2D Chamfer Oddity

[Steinwerks]

Had this happen to a deburring op after changing the tool from a 2-flute 1/4 sharp point chamfer mill to a 4-flute .038" tipped off chamfer mill. No other changes in toolpath definition, offset of .08". Flute length of .11" should be plenty, and it only happens on one area of the model. Previous tool has no problem with a normal toolpath using the same geometry selection and offsets.

[jeff.walters]

This looks like it can be caused by a couple of things, but most resembles what can happen some times when you try to comp from a spline. Can you either post your file or send it to cam.support? One thing that has helped in the past is smoothing.

[Steinwerks]

I can't send it unfortunately. I have seen the 2D Contour issue using splines from other people, but this is the first I've seen it, and like I said, it generates properly with the exact same parameters but with a slightly differently defined tool. Smoothing didn't affect anything.

I can say that the feature was created with a simple circle in a sketch using a Coradial relation, and from that a blind Extruded Cut.

[Laurens-3DTechDraw]

I think we have the issue of a designed chamfer and use of 2D Chamfer operation here.

Due to the way the operation looks at crashes of the tool with the model you can get such a weird path. If you change the angle of the tool for example it will probably go away. Changing the tolerance might do the same.

[lenny_1962]

I've had this happen quite a bit, try a 2D contour and use your chamfer tool instead of 2D chamfer.

don't know why but it usually solves the problem.

i had submitted this along time ago

[Laurens-3DTechDraw]

---

@lenny_1962 wrote:

I've had this happen quite a bit, try a 2D contour and use your chamfer tool instead of 2D chamfer.

don't know why but it usually solves the problem.

i had submitted this along time ago

---

Because 2D Contour just chamfers/follows the selected contour. Where 2D Chamfer looks at the 3D Model and tries to not collide with it.

[Rob.Lockwood]

i'm willing to bet an incredibly small sum that the 2d chamfer tool creates an offset from the contour where collisions are allowed, and trims everything where collision is detected outside of that.. Hence the issues shown here, sometimes the toolpath is within, sometimes it's not.. so you get sections of toolpath.

@al.whatmough did I reverse engineer it correctly? i've been stumbling over how this one works for a bit..

[lenny_1962]

this needs to get fixed, it's a pain when you have to cheat with a 2D contour using the edge or having to make a sketch so you don't hit then part with the cutter.

[al.whatmough]

@Rob.Lockwood  I can't give all the secrets away 🙂

But, yes you are on to something.

I will have rene take a look

[al.whatmough]

@Steinwerks if you can't share this file can you reporduce this on another part.

[Steinwerks]

@al.whatmough

I'll see if I can tomorrow. If not I can at least save the path with tool definition as a Template and attach that.

[Steinwerks]

Well that was easy. Must have something to do with the tool definition:

HSMWorks 2016 R3.40936

SolidWorks 2016 x64 SP3.0

[Steinwerks]

@al.whatmough @jeff.walters @dave.anderson

Anyone take a look at this yet? Just want to make sure it didn't get dropped. I can send it via feedback too if that's better.

[al.whatmough]

testing now.  sorry, I missed this yesterday.

[jeff.walters · Accepted solution]

I reported this to the developers on ticket CAM-4370.

I found that if you eight click on the operation and select Compare an Edit, then change the tolerance to .002 you will get an arc instead of the linearized moves.

[Steinwerks]

Thanks Jeff!

I always forget about Compare and Edit as I never use it, and didn't realize that Tolerance setting was in there.

[dave.anderson]

I did notice that modifying the 3.944 value to 4.00 then the issue goes away.

[Steinwerks]

---

@dave.anderson wrote:

I did notice that modifying the 3.944 value to 4.00 then the issue goes away.

---

Wow that's a weird one.

Now that I'm back at my desk, playing with the tolerance yields different results. The error goes away at .0016" tolerance, and the tighter the tolerance, the more jumps it develops. If I manually set the triangulation tolerance it gets real weird.

[Steinwerks]

It's popped up again, and does not seem to be related to the geometry size. Can't get it to go away at all anymore, nor reproduce it in a dummy part. I will send feedback though.

Tolerance at default .0004":

 

Tolerance set to .0016":

I'll note that this is in an assembly, but do not know if that has anything to do with the error.

@al.whatmough @dave.anderson

[keith.clausen]

Out of curiosity, is the model native to the CAD system or imported from another CAD system? As an example models imported from ProE can typically be plagued with anomalies.