Carbon Fiber filament wear on Rostock MAX V2 extruders

[VAXHeadroom]

All,
I have a client who wants exclusively carbon fiber printed objects, using probably on the order of 2-3 kg/week. What wear can I expect on the extruder nozzle (I know it's going to wear, just wondered how much)? Any other 'consumables' I need to pay attention to?

Thanks in advance.

[Jimustanguitar]

https://youtu.be/mk30MPLoYf4?t=2006" onclick="window.open(this.href);return false;

Check out what Sanjay and Josh from E3D have to say about it. I'd definitely advise getting a hot-end with a steel nozzle. Or order lots of spares and figure out a schedule for changing them.

[VAXHeadroom]

https://youtu.be/mk30MPLoYf4?t=2006

Check out what Sanjay and Joch from E3D have to say about it. I'd definitely advise getting a hot-end with a steel nozzle. Or order lots of spares and figure out a schedule for changing them.

Thanks! I was just about to post that same video link
https://youtu.be/mk30MPLoYf4?t=33m27s" onclick="window.open(this.href);return false; This time index is where they start talking about it.

[geneb]

From what I've been told, the carbon fiber filament doesn't add anything structurally to the part and may actually result in a weaker part due to the volume consumed by the fibers themselves.

g.

[mhackney]

Good timing VAXHeadroom. I've been printing CF PLA exclusively for 2 weeks straight. 2 KGs of ColorFabb CF XT-20 (XT) and 3Kg of Proto Pasta CF PLA (PP). I plan to write post on my findings. These are 2 very different types of F although they both have about the same % of fibers.

I printed all of this with a brass .4mm E3D nozzle. I measured the diameter of an extrusion before, multiple times during and after printing all 5 kg. I see no measurable wear at this quantity of filament. Which CF are you planning to use? Here are my impressions:

1) I can't really tell the difference in the final parts aesthetically or mechanically (simple fracture test). Both filaments look beautiful, a high tech matte black. The ColorFabb has a much higher extrusion temp so I suspect it handles temps above 200°C better than Proto Pasta if that is important. I also can't tell any difference in wear when I run my 3D printed fly reels on my wear tester (basically just spin them for days!).

2) The Proto Pasta prints at about 210°C beautifully for me. I get no blobbing or other print artifacts after careful testing and tweaking (see the link in my signature). Colorful prints at 245°C and comes out very viscous. It is very sensitive to having a good "sticky" surface to adhere to. I print on PEI in both cases. With PP, the PEI is 50°C. With XT it is 65°C. XT prints pretty well but does slowly build up blobs on the nozzle that transfer to the part. Nothing I tried prevented this but I was able to get pretty close. I polish my tips and even with that, there is sticking. The XT stuff is very viscous and sticky when molten so I suspect that is the issue, PP is much more fluid.

3) You really need to use a geared down stepper, especially with XR. I use a 5.18:1 geared stepper on the ezStruder. It's perfect. The stock EZ sturdier does not have enough torque. You might be able to sneak in a short print but it will eventually stall and skip steps, especially with the more viscous XT.

You also need to use the "Bowden to the Gear" modification to support the filament all the way to the drive gear. Both of these filaments are brittle and will snap in the path below the drive gear unless supported. PP is more brittle than XT in its filament state but I did not observe any significant difference in testing printed parts. This mod is the once that passes the Bowden tube through the PTC fitting (most of us use the E3D plastic fittings that do this) all the way through the black molded ezStruder parts right up to the drive gear. There is a discussion on ezStruder mods in the modification forum that Brian documented that includes this. It is critical for these filaments or you will be spending a lot of time disassembling your ezStruder.

You also need to make sure the filament path from the spool to the extruder does not put an inordinate bend on the filament. Again, it is brittle and will snap when yu least want it to (i.e. printing the last layer on an 8 hour print).

4) Both are much more sensitive to retract speeds. SLOW DOWN to 20mm/s retract and you'll be fine. Longer retract length than normal (1.5mm) also since the materials are more viscous.

