Stl errors noted in Prusa , solid inspectors says everything shiny. Cleanp3 says fied errors

When importing a STL file created by Sketchup Make 2017 into Prusa 3D Printer software version 2.7.4, Prusa fixes several “errors” automatically, but identifies 0pen edge errors that Solid Inspector 2 does not identify. Prusa can fix these errors using Windows software with just a click, but I’d like to fix the skp file where necessary. Since my last post, I’ve redone some of the model to cleanup issues, but freely admit there could be better ways to do things.

I ran Cleanup3 and solid inspector, back and forth, but I feel I keep chasing my tail. I fix one error and others popup.
How can I approach identifying and fixing the errors? The skp file is attached.

cap214.skp (1.3 MB)

I looked at the skp. Each profile has very many complex curves with many segments. Instead of trying to use the STL created file, it might be easier to model it in SketchUp in a simplified way. Don’t use so many curve segments. As it is now, it’s very complex and slow. I used the Dave Method but still could not identify the errors.

On load in 2024, SketchUp said if fixed a couple of problems. I also ran fixit 101 extension, and it too fixed a few things. Other than the excessive level of detail in most surfaces (as already noted by @Royce), there are two things I noticed:

  • A huge number of the edges in the model are perilously close to SketchUp’s lower tolerance for length, putting you at peril of strange errors when editing the model. I would recommend scaling it up by a factor of 1000 so that you can treat meters as if the were millimeters. An stl file never conveys units anyway, so you can export from SketchUp as meters and import into Prusa as mm.
  • You have multiple nested components and groups. Although Prusa may deal correctly with that, SketchUp and Solid Inspector 2 will not report objects containing nests as being solids. You might consider whether you really need TOOTH_EDGE to be a separate nested object inside tooth_ring, and THREAD_SUPPORT nested inside base2. The former seems unnecessary. The latter I can’t determine because I don’t know whether you planned to 3D print them separately and assemble later.
1 Like

Thanks for the input. I am planning to 3d print this eventually. Attached is a picture of what it should look like. The picture is a 1929 radiator cap of which I only know of two in existence world-wide. This cap will be used for a restoration car I am working on. Also, collectors may want a copy. so I’m trying to be as close to the original as possible. I started with a blueprint that only shows a top, side, and cross-section.

I also did scale it up when working, but scaled it down when exporting to Prusa so that it fits on the printer bed.

You don’t need to do that. As I wrote earlier, stl contains no units, just numbers. Prusa only knows what you tell it (or what it assumes) about what those numbers mean.

1 Like

I get it now. Thanks. I also didn’t think about printing it in parts and then assembling later. That might solve the “largeness” of the model issues. I’m just learning about the 3D printing process as well as learning Sketchup as I go.

Really appreciate the help and comments!


One additional note. I exploded all the components and groups and tried an export to STL just to see what happens. After 10 minutes, I cancelled Sketchup because it was just eating CPU. This sounds to me like a 2017 bug, but since the software is not current, I don’t think there’s anything to be done and I don’t use Sketchup enough to warrant upgrading. This 3D attempt is probably a one-shot deal for me, as a hobbyist.

Also, after running Cleanup, I got a series of messages saying “The plane equation for CFace (xxxxx) is not valid - fixed” . I don’t know what this means, but I’m going to punt and just let Prusa fix the STL automatically using Windows 3D. I don’t like to do this, because in my experience, these things always come back at the worst time to bite me!. However, I’ve done what I can to find out the issue.

Appreciate the help and comments. .

Thanks again.