when 3D modeling or 3D sculpting the body of any product, designers often need to build from trimmed surface edges to make srf blends. This may cause the model to become very complex. Such complexity may cause the srf analysis tools to project odd results.
Is there any method out there to reduce the complexity of trimmed srf edges for better looking surfaces?
Oliver
Is there any tool for surface smoothening?
Can we see a little more specific example of the specific issue youâre talking about? There are about a million options, but when results REALLY matter, you have to brute-force it, Apple doesnât make iPhones with âtoolsâ, itâs just days and days of point-pushing until itâs perfect.
Apple develops IPhones in Alias using G3 continuity. You are telling me that this kind of high-end products are modeled by cv (manual) manipulation???
Let me upload an exampleâŚ
sample.3dm (149.6 KB)
Umm, yes of course. Look at their job listings for âdigital sculptors,â thatâs exactly what they say.
wow interesting!!!
So what would be your concern about this edge? If you put a large fillet on it, it looks fairly clean for an intersected edge. If you want more control than that, thenâŚwell often what Iâll do is ReBuild the blend or the troublesome section of it, tweak it by hand, and ânail it downâ after with MatchSrf. Iâd be more concerned about the tolerance settings on that file, they seem pretty loose.
they say that tight tolerances is better for high end product design but why? looser tolerances gives you greater chances of surface matchingâŚ
I might consider it gold-plated unnecessary overdoing it, but if you search all the posts over the years asking how to do classic âiPhone corners,â you can see none of them really match it without some manual point-tweaking.
okok⌠goodâŚ
is it possible to srf match manually without running any commands or tools??
Not surfaces, not practically⌠I mean you can get close but the principle is about not just counting on being able to make a sweep or something and poof like magic itâs perfect.
Looser tolerances mean that the criteria for what is acceptable are looser, not that the surface is better. Looser tolerance can mean the surface if manufactured will be worse.
An important fact about modeling with NURBS. The intersection of two NURBS surfaces usually (there are exceptions) can not be exactly represented by a NURBS curve or other simple explicit mathematical equation.
For situations where the intersection is not exactly a NURBS cuve there is a solution. The intersection can be approximated by a NURBS curve within any desired non-zero tolerance, ie the difference between the approximate NURBS curve and the intersection can be made as small as desired but never exactly zero everywhere. The tradeoff with making the difference smaller is more control points/spans for the NURBS curve. This is one reason why tighter tolerances can result in more control points.
Any software which uses NURBS has the same situation.
Hi Oliver - Duplicating an edge and rebuilding the resulting curve is a legitimate workflow, but how close that new curve and surface come to the trim is dependent on multiple factors. Getting it spot on, within tolerance, not matter a great deal if there is to be a transition surface between the two in the end.
-Pascal
okay⌠I get the idea Thanks
@pascal, what other multiple factors are dependent?
many thanks
Select the intersection curve and apply the the âRemove multiple knots from surface or curveâ (! _RemoveMultiKnot) command to it. That will reduce its control points by about 50%. You will notice that in some areas the modified curve is a bit off from its original shape. Donât worry, that will be fixed in the next steps. The original curve had 51 control points before, now it has 24 control points.
Extrude the modified curve diagonally (not vertical, not horizontal - something in-between).
Use âMatch surfaceâ to match the new surfaceâs edge to one of the existing surface edges. Choose the âPositionâ and âRefine matchâ options.
Extract the surface edge and delete the extruded surface. The extracted curve is withing the tolerances that you set in the âMatch surfaceâ window during step #3. If you use the default setting â0.01 unitsâ, then the curve will have 26 control points.
Wish for Rhino 7: Make âMatch curveâ tool that works similarly to âMatch surfaceâ but using curves. 
@Rhino_Bulgaria, nice explanation as always! but thereâs something Iâm not quite sure about⌠the refine match? what is it?
regards
âRefine matchâ is an option inside the window of âMatch surfaceâ and its purpose is to add extra rows of control points to the surface so that it could closely match the target edge or curve.