Looking for suggestions - overhangs not looking good  

  RSS
mgoblue
(@mgoblue)
Eminent Member

Hey everyone,

I am having an issue where my overhangs look really bad (even when printed with support).  See attached photos.  The model in the attached photos was a large-ish chamfer printed with support.  I believe the angle of the chamfer relative to horizontal is around 35 or 40 degrees. I used the default 0.20 "Speed" profile in the Prusaslicer settings, and the PLA is the material that shipped with the printer (in July).

Any suggestions on how to fix this?  I need

Posted : 12/09/2019 12:52 am
mgoblue
(@mgoblue)
Eminent Member

Here is a link to an imgur gallery with bigger photos:

https://imgur.com/gallery/YJ4nqpv

 

Posted : 12/09/2019 1:40 am
bobstro
(@bobstro)
Famed Member

If those are fillets, they are going to work poorly on the underside. It's the nature of the current state of consumer-grade FFF printing. The angles don't look overly extreme, but you're clearly getting issues on the parts above the notch. Candidate problems are:

  • Temperatures: What temps are you printing at? If they are excessive, it can contribute to sagging on unsupported overhangs. Cool the nozzle temp or boost the part cooling fan. Are you printing in a warm room or enclosure?
  • Extrusion sizes: Larger (thicker or wider) extrusions are heavy and more prone to sagging. What parameters are you using to print? If the 0.20mm layer height profile with Prusa silver PLA and a 0.40mm nozzle, you should be able to get something better.
  • Look under Print Settings->Layers and perimeters->Quality settings and see of vertical shell thickness and extra perimeters if needed help.

If you'd care to share a project file (3MF) with your part as currently sliced, more help might be forthcoming. Be sure to zip the file before uploading here.

My notes and disclaimers on 3D printing and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan...
Posted : 12/09/2019 3:03 am
mgoblue
(@mgoblue)
Eminent Member

Thanks for the reply.

The surface showing the issue is a chamfer - not a fillet, which I thought would typically work better in this orientation.  My first print of this part had a fillet there, and it didn't look very good, so I did a trial with it changed to a chamfer.

Print temps are 215 first layer, 210 for subsequent layers, and 60 degree bed.  I am printing in a lack enclosure with the doors open.  Temp is about 75F ambient during print (inside enclosure).

I took the part and chopped a piece off (to disguise its function a bit - it is a work-related product), and sliced with the same settings.    The .3mf file is attached as a zip file for your reference.  Thanks for your help!

I am printing the file as shown, with the "extra perimeters if needed" option checked (this setting is not reflected in the file I attached).  The fan speed is already at 100%, so can't go any higher there.  Should I try at lower than 210 degrees?  

Posted : 12/09/2019 3:32 am
guy.k2
(@guy-k2)
Noble Member

Your attached zippped, 3mf file defnitely has round filets at the bottom edges, not angled chamfers. Large, rounded filets at bottom edges cannot print cleanly because they start as a near 180 degree overhang. Definitely change to a chamfer and don't make the chamfer more than 60 degrees from vertical. Some printers and materials will allow another 10 degrees, but 60 degrees overhang is pretty safe limit.

 

Posted : 12/09/2019 4:30 am
mgoblue
(@mgoblue)
Eminent Member

Yes, the fillet you show there is from one of the two short edges in contact with the bed. The edge that is giving me trouble is the longer edge, which is definitely a chamfer. The round fillets ironically print better than the chamfer in this case, but they are smaller. I think the fillets were were  3mm radius, while the long chamfer is 6 or 7 mm x 45 degree. I can double check the dims tomorrow morning. 

Posted : 12/09/2019 4:37 am
guy.k2
(@guy-k2)
Noble Member

OK, looked through your 3mf settings.

1. Change to a Quality setting rather than Speed

2. Increase your perimeters to 3 or 4. Two is too low to hide protrusions from infill creating surface defects. This is especially bad when you use a SPEED profile. The higher infill velocities create even more protrusion effects from the infill. Your thin walls leave zero place for those to be hidden and you see it as undulations in exterior surface.

3. Change to 30% gyroid fill. That and four wall perimeters will give you good strength while reducing plastic waste.

4. Change ALL the filets to chamfers. The chamfers are creating print failures that propagate into the following filet as print head makes its counterclockwise circuit. That means the first printed portion of the otherwise printable chamfer gets distorted.

 

 

 

Posted : 12/09/2019 5:49 am
guy.k2
(@guy-k2)
Noble Member

One more suggestion.

1. Using thinner printer layers yields better overhang performance by virtue of larger percentage overlap. Going to 0.15 mm instead of 0.2 mm means each shift outward is a smaller portion of the extrusion width. I have attached a 3mf with most of my suggestions implemented. Also printed it in PETG (with a PETG filament setting of course). 

Resultant part (without any supports) is improved enough to be usable but not perfect. Even with 0.15 mm layer improved overhang performance, filets will never come out perfect. If you really must have a filet, the usual solution is to model the bottom part as a chamfer at a printable angle and transition it to the filet curve when the flat chamfer meets the filet tangent.