5) Infill - this is an interesting one! You need to use an infill that layers over the previous layer exactly. Neither of these filaments bridges worth a damn, so infills that require bridging are problematic unless you slow waaaayyyy down (15mm/s infill speed). A typical crosshatch infill is a PIA. I ended up using Simplify3D's grid infill as it builds each layer on top of the previous. It was perfect for this fiber. None of KISS' infill types could be printed at reasonable speeds.

6) Overall, slow down. XT at about 40mm/s tops, PP at about 55 mm/s tops.

You might have an inkling to speed up when you see it printing smoothly but RESIST! I did that several times and each resulted in plugging. If you think PLA is hard to clean out of a nozzle, you haven't lived until you try to clean out these CF (especially the XT) filaments!

On the wear question, I suspect you can print 50+ kilos with a brass nozzle with no wear. The stock Rostock .5mm nozzle is probably going to handle even more.

cheers,
Michael

[mhackney]

Gene, the CF filaments I use do have better mechanical properties. They are significantly stiffer as measured on two home-brew tensile test machines. And, when the CF fractures, it is a very clean, straight fracture even if the fracture is not aligned with the extrusion directions of the top/bottom layers and infill. I found that very interesting. When it snaps, it makes a clean "SNAP" and it's broken.

Like all things in 3D printing, no one has really done scientific measurements on strength test. All sorts of factors like # perimeters, shells, infill type and % will have an impact. I tried to compare parts printed the same way and make "pallets" to test in every material I get. The CF filaments do create parts that are more rigid and more fracture resistant than standard PLAs.

[mhackney]

Sorry for the poor photo. These are PP parts I am printing now. The first shell over the grid infill at 50mm/s and 210°C with a .4mm brass nozzle (Kraken hot end). When done, the surface will look excellent.

[geneb]

TIL!

g.

[mhackney]

Then it's another good day!

[mhackney]

Here's a better photo showing the type of infill I mean. This is triangular infill from S3D. I think MatterSlicer does this one too.

[mhackney]

I edited the first post I made to include more info on the extruder - IMPORTANT STUFF!

[VAXHeadroom]

Excellent insights by everybody, thank you all so much!
I've only printed a few parts, but was able to use the same gcode as regular PLA - the prints came out beautiful.
One option for the nozzles - my brother has a full machine shop and suggested making a nozzle out of bronze - copper+silicon/tin - instead of brass - copper+zinc. It will be way harder and still transfer heat approximately the same. I was also thinking of getting a brass spare and having it titanium coated - should be cheap to get done and then it would last FOREVER.
But it sounds as though this may not be an issue - I'll have to keep an eye on my nozzle - it's a stock 0.5mm Rostock Max V2.
Also: would you think running the filament hotter would be better wear-wise? Seems to me it might be...

[Jimustanguitar]

was also thinking of getting a brass spare and having it titanium coated - should be cheap to get done and then it would last FOREVER.

Now that's an idea... I don't think anybody's plating them yet.

[mhackney]

Running hotter does lower the viscosity and I would expect less wear but might introduce other problems. Which filament are you using? They are very different.

[VAXHeadroom]

Running hotter does lower the viscosity and I would expect less wear but might introduce other problems. Which filament are you using? They are very different.

ProtoPasta bought through matterhackers. I'm running it at 230 and 80mm/s - I think that temperature is artificially high, I expect the actual nozzle temp is closer to 220. The reason I say that is if I set the temp to 220, the reported temp falls to almost 214 and when it does the plastic will hardly flow through it. At 230, the reported temp is 227(ish) and continues to flow fine, so I expect the thermistor, which is reporting the temperature of the hot block, is being fooled a bit being so far from where the nozzle gets cooled by the flowing plastic, so I'm just compensating seat-of-the-pants. I've been running FDM machines for 4 years so I have a pretty good intuitive sense of what's going on...
I'm planning on doing some fracture and crush destructive tests for my client soon, I'll report on them here when I get some data...

