How do I configure the slicer for this?

[neilhiggins 0]

The attached picture is Cura's preview of an STL file that I want to slice and convert to gcode. It's a cylinder with radial holes and a few other embellishments. For best print quality, the first layer, or maybe the first couple of layers, above each layer of holes will need to be printed concentrically, so that the bridges are as short as possible (radial bridging would go "from nowhere to nowhere" - disastrous). From there to the next layer of holes, just about any infill will be acceptable. I have browsed the options; so far I am not convinced that anything I have found will work as needed. Thanks in advance for any hints you can offer. Regards, Neil

[ahoeben 1,951]

You could set "Top/Bottom Pattern" to "Concentric".

[Slashee_the_Cow 270]

For @ahoeben's suggestion to work, you'd also need to set Top/Bottom > Top/Bottom Thickness to at least half your model's total height so that every layer is printed in a concentric fashion (assuming I understand the situation correctly, but with the cow-human translation going there's plenty of times I don't understand the situation).

 

It would really help if you could post a Cura project file (.3mf, get it ready to print then go to File > Save Project) or at least the model STL file.

 

Based on the screenshot it looks like your holes are pretty tiny - make sure you use the thinnest lines and shortest layers you can pull off with your nozzle (or switch to a smaller nozzle if possible).

Edited February 2 by Slashee_the_Cow

[neilhiggins 0]

Hi @Slashee_the_Cow. 

On 2/2/2024 at 4:45 PM, Slashee_the_Cow said:

set Top/Bottom > Top/Bottom Thickness to at least half your model's total height

Did you really mean that? A literal interpretation does not compute, meaning that you are writing off @ahoeben's suggested approach as a non-goer. I have attached a .3mf of the part for your further analysis and comment (at your option and convenience, of course).

 

Immediate thoughts: (1) Perhaps there is a legitimate need for a "professional" slicer product which (somehow) facilitates fine control of the tool path (and I don't mean editing GCODE, thankyou very much), or (2) Perhaps there is a need for a new option in CURA to cater for this kind of model, or (3) Perhaps there is a legitimate need for me to RTFM, if there is one and it contains any helpful content in this context. 

 

Small backstory: I made the shape by printing it as a stack of discs, glued together. My printer (Wanhao i3 mini) is so tiny that I actually had to print each disk as four dovetailed quadrants, also glued together. This model is a kind of experiment to generate a single-piece equivalent (which for obvious reasons I cannot print). The STL was generated by a bespoke Python program that I wrote (no CAD involved) . Another conversation on this forum discussed some early problems (inverted normals) which were subsequently fixed.

 

As with many STL files, there is a complete disconnect between the shape's triangle composition and the toolpath required to successfully print it, which puts a lot of onus on the slicer to work out the gory details. Perhaps it's too much to ask for your average free slicing software.

CE3S1_20240130-233109-V0-6-solid.3mf

[Slashee_the_Cow 270]

18 minutes ago, neilhiggins said:

Did you really mean that? A literal interpretation does not compute, meaning that you are writing off @ahoeben's suggested approach as a non-goer. I have attached a .3mf of the part for your further analysis and comment (at your option and convenience, of course).

Based on your image and text, I meant it as literally and honestly as I do that I'm going to get myself a piece of mudcake in an hour or two (I buy the giant ones from Costco, cut them up and freeze them).

 

Here's what it looks like if you have normal top/bottom settings and choose a concentric infill:

That gives each hole a concentric top, which gives you those long lines (along the Y axis in that picture) in midair, which exactly what you don't want.

 

But if we make the whole thing skin (FWIW I set Top/Bottom Thickness to 64mm because it's a round number (I like round numbers) which is more than half the model height (the important part):

It slices the whole thing like it would slice the circular top or bottom, not treating each individual hole cover as its own concentric fill. The downside of this approach, however, is that it makes the whole thing solid, because there's no infill in skin. That increases material use and print time (I'm not sure if those are important factors to you).

 

Fortunately I had a flash of inspiration: skin expansion!

That's with Top/Bottom > Skin Expand Distance set to 2mm (technically any value above half the distance between holes should work, but going over doesn't matter). Now instead of treating the top of each hole as a bit of skin, the skin has expanded enough that it fills that layer so it can print it as a full concentric pattern!

 

Important note: You'll need to change your infill pattern to something else. Concentric infill here will add next to no strength.

