Unbundling Vase Mode

[neobobkrause]

@mroek and @dorsai3d: Thank you both for your input. Thanks especially to @mroek. You're absolutely right. It's all so obvious how to generate multi-process jobs where a portion of the print has just a single shell. I'm going to play with the setting you suggest right away. However, the big practical downside of these single-walled prints is that the seams at every level are less likely to be watertight, which some applications require -- like a real-life vase with water in it to keep flower alive.

@dorsai3d: I want to apologize if I've offended you. That wasn't my intent. I'm just trying to have an open dialogue. But I might suggest that you not use the phrase 'ugly hack' to refer to somebody else's suggestion if you're going to be offended by somebody using the same phrase to refer to your ideas.

Though my sense of urgency for the 'unbundled vase mode' features has lessened based on @mroaek's input, (I don't actually need a watertight print.) I do feel the need to responding to @dorsai3d's suggestion that even a wall that's printed just a single shell thick is an object that has volume. I'll begin by agreeing with you. A manifold mesh like this is a volume having a thickness of zero -- if we are rounding our volume calculation to increments of the Extrusion Width.

Now let me disagree with @Dorsai3d. I suggest that a shape only has MEANINGFUL volume when the mesh is manifold and where the thickness is great enough to infill. A slicer can never infill a wall having a thickness of a single shell. Though I appreciate that @Dorsai3d and I aren't going to agree on this point, I see appreciable functional value in a slicer properly handling non-manifold meshes the same way it would process a zero volume manifold mesh. I won't bother justifying this opinion as I feel I've already explained why in earlier postings.

Thanks again,

- Bob

[dorsai3d]

@dorsai3d: I want to apologize if I've offended you. That wasn't my intent. I'm just trying to have an open dialogue. But I might suggest that you not use the phrase 'ugly hack' to refer to somebody else's suggestion if you're going to be offended by somebody using the same phrase to refer to your ideas.

I'd argue that it is inherently ugly; especially given backface culling. For example, this is supposed to be a symmetrical section of fuselage:

@dorsai3d's suggestion that even a wall that's printed just a single shell thick is an object that has volume. I'll begin by agreeing with you. A manifold mesh like this is a volume having a thickness of zero -- if we are rounding our volume calculation to increments of the Extrusion Width.

I... It's manifold, meaning it has a nonzero volume. That's part of the definition of the word! We're not rounding our volume to zero if it's extrusion width, we're looking to extrude the target volume described by the manifold model.

Now let me disagree with @Dorsai3d. I suggest that a shape only has MEANINGFUL volume when the mesh is manifold and where the thickness is great enough to infill. A slicer can never infill a wall having a thickness of a single shell..

Manifold. You keep using this word, but I don't think you're grasping what it means... Infill has nothing to do with the manifold-ness of a model.

http://blender.stackexchange.com/questi ... d-geometry
http://www.shapeways.com/tutorials/fixi ... old-models
http://3dprintingninja.blogspot.com/201 ... tmare.html

The goal is to create a physical model that matches the volume occupied by the 3D model created to describe the print. Infill is a convenience to provide structure while reducing print time, material usage, and/or weight.

Though I appreciate that @Dorsai3d and I aren't going to agree on this point, I see appreciable functional value in a slicer properly handling non-manifold meshes the same way it would process a zero volume manifold mesh.

Repeat after me: There is no such thing as a zero volume manifold mesh.

[neobobkrause]

I know that I'm frustrated that it's not currently possible to produce walls that are only 1 shell thick -- except by using Vase Mode.

But this is not true. If you set perimeters to 1 shell, you get exactly 1 shell. Without using vase mode. But it might require you to use solids, and not surfaces...

Here's the standard 20 mm box sliced with no infill and 1 perimeter:

box1peri.png

@Mroek: I was very excited to see your post the other day that contained a screenshot of non-Vase S3D gcode with a single wall. However I've been unable to generate just such a beast myself. Could I ask that you attach a factory file so that I can see your full process settings?

Thanks,

- Bob

[mroek]

Sure. I did nothing other than setting the perimeters to 1, solid top layers to zero and infill to zero. Just keep in mind that the model is a solid cube.

[neobobkrause]

@mroek: That's what I thought, but wanted to make sure.

The reason your print has walls that are only 1 shell thick is that the walls are perfectly straight. Try to produce gcode for a single-walled version of the this STL without using Vase Mode. (External link because it's larger than 5mb.)

Vase Mode allows me to print any shape, no matter how thick the walls are, no matter the external twists and turns. I like all that. And I like that the it's done without per layer seams. I'd be even happier (read: satisfied. read: less frustrated. read: capable of making my design intent reality) if the various features of Vase Mode were unbundled and if Vase Mode jobs could be included in multi-process jobs, with the features, defaults and restrictions I've described.

- Bob

[mroek]

It seems to work reasonably well, but if using bottom layers there are some extra printed lines near the bottom:

[neobobkrause]

@mroek: The takeaway is that S3D DOESN'T honor your shell count setting rigorously except in Vase Mode. That's what everybody is complaining about..

There is tremendous value in unbundling Vase Mode.

- Bob

[dorsai3d]

The reason it doesn't honor the perimeter count in this case is because your perimeter isn't wide enough to rest on the perimeter below due to the overhang, so the software is smart enough to add in a bit of extra plastic in an attempt to hold the whole thing together, rather than just squirting it out over open space. If you don't want it to do this, I think you can disable bottom solid layers and it might go away.

[neobobkrause]

I'd like to see any configuration that results in exactly the number of requested shells even when the perimeter has twists and turns. The Vase Mode algorithm consistently produces shells that rigorously match the indicated shell count. Vase Mode does many things right -- for those situations where you want those features. But those capabilities are currently locked up in this restrictive construct called Vase Mode. Vase Mode needs to be unbundled.

- Bob

[Lagbert]

I agree in the vase mode and "standard mode" algorithms should not be mutually exclusive.

There is no reason the X Y motion of the print head should stop during a layer change in "standard mode." We currently have "zipper" seams where the print head stops, proceeds to the next level, and then starts moving again. Obviously, prints with multiple islands won't benefit much from this, but many models would benefit from a pseudo vase mode.