Vase mode issues again. Glitches from nothing

[nerys]

OK now its glitching for no obvious reason at all. literally in the middle of a clean extrusion it just goes huh I am going to jump 1inch over here "just because" no reason visible at all in the model. slices fine in prusa slicer netfabb and prusa slicer say its clean no errors.

YES this is vase mode compatible. JUST LOOK at where the red travel moves and retractions are. there is nothing THEIR for them to exist. just trust me its vase mode compatible. its literally "what I do"

for some reason S3D is glitching and S3D is creating the glitches. Want to know how I know? Load my factor file. slice it. TAKE NOTE of where the glitches are. NOW rotate the model 90' and WOW all the glitches "move" a spot that was "fine" a second ago is suddenly "not fine" now. this means its not the model. its S3D pathing "creating" the glitch.

I wish their was a way to FORCE vase mode compliance. use my extrusion width. no adaptive anything. no smart anything. just "do it" detect a copllission? IGNORE IT ramming speed. shove through its an illusion :-) I suspect its trying to do something intelligently that would normally work fine on a normal model but its "breaking" my model for some reason.

Again. prusa slicer. clean. netfabb. clean. and unlike normal times when this happens (usually when a model is too complex in tinkercad and you do a MERGE it will "generate artifacts" but I don't SEE any of those in this model. its literally glitching in the middle of an arrow straight clean line......

Any ideas at all whats causing it?

[DrD]

Using the provided .factory file, the first thing that I'd noticed is that you have the nozzle wipe feature turned on.

Generally speaking, I don't use "wipe" when printing in vase mode as this mode should be a continuous line of
plastic being extruded without interruption.

Turning off "wipe" removes the unwanted toolhead movements that you're seeing. Do note however that there
will be a small number of travel movements (2-3 movements) across the inside perimeter.

Still better than what you were seeing prior to this.

D.

[parallyze]

YES this is vase mode compatible. JUST LOOK at where the red travel moves and retractions are. there is nothing THEIR for them to exist. just trust me its vase mode compatible.

Still you might want to check the model file again. Some parts aren't properly seperated from the
outer shell, causing additional loops.

In other places there's too much seperation, causing closed perimeter loops:

Also this model reports as 2 shells...

[parallyze]

Seems like half of the normals got flipped at some point. Maybe this happened when
using boolean modifiers to add the top/bottom threads...

Some of the faces have their normals properly to the outside of the model, some point inwards. This completely
irritates the slicer, this can also cause funny things like supports inside an object instead below....

So far I couldn't get ANY slicer to slice this in vase mode because of the flipped normals. I'd check the model files
from before the thread/merge processes if their normals are still aligned correctly to the outside of the model.

Edit:
Trying to render this in blender/3ds max or anything shows the problem a bit more clearly:

Due to the flipped normals programs have trouble defining the out/inside of this object, this is why the render
let's you see through the back wall there.

Another Edit:
Could somebody running S3D v5 export the STL file and attach it? Not running v5 myself and I'm not sure if
exporting using v4 might change things on my side, loading the factory file doesn't really work...

[DrD]

Seems like half of the normals got flipped at some point. Maybe this happened when
using boolean modifiers to add the top/bottom threads...

vasenormals.PNG

Some of the faces have their normals properly to the outside of the model, some point inwards. This completely
irritates the slicer, this can also cause funny things like supports inside an object instead below....

So far I couldn't get ANY slicer to slice this in vase mode because of the flipped normals. I'd check the model files
from before the thread/merge processes if their normals are still aligned correctly to the outside of the model.

Edit:
Trying to render this in blender/3ds max or anything shows the problem a bit more clearly:

vasenormals2.jpg

Due to the flipped normals programs have trouble defining the out/inside of this object, this is why the render
let's you see through the back wall there.

Another Edit:
Could somebody running S3D v5 export the STL file and attach it? Not running v5 myself and I'm not sure if
exporting using v4 might change things on my side, loading the factory file doesn't really work...

I've exported the STL model from the original .factory file in this post.

See attached.

D.

Tremendous Hillar-Bigery(1).stl

(1.56 MiB) Downloaded 366 times

[parallyze]

I've exported the STL model from the original .factory file in this post.

Thanks!

Normals seem more uniform with this one:

It does report as 2 shells, still. And while both of them can be seperated without problems they both
don't work properly in vase mode. Which is a bit strange.

I did look at the model in cross section and as far as I can tell it should work - there's one cut going
through the model from the center to the outermost perimeter, otherwise the cuts only create the inner
"fins". Some of the cuts are ignored when looking at the slicer preview - but this also happens in PS
and other slicers, this is what irritates me the most... also I'm still on S3D 4.1.2, so it's not only related
to v5 in this case.

Depending on how the model was created I'd check each part on its own before adding the threads
and merging top/bottom.

Interestingly just cutting out a part of the center doesn't work in vase mode either...

Pretty conviced there is something wrong in the model - but it may be as small as one or two faces with
flipped normals somewhere hidden in the masses...

Again. prusa slicer. clean.

Colonel Kira, do you have access to all the previous steps made to the model, could you test if the main body
without the threads on top does slice as expected?

And may I ask for your 3mf PS file that did work without problems??

[parallyze]

There's actually some intersections in the model... don't know why NetFab et al don't get triggered properly by
this, but this would explain why I can't get this sliced in vase mode in any slicer...

cross section in S3D, at some points at certain angles tiny lines pop out from the surrounding:

And a closer look at the mesh, inside:

...and outside:

(note the white lines which "disappear" shortly after the vertices, they're "overlapping" with another face, creating
a tiny downward crease and breaking vase mode compatibility)

Would love to see how this was sliced in vase mode in PS...

[parallyze]

...stupid 3 attachement limit ^^

In PS I can get this to slice either like this:

(slicing gap radius >= 0.01mm)

or like this:

(slicing gap radius 0mm)

So I am really curious how to get this one sliced properly in PS, as far as I can tell it doesn't
work anywhere using this very STL file... ^^

[parallyze]

This one might be slightly off topic, but I know from some threads of the past that many people are struggling a bit
to understand what's going on, usually they see an object with inner/outer perimeters and conclude it is not compatible
to vase mode.

This is the basic idea behind it:

You can not print a donut shape in vase mode. You have to switch between the inner and outer perimeter, therefore
this is pretty easy to understand.

Now let's cut a tiny (0.01mm) bit away from the donut, from the center to the outside:

...now the perimeters are connected, so there's no need for a retract. And because the gap of 0.01mm is almost
guaranteed to fuse on most FDM printers you can come away with this.

Now let's add another cut from the center of the donut, this time not reaching the outside:

This way you can create "fins" adding rigidity to an object.

And obviously you're not limited to two dimensional features, using angular faces you can also include details
without breaking vase mode:

Admitted, I don't see the point of the modifications on this very object - the cut will also produce some kind
of "seam" in the end (and retracting wouldn't need to cross perimeters). But depending on the material or
other unknown to us requirements retracting might simply not be an option to nerys here.

Hope this helps people reading this thread to understand what the fuzz is about on this very model

[DrD]

Thanks, parallyze for the in-depth info... you totally blew my mind on this one...

Now, my head hurts... lol

D.