@20ffb98d @5756e48c IME silk has abysmal cooling (ntm layer adhesion) properties.
@2b562a3e @5756e48c This is not a TPU blend, unlike most silks. I had this issue with regular PLA as well - that tail tip is kind of a tiny detail and is a bit obnoxious. Might knock the minimum speed down even more to try and cool better.
@2b562a3e @5756e48c Layer adhesion-wise, you gotta bump silks at least 10 degrees if not more (again, TPU blend forces it).
@20ffb98d @5756e48c I've had no luck with the ones I've used. Tried 260-300 even. It's like the silk finish is a layer on the outside that doubles as a release agent.
@2b562a3e @5756e48c Which ones have you tried? I've only really used the Polymaker silks, and while they did need a temperature bump, I never had issues to that level with layer adhesion.
@20ffb98d @5756e48c Recently YOUSU. Here is a clean layer separation made with no leverage needed (broke on first try with thumbs touching each other). https://media.hachyderm.io/media_attachments/files/111/034/929/040/781/025/original/fdfbef0adb08bbac.jpeg
@20ffb98d @5756e48c In that case you probably just need more cooling. Lowering min speed has diminishing returns IME because keeping the hot nozzle nearly in place for that long dumps as much heat into it as the fans are taking out.
@20ffb98d @5756e48c Another option would be printing 2+ at a time or a sacrificial tower just to get some time with nozzle away. Cura has a "lift head" alternative to this too but I find it oozes horribly.
@2b562a3e @5756e48c Sacrificial tower might work. I might also look into whether my fan ducting is suboptimal - I suspect it is, but CFD modelling isn't something I know how to do yet.