(edit: these temperatures/speeds are all from standard PLA, not the CF - I did try one CF part and used the same profile and it worked fine, but haven't really done much with it yet)

[VAXHeadroom]

OK - have several test prints done and one pic to share.

On the left is 80mm/s - might actually be slower than that since I had the cooldown parameter set.
On the right is 40mm/s - no other parameters changed. WAY smoother.
We did do a break test on one of these at the 80mm/s speed - fractured cleanly across layers as expected, but marginally strong enough.
I'm hoping the 40mm/s one is noticeably stronger. Will let you all know

*edit* Oh, and using Slic3r for this - I think it produces much better output than MatterControl/MatterSlice.

[teoman]

Carbon is much more thermally conductive so will cool the hotend down. By how much i cannot say.

[mhackney]

Having printed kilos of both CF filaments on the market I can say two things:

1) There is no noticeable effect on temperature or temperature stability of the hot end (E3D and Kraken)
2) I have not observed any wear on a brass .4mm nozzle (E3D & Kraken)

[geneb]

I think it's the formulation of the material you're printing, mhackney. The E3D guys did a presentation at MRRF last March. One of the things they showed was how a carbon fiber filament devoured the back end of the nozzle. The orifice didn't change that much but the space behind it was carved out. They'd frozen the hot end and then milled it in half to show the issue. It was pretty cool.

g.

[mhackney]

Most of my CF was the protopasta stuff and there is no doubt it extrudes at a lower temp and is much more fluid than the ColorFabb stuff. Being in the UK, the guys might have been using ColorFab XT CF and that sort of wear would not surprise me. I have printed a fair amount off the CF stuff though and so far have not had an issue.

I understand what they printed now, I can see how the fibers pushing into the slope down to the orifice could wear. By the time the filament gets to the nozzle, the CF fibers could be aligned and not as abrasive. Did they leave a presentation or paper that I could take a look at? Did they have data on how many hours of printing exhibits wear?

thanks,
Michael

[0110-m-p]

Didn't see this posted yet in this thread, but it is likely good information. Granted they are trying to sell parts, but the cross-sections they took of brass nozzles show pretty extreme wear with low hours.

http://e3d-online.com/is-carbon-killing-your-nozzle" onclick="window.open(this.href);return false;

[geneb]

That looks to be a much more fleshed out presentation of that than they did at MRRF. They covered a LOT of stuff in their talk, the CF issue was only a small part.

g.

[mhackney]

So this might be helpful to some and purely conjecture on my part until I experience it again...

I printed ProtoPasta CF PLA filament for about 2 weeks solid, 8 hrs a day to make the parts for my fly fishing reel kits. My big event was this past weekend. I shut the printer down before I left. I did not print all day Thursday, Fri, Sat, Sun or Monday. On Tuesday I ran into all sorts of problems with filament starving. The layers would go down with large gaps where filament did not extrude. Almost like there was grease or something on the PEI print surface. I suspected the extruder and/or hobbed gear. I tore them apart and things seamed better for 1/2 a print and then failed again. I ended up having to tear down my Kraken hot end. Inside I found a rock hard, black blob stuck to the side wall that took up most of the passage. There was enough space for some molten filament to get by but not a lot. I used a 2mm drill and carefully removed the plug by hand. It looked like the CF PLA. I shut down each night with the everything "loaded". I am wondering if maybe the CF PLA thermoset to create that blockage, it really was rock hard. Once cleared, I'm back to printing nirvana.

[EL Cuajinais]

I polish my tips and even with that, there is sticking.

Michael, thanks you for all the information you graciously share on these forums. I've seen this mention of polishing the nozzle on your guide as well. How do you do it? What tools/substances do you use to polish the outside of your nozzle?

[mhackney]

Polishing is actually simple and no tools or chemicals are needed! I simply rub the nozzle briskly on my pant leg - denim / blue jeans material works best but any stiff cotton fabric (old TShirt) will work just fine. You do not want to use any polish, it will plug the office and require you to clean it out. Brass can be easily friction polished by simply rubbing it on clean cotton fabric.

A polished nozzle helps prevent filament from sticking to it. It works!