(note that I can't add images due to points)
We've been performing some sensitivity analyses on various parameters to a very simple model for the purposes of shading impacts. The results we see typically align with our expectations, however strange behaviour manifests when the height and width of the wall increases.
We vary the height and width X of a wall from between 3 m to 1280 m.
The vertical wall is X m wide * X m high pointing north (southern hemisphere / Melbourne Australia), and a 1m x 1m window is in the wall.
The size of the window remains the same for this study, and its position remains the same in each run - 1 m up and 1 m left of the bottom-left corner of the wall. We model two scenarios - one with no shading fins, and another with shading fins on both sides. Then we collect the incident solar radiation on the window and calculate a ratio between shaded / unshaded results.
While the shading ratio remains 'stable' at small wall dimensions ( 3 m through to 40 m ), this changes as the dimensions increase. See below for an excerpt of the results. Remember - the window and its shading remain the same
W 3 m / H 80 m => 0.720
W 10 m / H 80 m => 0.720
W 20 m / H 80 m => 0.720
W 40 m / H 80 m => 0.721
W 80 m / H 80 m => 0.722
W 160 m / H 80 m => 0.722
W 320 m / H 80 m => 0.746
W 640 m / H 80 m => 0.796
W 1280 m / H 80 m => 0.894
My expectation is that no matter the size of the wall, the shading ratio would always remain the same.
It is difficult to determine what is going on here, though I'm assuming it's a product of the shading algorithm.
This model is being run with the 'FullExterior' Solar Distribution
