Fit variation, probably quantization
I'm doing a series of test prints to try to fit to a Lego 2x2 block. The goal isn't important, though.
I did a print with 4 samples with 4 different inside dimensions, 12.5, 12.7, 12.9, and 13.1, and found that it's between 12.7 and 12.9.
So I printed 4 more samples with 4 dimensions, 12.75, 12.8, 12.85, and 12.9. I was surprised to find that the 12.9 fit was different between the two groups of samples.
Finally, I printed a set of 4 12.8 dimension parts, and they had different fits. The obvious conclusion is that it depends on the X, Y position. I assume this is caused by the quantization that happens in slicing.
I did another test, and found that it seems to be associated with the relative position compared to the X, Y of the start of the print, not the absolute X, Y of the printer.
So (finally) my question.
If I want to print a plate of 3 objects that all have the same internal fit, do I need to set their positions to be a multiple of the extrusion fit (I'm using the default 0.45mm)? Is there some other way to make the objects consistent?
The new slicer has features designed to address this sort of issue. It would be worth trying it before making any specific changes.
Cheerio,
RE: Fit variation, probably quantization
The new slicer has features designed to address this sort of issue. It would be worth trying it before making any specific changes.
Cheerio,
I'll that, thanks!
RE: Fit variation, probably quantization
I upgraded to 2.5.0, it still behaves the same. When I print a series of objects with inside width of 12.95 in the STL, what gets printed varies from 12.83 to 12.88 mm.
I can correct if it's a constant factor, but I don't know how to correct when the printed dimension varies by where it ends up on the print surface.
So the error is approximately a quarter of the default nozzle diameter and varies by one eighth.
If you are 'measuring' the stl by the design diameter be aware that almost all design software creates a cylinder by approximating a polygon with its vertices on the designed circumference and slicer will take this as a centreline position for the extrusion.
You are not going to get much closer. Do these parts have to be snap-fitted repeatedly? They are likely to wear in use...
I would be inclined to calibrate out about half of the absolute error and wear the parts to fit by repeatedly fitting and removing them from a typical lego block.
Cheerio,
RE: Fit variation, probably quantization
So the error is approximately a quarter of the default nozzle diameter and varies by one eighth.
If you are 'measuring' the stl by the design diameter be aware that almost all design software creates a cylinder by approximating a polygon with its vertices on the designed circumference and slicer will take this as a centreline position for the extrusion.
You are not going to get much closer. Do these parts have to be snap-fitted repeatedly? They are likely to wear in use...
I would be inclined to calibrate out about half of the absolute error and wear the parts to fit by repeatedly fitting and removing them from a typical lego block.
Cheerio,
Thanks, Diem! That will be my plan B. I'm doing some experimenting to see if I can add a couple of minimal objects to force consistent location on the print plate, then ensure the objects are at multiples of the nozzle diameter.
Regards;
Rick