Poor horizontal-supported surfaces.

[Jules]

Oh, hey, I just sliced the one with the included support. You definitely don't want to use that. I was wondering why you guys kept referring to "bridging" when it shouldn't be bridging at all. Well, with the included support model, it doesn't know that it's support and is in fact trying to bridge on the bottom layers. Plus, the "support" is actually a mostly hollow box with beams (scrub the End slider at the bottom to see what the interior looks like) and it will be trying to actually run curves over thin air between the beams; with only three top layers the quality of the top of the support structure will be abysmal. Use S3D supports with 3 dense 80% layers and one top layer separation, and it should do a much better job. Although, it kind of sounds like you already tried that?

Totally agree with this - do not use the one with included support. If you run it through S3D with a normal "Generate Support" setting, it actually fills in all the gaps in the designed support.

Try just letting S3D generate the support for it, using jin's suggested settings for dense top layers - it's going to be a lot smoother on the inside.

I think the interior of that print is curved, not straight, (hence the "terracing"), and it's organic, not a straight angle, (hence the "wiggle"). That'd be a tough one to try to print.

Another thing, after S3D has created it's support, use the scrubber under the picture to see which direction the support has been created in. Ideally, you always want to angle your print on the plate so that the support runs at a 90°angle (perpendicular) to the direction that the first solid layer over it prints. That's not always possible, but i found that if you angle the print on the bed 45° from the x axis, you get the best support for any holes, whether they run parallel to the X axis or the Y axis.

The support for that particular print is definitely improved by the 45° rotation. (Eliminates the stringing in the long slots.)

[CCVirginia]

I reprinted this with a "process" layer for the problem area. I configured the outline shells set to 100 so there would be no infill in that area and also reduced speed a about 25% as well as set the supports as "insta" suggested. As can be seen from the picture it is much better, indicating that the way the slicer (S3D) handles this kind of situation could be better. I'm not even sure which setting is effecting the bottom of a suspended horizontal section (is it "bottom" or "Exterior Infill"?). Some other slicers may do better or perhaps there is some way to configure just this condition. My guess is that this is an infill problem. The print time increased to about 12 hours (I think it was 4 last time), not sure why as going from 35% to 100% fill for a few MM should not make so much difference.

So for now I consider this "closed", but if there are ideas for better handling of horizontal supported areas, let me know. And, thanks for the help!

[jsc]

Did you see my photos of my test on page 1? Were those substandard in some way?

[sthone]

100_4099.JPG (56.11 KiB) Viewed 11096 times

100_4105.JPG (57.97 KiB) Viewed 11096 times

Crappy pictures... but as suggested I tried the gauntlet upside down with S3d supports and overall I think it came out mush better. It still has some zits and things but that's probably just my ooze setting and such.

-Steve

[CCVirginia]

jsc: I noted your infill is lengthwise where as mine is angled - that may have helped. It still looks like the layer is not fully filled, it is probably good enough but I'm trying to understand what is going on. Do you know why yours looks better than mine?

sthone: that does look very good - I always get a very rough surface over the supports, yours seem to have come out much smoother than mine ever have. I was tying to keep the top nice. In fact, your entire print looks smoother.

[jsc]

I don't know why your angles would be any different. The bottom layer infill takes the same direction regardless of the print orientation, because it treats each terrace as a bridge (contrary to some of my earlier statements). You can see what it's doing in the gcode preview:

gauntlet.jpg (322.27 KiB) Viewed 11079 times

The blue indicates a slower speed, which means it's trying to bridge. It's taking the shortest path, which is generally vertical. Does your preview look like that?

What do you mean by it doesn't look fully filled? You mean you can see gaps between the threads? That happens when you do any bridging, the lowest layer doesn't usually get pulled completely taut so ends up a little loose. In fact, I purposely set my bridging extrusion rate lower, which in theory should make the bridge tighter, but also makes it less "full". It's all tradeoffs.

[CCVirginia]

With S3D Supports- rectilinear

WithS3DSupportRec.PNG (342.03 KiB) Viewed 11031 times

[quote="jsc"]The blue indicates a slower speed, which means it's trying to bridge. It's taking the shortest path, which is generally vertical. Does your preview look like that?
quote]
Yes and no. When I print without supports it shows the bridge speeds. With S3D supports it does not - your original assumption.

With S3D Supports @ 90 degrees - no bridge speed.

WithS3DSupport90d.PNG (349.25 KiB) Viewed 11031 times

I don't see any way to control the layer over a support.
I also note the direction of the supports changes with the direction of the print on the build plate and there is no way to change this, an active topic on the S3D forum.

While we can work around this I am still somewhat convinced S3D is not slicing this as well as it could, I started a thread on the S3D forum here:

https://forum.simplify3d.com/viewtopic.php?f=9&t=3357

When I print with "rectilinear fill" it still comes out angled - so I guess I could change the fill angles for the whole model, but it is not taking the shortest distance.


This is a bridge (without supports)- it would never work this way:

Without supports

WithoutS3DSupportRec.PNG (311.24 KiB) Viewed 11029 times

When I get some time and profiles for some other slicers I may compare them. Right now the day job is calling!

[jsc]

Your "supports" picture shows a very large support spacing. You need to use at least 3 dense support layers of 80% or more to get a good support layer if you're going to put curves on top of it. I also prefer a 2mm grid if the model is complex. Also, I have my bridging speed set very low, like 25% of my usual 80 mm/s, which is why there is such a speed differential in my screenshot.

[CCVirginia]

Your "supports" picture shows a very large support spacing. You need to use at least 3 dense support layers of 80% or more to get a good support layer if you're going to put curves on top of it. I also prefer a 2mm grid if the model is complex. Also, I have my bridging speed set very low, like 25% of my usual 80 mm/s, which is why there is such a speed differential in my screenshot.

Yup, I'm also using 25% for bridging - so it does not look like it used bridging over supports. Perhaps you tricked it into going slow some other way?
I think I'm getting "profile overload", I have dense support layers someplace! I have not done the support spacing - good hint.
Wow - you run fast - I am starting with 3500 mm/min (so 58/s).

But, none of this impacts the infill - which seems to be the issue.

[jsc]

Zip up your factory file and attach it so I can see what you're doing differently.