I couldn’t find relevant thread about this issue. In particular cases, “Point in Brep” component returns true for points that are actually outside. This occurs for solid surface with flipped normals - in my case, I have to either flip surface or flip the direction of the curve used for revolution. It took me quite a time to figure it out as there’s no warning message. Maybe this could be improved somehow @DavidRutten?
PointInBrepIssue.gh (13.4 KB)
I thought that has to do with the relation of the curve and its area centroid.
But then I moved your curve to the positive side of the XY plane and made sure the curve direction is the same as some other reference curves.
PointInBrepIssue.gh (33.0 KB)
Thanks you both.
I’ve created RH-61806: Grasshopper: RevSrf creates flipped solid
Apparently, the location of the input curve has something to do with this.
Thanks for investigation It definitely is related to the direction of curve from which UV (and thus normal) direction of surface is defined. Flipping curve of flipping resulting surface helps. But there still should be some option to prevent this issue.
- Get bounding box of the shape
- Create one point that is deffinitely out of bounding box
- Check whether the point is outside brep or “fake” inside brep
- If it’s “fake inside”, invert all other booleans, else return original values