Cura 14.01 strange slicer behavior

[shurik 87]

Hi there,

Just wanted to prepare a simple stretchlet for print:

And this is how it gets sliced:

I played with different combinations of "Fix horrible" settings, most of them had no effect. One of the combinations printed everything as a solid block with infill everywhere.

This weird slicer behavior was aslo experienced on some other models but there it was neutralized by the "Fix horrible" play. Here it can't be done.

Can anyone help please? :ugeek:

 

[achillx 0]

Haha.. 1 minute time estimation!!

Can you share the file so I can see what it does in my computer?

 

[gr5 1,707]

Cura prints solid things. Not things infinitely thin. The walls of this have no thickness. Not even a millionth of an inch. So from that prespective Cura did a close approximation. Damn close. Within a billionth of an inch accuracy, lol!

Try this model:

http://www.thingiverse.com/thing:18661/#instructions

Set infill to 0%, uncheck fill for top and bottom. That should be it.

I don't recommend spiralize - don't check that for this as spiralize tries to make the bottom water tight by adding a fillet.

 

[shurik 87]

Yeah, one day we will reach the point where 3D printing will be done in just one minute...

Here's the link: http://sdrv.ms/1aUNs5M. It was taken from here: http://www.thingiverse.com/thing:13505/#files.

gr5, you beat me by a few minutes! :-) Well, just for the challenge - what should be done with the one I tried to make?

 

[gr5 1,707]

If you read the instructions for that particular thing in thingiverse it recommends you download "stretchlet1" and turn off infill and top and bottom.

 

[shurik 87]

gr5 - thanks! Was able to slice the stretchlet1 properly, but only using the "spiralize" - no other way.

 

[achillx 0]

@gr5

I looked at the file that Shurik tries to print in Rhino and it's not infinity thin.. It has a thickness of 0.4mm.

So in theory if you set the shell thickness to 0.4 you'll get a one pass for each layer but... nope..

I did a simple model to test Cura and I get wrong (for my perception) results.

https://www.dropbox.com/s/oq3mh52oa3rerw4/Cura_shell_test.stl

The first (small one) has a shell thickness of 0.4mm

The second 0.8mm

and the third 0.84mm

So with Cura shell thickness set to 0.4 the first one is not get sliced.

As for the other two it seems that 0.4 and 0.8 Cura shell thickness produce the same toolpath.

In theory if I set the thickness to 0.4 I'll have a 0.2 offset from the perimeter of the mesh and if I set it to 0.8 I'll have a 0.4 offset which will result in a single line in the middle of the slice.

 

[gr5 1,707]

Don't confuse nozzle width with shell thickness.

Nozzle width sets the offset from the edge - not shell thickness.

If you set shell thickness to .8 (double nozzle thickness) it tries to do two passes on the *outer* surfaces. But if your part is less than 1.6mm it can't fit 2 complete passes.

 

[shurik 87]

Well, I got stretchlet1 but it didn't print well - the inner edges are combined together, so it cannot stretch. Any other ideas perhaps?

 

[ian 32]

idea might be to print a little cooler. i tried printing this with 240 and the edges welded together.. so no go..

also make sure your bed level is really nice and flat... a good start to this model really really helps !

Ian :-)

 

[achillx 0]

Don't confuse nozzle width with shell thickness.

Nozzle width sets the offset from the edge - not shell thickness.

If you set shell thickness to .8 (double nozzle thickness) it tries to do two passes on the *outer* surfaces. But if your part is less than 1.6mm it can't fit 2 complete passes.

 

You are right..my mistake.

Although if you have an object with 0.4mm thickness and set the Cura shell thickness to 0.4mm it doesn't slice it. if you set it to 0.399 it does.

I know that probably the two toolpath lines overlapping and cura somehow does this but it shouldn't.

In every code you have some exceptions, that's in my opinion one of those exceptions. A proposal is to merge overlapping toolpath lines.

Also..I know it's a NURBS to mesh conversion problem take a look at the images bellow.

All objects are in pairs..the left one has 0.4mm thickness and the right one 0.41mm.

You can also see the issue I described in the beginning..

 

0.4mm

 

0.399mm

 

[gr5 1,707]

Well, I got stretchlet1 but it didn't print well - the inner edges are combined together, so it cannot stretch. Any other ideas perhaps?

 

I'm not a big fan of spiralize. Did you tell it shell thickness .4 or .8? In your screenshot it is .8. If you did non-spiralize, then it makes 2 passes and on the second pass it will connect the inner edges. So that's a problem.

In spiralize mode it guarantees only one pass but you ask for .8mm so it overextrudes 2X so that you get a .8mm wide bead out of a .4mm nozzle (it can do that because there is a flat area on the tip of the nozzle).

This *also* connects the inner edges. So in other words it is important to do .4mm shell either way.

 

[shurik 87]

gr5 - you were 100% correct. My shell was default 0.8mm. Once lowered to the recommended 0.4, everything went beautifully fine. Thank you!