Print Dimension Tuning - Feature Request
Posted by to-the-nth
|
Print Dimension Tuning - Feature Request May 19, 2014 02:31AM |
Triffid Hunter calibration algorithm for E steps Fine Tuning talks about tweaking the flow rate until a solid fill is solid and 95% infill still has some gaps.
However it seems that once this is tuned, we are still missing a degree of freedom to adjust for the final wall thickness of the perimeter.
In the diagram below, we start with a 3 wide perimeter with minimum flow so the adjacent filaments are just touching. As the flow rate is increased to get better adjacent filament contact, the wall thickness grows:
-width of first filament increases
-2nd filament gets printed somewhat to the right of the nozzle centerline because it is pushing against the enlarged first filament and the 2nd filament is now larger.
-3rd/last outer filament again is pushed even farther from the center of the nozzle because it is pushing against the already offset filament 2. The third filament is larger because of the greater flow rate. And of course the right edge of filament 3 is unconstrained.
Experimentally with a 0.4mm nozzle, 0.4 infill, 0.4 perimeter extrusion width, 0.3mm layer height, 10mm test cube. I see the cube grow from a nominal 10mm with a single perimeter to 10.27mm with 3 perimeter layers. Based on E steps fine Tuning, the flow rate is still on the low side. Increasing the flow rate to get good solid infill would make the part grow more.
The extra 0.27 means that each perimeter is 0.135mm over size. One fix would be to provide an offset adjustment knob that shifted all the perimeters inward. Right now the only option folks have if a perimeter is to thick is to redraw the orginal cad.
I started exploring writing a script that would post process a gcode file to reduce the flow rate on the outermost perimeter to reduce the size. Slic3r can independently set the speed on external perimeters so I was going to key off of filament printed external speed and hack the flow rates. On the 10mm cube, there is a bug or strange interaction between small perimeters and external perimeters. It seems that small perimeters are always being used on the 10mm test cube and the external perimeter speed setting has no effect. I would have expected that the actual speed would be the min of external and small if the perimeter falls in to both catagories, but apparently the small setting overrides external. External speeds start to appear in larger test objects.
However it seems that once this is tuned, we are still missing a degree of freedom to adjust for the final wall thickness of the perimeter.
In the diagram below, we start with a 3 wide perimeter with minimum flow so the adjacent filaments are just touching. As the flow rate is increased to get better adjacent filament contact, the wall thickness grows:
-width of first filament increases
-2nd filament gets printed somewhat to the right of the nozzle centerline because it is pushing against the enlarged first filament and the 2nd filament is now larger.
-3rd/last outer filament again is pushed even farther from the center of the nozzle because it is pushing against the already offset filament 2. The third filament is larger because of the greater flow rate. And of course the right edge of filament 3 is unconstrained.
Experimentally with a 0.4mm nozzle, 0.4 infill, 0.4 perimeter extrusion width, 0.3mm layer height, 10mm test cube. I see the cube grow from a nominal 10mm with a single perimeter to 10.27mm with 3 perimeter layers. Based on E steps fine Tuning, the flow rate is still on the low side. Increasing the flow rate to get good solid infill would make the part grow more.
The extra 0.27 means that each perimeter is 0.135mm over size. One fix would be to provide an offset adjustment knob that shifted all the perimeters inward. Right now the only option folks have if a perimeter is to thick is to redraw the orginal cad.
I started exploring writing a script that would post process a gcode file to reduce the flow rate on the outermost perimeter to reduce the size. Slic3r can independently set the speed on external perimeters so I was going to key off of filament printed external speed and hack the flow rates. On the 10mm cube, there is a bug or strange interaction between small perimeters and external perimeters. It seems that small perimeters are always being used on the 10mm test cube and the external perimeter speed setting has no effect. I would have expected that the actual speed would be the min of external and small if the perimeter falls in to both catagories, but apparently the small setting overrides external. External speeds start to appear in larger test objects.
Sorry, only registered users may post in this forum.