Carbon Fiber Filament: Super-Material or Super-Hype?

[@justinmoritz6543]

Out of all of the 3D printing channels, you are among the betters ones out there. Very good, accurate information and not the normal pseudo engineering/science that is ALL too common in consumer 3D printing. You stand out as having genuine thoughts and original creativity. Excellent content, keep it up!

[@NathanBuildsRobots]

Thanks for watching, and for the generous comment! I try to reference real engineering principles wherever applicable.

[@McJiver]

I believe i saw a video from Hoffman Tactical and he heat dries his CF nylon at a similar temp for 12 hours. I thought that was insane but this proves a notable difference even if the test was a bit muddy.

[@gottmituns3225]

Kind of a P80 filam❓ That what i''m looking 4

[@TuncayAyhan]

Your channel deserves way more subscribers! Can't believe you are at 2.31K subs.. The video/production quality is high af!

[@NathanBuildsRobots]

Thanks! I'm definitely in a phase where I'm figuring out what "works" for video production. I'm going to be grinding out 1 video a week for a while, then start doing some larger projects - stuff that 3d printing people will enjoy, but will also appeal to a broader audience. Hope you stick around for the journey!

[@marksdroidx]

Agreed. He is quickly becoming my favorite 3d print channel.

[@Scottthebott]

Do you have any videos on how to anneal the PLA? First day with a printer and could sure use a tutorial 😅

[@NathanBuildsRobots]

Basically just cook it for a little while. CNC kitchen might have a video on the process.

[@Scottthebott]

@@NathanBuildsRobots I ended up on CNC Kitchen last night and wow, I did not realize how many methods there were for this process. Very interesting info. Thanks man

[@MageLeaderInc]

The one thing I would say is spend more time with the nylon carbon fiber to dial in the settings. You can get much smoother prints than what you're currently getting. If it's dried completely and calibrated specifically for it. I would set it up so you can keep drying it while printing as well.

[@NathanBuildsRobots]

I got some from phaetus that arrived dry and printed out super strong. Process is everything with this stuff!

[@synapticaxon9303]

Yeah, and then you spend some fraction of spool dialing it in and print a few things and you're out. :) That'll be another $100 plz. Worse than a boat!

[@Soravia]

Nylon lets off steam from absorbed moisture, and weakens the material due to voids from moisture let out.

[@AndrewAHayes]

I designed a drone chasis but knew the CF filament was not strong enough so I printed in ASA and wrapped the parts in CF matting and epoxy and drew in the epoxy in a vaccum drum, the parts are very strong although I have managed to damage some in some extreme crashes

[@NathanBuildsRobots]

Continuous fiber CFRP is much stronger than the chopped fiber stuff used in most 3d printer filaments. At the end of the day, the nylon matrix ends up taking most of the load, so the carbon fibers are more of a stiffening additive than anything else.

[@SquareOneFPV]

Hi Andrew, I'm thinking about 3D printing a frame for a very light weight 1s toothpick. I know buying a cheap real carbon one is probably better, but I'm new to 3D printing and making designs and this would be a fun project to keep me busy this winter. What filament would you recommend, or do you have any tips? Thanks. Oh and great vid Nathan, you got a new sub!

[@NathanBuildsRobots]

Thanks. Feel free to chime in here Andrew, but I have thought about making a carbon fiber quad and will share my thoughts really quickly: I wanted to try friction welding little carbon fiber rods together along with 3d printed parts. It would be tricky but could be very strong and light kzbin.info/www/bejne/gXy9hppreNp6erc amzn.to/3EVBvM8

[@synapticaxon9303]

I found that I could not get flat discs for a tesla turbine with PA-CF but they came out great in ASA, nice and smooth and stiff, no warping! On the other hand, PA-CF is the strongest for me when making things like screws or inserts, thin rotating pins, etc.. I also find that PA-CF has anisotropy, I'm getting way stronger results in a single layer vs vertically. This has changed the way I orient parts and so on. For my turbine, I ended up making the shaft and casing and misc parts out of PA-CF, and using ASA for the discs, and I think it's the best of both worlds. My main takeaway is that there isn't a one-size-fits-all material, and the future is multi-material. Use the right material for each part.

[@NathanBuildsRobots]

True, and multi-orientation. Non planar printing can make up for some of the deficiencies of interlayer strength for CF reinforced plastics. The issue is that there are basically no fibers oriented in the z direction, and they aren’t connecting layers together. There could be a stitching method to pull fibers into a vertical (along Z) orientation.

