sojhalb / curaengine Goto Github PK

Fiber cut seems to be put into the wrong spot on some layers with the fiber cut length set to ~8cm.

Currently /dev 1074ea still outputs X axis labelled G-code instead of the expected B axis

Basically the same issue as the test cube but with a more complicated shape the cylinder with window.

Fairly sure that some compensation is in place for fixing the extrusion lengths for higher radii, but it doesn't stop the layers from separating at around 20+

Despite all the fiddling with settings slicing through Cura doesn't work particularly well

Especially for the longitudinal seam, the lack of overlap makes the seam quite weak. The slicer attempts to make the exterior dimensionally accurate and the seam is treated at exterior so it's technically not overlapping at all.

Next step for this is to run cura with external backend and check to see if the angle numbers are really getting into the backend

Fiber for the test cube has some issues not placing the cut at the right spot with the new ~8cm fiber cut length

To be confirmed, in the test print it seems okay but the next step for this issue is to check the extrusions for rings on low versus high layers

Layer 25 (24 in g-code) of test cylinder and layers 43 and 44 (42 and 43 in g-code) of test cube have small bad extrusions sticking out. More details to come

Layers 99+ for test cube print show some artifacting which is thought to be because there are a large amount of travels (and therefore cuts) being inserted for relatively small amounts of extrusion.

There is still a seam and that's unfortunate.

So I added the segmentation to B extrusions into the FlavorParser of Cura but if you slice from inside Cura it doesn't go through the flavor parser it just kind of shows up in Cura. Probably should have seen that one coming.

The infill is calculated assuming that the X axis is in millimeters while it is actually in deciradians. For a completely filled in layer (62.8... deciradians) it is only accurate when the surface is exactly 62.8 mm, or at a radius of 20mm. Lower than 20mm the surface will be smaller than 62.8 deciradians and the infill will be too dense, higher and the infill will not be dense enough.

Should be in the Cura issues, but basically cura should cut up the lines before sending them to the shader not in the shader

Could also be a firmware check to see if the tool active is the fiber tool, and then do the fiber cut only if the fiber tool is enabled

Seems like Y walls, especially the square window in cylinder with window, don't have the right B coordinate