Question about retraction distance and stringing/print quality with PETG
I've tried printing the Basic Stringing Test on my CR-10S using PETG (250 °C nozzle, 80 °C bed) with different retraction distances leaving all the rest unchanged.
In the image below you can see 0 mm, 5 mm and 10 mm retractions (in order left-to-right).
The result stunned me because I would have imagined that raising retraction distance would have diminished the stringing.
Why is that?
I know that having too high of a retraction can lead to other problems (namely under extrusion and possibly heat creep). The stringing surprises me too.
Yes, it's very counterintuitive! Should I go beyond 10mm to see what happens?
@FredoCorleone No, but lowering the temperature also reduces strings, PETG can be printed at a lower temperature.
You cannot endlessly increase the retraction distance, doing so leads to different problems as you encountered. As a rule of thumb, the retraction distance should not exceed the length of your nozzle. Depending on the type of extruder, many printers use a value between 2 and 7 mm (e.g. the Ultimaker Cura retraction length is 6.5 mm at 25 mm/s, this is for a Bowden style extruder).
In fighting stringing (in which PETG is very subjected to happening) retraction length is only one of the parameters that you need to adjust to optimize the printing process. Other important parameters (besides retraction length) for stringing are:
- retraction speed
- travel speed
- minimum travel distance
Furthermore, the type of extruder (direct or Bowden) is also important and other parameters that may reduce the stringing are:
- coasting (stop extruding before the end of the print line is reached using the residual pressure in the nozzle to extrude the last filament)
- pre-loading/priming filament after a move
Usually it suffices to tune the first 4 parameters. As you see in your results, the retraction of 5 mm is close to the result you would expect, increasing it, may lead to clogs and in your case leads to a sort of stringing. This latter result is probably due to the fact that the filament does not reach the print in time (you definitely see under-extruded cylinder columns), while leftover pressure after printing the column layer oozes some of the filament creating strings.
Just play with the retraction length in steps of 0.25 mm (around the 5 mm) and you may try to change the speed with 5 mm/s increments if necessary.
Another thing you should add is "combing", my printer was not retracting filaments in some occasions because that setting was enabled.
@FredoCorleone If combing is "off" the material will retract and move straight to the next position, otherwise it will use already used paths. This can indeed cause less strings. But in this case it does not matter as it concerns travel speed in between areas where there does not exist a printed structure, it is movement from one to the next tower.