[@synapticaxon9303]

@@NathanBuildsRobots Exactly, this is why I think 5d printing or whatever they call it is the next big thing. We must go beyond x/y/z. It seemed hard once upon a time but lately I see lots of people doing it. I think it will kind of become table stakes in the coming years. The newer filaments and increased capabilities elsewhere will necessitate it.

[@sentinelau]

Nothing better than recording a video whilst eating some fried chicken, am I right?

[@NathanBuildsRobots]

Can't film today, Chik-Fil-A is closed :(

[@gummywurms226]

Not all CF Nylon is the same. I would try again using a 25% CF Nylon.

[@NathanBuildsRobots]

I completely agree. I tested out some cf nylon from Phaetus and it blew everything I have printed before out of the water. I have tried re-drying and re-printing this Sainsmart stuff and unfortunately it just does not hold up very well.

[@tazanteflight8670]

Adding a desiccant to a roll of nylon is pretty funny, considering the nylon is more hygroscopic. The poor desiccant will starve.

[@grathado28]

this video has the appropriate amount of comedy 🤣

[@NathanBuildsRobots]

Whose ready for CF reinforced wing night?

[@rsilvers129]

Thanks for air fryer tip. Also you can set Excel to show only 1 decimal.

[@NathanBuildsRobots]

You don't think my measurements are good to 15 significant figures?

[@koubenakombi3066]

Great tests! Great help to choose filaments! Thank you! After heating PETG in the air-fryer, I would not make any french fries in there again!

[@NathanBuildsRobots]

Ok, hot winds only from now on! I have since tried some higher quality CF Nylon and it definitely lives up to the hype. It's over 2x stronger than PLA.

[@synapticaxon9303]

One time I also accidentely printed PLA at PETG temperatures and was amazed at the resulting hardness of the printed PLA. It was MUCH stronger than PLA at PLA extrusion temps.

[@NathanBuildsRobots]

Cool, hotter = better layer adhesion because things are more melt y. Will take a toll on overhang and bridging performance but for strength that is a hot tip!

[@synapticaxon9303]

@@NathanBuildsRobots Happy accident, I try not to get too crazy with overhangs and bridging. ;)

[@noway8233]

Overextrusion efects

[@birdmananyweather1681]

The filiment on the fries in the air fryer..... I lost it 😂😂 keep up the vids my dude. Got my sub!

[@patrickspapens5497]

Would you think there would be a difference between air frying the filament or oven drying? Air frying has a bit more air circulation so it might be faster in transporting any water away?

[@NathanBuildsRobots]

I would think so. My oven is a convection oven so it would probably be similar in time, but it’s good to have separation between food and experimental plastics

[@Gojira_Wins]

​@NathanBuildsRobots Just to point it out, Air Fryer (Frying) is a brand name. All the brand does is sell convection ovens. So your convection oven is the same as their "Air Fryer".

[@patrickspapens5497]

@@Gojira_Wins not entirely though. Airflow in an air fryer is very different than a convection oven

[@ovDarkness]

Test show errors in slicing and/or printer settings. PETG shouldn't shatter. Probably too much part cooling. PA-CF should be printed in heated chamber. (70-80C ballpark). PLA should be annealed in much higher temperature than 50C. You not only want to relieve internal stresses. You want it to crystalize.

[@NathanBuildsRobots]

The PETG may have broken with a single fracture, but with the way I had it set up, the shoelaces acted like a slingshot and threw the pieces at the ceiling/floor, and probably shattered at that point. PETG is much more brittle than the other materials I was testing, because the rest survived the same post-fracture impact. PA-CF can be printed at room temperature. It may not be optimal, but it was about as easy to print as PLA. The stuff from Sainsmart had much worse strength and quality that other PA-CF filaments I have tried. Interesting to hear that about PLA annealing. I don't have much experience with annealing in general. With steel you can look up very good data and recipes on annealing. With plastics and 3D printing I had trouble finding good sources on how to do it, and what temperatures/heat soak intervals. Do you know of some good resources for that?

