Honeycomb infill weirdness

[JimG]

I have a strange thing going on when I print flat pieces with ABS. For instance, a 2.0 mm flat base for an electronics enclosure gets printed in this order on my rig (10" i3v):

1. Base layer, 0.4 mm thick, printed on 45 degree diagonals
2. Layer 2, 0.2 mm, on the other 45 diagonal.
3. Layer 3, 0.2 mm, back to the first 45 diagonal
4. Layers 4, 5 and 6 are honeycomb infill
5. Layer 7, on 45 degree diagonal, has to "bridge" the openings in the infill. This is where the excitement begins...

Instead of sagging down into the honeycomb voids, the layer 7 threads bow upward. My first impression was that heated air was getting trapped and puffing up. But now I don't think this is the explanation because the worst threads for bowing occur before an individual honeycomb cell is even starting to be closed up.

This bowing up is worse towards the middle of an object, where the diagonals are longer, than nearer the corners where the diagonals are short. This layer ends up having the texture of alligator skin, with lots of little sharp peaks.

The ABS cools in the upward bowed position. When layer 8, which swaps diagonals and crosses at 90 degrees, goes over the layer 7 threads, the nozzle hits the hardened plastic. I can hear and feel the collisions. Layer 9 swaps diagonals again, and is not quite as rough as the previous two. But the finished surface does not look very good.

My present working theory is that hot end (0.4 mm hex) is becoming cooler over the course of placing the long diagonal threads. This is based on the observation that the short diagonal threads don't have the issue. There is a brief stop at the end of each diagonal where there is no filament pulling heat out of the hot end. (I have printed the ABS up to 250F and still had the problem, FWIW.)

I'll also mention that the problem is much worse with ABS than PLA. My PLA bases are a little rough, too, but much more acceptable than the ABS.

No doubt someone else has seen this. What is causing it? How do I fix it?

Jim

[AbuMaia]

Do you have a fan trying to cool the extruded plastic? I've experienced the same thing with the bridges over the honeycombs puffing out, and thought it was from my print fan blowing air into the cells. That would explain how they bow out before the cells are closed off.

[JimG]

No cooling fan. There is a fan on the extruder but it is not directed at the printed object.

Jim

[gmay3]

That is very strange! I don't think I've had a bridge curl upwards towards the nozzle before.

Assuming you have the makerfarm supplied extruder fan, one thing I might try is to reverse the direction of the fan by unscrewing it and flipping it. My thought is that if your fan is blowing from the front of the extruder to the back of the x carriage, a small amount of air could be leaking down on to the print. I know this happens because if I have my extruder in the middle of the bed with the fan on, the bed will not heat up since the air is cooling the glass. Try flipping the fan so that the air pulls from behind the extruder and exhausts out the front of the printer.

[JimG]

I will try this. But that doesn't explain why the problem only happens on the long diagonals.

If I change the object by adding a hole in the middle of the flat piece to cause the diagonals to be shortened by half, then there is no issue on the shortened diagonals. The full length diagonals on either side of the hole bow up as usual. So that suggests it is not the fan, nor the geographic location on the bed, but rather just the length of a single extrusion thread.

Jim

[gmay3]

Hmm, do you notice that the thread being extruded is coming out cleanly without sticking to the side of the hot end tip while bridging?

[JimG]

I have tried to look closely at this, but couldn't really tell. It does behave like the ABS is glopping up as it comes out on the longer threads. But it doesn't do this on the shorter threads??

The cone of my hot end does have some ABS stuck to it (the result of an unfortunate break in USB comm's in the middle of a print). But as best my eyes can tell that stuck-to-cone ABS is not contacting anything on the print.

Jim

[gmay3]

Yeah I guess what I had meant is if the thread was sticking to the angled sides of the cone shaped nozzle, which it seems like it isn't.

You could try varying the print speed knob during this problematic layer to see if an increase or decrease of print speed makes this bridging more flat.

Also, if it's possible, maybe a picture of the issue would reveal something we didn't think of before.

Good luck!

[CalifDan]

I was curious what the infill settings are. If the infill is low, perhaps the layer is not adhering to the lower layer. This would be more evident on a longer run than a shorter run. It might even be a temp issue if it is lack of adherence.

[JimG]

Infill is set at 20%, honeycomb. I have been thinking about trying the rectilinear pattern to see if the behavior is any different.

Jim