Accuracy of the Core One L
Hello, everyone. I have another question.
How accurate should the Prusa Core One L be? I printed a cali -flower square with 250+250 mm, which is 249.7 mm to 249.8 mm. Also, it's not perfectly at 90°.
I have a diagonal difference of about 2 mm... I bought this machine instead of a Bambu because I wanted maximum accuracy. Have I been wrong?
I mean, I love this machine. Am I expecting too much, or is this something I can tune?
Greetings
RE: Accuracy of the Core One L
This does not measure the mechanical accuracy of the printer but how much your plastic is shrinking when it cools. When you used the Califlower website and made the 10 sets of measurement, what did it say the skew angle was?
RE: Accuracy of the Core One L
Hi, and thanks for the answer. I wrote it wrong. The diagonal difference is about 0...!!
I think it's all just the shrinking of the filament. I crossposted this on Reddit, and all people said that everything is perfect. In case I need 100% accuracy, I'm going to adjust the shrinking in the filament settings. I've been printing the overhang bench from Printables, which has been quite accurate (100 mm has been 100.01 mm).
So everything is fine, and I know that I can be really happy with my printer 🙂
Thx
Although the machine itself operates to a higher precision your expected limit of accuracy, after you have corrected/calibrated for thermal contraction, is half an extrusion width - a fraction larger than your nozzle radius. So with a 0.4mm nozzle you can expect to see 0.225mm variation by default. In practice you can improve on this with careful design but you must account for filament squish and the stress displacement on corners and curves.
Cheerio,
Cheerio,
RE: Accuracy of the Core One L
I am currently tackling the problem of shrinkage, and I found that overall dimensions and the dimensions of holes shrink at different rates. Unfortunately, PrusaSlicer does not have separate settings to address this. However, OrcaSlicer offers dedicated options called X-Y hole compensation and X-Y contour compensation.
Unfortunately it's difficult to predict the degree of contraction in any but the simplest cases. Contraction along the lay is always greater than across and, of course, layers interact.
If the finished dimensions really matter then the two approaches are:
Test print, measure, calibrate into the CAD and iterate until satisfactory - but don't expect ultra precision to survive a changing to a different batch of filament.
or: Design blanks to be printed a fraction oversize and then machine to fit.
Cheerio,
RE: Accuracy of the Core One L
I first tried the approach of “test print, measure, calibrate in CAD, iterate.” I was getting better results, around ±0.1 mm, but it still wasn’t very consistent or reliable..
Now I’ve shifted to an approach where I oversize the features and then machine them to fit. We want to machine holes, so I need to add more perimeters. However, it’s not possible to increase the number of perimeters only for holes, because in PrusaSlicer this is a global setting. As a result, PrusaSlicer adds perimeters everywhere, which significantly increases material consumption and print time, which is not ideal..
Is there a way to set a different number of perimeters for external walls and for holes?
Use modifiers: right click on the part - add modifier, adjust size and position, right click on the modifier - layers and perimeters, set.
Cheerio,