How do you assure that you print layers that bond correctly?

Question

I'm fairly new to 3D printing but I'm getting the steam up and I chug out pieces without much hassle. One thing though is that I have the impression that sometimes, parts are easily broken, and I suspect bad adhesion between layers.

I'm printing eSun PLA at 208 °C (212 °C first layer), bed at 50 °C first layer then 30 °C.

When I print the Benchy boat, it feels extremely sturdy (even the small chimney is unbreakable by hand), but if I make, say, a 50 mm diameter cylinder with 4 mm wall thickness, It breaks fairly easy along some layer. I don't see any specific error, it "looks" okay.

So my question is:

How do you assure that you print layers that bond enough?

[Edit] I already know that might actually cause bad bonding (low heat, underextrusion, ...) but I'm looking for a way to see when it happens.

Here is an image from when I stopped mid print (you can see the little ooze string just at the start at the crack) for checking dimensions. The next layer didn't bond well because, I guess, the already printed part had cooled down when I un-paused the print (say 1 minute later) or maybe the z-axis went off a bit when I touched the build plate.

Is the only way of knowing layers bonded correctly, to try to break the part apart?

Answer 1

I think the reason why the large cylinder is breaking much easier than the smokestack would is because of leverage. If you print the same cylinder at a much smaller scale it might be more difficult to break it. Think of taking a wooden stick, if you try to snap it by holding both hands near the center, it would be difficult but if you push on the very ends, the leverage will make it easier to snap. I have the tug boat on my desk right now and I was unable to snap it, instead I broke the cabin section just right above the steering wheel.

Now to answer your question. One of the faults in larger prints having weak layer adhesion could be a number of things. I think a common fault is under extrusion. If the nozzle has a high enough gap from the previous layer and it extrudes just enough filament to touch the layer but not necessarily push it snug, it could be a weak bond. This is like tape, usually when you want to make sure it will stay stuck, you press hard and rub it in. So I imagine filament shouldn't loosely flow onto layers but instead be pressed hard against the previous layers. Visually this is hard to judge but I noticed getting my first layer very close to the print bed makes every other layer stick very stronger on the previous one. You get a small skirt on the first layer but it does tend to give me a stronger bond between layers. You can easily slice off the skirt with a blade. Increasing the extrusion multiplier could give the same affect without needing to adjust bed height/leveling, but this can end up giving you some nasty walls.

Another thing I recommend is to check your Z-axis. This is difficult to measure and best to find out by print quality on tall structures. On my first printer I had the issue where my threaded Z-rods would bind and cause one of the two rods to get a slight bit ahead. This caused a layer split focused on the side opposite of the faulty rod. This made many prints very prone to breaking at that point. I had to rebuild my Z-axis assembly to correct this.

Print temperature can help, try raising it up by 5 to see if this improves the bonding.

Lastly, try another filament. Filament tends to go bad due to moisture exposure, that is why usually you receive it airtight with a small bag of silica to absorb moisture. If I use filament which has been sitting out for months, just squeezing my part separates the layers in almost a slinky kind of way.

Answer 2

No matter what you do, the adhesion between layers will never be as strong as the adhesion in the direction of layer application. It's analogous to a wooden baseball bat (for those old enough to remember them :-) ). If you apply force perpendicular to the "grain," i.e. growth layers, it's strong. If you apply force along the grain, it'll snap pretty easily.
Now, if you print a cylinder vertically, it snaps between layers. If you print it horizontally, the layers run along the cylinder axis and it won't break. That's of course a much more difficult print with extrusion systems, and cosmetically less desirable as well.

Answer 3

(Note: I realize this question is old, but I came across it looking for information on related topics and it didn't seem to have any sufficiently good answers yet.)

To me, this looks exactly like what I'd expect printing with layers almost as thick as the nozzle diameter and insufficient hotend temperature or excessive speed. I've had problems like this at sizes 0.3 to 1/3 mm using an 0.4 mm nozzle even at my usual PLA temperature of 210, and can imagine it would be much worse at lower temperatures. If you didn't see the problem on benchy with a smaller cylinder, it's probably either because acceleration limits (which will be dominant on small details) prevented reaching high speeds, or (less likely in my opinion) because the deposited material had not had as much time to cool before the next layer was reached.

In order for layers to bond the newly deposited material needs to be pressed against the material already present, while melted, with sufficient heat transferred into the already-deposited material to make it amenable to bonding (naively I would guess this means it needs to reach around the glass transition temp, but that may be wrong). With layer thickness near the nozzle diameter, the nozzle will be putting less pressure on the existing material, making this harder. Also, if there has been underextrusion anywhere in the supporting wall below, the already-deposited material can simply compress downward into the gaps when new material is extruded against it, rather than them getting pressed together.

For 0.3 mm layers, I'd go with the highest temperature the filament manufacturer rates it for, or even slightly higher if you have reason to believe the material is okay at higher temperatures. Some people say lowering the fan speed or turning it off is an option, but I've had really bad results with precision and/or stringing whenever I try that.

Significantly lowering the print speed is also an option. That will greatly reduce the chance of underextrusion, and results in more heat transfer from the nozzle into already-deposited material. However lowering the speed generally defeats the purpose of 0.3 mm layers; you could instead just drop down to 0.25 mm or 0.2 mm and get much better quality at the same total print time.

Answer 4

From my experience the most profound difference is in the material. Especially with eSun PLA that I have used over 10 spools I have found huge inconsistencies. In one spool the printing adhesion was fine on other spools it was very very bad. My first advise is try another branded filament, I would recomend Colorfabb, Polymaker and Formfutura for PLAs. Secondly if you want to use the eSUN try higher temperatures, go above 210 or even 220.