Merged inner wall or wrong calculation of thickness  

That looks right for Arachne.  The default extrusion width with a 0.4mm nozzle is 0.45mm so 2.25mm thick walls with the Classic perimeter generator should give you what you're looking for.

Cheerio,

RE: Merged inner wall or wrong calculation of thickness

Doesn't matter if I use Arachne or Classic. The final result is the same.

I tried 2.25mm thick, like you suggested. Classic made a thin line in the middle of six walls, while Arachne appears to be correct.

But I still don't understand why Prusa can't do 5 walls. Classic or not. 5 x 0,4 = 2mm.

RE: Merged inner wall or wrong calculation of thickness

What is your extrusion width?

whilst your nozzle may be 0.4mm diameter, what is programmed to come out of it, may not be 0.4mm diameter.

doing maths with the wrong dimensions will give you 'correct' wrong answers

this solid block, with 5 perimeters and no infill gives the sort of geometery you are looking for but the wall thickness will be 5 x extrusion width not n5 X 0.4mm.

using this method, if the important feature is 2mm wall thickness, you can increase the extrusion width to 0.5mm and use 4 perimeters

regards Joan

RE: Merged inner wall or wrong calculation of thickness

Thanks Joan!

Your example is different from what I'm trying to achieve. I can do a solid block and get the same results as you, but if my object is 2mm thick, Prusa can't do what I want and that's what I'm trying to understand.

This is a example of the object

I'm using 0.4 extrusion width and 3 perimeters

Slicing with Arachne gets the change of flow and merged inner walls

With Classic it squeezes two walls

And in Cura I can get the expected result using the same configuration

RE: Merged inner wall or wrong calculation of thickness

If I change the parameters using @diem information, I get the exact same results.

Using 0.45 extrusion width and the same 3 perimeters. Object is 2.25mm thick, which is 5 x 0.45

Arachne - thicker inner wall and change of flow

Classic - two inner walls

RE: Merged inner wall or wrong calculation of thickness

Perimeters are not laid down side by said at their extrusion width.  If they were then they would not adhere to each other and parts would be weak.  They are laid down with a slight overlap.  The math is a bit complex as the cross section is oval and there used to be a page on it on the PS github wiki but it seems to have gone walkabout.  There is a similar one on the Slic3r project https://manual.slic3r.org/advanced/flow-math   

Classic with collapse perimeters which tends to lead to a bit of over extrusion, Arachne will increase flow.  Sometimes this doesnt work too well and you can get slight gaps, especially in corners. 

Also the amount of overlap is not configurable in PS.  Some other slicers do allow you to control that.  At least to a certain extent.  Super Slicer for example has its various experimental Perimeter overlap settings. 

RE: Merged inner wall or wrong calculation of thickness

@caiokao
try 3 perimeters, Classic, External perimeters 0,49 and Perimeters 0,48 on your 2,25 wall

When doing thin solid walls there is a need to experiment with line width to get there

@caiokao
Object is 2.25mm thick, which is 5 x 0.45

@joantabb
if the important feature is 2mm wall thickness, you can increase the extrusion width to 0.5mm and use 4 perimeters

@diem
The default extrusion width with a 0.4mm nozzle is 0.45mm so 2.25mm thick walls with the Classic perimeter generator should give you what you're looking for.

It is not that easy, filament lines are squished together, not laying side by side like rebars.
VERY good information -> https://manual.slic3r.org/advanced/flow-math
From the above document:
"Path spacing is thus: spacing = extrusion_width - layer_height * (1 - PI/4)"

RE: Merged inner wall or wrong calculation of thickness

You seem to be trying to defy mathematics to prove a point
A 0.4mm nozzle will not satisfactorily print a o.4mm  line with a 0.4mm nozzle  and bond to adjacent filament lines, there will be insufficient inter line bonding
asking slicer to use 3 perimeters with .4mm extrusion width  in Prusa slicer, will not produce 5 lines of filament on a sold item, Prusa slicer will use three [erimeters on each side and adjust extrusion width to match. 
Prusa slicer will not accept 2.5 perimeters to force 5 perimeters into 2mm wall thickness. 

Prusa slicer WILL accept 2 perimeters with 0.5mm extrusion width, and print solid

see above 2x2x0.5mm = 2mm...

prusa slicer will accept 3 perimeters of 0.333mm with a 0.25mm nozzle

2x3x0.333mm, near enough =2mm

RE: Merged inner wall or wrong calculation of thickness

@caiokao
try 3 perimeters, Classic, External perimeters 0,49 and Perimeters 0,48 on your 2,25 wall

When doing thin solid walls there is a need to experiment with line width to get there

That gave me the exact result I'm looking for. Five "walls" in total and no percetable change in flow rate.

Now I'm wondering why, and I'll check the link you posted. Thanks!

RE: Merged inner wall or wrong calculation of thickness

@joantabb

Prusa slicer WILL accept 2 perimeters with 0.5mm extrusion width, and print solid
-IMAGE-
see above 2x2x0.5mm = 2mm...

Yes, PS will accept it, WITH Arachne. Outer perimeter is 0,5mm, inner perimeter is 0,564
(2 *0,5) + (2 *0,564 ) = 2,128 Squeeze them together and the wall will be 2mm

prusa slicer will accept 3 perimeters of 0.333mm with a 0.25mm nozzle
-IMAGE-
2x3x0.333mm, near enough =2mm

Yes, PS will accept it, WITH Arachne. Four outer perimeter is 0,333mm, two inner perimeter is 0,420 (0,15 layer height)
(4 x0,333) + (2 x0,420) = 2,172 Squeeze them together and the wall will be 2mm

@caiokao
At the bottom of "Print Settings" -> "Layers and perimeters" -> "Vertical shells" -box you have "Recommended object thin wall thickness" with current layer height and perimeter width settings. 
IF it is possible to design with these numbers in mind I recommend you do it.
I try to stay away from Arachne as long as I can, If I have to use it, I'll of course use it. (And set "Perimeter distribution count" to one lower than "Perimeters")