[@PaddyBoyBKLYN]

After its dried then printed, i wonder if sealing it with Cerakote would allow it to maintain its dried strength. 🤔

[@NathanBuildsRobots]

Nylon loses some strength and becomes more flexible and less brittle as it absorbs moisture. It’s not necessarily a bad thing. And moisture will deep in over time, it’s almost impossible to keep it out. Even if you seal it it just takes one crack or scratch for the water to get in. And over enough time, the water can just go straight through the sealant. What is a bad thing is printing it while the filament is wet. It puffs up like popcorn and loses about half its strength. You get a slight reduction in density, but the strength loss is much greater (something like density^2.3)

[@oxpack]

I find cf-nylon near impossible to print. Pain to switch nozzles and also wears out the extruder brass gear. The added carbon by definition has to lower the layer bond strength as there is less media to attach to each other when the one melted layer goes down on the next. Maybe if you spend days tuning the settings it works but at that point could have just machined a much stronger and better looking part.

[@HiTechDiver]

So much for tensile strength, but that's only one factor in material properties. How about crush resistance and breaking strength. We print CF Nylon at our college, and it is superb.

[@NathanBuildsRobots]

It depends on the blend. This was some shit tier carbon fiber blend. Good stuff is 2-3x stronger than PLA

[@98f5]

I used this sainsmart nylon 6 cf to print most of one of my vorons and it still is super strong 3 years later.

[@marshalllapenta7656]

This is what I was thinking about using carbon fiber, for it's strength properties. Yet using nylon to bring the cost down. Keep making videos like this....

[@jonathanlarsen4177]

I've heard good things about CF-PET (not PETG), apparantely it doesn't creep (as much?) as CF-Nylon. This was a big problem for CNC Kitchens Voron 0 build

[@NathanBuildsRobots]

Yeah I saw that video. CF nylon is a pretty weird. Since it is a composite it can do stuff that you wouldn't expect from a solid homogenous material like the other plastics.

[@chengxiong8543]

interesting test but what is being measured is ult str not tensile str, esp for polymers tensile str can be difficult to calculate as exhibit non linearity. tensile str to be measured using stress/strain incremental data and linear curve fitting.

[@NathanBuildsRobots]

Typically I use the term "ultimate tensile strength" or "UTS" to describe the maximum stress achieved by a material during tensile testing. Looking at the literature, it is more often referred to as simply "ultimate strength" as you point out, but both terms are used. Yield strength is the stess at which the material starts to plastically deform. It is harder to measure, requiring precise displacement information. The Young's modulus, or modulus of elasticity, is the slope of stress vs strain in the elastic region of a stress/strain curve.

[@MRMIKE276]

Excellent tests. Very scientific and I can tell you are either an engineer or going to school to be one. Subscribed :)

[@Eronpas]

I hope you engaged the safety -squints

[@NathanBuildsRobots]

I engaged the "20 foot wire and hide in the other room"

[@proterotype]

I Love nerdin out on this

[@alexfeatherstone6676]

PETG usually only shatters when its cooled too much, too much cooling it becomes brittle

[@dosomisolami]

Thanks for all the hardwork :) i think it's better add material change its shape instead of failure load and can you add abs too?

[@ecds]

Great but… material strength it’s one of the characteristics... other one is temperature stability, it’s necessary in mechanic parts… in that situations it’s where pacf it’s one good filament (compared with pla or petg)

[@parvariramin]

Which one is stronger against pressure and direct sunlight? Is it also of good quality in printing?

[@NathanBuildsRobots]

Your biggest issue with handling constant pressures in direct sunlight is going to be creep. CF nylon has good print quality if it is handled properly and printed in a dry state.

[@Soravia]

You used a junk 25% CF Nylon. Quality 25% CF Nylon costs 4 times the price of $60+ / kg price of Sain Smart brand

[@NathanBuildsRobots]

I agree. I redid this test on some CF-Nylon from Phaeturs (which is just a rebranded filament as far as I can tell) but it was high quality stuff. It broke my test rig when I tried to pull it apart, so it was well over 2x stronger than anything I tested in this video. The footage of that is in my Phaetus Rapido video if you want to see it: kzbin.info/www/bejne/gJqcqJylp9OknpY

[@tgirard123]

I feel like the Annealed PLA shouldn't count as much because you rarely get a usable part when you anneal PLA. It just shrinks and warps too much. I'm a little surprised about the Nylon performance. SainSmart is not a real name brand that I would use for Carbon fiber Nylon parts. I would love to see if you got any differences with, say, Essentium HTN CF25 CarbonX™ PA12+CF FIBREX™ PA6+GF30 These are pricey but are well known and proven All this BS aside, I really liked how you did your tests

[@NathanBuildsRobots]

Thanks! I got some CF Nylon from phaetus and it destroyed the Sainsmart stuff. It's over 3x stronger. So idk what's up with the sainsmart filament. I think I needs to be extensively dried, or it's just low quality.

