Bug: No adhesion in between perimeters of multi part models
 
Benachrichtigungen
Alles löschen

Bug: No adhesion in between perimeters of multi part models  

  RSS
Chris
(@chris-2)
Eminent Member
Bug: No adhesion in between perimeters of multi part models

Hi all,

I reported this bug on github, but I wonder if anyone else has noticed it or has a workaround.  I'm printing a multipart model.  Where the parts have layers in common, adhesion is great.  Where the parts only have perimeters in common, I'm getting little or no adhesion.  It turns out that Prusa Slicer is not calculating perimeters correctly for multi part models which is resulting in weak adhesion between model parts. Here are shots of a model sliced as a single part and as a multipart to compare how Prusa Slicer is calculating line widths. In the single model version, Prusa Slicer creates three perimeters. The exterior perimeters are about 1.2mm and the interior perimeter is about 0,9mm with a little extra at the corners.

When the model is divided into three separate parts (each perimeter is a part - I have PC-GF sandwiched between two perimeters of PC), Prusa Slicer calculates much lower line widths. The two exterior parts are given about 0.9mm line widths and the interior part is given the same, about 0.9. This results in much less material being extruded and little or no adhesion between perimeters.

Any ideas on how to change this?  I played with the extrusion multiplier to force extra material, but that messes up the model in other places.

Thanks

Veröffentlicht : 31/01/2024 6:26 pm
Diem
 Diem
(@diem)
Illustrious Member

I consider this a feature, not a bug.

Print-in-place mechanisms rely on non-adhesion between adjacent parts.  If adhesion between adjacent exterior perimeters were to increase then the tolerance between print-in-place parts would have to be widened and the resulting slop/looseness/backlash would render many of them impractical.

Where it matters that the connection is strong it helps to design interlocking interfaces - eg., PLA and PETG do not adhere well to each other but if the connection follows a jigsaw-puzzle curve they are not coming apart without breaking something.

Cheerio,

Veröffentlicht : 31/01/2024 9:41 pm
Chris
(@chris-2)
Eminent Member
Themenstarter answered:
RE: Bug: No adhesion in between perimeters of multi part models

Fair enough - but it's only a feature if that's what you are trying to accomplish.  One of the advantages of the new tool changer generation, like the XL, is that you can more freely mix compatible materials for engineering purposes because you waste a lot less material.  Polycarbonates are a great example.  In my case I'm making a lamp and need a single perimeter of white PC for light diffusion, and then 2 layers of clear PC for durability.  I can't use physical structures to create the connection.  I haven't been able to find a good way to manage the overlap of the perimeters to create the adhesion that I need.

Veröffentlicht : 31/01/2024 11:16 pm
Diem
 Diem
(@diem)
Illustrious Member

I can't use physical structures to create the connection.

Why not?  There is always another way to do it.  Incorporate a few buttons at the bottom of the outer later to penetrate through the interlayer into sockets in the base and perhaps repeat at the tip of the cone and most of the work is done.  Or redesign internal ribs in the corners...

If this is a customer's specification throw the problem back at them, tell them they have to design for the medium. 

Only if this is meeting a legal specification (eg., Maritime Navigation Lights) must you really search for workarounds, the great advantage of 3D printing is iterative design and testing.

Cheerio,

Veröffentlicht : 01/02/2024 1:22 am
Snuffleupagus
(@snuffleupagus)
Estimable Member
RE: Bug: No adhesion in between perimeters of multi part models

@chris

 

Have you tried using "Classic" perimeter generation?, I see from your file that you have the "Arachne" generator selected. It may not be the best for this project as it has the issues you describe even on some of the most simple objects.

Veröffentlicht : 01/02/2024 6:25 am
BaconFase
(@baconfase)
Reputable Member
RE: Bug: No adhesion in between perimeters of multi part models

What if you did something like:

Import the 'solid' model to use for the CF bottom and the inner/outer transparent parameters. Make the pyramid section hollow (for the inner PCGF).

Import the '3 layer' model as a separate model from above. Exactly overlap it with the other model using coordinates. Delete the inner/outer transparent parts. Add a single layer small slab to allow the pyramid to float, then you can delete the CF bottom part as well. This is now effectively just the inner PCGF floating in the air, but in a desired position and as a separate model from the rest.

You may need to tune the perimeter widths in case of too much overlap, but fiddling comes with the hobby. For whatever reason PS doesn't seem to want to generate the middle part if it is set to 1mm or above.

I also had to change a few things, like going back to Classic instead of Arachne, and other stuff to remove the purple solid layers so the transparent parts would print hollow. Fix your seams. Maybe other things to fix. There might be actual overlap somewhere, I didnt pick through the haystack.

But you get the idea.

justDumbOverlap

 

XL-5T, MK3S MMU3 || GUIDE: How to print with multiple-nozzlesizes do read updated replies || PrusaSlicer Fork with multi-nozzlesize freedom || How Feasible is Printing PETG for PLA supports on XL very

Veröffentlicht : 01/02/2024 6:36 am
Chris
(@chris-2)
Eminent Member
Themenstarter answered:
RE: Bug: No adhesion in between perimeters of multi part models

Thanks all for your help.  I did back into a solution. 

The classic slicer has the same problem and also struggles with precise line widths for one perimeter models.  Arachne is really good at that.  The mechanical solution doesn't work for this project because it's a lamp.  So - any variation in the transparent or white pieces is visible in the lamp - which is undesirable.

What I settled on is the extrusion multiplier.  If the outer PC-Clear perimeters are at 0.9 line width and the inner PC-GF line width is at 0.6, the math puts you at 1.33 or so for good adhesion.  I tried 1.3 and there was very good adhesion, but not perfect.  At 1.4 the adhesion is perfect and when you break the part the perimeters do not separate.  It's very cool to see the laminated piece work out with clear on the in/outside and the white layer in the middle.  Strength wise it behaves exactly as I expected.

To account for the over extrusion in other places I will use a single perimeter setting, zero top and bottom layers, and rectilinear infill at an appropriate percentage (something around 80%, I haven't tested yet).  This should leave spaces between the lines to account for the over extrusion in the more conventional parts of the PC-GF model and create something pretty similar to a monotonic top and bottom.  This won't work for everyone but it should solve my issue.  If we had per model extrusion multipliers that would make it a lot easier.

 

Thanks!

Veröffentlicht : 01/02/2024 7:33 pm
Chris
(@chris-2)
Eminent Member
Themenstarter answered:
RE: Bug: No adhesion in between perimeters of multi part models

I meant to say I only applied the 1.4 extrusion multiplier to the PCGF part.  I can't find a way to edit my original post.

@BaconFase, I'll also try splitting the PCGF part and floating the single perimeter part with a minimum line width and/or wider geometry in Arachne.  That also may be a good way to resolve not over-extruding for the rest of the part.  My model is more complicated than the model in the project files.  I just made this simple model to illustrate the point more directly and to test adhesion in a print instead of printing a full prototype piece of the full model - which would be wasteful.

Veröffentlicht : 01/02/2024 10:41 pm
Teilen: