A print constantly fails in an odd way

Hey,
I think I’m hitting some slicer or firmware bug. I’m using X1C and Bambu Studio 1.9.7.52. I have one print that constantly fails in the same way even after re-slicing:


I was trying to track this down by inspecting thelayers in Bambu Studio after slicing. I also tried to use external code viewers, but I can’t find the reason.

I’m trying to print this:

Observations:

  • it seems to be printing up to and including layer 17
  • where it fails is, I think, layer 18
  • notice odd issue in the inner ring - there is this part that looks like thinner filament lines
  • the last printed layer in the outer ring does not have the walls printed but it does have the solid infill printed which comes after the walls
  • the print somehow fails completely after that
  • the print always fails at the exact same way

I’m attaching the model
ams-qk-f.3mf (896.4 KB)

Please let me know if you have any idea what’s going on. Thank you.

PS. Spaghetti detection never detected this.

- Daniel

Welcome to the forum. Thanks for the pictures and .3mf file as they really helps for explanation.

As you observed, the issue happens at layer 18.

My guess is the model is failing because of a combination of a couple problems.

Low wall count that is trying to print on a tapered cylindrical surface. 2 walls doesn’t give much if any surface for the next wall to print on. Being a cylinder feature doesn’t help. Once a wall fails there is no recovery and you get spaghetti.
image

To fix the issue I would,

  • Up the wall count to 3-4
  • increase the infill percentage and change it to gyroid or cross hatch.

image

Here is the modified file.
ams-qk-f_mod.3mf (917.5 KB)

Spaphetty detection struggles with dark filament.

Hey,
Thank you for the answer!

What I forgot to mention is I printed it once successfully, but I did not save the 3mf to have a comparison and I do not know of a way to retrieve my past prints.

I will definitely try to increase the number of falls, but I’m trying to find out if the issue is indeed in the model or if it’s a bug. I printed numerous cylinder-shaped models with similar settings with no issues (even ball-shaped without supports). Also, what made me curious is why it’s a complete spaghetti - I would assume at least some parts of the model should “try” to print, like solid infill on top of existing infill, or the top surface on the solid infill (which layer 18 is for the outer “ring”. Also the infill in the center should continue printing, or really some other parts of the model, yet it’s a complete spaghetti. What’s more, the outer wall (around the whole model) did not print, which I guess could be because of what you mentioned, but the solid infill did:


and then nothing on top of it anymore.

There is also this:


(which I do not see how this could happen by examining the sliced model)

Or this:

So, the fact that:

  • it always breaks in an identical way
  • it’s a complete spaghetti (like it’s skipping a layer?)
  • no filament is extruded onto the printed part in any part of the model
  • I printed it successfully once before and it was a flawless print (but I do not have this 3mf, don’t know the difference(s))
  • there are other imperfections I can’t explain
    made me think it may be some sort of bug.

I’m pretty sure I may just change some setting or settings and make it print, but I’m also trying to figure out what’s the exact reason of a fail like this (where at some point nothing more is being printed on top of the previous flat surfaces)

- Daniel

Change the infill from grid to something else like gyroid.

Grid infill is prone to causing the head to impact the model as the lines crisscross over already printed lines causing a bump each place they intersect.

Yeah, rectilinear infill can cause the printing to fail by kicking the model off the build plate, but here the model was still sticking to the bed, it did not move and was sticking still to the bed (rectilinear infill caused several failed prints for me in the past). The reason I’m not using gyroid, even if it’s the best infill IMO, is it’s pretty “violent” to the printer while printing with high speeds and causes a lot of vibrations and loud sounds (my wife almost hated me on a 2.5-day long print with gyroid infill :stuck_out_tongue: ). I’m keen to try things, though, I may try gyroid, although I do not believe infill is a problem here.

- Daniel

Interestring, I just found print history in… Bambu Handy (is it not available in Bambu Studio at all?). It looks like you can re-print the prints up to 3 months? I do not see any older prints, but the second to last on this list was this successful print. I started re-print in Bambu Jandy and once it downloaded, I removed the SD card and connected it to my PC to retrieve the 3mf (with gcode and settings) from the cache folder. So I have the gcode and the settings now of this successful print (and the settings look identical):
ams-qk-f.3mf (2.7 MB)

I’m going to:

  • re-print this one to see if it succeeds as it did before
  • compare gcode of this print from 3 months ago and the current code (the one that fails) to maybe spot the problem
    I also believe I was using an older Bambu Studio version then, so I may try to downgrade.
    I’m really curious why it fails now in this odd way but printed without issues back then.

- Daniel

So I re-printed the successful one (from almost 3 months ago) from the print history using Bambu Handy and what do you know:

The imperfection in the inner ring that I mentioned is also not present as well as other imperfections - photos taken at the beginning of layer 18 on a running print, I was watching it closely this time to see what’s going on if it fails:


So, as I mentioned earlier, I took the 3mf file from the cache folder from the SD card, unpacked it, and retrieved gcode file. I then exported gcode from the failing 3mf in Bambu Studio (the recent one that I was using that failed 3 times in a row in an identical way, I’m using Bambu Studio to run the prints through the cloud). I compared both files using a diff tool I was expecting to find some differences that would explain this behavior. What’s more, I expected the gcode to be this different that I would have a hard time to diff between the files. To my surprise the gcode of both was identical except for the filament settings - previously I was using the Generic PLA profile, but later I created a filament-specific profile (DevilDesign PLA) using the manufacturer’s specification pdf. I printed so many things using this profile without any issues. Anyway, the only differences in gcode are these related to the filament profile:

Old gcode (working) New gcode (failing)
(empty) ; different_settings_to_system = ;compatible_printers;filament_max_volumetric_speed;
; filament_colour = #FFFFFF ; filament_colour = #000000
; filament_ids = GFL99 ; filament_ids = P0f3f78c
; filament_settings_id = “Generic PLA” ; filament_settings_id = “DevilDesign PLA Standard @Bambu Lab X1 Carbon 0.4 nozzle”
; filament_vendor = Generic ; filament_vendor = DevilDesign
; nozzle_temperature_range_high = 240 ; nozzle_temperature_range_high = 235
; nozzle_temperature_range_low = 190 ; nozzle_temperature_range_low = 200
; temperature_vitrification = 45 ; temperature_vitrification = 50
M620.1 E F299.339 T240 M620.1 E F299.339 T235

That’s all, these are the only differences between the perfectly fine print and the failing print with lots of spaghetti. There is no difference in gcode in the actual layers.
I feel stupid, but maybe you have some idea. I still feel it’s a bug, likely in the printer’s firmware, can;t explain this in any other way.

- Daniel