I am a bit lost in optizing my print results.
At this point I do not know if the issues come from my slicer, marlin, hotend, filament or printer hardware.
Solid (20% infill) 20mm outer diameter, 16 mm inner diameter. https://dl.dropboxusercontent.com/u/56218477/Circle.JPG
However when printed:
the outer diameter: 19,9 mm (should be 20 mm, which is quite OK)
the ring width: 2,2 mm (should be 2 mm)
the inner diameter: 15,5 mm (should be 16 mm)
(had the same dimensions of the inner diameter with slic3r and my original hotend)
Thin walled ring 20mm outer diameter, 16 mm inner diameter. https://dl.dropboxusercontent.com/u/56218477/Inner_outer_circle.JPG
However when printed:
the outer diameter: 20,1 mm (should be 20 mm, which is is quite OK)
the ring width: 2,4 mm (should be 2 mm)
the inner diameter: 15,3 mm (should be 16 mm)
wall thickness: 0,6 mm (expected 0.4 mm)
nozzle size = 0.4 mm
layer height = 0.15 mm
How can the outer diameter almost spot on, and the inner diameter so off ?
Wall thickness of x * nozzle size does not seem applicable… Or does it ?
Why is the wall thickness not the same as the nozzle diameter ?
Can this be adjusted or compensated in cura ?
Is this something that needs to be contributed to the slicer as both cura and slic3r cause the same results ?
Issue 2: Artifacts
I printed a couple of temperature test towers to determine the optimal print temperature.
All towers had stringing, corner overshoots and wavy lines.
Before I printed my last tower I optimized the retraction using the above descibed test, and had no stringing whatsoever.
I also changed my xyjerk from 10 to 5 mm/s2 and my accel x & accel Y from 2000 to 1000 mm/s2.
However none of it made a difference, see images:
nozzle size = 0.4 mm
layer height = 0.25 mm
I am not an expert and I have only just fitted e3d v6 extruders on my K8400 but I discovered that when I was printing an a small object with a 4mm hole through it and a recess to take a hexagon nut that the internal dimensions of both the hole and the hexagon were too small. I checked my cura slicer settings and discovered that the flow was at 100% so I reduced it to 85% I also reduced the retraction to 2mm and when I printed the object the dimensions were fine.
Before I did the above I also ran a check on different temperatures and discovered that 85 degrees seemed to work best with the PLA filament I was printing with
Inside diameters being too small is a well known issue.
STL files define everything as triangles and so arcs and up as a series of line segments. Imagine a circle defined with just three line segments. The points of the resulting triangle are on the required diameter, but the effective resulting hole size is defined by what will fit inside the lines. The more segments, the closer to a true arc you will get, but the files can get far too big.
A second problem is that even if you coded arc moves (G2 and G3 are implemented in Marlin) the hole will shrink as the filament is ‘dragged’ around the hole.
Resonance artifacts are a problem on even the high end professional machines. Stepper or servo motors act like springs (you can push them out of position) and so do the belts and printer frame.
@rrekwa:
I experimented with the flow and wallthickness in cura.
And have the feeling that cura has a bit of a mind of its own.
To test I used a 20mm square with 0.4 mm wallthickness.
did the following tests: (layer thickness 0,15mm)
Cura shell thickness: 0,3mm 85% flow -> actual 0,6mm
Cura shell thickness: 0,39mm 100% flow -> actual 0,8mm shell is irregular
Cura shell thickness: 0,21mm 100% flow -> actual 0,55mm shell has holes
Cura shell thickness: 0,35mm 85% flow -> actual 0,65 / 0,7 mm
I heared somewhere that there was a rule of tumb that for good layer adhesion, states that for a 0.4mm nozzle the extrusion width should be 0.5mm.
However I am not able to get my extrusion width under 0,6mm in a reliable manner, let alone to 0,5 and 0,4 mm.
I expected to have noticable effects when dropping the flowrate by 15%.
The shell thickness in cura is also some blackmagic factor, as If I set 0.4 mm cura does not slice, so I need to set other factors, but 0,2mm is the not allowed as well… ??
I guess that the only way to deal with this is print a hole calibration test to determine the appropriate hole corrections, and apply this to the designs.
[quote]I experimented with the flow and wallthickness in cura.
And have the feeling that cura has a bit of a mind of its own.
To test I used a 20mm square with 0.4 mm wallthickness.
did the following tests: (layer thickness 0,15mm)
Cura shell thickness: 0,3mm 85% flow → actual 0,6mm
Cura shell thickness: 0,39mm 100% flow → actual 0,8mm shell is irregular
Cura shell thickness: 0,21mm 100% flow → actual 0,55mm shell has holes
Your correct about the extrusion width. It is important for the filament to be forced down onto the previous layer for it to bond. To achieve this the tip of the nozzle is flat, an excess of filament is extruded and forced into the gap between the previous layer and the flat of the nozzle. This forces the filament into a () shape as opposed to a () shape and as you can see makes it wider than the nozzle.
If you extruded 0.4mm with a 0.4mm nozzle you get just get cylinders layed next to and on top of each other with minimal contact or bonding, as opposed to squashed cylinder you would get extruding 0.5mm. Sort of like this:
oooooooooo ()()()()
oooooooooo ()()()()
I never design a print less than: nozzle diameter x 1.25 or 2 x nozzle diameter x 1.25.
So with your 0.4mm the thinnest wall you’ll get is 0.5 or 1.0mm for twin walled items.
For wall thickness above 2, the printer can print the outside ones first the force filament into the gap between them ()(___), so it could fill in a 0.2mm gap between two 0.5mm. This can cause its own problems though as on small objects if the first two passes haven’t cooled the filling in can force them outwards.
Try your test square again at 0.5mm wall thickness and you should be able to adjust the wall thickness through your slicer. It is possable to print smaller than your nozzle but basicaly dont. If you want thiner walls get a smaller nozzle.
