Notifications
Clear all

dimensional accuracy of printed parts  

  RSS
xenon
(@xenon)
Trusted Member
dimensional accuracy of printed parts

Hello together,

I'm trying to get used to the XL and its quirks.

I made a simple part to print on both printers and measured the distances from the first "tooth"... the dimensional accuracy is all over the place and way less predictable than on the i3:

same roll of PETG (not that it should matter), quality-focused settings for a smooth as possible outer layer (print outer layer first), exstrusion factor 1 (which underextrudes slightly but results in a cleaner print). .6 nozzle on the XL & .4 on the i3 (also should not matter).

i can squeeze the parts together in all combinations, but if i dont use the full length of the XL part they slip together effortlessly (as they should, both being slightly underextruded).

preview of the part:

 

any solutions?

heat the entire bed?

something belt tension? (it is in the OK zone on the small prusa tension gauge)

 

Posted : 17/09/2023 6:42 pm
Diem
 Diem
(@diem)
Illustrious Member

It looks like you need to apply calibration factors to the i3.

With a 0.6mm nozzle your repeatable target without special measures should be within half an extrusion, about 0.35mm so you are getting close.  Have you tried theXl with a 0.4mm nozzle?

Cheerio,

Posted : 17/09/2023 8:53 pm
xenon
(@xenon)
Trusted Member
Topic starter answered:
RE: dimensional accuracy of printed parts

the .4 nozzle x should be here tomorrow, gonna test it for sure. i dont want to make a mess in a heaterblock with the nozzle adapter...yet.

and the i3 is the precise one and hardly needs any additional tuning (extrusion factor 1.02 is all it needs for correct extrusion width) while xl is all over the place (and needs 1.08 on the .6 nozzle to achieve correct width). it is a slightly bigger nozzle but that is still terrible.

hopefully it is just a nozzle thing...

Posted : 17/09/2023 9:21 pm
Diem
 Diem
(@diem)
Illustrious Member

the i3 is the precise one and hardly needs any additional tuning

Of course, it is well bedded in by now and the XL is still new and settling down.  BUT those gains suggest you need to use small calibration factors in X and Y to make the traces flat - much of it may be thermal expansion/contraction.  In one sense the XL is more accurate as the errors tend to correct over longer distances.

Cheerio,

Posted : 17/09/2023 10:02 pm
xenon
(@xenon)
Trusted Member
Topic starter answered:
RE: dimensional accuracy of printed parts

same issue with the small nozzle, of course.

it loks like a basic movement system issue...like the belts are irregular in some places or the toolhead is being pushed by its wires.

Posted : 19/09/2023 10:50 pm
xenon
(@xenon)
Trusted Member
Topic starter answered:
RE: dimensional accuracy of printed parts

also printed one rotated 45° to maybe narrow it down to one belt/motor....the deviations are massive, a 150mm dimension is 0.7mm too small.

judging from the display during the print, the entire bed was heated evenly already...no chance for improvements there.

Posted : 19/09/2023 11:05 pm
Thejiral
(@thejiral)
Noble Member
RE: dimensional accuracy of printed parts

0.7 mm per 150 mm is a shrinkage of 0.5%. That is actually still within the shrinkage range of PETG. Maybe you get different results on other printers but it also depends on the precise conditions. What is the shrinkage if you print it turned by 45° or 90°? Does it differ?

I assume you separate the shrinkage error (ie "k") from the over/underextrusion error (ie "d")?

Mk3s MMU2s, Voron 0.1, Voron 2.4

Posted : 20/09/2023 11:48 am
xenon
(@xenon)
Trusted Member
Topic starter answered:
RE:

the .7mm undersize over 130mm length happened diagonally (bottom left to top right). the other angle (top left to bottom right) is off by "only" 0.4mm.

material shrinkage / temperature expansion isnt enough of an argument imo...my i3 y axis is running circles around this, and the deviations behave in a way you expect (linear), not like your belts have varying pitch in places.

error from underextrusion is only some 0.04mm - i didnt adjust for it because it is very little but it results in cleaner prints, and hardly affects dimensional accuracy over longer distances.

Posted : 20/09/2023 2:28 pm
MME
 MME
(@mme)
Reputable Member
RE:

Are you letting the print and bed cool all the way down before you remove the parts. I know this is a simple thing but it does make a difference In the shrinkage on long parts 

Posted : 03/10/2023 1:45 am
xenon
(@xenon)
Trusted Member
Topic starter answered:
RE: dimensional accuracy of printed parts

yeah the bed was mostly cooled down, but from my experience it hardly matters.

Posted : 03/10/2023 3:51 pm
Share: