2

I was printing test cubes to adjust my extrusion ratio, and had ugly walls. I noticed that the pattern on the walls was always the same, and disappeared if I chose 0 % infill (gyroid from PrusaSlicer).

What parameter do I need to change, in order to avoid the marks on the outside?

enter image description here Top two cubes with no infill. Middle row with extrusion multiplier from 90 to 75 % (!) in decrement of 5, temperature 215 °C. Bottom cube at extrusion 85 %, temperature 190 °C.

Here is an example of when this can turn out to be a real problem, the inside floor of the part shows on the outer wall:

enter image description here

I was surprised that the extrusion multiplier had so little effect, so I tested 50 % just to see that it had:

enter image description here

More info: this was done on a Prusa i3 MK3S+, and sliced in PrusaSlicer.

I can't seem to find how to show the seam anymore, but I do see something interesting:

enter image description here

Even without the layer change, there is a hop in the corners that really look like they could be the cause: enter image description here

Changing to random Z seam fixes it in the slicer. I have yet to print it.

My retraction parameters (no Filament override):

enter image description here

I performed the Linear Advance calibration from marlinfw, for good measure. Now I got a new value, I'm pretty sure it was a clear 0.05 before, now it's 0.09.

The calibration I made 10 days ago (chose LA K=0.05): enter image description here

The calibration I made yesterday (chose LA K=0.09): enter image description here

I also tested the next PrusaSlicer release, 2.4.0-beta3. The seam moves quite a bit more from layer to layer, so the issue is not as visible. I'm surprised that LA has so little effect: enter image description here

Interestingly, slicing with Cura defaults plus coasting (retraction 0.8 mm, 35 mm/s) gives me much smoother walls and not the artifacts above. I'm not saying this isn't luck, and that I don't still have a backlash issue, but this tells me there's something I could do about it in PrusaSlicer. enter image description here

Gauthier
  • 161
  • 4
  • 1
    It looks to me like the seam point is changing between layers depending on how the infill paths fall, and you have a bulging seam. Can you confirm this by looking at the slicing output or by switching to a "2D" infill (where all layers will look the same, e.g. "triangles" or "grid") and seeing how the results change? – R.. GitHub STOP HELPING ICE Dec 07 '21 at 16:21
  • 1
    It also looks like you may have backlash on X and/or Y axis from loose belts. – R.. GitHub STOP HELPING ICE Dec 07 '21 at 16:21
  • Did you mean flow modifier with `extrusion multiplier from 90 to 75 %` or is this the infill percentage? If it is a flow modifier, the decrementing does not seem to have much effect nor is it applicable for troubleshooting the issue. No offence, but even the no infill cubes seem to be of very good quality (so actually not good). – 0scar Dec 07 '21 at 21:17
  • I meant extrusion multiplier (it is called flow modifier in Cura, I believe), I added a new image at 50 % to show that the parameter does have effect. Do you really mean "very good quality"? The no infill cubes are ok, but I hoped to be able to get the walls smoother. Anyway, the smoothness is not as high on my list as the wider ribs that appear with infill (and inside floor). – Gauthier Dec 08 '21 at 13:02
  • 1
    Extrusion multiplier is not your problem. Set it back to 100% so it's not distracting from the problem and start looking at the things that might be causing this. Indeed the quality is not good, but it should be possible to track down what the problem is. – R.. GitHub STOP HELPING ICE Dec 08 '21 at 13:44
  • Ok! I was given the recommendation to reduce extrusion to solve the ribs. Any hint as to what things might be causing this? How to prevent the inside structure from influencing the outside wall? – Gauthier Dec 08 '21 at 14:05
  • More walls, outer wall first may work. But, your machine doesn't print well, even without the infill, hence @R.. mention to the belts. Edit your question to include the printer brand, or if self-designed, a photo. – 0scar Dec 08 '21 at 15:20
  • @Gauthier: Look at the suggestions in my first and second comment. – R.. GitHub STOP HELPING ICE Dec 08 '21 at 15:29
  • It looks like you are right about the seam, see my edit! Do you think that the "bulging seam" is caused by the backlash, or something else? I expect the result to look better after randomizing the seam (or aligning, maybe), but there is still something wrong with the bulge. I've done Linear Advance already. – Gauthier Dec 08 '21 at 17:20
  • Bulging seam could be caused by not having Linear Advance tuned well, by having retractions too slow (allowing oozing during retract/unretract), or perhaps other things. What is your retraction length, speed, and acceleration? I think you do have backlash (loose belt) problems too that might be making this more noticable, but that's a separate problem for the most part. – R.. GitHub STOP HELPING ICE Dec 08 '21 at 18:07
  • @R..GitHubSTOPHELPINGICE Post updated with retraction settings. These are the Prusa default. – Gauthier Dec 09 '21 at 12:40
  • Retraction settings look pretty much ok, but there's a hidden parameter not shown, the E-axis acceleration limit. If it's very low (below a few thousand) the retraction will be very slow because there's not enough distance to accelerate to the desired 35 mm/s speed. – R.. GitHub STOP HELPING ICE Dec 09 '21 at 13:32
  • @R..GitHubSTOPHELPINGICE Right, it's at 5000 mm/s^2. Also, maximum E jerK: 4.5 mm/s. – Gauthier Dec 09 '21 at 14:41
  • I just came back to say I'm sorry I don't have any more good suggestions at this time. This is such a well-asked and well-documented question. Your retraction settings all sound very reasonable and unlikely to be part of the problem. If you haven't checked for play in the belts yet, that's probably your best bet. It could cause erratic seams dependent on where the XY motion last changed directions. – R.. GitHub STOP HELPING ICE Dec 11 '21 at 16:21
  • @R..GitHubSTOPHELPINGICE I'm really grateful for your help, I really appreciate it! – Gauthier Dec 12 '21 at 11:13

0 Answers0