Support contact Z distance / Supported surface optimization
 
Notifications
Clear all

Support contact Z distance / Supported surface optimization  

  RSS
oliver.n
(@oliver-n)
New Member
Support contact Z distance / Supported surface optimization

Dear all,

I have tried to optimize the surface finish of supported layers above removable (non-soluble) support structures in Slic3r PE 1.38.5. With the standard settings I had the problem that the first supported layer wouldn't stick to the support at all, causing the perimeter lines to be deformed (dragged by the nozzle).

After 14 iterations I arrived at a setting that would produce acceptable perimeters of the first supported layer and a reasonable supported surface quality. The test objects were printed with PETG (white, from Filamentworld.de) at 215°C nozzle temperature, 0.2mm layer height on a Prusa i3 MK2S with a 0.4mm Micro Swiss nozzle, a modified nozzle fan ( https://www.thingiverse.com/thing:2276997 ) and the following parameters

Support Contact Z distance = 0.10 mm
Support Interface Layers = 10
Support pattern spacing = 0.2 mm
Extrusion width of support = 0.35 mm
Bridge flow ratio = 1.0 (!!)
Bridges fan speed = 50%

Finally, setting the bridge flow ratio to 1.0 solved most of the problem. Still, the perimeters of the first four object layers above the support are not very good and the object was support was very easy to break away. The slicer actually created a distance of 0.5mm between the z values of the last support layer and the first object layer. I guess that a smaller distance between those layers would make the support a little bit harder to break away but would improve the print quality.

In the following thread a similar concern is discussed: https://github.com/alexrj/Slic3r/issues/2930

Apparently, in bridge mode Slic3r will attempts to extrude a thread with a diameter equal to the nozzle diameter, explaining the 0.5 mm actual distance. With a bridge flow ratio < 1.0, however, the extruded thread should have a smaller diameter. Hence, in order to get the correct contact distance, one has to set bridge flow ratio to 1.0, which will impair bridging performance for unsupported parts of the object. Furthermore, the slicer will not create the correct support pattern for support contact z distances < 0.10 mm (but still > 0). Setting this variable to 0 (for soluble support) will disable bridge mode and cause strong adhesion between the support and the object.

It would be great if one would have more freedom to fine tune the contact distance to get smaller z distances between the top support layer and the first supported object layer. In the above-mentioned thread this was solved by making changes to "SupportMaterial.pm" (function "sub contact_distance") , causing the program to use "$layer_height + $extra;" as actual z distance instead of "$nozzle_diameter + $extra". With Slic3r PE 1.38.5, however, I cannot find this file or this function - maybe because supports are generated differently in the Prusa edition? On the other hand, I would not like to give up all the benefits of the Prusa edition.

Is there any way to solve this problem? I got inspired by a very good supported surface quality I got with another single filament FFF-printer (Markforged Mark Two) and believe it should be possible to get close to that with the Prusa i3 by fine tuning support settings.

Kind regards,
Oliver

Opublikowany : 05/02/2018 12:00 am
waino
(@waino)
Eminent Member
Re: Support contact Z distance / Supported surface optimization

Unfortunately, this is exactly the same problem that I've been trying to solve as well.
Based on my own experience, basically any model that has any hanging details is most likely going to fail when sliced with Slic3r due to the fact that the below support is not going to touch the object firmly enough.

For example the tail of the horse in this example:

https://www.thingiverse.com/thing:182335/#files

Will always fail with Slic3r (Prusa Edition) as there is too high gap between the below support and the bottom area of the tail and therefore the tail will eventually fall of in the middle of the printing. I have tested the above example using Cura v3.1 and the print was succesfull.

Opublikowany : 12/02/2018 5:20 pm
Share: