, not the true length. Using the true length is theoretically correct, but using the horizontal projection gives you comparable results 99% of the time and is usually faster.

No, the collar tie is not a support.

2) You should NOT birdsmouth or otherwise notch a rafter at the top. Provide a beveled plate on top of the steel beam. If you're really concerned about the beam flange, then determine the line load required to cause yielding in the flange...a wood structure usually won't get there.

 

[LittleInch]

A pity you didn't draw the rafter ties a bit more to scale in location.... so they're basically a little more than half way up the rafter?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[DevinTheStudent]

Drawing is to scale, each square in the drawing is roughly 2ft

 

[Ron]

A1 to Q1....No, the collar tie is not a support. It is there to prevent the rafters from putting too much lateral thrust on the walls. It is simply a ceiling joist that is raised to a high point in rafter system instead of being on top of the wall where it would be tied to both the wall and the rafter, making it a more robust system. "1503-44" explained the thrust; however, in my opinion the collar tie does not create a moment condition.

A2 to Q2.....If there is an appropriate steel strap or appliance to connect the rafter to the top plate, birdmouthing is not necessary.

In theory, if your ridge beam has sufficient capacity and little or no deflection under load, the collar tie is not necessary. In practice, it is better to have the collar tie since even a relatively small deflection of the ridge beam will impart thrust to the top plate.

 

[IceNine]

Collar ties at the location shown will do next to nothing in reducing the thrust at the top of walls.

 

[LittleInch]

My apologies, I read your 13 ft on the drawing as 18 ft....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[dauwerda]

There have been multiple posts on the subject of collar ties (and the differences in verbiage in different parts of the world).

In general (and using the definitions used by the IRC), a collar tie is placed in the upper 1/3 of the rafters and is there to prevent issues with separation due to uplift as noted above. It is not for dealing with rafter thrust, that is a rafter tie or ceiling joist, which is required to be in the lower 1/3 of the rafter (again per the IRC). A collar tie is not required if other means are used to prevent separation due to uplift (like a ridge strap).

Your detail shows a steel ridge beam, so I would assume that beam is sized to carry the vertical loads at the peak and rafter thrust is not an issue for you (therefore no rafter tie or ceiling joist is needed).

The collar tie is not a support. Also, as noted above, the rafter span is typically taken as the horizontal projection of the rafter. So, your rafter span would be the horizontal distance from the wall to the steel ridge beam.

I would add triangles (beveled plate) below the rafters at the steel beam as noted on your sketch. A birdsmouth at the top is not a good idea as it will greatly reduce the shear capacity of your rafter.

 

[1503-44]

Ron, the moment is at the 10'-6" position of the rafter.. is what I meant.

I also thought it was 18ft. It still looks like 8. Better do some work on 3s & 8s.

 

[CivilSigma]

I would like to add that, in Canada, Ontario, if the roof slope is greater than 1:3, and the collar tie is or larger than 2x4, then it is assumed to provide an brace point to reduce the rafter unsupported span (

http://www.buildingcode.online/1868.html)

So, for your question No.1, I would say: Yes, the collar tie will reduce the unsupported span (if rafter slope is greater than or equal to 1:3).

 

[DevinTheStudent]

Thanks everyone for your input glad this allows for a good discussion!

 

[Ian 2017]

I have a similar issue. Birdmouthing transfers the load vertically to the top flange of the beam, otherwise it’s not doing much. I appreciate phamENG concerns of shear. This is done at the wall end of the rafter but we more area to work with: The beam is installed and would make this very difficult anyway.

That considered, I’m proposing to add 2x4 ties above the top flange of the beam such that the 2x4 bears across the entire top of the flange.

Thoughts appreciated!

 

[LittleInch]

Please start you own thread and add some drawings.

Hijacking someone elses thread is frowned upon. Link to it by all means.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[XR250]

I ran it in my 2D frame program. It does work but it puts a huge compression in the collar tie that resolves in the rafters wanting to pull away from each other above the ridge. If you are able to support the collar tie verically from the ridge and it is stout enough to take the vertical load of the rafters then it removes this load. You could also tie the rafters together above the ridge to resolve the collar tie compression.
There is never a free lunch with compression collars.

 

[1503-44]

Where do you ever get free lunch? Only at mom's house.

 

[XR250]

...and that isn't always "Free".

 

[southard2]

I'd like to reiterate what another poster stated already. If the ridge beam is not supported than you need to have continuous ceiling joists. I've seen far too many A-framed rafters or trusses thrust the tops of exterior walls outward. I can think of at least four instances where I have seen this. And I suspect it happens a lot more than people think. A frame trusses simply do not have enough stiffness.

Never mind sorry I see that your wrote that it is a steel ridge beam so I assume it is structural. As far as the collar tie a lot depends on the loading. If your loading on each side of the roof is balanced than it will be in either compression or tension. But with unbalanced loads I suspect the heavier side will transfer some load to the lessor loaded side.

Honestly if this was a real design my advice is always keep it simple and design those rafters to bear all the weight by themselves. Eventually someone will cut the ties, leave them out, nail them improperly, etc....

John Southard, M.S., P.E.

https://www.pdhlibrary.com/