Horizontal-supported surfaces.

[CCVirginia]

On several models that have horizontal or near horizontal overhang I am getting an odd pattern of single threads, even with support (See picture - center of Guantlet). The horizontal area over the overhang is VERY rough. It seems like the way S3D handles this situation is not the best. I have tried several settings with little impact.

Problem over supports

The model I am using to illustrate this is the "Gauntlet" from here: http://www.thingiverse.com/thing:596966, from the e-nable initiative. The first picture uses the "built in" supports, but using S3D supports was about the same. Note the "wiggle" pattern and that even on the same layer there is not enough plastic to cover the holes.

I reprinted this model with a process just for these layers using 100 outline shells so there is no infill, I also reduced speed about 25%. This uses S3D supports.

Better

It is now much better (not ideal, but better), indicating some kind of infill problem. Perhaps the s3D algorithm could be improved for this kind of situation. If there are suggestions to make it better without a new (and very slow) process, please post!

Using: Makergear-M2, V4 Hot end. Simplify 3.01

[ClassicGOD]

With default settings and positioning of this model looks like S3D is generating support that is parallel to the bridging so it fails. Set your dense support layers to 3 or change your support angle to 90 and try again.

[KC_703]

+1 on rotating the support angle so its not parallel to the required bridging.

Can also experiment with dense layers to create more of a "platform" for the first layer of the overhang.

[CCVirginia]

I suspect this is an "infill" issue, not a support issue for 2 reasons:
1) The much better print used the same support but changed the outline shell layers to eliminate infill. While there are a few rough spots (as would be expected), this print has a "tight" bottom.
2) There are 2 versions of this model - one with support in the model and one without (which I printed using S3D support) - they look about the same.

So, my conclusion is that S3D is not printing a viable pattern for the suspended/graduated bottom and that this is an issue with the slicing algorithm. It is not clear that there is any way to impact such a suspended layer without changing all of "infill", which does not seem the right tuning parameter for this common situation. This is really a special kind of bottom. Regardless of the bottom, the graduated pattern (which is tough to get right) does not have an effective fill pattern. Since the 100 outline shells fixed it, we can see that a solution is possible.

So my request is to address the slicing algorithm. My "work around" question is: what is the best way to improve the pattern for such a graduated surface, particularly on a suspended bottom?

[mpdcdc]

I'm seeing this issue too. The base and top surface of my prints are great, but wherever there is an underside that has been supported by the generated supports, the finish is horrible, and not because the supports have been removed, but because the pattern used isn't the same as it is elsewhere for a surface or an infill - it looks like loose random fill within the bounds of the outline.

Any suggestions?

I'm using a Taz 5 with fairly cheap 3mm PLA. The finish elsewhere is generally excellent - it's just the underside of overhangs that have been supported that don't look to be finished properly.