Chamfer edge inconsistent

#1

I have always had a strange issue when chamfering edges in rhino. In particular for chamfered holes.

I have tried the various options under edge type, however the inconsistencies remain.

here a screen recording showing the variations.

is this some sort of tolerance preset that I can change?

#2

Hi Andrew,

I do not see that issue here in a test.
Could you upload the file that is showing this behaviour?

-Willem

#3

It looks like SmartTrack is on so your cursor tried to “prefer” orthographic directions.

The ChamferEdge tool seems to be working as expected.

Please describe the problem you’re trying to solve.

#4

Hi Andrew,

Looking at the video again it seems this part is very far from world origin
That could be the cause of rounding errors like these.
What if you test this with the same object close to the world origin?

_Willem

#5

Hi, thanks.

Please see attached a file that I just made for reference, where again I have the same issue.

I have made a 10mm hole inside a ‘plate’… I proceeded to chamfer the edge of the hole, and when I dimension the outer edge of the chamfer, I have inconsistent dimensions… somehow it looks like it is finding or referencing different centre points when dimensioning.

here also a screen grab…

52
Screenshot 2019-10-25 at 14.52.52.png2310×1806 1.01 MB

rhino _ chamfer edge inconsistencies.3dm (5.1 MB)

#6

Hi Willem,

I just made and posted a new part where again I have the same issue where I have places the part at 0,0,0 coordinates.

however, would it not be strange that this happened even if very far away from 0 world coordinates?

#7

I cannot reproduce the bad result here.

this could be bug specific to your build.
You can change the category to Rhino For Mac to find some testers over there.

image
image.png957×545 97.8 KB

_Willem

#8

Hi Willem,

Thats so strange… I had assumed that this was an inherent issue in rhino… good to know it is not… but now to find how to resolve…

Could you please explain a bit more what you mean by

I am currently running on V5.5.4 on Mac

#9

One more note on this, I just tested on a colleagues computer, runnings rhino 5.5.4, whom also produced the same bad results.

#10

The ChamferEdge command (like many others), follows the edge using a surface fitting mechanism that says within the model tolerance. This explains what you’re seeing as a small change in the resulting radius. A tighter tolerance will result is a more complicated edge curve, but will reduce the variation in the dimension.

If have a clean circle is the goal, then don’t use ChamferEdge.
Draw circles and created surfaces using them with Loft instead.

#11

Hello - this is something that just got better in V6 compared to V5. V5 Windows also produces an approximate circle, V6 a true circle at the edge.

image
image.png882×292 45.9 KB

I have not tested V6 mac but I assume it is the same as Windows.

-Pascal

#12

Here’s a 5.5.4 Mac file with two examples.
The one on the left uses ChamferEdge.
I used circles and Loft for the one on the right.

Pascal is right, V6 tightens this up by using better intersector code.
CE-vs-Loft.3dm (3.8 MB)

#13

Hello,

OK, Yes, I just tried ChamferEdge in V6, which produced true circle as you say…

Unfortunately though, I still need to use V5 as the print to PDF in V6 produces distorted crvs.

here a post on that:

I believe @stevebaer is looking into the print to PDF setting at the moment, so hopefully I can migrate back over to that soon.

#14

HI John,

Thanks… I have in the past created a loft / revolve to produce a clean, true output, whoever this does use a lot of extra processes / commands to produce…

I have just tried in V6, as @pascal mentioned, which works well.

#15

For V6, the developers tried to add “special case” examples (like using circles), instead of using the surface intersection fitting that works for irregular shapes.