Does Slic3r ever change your shape?

[Jimustanguitar]

I've been fighting with Slic3r all day, and I wondered if this is a frustration that the rest of you experience too. You can see my original shape here:

[img]http://i.imgur.com/SWuGWct.jpg[/img] [img]http://i.imgur.com/oS16JkY.jpg[/img]

And here's what Slic3r turned it into:

[img]http://i.imgur.com/eKiAa1V.jpg[/img] [img]http://i.imgur.com/a7NwNUN.jpg[/img]

Any ideas?

[dsnettleton]

I'll assume you've checked your infill settings and they're all kosher. Slic3r does indeed do funky slicing sometimes. I've found skeinforge to be a little esoteric for my taste, so I usually try an older version of slic3r when the current one isn't working so hot. First, though, you may want to try running your .stl through http://cloud.netfabb.com/. Even for proper models with no holes in them, this can work sometimes, as it reconfigures your vertices and triangles.

[Jimustanguitar]

I'm doing about .4 infill... What setting would effect this? Here are my Slic3r Settings: http://imgur.com/a/p6DRp



Also, some more messed up shapes

[img]http://i.imgur.com/eAfPJab.jpg[/img] [img]http://i.imgur.com/ZQ2uCvq.jpg[/img]

[Jimustanguitar]

I am using Sketchup.... Is it possible that something is being lost in translation with the STL converter plugin?

[dsnettleton]

The .stl file seems to be passable on the surface, since you're getting a good preview of it in Repetier before slicing. Are you getting any errors in your log?

[DaGroundZero]

I export the .stl from sketchup and run them thru netfab to look for holes. I do an automatic repair and export the fixed .stl if I need to. I've not noticed this yet.

[Jimustanguitar]

I'll have to check out netfab, thanks for the tip. I've been meaning to try this service too: http://willit3dprint.com/

[Polygonhell]

FWIW if your seeing the issue commonly it probably has to do with your workflow in the package in question, and you might want to try and figure out what it is your doing that is causing the issue.
I know a friend of mine was overlapping geometry to get the shape he wanted, so a model would look correct, but it makes the resulting output none manifold and what you get out of the slicer at that point is a complete crapshoot.
It could also be the normals are inverted in the model.
Slicers expect well formed models where well formed is a continuous mesh of the exterior, with no overlapping or vanishingly small polygons, the latter can be hard to avoid.