Seemingly independent of anything I’ve found to adjust so far, even when I get the flow rate way, way down (far too low overall), I am still getting so much extra plastic on that overhang that even after a few layers it looks awful:
Has anyone ever printed a model like this on the Taz? With 0.1mm XY tolerance I would think it’d be doable, and clearly others with different equipment are printing it fine, one even with slic3r?
Are you using a fan? ABS or PLA? If PLA you’ll really want to install a fan. Your prints will improve tremendously. Ideally the fan would blow on the top couple of layers of the printed object, not just on the nozzle as some mounts are trying. You’ll find that it may not be too much extrusion, but when it’s printing in air it’s not cooling fast enough. It can get pulled back in when the print head moves back. It’s hard to tell without seeing your gcode file/Slic3r config.
Ya, it almost looks more like the overhang is curling up and looks like too much material, but that is because it isn’t flat. Supports or a fan would probably be best for this part.
Printing with PLA. I did build and install a fan. See Taz fan mount! - would love your comments on the adequacy of this particular fan mount or any suggestions for a different one. I fear anything attached the nozzle will spend as little time on that drooping overhang as the nozzle itself does. Which is to say, very little at all. Have also tried larger fan aimed at the print bed - results not good.
When I tried this model a couple of weeks ago I found that the rings printed rather closely and needed to be cut apart. Support removal would be trivial. Change the support spacing to 4mm and it’s sparse and easily removed.
Supports it is! The fact that you got it to print at all is good to know.
There’s no way you still have, or would be willing to share, more details about how you did the print, that slic3r config or gcode, is there? Any clues would probably be very helpful.
Even without the overhang problem, the rings almost always fused at the bottom. I was able to get relatively clean separation on the upper layers, but I kept running into what felt like the lower limits on flow rate out of the 0.5mm nozzle. Saw a lot of inconsistent flow and jams. How extrusion speed interacted with temperature is something I’m still trying to understand.
Thanks again for the advice. The generated support worked admirably. I try to avoid using ABS slurry, but that further reduced overhangs simply by keeping things firmly attached. I was sure I had it this time:
bracelet_0.5 v33.ini (2.81 KB) Rotator_Cuff-XXL - v33.gcode.zip (1.7 MB)
But the result is just a bit too fused together. The distance between the parts doesn’t seem to be consistent around the ring, and there are two or three spots where it’s bad enough to be beyond my carving abilities.
I’ve been playing a lot with layer widths, thinking that this was the secret to fine-tuning the real distance between the parts. Do you think that’s likely to be the answer? Or should I look at something more basic, like extrusion multiplier?
Looking at the latest one, I would say 2 things need to be adjusted. Your infill needs to be set higher to avoid the middle void you ended up with, and your filliament needs to be measured with calipers and the exact diameter entered into prointerface. You’re pretty close, but I think you are getting a little bit of overextrusion in the corners. For the areas where it is out of round, you might try adding some belt tension. Maybe try printing it with 3 shells instead of the usual 2 as well. I find that helps on some things. You might also be getting some first layer squishing on the right hand side of the part, which could indicate a bed slightly out of level.
I’ve attached an example. I tried all kinds of things, and got all kinds of dramatic variations in the print, but never to those sloppy overhangs. I did not try enabling supports. Since this is designed to print an interlocking part, I was afraid the supports would get into the spaces between the rings and mess it up. Can give it a shot, though.