Ask Hackaday: What’s Up With This Carbon Fiber Printer?

The Hackaday Tip Line has been ringing with submissions about the Mark Forg3D printer, purportedly the, “world’s first 3D printer that can print carbon fiber.”

Right off the bat, we’re going to call that claim a baldfaced lie. Here’s a Kickstarter from a few months ago that put carbon fiber in PLA filament, making every desktop 3D printer one that can print in carbon fiber.

But perhaps there’s something more here. The Mark Forged site gives little in the way of technical details, but from what we can gather from their promo video, here’s what we have: it’s a very impressive-looking aluminum chassis with a build area of 12″x6.25″x6.25″. There are dual extruders, with (I think) one dedicated to PLA and Nylon, and another to the carbon and fiberglass filaments. Layer height is 0.1mm for the PLA and Nylon, 0.2mm for the composites. Connectivity is through Wifi, USB, or an SD card, with a “cloud based” control interface. Here are the full specs, but you’re not going to get much more than the previous few sentences.

Oh, wait, it’s going to be priced at around $5000, which is, “affordable enough for average consumers to afford.” Try to contain your laughter as you click the ‘read more’ link.

That’s all we know. There’s no mention of how carbon fiber get into the printed objects. Is it embedded in a plastic filament, like the previously mentioned filament, or something far more exotic? If the former, the idea of printing with carbon fiber doesn’t make sense except in a very few niche cases.

I’ll be one of the last people on earth to claim experience in composite fabrication, but I have built a few things with resin and cloth, and know a little bit about why 3D printed carbon fiber parts is a really dumb idea. With composites like CF, fiberglass, and Kevlar, all the strength is derived from the bias, or the direction of the weave. If you want to make a part strong in one direction, lay your cloth down at an angle and slather on some epoxy. Need it stronger in two directions? You need two layers of cloth, at least.

This is rudimentary stuff, and doesn’t even touch on vacuum bagging, and degassing the resin. There’s a whole host of other things that make carbon fiber crazy strong, and you just can’t do these things by hitting a ‘print’ button.

That being said, there are probably less than a dozen people on the planet that know how this printer actually works. Is there some magic in the way Mark Forge filament is produced? Are there special slicing algorithms that capitalize on the unidirectional strength of CF? Is this just a tool for printing things with carbon fiber for the sake of printing things out of carbon fiber? I don’t know.

If you have any information on how this printer actually works, post it in the comments. If you have any materials testing data on a part manufactured on a Mark Forged printer, post it. We’re dying to know if this is actually a cool product, or just a cool business model and a bit of PR.

40 thoughts on “Ask Hackaday: What’s Up With This Carbon Fiber Printer?

  1. Makes sense , Run a line of cf fiber to replace filament. Then it mixes with epoxy as it is extruded out.

    Normal procedure :
    layout piece of fiber,
    squeegee lots of epoxy over it

    1. What you are looking for is an epoxy that can be heated into active form while being solid before so it can be used as a sabo. Then as the ‘filament’ passes the heater it would make epoxy, expose the cf thread and lay it down.

      1. Even if this did work, you lose a lot of the normal benefits of CF construction by treating it like a 3D printed filament. You’d at least want this to handle fabrication like a winding or wrapping machine to get much usefulness out of it. That could potentially be useful, a CNC lathe that can do prepreg CF filament wrapping… with built in laminate strength calculations and wrapping software…

  2. Carbon fibers in a material would only make the finished product strong within each extruded filament, right? I mean, you’d need the part to reflow and have fibers connecting between layers for there to be any strength between adjacent print lines, right?

    1. They have an example where CF is extruded alongside a (PLA | Nylon) part as a ‘dual color’ part. I *guess* you could have the plastic part serve as a mold, heat the whole thing up, reflow the CF filament, and break off the mold. That just seems a little insane, though. Given the melting points we’re talking about here.

        1. “f it were really a carbon fiber product manufacturing device, the correct name would be a loom, not 3D printer” Why wouldn’t mold be an equally correct name? In the event there’s a need. no doubt a 3D printer that produces carbon fiber reinforced plastic part t it will be designed. Most likely it’s going to be more complex than printing plastic, and less complex that that Lexus loom.

  3. Good comments. I had the same questions when thinking “carbon fiber printer”. I’ve worked in fiberglass for a number of times (automobile body work introduced me to it, used the things learned there in other fabrication projects), and you are absolutely correct – to get any use out of “carbon fiber”, it has to be a weave.

    A lot of “carbon fiber” components we see produced today are basically a CF (or simple substitute) overlay to give it the “look” of CF, without actually having to manufacture the whole thing in CF. “3D printing” just won’t work for anything more than a “novelty” when it comes to CF.

    If, on the other hand, this is a loom that weaves it in place, then we might be talking about something that changes the game of production.

    1. Carbon reinforced plastic uses short carbon fiber filaments to significantly increase strength and stiffness for moldable parts so I don’t agree that weaving is a requirement. For high-performance parts the problem is to remove all excess binding material between carbon fibers but those designs aren’t really in the potential market for something like this.

    1. I think it would be more analogous to putting rebar into mud to form concrete. Yes, the CF strands are nice, but you’re not using them in any way that would give you any serious strength.

  4. There are a few clues. One of which is that it emphasizes that the fiber is continuous, not chopped. So my guess is that it has long fibers in some sort of polymer matrix. From the movie, it looks like it lays the CF layers down as flat ribbons. So maybe that is how they cam realize their claims.

