Tied roof design module is glitchy

For anyone using the rafter tie module - please be aware that the deflection checks will be based on the minimal span that inevitably results along the rafter from the end support to the first tie location. Because the module doesn’t seem to allow the bottom tie to be exactly at the bottom (see below), it can be raised up an inch or two to enable a reasonable calc of the rafter section. However, that 1" or 2" resulting bottom length will show an excessive deflection issue. To correct this, I typically change the deflection limits from L/360 or L/240 to a very high limit, such as L/10.

For example, I have a current rafter design with 13’-6" horizontal distance form eave to ridge and two ties - one at the bottom and one 3ft down from the top. The calculated long-term deflection is 0.00433" or L/500. That tells me that ClearCalcs is actually checking deflection of a 2" member span, or the length at the bottom of the rafter. The problem might go away if the tie was right at the bottom of the rafter but then I get an error saying “Multiple supports share the same location; please remove any duplicate supports.”

Note that the ClearCalcs rafter span “L” also includes the eave extension. Set the eave size to 0ft to avoid this issue.

Hey Jim! Big thanks for sharing your thoughts and experience here.

The deflection on eaves is definitely something we want to handle better. As you mentioned, with the short eave span, L/240 on a 2" span is about 8 thousandths of an inch… Not very meaningful!

I’d be curious to hear how you (and others here) handle deflection with roof assemblies. In the image below, which would you typically consider?:

  1. Absolute deflection at ridge (for your L/ value, would you use L = total roof span?)
  2. Rafter deflection (would you use L/ based on total rafter span or based on span between wall/ridge & tie?)
  3. Deflection of roof eave (checked against 2x the L/ value since it’s a cantilever)


It’s not even the deflection of the eaves that is the problem, the program seems to be calculating the deflection between the wall support and the point on the rafter where the collar tie is located. Somehow, the deflection and the L/xx check are not performing as they should.

My gut feeling is to keep the check as if it were for a regular simple span rafter with L being the value from the support to the ridge. Check the deflection in the middle of the rafter and compare to L/180 or L/240, depending whether the tie is at the bottom, or the tie is raised, respectively. The collar tie affects the rafter deflection, but the deflection check can still be simple.

It seems that the programming tried to do too much by checking L/240 for the short section of sloped ceiling with sheetrock on it when that check is probably not going to yield reasonable results for such short distances.

Thanks Jim for the detailed response and understood about the L value not matching what you’d expect with the collar tie. I’ve filed a ticket on our end to fix this.

For the eave deflection, I just double-checked and we do have an option to disable the eave deflection checks, which should at least help you avoid the first issue.