simplify 3d v2.2.0 just released

[jimc]

thats very strange. i have never seen it do that.

[Tim]

I've seen that support thing. I didn't realize it was a bug. The first time I printed one, I was surprised it turned out fine. The first layers were a mess, but it managed to cling to a wall and sort itself out by the time it was needed.

Depending on how far out the support extends over the air, yes, it can recover. I printed one set of chess pieces that way, and they all turned out okay with only a bit of mess around the bottom of the support. But the worst-case scenarios make it pretty clear that it's a bug, and all they'll do is end up dripping filament all over the print.

[jimc]

so i got your pawn model from thingiverse and played with it a little. i tried a bunch of different support settings and what i found was that the grid resolution is too high and or you have inflation distance turned up. the inflation distance just takes whatever the support is and just blows it outward. the larger the grid pillars are also the larger the overhang from the model. the pawn is a small model so i turned the pillars to 1mm for a more accurate support and it worked fine.

Screen Shot 2014-12-09 at 11.58.13 PM.png (416.6 KiB) Viewed 7055 times

[jimc]

here is the exact same thing only changing the support back to the default 4mm. the grid is just too big and causes overhangs.

Screen Shot 2014-12-09 at 11.59.50 PM.png (417.66 KiB) Viewed 7055 times

[Tim]

As I said, the S3D guys told me to reduce the pillar size (I never messed with the inflation distance), and that does alleviate the problem somewhat.

But in the attached picture, which is just a closeup of what should be the same slice of the same object as you did, you can see that the problem doesn't go away, and there are still issues like parts of the bottom of the support pillar which are in no way attached to the structure---they are just free hanging in mid-air. In other places, there are weird gaps in the support structure.

I have seen other objects in which the bottom of the pillar follows the profile of the part underneath the support structure. If the object surface falls away at an angle, then the support structure falls away at an angle, too, maintaining the separation distance that is plugged into the support profile. If you look at the example slice on the S3D tutorial page with the Easter Island head, that's certainly working the way one would think it ought to. It's undeniable in this case (although, for reasons I don't understand, everybody except me seems to be denying it. . . isn't anybody going to back me up here?) that the bottom of the support pillar has been made flat, which violates the separation distance rule. It's just broken.

[jimc]

hmm sorry i didnt notice you said they already told you to turn down the resolution. i kinda figured you would have tried that already.

[jsc]

I agree with you, Tim!

What looks like is happening is that the support is calculated in blocks of your grid resolution, and the bottom "attaches" to whatever part of the model appears below in that grid. If it's just a teeny bit of the model intruding into the grid block on a sloping surface, the rest of the flat bottom of the support start will be suspended in mid air.

Adding some inflation gets rid of some of the missing layer in the support: