Printing Carbon (PA6-CF) E-Bike Parts

[TexanInTennessee]

Been trying to print some components in PA6-CF. The printer is a Bambu H2S. The filament is a Bambu filament, using Bambu presets. The print head is a 0.6 tungsten hardened high flow head, brand new.

Since I swapped the print head for the recommended one, I get prints like this>



When I was using the standard 0.4 head I had slightly better success but still many prints would turn to spaghetti.

Grok is useless so far, I was hoping if some of you have gone down this path could save some time with the proper adjustments?

 

[Hummina Shadeeba]

I’ve got two giant rolls of carbon filament that never print well and warp. My success for everything else difficult to print relies on printing very slowly especially at the beginning and an extra hot bed and closed chamber with a big brim

 

[TexanInTennessee]

I’ve got two giant rolls of carbon filament that never print well and warp. My success for everything else difficult to print relies on printing very slowly especially at the beginning and an extra hot bed and closed chamber with a big brim

I have not tried a brim yet. Will tray that with models where the first layer did not set up. Also have not tried a glue stick.

This model, I've been able to successfully print with the 0.4 head and standard settings it was just low resolution. With the 0.6 it has extra buildup on the print.

For speed, do you slow down the first layer print? All layers? Grok had me try slowing from 600 ms to 200 ms but the test print looked exactly like it did with the vanilla settings.

Grind Hard Plumbing, I've seen them print carbon pieces that look like PETG, and they have the same printer I do. Wish I could ask them.

 

[harrisonpatm]

Couple things come to mind.

First, this looks like a small part overall. When you print something small like that, at the high heat needed to successfully flow PA6, you're gonna be dumping more and more heat on the previous layers, which will make it harder for the first layers to solidify. When you print large things, the prior layers have time to set before another hot layer gets dumped on top of them. Printing small things, not so much. If this is the case, I don't know what slicer you use, but you might be able to fine a parameter called 'minimum layer time' or something like that. For example, you could set this value to 60s. If the layer only takes 10 seconds, the print head will just stay in place or move to the side for another 50s, letting the first layer cool and set before moving on.

Second thing that might be the culprit is the nozzle materiel. You selected tungsten for its resistance to abrasion, I presume, because you're printing with CF. The downside is that because tungsten is so dense, it acts as more of a heat sink than you want a nozzle to be; the nozzle isn't getting hot enough, or isn't consistently holding heat. In theory, while tungsten is supposed to be of similar heat conductivity than brass, in practice the extreme density makes heat transfer difference. Remember that the thermistor measuring nozzle temp isn't actually on the nozzle, it's usually placed in the heat block, assuming heat transfers to the nozzle nicely. If I'm right, and you don't mind an investment, you could see if Diamondback makes a nozzle for your machine. Machine diamond is not only harder than tungsten, but it is a fantastic conductor of heat. To test at least, you could switch back to brass/copper/stock nozzle to see if there's a difference.

Last suggestion I would make is to try PA12 instead of PA6. I found print consistency to be a bit better. That might be personal preference or machine difference though (I use Creality K1C, and various modded Ender 3s).

Good luck, looking forward to seeing what works for you

 

[harrisonpatm]

BTW, I use PA12-CF for my front brake hand lever, for two years now

1989 Kawasaki EX500 (Ninja) - EV conversion, finished but still refining

All good ideas, but mostly it's bad light because it's my cellar-access, dirt-floor basement, whose only light is a bunch of random LEDs I've already put up from different pieces of scrap over time. I've reached a point in my workspace of diminishing returns regarding how much upgrading I'm able...

endless-sphere.com

 

[TexanInTennessee]

Great tips. I will try them, thank you.

I don't recall seeing minimum layer time in Bambu studio (my slicer). I will look for it or a similar setting.

It makes sense what you are saying about the small part.

With other parts I've tried in PA6-CF the first layer won't adhere at all.

With $80 sunk in this filament roll it would be nice to get some more parts out of it.

 

[harrisonpatm]

With other parts I've tried in PA6-CF the first layer won't adhere at all.

In that case, the recommendation is usually glue stick. Perhaps a glass bed

 

[A-DamW]

Grind Hard Plumbing, I've seen them print carbon pieces that look like PETG, and they have the same printer I do. Wish I could ask them.

I'll bet you could do some creative searching of the web and find some contact info.
I find that if I ask nicely, most human beings are amenable to sharing.

 

[Apartunis]

If you need structural integrity then consider subtractive manufacturing instead.

If you want to make earrings and weed grinders then carry on.

 

[harrisonpatm]

If you need structural integrity then consider subtractive manufacturing instead.

If you want to make earrings and weed grinders then carry on.

OP posted in 'members hobbies and interests' about his hobby and interest and asked for specific help regarding it. Opinions on whether said interest are worthwhile weren't asked for. Please refrain from posting like this, it's toxic and rude.

 

[scianiac]

I suppose I have some experience in this field, I have printed probably idk 100kg+ of various CF nylons. The picture is not good enough to provide too many specifics.

