I am currently designing a rooftop geometry, based on combined cylindrical surfaces.
When I unroll these surfaces, although they are mathematically developable, I noticed, that there are area deviation from original to unrolled surfaces… Is it a bug? One surface generates a deviation of 0.05% multiplied by 80 times gives me a deviation of ca. 4%… To much…

Why is there anyhow a deviation although mathematically they should exaclty match…

If multiple surfaces each have a deviation of 0.05%, then the total deviation will be 0.05%. Percent deviation does not simply add for multiple surfaces.

Total percent deviation of multiple surfaces = sum of the area changes / sum of the original areas * 100

Percent deviation = Change in area / Original area x 100.

Change in area = Original area * Percent deviation / 100

Example: 80 surfaces, original area of each surface is 50, percent deviation of each is 0.05%

Most likely your surfaces as modeled are not exactly developable. Post a .3dm file with an example of a surface which has the deviation. You can use the vertical arrow above where you type your post.

Attached you get the 3dm file. All surface are based on ExtrudeCrv. So should be 100% cylindrical.
I tested: If i cut the cylindrical surface in a diagonal angle, than it generates the deviation in area calculation between original and unrolled srf.