Dimensions of objects change without user interaction

Dimensions of objects change without user interaction

m.zentveld
Observer Observer
382 Views
5 Replies
Message 1 of 6

Dimensions of objects change without user interaction

m.zentveld
Observer
Observer

Hi, wondering if anyone has come across this and has a suggestion. I'm finding that after working on a design for a while I'll use the inspect tool to check a distance and notice that the object sizes have change by a small fraction. Even on simple parts. For example I have a cube that is 106mm in each direction after working on other parts of the design I come back and re-check the dimension and it's now 106.0000000000000142 mm.

I can't get it back to where it was either, rolling back the history or re-editing the original object doesn't put it back to 106mm. This happens on new designs as well as existing ones and I've tried reinstalling Fusion 360.

 

It seems to be related to saving the design, it seems to be fine if I never save the design (and turn off autosave), however that's not an overly practical approach. Having those little fractions completely messes with joins.

 

Thanks.

0 Likes
Accepted solutions (1)
383 Views
5 Replies
Replies (5)
Message 2 of 6

davebYYPCU
Consultant
Consultant

Change your display settings for number precision, to 3 decimal places, 

 

Might help....

0 Likes
Message 3 of 6

etfrench
Mentor
Mentor

How does it affect a join?

I have the decimal precision set to one more place than I can measure.  For Imperial it's 5.

ETFrench

EESignature

0 Likes
Message 4 of 6

m.zentveld
Observer
Observer

The problem is that once you get into complex join patterns across lots of parts you get join errors as the faces don't perfectly align, that .0000000x of a millimeter is enough to cause the join compute to fail.

0 Likes
Message 5 of 6

jeff_strater
Community Manager
Community Manager
Accepted solution

that should never be the case.  All geometric modeling programs work within a tolerance.  The modeling tolerance for Fusion is 1.0E-09, and the assembly/joint solve tolerance, I think, is 1.0E-06.  A variance in the range of 1.0E-16, as in your case, should never cause a failure.


Jeff Strater
Engineering Director
0 Likes
Message 6 of 6

m.zentveld
Observer
Observer

Thanks Jeff, it turns out the issue I was having was unrelated to those fractions of a decimal in the dimensions. I've run through an updated model and the joins compute successfully (i.e. within the margin of error).

So it simply appears to be a cosmetic thing with those dimensions and as pointed out by other users I can simply adjust the number of decimals that are displayed.

 

Appreciate the responses.

0 Likes