Does face deletion break parametric history?

[oprcmr00]

Hello. I don't understand why this doesn't work.

1) I created a 3D sketch with a closed 3D profile

2) Patched it

4) Thickened it

5) Deleted the outside face (created by "Thicken") to re-patch it with a center rail, in order to create a bulge in the thickness of the model → finished with other operations.

6) Went back to the original sketch and moved (as an experiment) an unimportant CV from one of the original splines by an unimportant amount: 1mm. → the patch broke.

I come from 4 years of direct modeling and always avoided parametric because it was a stick in the wheel the entire time to say the least. For the more parametric-like things I preserved sketches and re-lofted as needed. Deleting surfaces is very important because in that way I can re-surface, varying the shape of objects, or insert foreign component into surfaces harmoniously by cutting the surface with a curve, then deleting the middle face, then lofting, filleting, etc. 

This is not working, I have to do that operation. Please give an (or any) advice...

Pictures (in book-style order): the final object; editing the first sketch; the broken object; the related warning.

 

 

 

 

 

 

 

[kacper.suchomski]

Hi

Complex curves projected onto another surface create curves that are not perfectly tangent, which in turn creates sets of edges on the model.

Editing the sketch edits the set of generated edges.

The new edges are not included in the operation because they were not specified.

Yes - you are working on complex geometry, and editing will generally require you to repair the geometry.

[TheCADWhisperer]

@oprcmr00 

Sketch2 should be a 2D sketch independent of the prior history surface geometry.

[oprcmr00]

Thank you. Still there are mysterious things happening here that I don't understand (and therefore can't fix). Can you check this? ↓

Incredibly, after moving that "most unimportant" CV by that "most unimportant amount" (1 mm), after re-patching (successful), re-thickening (successful), at the moment of the deletion of the external face, THE INTERNAL FACE IS DELETED INSTEAD. Why? Subsequently to that everything downstream fails.

[oprcmr00]

Thanks for exposing your thought process. I'm well in the direction of ditching parametric. My idea was to create an anatomically-conforming structure with surfacing tools. It is then to be worked on by adding attachments, bolt holes / slots, sliding parts, etc. - e.g. imagine an "helmet with tools" to render the idea. I was thinking that parametric would save me some time as I could be able to go back and auto-re-surface by draggin CVs. It doesn't seem to be the case. 

[TheCADWhisperer]

@oprcmr00 

I do all my design work as parametric.

Worked the first 14 years as non-parametric and never looked back.

All 3D sketches are the result of Intersections from 2D sketches...

[TrippyLighting]

---

@oprcmr00 wrote:

... It doesn't seem to be the case. 

---

Can you demonstrate the behavior in a screencast?

This geometry is simple enough that things shouldn't break.
.

[oprcmr00]

Not equipped for screen-cast at the moment. But you can see that the sequence of the first features is already mind-blowing. I have a fully closed 3D sketch, patched, mirrored, thickened, then deleted a face. Then as an experiment a random CV belonging to the first sketch was moved of 1mm. Result: the wrong face is deleted. 

↓

↓

 

 

 

 

 

 

 

 

 

 

[TrippyLighting · Accepted solution]

When you mirror the 1st patch, the two surfaces are not tangent, and that is ultimately the problem. The issue here isn't the parametric nature, but your modeling technique.

I would create the central sketch as a 2D spline on it's own plane. Then you don't need a sweep, you can use a simple surface extrude.