I would like to create the side of watch case with a domed center but ending flat the lugs. I don’t know which tool would give me the best outcome. Any advise would be appreciated.
Untitled.3dm (435.9 KB)
One way would be to use DupEdge to duplicate the top and bottom edges into curves. Then make a second cross-section curve that is a straight line (wherever you want the piece to turn flat). Then do sweep-2-rails from the curved cross-section to the flat cross-section. Then do sweep-2-rails from the flat cross-section to the point on the end. Then, of course, you could either delete the original piece or hide it on another layer.
watch side2.3dm (419.5 KB)
thank you, that is what I actually tried to do. However I see that the mdoel provided by the client is ending in a tip but it ramins everso slighlty domed. I attach the surface. It seems they used several patches and then trimmed the final shape. This gives a very good continuity, but i don’t know where to start
production model.3dm (172.7 KB)
It very much depends on which curves/sections you must adhere to, and which ones you are free to change. For starters, you can see that the upper/lower longitudinal boundary is made from various incredibly dense curves. Why that is so, we can’t know. Also you need to decide at which longitudinal point your surface(s) shall be linear vertically.
If you can rebuild the upper and lower longitudinal boundaries as simple degree 5 single span curves, and the crown and line sections as degree 2 single span curves, you can then use Sweep2, and then extend that end bit and trim off what’s not needed. See a quickly futzed version below.
Now I have to build what is called a lyre lug (the surface going inwards). However I can’t matcht the edges between the upper surface and the lyre lug. I feel the furface of the side of the case and the top surface are okish modelled. However I have not much control over the lyre surface. How should I approach this?
Lyre Lugs.3dm (141.7 KB)
The main surface from the bezel into the lug shown here seems overly complex. It has a high number of isocurves, which is not the best starting point to build the secondary surfaces from it or towards it. In NURBS surface modelling, you first make sure you have simple higher degree single span curves with good curvature, to then be able to build simple single span surfaces, which, in turn, makes the final G1 or G2 filleting more straightforward. Also, I don’t know what your design intention is, meaning where you must have a G0 surface transition, where it should be G1 or G2, or where the curves are allowed to be wobbly as they are now, see below.
In general, you mentioned above that someone will re-do the entire model in SolidEdge afterwards anyway, so why bother?
Important if an exactly flat portion is needed: A single span curve will never have one part which is exactly straight and another straight which is not flat. A single span surface will never have one part which is exactly flat and another part which is not flat. That is how the math of NURBS curves and surfaces works.
If “close” to flat is sufficient then it may be possible to get “close enough” using single span surfaces. In some situations, automotive exterior surfaces for example, exactly flat or straight may usually not be desirable.
Yes, all of one of the single span surfaces is flat. None of the other single span surface is flat.
.
Why are you trying to match edges? Aren’t you going to eventually want to round those edges?
If you are planning to not leave these edges as sharp corners then those edges won’t even be on the final model. Everything you are doing to get the edges to match will just make it more difficult to round those corners later.
It’s common for edges of watch cases to be left “sharp”.
Hello Lucas,
have no time to make this exercise but if you like to deform a surface and get nice look best think is to use SubD. It is easy to operate and is giving nice result.