During a functional print of a spindle, the very top of the spindle was higher than any other object on the build sheet, and the layer time got very low, about 4-5s. This is the result (here is the issue recreated with just the tip, one of several attempts to fix this problem I tried):
The 3MF for the above is here:
VSpooler-Barrel-Cam&Gear-cut.3mf (953.3 KB)
So, I tried a few different things. It seemed that no matter what I tried setting for the “part cooling” settings in Bambu Studio, the cooling showed as the fastest speed in the “Preview”, but when it printed, the part fan didn’t actually switch on. I cracked open the GCode from the print and found that there was nothing but “M106 S0”. I then tried slicing it with OrcaSlicer, and it did emit GCode commands (M106 S252 IIRC) corresponding to the “Fan speed” preview, but the part fan never showed as running on the display or via remote management (although for a number of these prints, I didn’t catch it while it was printing, so I can’t be 100% on this). A lot of this also seemed to be down to the print speed, so I tried lowering the minimum print speed from the 20mm/s default to 10mm/s, and this did help slow down the layers a bit.
Am I missing something I need to do in order to enable part cooling? I’m printing with Bambu ABS Red 40200.
In the end I added another structure to the print which forces the last 8mm or so of the spindle to have longer layer times, at the expense of $1 of filament for a model I don’t need. Obviously, this is an ugly workaround. This seems to be a slicer bug. If I’m right about that, here’s my suggestion for a fix:
- if the layer time is low, it’s very likely that some very small parts are being printed. The printing should slow down: not just print speed, but acceleration, jerk, etc. IOW, it just needs to chill and print those fine tops of the model correctl;y This could be accomplished from the slicer.
- If the time is still too small, insert print pauses à la “G4”. The command would probably have to be carried out with the toolhead over the purge chute, and resume with enough time to run the nozzle over the cleaning roller. Maybe noisy, but at least it would make the shape I hit print on, instead of the garbled mess I got.
- (may or may not help) also consider using the two thresholds, and fitting an appropriate curve to them, so that the fan speed is set per layer at a variable rate when it is between the first threshold and the second threshold.
- ask for two more parameters from the material data sheets: Tc (crystallization temperature), and C (specific heat capacity), and automatically calculate the correct fan speed live, based on the chamber temperature and surface area of the new layer.
Anyway, if anyone has the obvious, simple answer I missed, please share iit!