Strange behavior when cut into layers

Hello,
When generating gcode layers (Luban 3.80) between layers 382/383 and 444/445 there is quite a big jump between layers in the simulation. I present the case in the video posted on YT https://youtu.be/Gqw0K2N-554

STL file downloaded from: https://www.thingiverse.com/thing:3895754/files

I know it’s just a simulation, but 8 hours of printing for the test is quite a lot.

Has anyone experienced anything similar? Could it be just a wrong simulation and the generated G-code is correct.

Best regards

I’m running Luban 3.3.3, and I see something similar on that model, although it happens at different layer numbers for me. It doesn’t happen on another tall model that I have successfully printed. The issue doesn’t occur on your model if I use the stock “Normal Quality”.

After playing around with it, the anomaly occurs when I enable “Z hop when retracted”. If I disable that again, the discontinuity goes away. So I think it’s a real thing that will cause a problem when you print. I’d suggest upgrading, but given that the issue has been around since at least 3.3.3, an upgrade is unlikely to fix it. I have a v1, and I’ve never needed to use that feature. Without it, you might get some stringing between the features on the base, but that’ll be easier to clean up than the supported surfaces.

I would suggest that you should plan on being around once it starts working on the horn. That narrow of a feature is going to print faster than the PLA can cool off. With the angle of it, it’ll likely droop and spaghetti. I suggest adding some active cooling – my preferred method is blowing on it. It should go pretty quickly once it gets to that point, so you shouldn’t hyperventilate. :smiley: Luban never gets the time estimate right, but I’ve also never seen it underestimate. I’d set an alarm for 8hours and 27minutes, then come and watch it.

I guess this is a render issue of Luban.
You could preview your gcode with other slicer software to double check and or upload the code here/elsewhere than I could do that for you.

I don’t have another slicer. If you provide an email address to which I can send the gcode, I will send it.
The file is over the limit I can post on the forum.

Upload it to wetransfer or a other cloud hoster.
You are welcome!

@Gorylsam86, if I return the model to the default 100% scale, I still see the issue. Both in the head and in the horn, pretty much at the same spots in the model (but obviously a different layer). If I scale the original to 50%, the change is even more obvious, and the GCode is small enough to upload (and will only take 25min to print).
Unicorn_Casual_-_Decorative_Base_Anomoly.gcode (2.2 MB)

Between layers 93 and 94, there is a very large jump. The preview renders 8 additional layers instead of 1 additional layer. Layers 109 to 110 also adds 8 layers.

@xchrisd, if I slice in Cura 4.7.0 w/ Z Hop when Retracted, I don’t see the issue. I didn’t match the print settings 100%, since I was able to reproduce in Luban regardless of print quality or supports.

Here is the link:

Hey, i have had a look at your gcode and see no strange behaviour.- i guess its a luban render issue.

@parachvte, may you have a look at? The files are in previous posts.
I can reproduce under win10 latest updates and Luban 3.8.0, the issue occurs only if previewing the layers one by one, like the youtube video in the first post.

1 Like

The 50% version was only 25 minutes to print, so I gave it a try.

It worked fine, although I did get drooping on the horn.

From how Luban rendered it, I expected the head to stop printing right around where the jaw and head connect, with a bunch of loose filament stuck to the side of the neck. I don’t see obvious lines or holes in the affected areas. Looks good to me.

Ok thanks for the sample.
Looks like Luban is misrepresenting the simulations. I will try to print on the settings shown in the film.
We will see if the print is successful.