GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Question for Dana

Michael Lind | Posted in General Questions on


In  Tyleralex1’s  comment # 3    polyiso roof insulation

You said  “timber-screwed to the structural roof deck (not to the rafters)”
what is the reason for this

Mike Lind

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Roger Berry | | #1


    My unvented exo-inusulation roof used nailbase panels up to 7 3/4 " thick EPS for the most part. I had to fudge thickness vs R on one roof and used 2" iso under 4" nailbase. In all cases the crew was very good about sinking the 8-10" screws into the framing. The pull resistance of timberscrews set in nominal 5/8 ply decking would not likely match that of being set 1 1/2" into the framing elements. I would not ever rely on the decking to hold timber screws.

    I can't address why Dana may have said to place the anchors in the deck material. I do know from practical experience that the screws do thermally conduct significantly. Picture attached.

    I suspect that long screws hanging out into the bottom side of the deck sheeting would create potential condensing points. Would it make a lot of water collect? Good question. I don't want to get the wrong answer about that in a part of the house I hope never to see again.

    Exo-insulation with foam of any type means setting up for drying to the interior as the moisture control failsafe. I trust my air control details for outside air, but the interior air is only restricted by drywall and paint. If for some reason I was to build up a lot of moisture load in my air, the batt insulation under the deck would permit that air to contact any screw points piercing the deck. I checked for such misses and found only one out of a few hundred long screws. So I guess I won't ever know if my thoughts can be certified.

    Even embedded in the rafters, enough interior heat escapes to make the dimples in the snow cover. Despite much commentary to the contrary, I find that a 4% heat loss over a 4x8 panel seems credible. by my calculations. Stainless steel screws would cut that. For my next build, current thoughts are taking me toward a multi-layer foam system to blank out the screw losses as well as reduce the need for very long screws. I may crash and burn on the labor side.

  2. Michael Lind | | #2

    I know that several nailboard Mfgs show nail patterns that do not line up with normal rafter spacing

    1. Expert Member
      Dana Dorsett | | #3

      Instructions from some of the nailbase panel manufacturers explicitly state it should not be screwed into the rafters.

      I don't know for sure, but I think it's to avoid splitting the rafters from stresses from seasonal dimensional changes in the foam or nail base layer that differ from the seasonal changes in the structural roof deck. Screwing it into the roof deck delivers the pressure any strain or creepage to the deck, which distributes it to the rafters over much broader area.

      A related alternative theory might be to keep the stresses from snapping the screws. Roof decks and wall sheathing are normally nailed rather than screwed because nails will only bend rather than snap under shear stresses, whereas screws are much more brittle and prone to breaking.

  3. Michael Lind | | #4

    So on a hot humid day the top panel expands 1/8 of an inch [ any more and it would buckle ] and the bottom panel stays put, in 1 inch of foam that is a lot of back and forth bending in the screw. At 4 inches of foam, significantly less flexing. Makes sense.

    I get the impression that timber screws driven 1 to 1.2" into 3/4 CDX makes a very strong connection

  4. Roger Berry | | #5


    I just reviewed the installation guides on three different manufacturers of nail base and the literature does not address underlying structure beyond concrete decking (typically commercial only). The general expectation seems to be that nail base will be used over steel decking or 1" wood decking. Again, largely commercial situations. The charts do cover (nominal 5/8) 19/32 plywood, which on most charts requires more fasteners per panel. 15/32 ply might well require more fasteners still.

    Many charts and diagrams are based on a two foot pattern that is agnostic on framing under the field. It could be taken to mean "don't hit the framing" to 16" framing centric building, but I suspect it has more to do with locating the fasteners across the panel for convenience and lateral load distribution.

    Atlas lists its screw fasteners for the type of material being screwed to along with the roof pitch and snow load parameters. Fortunately for me, the very low pitch I have did not require my putting in 35-40 fasteners per panel. That would have been quite the radiator array.

    I did see references to stacking nailbase in 2" layers to reduce the thermal bridging, so I guess I was late to the game on that one, though I can't say that four layers of nailbase would be good cost choice versus the heat loss of the long screws.

    If you are screwing into 3/4" nominal ply decking you will find substantially more "bite" than with 1/2", especially if you hit an internal void. A .19" body diameter on the screw means your side support ratio goes to almost 4 to one in true 3/4" material. Plus more threads are engaged with the material to increase pullout strength. Now that 1/2 panels are on a weight loss program, the support of body and thread gets that much more tenuous.

    Your perception about relative displacement of the screws with 1" foam is on target. Metal roofing is popular in our area, but the face screwed kind is known to wiggle the screws loose over time. The expansion shifts over a 20' run of roof panel are quite large when the sun makes the panel 135F which then drops to 45F around 3AM.

    Not being fond of heights or maintenance that can be avoided, I went with the clip style standing seam roof panels. I tell myself that the expansion and contraction is gliding over the clips that are screwed to the roof deck thus alleviating the push-pull on the screws.

    Not clear on your project, but I would look closely at your pitch, snow loads and deck thickness when making final fastener selection.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |