Plate weirdly eaten up (with CR-30 profile)

[Xowap]

Hi all,

I’m getting the weirdest issue trying to slice my piece. I’ve put side to side the STLs and the preview:

As you can see, the “base” of this print is getting completely eaten up in a weird weird way.

Here’s the OpenSCAD project I’m using GitHub - Xowap/cr30-calibration: CR-30-specific calibration shapes

Thanks for help

[stewart]

Can you drop the STL files here? I suspect they’re not being generated manifold.

[Xowap]

There you go: temp-bridges.stl - Google Drive

I have no idea what it means or how to fix it

[stewart]

It means that the object is constructed by packing a lot of objects against each other creating interior shared faces. What it should be doing is unioning these objects such that there is one “water-tight” body with a single interior void. Having a body with internal faces cutting through is an invalid / non-manifold situation. Because there are many ways to interpret this condition, each slicer may or may not process it properly.

[Xowap]

I suspected something like this.

Do you know how I could get OpenSCAD to conform? Maybe ask it to bleed epsilon into neighboring shapess? I thought the union operator would take care of detecting such things, but I guess it’s not that obvious.

[stewart]

I haven’t used OpenSCAD, unfortunately.

[Xowap]

Thanks anyways, that was very helpful!

I’ll comment if I find any solution

[patrice]

Hi Xowap,
I use OpenSCAD a lot, can you attach your scad file so I can have a look in it ?
Patrice

[Xowap]

Sure!

Here you go: https://github.com/Xowap/cr30-calibration/blob/develop/src/temp-bridges.scad

Thanks

[mcdanlj]

Try using the development version of OpenSCAD, then turn on the Manifold kernel: Edit → Preferences → Features → manifold

Then if it doesn’t generate manifold objects, I believe that the OpenSCAD developers would appreciate a bug report including the .scad file.

[Xowap]

Thanks for the tip!

Still doesn’t work however, I’ll forward this upstream