Perimeter wall detachment

[CSharp 0]

Good morning everyone,
I recently (and with satisfaction!!) switched from Simplify 3D to Cura but I have a problem that I can't solve and it's what you see in the photos, it seems that the internal perimeter of the holes is detaching. Using Simplify I don't have this problem so the problem must be some parameter of the slicer but I don't know which one.

My printer is a Sharebot Viper (i.e. a rebranded FLASHFORGE GUIDER IIS..).

Do you have any ideas?

 

Thank you

 

[Dustin 178]

They are detached because your are significantly under extruding..
Your top surface is actually equally as bad and shows this as well.

Your either printing too cold or your flow rate is too low (or both)

[GregValiant 1,455]

Also, make sure your wall ordering is "Inside to Outside".  Some under-extrusion coupled with the outer wall going down first can drag the too-skinny filament across the middle of the hole instead of it staying where it was put.

[CSharp 0]

12 hours ago, Dustin said:

They are detached because your are significantly under extruding..
Your top surface is actually equally as bad and shows this as well.

Your either printing too cold or your flow rate is too low (or both)

 

And what do you suggest changing in the print profile?

 

• The filament size is correct.
• The nozzle size is correct.
• The wall ordering is "Inside to Outside".
• The temperature is correct (and using Simplify 3D with the same temperature I have no problems)
• With the THIN WALL TEST I have a correct result.

I attach the profile I am using

MyDraftProfile.curaprofile

[GregValiant 1,455]

One thing to consider is that Simplify, PrusaSlicer and Cura calculate volume-of-extrusion slightly differently.

• Cura considers a simple rectangle LayerHeight x LineWidth x Length.
• PrusaSlicer considers that the left and right wall of any extrusion will be a radius
• What I've read is that Simplify adds a simple fudge factor of a couple of percent to the LH x LW formula.

The differences are subtle, but there.

I've looked at your profile and there are a couple of things that are frankly personal preference, but it's because I didn't like the results I got when I use them.  "Coasting" is at the top of the list and "Retract Before Outer Wall" is on the list.  They both effect the start of the next extrusion.  My Ender doesn't like them.

I get filament from Matter Hacker.  It is consistently 1.72 diameter.  I enter that into Cura's Printer Settings because at 1.75 diameter I would be suffering about 4% under-extrusion right-out-of-the-gate.

 

Top skins are the best for checking "flow".  I picked up a microscope at a garage sale for cheap but a decent magnifying glass will work as well.

What @Dustin and I are seeing in your image are gaps between the extrusions of the top skin.  If you zoom in you see that the edges of the extrusions are not welded together and you can see the cross-hatch pattern (of the layer below) through the gaps in the top layer.  That is definite under-extrusion and when you add a bit of material shrinkage, you get the gap you see around the hole and material being pulled into the center of a circle.

 

The simplest thing to do is increase the flow in Cura.

The project below has a FlashForge Dreamer NX as the active printer as it is the only FlashForge printer that was available in Cura.  There are three identical models with different Per Model Settings.  The model on the left has the flow at 100%, the middle model has flow at 110% and the right model has the flow at 115%.  I set the Line Directions to [0, 90] to make it easier to check while they are printing.

Open the project and switch to your printer.  I started out with your MyDraft profile but go over all the settings to insure they are what you have been using.

Watch during the print and when they are done check the top surfaces with a magnifying glass and see what you see.

I don't know your printer, but pretty much all Creality machines must have the E-steps/mm calibrated in order to produce the correct volume of filament for extrusions.  If you can or have calibrated the E-steps then DO NOT use a "single wall calibration cube" to further alter the flow.  Just insure that when asked to deliver 100mm of filament the extruder does actually deliver 100mm - no more and no less.

 

 

Flow Test Cubes.3mf

 

EDIT:

This is an exaggeration but it's what I'm seeing in your images.  Notice that you can see the layer below through the gaps in the top layer.

Edited November 21, 2023 by GregValiant

[CSharp 0]

Wow, I had no idea about these differences in the calculation of flows between the various slicers.
I did the test you suggested with the flow at 110%, I would say that I no longer have problems with the loops detaching and in the top you can't see the layer underneath.
I will try with intermediate values between 100 and 110 to see which is the best because the dimensional accuracy is a little worse but I will compensate in another way.
 To report this change in my profiles I have to set all the parameters relating to the flow in the material sheet to 110, right?
Should I leave the "Flow Rate Compensation Factor" parameter in the Experimental tab at 100%?

 

Thanks for the tips

Edited November 23, 2023 by CSharp

[GregValiant 1,455]

Yes - all the flows should be the same.  I have my printer tuned to PLA but I have found that PETG likes the flows at 105%.  I could change the E-steps/mm to account for that, but then I'd just be moving the flows down to 95% for PLA.  So it just doesn't matter.

Because the layer height doesn't change and the extrusion length is what it is - when you make a flow adjustment the Line Width changes.  So you could get some variation in the dimensions.  At 0.40 line width if you change the flow by 10% you will get a change of 10% to .44 for the Line Width.  For an outer wall that has an inner wall already in place next to it, that extra .04 of width is all to the outside.  That is not a big change but is is a change of possibly .08 across the print.  (I see more change than that if I put the print in the sun.)

 

I'm not happy with the Flow Equalization Ratio.  It makes changes to the Print Speed to account for the flow changes that come from Cura varying the line widths and I'm seeing really out-of-whack speeds because of that.  There are several bug reports on Github in regards to the speeds.

You can just leave it at 100% since it doesn't really have anything to do with the general flow, just flow that is being changed due to line width variation.  I set it to zero because I'm mad at it.  I haven't seen any real difference in the prints but I don't get print speed adjustments to the 300mm/sec range anymore.