I sat on this for the last days and couldnāt give up. I bought a dryer that is capable of 80c and dried the PETG again for 8h @ 65c. The strength and appereance was the same. Next I calibrated everything again and got the best results with the following settings:
Nozzle temp: 250c
Bed temp: 70c (75c for taller stuff)
Flow ratio: 0.97
Pressure advance: 0.0395
Volumetric speed: 13mm²/s
Retraction length: 0.2mm
Min fan speed threshold: 10% | Layer time 30s
Max fan speed threshold: 20% | Layer time 12s
Following that I printed the test piece again, the strength was (minimal) better, but still very brittle. So I began to read some articles about PETG and factors that influence the strength. Below a summary of tips I found online and want to share for other people that might have the same problem as me.
Wetness:
The PETG should be fully dried, because wetness reduces strength alot (among reducing other things like stringing, holes, etcā¦).
Speed:
Many people say, that PETG should be printed at low volumetric speed and adjusted general speed (like outer wall, inner wall, infill). While this is kinda true, I tested the strength for several (volumetric/general) speeds and couldnāt find a meaningful difference between them.
After running the max flowrate calibration I reached a maximum flawless volumetric speed of 16mm²/s (I settled for 13mm²/s just to be safe).
Temperature:
Alot of people stated the temperature is too low for good layer adhesion and therefore should be raised. After several temperature tests together with adjustments to flow ratio and speed, I got the best results in the range of 240-250c. The layer adhesion was the best at 250c.
Note that if you want to run at higher speeds a higher temperature helps alot, but brings cosmetic problems like stringing/distortions.
Fan Speed and Layer Time
Now this is were changes finally showed a meaningful difference in strength. After reading so many comments online, trying all of the above, the comment by @Bjorn inspired me to try fiddling with these settings.
I printed multiple tall thin pieces and changed the Maximum part fan speed by 10% every print, starting at 100% (Min fan speed was set at 10%).
The lower the fan speed was, the stronger the parts were. At 20% max fan speed the part was finally as tough as I needed it to be.
If I had printed an other model, with a higher layer time (above 12s), I wouldāve never run into these problems, because then the fan would run at 10% on provide good enough layer adhesion.
Before these adjustments the test part just snapped like a thin piece of wood, now the print just bends and shows a little crack very late, while not breaking at all.
Conclusion
If you want to print a strong small part, adjust the fan speed and layer time accordingly. Thatās also true if your print has small features, that need a good layer adhesion. Here you can adjust the gcode for these layers (M106 P1 SXXX, where XXX is a number between 0-255. 255 means 100% fan speed, while 128 is equal to 50%). Just remember to adjust the part fan before the other layers that donāt need the additional strenght).
I donāt know if anything I write here makes sense, but that is what worked for me. Maybe I can help someone with this write-up.
Next To-Doās
Next up are test prints for overhangs and bridges, to see which settings perform the best here, while still maintaining a good strength.