Filamentlehre d181|17
As the Filaments diameter and roundness is one of the major error source in FDM-printing i thougt it would be nice to have a filamentgauge. While I still think the approach is not too bad, it turns out that the precision is far from what we would need - even if made from steel you would only detect an error of 0.1 but not 0.01mm (or this become 10× bigger). An optical inspection (2×macro cameras) would probably be a better way but is expensive and complex. Hence a pure lever system with hall sensor and magnets is not working, at least not 3-D printed — however as it looks nice i am still posting my failed approach…
(you may notice that some blades are without bevel - these need to be changed)
It’s definitely pretty even if it has turned out not to work well.
thx, the lever amplify only 1:10 and as the printing itself has an inaccuracy of 0.1 you will only detect very strong errors - with the quality of filamens today this is of scale - and a 0.1 error in diameter would mean a 10% error in flow - which is why this may work as a grasper but not as a gauge.
Really neat even if it didn’t work out. It reminds me of Nikola Tesla’s tower.
Yes, didn’t saw that but you are right!
And if no filament is in the legs are more spreaded like it.
http://thinkorthwim.com/wp-content/uploads/2007/01/tesla-wardenclyffe-tower-shot.jpg
I like that idea a lot actually. To bad it has only 0.1mm tolerance
@Matthew_Del_Rosso if build precisly (steel) it could work but it will be hard to see that the distance between opposed arms is changing only 0.1mm for 0.01mm diameter change. But with a transparent screen with a circles on it you could check for roundness. A slight modification is needed - the filament need to go throug another hole in front as the filaments are rolled hence curved and need to be centered and straightend. Also the screen need to be mounted on a stable tube.
“As the Filaments diameter and roundness is one of the major error source in FDM-printing”
Not sure where you get this knowledge from. But in the last 3 years, diameter and roundness have caused virtually no problems for us.
Slip due to different material hardness is a much bigger factor.
@Daid_Braam if you only have high quality filaments - sure … Else it is simple math load your printer with a 1.6∅ filament and see what happens.
But i wonder about your problems - I never had slipping filament - not with TPU, nor Flex, PET, PC, or POM, not with TEC not even with BDP - you should change the extruder gear if you have those issues.
I’m using 2.85mm filament. But slip isn’t always visible, we are actually measuring now with a high resolution encoder on the filament, which shows slip of the filament depending on the extrusion speed.
well you have much bigger forces to transmit with those diameter, more if using a small nozzle. If you let the filament run curved on a barrel/wheel where it is pressed with 3 bearings and the wheel has a groove or toothed - and your stepper is attached to that wheel not the filament … you will have much more contact surface to transmit the power. Also this explain why you didn’t have diameter issues … the area is exponetial to the diameter so the same error causes much less issues on a bigger diameter.
Is there a way to “calibrate” the filament before going into the extruder? I was thinking on something like a drawing die, to calibrate roundness and diameter…
For sure, but this would cause enormous waste if you strip down your filament permanently. (if you preheat it and reshape it that wouldn’t be easy.)
@Ulrich_Baer i understand, but i was thinking about reduktion a very small amount. for example my PLA filament is in moment 1,71-1,73mm (also not very “round”) - for that i take a drawing die about 1,65mm (polished and hardend metall) to get a round an calibrated shape. hole process happen “cold” after unwinding from the spool, and before going to the extruder…I was thinking to use something like a belt caterpillar after the die to draw the filament trough the die
(sry for my bad english 
I think it would be easier and cheaper to find a different filament source than to re-draw oval material to a more consistent size and shape.
@Karl_Schlacher lieber in deutsch? G+ übersetzt das ja so halbwegs… Wie gesagt das geht auf jeden Fall nur ist eben bei einem Kreis grade aussen am meisten Material. Ich habe das grade mal durchgerechnet und du würdest im Durchschnitt 8% Material verlieren und das obwohl der Unterschied im Durchmesser nur 4% sind. Ich würde empfehlen so eine art rotierenden Anspitzer zu verwenden.
Deutsch passt besser…
I think I understand now…BUT…I believe you will not loose any material . The material goes in the length, and you have a well calibrated filament. I observed this in a actual project at my work, where we calibrate a aluminum tube after welding.
Also the force to make something like this should not be so much…Rp0,2 is about 30N/mm² when you have solid PLA, elongation at break 3-6%…You have to overcame the Rp0,2 during the drawing process to get a “cold solidifcation” in the filament. Example (1,72mm²*pi)/4=2,32mm² * 30N/mm² = 69,60N = about 7kg…
Or i understand something wrong…
Ok if you speaking of reshaping it by forming/ plastic deformation - You wouldn’t loose any material. You should use a warm air preheating (~50°C) as PLA is very brittle and probably your draw plates hole would be a little bigger than your filaments target diameter as it get streched a little further. Sounds a nice project! The big advantage would be that all different filaments would get a uniqe shape for your printing - so also if this process wouldn’t be perfect you will have a constant filament to work with. If using different materials, they may react differntly to the treatment.
I think you would need to replace the hall sensors with inductive ones.
https://de.wikipedia.org/wiki/Hall-Sensor … and i would need 6-12 which should not interfere much with each other . Can you please elaborate a bit further why inductive sensors would be the better choice?
