FG in 2014: differing lengths for symmetrical parts

[graemev]

I seem to be having difficulty getting dimensionally identical, though symmetrical, parts to come up with the same BoM length in frame generator.  Specifically, the legs of the sawhorse: lefties & righties differ by 0.067" despite measuring identically as modelled.  For my current purposes I can override the values in the BoM on the drawing, but would like to think that it shouldn't be necessary to double check every FG generated part for correct lengths in every frame I create hereafter.  (I'm already getting a premonitory headache just thinking about it.)

Bug?  Bad technique?  Bad haircut?

[cbenner]

Hi,

Let me ask you a question, did you do the miter before or after the trims?

I removed the miter, verified that both legs were the same length, then redid the miter, and it all worked fine.  Generally in FG, it is best practice (IMHO) to perform trims before miters or notches.

I can't say about the haircut unless you post a picture.  🙂

[graemev]

Hmm...  Tried that, no dice...

Stand by for updating the program to the latest SP & Hotfix...

[graemev]

Latest updates (Build 246, SP2) installed & confirmed.  Still no joy.

Edited to add:  Tried removing all end treatments, Trim/Extend, then Miter.  Still the same result, though it seems to be at the Miter stage that they diverge.

[graemev]

Well... progress in diagnostics, but still no cure.

It seems the material length reported in the BoM is the length after the first, if any, angled cut at each end of teh member.  This is a good value for the shop to work with, so thanks Autodesk for that one.  The problem I'm having, it would appear, is that there seems to be a difference between the members as to whether the Trim/Extend feature comes before or after the Miter feature.  In the individual models there would appear to be no difference, neither as opened nor as edited in place.  In the BoM there is, and the difference is accounted for by changing the sequence of operation, which I appear not to be able to control.

I have now tried editing one leg of a pair with the frame generator, selecting the opposing attachment point of the profile and rotating 180°.  The edited member updates to the correct length.  Yay!  The other leg of the pair updates, of course, to the wrong length.  Boo!  It would appear I can have it any way I want it, so long as half the legs are wrong.  *sigh*