    1. Where was the emphases that that fiber is continuous? The video clearly states short fiber, as does printed formation,where is says it’s chopped. Even if it lays a strand of fiber in each layer, in the end the strength is going to depend on the adhesion between layers

      1. On the homepage.

        One part. Thousands of Continuous Carbon Fibers.

        The incredible strength of carbon fiber comes from the long, continuous strands that carry load down the entire part. This is why space shuttles, rockets, and Formula 1 cars are constructed from continuous strand carbon. And it’s how we print. Don’t settle for plastic with a dash of chopped carbon fill. Longer is stronger.

        1. The Space Shuttles were 1960’s aluminum airliner construction technology, wrapped in a heat shield and made to fly at hypersonic speed. The USSR’s Buran was a bit more advanced, their tile carving system was CNC vs NASA’s manually operated 3D tracing pantograph.

          To extrude plastic with a continuous fiber embedded, it would require a cutter at the nozzle for breaks in extrusion.

  5. It could be just that the plastic filament they used is pre-impregnated with chopped short strand of carbon fiber. Chopped fibers don’t have the same strength as a continuous one typical of CF construction, and can even weaken the plastic matrix that carries it.

  6. It isn’t like that kickstarter, which is little bits of CF mixed with PLA, its a thread of CF coated in binder. You can see the second extruder head has a much smaller heater because there is much less plastic to melt. It lays a long continuous thread of CF into a part, which gives it a lot of strength. CF is not strong because the fibers interlock, the strength comes from the binder having the fiber to grab onto instead of all the strength coming from the cross sectional area. Carbon fiber is more like cement with rebar than vitrified fabric.

    1. also the relatively thin coating on the CF and the fact that its plastic means you don’t need to degass, and you get a decent ratio of binder to carbon fiber. They claim better than 6061. If they had lead with that they would have faced a lot less skepticism.

  7. From the video it does not look as though they are really doing anything special. The extruder looks like any modern reprap or derivative. They have used what looks like an Igus rail and linear bearings from the X axis (and presumably the Y). That’s a nice upgrade but not huge.

    Laying fibers down this way is interesting and kind of cool considering all of the advances in molded Fiber Reinforced Polymers (FRPs… search “thermoplastic matrix composites”). I just don’t think you would reach any appreciable strength or stiffness increase as your weak point is still the low shear strength between adjacent layers.

    Stiffness in traditional composites come from longer length fibers. The very short sections in the “carbon fiber” impregnated PLA materials really won’t do much for you. The bulk modulus may be higher for the material but you won’t see anywhere near the huge anisotropic stiffness increases you would get from long fibers.

    This looks like it is laying down a continuous fiber and then printing out a thermoplastic (PLA/PA). I could see sheets of material actually being quite strong but an actual epoxy matrix composite is going to be way stronger/cohesive and stiffer using a weave/stitched fabric. If you want to make carbon or glass fiber parts using 3D printing it would make way more sense to just print molds and buy fiber matte and epoxy. I was actually surprised at how few people are doing this (ex:

  8. carbon fiber is electrically conductive and could easily replace corrodible bonding jumpers and thinner as the bonus. Also as a deposited material be the first stage of laying down additional more conventional layers without the initial accounting for much thickness. Aluminium typically is not good when used with carbon in aviation. If a layer is ‘print’ bonded, it may reduce the downside.

  9. I’ve done quite a number of CFK/GFK parts myself (one is my recumbent, which has about 50.000km of total millage) and some some knowledge from university.
    As mentioned above he most “critical part” of laminating in general is to keep the matrix (resin) as thin as possible between the fibers (not talking about the strand, but the fiber, the stuff that’s a lot smaller than a human hair). That’s what all the vacuum and autoclave stuff is for: pressing the fibers as close together as possible, sucking out the useless resin.
    This means the “must be a weave fabric” part is not entirely true. For (partly) handcrafted parts it’s just the best way to get done at all (the whole stuff gets REAL slippery, when soaked with resin). OK there are unidirectional fibers with just a very thin glass fibers in 90° to hold the other fibers in place. The high end parts are most likely layers of unidirectional fibers in directions, with serve the use case most (in the sense of which forces to withstand).
    To the printer: If you have a program with which you can influence the direction in which you lay down the endless fiber, it would be a good, strong part, if carefully designed (meaning the direction of the fiber). BUT you will lose most of it, because of the way to much resin between the single layers (maybe even all). So the single layers will get very strong, but that makes not much sense, because you loose the benefit of the all-to-gether strength of the layers.
    What could be a use case is very tiny parts, where you use a single strand of fiber as the finished part. Given the fiber is soaked with the (thermoplastic) resin properly in this procedure (which a doubt a bit)

  10. And you’re loosing the advantage to build a curved part, where the fiber follows that curve (in a mold), which is most likely what you want anyway to get the srength from the shapedesign also.
    And one thing to the fabric. Most “shiny, glossy” parts, like bicycle frames and such, have a fabric carbon layer which is mostly decorative (yeah I know not entirely true, but to good parts)

  11. I can’t see how the volume fraction of carbon in this would be anything worthwhile compared to traditional prepreg autoclave materials. But compared to say a chopped carbon sla print, it would be leaps a bounds better. Assuming they are coating continuous carbon strands with plastic as they go and orienting them in the same direction for each layer this is actually a pretty solid idea if economical. Of course any turnaround areas need to be outside of anything your loading in the printed part…

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.