I’m working on a reverse engineering project to recreate a plastic piece. I was provided with a 3d scan od the object, but it can’t be used for CNC due to it’s quite crazy hidden geometry.
Also the scan contains some defects in painting that were visible on the physical element. So I recreated the model using a lot of sections and a loft (getting rid of the defects in the process, yay! )
The final part I need to do is to make it into a solid.
I created a nice round seam between the top and the walls using a Blend surface command (piece by piece i sliced a copy of the top and botom into smaller sections and then merged the created seam, wouldn’t wok other way), so I thougth that it would make a perfect connection but when I try to Merge the surfaces… I get the “edges too far apart” error.
When I set the tolerance to 100 it worked but crated a twist. Like It didn’t want to merge the surfaces over the smallest possible distance.
I’m stuck over this for a few days already. Blease help me solve this problem.
I have a feeling that it’s some basic mistake that I’m maiking in the process or I made during creating the geometry.
I deleted the surface you had and did the BlendSrf and seems to join ok.
Used Tangency continuity not curvature.
Not sure if the surface is ok for your needs though…
Hi Toto - why merge? Join will work if you MatchSrf from the blend to the lower surface with Refine set, but I’m not sure why you want to MergeSrf here.
So Toshiyaki, thanks for the fille, but something funny happened on the “coners”.
Did you have the same on your PC? Do you think it happened only in 4.0 version? Maybe the algorythm was improved in rhino 5.0. but that’s just a guess.
Pascal so i think you fixed my problem
I’m honeslty not very experienced in rhino (3Ds Max was my thing) and untill now I used to make only visual art for architecture an civil engineering and now I’m starting to work on some models for CNC miling and that’s a whole different story.
To answer your question. I’m taking great care in making the model very “airtight” to get all the edges to match perfectly and so on. I just figured that merging the surfaces would be the best option.
MatchSrf + join worked perfectly. I didn’t try the Refine setting before - I think I need to educate myself on the surface tools a lot more.
So for now I’m good! Thank you guys!
As a form of payment plese let me entertain you:
This is my dog in a hoodie
I’m having trouble understanding the bounary arrows.
Can any of you guys point me in a direction where can I learn about them?
And how are they called?
I’m pretty sure that the fact that they are pointing in the oposite direction at the corners is the reason why things go crazy right there.
At the edges where the edges are split there are kinks, so no a smooth curve.
I’m trying to find a way to remove that, but haven’t suceeded yet… @pascal help!
That’s correct. Your top surface is degenerate because it has U and V going in the same direction instead of crossing directions. The most robust NURBS surfaces have the UV isocurves crossing at 90 degrees.
You have fairly simple shapes here so you should be able to make simple surfaces. And don;t be afraid to use trimmed surfaces.
In the enclosed file I made the top surface as a loft and then used filletsrf to make the sharp transition that connects the bottom and top surfaces. . model reversex.3dm (203.1 KB)
If you start with a simple surface you can use point editing to make the surface conform better to the scan data, You can use history and point edit the input curves or directly point edit the surface. Knots can be added where you need better control of detail.
