Y-Axis Print object too small
Hello everyone,
I need your help because I can't get any further on my own.
First of all, I have been printing for 8 years and have already gained a lot of experience in 3D printing, slicers and calibration.I recently cleaned the dust off my old MK2s and overhauled it. New bearings, new print parts, new extruder parts. All in all, the printer runs very cleanly, smoothly and easily. It is firmly assembled, the belt tension is good and the self-calibration was perfect except for a minimal degree of deviation.I have also calibrated the current PLA filament and can print with high quality.
The problem is that the printed objects are too small in the Y-axis direction. The X-axis fits perfectly! The undersize can only be found on the Y-axis. I have printed different parts and also printed the following calibration pattern:
https://github.com/dirtdigger/fleur_de_cali
I have also attached the Excel list so you can see my measurements and deviations in the Y-axis.It can't have anything to do with the shrinkage of the part as the X dimensions are perfect. I have already thought about whether it is the stepper motor steps per revolution, which as far as I know are 100 for X and Y as standard on the printer. However, it makes no sense to adjust these values as nothing has changed mechanically in the transmission ratio.I also have no settings in the slicer (Orca slicer) that change my dimensional accuracy.
Can you please tell me what the problem could be or what else I can check?
If there is no other way in the end, I would also adjust the steps per revolution for the motor on this axis. However, that would be my absolute last idea.
Thank you very much!
Tim
So your greatest error is almost one extrusion width undersized. Your base unit of accuracy is half an extrusion width as the extrusion path is defined by the centreline of the extrusion.
Also, this is a mesh structure so errors in one dimension may be compensated into another by, er, squishing. It's possible that the X dimensions are being maintained by compressive forces from the contraction in Y.
The thermal coefficient of expansion for pure PLA is about 8.5 * 10-5 per degree K (effectively °C) but 3D print extrusion tends to align the molecules adding a directional bias and as soon as you add any inclusions, eg pigment, you modify it, sometimes drastically. Simple calculation would predict a contraction of about 1.8mm for your longest dimension - but the biggest measurement gives less than the 0.45mm extrusion width...
To get a more accurate measurement of the machine's inherent accuracy lay down two short, unconnected, parallel lines '| |' on opposite side of the print sheet and measure how far apart they are - then factor in the contraction of the metal print sheet.
I suspect your printer is well within design limits.
For any dimensionally critical parts you have two options: Print, measure, calibrate and repeat until acceptable or print oversize and machine to fit.
Different parts will require different calibration.
Cheerio,
RE:
Hello Diem,
thank you very much for your detailed answer.
To test the theory, I printed a cube in vase mode. Then I measured the wall thickness with the micrometer screw. I got the set 0.45mm up to 2/100. According to this, the extrusion multiplier is perfectly adjusted and there is no under- or over-extrusion.
Then I printed 4 lines on the print bed. Two for X and 2 for Y, each with a distance of 170mm. I then measured these lines as well as possible with the caliper.
The distance on the X-axis is 170.08mm. A slight oversize that shrinks when cooling on the object and explains why the X values in the calibration print fitted perfectly. It looks different on the Y-axis. There I measure 169.75mm.
The belt is tight and new. The original belt pulley is on the stepper and has no no noticeable tollerance. The Y-axis has new bearings and the new retaining plates.
For me, this means that the motor does not move enough steps to reach the distance of 170mm.
How can this happen? Do you have any more idea?
Sincerely
So your error is 0.25mm with a default extrusion width of 0.45mm - roughly half an extrusion or a 0.15% difference. This is within the expected range and more accurate than most people could cut wood, or even metal.
Next you repeat the exercise for a couple of different lengths and determine whether it is a purely percentage gain or if there is a constant component.
And apply the calibration to any part for which the dimension matters. Simplistically you resize the cube to 170x170.25mm or another to 100x100.15mm
.. then you trial print, measure and adjust.
Cheerio,