[@noway8233]

Thats why yuo need to rescale yuo model at abuot the same srinking , then yuo print, then yuo anhinling, then yuo mesure

[@stefanguiton]

Great testing! In terms of future video ideas, I wondered if you have ever seen this before: Someone managed to reinforce 3d prints with fibreglass and a soldering iron to bond them together( kzbin.info/www/bejne/fWKrhoikmNaWoLs - "3d Print Welding and Fibreglass" ). Would be cool to see some of your professional testing methods on the process to see how much stronger it would be vs normal filaments. Keep up the great work as always!

[@NathanBuildsRobots]

That sounds interesting, is basically a DIY version of the mark forged continuous fiber printing. There are some academic papers on its tensile strength, it's crazy strong. markforged.com/materials/continuous-fibers/continuous-carbon-fiber

[@NathanBuildsRobots]

I have worked at a ton of places where I could have gotten this type of woven fiber in Kevlar, glass, and Carbon Fiber in the trash. I should have stocked up on it! Next time I find some in the trash in totally going to grab it.

[@iamdmc]

PA6-CF20 is much stronger than whatever you used. Sainsmart? never heard of them

[@NathanBuildsRobots]

Yeah, I tested another CF nylon and it was over 3x stronger. Sainsmart's CF blend is not so smart.

[@iamdmc]

@@NathanBuildsRobots which one did you try? I've got an esun PA6-CF15 (the newer "upgraded" 750g spool) on the way. By the way thanks for making the S1 silent mods. I've done on my S1 and goddamn it's so quiet. Thing of beauty

[@NathanBuildsRobots]

@@iamdmc Nice. I used Phaetus PA-CF blend. It printed at 300C, so I'm guessing it was PA-6. Phaetus is just a label, so for all I know the esun stuff is the same thing.

[@philsmith6784]

Try CF PLA/CF HTPLA

[@NathanBuildsRobots]

CF PLA is probably insanely stiff. I will reach out to some companies to see if I can get some. I watched a CNCKitchen video on CF Polycarbonate. He said it was one of the strongest filaments he's tested and it was something like 9000 psi. That's about what I was getting (within 25%) for virtually every type of filament I tested - so it seems like strength is pretty even across the board.

[@andrecook4268]

Noice.

[@Bdickey]

lot of work put into this just to botch the testing and then submit the results anyway rather than correcting the mess ups despite being aware of them

[@NathanBuildsRobots]

I’m so sorry, would you like me to get a universal tensile tester with a calibrated load cell and follow the ASTM standard? Because I can get right in that for you jus pay me $200/hr

[@tjohnson4062]

Absolutely worthless for information. "This might be that and that might be the other", all the while there's incomplete extensions and incorrect application. I have an idea for your testing... Try and do a more scientific approach. Unless you were just after the free filament then you did it perfect.

[@NathanBuildsRobots]

Would certify a bridge or plane with this data? Hell no, but it was fun and I learned some general information about the strength of materials. Would love to get some real data using a proper universal testing machine, but as you can tell by my production value, this is not a high end lab setup.

[@chrismauck2710]

Absolutely worthless information. None of your tests have a unified control variable. What were you even testing? Every single CF Nylon sample you "tested" was different. Missing layers, poor adhesion, not dried ect. Not only that but it looks like you didn't even bother dialing settings for the CF Nylon to begin with. To top it all off, the excel plot you made didn't even have the materials labeled (or even averaged based on material) on the X-axis making it more difficult to even understand what it is you were trying to test.

[@NathanBuildsRobots]

I am documenting my learning experiments on YT with a budget of 0 dollars and not enough time. If I were trying to get my lab NIST certified, publish my findings in an academic journal, or was qualifying these materials for a safety critical application, I’d put a lot more effort into it. The load-cell I’m using is nowhere near fast or accurate enough to get good data, even if my samples were prepared in a professional lab. In any case, I hope you learned something despite the imperfect dataset.

[@NathanBuildsRobots]

Btw, this sainsmart cf nylon is crap, very hard to print with. My samples printed with another brand came out perfect 1st try, it’s featured in the video I did in the phaetus rapido. Printed some good cf nylon and it worked much better.