Also you see the usual Prusa artifact that appears on the exterior at the level of interior solid surfaces. That shows up in the chamfered portion.

 

Click for larger image. When adding media on this forum,  always checkmark link to "media file" to allow forum users to click and see original, full size image.

 

Posted : 12/09/2019 8:00 am
mgoblue
(@mgoblue)
Eminent Member

Thanks for the suggestions!

The print I ran overnight with the "extra perimeters if needed" option checked actually turned out a bit better.  Just for my own sanity, I am running the print again with that setting turned back off (and I am printing the cut off version just to make sure the change to the STL didn't change anything).

I will probably be running a few more tests with varying numbers of your suggestions implemented to see the differences.  Here is a side by side photo of the original part I took the pics of (on right), and the new one (on left). I see an odd defect on both parts, where it looks like there are lines of extrusion that just stop.  There is a small one in the upper right corner of the improved part (arrow in additional image).  I can't tell if this is part of the supports that didn't peel off, or if it is a defect.

 

This post was modified 1 week ago by mgoblue
Posted : 12/09/2019 1:18 pm
richard.l
(@richard-l)
Member Moderator

Wouldn't it be beneficial to use Variable layer heights? I could see .08 layers with Extra Perimeters until it gets to the flat section. It would increase your print time but should give a much better result.

...
Posted : 12/09/2019 1:58 pm
mgoblue
(@mgoblue)
Eminent Member

I might give that a try too - I just sliced with the bottom section at higher detail and the slope at the top with higher detail, and it looks like the slicer doesn't find a need to add support for the big chamfer anymore.  (I have changed the model to only have chamfers now - no more fillets on the short sides, and i incorporated all the suggestions of guy-k2.  I am running a print right now with his suggestions all rolled in, and I will run the variable layer version next.  I hadn't tried variable layers until just now, so still figuring out the interface there.  I wish there was more control on the amount of detail added (effectively determining the layer thickness at those areas), but it is tricky to get it uniform through the section I am editing.

Posted : 12/09/2019 2:31 pm
mgoblue
(@mgoblue)
Eminent Member

Update:

Implementing the settings suggested by guy-k2 helped a fair amount.  The poor overhang appearance improved noticeably.  One thing that actually got worse, however, was an increased amount of warp in the part.  The surface that was on the build plate has a noticeable curvature now.

The trial I did with variable layer thickness also looked pretty good, but the unsupported overhangs are a bit more rough/fuzzy.  As mentioned above, Prusaslicer seems to think that the supports are no longer needed, even though they were still turned on in the settings.  This test used all the same settings that were suggested by guy-k2, with the addition of the variable layer thickness.  This print did take about 30 minutes longer, so not sure if that would contribute to warp.  

 

Here are some more photos.  Right side part is older settings (2 vertical shells, etc. - basically my original settings), center part is the first try at guy-k2 settings (4 vert shells, quality profile, .2 height, all chamfers removed), left part is the same settings as center part, but with the variable layer height applied and NO supports.  Notice the warp at the outer edges on the left and center parts, while the right part is quite flat.  The warp seems most noticeable on the part printed with variable layer height - even in the second image (on the left front corner of the part).

Posted : 12/09/2019 6:59 pm
Robert-mm200
(@robert-rmm200)
Reputable Member

My two cents: Warp is lose of adhesion at the edges.

Try it with a brim. If this is PLA - use a PEI sheet.

Posted : 12/09/2019 7:27 pm
guy.k2
(@guy-k2)
Noble Member

Warping, probably got worse with the longer print time. This is more likely if your print bed isn't sticking well. It is a pretty small part. So, I would expect it to have minimal warping issues if the 1st layer is sticking properly

1. Wash your print plate with Dawn and hot water. Don't touch the surface again with your bare hands. Dry and hold with clean paper towels. Heat and dry on print bed at 85C for 10 minutes. Immediate drying and heating are to avoid the risk of rusting the plate. You typically only have to wash the plate once to remove the factory residue or when the plate gets really dirty again.

Between prints, never touch the bed with your fingers. Use a plastic scraper and wipe down with isopropyl alcohol.

Lowering Live Z another 0.25 mm may also help, but most likely your plate needs its initial water/dawn cleaning. Yes we know Prusa's instructions say otherwise. Real world is the plate needs it once to become really usable. 

 

 

Posted : 12/09/2019 7:30 pm
mgoblue
(@mgoblue)
Eminent Member

I will give the sheet a good cleaning. All the prints I have been doing have been on the exact same spot since I am working with the same slicer setup and just changing settings. 

Posted : 12/09/2019 7:41 pm
guy.k2
(@guy-k2)
Noble Member

Yes, definitely clean it.

Also, remember you can drag the item to a slightly different location before slicing. That avoids wearing down one spot of the print plate. You'll also notice (if you used my actual 3MF) that my custom startup code deposits an easier to remove priming line. Give it a try if you have not already. The startup sequence is different and also takes measures to stabilize PINDA temperature and avoid reposting zits during bed leveling.

This post was modified 1 week ago by guy.k2
Posted : 12/09/2019 7:56 pm
niggybrk6 liked
Share:

Please Login or Register