First and the most important, to get any quality the filament has to be dry. If you aren't printing it out of a drybox it's probably already wet and will just make ugly prints.

Bed adhesion will always be tricky, I've got some pretty reliable setups using some exotic print beds but you still need to use an adhesive. I use PVP-K90 in 50/50 water/iso and that is the best I've found so far. Print bed temp, hotter is better but honestly it doesn't have as much of an effect as you would think. Brims or hold down pads if you have to. Set the first layer speed to slow, like dead slow, 15mm/s or less.

Print temps you want hot, this helps with bed and layer adhesion, 280 in that area.

Print speeds need to be slower than other materials but that is probably in the print profile already.

The actual type of nylon doesn't make much difference in terms of printability, unless it's PPA which has more restrictions. The final properties will change a lot between PA6 and PA12 for instance but they both need to be dry to print. If you are relying on PA12 not absorbing as much moisture to get good prints it's already too late. It needs to be dry and stay dry. My nylons never leave their sealed dryboxes and they are all filled with molecular sieves to keep humidity near 0% and still I sometimes have to redry something if it's been sitting awhile.

Honestly though at this point there are so many good materials depending on what you need there are other materials that get pretty close or are better in some material properties. None quite have the specific balance that nylons have but depends on what you need from the part.

 

[BlueSwordM]

I suppose I have some experience in this field, I have printed probably idk 100kg+ of various CF nylons. The picture is not good enough to provide too many specifics.

First and the most important, to get any quality the filament has to be dry. If you aren't printing it out of a drybox it's probably already wet and will just make ugly prints.

Bed adhesion will always be tricky, I've got some pretty reliable setups using some exotic print beds but you still need to use an adhesive. I use PVP-K90 in 50/50 water/iso and that is the best I've found so far. Print bed temp, hotter is better but honestly it doesn't have as much of an effect as you would think. Brims or hold down pads if you have to. Set the first layer speed to slow, like dead slow, 15mm/s or less.

Print temps you want hot, this helps with bed and layer adhesion, 280 in that area.

Print speeds need to be slower than other materials but that is probably in the print profile already.

The actual type of nylon doesn't make much difference in terms of printability, unless it's PPA which has more restrictions. The final properties will change a lot between PA6 and PA12 for instance but they both need to be dry to print. If you are relying on PA12 not absorbing as much moisture to get good prints it's already too late. It needs to be dry and stay dry. My nylons never leave their sealed dryboxes and they are all filled with molecular sieves to keep humidity near 0% and still I sometimes have to redry something if it's been sitting awhile.

Honestly though at this point there are so many good materials depending on what you need there are other materials that get pretty close or are better in some material properties. None quite have the specific balance that nylons have but depends on what you need from the part.

Yeah, I'm surprised everyone else missed this.

Nylon and TPU purely REQUIRE drying.

 

[TexanInTennessee]

Set the first layer speed to slow, like dead slow, 15mm/s or less.

Great advice, all of it, and especially this. I never would have thought to slow it that far.

There are several issues I'm fighting. Plate adhesion is just one of them. The print I'm using as a test is a good choice because I was able to print it with the standard .4 head several times, just with grainy results. This model does not give that same issue.

Because of that, I'm prioritizing the print quality before I start thinking about bed adhesion and printing other models.

I kept thinking the flow dynamics calibration was my hurtle, it deposits/leaves extra material on the print. Tried adjusting it but did not see a change after a certain threshold so I don't think that is the issue.

The prints are coming out strong just messy with too much extra debris on it.

 

[scianiac]

Great advice, all of it, and especially this. I never would have thought to slow it that far.

There are several issues I'm fighting. Plate adhesion is just one of them. The print I'm using as a test is a good choice because I was able to print it with the standard .4 head several times, just with grainy results. This model does not give that same issue.

Because of that, I'm prioritizing the print quality before I start thinking about bed adhesion and printing other models.

I kept thinking the flow dynamics calibration was my hurtle, it deposits/leaves extra material on the print. Tried adjusting it but did not see a change after a certain threshold so I don't think that is the issue.

The prints are coming out strong just messy with too much extra debris on it.

Yeah the flow inconsistencies are from water boiling in the filament, pushing out extra filament and then leaving a gap where the steam bubble was. Most CF filaments though will have a rough texture, some more than others, depends on how much and what type of carbon they put in. There are also a lot of effects with the viscosity, temperature, print speed. Nylon is more viscous and you fill it with a bunch of carbon you only make it thicker so if you try to print too cold and too fast it's going to be ugly. There also isn't a lot of disadvantage here to just going slower overall. It takes quite a small part with zero fans in a pretty hot chamber to overheat a CF nylon part (make it all melty from to little layer time).

 

[Apartunis]

OP posted in 'members hobbies and interests' about his hobby and interest and asked for specific help regarding it. Opinions on whether said interest are worthwhile weren't asked for. Please refrain from posting like this, it's toxic and rude.

I offered a couple of worthwhile applications for the tech in question. Trying to make nonstructural parts structural by adding oooh carbon fiber in an unuseful way is not going to work. For that you have to start with a structurally sound material.

