temptest.rar(1.7 MB)
Bad surface will happen, when make fillets to the four sides marked out in the attachement with fuction “CRhinoFilletEdge::FilletEdges”, fillet radius as 0.3
Q1: Generally, what causes a bad surface? And what is the special reason in this case
Q2: is there any solution to optimize the original surface, to avoid broken face when filleting?


please help me to see the problem.
thanks very much


I find another situlation.

if I filletedges not just these four edges, for example: add another two edges,the filleted surface will not a bad surface.

(Dale Fugier) #4

@chuck, when you get a chance can you look at this?

(Chuck Welsh) #5

I don’t see any marked edges, so I do not know which ones you are trying to fillet. Once I know that, I will be better able to help you.

There are a few problems with this model. One can be seen by calling the What command. There is at least one edge that has a tolerance of 0.07, which is 70 times as bad as the file tolerance of 0.001. Also, fillets do not work well across edges where the surfaces are nearly tangent but not quite. The detail on top looks like it was made of many small segments that are not quite tangent. It would be better to make it out of a few smooth curves.

Chuck Welsh
Robert McNeel and Associates


hello,chunck,thanks very much
the 1.bmp file have four marked edges.


another question:
this model was built using sdk function(loft ,sweep2, patch and so on sdk function) quite.I know it is not very smoth,but I don’t how to make my model become smoth using sdk function automatic,not using command manual work,using this function before,I have using rebuild funtion to rebuild my curves


this tolerance is produced because I need use patch to creat one of the face.and patch alway have a large tolerance.
I don’t know how to reduce this tolerance

(Chuck Welsh) #9

Sorry, I missed the bmp file. I ran the fillets in sr5 and sr6. Sr5 made a mess. Sr6 worked correctly. Which are you running?



I ran the fillets in sr5

(Chuck Welsh) #11

Great. I did fix some fillet bugs for sr6. This must have been another example of one of them. Please try sr6 and let us know if it still doesn’t work for you.

Chuck Welsh
Robert McNeel and Associates


ok,thanks very much


another question:
as you see,my brep’s edges has a larger tolerance,because some of my surface is built with patch ,so how can I reduce the tolerance ,the best reslut is the brep’s edges can match the curves using to pathch.

CRhinoFitPatch question

when I update my rhino to sr6,if I also need update the sdk?
I just find the latest sdk as follow:
Rhino Sdk 5.0 64-bit (VS2010) 2013-05-24 (95.6 MB)


Does RhinoV5 support G2.G3 type fillets?


I update my rhino to sr6 and sdk to Rhino Sdk 5.0 64-bit (VS2010) 2013-05-24
I find the when I using filletedge command to fillet these four edges,it work fine.
but when I using sdk function to fillet these four edges still will produce bad surface.

(Dale Fugier) #17

Yes to G2, that is just BlendEdge. No to G3.

– Dale

(Dale Fugier) #18

Without your source code and your model, we cannot help you with this issue. Can you provide working source code?


I was not thnking on bend edge command, but rather use command fillet and choose what type of radius do you want…circular, G2,G3…
This is still not possible in Rhino?


filletEdges.rar(1.7 MB)
Dear Dale,
thanks a lot for your kindly reply, kindly have a double check of my questions bellowing:
filletedges.rar contains four bmp, one pattern and one Source code
BMP1: showing the original four edges of the target fillet
BMP2: if use the command filletedge to lead the fillet, will have overlap face
BMP3: if use code sdk, fillet will have obvious break.
If add an extra edge, as BMP4 shows, then the command filletedge could make a good one, while code sdk leads to overlap face
There are 3 brep in the pattern file. From left to right, is the brep of my target fillet, brep resulted from command filletedge, and brep resulted from code sdk