(highlighting with bold is mine)
Originally Posted by
Bob Knauff
I have become convinced through research and interviews with contractors and repair type folk that some of the deflection is due to improper installation practices by the installers of the roof sheathing.
...
It is always recommend that the installer stand on the sheathing directly at the rafter to promote good sheathing to rafter contact as it's being fastened but if the sheathing is a thinner type, it was made quite clear that if the installer stood to either side of the rafter while nailing, the sheathing may sag and create a dip in the surface plane of the roof after being nailed in place.
Bob,
Improper installation practices are a definite contributing cause, but not the full cause.
The thinner (less than 5/8") sheathing is the main cause.
The rating on the sheathing for allowing a 24" span for 3/8" sheathing does not (for any span rating for any wood, including structural panels) make ANY consideration for "building aesthetics". The span rating is simply a structural rating of what it will withstand within the limits of its design, including safety factor.
Considering 'H' clips as "edge support" helps attain the design load rating, but, as we all know, does not help keep thinner roof sheathing 'in plane'.
The fact of using thinner roof sheathing panels is the main cause (by far) for roofs which sag or deflect 'out of plane'.
Getting back to installation problems, one big installation problem is with the span rating and cutting of the panels.
APA recommends no less than 1/2 a sheet be used ... that's no less than 2' x 8' or 4' x 4' ... and remember, like lumber, structural panels are rated for installation over two spans, three supports - MINIMUM.
Those small pieces, they are just a no-no, but it is down ALL THE TIME.
Now, though, we must get into the roofs in the photos which show an 'out of plane' area which is larger than the spacing of two trusses or two rafters ... that now becomes an installation issue, installation of the trusses or rafters, where they are not installed 'in plane' across the top of the top chords of trusses or where rafters are installed which are not 'flat In plane', i.e., the rafters have a crown to them.
If the supports are not 'in plane' how can the roof sheathing which is nailed to those supports be expected to be 'in plane' when they are installed?
With trusses, even being made in a truss 'factory' setting, 'identical' trusses are not "identical", some slightly longer, some slightly shorter, some slightly taller, some slightly less tall, set one of each side-by-side and the top chords will be 'out of plane' from the start.
From there, it's all downhill.
Kind of like what I was telling a window manufacturer two days ago, if they keep changing their engineering to match the installation instead of changing the installation to match the engineering, when it goes to court, their case will be like a soap box derby racer ... it will go downhill FAST.