In the first image, there is a portion of a layer where the slicer has decided to put wall loops around that portion. In the second image, you can see the next layer where there is a void in the same location.
There doesn’t seem to be any good reason to put loops there. It just gets covered up anyway. Is there a way to get the slicer to not do this and just make that lower layer continuous?
First, thanks for posting clear screengrabs, this is enormously helpful in taking the guesswork out of the equation.
Can you show more of the model? I ask because it looks like there might be some internal structures that are above the areas of concern and these might be influencing the model.
However, I do see one item that can try out from the start.
If you look at the corners of the inner walls, this indicates that you are likely using standard wall generators. Consider using Arachne as this will “tuck” the filaments in tighter to the nooks and crannies.
If you want more feedback and care to upload the 3MF, then the members of the community might be able to directly inspect the model. What I would look for are the aforementioned structures.
BTW, I would be happy to upload the model, but it’s 54 megabytes.
There’s definitely something going on in the model. Can you show the model before it is sliced so that we can at least understand what the overall objective is?
Also, is this an STL or STEP and what was the model source? Was it a premade model you downloaded or something you have design control over?
It’s an STL. The STL is 39MB. It compresses to 5MB. It comes from a Python script I wrote to make lithophanes. The forum will not allow me to upload either stl or zip files, unfortunately. I’m attaching an image of what you’d see if you printed this and held it up to the light. (No, I’m not obsessed with Physics Girl. I printed Hank Green first. @theosib.bsky.social on Bluesky)
OK. That explains a lot. Thanks for clarifying that.
Unfortunately, there likely isn’t a solution that will appreciably change a model of this type. That’s because there are too many underlying geometry features and now that I understand what your objective is and the source, I’d say you probably won’t be able to completely get rid of those holes.
But fear not. There are things you can try.
I don’t want to fill the hole. The hole is fine. I just want the lower layer to be entirely rectilinear instead of having superfluous walls in it. This may improve a little bit how the light passes through the filament.
Ah… that makes sense. I wasn’t thinking in terms of light transmisivity.
Have you tried scaling the STL in the vertical Z axis? Even a 1% scaling, smaller or larger will have a moiré affect on how the algorithm lays out the mesh.
However, one thought just occurred to me and I tested it in the slicer. You can eliminate or reduce inner walls by reducing the wall loops to 1. Here’s what happens. Notice no more inner walls. But this still does not fix the geometry problem.
The layer heights are exactly tuned to yield the necessary translucency. The first layer is 0.1mm, and the rest are 0.04mm. I once tried setting the loops to 1, but they came off the bed on the first layer. The little outer borders just sit there too long before being attached to more filament that they have trouble staying attached to the bed.
I set the infill to 100% rectilinear, the top and bottom shell thicknesses to zero, and the wall loops to zero. This eliminated those internal walls. For some reason that I can’t figure out, there’s still a wall around the exterior of every layer.
I found an option on the Quality page for applying one wall to the top layer. Turning that off eliminated the walls entirely. The next thing I noticed is that some parts were sparse infill and some were solid infill, which had a different shape. I changed the “Minimum sparse infill threshold” setting to zero, and now everything is 100% sparse infill (which is set to rectilinear at 100%).
That’s awesome and thanks for sharing the solution here.
Others will benefit if they run into this issue.
