Title says it all. As pictured, PA-CF print just completed. Attached to it, a small Neodymium magnet. I was pressing the magnet in to a pocket in the print, which was a little too small. When I let go it slipped out of the pocket and grabbed on to the adjacent surface.
It’s Nylon and Carbon Fiber. Where’s the iron coming from?
Neither Carbon Fiber nor PA are magnetic, paramagnetic or diamagnetic so the only explanations could be:
If I had to guess I’m banking on 4.
Is it only that particular part of the model that the magnet is attracting to?
The magnet is attracted to every part of the print. It isn’t a very strong attraction, the magnet easily gets knocked off. But if I put the magnet on the table and then bring the printed part close enough the magnet does exactly what you’d expect it to do if the print was metallic. The print itself is about 150mm in length and the entire thing is attractive to the magnet.
I spent about 30 years of my working career designing and manufacturing Hard Disk Drives. I am not a beginner when it comes to magnetics. 
My guess, the filament is formulated to include iron for some reason. I have been 3D FFF printing since 2010, but I have never made the effort to learn anything about how filament is formulated and manufactured. So I have no idea why it might be “doped” this way.
Like I said, too bad it’s so weak or the property might actually be useful.
I’d say the manufacturer of the filament adds in metallic particles to aid the carbon fibre effect (surface finish and sheen). And it also probably adds a little profit, as CF is light and the spools are sold by weight.
“I don’t think it can be impurities”
“the filament is formulated to include iron for some reason”
PA-CF is clearly two materials fusing into a composite, digress your perspective on this please.
It would have been nice to know that the entire filament was semi-magnetic in your original post, it sounded a lot like it went to a specific spot, sorry for any confusion that may have caused. A little odd that you describe it repelled from one side, and stuck to the other, when we are talking about iron which would prefer proximity wake field reaction instead of the opposite surface some mm to m away. You know the old laws of attraction deal.
It sounds like there is a density or distribution of their doping going awry if it preferred one location over another. Poor doping mechanism.
I think it goes without saying if the company does not disclose that their filament is doped with a conductive material you are looking at a potential class action lawsuit for failure to disclose the full properties of your filament. You can mention your material is semi-conductive in areas and not disclose the exact formula or process.
I can only imagine a spec sheet for NASA calling for pure PA-CF only to find it retained a static charge and created a dissimilar metal situation aboard some equipment. There would be literal hell to pay.
Language barrier, I suspect. I didn’t say “repelled”. I said the magnet fell out of the pocket (more accurately, I never fully seated it in the pocket because the pocket was a little too small) and then adhered to the surface of the print adjacent to the pocket. The magnetic attraction isn’t sufficient to hold the magnet against anything but the slightest acceleration, it falls off easily. But it holds enough to retain the hold when touch the magnet with the print, the magnet comes away with the print. But I can’t stress how weak the force is. If it wasn’t a Neodymium magnet, I doubt it would generate enough force to adhere.
Carbon Fiber is, by definition, conductive. So doping the filament with additional conductors as well as CF doesn’t “move the needle” on liability. It was conductive before and it’s conductive afterwards.
But I have to wonder how the filament will “age” if it’s got particulate Iron blended in, will it degrade/rust with exposure to the environment? Especially encapsulated in Nylon, which is hydroscopic to begin with. Iron seems a curious choice.
This is the first filament I’ve used in decades, that wasn’t advertised as metallic, that was magnetic.
Fair point I wasn’t thinking about CF in particular, just a slip of the mind I suppose.
I think oxidization on the surface and in micro fractures is the first thing I would consider being an issue with aging.
We also have to consider the thermal properties of CF versus PA versus FE or whatever ferrous material they are using.
If they indeed used Iron you could test this by artificially creating the right conditions for oxides to deposit on the surface. I am truly curious about what kind of dope they are using now.
I agree your findings are the first I’ve ever heard of this specific situation.
