Dimensional Resolution

[Rdrost 0]

Hello, I am wondering what the dimensional resolution is (I have a ultimaker 2+ extended). For example, on a drawing of a specific part - what would I say the tolerances are for a certain item? I know it depends on the orientation it is printed in and that the z axis resolution (the height) will be +- the layer height. I would also assume it depends on the nozzle size as well?

[Nicolinux 288]

Hi,

I don't know if it depends on the nozzle too. If you print wit different nozzles, but at the same layer height (within the boundaries of what's possible), I would guess there should not be any differences.

Check out the latest blog post. The last paragraph talks about tolerances for holes:

https://ultimaker.com/en/blog/20113-3-design-tips-for-fdm-3d-printing

[yellowshark 153]

I normally see an accuracy to within 50 microns; quite often I will see better, 40,30,20 microns, but I guarantee 50 microns. Occasionally it is worse but I have not figured that one out. I have a suspicion that the z-axis might be affected by the nozzle to bed distance on 1st layer but that is only a suspicion, never tested it.

I also have a suspicion that the whole thing can be affected by your setup; not the Cura settings but the environment such as internal (cabinet)and external(cabinet) temps., probably humidity. Also by the PLA used.

Of course the whole thing is affected by shrinkage and also by geometry(I do believe) ( and I am not referring to circles here) and yes as already said circles are an issue (which can be overcome).

I am testing at the moment that if a piece has an unacceptable error then by adjusting the model dimension(s) to get it right, it should then be repeatable; i.e. if I adjust a dimension to correct an error then the new “correct” result should be repeatable every time I print that piece.

I am also testing the effect of nozzle dimension. I did a test very recently on this, which suggested that setting the Cura nozzle size to .35 gave better dimensional accuracy that .37,.4,.43,.46. BUT I need to more testing to confirm whether that is correct or just an aberration.

And finally do not forget when you are working at tens of microns the absolute accuracy of your measurement can be highly questionable! I normally get my son, a trained mechanical engineer, to do the measuring.

[yellowshark 153]

....oh, and whilst I have only done limited comparison , results so far suggest that injection moulding is more accurate than FDM printing