Overhang angle for supports not accurate?
I have run a number of test and am getting very strange results. I have "Generate support material" checked and "Auto generated supports" checked. I built a set of blocks in Fusion 360 that go from 0 degrees up to 60 degrees in 5 degree increments measuring the angle from the build plate. I have also run the identical test using overhang tests in Thingiverse. I need to set the "Overhang threshold" to the following values to trigger the respective block:
- 21 degrees = 15 degree block
- 26 degrees = 20 degree block
- 31 degrees = 25 degree block
- 36 degrees = 30 degree block
- 44 degrees = 35 degree block
- 51 degrees = 40 degree block
- 57 degrees = 45 degree block
- 62 degrees = 50 degree block
- 66 degrees = 55 degree block
- 71 degrees = 60 degree block
I am totally perplexed by this behavior? Is the threshold and estimate? The supports seem to trigger correctly in Cura when I try the same test using my model or an imported overhang model. What am I missing? Thanks in advance. I love this slicer, just trying to make sure I am using it correctly!
Best Answer by --:
It's a known defect in Prusa Slicer. There are bug reports about it on GitHub. And you've only discovered the tip of the iceberg; it is a lot worse than this. lol.
RE: Overhang angle for supports not accurate?
@tim-m30
Thanks Tim, I was afraid of that. I suppose I can work around it for now.
RE: Overhang angle for supports not accurate?
That last one I fought was when it supported areas under what I had set, and areas that started above what I had set. So there were gaps in the middle of curves that should have been supported. Adjusting the angle until you get what you need is the work around. But there are other things to watch for, like single support towers that are only 2x2 or 2x3 from the bed to something 150 mm off the bed... these will usually fail, so review the supports carefully for these types of hidden hidden gems. One slice even gave me a single extrusion support that was 80 mm high (one extruder width row of plastic, 3 mm long - lol - and it was supposed to stand up for 80 mm).
That said, while not perfect, it usually does the job fairly well. I find using 0 (auto) works most the time, and low values like 4 to 8 work to get me the basic must have supports (all the layers where the slice starts new regions of a part mid-air).
RE: Overhang angle for supports not accurate?
So this is clearly a bug, it has been reported on Github, no one from Prusa has responded to either this thread or the thread on Github....is their any plan to fix this or do we just continue to eye the overhang threshold since the slicer seems to either ignore it or calculate it incorrectly?
RE: Overhang angle for supports not accurate?
So this is clearly a bug, it has been reported on Github, no one from Prusa has responded to either this thread or the thread on Github....is their any plan to fix this or do we just continue to eye the overhang threshold since the slicer seems to either ignore it or calculate it incorrectly?
GitHub puts it in front of Prusa, and if they elect to, they put the fix on their roadmap. Actual time to implementation depends on their evaluation (perhaps they believe other factors are at play) and priorities (big push on to support the Prusa Mini). I'm waiting for the filament sensor bug to get patched myself.
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