I have always struggled to keep the hose in small infusion buckets. Every now and then a clamp slips or I bump a hose and it pulls out and air goes in and everything gets messed up. Besides the bucket screwed to a board (with a fender washer inside the bucket) to hold pots of resin from tipping over – which is my favorite – this is a positive new development:
I printed this from PLA and am a little worried about the thing snapping off at the lip that holds the bucket lid. So far though it works pretty well! Can’t bring myself to use the Markforged and make a stronger one – it’ll be like five bucks! This one on the Fusion3 with PLA is 80 cents or so.
You can see if featured in the videos for Laminate Samples #7 and #9 in the material library.
If you want to print one – check out GrabCAD and search for “infusion”. This won’t be the last version, and you’ll have to scale it to use the right size hose – it’s set up for 3/8″ (9mm) OD hose – but it might be useful.
Over the last five years, 3D printing with long-fiber composites has become a thing that actually happens. The combination of automation, materials technology and the maturing of the 3D printing industry has driven the first round of companies developing and marketing this technology. Dozens of patents have been granted to these first-movers in the the automated, moldless composites space and it seems like there is some overlap in the different approaches these companies have taken – but also a lot of variety. Here’s a look at the most prominent participants – how they compare and where they stand in 2019.
Of these companies, several focus on thermoplastics and others on thermoset resins. The most established is probably Markforged, which has been shipping usable composites 3D printers since 2014. Continuous Composites and Moi use similar robotic arm-based mold-less “printing in thin air” processes, seemingly with thermoset resins. Arevo has shown larger scale multi-axis thermoplastic machines – and has demonstrated a printer bicycle frame. Orbital Composites released the Orb-1 based off a Kuka robot – but it won’t be available until early 2020. Desktop Metal came late to the 2019 party with their “Fiber” system, which looks like it’s aimed right at Markforged… and almost more interestingly it will be sold only as a “hardware subscription.”
The two issues I see are related to patents and to the need to compete with already-existing composites technologies. Many of these companies have patents that seem to my untrained eye to overlap and pretty serious ways. They also overlap onto existing automatic fiber placement (AFP) and automatic tape laying (ATL) technology – which are just different flavors of automated lamination. Both Markforged and Continuous Composites have heavily invested in securing patents on their technology. To the best of my knowledge, nobody has made a fuss over patent infringement yet, but it’ll happen once there is a market to defend.
The market will only really exist as a meaningful alternative to traditional composite technologies once the process is much better developed. This may be fast for some applications (the Markforged, Anisoprint and Arevo machines already have end-use applications) and may take way longer for others. Additively manufactured metal structures that can provide weight savings through generative design, hollow areas and lattice infill may be more of an immediate challenge to traditional high performance composites.
Markforged:
The Markforged system has proven itself useful (I have one and it is pretty awesome) but it is really a half-step toward printing composite structures. The fiber is only laid in one plane. It can be very useful for the right type of part – but more as a metal alternative or as a stronger alternative to other 3D printed options. Markforged has numerous patents relating to thermoplastic printing with fiber and multi-axis printing but have also entered the metal additive market which may have more potential in the short run. Markforged gets plenty of attention but here’s a video overview of their process:
Continuous Composites:
With it’s first patents filed in 2012, Continuous Composites is a pioneer in printing unsupported parts using fiber in a thermoset resin matrix. Cured with UV light or heat as it leaves a robot-mounted extruder, the impregnated fiber bundles cure in mid-air – a process they call Continuous Fibre 3D Printing (CF3D). The potential for certain types of parts is clear, but so far there is no evidence of printing in multiple axes – though this is probably more of a fixturing and software problem than a limit to the technology. This is an impressive company and along with Arevo (which focuses on thermoplastic resins) Continuous Composites looks like a leader in the race to print strong free-form composite structures. Here’s a video of their technology in action:
Moi Composites:
Founded in 2018 as a spinoff of Politecnico di Milano University in Italy, Moi Composites has similar technology to Continuous Composites. Their Continuous Fiber Manufacturing (CFM) technology uses thermoset resin mixed with glass fiber and cured as it is extruded. They seem to have the ability to lay fiber at different angles and to build relatively complex parts in space. This video shows their technology in 2017:
Arevo:
The Arevo additive system looks kind of like a huge Markforged print head attached to the end of a robot arm. Their whitepaper describes a process called Direct Energy Deposition (DED) in which a laser heats a thermoplastic filament and a compaction roller consolidates the filament eliminating voids. They claim a 50% fiber fraction – so half the weight is carbon fiber – which is really good. They seem to be focused on an electric bicycle as their first product… but the aluminum fixture in the video looks suspiciously like a mold! This is very close to an automated fiber placement process that companies like Automated Dynamics and others have been using for a long time – the laser, the thermoplastic prepreg and the consolidation roller. Just because it isn’t really that new doesn’t mean its not cool!
Anisoprint:
The Anisoprint printer and prints seem very high quality and similar in size and output to the Markforged system. Apparently they have an option to use basalt fiber – which is cool. The process seems to combine fiber with filament during the extrusion process rather than extruding plastic and then extruding a “prepreg” tow in a separate operation like the Markforged. A 2019 article in 3DNatives shows some non-planar extrusion and gives more information on the technology. It will be interesting to see if they can develop full 3D fiber printing with thermoplastics. The software looks good:
Orbital Composites:
In 2015 and 2016, Orbital Composites made a splash with the idea of 3D printing composites in space. Now in 2019 they have released the Orb-1, which appears to be a FDM / thermoplastic composite printer based on a Kuka robot arm. It’s a neat idea but quite expensive ($99,000) – it also has a huge 1m x 1m x 1m work envelope. It will be neat to see a video of the Orb-1 in operation.
Desktop Metal:
Desktop Metal unveiled the “Fiber” system in late 2019, and it isn’t available yet. The engineering looks very solid, and it uses a variety of thermoplastic resins (PEK, PEKK, PEEK…) and a separate “Micro Automated Fiber Placement” head that lays actual “tapes” of fiber reinforced plastic. I haven’t seen or held a part, but expect that they are similar (maybe better) to that Markforged parts. Higher performance plastic resins will give better stiffness and higher temperature handling. Most interestingly, the printers will be “sold” by subscription only. Apparently the price will be around $3500 per year. I’m excited to learn more about this one!
Magnum Venus Products posted a video of their large format 3D printer – one of the first to use thermoset instead of more common thermoplastic materials. The video only shows simple parts, no long fiber reinforcement and no post-machining, but it is an exciting development for automated composites.
Most of the other large additive options (BAAM and LSAM) use thermoplastic feedstock, often mixed with a short fiber reinforcement. The output is amazing, especially when machined and primed, but thermoplastic feed-stock is expensive especially if high temperature service is needed. The inter-layer adhesion isn’t ideal (“welding” hot material to cooler material as it is extruded) and shrinkage is a problem. On the bright side it is all recyclable – in theory.
The argument for thermoset based material is that it can be formulated to perform in various ways and the shrinkage is controllable. The cured parts are strong in the z-direction because the beads are bonded together. Unfortunately it is not a recyclable end product. As this technology becomes more mature, it will be interesting to see where each system finds it’s sweet spot!
I’ve had really good luck gluing Markforged Onyx (Nylon) to composite and metal with Plexus MA830. Probably lots of similar methyl methacrylate adhesives would work well too!
