Hello parilla88,
If I am interpreting your images correctly, the spun profile becomes less accurate in the region with arcs as the tolerance is increased. I have a couple of question:
1. Is that interpretation correct?
2. does the spun profile vary more than tolerance from what you would expect it to be?
All the best,
Tessa

Hi,
1. Given that the shape is cylindrical I would have expected the tolerance to have no bearing on the spun profile.
2. Increasing or decreasing the tolerance is not proportional to the results. A tolerance of 0.01mm appears correct where as both 0.1mm and 0.001mm are not. So the spun profile resolution is not improving or degrading as the tolerance is increasing / decreasing.
Hi parilla88,
The incorrect spun profile at tolerance 0.001 does look like a bug, and if you send me the fusion file, I can open a bug report for that.
The outcome for tolerance 0.1 is not ideal, but the algorithm does work to the tolerance and if the tolerance is large, it can give seemingly undesirable results. It aims to fit the largest arc it can within the tolerance. Some customers prefer this behaviour, and some point out that it sometimes fits undesirable arcs. We are currently making what we feel to be the best compromise between those two points of view.
Do you have a particular reason for wanting to use a large tolerance?
All the best,
Tessa

Hi Tessa,
File is attached in first post.
No reason to use such a large tolerance, but I do have to play with the tolerance on different models to get a good outcome, which becomes time consuming.
Thanks.
Hi parilla88,
The case where the profile is wrong for 0.001mm tolerance has been fixed in an internal development build, and will be coming out in a customer release in due course.
All the best,
Tessa
