I have a problem with inside cutouts printing smaller than I'd expect.

3D Printing
, , ,
1

I have a problem with inside cutouts printing smaller than I’d expect. I have a test part which is a square 1.000" x 1.000" with a .500" x .500" square cut out of the center.

I’ve dialed the printer with M92 and have the outer dimensions nearly perfect - x and y are printing at 1.001" and 1.002" respectively. However, the inside square is only printing at .975"-.980".

I’ve dialed my extrusion back until I can start to see visible indications of underextrusion, and tried with both Slic3r and Cura. Filament measures at 1.75 mm very consistently with calipers.

Any idea what I can do to improve this? Printer is a stock late 2013 Printrbot Simple.

Thanks!

  • Jason
2

I see this too. About .1mm error rate on interior spaces. I generally drill out or thread holes for fasteners, or when interior cavity dimensions are critical (ARLower, for instance) I alter the model to have a larger cavity, then the error is compensated for. I think that the slicing software is putting the innermost perimeter of the interior cavity just inside the volume of the interior space, which is how you would want the extrusion placed for an exterior perimeter. But that is only speculation, as I’ve never looked at the slicing code.

It also could be ABS shrinking while cooling. I think the average rate is about 2-4%

3

I think what’s happening is that your extrusion is wider than you think (larger than the nozzle). You compensated by adjusting for the outside dimension, but this pushed all of the error to the inside dimension.
I have no fix for this (yet), as I suffer from the same problem. Maybe someone can help us both out.

4

Don’t “calibrate” your printer by changing steps/mm. The dimensional error is inavoidable on FFF printers if your slicer doesn’t account for it - Skeinforge had the inset plug-in for that, but neither Slic3r nor Cura have something comparable.
Until they get that functionality, you’ll have to design your parts with those tolerances in mind.

5

Thanks, everyone, for your feedback.

Thomas, is there a preferred approach to getting the external dimensions correct? This looks to me to be a pretty standard approach, I’m curious what a better way is.

6

I’m also curious to understand what drives this limitation in FFF printing. Could you elaborate or point me at a reference?

7

Thank you! I’ll give Kissslicer a try.