I have few questions how the edge tolerance is decided for the feature which is being obtained after series of operations.
For instance, I have a sketch, which is used for trim sheet (tolerance is set to 0.0005 in trim sheet UI). after performing it, I'm thickening the sheet (tolerance is set to 0.0001 in trim sheet UI).
I was hoping the tolerance of the resultant feature would be 0.0006 (Cumulative tolerance), instead the information window had shown the result as 0.00055.
Could anyone explain the methodolody behind this calculation. (When series of operations are used)
thanks in advance
I think that depends on the type of operation, developers could explain that.
For example, when performing Sew the tolerance of the common edge (1st selected) is the radius of the
"tube" touching the adjacent edge (2nd selected). That was beautifully explained by J.Baker some
years ago here (I should also have an explaining picture that I made for my users though).
Why were you hoping the tolerance was 0.0006 instead of 0.00055 ? Smaller is better.
I'm happy someone ask this question though, which is rarely raised, I wish users over
here asked themselves such a question more often as this is a very basic issue in
freeform modeling, the handling of tolerances. Should edge tolerances go out
of control during modeling then the model is ruined almost beyond repair.
I learned that the hard way when I made my first FF model, as this key topic
was not covered by the freeform class I attended.
Hope someone from development answer your question, I'm also interested in that.
I agree with @Maurizio on how this subject (tolerance) is very important, as well as I agree that it is a little discussed subject.
In NX the edge tolerance is used in many commands. By default I use 0.001 mm (0.000039 inch) in modeling preferences but I change this value in some features such as Offset in Face curve (increasin it).
This means that the resulting curve will deviate from the original edge or curve of a higher value, but the feature I build on that curve, such as a through curve, will still have a tolerance of 0.001 mm
I would like to discuss this topic with other users.
well, this has not always been the case.
If you made such Through Curve feature with an old NX version (NX4 for example)
with Arclength alignment you got edges tolerance twice as much as the modeling
tolerance you set in preferences.
When I noticed that and asked for the reason an explanation from development
arrived telling that this setup was a good compromise and that I shouldn't care too
much about this.
Such example was useful for me to scare users about how little it took to lose their safety
margin before eges become too loose, and warn them about taking care of this
(perform Examine Geometry every 10 minutes or so during modeling work, to
say one). The safety margin is said in NX documentation to be 4-5 times as much
as the general modeling tolerance : you made just *one* feature and you lost
half of this at once.
Now NX behaviour has changed about this, as the underlying working of Mesh commands
has been rehauled several times over the years, as we users noticed, but the point remains
of taking care of edge tolerances so that the natural increase of tolerances doesn't go out