Squirrels, birds, and bees Oh my

[bob-n]

Someone asked about lateral expansion. According to below website, aluminum expands roughly 24 parts per million per degree C.

https://www.alumeco.com/knowledge-technique/general/thermal-expansion

If you install at 20 C = 68 F and the panels can cool to -20 C (-4 F), then a 78" long (2 meter long) panel will shrink 78*0.000024*40=0.075" (1.9mm.) If the same panel heats up to +60C (140F), then the panel will expand 0.075" (1.9mm). A good installation will allow for at least +/-0.1" or +/-2.5mm expansion in the long dimension, either with gaps or flexible mounting. One common technique is to space the panels with a scrap of 1/4" or 6mm plywood or similar size bolts during installation and remove the spacer after tightening down the panels.

[J.P.M.]

Someone asked about lateral expansion. According to below website, aluminum expands roughly 24 parts per million per degree C.
https://www.alumeco.com/knowledge-te...rmal-expansion
If you install at 20 C = 68 F and the panels can cool to -20 C (-4 F), then a 78" long (2 meter long) panel will shrink 78*0.000024*40=0.075" (1.9mm.) If the same panel heats up to +60C (140F), then the panel will expand 0.075" (1.9mm). A good installation will allow for at least +/-0.1" or +/-2.5mm expansion in the long dimension, either with gaps or flexible mounting. One common technique is to space the panels with a scrap of 1/4" or 6mm plywood or similar size bolts during installation and remove the spacer after tightening down the panels.

Or, 24 parts per million per deg. F.

But there's a little more to it than just the thermally induced dimensional changes of one material or component alone.

As one example only, consider what happens to the panel glazing which has a coeff. of linear thermal expansion that's about half that of the aluminum frame. Reality there is that the glazing seals handle most of the temp. induced differentials in assembly length dimensions between the two materials, but the alum. will put a compressive stress on the glass through the seals when the panel is "cold" but that stress will (hopefully and mostly) be absorbed by the seals in spring like fashion. The thing that usually/often kills glazing seals is the cyclic nature of the stress/movement over diurnal or more often thermal cycling over many years as well as sealing material that loses its flexibility.

Most racking components including posts and clips are considered "fixed" points at the anchorages in a design sense, but they can rotate (some) at the panel and/or rail attachments thorough a moment and retain their strength. A 20 ft. long array dimension going through 70 F of temp. excursion will change dim. by ~ +/- 3/8" or so, or ~ +/- 3/16" at either end perhaps several times a day, for 20 or so years. Most such fixations don't fail because what they are attached to and how they are attached do have some flexibility. That is, the materials they attach to - usually/often wood or if metal, some moment allowing arrangement such as a cantilever and will act somewhat like a spring to absorb the movement.

One other thing to consider, among many others, is that the fixations at the roof such as posts, which are considered fixed at the roof for stress design purposes, will have less thermally induced angular rotation on them as they get longer, but usually there's enough actual flexibility in the roof materials, including rafters and decking to take up for most such thermally induced dimensional changes.