Localised poor layer adhesion

[russelle]

Hello, I’m hoping for some help. I’m prototyping a part with a split pin. When I print it, there is a visible “groove” around it, corresponding in height to the top of an nearby feature. This is a weak point and the pin is easily fractured at that plane, by squeezing the pin.

But here’s the weird thing:

• the model is perfectly mirror-symmetrical from front half to back half, but printed normally it is always the front pin half that fractures, not the rear pin half.
• if you rotate the model 180 degrees, the same half-pin breaks, even though it is now facing the back of the plate
• if you rotate the model 90 degrees in either direction, the other half-pin breaks
• both pins will break at that layer with enough force but the pattern above is very consistent about which pin is much weaker than the other.

This has been repeatable over many test prints with different settings and different numbers of copies printed at once (from 1 up to 40).

Here is a photo illustrating the part and the behaviour. The black marker marks denote the part facing the front of the plate. The red marker denotes which half-pin fractured on squeezing. The part at the centre rear is the clone “printed normally” i.e. not rotated. For scale the base of the part is 8 mm x 16 mm.

This is the best close-up I have of the failure. (It’s difficult - it’s tiny.) It is a slightly different version of the part but the failure looks the same. You can see its appearance is consistent with poor layer adhesion.

The samples above were printed as follows:

• 3mf (saved from Bambu Studio so should have settings attached):

S0019.3mf (45.2 KB)

• A1 mini with 0.2mm hotend
• Bambu PLA Basic
• Standard settings for Bambu PLA Basic

I have tried:

• various setting changes (as prompted by AI) designed to prevent a change in layer time or in speed either side of the top of that tube on the right: it still happened
• having very long layer times for full cooling, by printing 40 copies at once: it still happened
• changing the height of the tube on the right: the fracture plane of the pin followed it up or down
• inspecting the slicing. The layout of the slicing of the two pin halves has perfect mirror symmetry.
• looking at the timing/ordering of the printing of each pin half and the direction the walls are drawn. I couldn’t pick up anything here that would explain what I’m seeing

The fracture itself is the problem but I figure the strange, consistent pattern about which half is by far weaker must be telling us something.

I’m new to 3D printing and would love some help with diagnosing and fixing this problem!

[MaxThreeD]

Hi ruselle,

Even with 40 copies, the overall layer time once it clears the height of the thicker tube drops drastically, eg. from 140 to 40 seconds. 40s is way over the 6s minimum layer time in filament settings. so the speed adjustments don’t kick in. But I’m guessing the layer bonding is still different enough to make a difference, given how consistent you say the break location is compared to the thicker tube.

Which tab breaks is likely a red herring (though a curious one indeed)… there’s no real difference except maybe where the seam lands, or the order it is printed in exactly (especially if it’s that sensitive to bonding vs. time as demonstrated above). You say that half is “far” weaker, but I wonder how you determine that… bending each backwards individually?

Anyway, you’re right that the fracture is the real problem. I’m not sure PLA is going to cut it for this, at the size you want, with the 4 thin walls… that’s a tough ask anyway, IMHO. The bonding would have to be top-notch, at least. General recommendations for improving PLA bonding is to increase nozzle temperature and reduce printing speed, which may certainly help in this case. You could try thicker walls of course, or at least a thicker line width.

HTH,
-Max

[EnoTheThracian]

Not at the laptop atm, so can’t look at the 3mf unfortunately.
Pure speculation but easy to check in the slicer Preview:

Assumption 1:
In this case, weak layer adhesion is primarily related to layer time or more precisely: Local time between layers.

Assumption 2:
Path planning leads to a longer time between layers for one half of the pin than the other.

Check:
Locate the failing layer interface in the slicer preview. Use the bottom slider to check the paths of the layers above and below the fracture. I consider it likely that depending on part orientation, one pin half has a longer time-between-layers than the other.

Redemption recommendations:
There are limits as to what is possible with FDM. In particular regarding layer adhesion where PLA is actually pretty good (very few materials are able to beat it in z-impact. IIRC, only TPU and PA6-CF).

So I think that @MaxThreeD already highlighted the key items of line width (also consider nozzle size), nozzle temp and getting closer to the minimum layer time.

&