Why can't I print this (bigass picture warning)

[insta]

IMG_20150513_084552.jpg

(1.1 MiB) Downloaded 617 times

IMG_20150513_084535.jpg

(1.29 MiB) Downloaded 617 times

IMG_20150513_084531.jpg

(1.35 MiB) Downloaded 617 times

IMG_20150513_084528.jpg

(1.03 MiB) Downloaded 617 times

2015-04-26.jpg

(546.29 KiB) Downloaded 617 times

As you can see from the last picture, flat-plane extrusion is dialed in nicely. Why am I having such trouble on the edges of the part? This is like the 4th time it's done it, on just this part. It's not a damaged STL, I've run it through Netfabb with no errors. The GCode looks right, as far as I can tell. Perimeter walls are dialed in to within -0.01 / +0.00mm of expected.

If anyone wants the STL I can share it for troubleshooting, but it's not something I'm going to freely distribute as it's not mine.

[Slipshine]

shrunken pictures to help.

pic1.jpg (34.02 KiB) Viewed 14548 times

pic2.jpg (37.12 KiB) Viewed 14548 times

pic3.jpg (39.52 KiB) Viewed 14548 times

pic4.jpg (34.5 KiB) Viewed 14548 times

pic5.jpg (42.75 KiB) Viewed 14548 times

[Bratag]

What orientation are you printing in? Does it make a difference if you flip it over (assuming it would be a lot more support needed).

Does look decidedly strange.

[insta]

It doesn't make any difference what orientation it's printed in

[Bratag]

It doesn't make any difference what orientation it's printed in

Hrmmm. How many perimeters are you using. My thought is that its trying to squeeze and extra perimeter in while at the same time trying to step over and up. If you drop to 1 perim does that make a difference (may not be ideal for the structural integrity you are hoping for)

[jsc]

Send me the file and I'll try it.

[jimc]

it looks like way too much heat. anything odd with your heat settings? what plastic is it?

[insta]

It's white ABS @ 220C with HIPS support (black). It's 3 perimeters, 20% rectilinear infill in slic3r @ 0.1mm layer height.

jsc, PMing you the file now...

[jsc]

Well, here's a few thoughts after attempting the front shell.

First of all, the model quality itself is not that great. The mesh looks low resolution, and S3D detects some non-manifold edges. There are strange artifacts here and there. The wall width is inconsistent, causing some sections of the wall to go from infill to perimeter and back. I think the main cause of the bad rash effect is due to getting machine-gun infill in overhanging sections where the perimeter is none-too-solid to begin with. I see several areas where the width has dropped enough to go to a single perimeter with infill; that's going to have a visible effect.

On the front shell, especially, there's that large cutout, too, that is causing issues. I'm not doing this as a dual extruder print, so with the separation required, the entire area above the support in the gap is curling up. It's acting, not unexpectedly, like a large flat print that doesn't have good adhesion to the "bed", and there is quite a bit of curling/sagging for about 5mm over that area. If you're doing this with HIPS support, make sure you specify zero separation and nearly solid interface layers.

Also, 220 seems quite low for ABS, which I usually do at 240-255. Even PLA I do at 225.

(What in the world was used to model this? Why was the modeling done at a 30 degree angle?)

If I wanted to get a nice print out of this, I'd revisit the model. It's not high enough resolution for a prototype, but has too many difficult features for a 3D print. Not every shape can be easily 3D printed. High degree overhangs bad; inconsistent wall widths bad; large unsupported areas with exterior overhangs bad. It might almost print better if you flipped the whole thing upside down and just stuffed it full of support.

[jsc]

I mean, just look at this thing: