We have been grinding up our rafts, supports, and failed prints from normal Gigabot to print on Gigabot X. This is a GBX test print using regrind silver PLA from normal Gigabot prints. It is spiral vase mode (1 outer perimeter with no top, bottom or infill) cylinder that has some defects along the side. What caused the defects and how do we prevent them?
Irregular granule size, or inconsistent feeding into the extrusion screw. Prevent by screening / sorting regrind, or re-melting and extruding it into pellets. Increase size of flutes in screw to make sure that granules fill flutes consistently.
The above print was printed with PLA regrind that was sifted to be 3 - 4 mm in length (particles <3mm removed with a 3D printed sifter with 3mm diameter holes, and particles >4mm removed with a sifter with 4mm diameter holes). We were able to achieve a similar print with the same sifted particle size distribution, but without the surface defects. In fact, we printed some cylinders without any regrind sifting at all (so 0 to ~5.5mm), and we were able to get prints without the surface defects. So there's something else at play here >:)
Are the defects distributed over the entire cylinder or are they biased to a direction?
They also occur on the side of the cylinder opposite the side pictured.
Movement speed too high for the mass of the hot end with the pellet extruder? That doesn't make a ton of sense since it's a circular print and probably never got going that fast in those two directions anyhow.
Since it was printed in vase mode, the speed of the extruder at any one time should be the same throughout the print.
Loose Y axis belt
Really good guess, Stuart! I might have guessed that too if I hadn't been testing one specific variable, and the defects disappeared in my other trials of this print when I altered that variable.
Here are some ideas:
H1: Tool head gantry rigidity and/or runout in the linear guides allows the extruder tip to have play in the problematic axis (Y, is it?); possibly exacerbated by forces on the extruder head such as wire conduits or flex-hoses which remain anchored, or hopper whose mass causes a bending moment during acceleration, and whose mass varies during the duration of the print.
H2: If not a loose belt, then propose it could be belt-stretch.. especially if that axis's belt is working hard against a particular load such as those in H1, or if there is a bind in, for example, the linear guide(s) associated with the problematic axis.
H3: If cooling fan was used, and if the cooling air-flow is directional, hypothesis is that when direction of extrusion is parallel to direction of cooling air flow, the cooling rate of the just-deposited plastic is different than when cooling-air direction is orthogonal to the direction of deposition.
H3 is starting to touch on the issue... This PLA print was printed without any part cooling fans. I was able to get rid of the defects without adding part cooling by changing a different variable.
Another hint is that the bottom section of the print also has poor surface quality (see picture). It doesn't look exactly like the surface defects along the sides of the print, but it's related.
The answer may be simpler than you think.
Seems like the layers nearest the bed are influenced by bed heating moreso than subsequent/higher layers. If poor quality occurs near the bed then it infers we'd need to reduce the rate of depositing thermal energy into the part through a number of means including reduced extruder temperature, increased cooling, reduced extrusion speed (thus increased dwell-time from layer-to-layer), shorter layer thickness or smaller nozzle diameter (thus slower volumetric flow rate of material; less BTU/hr deposited onto the part).
Are any of these how you solved it?
Congratulations, extruder temperature is the answer! We will send you a coupon code to redeem your prize of a 5lb spool of PLA.
I was trying to optimize the heat zone temperatures (there are 3 heat zones on Gigabot X), and the highest temperature trials produced the surface artifacts in the picture. PLA especially can warp and ooze at higher temperatures.
Gigabot X also has a vertical compression screw to extrude the plastic, and sometimes that screw falls if there isn't enough upward pressure generated from extrusion to hold it up against gravity. This can happen when the temperature is way too high, and the plastic's viscosity is really low inside the extruder. That could explain why the surface defects look like over extrusion.