The issue with not disclosing the material contains metal is still a primary concern that we both have a general consensus that aging of the product should be scrutinized. As well as any other underlying variables.
If I get my hands on some PA-CF I’ll make sure to run a test and return here.
In case another data point helps, I have PA-CF that came with my X1C, and a part that I printed some time ago.
It has no attraction to some rare-earth magnets that I have.
Elemental Iron is strongly magnetic. But Iron Oxide (rust) is negligibly magnetic. So if there is Iron and it does oxidize with time/exposure, printed “some time ago” may not be a valid sample for this test.
So… Who here wants me to put a piece of the PA-CF under a laser spectrometer? Mind you it would need to be more than a trace of iron for me to see it in the measurement. And I’ll just get elemental data, none on the actual molecule structure ( needed to differentiate some polymers)
Side note, since the CF is quite abrasive… I wonder if it isn’t just caused by the tiny fragments grinded off the nozzle?
Are you using te hardened steel one?
If you have it, I prefer SEM/EDX.
I am using the hardened steel nozzle. And I did consider that as a source, briefly. But PLA-CF doesn’t do it. Only the PA-CF (based on my limited sample of BBL filaments). And if the nozzle was being worn away enough to “salt” the extrusion with a meaningful amount of magnetic particulate (Stainless isn’t very magnetic, BTW), by now I expect the nozzle size would have gotten quite a bit bigger and I’d know it.
Unfortunately not, only have access to a UV laser spectrometer…(Keyence EA module ). Should give a fairly decent idea though, just not for very small trace amounts.
Fair enough, was just a though, especially if you weren’t using the hardened nozzle with CF fillament.
I would also expect that you’d have noticed a nozzle issue with that as it wore down…
Last thought and somehow harder to test unless you have an ionizer… PA is very susceptible to static charge build up, maybe we aren’t seeing magnetic forces here, but pure static adhesion?
Maybe see if a small piece of paper also sticks to the surface…
Definitely magnetic. No question about it. For the magnet to stick with a static charge, the charge would have to be pretty high. Rubbing a balloon on your wool sweater for a few minutes wouldn’t be enough charge to get the magnet to attract to the balloon… at least, it never has never been enough for me for anything besides picking up another balloon or cat fur. A small piece of metal weighing a few grams? Uh-uh.
We need an aqueous bath of Oxine and iso, and a UV light. Or the alternative method that uses a similar composition, but does not require UV light in order to visualize the very minuscule particles we are talking about here. If you happen to have access to lab grade chemicals.
Otherwise maybe just get a good old bucket of salt water and maybe ionize the material a bit to jump start the process.
There may be a trace color change so I would use a very non red, light colored filament to see, otherwise use a white container to watch the sediment layers for any kind of proof of oxides deposited.
Ahh, one more testing method I just thought about, process the filament into a finer particle slurry, and use an iron testing kit that has a decent sensitivity spectrum. You would want to use distilled water or so I am sure, to remove the minerals that may ruin the test.
If it was a few years ago, I’d have access to everything we’d need for this analysis - SEM/EDX, FTIR, Raman, Auger, Gas Chromatography, Liquid Chromatography, etc., etc…
But I work for someone else now and while they use the same equipment it is not accessible to me for personal projects.
I don’t really care that it’s magnetic. I was just surprised to make the discovery and thought I might share that with the forum. The filament works great, I have no complaints.
Could maybe try and dissolve the nylon to leave the CF and any metallic particles behind?
So as I read the original post I happened to have a couple of samples of prints in two types of PA-CF (Esun EPA12-CF and PAHT-CF) I also happened to have a stack of 8x3 magnets handy. Both the samples are 6.9mm thick and almost solid
I wouldn’t have believed it if I hadn’t tried it myself but there is definitely some weak attract / repel going on between my samples and the magnets. Running the magnets a few mm over the surface I can feel the push and pull - with the effect more noticable when moving perpendicular to the layer lines.
Edit: It is just enough to pick up the magnet
