Hey all.
I have a problem related to 3D printing that I am hoping for some ideas, perspectives and possibly a fix for. After a week of experimentation I am no closer to a solution, so hence I come here in search of answers. Below is a somewhat detailed description on what I am observing.

Problem:
It appears as though specific layers in a model I am printing are deformed for some unknown reason. I have tried several things (see below) to figure out what is causing this issue. It should be noted that I have printed about 50 different models, both toy models (e.g. cats and dogs) but also robotic parts with requirements on dimensions, structural integrity, etc. So far, the machine has performed well.
Fig 1.jpg
Fig 1: Side view. Note how the layers are malformed. It could be that only one layer is malformed, and the following is deformed because they miss support.
Fig2.jpg
Fig 2: Note also in the inside corners that the melted filament is dragged across as a spider-web almost.

I am not trying to print a somewhat complex model. I have previously printed an almost identical model on a different machine (a Cube 3D), but now I am using a FlashForge Finer. Its not the exact same model, but the complexity is similar.

Model details:
The thinnest wall is 3mm, with a lot of holes and some over-hanging parts, some of which requires support. I’m using auto generated trees, with some unnecessary trees cut away. I am happy to post-process the part with sharp knife and a drill-bit to get the dimensions correct. Looks/esthetics is not significant here (it’s a robot part)
Model is 100x100x120mm. Maximum for machine is 140x140x140


Specifics:
Printer: FlashForge Finder
Slicer software: FlashPrint 3.17.0 32 bit (was using 3.16 before and had same problem)
Firmware: 1.5.20170419 (was using older FW before and had same problem)
Material: PLA 1.75mm Temp range 190C - 220C

Significant settings (a lot can be configured on the Finder, but these are the settings I think is most relevant)
Speed: 60/80 mm/s
Infil: Hexagon 15%
Temprature: 220C (225 for the raft)
Layer height 0.18mm (nozzle diameter is 0.4 far as I can work out)
Perimiter shell count 2
Combine infill layers: 2 ( combines two layers on the inside of the model. I tested with this set to 1 but that didn’t help)
Extrusion: Speed 109% (default for this machine, worked well in the past)

What I have done so far:
I have tried printing 5-6 times with different tweaks to the settings. I lowered speed to 40/40 (print/travel) but that didn’t help. I stepped temperature up to 230 (thinking that this could be kind of delamination), but that did not help. I tried modifying the model, actually redrawing it fro scratch, exporting a new stl and then slicing this anew, but that gave the same results.

Next I suspect that it is the vertical z direction that has some kind of mis-step, since the malformed layers are at somewhat regular intervals, and at the exact same location on each try.
Fig 3.jpg
Fig 3: To check z direction I printed a 120mm tall 30mm cube. The result is a perfect print. Next, I made a similar structure but hollow with 3mm walls (worst case from the model I have problems with). Additionally I made gaps in the sides, as shown above. This model also comes out perfect. Finally, I put a lot of holes in the thin parts of the object as shown above. Still the model is printing perfectly.


The next thing I tried was lowering the temperature, suspecting that the filament is dragged around because its not solidifying fast enough. I previously tried with a large external desktop fan for cooling (another recommendation from web serachinges), but the fan did not help at all. So then I tried lowering extruder temp to 200C (from 220C which is default).
Fig 4a.jpg
Fig 4a: The result of lowering to 200C is a disaster. The problem is significantly worse, and in fact the model structure collapsed on the printer. Previous attempts at least held together until I touched them.
Fig 4b.jpg
Fig 4b: Closeup of the stringing from my 200C temp try.





PLA material:
One potential (again from various web searching) source could be non-uniformnes of the filament. Hence I took measurments from one roll of filament (Ive used 4 different roles on this experimentation so far), using a micrometer at 10cm intervals.
I found an average filament width of 1.749 with a standard deviation of only 5 um (micrometer). It seems that the filament dimension should be well within spec.



Discussions and thoughts:
1.) The bottom of the model prints without fault. Its only at specific layers this problem happens. It looks to me, and my admittedly somewhat untrained eye when it comes to 3D printing, as though the problem is related to large changed in surface area between adjacent layers. From a control engineering perspective its somewhat reasonable to equate surface area to cooling load on the extruder heating element, thus step changes in the surface area could be expected to cause temperature control problems. Could this be what I am seeing?

2.) Could this be a problem with the slicer software? I have not found any other software that can slice for FlashForge Finder, without setting it up as a “custom device”. This requires more knowledge of the inner workings of 3D printers and the various gcode formats then I currently possess.

3.) Could it be a low temperature problem? Since dropping temperature made it substantially worse, perhaps increasing temperature is the right thing to do? If so I would need to go above the recommended range for the PLA filament. I already tried 230 without any noticeable improvement, but that was also running a fan. Perhaps running 230-240 without a fan? This is PLA so such temp ranges seems extreme... Could it be a fault with temperature sensor on the extruder, thus controlling the temperature at a to low setpoint?

Conclusion:
I have tried every test model I can think of as relevant to the problem I am seeing. I tested changing both printing speed, extruder temperature and adding external cooling from a powerful desktop fan (which I then ran at different settings to experiment with cooling pressure). I even re-draw the model with some differences in structure to see if the problem is in the design. The model I am printing is part of a robot so the room for experimentation in design is limited. It may be that both my designs therefore has the same problem, hence I can not rule out model as the source of the problem, however unlikely this seems. I have not been able to try other slicers.

Any ideas? What more could I try here? Any settings to check (and/or post here to further describe the problem).

Any help would be greatly appreciated. After 7 days of testing I am no closer to a solution...
Best regards,
Ole