Hello, I am finding it very difficult to smooth the curve shown.
The middle section was added as a random curve to join the two existing ones and now needs editing to create one smooth curve.
The main problem seems to be that dragging the vertex points causes very severe, almost uncontrollable changes to the shape.
I cant arrange the curvature comb densities for each part of the curve to be the same, nor can I get the polygons with the vertex points to look anything like in the help files. On the problem curve they lie almost on the curve itself.
I suspect this is all related and exaggerating the sensitivity of editing the centre curve.
Am I missing something with the set up which would make it easier as it feels impossible at the moment?
Solved! Go to Solution.
sorry, but I cannot get Your request.
As shown in the picture, Your curves already are connected tangent AND curveature consistent.
What more on smoothing do You want?
The curvature of Your curves is changed from curvature left to curvature right in a linear way.
Will this not be enough in smoothness?
to explain what i said before this picture should help.
It shows similar curves but "only" connected tanget and not curvature consistent.
The image shows that the curvature on curve left will not be the same as on curve middle.
But nevertheless the are tangent, so there is no buckling between them.
Both have the same direction, but not the same curvature
In Your example the curves have both equaly, direction and curvature
SE is not a "class A" surfacing software where the reflectance of a surface can be controlled. For that you want Imageware (add on to NX), or at the low end Rhino.
In general you need a Nurbs oriented software to have details curvature controls.
to this IMHO I can not agree totally.
Class A - in my very past years (1980) were defined that You will be able to controll the second derivation of a curve, which is done when I have controll to the curvature.
With SE I will be able to talk about the curvature of a curve and of a surface.
There are also tools within inspection to show the behaviour of the surface to surface connections using zebra stripe shading and curvature analysis.
I will agree that a more expensive software will have more and advanced functionality for dealing with tangency, curvature, etc. but Yoou also have some functionality here within SE.
Due to the fact that SE is not intended to be a surface CAD system for automotive, aerospace, etc.
the range of this functions is not as great as it will be within NX, CATIA or if I remember ICEM SURF, the grandparent system of all class A modelers.
But depending on what You have to create and to design SE is not as bad as it is said to.
Thank you for the replies.
It seems I have not properly understood what the curvature combs are showing.
I was under the impression that for the best surface smoothness, the curvature combs themselves needed to be smooth as well. The examples in SE seem to be.
In my example I was worried by the 2 sudden changes in the line joining the tops of the combs.
It did puzzle me why these “kinks” were there, when as you said I had specified tangency and equal curvature.
As I say, I think I need to study curvature combs in more detail. I do not find the help files in SE very helpful in this respect.
Do you know of any examples or documents that explain how to use these combs properly?
Thanks again, SPCAD.
Hello again Wolfgang,
Looking at the attached document, it seems in my example I have G2 curvature continuity between the two curves.
As I said, I thought I needed to get the combs smooth as in the G3 curvature continuity example.
However, the text accompanying the examples of G2 and G3 are very similar and does not really explain the differences or benefits.
I think this document is an AutoCAD help file. There is more information about SolidWorks and AutoCAD than SE. I suppose the theory is generic, but it would be nice to find some better help specifically for SE.
I'm not quite sure wether there is an explanantion within Solid Edge documentation or not.
But curvature contour IMHO is easy to interpret.
Try to imagine a simple edge wit a constant radius rounded.
The falte plane on one side has a curvature of zero, it is "plane"
The constant radius has a constant curvature which is the radius
So the curvature jumps from 0 to radius and back to zero.
The curvature contour therefore is not continuos.
If You change that connection for curvature continuos, then You can see, that the curvature curve changes from the high radius to the zero flat in an small area but not instantly.
This IMHO is what G2 means!
I never ever have heard that somebody had a G3 request for modelling.
Question here als would be,. "Whatfore?"
Good example, its quite a nice way of explaining what the comb shows. In future I need to think a bit more about what the comb is actually showing, clearly, it does not have to be smooth in itself.
Therefore I agree that what I had originally should be perfectly fine. I am sure G2 is more than good enough in this application.
Thanks for the help.