Parallel vertical walls retraction when connected by internal parts and some supports missing  

Unfortunately only one of your included files extracts cleanly, the jpeg ending 07, the others fail and the stl is empty ...

From the one picture I can open I wonder if you would get better results if you increased the number of perimeters but this is mostly guesswork.

From bitter experience we generally suggest old fashioned zip archives here.

Cheerio,

RE: Parallel vertical walls retraction when connected by internal parts and some supports missing

7zip seems to be able to cope with the archive.  Although I do wish you had included a saved project file.  Without it we don't know ANYTHING about your setup.  What printer, what filament, what profile were you using ?  With one we don't have to play 20 questions and we get the model you are having trouble placed exactly as you had it.

Anyway if I had to print that model using PS2.3.x as a slicer I would first try as I've done it on the attached project.  Its fairly standard, although I've set the printer to be a basic MK3 without knowing what you are running. Just change the printer to yours and the filament is my normal PLA profile so that might need changing too. 
The bottom of the bridges (what you call the interconnect) are not going to looks great with PS2.3.x or earlier but will print out better that you pictured as set in the project.

I'm guessing but  you mention 'missing support' , I suspect you mean the support under the bridging interconnects.  Well by default PS has an option set called 'dont support bridges'.  When enabled it will not place support under bridges, as technically it doesn't need to.  Sometimes for a better surface that you might want to.  So I've turned that option off in the project.  
Not sure what you mean by retraction as I cant see what that setting has anything to do with your pictures?

Personally I would be using PS2.4 even though its in beta as the snug supports make it more efficient to print and I could have selected thin bridges instead of turning off bridge detection. 

Anyway have a look in the attached project and study the preview.  Then try a print.

RE: Parallel vertical walls retraction when connected by internal parts and some supports missing

Hi,

I attach the new saved project with a couple pixes here. With your hints I got the supports under the bridges and so the bridges are better (but not perfect).

However I still have the recess in the lower end of both the walls as you can see in the attached pixes.

This time I sliced with PrusaSlicer 2.4.0 beta 1.

Those walls should be just vertical ones, without thouse recesses in the lower end.

The recesses disappear if I print without the internal bridges, so the recesses must be related to those bridges but I can gigure out why.

Obviously the OpenSCad project is exactly the same: it only with the bridges added or not. The wagon walls are crafted by subtracting a smaller parallelepiped from a larger one. So no reason for those recess to exist! I enclose the .STL for best clarity.

Posted by: @neophyl

7zip seems to be able to cope with the archive.  Although I do wish you had included a saved project file.  Without it we don't know ANYTHING about your setup.  What printer, what filament, what profile were you using ?  With one we don't have to play 20 questions and we get the model you are having trouble placed exactly as you had it.

Anyway if I had to print that model using PS2.3.x as a slicer I would first try as I've done it on the attached project.  Its fairly standard, although I've set the printer to be a basic MK3 without knowing what you are running. Just change the printer to yours and the filament is my normal PLA profile so that might need changing too. 
The bottom of the bridges (what you call the interconnect) are not going to looks great with PS2.3.x or earlier but will print out better that you pictured as set in the project.

I'm guessing but  you mention 'missing support' , I suspect you mean the support under the bridging interconnects.  Well by default PS has an option set called 'dont support bridges'.  When enabled it will not place support under bridges, as technically it doesn't need to.  Sometimes for a better surface that you might want to.  So I've turned that option off in the project.  
Not sure what you mean by retraction as I cant see what that setting has anything to do with your pictures?

Personally I would be using PS2.4 even though its in beta as the snug supports make it more efficient to print and I could have selected thin bridges instead of turning off bridge detection. 

Anyway have a look in the attached project and study the preview.  Then try a print.

 wagon_211031.zip

RE: Parallel vertical walls retraction when connected by internal parts and some supports missing

The reason its doing that is the transition from using gapfill to perimeters.  There's lots of threads here, details 'buldge' - look them up.  Yeah the spelling funky but that's the original poster for you.  Its a really common problem, happens on lots of models where perimeters transition to other types like gapfill or solid infill/top etc.  Like the Benchy hull line.  Same fundamental problem.  

Turning OFF detect thin walls will help, you should see it start to use perimeters and the gapfill will go.  That removes one source of transition.   In the project I attached I had it turned off and there wasnt gapfill.  Did you try printing the one I attached after setting the printer/filament to your own ?   Best 'fix' is to make your model walls thicker so that it can always fit perimeters into the space.  

You could also try external perimeters first on those layers using a modifier.

If you are using 2.4 beta then also turn OFF Thick bridges. and also change your support type from Grid to SNUG.

RE:

Posted by: @neophyl

The reason its doing that is the transition from using gapfill to perimeters.  There's lots of threads here, details 'buldge' - look them up.  Yeah the spelling funky but that's the original poster for you.  Its a really common problem, happens on lots of models where perimeters transition to other types like gapfill or solid infill/top etc.  Like the Benchy hull line.  Same fundamental problem.  

Turning OFF detect thin walls will help, you should see it start to use perimeters and the gapfill will go.  That removes one source of transition.   In the project I attached I had it turned off and there wasnt gapfill.  Did you try printing the one I attached after setting the printer/filament to your own ?   Best 'fix' is to make your model walls thicker so that it can always fit perimeters into the space.  

You could also try external perimeters first on those layers using a modifier.

If you are using 2.4 beta then also turn OFF Thick bridges. and also change your support type from Grid to SNUG.

Hi, I followed your hints: enlarged perimeters vertical shells to 5 (up from 3),  turned OFF "detect thin walls", enlarged .5 mm the walls (can't make it larger), set "thick bridges" to OFF, turned support to SNUG (btw I found they're much harder to remove then grid, it's really hard to remove the last layers which is strongly sticked to the print first layer).

I tried reducing the bridges penetration into the walls to .5 mm, too.

You can see by the attached pixes the issue stays there and it's closely related to the internal bridges: the bulge starts at the layer they start.

I attached the .3MF, too.

I wasn't able to print with your settings because my printer (Modix BIG 60) goes crazy from the beginning also if I set my usual g-code flavour.

RE: Parallel vertical walls retraction when connected by internal parts and some supports missing SOLVED!

Hi, I wondered if the issue was due to the new versions of PrusaSlicer, so I tried slicing with 2.3.1 and it worked!

There's still a very subtle imperfection but you can barely see looking from the right angle.

So it has to be a bug from version 2.4: let's hope someone of the developers will fix it.

Thanks to everyone for your time and suggestions!