There’s a bit to unpack here…
TL;DR; Use a filament that has been tuned for your printer —OR— Experiment with print speeds, nozzle temperature and flow rates … BUT don’t mess with trying to improve fan cooling mechanisms. Those are just fine and you should not believe anyone who tells you otherwise. The Devil is in the details, here.
Firstly, that model is expected to fail beyond about 50º on most printers. It’s a “print until failure” type torture test. However, it’s not that simple. For instance, BambuLab’s PLA basic Jade White print of that model manages to fully complete on my X1C — if I use the PEI smooth build plate. The textured plate, has a harder time holding onto the part as it starts getting knocked around, not that getting knocked around at all is what we want.
There’s more to consider…
If that particular type of model print is taking place with any amount of over-extrusion (more “squirt” than ideal for this specific print), it will likely fail sooner. The additional extrusion causes added stresses in the vertical direction as the print grows taller, also contributing to curling during cooling, etc, to lesser or greater degree depending on specific filament shrinkage properties and what not.
Ambient air temperature and humidity also play a role. Aside from filament moisture issues (unlikely significant factor in this case) they contribute to the fan’s ability to cool. Ironically, dry air has less ability to cool than moist air due to the chill factor.
While on the topic of moisture though, water held in filament causes the production of tiny bubbles of steam when heated by the nozzle, which can result in what can appear to be mere over-extrusion when it’s just a small amount. I’m getting too far into the weeds and babbling science over good advice here. Suffice it to say, there are several variables at play, and understanding how to manage them all isn’t necessarily straightforward — but moisture content in PLA, especially plain white, is seldom actually an issue.
That all said, the actual photos you shared of that specific overhang test model are roughly on par for an average printer using an average filament, give or take the above considerations. People building custom printers with many hours of tuning can certainly get much better results for that particular model. On the other hand, the same excessive cooling for general printing is likely to produce weaker parts due to over cooling and the resulting lowered layer adhesion, for example. Clearly, there’s a balance to be struck.
Since you are using the “Generic PLA” settings with an unknown (to me) filament, it’s difficult to suggest specific remedies with confidence. While the filament is likely fine for most purposes, it may not be as suited to prints with large overhangs as others. Spending many hours adjusting flow rate, cooling and print speed settings will almost certainly yield improved results, but that can be quite tedious. I’m OK with such things on other printers but I hold the opinion that an expensive BambuLab printer’s very purpose for existing is to prevent all that work. I’ve never had to tune anything on my X1C when using BambuLab filaments but on my other printers, it seems nearly every damned model I print ends up needing some kind of tweaking, due to over compensating to fix some issue that then creates another in a different scenario. Just saying.
Print speed has a significant impact on this type of print. The sooner the print head returns to lay a line atop another, the less time the lower layer has to cool. Your photos seem to indicate this as an issue — aka lack of cooling BUT do not jump straight to blaming fans or cooling ducts — a common mistake. Others may suggest printable upgrades for “upgrade” cooling ducts and other modifications. I urge you to ignore those. They won’t help and will waste a lot of your time. These printers work very well as they come, under most conditions. I believe you’ve been unlucky with the particular combination of filament material, model shape, and print speed in this case.
For a potential quick fix without changing filament brand etc, try lowering the print temperature by 5 to 10ºC and also reduce the Flow rate (not Max flow rate) by about 10%. Ten percent may be too much, but it will quickly show if it was a significant contributor to the issue. Under-extruding produces weaker parts with voids between lines but also solves many other issues, helping to isolate primary causes.
While I am not a BambuLab fanboy (yet :P) and would stop short of saying, “Use the officially tested filament!” there is some value in such a suggestion when starting out with these higher-speed machines.
Hope that helps.
EDIT:
For reference, here’s a print I just did of a similar model on my X1C using BambuLab PLA Basic Jade White on a PEI smooth plate, using all standard settings, including a bed temp of 55ºC. I usually change that to 65ºC for better adhesion but forgot this time. (I never use glue on PEI surfaces, unless I need LESS stick, like for large footprint TPU prints.)
This was printed a “Normal” speed, in the center of the bed and seems to have completed the 60º section before getting knocked off. Had I done something to hold it down better, it likely would have completed — not that such would prove anything. The overhang quality up to 60º is notable however, it looks like a physically much smaller print than yours and thus has much less opportunity to build up over-extrusion artefacts.
I would be happy to print the same overhang test as yours for comparison’s sake, if you like. Just point me to it. Maybe even more data could find a way to be actually useful? (I’m aware my post is very long and probably not all that helpful. My apologies!)
Here’s another one, this time with 10% additional flow rate (0.98 to 1.078) …
Here’s one printed at 20% over extrusion – flow rate 1.176 …
Each ten percent increase has resulted in the print reaching a few layers less in height before curling and being knocked from the bed.
You can also see from this that over-extrusion alone does not cause some of the other artefacts your are experiencing — which is the real reason I wanted to produce these three prints to show — remembering that this “proof” is only valid for this specific filament, with this specific overhang model.
