MakerBot Industries

Nope I can't say I have tried that. But I wouldn't think that would be a problem with this object because the space between the hole and the outer wall of the object is not really that small. So SF should not be omitting infill around that area.

Ill give a zero shells try a bit later at this point it wont hurt to try.

I should add it is only doing this on the top two infill layers. The bottom two are fine when I watch them print.

So I tried the no extra shell print just for kickers, and I must say I was pretty shocked. Now let me just state that I know about the SF bug where if you have too many shells enabled it will skip the infill and the shells in parts of the print that are thin. However with this print it doesn't look like its skipping anything its just not filling appropriately.

Here is the way it printed with zero extra shells. Don't mind the overlap in the top right I forgot to lower Infill perimeter Overlap (ratio) before I hit generate gcode.

I find this really puzzling because there is clearly enough room for one extra shell to be there even two if you really wanted. But the infill is suffering once you turn on extra shells. It would seem that the extra shells part of SF is more broken then I had previously thought. Are most of you guys printing with zero extra shells all the time? If so what are you using for infill settings to make objects strong.

I have basically gotten my prints to the point that I'm just nit picking. The biggest thing for me was getting the first layer height just right so that it was in sync with the rest of the layers in the object. That way you don't have the first 2-3 layers sort of drooping outward on your print. Bellow is a side picture of the above print that shows all the layers evenly sized. The only thing I wanted to perfect was the Infill and getting as smooth of a finish as I could.

Definitely printing to be proud of.

I am just seeing one side of the printed part so it maybe happening on the other side, but in my prints there is always a tiny gap where the outermost layer starts and ends (or where the filaments comes out from the next to last to the last outer layer and then goes inside to fill it).

It may be just a coincidence that there are no such gaps on this print or that you actually have found a way to eliminate them. I do not have any clue what needs to be changed in SF in order to reduce them or make them less visible, let alone make them go away for good. They sometimes line up creating a vertical gap or are spread around the part nearly at random - one such tiny gap for every layer.

By the way, the layers on that yellow part are indeed beautiful.

@Riftwire - I dunno, I have seen the problem before, and turning off extra shells fixed it. I guess extra shells is a little more bugged than just the too thin part thing. However, it does not do this all the time. I have not been able to determine the set of circumstances that causes this issues to appear. But it does always seem to happen to holes like the ones pictured. Kind of weird, because the next object you print, that has a similar hole, will print just fine too.

Have you tried generating the object with SF-39 or later?

Oh, and to answer your question… no, I usually leave shells on. Objects break too easy without it. I, just like you, and I assume most others, only turn extra shells off when the bugs appear.

Yeah it seems that the shells part of SF is more broken then I had thought. I talked with Nick again today about this and Matt the repg guy and we were looking at the gcode file in some visualizing software and couldn't see where this was happening.

We have a bunch of theory's but I need to test some out and see what happens. Its pretty strange because at least from past experience it shouldn't be happening on a part like this as there is plenty of space between the perimeter and those holes.

I have not experimented with any other versions of SF I only have 31, 35, and 40 available to me. 40 changes the mathematics a bit and would force me to completely relearn a bunch of stuff so for the time being I have just stuck with 35.

Well, I have seen it before. At least twice, on parts where the "to thin" bug wouldn't apply. So there is definitely something going on. I didn't do much investigating though, I just turned extra shells off and forgot about it. I have no idea, if the times it occurred for me, if it would have shown up in a gcode simulator. Strange that it didn't for those guys. For all we know, it could be an issue that has been fixed. I haven't tried 40 yet either. I tried 39 for a few minutes with repg24, but reversal was replaced by dimension and that stuff doesn't work without 5d supporting firmware.