Thin feature (hexagon) first layer issue

[CulPrint 0]

Hi

 

First of all, thank you for the awesome product!

 

Now, to the business. I'm printing RPi 4B case and it has hexagons as vent holes (https://www.thingiverse.com/thing:3793664). I'm unable to print the first layer of hexagons if I have initial layer horizontal expansion in "correct" value and I have to set it to 0. Closer to the good value the first layer of hexagons won't even slice and any negative value that slices the hexagons causes adhesion issues. Those hexagons seem to have 0.5 mm wall thickness and I'm using 0.4 mm nozzle. I already have 150% initial layer flow and 140% initial layer line width to get very good fist layers and adhesion to PEI bed. Any Ideas? Without expansion corrections it print somewhat good.

 

Cura 5.2.2.

 

How about support blocker style any feature blocker which can be drawn on first layer as polyline/B-spline etc.?

Edited August 8, 2023 by CulPrint

[gr5 2,270]

6 hours ago, CulPrint said:

I'm unable to print the first layer of hexagons

I'm confused if your issue is that the printer doesn't print them very well (or at all) or if you are having a problem with the slicer.

 

Please show what you mean with either a photo of the first layer or if you never actually print (because the issue is clearly cura) then a screenshot of PREVIEW screen in Cura.  And actually please also post the project file: do menu "file" "save project" and post that file.

[CulPrint 0]

Sorry for the confusion. If my initial layer horizonal expansion is in good adjusted value, let's say -0.40, the slicer does this:

 

 

And the second layer is the actual first layer that prints hexagons, which naturally does not adhere to the bed:

 

 

I need to adjust the initial layer horizontal expansion to 0 in order to actually print anything in the first layer of hexagons:

 

CCR10S_BOTTOM.3mf

Edited August 9, 2023 by CulPrint

[GregValiant 1,415]

You have a lot of settings that are simply making this difficult to print.

The ribs of the hexes are only .50 mm wide.  You have Hole Horizontal Expansion turned on at .08mm so now the ribs are only 0.34 wide.  Your Initial Layer Line Width of 140% * 0.40 Line Width is .56 wide.  That's why nothing is showing up on the first layer...You are trying to print .34 features at .56 line width.

You need to get rid of that 140% Initial Layer Line Width.  It's just too wide for the fine features of this model.  Drop it to 100% and go to 0.50 Line Width for everything.

Turn off Hole Horizontal Expansion as well.  It works well for "circles" but those hexes are  vertical holes as well and they don't need it as the extrusions won't be getting dragged into the center because there is no real center.

You have Coasting enabled.  It's OK (I guess) for long runs but this model is all about the hexes and they just don't have any long runs.  You will end up with a lot of dry starts to extrusions using coasting on this model.

 

This is the "Standard" profile with 0.50 line width and 3 walls.

 

[CulPrint 0]

It did print just fine with the original profile when I set the initial layer horizontal expansion to 0. I also printed it with the changes you suggested and that works too. My original initial layer line width was remnant when I was using the previous bed material, which gave some issues without those changes.

 

Thank you for the comments!

 

For future reference, if I have fine features like this and holes that need to be dimensionally accurate, what should be done? Without changing the model, how to print holes (like screw counterbore) well and avoid elephants foot?

[GregValiant 1,415]

"... if I have fine features like this and holes that need to be dimensionally accurate, what should be done? " 

The absolute best thing you can do is give Cura the model you want printed.  Not a model that is "close except for this section over here..." but the model you want to end up with.

 

Small holes are always tough but if you want a hole to finish at 3.8 and it's closing up do to the "Snot Factor" pulling the molten plastic in towards the center, make the hole bigger in the model, or run a drill bit through it.

 

I printed a hole chart from 1mm up to 30mm and what I found was that as the holes get larger the amount of HHE required goes down to 0.  Instead of using the results from that to plan the HHE in Cura, I use the results to make my holes the correct size in the model so that they come out correct in the print.

 

HHE doesn't work right if there are cross holes and it works on any vertical "hole" no matter what the shape.  Rememberr ... in the actual model file all holes are "faceted" so nothing is truly round and Cura can't tell the difference between a 25 facet "supposed to be round" hole and a 25 facet oblong pocket with a couple of squared corners.

 

There is a new setting for "HHE Maximum Diameter" that can help when a cross hole gets in the way or just cut out HHE if the hole size is above the setting.

[CulPrint 0]

3 hours ago, GregValiant said:

"... if I have fine features like this and holes that need to be dimensionally accurate, what should be done? " 

I printed a hole chart from 1mm up to 30mm and what I found was that as the holes get larger the amount of HHE required goes down to 0.  Instead of using the results from that to plan the HHE in Cura, I use the results to make my holes the correct size in the model so that they come out correct in the print.

 

That's quite interesting. With one certain PLA+ I'm using, the larger the hole, the larger the hole deviation from the model. But when calculating the "relative deviation", that is, deviation from nominal divided by the nominal size, the larger the hole, the smaller the value. Below is table of printed circles and their nominal size and actual size before horizontal expansion adjustment. Clearly there isn't linear correction that can be applied.

 

Nominal	measured	diff	diff/nominal
20	19.70	0.30	0.015
15	14.83	0.17	0.011
10	9.86	0.14	0.014
5	4.91	0.09	0.018

 

I won't be using this filament after the spool is empty, but it has had the worst warping/shrinking I have ever used. Seems to be pretty much filament dependent phenomenon.

[GregValiant 1,415]

It's dependent on the mechanicals and settings as well.  Big nozzles and tall layer heights are worse as the molten filament isn't controlled as well and it wants to be pulled to the middle.  High resolution model files also make the situation worse because the filament just doesn't want to curve, it wants to stay in a line behind the nozzle.  So a rectangular shaped hole doesn't have the problem as all the runs are straight.  As facets are added and the holes starts to get closer to a circle the problem gets worse.

Another clear cut case of "Everything affects Everything".