Houston, We Have a Problem...

[gwhite]

I've had a MakerGear M2E for about 2 months, and have been printing away without any major fails. I decided to try to make a vacuum nozzle that would allow me to suck a good bit of the air out of a large zip-lock bag for filament storage. I've tried commercial "space bags", but they are all way too large. In any event, the model looks like this:

I fired up the printer last night before going to bed, and everything appeared to be starting out just fine. When I got up this morning, I went down to the shop anticipating a lovely prototype, and I found this instead:

I checked the S3D print trouble shooting guide, and I didn't find anything under "rats nest"... My best guess is that it lost bed adhesion when it was about 5 mm up. After that, it was just squirting filament in the air, towing the mess around the build plate. I found the two pieces from the start of the print (on the lower right) off the bed in the back. The skirts were misplaced as well. I've been printing using hairspray on the glass build surface, with good success. I've even printed a couple iterations without refreshing the spray, and in this case, I was printing in areas that hadn't been used before, but were ~ a week old.

There was one huge solid lump that looked like it had spent a while stuck to the nozzle, which eventually fell off. The nozzle had some filament curlicued around it, but there was a good coating of melted filament all over the hotend:

The only other factor i can think of is that the filament (eSun PLA+) had been on the machine for a week or two. My shop is kept at around 40% relative humidity, and I haven't had any weirdness I could trace to filament moisture before. However, this is about as long as I've had filament out.

I cleaned everything up, and am printing the smaller piece now. This time, I flipped the build surface to print on polyimide, which generally has really good adhesion with the PLA, but is harder to get parts off of. The print is about 1 cm up, and so far it looks fine.

[gearsawe]

looking at the last pic. something does not look right how the filament is coming out. Filament should eject straight down out of the nozzles. It should not curl back around or appear to be shooting off to one side. This could be the result of a damaged tip or a small clog.

Second This is the reason I dislike printers which move the entire bed. As parts get taller they came be shaken of the build platform. Prefer those which only move the bed in the Z axis. Now seeing some which don't move the bed at all but move the entire gantry up and down.

That is a great looking rat nest BTW. They are very hard to model and print.

[gwhite]

Well, attempt #2 was interesting. It got up to a couple cm, and everything was fine. However, because of the way it is sliced (and because I have ""avoid crossing outlines" turned on), it occasionally moves along the perimeter at a high speed to get to the next place it wants to start. I noticed a crackling noise that sounded like a small version of the creaking you get when you pry a part loose. I think the part had warped slightly, lifting a bit, and the nozzle was dragging. Each time it made a fast traverse, it was slowly knocking the part loose. I had to go out for a bit at that point, fully expecting to find a mess when I returned. Sure enough, this is what I found when I got back:

It clearly popped loose, and as the extruder went round & round, it waltzed the part off to the side. I expect this problem gets worse & worse with tall parts, because the leverage to knock it loose gets higher & higher as the part grows.

I ran it again, but I added a lift during moves so it won't drag. I also cranked up the bed temperature from 60C to 65C. That worked fine. No Cheez-Whiz effects, and the print was still securely attached to the bed at the end of the print.

I have mixed feelings about "avoid crossing outlines". The preview shows that it cuts way down on a lot of seemingly random traverses across the open middle of the part. Instead, it looks like it's just doing lots of random traverses, but over the part instead. What is annoying is that there is no need for traverses at all, but with a circular part it insists on doing the infil in a couple separate pieces for each layer, instead of doing them in one pass. I have the infill set to rectilinear, and that means there are places where the infil is aligned with the walls, and other places it's perpendicular. For tubular parts, it would be nice if there was a "radial" infil option. I'll have to play with some of the other infil pattern settings to see if I can find one that doesn't involve lots of unnecessary traverses.