Would it be possible to design an anti-warping infill that makes it easier to print large parts with ABS by reducing internal horizontal forces? I suggested this on the slic3r issue tracker, but it’s really just pie in the sky. Any thoughts?
https://github.com/alexrj/Slic3r/issues/1462
I found the thickness of the solid base layers to have a significant impact on part warping. The thicker the solid part, the greater the warping. It might help to have a gradual transition from 100% infill to the internal infill value…
I’m finding that making the floor very thin on large cross section parts helps. Im talking 2 layers thin, with 15-20% infill over it. The infill doesn’t dream to pull much, it’s the solid plastic. In kisslicer I’m still trying to figure out the settings. There’s a setting that controls perimeters, then there’s a global thickness in mm. I changed that to .04 for 2 layers, but what would happen if my perimeter’s was set to 4. Would it give me 2 floors and ceilings, and 4 perimeters?
Don’t you think it would better to go straight from 100% to normal infill. I honestly think that 100% to 0% would be the limit of no horizontal shrinkage forces, and a solid part would be the limit of max horizontal shrinkage. Thoughts?
There is no fill that would in the general case prevent an arbitrary shape from warping. If your perimeters have long straight runs, there will be warping and even delamination if the part is big enough. If the fill and perimeters could be corrugated, the bends could absorb the shrink that causes the warping.
Theoretically you should adjust the extrusion rate only at infill. (So, more plastic to keep the “pull” tension away)
Another way is to print the infill slightly bigger …
I’ve experienced less warping since switching from rectilinear to honeycomb in Slic3r; haven’t done any extensive A/B testing to prove it, but it seems to make a difference.
@Jason_Gullickson that makes sense. The zigzag allows stain relief for the shrinkage
Jason, that sort of makes sense, as the tension in a layer filled with a honeycomb pattern might be relieved compared to when the infill is made up of long straight lines that shrink as they cool. But it’d be interesting to improve even further on this.
As I mentioned in the issue, it appears to be possible to design parts for less warping, and it’d be nice if something with a similar effect could be done with clever infill, without redesigning the part. http://hackaday.com/2011/11/15/a-technique-to-avoid-warping-on-large-3d-prints/
Yeah @Ketil_Froyn I really like the idea of having the slicer dynamically tweak the infil density and pattern to reduce internal stress without altering the overall model. Would be some fun programming 