Print Core - Wall Delamination Issue

[PrintPlace 0]

We are a BE 3D Printing company operating a line of UM's - We have been active in the 3DP industry for about 10 years and have been working with UM's during (almost) the same amount of time. We keep our equipment in pristine condition, we perform periodic best practice maintenance and have been able to perform (most) of our troubleshooting and problem solving thanks to the information found within the UM community, the expertise of UM team members and our own experience. However, we've been experiencing one specific print core issue that we ourselves haven't been able to solve and neither have been able to find any information about, which is why we resorted to posting this question instead.

The component in question is a standard AA 0.4 print core, which we only expose to standard, non-abrasive, high quality materials, i.e. PLA, PET(G) and ABS. I have always been told that the lifetime of a standard AA 0.4 under such usage behaviour approximates infinity (well, not infinity, of course, but a very, very, very long time). However, what we've been noticing in practice, is that the print behaviour of some print cores tend to change over time. I wouldn't call the change sudden, but I wouldn't call it gradual eather, sometimes it occurs over the period of, let's say, 5 decent builds, where the first one shows no issues and the fifth one clearly does. The issue itself manifests itself as wall delamination - We did some tests where we printed a (PET(G)) test part with the suspected defective print core (lifetime 41 days) and the output clearly shows wall delamination (We're only talking about X-Y outer wall delamination, not Z delamination - I manually seperated the outer walls a bit more to make it more apparent on the picture below):



We then printed the same test part on exactly the same set-up, but with a working print core (lifetime 57d), and the output is printed perfectly:

 

 

Both components beside one another, to make the difference very clear:

 

Under the assumption that the manual build plate calibration was more or less identical between both test prints (this is something we made sure to be the case), this could suggest that the root cause is indeed to be found within the print core set-up, but then the question is, what could possibly cause the print core to behave accordingly? A heater sensor test shows positive results. It loads material without any problem, so there's no real need for any hot or cold pulls. The print core is designed in a fairly simple and straightforward fashion and we don't immediately see where things can go wrong within the set-up in such a way that it would result in wall delamination (except in those areas that are harder for us to test thoroughly, such as a malfunctioning of the board...?). 

This is also not the first time this happens, we noticed this behaviour a couple of times before, on other machines, with other print cores, that we eventually took out of order, but this is now becoming an apparent structural phenomenon. One can argue that it's faster to just buy a new print core. It might be, but this has always been about understanding what is actually happening (which is what drives most of us here, especially the veterans, I presume) 🙂 Looking forward to reading your input! 🙂 

 

P.S. Now that I'm posting anyway - Big shout-out to gr5, your many contributions to this community have proven to be very valuable more than once. Keep on rocking! 

 

[gr5 1,841]

The only thing I know for sure is that the gap in the walls, the "delamination issue" is underextrusion.  That is a very common sign of mild underextrusion.  Even if the part looks perfect everywhere else.  And the part doesn't look fine everywhere else - I see underextrusion on the top layers as well.  You can find the ratio of the gap to the lines by measuring with a micrometer on a blown up version of your images and figure out what percentage underextruded.  I'm going to guess 15%.

 

It's nice to know it's in the print core only although that isn't 100%.  A good feeder should overcome some of the deficiencies in the core.

 

Just to further eliminate the feeder - I recommend fighting the feeder.  So lift the lever and pull the filament half way down the tube then on the touch screen do "MOVE" material to energize the feeder and then fight the feeder by pulling down on the filament.  You can even turn the dial in the MOVE menu while fighing it.  Try to get it to skip.  You shouldn't be able.  The S5 feeder can pull about 15 pounds or about 7kg.  This is rough.  5lbs of force should be enough (barely) for good prints normally.  10 pounds is excellent.  15 pounds is nominal.  So do this quick test.  Maybe grab a weight so you can compare the amount of force against a known weight.  Or you can grip the filament to a weight and let the weight do the pulling.

 

Also hopefully you are doing default temps and speeds.  Printing too fast or too cold can cause underextrusion.

 

And it could be partly a bowden issue.  You should replace those every 1000 hours of printing or so I think.

 

That leads us to the core.  Really I expect it is buildup on the inside of the nozzle in that narrow passage.  That passage that is only 0.4mm in diameter may now be 0.3mm.  I'm not really sure how to clean that out.  Cold pulls comes to mind but that gunk can be stuck.  You could take the core apart and burn out the nozzle.  It will be weaker (the brass no longer tempered if it gets above around 500C) but the brass is stronger than necessary anyway.  I have a youtube video.

 

Anyway, if it were me, I would just increase the flow to 110% and increase the temp by +5C.  All of which you can do from the tune menu live.  This should help you get another 100 hours of prints maybe.

 

Or you could use a hypodermic to scrape the inside of the tip of the nozzle while it is hot on the printer or right after a good cold pull that leaves the nozzle tip empty.

 

Another possibility of course is the temp sensor.  Any increase in resistance (contacts where print core meets the printer for example) will make the temp look high which means it will print colder than desired.  You could try this video which is for PLA to get an estimate of the temp.  I don't know any other way to check the temp.

 

PLA softens at a much more consistent temperature than you might expect: