I have problems with layer widths and uneven outer walls on my 3D prints. Sometimes layers are squeezed and sometimes pushed outside. I noticed that these problems happen when there are retractions on layers. I don't have problems with round and simple objects without changes on layers or where all layers are identical.

Image showing uneven layers

I think that this can be linked with pressure in the nozzle or retraction. It looks like sometimes the pressure is too high and it's pushing too much filament and sometimes it's too low and it's not pushing enough filament. In this picture, layers in the first circle are squished in and it looks like there is not enough filament being pushed from the nozzle. In the second circle layer is pushed out and is wider than other layers. Looks like there is not enough filament being extruded at the start of the layer and too much being extruded at the end.

Image showing two areas: one with layer pushed in, and one with layer pushed out

This problem happens always on the same layers, where there are some retractions on these layers. I tried to print the same objects multiple times and it always occurs on the same layers.

Image showing squished in layers

Image showing squished out layers

What have I did to fix this problem:

  • calibrated e-steps and slicer flow,
  • tightened belts,
  • different slicers (Cura, PrusaSlicer),
  • different filaments (e.g. PrusamentPLA),
  • disabled combing in slicer,
  • printed with and without infills,
  • different retraction distance (from 3 mm to 8 mm),
  • different retraction speeds (from 20 mm/s to 80 mm/s),
  • different retraction accelerations (from 500 to 1500),
  • different hotend temperatures (from 200 °C to 230 °C),
  • slow printing speed (up to 15 mm/s),
  • calibrated K-factor (also tried many values from 0.0 to 2.0),
  • calibrated junction deviation (also tried many values from 0.0 to 0.3),
  • printed with and without cooling.

I think, that disabling coasting and changes in retraction settings helped a little, but not too much.

My setup:

  • Ender 5, SKR mini E3 V1.2
  • Capricorn PTFE Bowden tube,
  • printing with PLA, but this problem also occurs with PETG.

The most similar issue I've found is this question, Inconsistent Layer Issues, but this didn't resolve my problem.

How can I get rid of inconsistency in layers widths and get smoother outer walls?

I calibrated K-factor with K-factor Calibration Pattern and discovered that lines are always thicker after retraction at the start of the line and thinner at the end of the line. Then I generated and printed test files to confirm this. My lines are always thicker at the start of the line and they get thinner later. This is the problem presented in the second picture.

I printed 3 cubes (dimensions of a cube: X=0.2 mm, Y=100 mm, Z=10 mm). This is the result.

Result of printing 3 cubes (X:0.2 mm, Y:100 mm, Z:10 mm)

This is the best representation of my problem. The order of printing was as follows: 1 -> 2 -> 3 -> 4 -> 5 -> 6 -> 1 -> 2 etc. There are under extrusions at the end of the second and third line (points 4 and 6) and under extrusion at the start of the first line (point 1). These under extrusions are before (points 4 and 6) and after (point 1) travel moves.

What can cause this problem?

  • 1
    $\begingroup$ 200 is absolute min I'd ever consider for PLA, and 230 is what I normally print roughly-pure PLA at. $\endgroup$ Sep 29 at 0:19
  • $\begingroup$ Are the locations of the inconsistency consistent between two prints of the same gcode? Or do they differ each time you print? One stupid thing it could be is filament with really bad diameter consistency. $\endgroup$ Sep 29 at 12:15
  • 1
    $\begingroup$ They are always the same locations. Exactly in the same places. I tried to print multiple times the same gcodes and with different filaments. It's not filament related. $\endgroup$
    – kosteklvp
    Sep 29 at 12:41
  • 1
    $\begingroup$ Based on the latest pics, I wonder if you have the bowden tube misinstalled (not butted against the nozzle correctly, or otherwise). $\endgroup$ Oct 1 at 20:11

I noticed that these problems happen when there are retractions on layers. I don't have problems with round and simple objects without changes on layers or where all layers are identical.

Based on the above, this is not a heat related problem. It's likely inconsistent amounts of material lost from the nozzle just prior to the affected walls. This could be due to oozing during travel without retraction (make sure combing is either off or has a "Max Comb Distance With No Retrace" set very low, < 1 mm) or due to inconsistently unsupported infill structure. One cause for the latter is discussed on this issue report I filed for Cura, though it could happen in other ways as well. This is particularly bad with bowden extruders; ever since I switched to direct drive it mostly went away. But to avoid it with bowden, I had to turn off zig-zaggify infill and use only infill patterns where each layer fully supports the one above it (e.g. triangles).

  • $\begingroup$ I edited my question. I was reffering to combing (not coasting) under things that I've already tried. I added also subsection about infills. I've tried printing with and without infills. $\endgroup$
    – kosteklvp
    Sep 29 at 8:06
  • $\begingroup$ @kosteklvp: You should edit that to fix it because coasting and combing are completely different things. Can you confirm that you have coasting off? It's a very problematic hack and should not be used. $\endgroup$ Sep 29 at 12:14
  • $\begingroup$ Yes. I meant combing, don't know why I've written coasting. $\endgroup$
    – kosteklvp
    Sep 29 at 12:38

That kind of issues can be related to multiple causes:

  • Heated bed operated in "bang bang" mode (heater on until set temp reached, turned off until set temp minus threshold is reached, and so on), which causes periodic expansion and contraction. It is solved by using PID algorithm for bed heating

  • The slicer is set with a minimum layer time and the part is not big, resulting in a lower extrusion speed for smaller layers, and higher extrusion speed for larger layers (of course up to the value set in the slicer). When you have a Bowden system with single gear (like Enders do), the extrusion is not linear: the more you ask, the bigger the gap between what you ask and what you get. See this (orange line, from Testing bimetallic heat breaks):

    Flow rate and speed

    In other words, slower (smaller) layers extrude more plastic and look fatter. The way to solve is to get a dual drive extruder and by improving the quality of the hotend, for example by installing a bimetallic heat break (see grey line above).

In your case, image 1 and 2 appear to be related to variable extrusion, and image 3 to bed heating.

  • $\begingroup$ Each picture refers to the same problem. Image 3 pictures a top of a bucket, which has the same problems with layer widths and is not related to bed heating. Maybe I didn't showed and explained it exactly how I wanted. There is a lot of sense in second proffer and it may refer to my problem. $\endgroup$
    – kosteklvp
    Sep 28 at 17:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.