How did that happen?
So, a weird thing happened to a print I left overnight.
All those layers were supposed to be on top of each other, but somehow at some point there was a ~22mm displacement in both X and Y axes.
What could have caused this? The piece was still well stuck to the plate, the plate was at its proper position on the bed, and the printer reported only an extruder clogging (which happened long after the displacement). I saw nothing that the head could have bumped into.
I'm wary of trying to print again because this is a long print with XT-CF20 which is quite expensive, and I have no idea what caused the printer to do that.
RE:
Simply post an image to the Prusa site, don't use 3rd party hosting.
What everyone else sees:
RE: How did that happen?
Darn that's devious, I see the proper image. Thanks, I'll try to post it again (I can't seem to find a way to edit the first post).
RE:
There are a few things that can cause print shifting. Collisions the most obvious. Power fails, another. A third is using a host, like Octoprint to send files from a server to the printer: one lan blip or slow i/o and the print is toast. Since we weren't there when it happened, it's down to guesswork.
An image of the entire part might be helpful to look for any small melted spots that would indicate a collision. And anything that would help get a scale of the part.
Common issues that cause collisions are part warpage and a part corner lifting off the print sheet. Since you're using a higher end filament, are you in an enclosure?
RE: How did that happen?
There was probably no power fail, I don't know if the printer would have reported it but I have lamps that turn on whenever there is a power shortage (even a microcut) and they were off. It was printing from the UBS drive, no octoprint.
There was no warping, the bottom was flat and stuck to the bed so it must not have been knocked and displaced - unless it's possible for it to stick flat again after being moved?
I do have an enXLosure but it was open (doors and lid) as I usually leave it.
As for other pictures, unfortunately I threw away the part but the horizontal bar is about 80x20mm, the total height was I guess less than 10mm, the shift is 22x32mm (not 22 in both axes as I mistyped in the OP). I've had some small displacements on thin, tall pieces as the nozzle moves the upper part, but that part is low and flat, and the shifting is really big. Very puzzling.
RE:
The infill of the lower half looks pretty weird, almost like the printer wasn't moving upward. Any chance something was blocking the Z axis? It could be Z stuck for a few passes, then started printing again a few layers above the part ... but that usually gives more spaghetti.
Actually looking at the top part, there seems to be a bevel to the infill, right side weird, left side not so much. A stalled Z motor would do that.
RE: How did that happen?
I think there were only a few lines of infill on the lower part, the grid is pretty shallow. You're right about the top, the right side looks weird. Maybe a trick of the light?
I checked again, but I couldn't see anything around the printer that could have crashed on the head.
I tried another print, a thinner PCTG part printed with another extruder, no problem. So if there's a mechanical problem, it's in the toolhead.
RE: How did that happen?
I'm thinking its in the Z axis, but really hard to say. Might be something in the umbilical binding, a broken wire, cat jumping onto the printer...