 

[TPEHAK]

I printed a horn cowl out of this filament on Bambu X1C printer. It printed without issues with default settings. It required clean plate and wide brim with 0 gap though to prevent lifting from the plate. The nossle is regular hardened 0.6mm nozzle.

https://grabcad.com/library/buell-xb-horn-cowl-mount-t-rex-frame-slider-left-1

I also printed some small parts out of this filament with X1C printer and regular hardened nozzle with default settings without issues. Clean plate and wide brim with 0 gap is must too.

I also printed large parts out of glass filled nylon which behaves similar to carbon filled nylon. Default settings, same printer, same nozzle, wide brim with 0 gap and freshly washed plate before printing. No issues.

Luna M600 Ludicrous V2 reverse engineering and firmware making

It looks like we have a situation guys! If I understand that VESC firmware architecture correct, Luna, or Marcos or whoever is responsible for the firmware did not release the firmware for the new version of the M600 controller (the one I have)! And the released VESC firmware is not compatible...

endless-sphere.com

The geometry of the parts I printed was optimized for FDM 3D printing though (no undercuts over 45 deg). Maybe your parts need to be optimized too. You also need to make sure the part stays adhered to the plate along the whole surface during the print (freshly washed with dish soap plate and wide brim with 0 gap is a must).

 

[TexanInTennessee]

I'm giving up on PA6CF. For the things I want to print I believe ASA might do just fine.

I should have done much more research before getting it for this application.

I assumed I could get it to print, perhaps so, but I'm done running test prints and am ready to move on with my life.

You know what they say about assuming?

But it's okay, I'm used to feeling like an ___, father of girls.

 

[TexanInTennessee]

I'd like to see someone make use of this roll, would make me feel better than thinking about it taking up space on my filament rack.

The first person who submits their full address to me in a DM, I will send you this roll for free and will pay for shipping to the continental 48.

It's a mostly new roll of 1KG Bambu PA6-CF, and it will only collect dust here.

But be patient with me, I live on a mountain and get to town once per week.

 

[scianiac]

I'm giving up on PA6CF. For the things I want to print I believe ASA might do just fine.

For all the nylon I print, it's basically all for products, pretty rarely do I find any of my personal designs require anything beyond ABS strength wise. Partly because I'm cheap but also like ABS/ASA are just really nicely balanced materials. If you need more stiffness the carbon ABS/ASAs are very nice. Lately I've been switching to a lot of the polymaker stuff. If you need more impact resistance (even though ABS is pretty good) there are some interesting super hard TPUs that are almost indestructible.

While I haven't had a chance to test it yet, looking at the testing coming out I think PET-CF/GF (not PETG) might be in a sweet spot, properties comparable or only a little worse than most CF nylons but easier to print, not as moisture sensitive, available in colors. It's got very high stiffness, strength, layer bonding, low creep, decent temp tolerance. Only place it falls down a little is impact resistance but it's still not bad and that depends a lot on what nylon you compare it to. Print wise PET has worse warping that PETG but the fibers should stabilize that nicely and bed adhesion should be no problem.

 

[TexanInTennessee]

I normally buy Polymaker. I was hoping the presets for the Bambu filament on my Bambu printer would help things as far as PA6CF goes.

Bambu is always out of filament too. Polymaker is 15-20 bucks on Amazon, Bambu is 30 (PLA). Easy choice.

Also like Sunlu silks, they're brilliant colors.

 

[TexanInTennessee]

Should have posted sooner but the CF filament was claimed awhile ago. Thanks

 

[harrisonpatm]

While I haven't had a chance to test it yet, looking at the testing coming out I think PET-CF/GF (not PETG) might be in a sweet spot,

Can confirm, I got a roll a month ago to try out and I'm very pleased. Started looking into it because I got into recycling PET bottles for fun using my dead Ender 3, and I was not expecting PET filament to be as nice as it was, even homemade. So I went looking for other purpose-made PET filaments and settled on Fiberon PET-GF, red. Really pleased with the results. Prints hot (280), but otherwise I don't have any issues with it. It's not even terribly expensive.

Replacement chest freezer handle and some aesthetic prints for my kids.

 

[TexanInTennessee]

Nice prints. PETG prints well for me.

Thinking about ASA next. Anyone print with ASA? Or am I getting into another fiasco?

Have only done PLA, PETG so far successfully.

ASA is supposed to be more UV stable but prints at slightly higher temp than PETG, is reportedly slightly more troublesome to print.

 

[TexanInTennessee]

Can confirm, I got a roll a month ago to try out and I'm very pleased.

Would you say it was plug and play with the filament or did you have to fiddle with speeds and temps and set things in a custom way to make it work?

 

[harrisonpatm]

Would you say it was plug and play with the filament or did you have to fiddle with speeds and temps and set things in a custom way to make it work?

I did have to dry it a little, and I did have to tune in printhead temp, but I'm used to that when I try out a new filament.

In general, i do find hot-printing filaments to be more finicky, at least a little bit.