Need help with external thin walls connection  

Please save your project as a .3mf file

Files > Save Project as

Zip the .3mf and post it here. It will contain both your part and your settings for us to diagnose.

Cheerio,

RE: Need help with external thin walls connection

Many thanks for your quick reply. Here is the archive containing the project file.

You are using Arachne.  Change the perimeter generator to Classic and tick 'External perimeters first'

Cheerio,

RE: Need help with external thin walls connection

Many thanks for the suggestion. I'll give it a try at once (currently printing batches of these parts) and report the results here.

Best.

RE: Need help with external thin walls connection

I just printed a part with the suggested settings. The visual aspect is perfect now.

But the thin wall are now 0.1 mm thicker than in the version using Arachne, where they have the exact dimension as defined in the CAD model. This can be a problem since it will add up with the neighbor parts when snapping with using the embedded magnets and I'm afraid the 0.2 mm increase in thickness making the bind weaker. 

I'll print several instances with this setting and check how they behave. I'd like the bind not be too much weakened because the part aspect is way better now.

Anyways, I thank you a lot for putting me on the right track. Arachne setting being the default in the print profile, I would never have thought about tweaking there.

RE: Need help with external thin walls connection

Yes that's the case with classic.  If you look you will see its actually 2 collapsed walls.   This leads to the thicker walls.  You have the option to 'Detect thin walls' which has always been an option with classic slicing but if you enable it you will notice that you have the same issue as with the slicing set to Arachne.

I think I know how you want it to slice, where it does the outer perimeter and then draws in the 4 corners, but there's no way in PS currently to make it do what you want.  It just wont.

RE: Need help with external thin walls connection

@Neophyl I effectively noticed the 2 head paths along the walls when looking at the generated GCode, and the inevitable dimension increase is not that much a surprise for me.

I was hoping the inner path being done with some under extrusion so that to reduce the overall thickness enough and stay close to the design value. But this not an easy task, and we're touching the limits of FDM here. One must account with a dimensional precision in the 0.1 mm range at the best. Anyway, the Prusa MK3 impress me WRT the dimensional precision, this result having been obtained without any calibration obscure parameter tweaking, and knowing that both units I work with have been assembled as kits. 

I'm about to order a Core One and hope having as much fun and satisfaction with it as with the MK3.

RE:

@epas

I didn't read  everything posted in this thread, maybe I missed some help offered here.

However, I took your file an dragged only the model into PS 2.8.1 ,(Default settings) in the end the result was quite good.

I used the "Classic" perimeter gen. and changed the extrusion value with modifier for the Object "chocolate - body - v2.stl", to "external perimeters" 0.35 and that was it, it printed great, no issues at all and the fitment between "chocolate - top - v2.stl" and the walls of chocolate - body - v2.stl. were very nice and as you might expect from a print f this nature. average wall thickness was 0.61. Very acceptable for the materials proposed.

Regards if you need more info Just ask.

Snuffleupagus

RE:

@snuffleupagus Many thanks for providing your contribution. The cause of the defect has been spotted by Diem in his earlier answers and is effectively related to the Arachne mode.

Disabling it produces continuous perimeters without the artefact. I didn't pay attention to this setting, since the Arachne mode is selected by default in the standard profile as it can bee seen here (PS 2.9.0) :

(the not truncated profile name is visible when clicking on the image to display it separately)

But even if the result is the expected one without any defect, as I mentioned in a previous post, it works at the price of the wall thickness being 0.1 mm thicker than the target dimension. I still have to check that this will not cause some trouble when using the parts, because increasing the distance between magnets by 0.2 mm (0.1 mm for each of the two facing walls) could weaken the bind too much. It would thus be a matter of compromise then : bind solidity versus visual aspect (knowing that nobody has asked me about the artefact until now, maybe thinking that "it's by design" 😉).

RE:

Interesting, I printed two of these using my settings on an MK3 and on an MK3s and had real world results of 0.61mm thin walls meaning that using those settings it printed perfectly with a discrepancy of 0.01mm measured with engineering spec'ed calipers. When we talk of tolerances of that minuit amount they become insignificant based on the flexibility of the materials used. the tolerances were more then adequate and the detents worked perfectly in these circumstances.

Good luck to you, I wish I could have been of more help.

RE: Need help with external thin walls connection

minuit is supposed to be the word Minute*

Not happy with this forums edit policy's

RE: Need help with external thin walls connection

to "external perimeters" 0.35

Epas,  Snuffle also changed the extrusion width, so that may give you the precision you need.

RE: Need help with external thin walls connection

The 0.35 mm setting is a good suggestion and may get me closer to the desired effective thickness. I'll give it a try ASAP and report the results here.

I confess having never touched to default settings for this kind of advanced parameters until now, the precision of the produced parts having been in the expected tolerances (even with quite complex functional ones). This is a new level of mastering I need to experiment with now. 

Anyways, many thanks to all of you guys for your time experimenting and providing me such valuable advice. 

RE: Need help with external thin walls connection

@epas

Much more help can be offered, however you must understand that the default settings in prusaslicer, are only the defaults, because there has to be defaults.

This in no way suggests that the defaults are good for every scenario, in fact the defaults are rarely the answer. The sooner you discover this and move forward the more you might realize that you need to educate yourself beyond the defaults, and the more success you will find in what appears to be a hobby for you. Prusa sells their printers based on the false idea that anyone can just buy one and print anything. It's a lie, education is required, more then just a little.

Their are several folks here that are more then just a little educated, however there are also many more here that don't know and make suggestions as if they know, you should watch out for this as you proceed with your education. Also I would strongly caution you to watch out for whatever the latest version is frankly 2.4.2 is currently one of the greatest versions of this software ever made, while newer versions offer "some new perhaps" better features?, many times the new versions come at a loss this stems from prusas desire to sell more printers, not from a desire to improve the community.

keep all this in mind as you proceed, things aren't as portrayed.

Regards

RE: Need help with external thin walls connection

Hi,

Just tried the 0.35mm setting for external perimeters and here is the report as promised. The wall thickness is now spot on the expected value (measured with a precision caliper)  👍. And the artefacts have gone thanks to the disabling of Arachne mode. The parts are 100% perfect now.

@snuffleupagus Thanks a lot for the suggestion, and I really appreciate your latest feedback. I confess not having played with defaults until now because, although not used for a business type activity, the printers are involved in a quite sustained production for a non profit educational organisation I volunteer in. This being done on my free time (at least while I was still in activity), I hadn't enough time to experiment with advanced tuning. Although exciting and intellectually very rewarding, it's basically a trial and error process, which is quite time consuming. BTW, until now these defaults have worked quite well WRT my needs and expectations, so I chose to play on the safe side by applying the good old "if it ain't broke don't fix it" principle 😀 .

I'm really grateful to everyone having contributed to this thread for having solved my issue and making me a bit more educated in the domain. It will probably be useful too for other users in a similar situation.