 

It'll do that for however many layers you have Top/Bottom > Bottom Thickness > Bottom Layers (and its counterpart for top layers). By default that's four layers each, so it's doing twelve layers of infill total. If you'd prefer a bit more infill, just lower the top and bottom layers but assign values to Top/Bottom > Top Surface Skin Layers (how many layers it will do on the top) and Top/Bottom > Bottom Thickness > Initial Bottom Layers. By default they'll follow the normal number of layers but if you wanted to for example reduce it to two layers above and below the holes for more infill you can still have enough layers for a decent top and bottom.

 

I would also like to give you the award for "model which takes the longest to slice, under 20,000 verts". That is both a joke and the truth (the best kind of joke).

[GregValiant 1,258]

I thought I'd stick in my 2 cents worth...

Concentric Top/Bottom with Concentric Infill can allow a lot of the first skin layer to fall into gaps in the infill.

 

I think I'd just adjust the "Skin Expand Distance" to 2.0.  That would work with "Lines" for top/bottom and a grid infill.

 

It would likely be neater with Concentric top/bottom but my preference is the added strength of alternating directions of the extrusions.

 

 

 

Edited February 20 by GregValiant

[Slashee_the_Cow 270]

Now that I've managed to get in before Greg, to answer some other particulars:

1 hour ago, neilhiggins said:

(1) Perhaps there is a legitimate need for a "professional" slicer product which (somehow) facilitates fine control of the tool path

Based on the price of Simplify3D, it's gotta be either really good or really overpriced. Prusa Slicer has more options than Cura in some categories but its UI looks like it was designed by someone who only knows how to program a slicer (that is to say: needlessly complex and obtuse).

 

1 hour ago, neilhiggins said:

(and I don't mean editing GCODE, thankyou very much)

I'm not sure anyone would suggest hand editing the amount of gcode involved this project, maybe unless you only needed to do a global search & replace. Some projects I've made, I've needed to write single-purpose post-processing scripts.

 

1 hour ago, neilhiggins said:

(2) Perhaps there is a need for a new option in CURA to cater for this kind of model

Here's the feature request page.

 

1 hour ago, neilhiggins said:

(3) Perhaps there is a legitimate need for me to RTFM, if there is one and it contains any helpful content in this context

I'll admit I've yet to RTFM but I doubt it would cover a situation a situation as specific as this. It's the sort of thing you learn over time, especially if you help out on the forums and figure out how to do all the different things people want to do 😉

 

1 hour ago, neilhiggins said:

The STL was generated by a bespoke Python program that I wrote (no CAD involved) .

Points for effort there. I've written some single-purpose Python programs before but generally only to automate a simple (but annoying) task. I could probably do this in twenty minutes in OpenSCAD or half an hour in FreeCAD but I wouldn't learn anywhere near as much about how this stuff is actually done.

[neilhiggins 0]

Dear @Slashee_the_Cow and @GregValiant,

 

Thank you again for your time, and the benefit of your experience.

 

On 2/20/2024 at 9:53 PM, Slashee_the_Cow said:

I meant it as literally and honestly as I do that I'm going to get myself a piece of mudcake in an hour or two

Yum. I hope you enjoyed it.

 

On 2/20/2024 at 9:53 PM, Slashee_the_Cow said:

Fortunately I had a flash of inspiration: skin expansion!

What can I say? Good get! The big issue with applications like this is how the numerous parameters interact, not something that a newbie can necessarily intuit. Try and fail, try and fail, try and ... succeed? (or ask an expert)

 

On 2/20/2024 at 9:53 PM, GregValiant said:

It would likely be neater with Concentric top/bottom but my preference is the added strength of alternating directions of the extrusions

I agree, with the reservation that ANY direction will coincide with "radial" at two opposing angles, causing the fill to fall through the cracks. So unless I have misinterpreted your post, the optimum solution would appear to be a concentric layer topped by alternating directions.

 

On 2/20/2024 at 9:53 PM, Slashee_the_Cow said:

I would also like to give you the award for "model which takes the longest to slice, under 20,000 verts". That is both a joke and the truth (the best kind of joke).

Not an objective, and by the way, it slices MUCH more quickly under Cura 4.6. Something has changed: Why, and whether for the better or worse in the long run, I cannot say.

 

Anyway, I am very grateful for your assistance, guys. Many thanks, and best regards, Neil