Solved! Go to Solution.
Trim Sheet I added, using Curve On Surfaces and Through Curves.
During Single Curve selection I use the Filter STOP AT INTERSECTION.
Hi @Cesare Thanks!
Since the Through Curves surface is tangent to the two upright planes, the two curves on the surface, either with G1 constraint or not, should be always tangent to the planes. Additional boundary constraints seem to be unnecessary, as in @MATHI_CAX_TRG's solution, in which the two curves on surface are replaced with two trimmed edges.
What exactly is your intent of creating a surface on top of a surface that has the exact same continuity? The Through Curve Mesh is redundant, isn't it?
Most methods would stop at using the Curves On Surface as boundaries for Trimming the Through Curves surface, not to create another surface on top of another.
I didn't have to add any additional steps to the Curves On Surface to get the TCM to work. I set the Surface Constraint Direction to Isoparametric then I picked the first COS point on one planar surface and assigned G1, picked the second point on the TC surface then the last point on the 2nd planar surface and assigned G1. When I adjusted the start and ends of the COS, I made sure to drag them towards the planar surfaces such that they were lying on the edge of the TC.
Hi @TimF I just see your video. Thanks!
In my modeling,
I found that the failed surfacing with TCM is due to Curve A and reproduce the issue in a simplified model.
Your solution adds additonal contraints to and modifies my two curves on surface. In @MATHI_CAX_TRG's solution, my two curves on surface are simply replaced with two trimmed edges and then TCM works. The edges are in fact shapes copies of my curves. Therefore my curves are kept intact. I don't know why that works.
Attached is a part file by @MATHI_CAX_TRG's solution.
I couldn't use your original curves because I'm using 11.0.1 and get a warning about features being in the future when I edit them. To edit them, I have to move the endpoints and I knew you didn't want that either. The method was what was important....additional features aren't necessary to get the same result you were seeking - that was the point.
Again, the difference between just trimming the TC surface and creating a new TCM is zero in most cases.
Also, the INPUT curves for the TCM MUST HAVE G1 built into them and your COS's DID NOT - that's why the ECM fails.