Hi,
I think, it's a real unsolvable problem.
The nature of the face is to be a face with a surface and a perimeter.
Here, if you fillet with 2.5 the face is becoming null (a single point which is another nature of object), so it's like willing divide by 0.
the only possibilty to be near with filleting, I think, is cheating with 2.499.
You will have on UI a 2.5 written but in real a microface is done so you don't have a perfect dome like you want.
The problem, it seems, is you want a modification on the nature of the top face when the top is a single point.
So you can design it but you have to model it without fillet and model with a single top point, so the nature of the top is a point and not a face.
You can use :
- a rail revolve to do it with a square as rail curve with a master curve beginning from the "center top" you want.
- a network job with edges you want. Edges made with a master curve and a array/circular transform (to have your 4 edges of the dome)
- perhaps another exists...
If the program search during a filleting a aberration on a single point face to fix it, the problem is the same with a face becoming a single line (case of filleting rectangle 5x10) or a face becoming a curve on others cases.
Filleting is a very cool tool, but it has some limits.
If we want go beyond the limits, we have to search others solutions.
Have fun !
|
