Code evaluation of rafter, joist, and collar tie geometry

[charliealphabravo]

Hi all,

I have been asked to do a code evaluation for a conventionally framed residential roof (4:12) in Alberta Canada that has failed under snow load. The snow load was most likely at or just past the design load.

I am using the 1965 NBCC in the review. The code gives maximum rafter spans, say for example 11'10" for a 2x6. The rafters used were 2x6 spanning 15' to an unsupported ridge plate. The code also requires all structural members to safely support the design loads and prescribes the number/length of fasteners for each joint (rafter, ridge, ceiling joist, collar tie).

I have some general concerns with doing a code evaluation for a structure this old...but my specific question is whether it is appropriate to identify the rafter span as a code deficiency when the accepted practice of the day may have been to use collar ties to increase the effective span of the rafters. The collar ties where 24" below the ridge so that the 11' span could have been technically met if the collar ties were viewed as "supports".

It is clear from a modern analysis and literature review that the rafter set does not work (connection forces, unbalanced load at the ridge intersection, no fasteners at the rafter-to-1x6 ridge plate, etc) and this is also why the roof is failing and why the construction practice was abandoned. But is it inappropriate to use modern methods and knowledge to show that the structure did not meet code in this particular case?

Thanks to all in advance.

 

[dik]

It's been a decade since I did up a report on the failure of conventional framing... will see if I can dig up some info... surprising matter, and observed in the 'punching' failure of the collar ties through the roof surface that they can also be loaded in compression depending on the geometry...

If the thing has been in service for the last 40 years, why has it failed?

Dik

 

[charliealphabravo]

This is my first winter here but I understand that there was a lot of snow in the area last winter. This is consistent with the records of Environment Canada. I suspect that the roof has been getting progressively worse over the years and that this past season was the last straw.

The roof has not collapsed but the ridgeline is sagging up to 5 or 6 inches, particularly where the two gable ridges intersect. There is little or no support at this location except for short posts or struts under the 1x6 valley rafters. I don't think these systems works on paper but they seem to survive the years on redundancy, friction, opposing forces, and reserve strength.

 

[dik]

The roof often acts as a 'folded plate' type of structure... sags often occur because the eaves push out due to connection problems or the removal of collar ties/ceiling rafters...

Dik

 

[woodman88]

Attached is an article on a similar problem with a framed roof.

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[dhengr]

That roof lasted forty or fifty years, probably under some similar snow loads as those of last winter, check this snow load history out. Those old carpenters and codes were not working in a vacuum, it is surprising how much they knew from their building experience, and earlier failures in what they tried to do, etc. You have to hedge your report to the extent that you can’t fairly expect them to have known what we have learned over the intervening 50 yrs. The codes have changed radically in that time, the materials have changed in quality, etc. and all of this must be taken into account. You have to look at the codes of that era to see the standard level of practice of that time. Look at old plans and designs of that era too. We can analyze structures far more completely and accurately than we could 50 yrs. ago, modeling them correctly may be another matter however. For the most part, the question should be did the roof structure meet the codes and building stds. when it was built. This is not intended to excuse a joint which failed and only had one nail in it when there should have been three or four, by the then current stds. This type of give and take, or understanding of the progression in our knowledge and methods, must be included in your final report.

 

[msquared48]

Quite frankly, I am surpried that they tabularized a collar tie arrangement.

That being said, I am further surprised that 2X4's would work for rafters considering the edge spacing requirements and the tension force generated with raised collar ties rather than resting on the exterior wall plates.

Could be that this was just a failure waiting to happen.

Mike McCann
MMC Engineering

 

[charliealphabravo]

Thanks all,
The span tables were based on actual clear span, not a collar tie arrangement, but today I discovered a section in the same code that explicitly defined collar ties as a support for the purposes of determining rafter spans. In other words, it looks like the rafter spans in my case technically meet code so my dilemma is solved for the moment.