I showed this fantastic-looking dual-color filament in a recent video that has recently popped up at quite some vendors here in Europe and the US.

Check out the Dual Color Filament (Affiliate):

Redline Filament (GERMANY): https://tidd.ly/3nZafag

3DJake (EUROPE): https://geni.us/My3w8NW

Matterhackers (USA): https://www.matterhackers.com/store/l/matterhackers-quantum-pla/sk/M4VVM2GS?aff=7479

When you print this material, you end up with different colors on your part depending on the angle and side you look at it. The filament itself is made by co-extrusion, so using two extruders that feed into a special die, out of which you get a filament strand that’s half one and half another color. During printing, the filament flows in a laminar way through the hotend, and because there is no turbulence, it comes out of the small printer orifice, just as it went in.

This is how it looks when used with a simple V6 nozzle during purging and on a printed part.

I’ve also recently played around with high flow 3D printing nozzles, namely Bondtechs CHT and a DIY version where I drilled holes into a nozzle and soldered copper wires in. Both of these methods function in a very similar way. Polymers are a very bad conductor of heat, so when printing fast, the core of the filament may still be solid, which causes problems during extrusion. In kitchen-terms quite similar to the frozen bread rolls that I put in the oven every morning. If I don’t wait long enough, the center is still frozen while the outside is already crispy.

Increasing temperatures can help a bit, but you might even get to the point where the material degenerates on the outside before it’s melted on the inside. This is where the high-flow nozzles come in. Due to their internal structure, they split up the filament and basically heating it from the inside. The distance from the heating walls to the center of the material gets smaller so the heat needs less time to travel to the core.

This is exactly what made you all curious! The dual color filament relies on the fact that it perfectly flows through the nozzle without any turbulence that could mix the colors up. So let’s test that and print my Mini Stefan figurine first with my DIY high flow version, then Bondtechs CHT nozzle and finally my Über-Volcano!

On my DIY high flow nozzle, I drilled a small hole at the threads, pushed a piece of copper wire through, and soldered it in place to seal it and increase the heat transfer. Here, the filament gets separated into two individual strands, needs to flow around the piece of wire, and then connects again below. Purging some didn’t show any mixing in the material that came out of the nozzle, and also, the final part looked perfectly well.

With Bondtechs CHT nozzle, the filament gets sliced into three individual strands by the blade-like structure, then travels through channels and meets up again before the tip. I couldn’t spot any mixing on the purged material. The final model looked indistinguishable to the reference, which means the flow is still really laminar with basically no turbulence.

All right, let’s step it up! Regular nozzles aren’t really that long to add a bunch of obstacles to introduce mixing and turbulence. So let’s take a much longer Volcano nozzle and take the mesh nozzle concept to its limits. During my DIY high flow nozzle video, the idea was that adding more wires to a nozzle increases heat transfer because we increase the surface area, melt faster and therefore improve flow rate. These wires could potentially also act as mixing elements. In industry similar concepts are applied. Adding such obstacles in a flow path is also used to increase turbulence which you might know from these syringe mixing tips found on two-part epoxy that make sure that resin and hardener get properly mixed before coming out of the nozzle. We can find similar structures in polymer extrusion where static melt mixers are put into the flow path to homogenize the liquid material further. Let’s see if we can do the same on a 3D printer nozzle. The length we have available should be enough for six wires which are all 120° rotated to each other.

I learned from last time that drilling within the threads isn’t easy because the bit wanders off and results in non-centered holes. This time I tried to do it properly that so that it’s easier to watch for real machinists. Therefore, I put my new Mekanika EVO CNC router I’m currently assembling for a review to a simple test and machined three sides of the threads flat to have an even surface to start the bores. Since the wires will reduce the effective flow area and therefore additionally constrain the flow, I increased the internal diameter of the nozzle from 2 to 2.4 mm on the length where the wires will be.

Then I used my ancient vernier height gauge to mark the height of the bores and used the world’s tiniest center punch to create indentations at these locations. Using this approach made drilling the 1 mm cross bores so easy, and they all ended up perfectly centered, which made me at least a bit proud. Then I cut short pieces of 0.8mm copper wire and stuck them into the holes. I preheated the nozzle with hot air because the brass sinks a significant amount of heat and makes soldering quite difficult. Even though I preheated the part, this process was a pain, and I wasn’t proud of the end result. To get a working nozzle again, I snipped off the ends of the wire and re-cut the threads with an M6 die. Not the prettiest, but let’s see if this Über-Volcano works!

And well, the tests didn’t last that long because the nozzle was horribly leaking plastic. I first thought it wasn’t properly seated to the heatbreak, but after checking that and still having the same problems, I realized that some of the solder joints were probably leaking. For once, I used 1mm holes instead of 0.8 mm to make inserting the wires easier, but they didn’t, therefore, seal on their own anymore. The 220°C I was printing at might also have just been just too close to the melting point of the solder if temperatures fluctuated. Since I couldn’t just resolder the joints with all the plastic inside, I burnt out the nozzle with a torch, effectively getting rid of the old solder and wires, and resoldered everything. Nothing leaked this time after heating the nozzle to only 210°C, and even despite the six wires in the flow path, it didn’t seem like a lot of mixing was going on. Also, my Mini Stefan printed really well, and there even wasn’t a lot more stringing visible than with a standard nozzle. Believe it or not, the dichromatic effect on the part stayed the same, which shows that even though the flow path diverts a bunch of times, the flow stays nicely laminar. No turbulence and no mixing.

This shows why double screw extruders are a thing in polymer processing. Mixing highly viscous substances like a polymer melt requires a lot of effort and can’t be done by a melt mixer in many instances. It’s very hard to introduce turbulence in high viscosity fluids. That gets even clearer if we also look at the formula for Reynolds number, which is a quantity that is calculated to determine if a flowing fluid is laminar or turbulent. Low Raynolds numbers suggest a laminar flow. Values above a certain limit are usually turbulent. Due to the big viscosity number in the denominator, Raynolds numbers of polymer melts are very small and therefore suggest laminar flow. The smooth wire and the center placement in my Über-nozzle are not enough to introduce turbulence, and therefore the dual-color filament still works here, which is quite impressive!

Before we finish up, I, of course, also had to test the melting performance of this beast, which was really impressive! Keep in mind, these results are all for 0.4 mm nozzles so not fully comparable to my last tests where I mostly worked with bigger tip diameters. A standard V6 performed the worst, the standard length CHT performed basically the same like a longer Volcano nozzle and the Über-Volcano beat everything by a wide margin! The molten material came out super smooth all the way up to 40 mm³/s which shows the outstanding melting performance due to the wires in the flow path. I honestly would have expected that it performed worse due to all the additional flow resistance! It might have performed even better if the Hemera extruder wouldn’t have skipped so I’m exited to see what we can achieve with the new large gear extruders or maybe even Proper Printings Belt extruder! I have to point this out thought that adding heat-conducting elements like wires into a nozzle is patented by 3DSolex, so check your local patent laws if you want to play around with this concept!

So there we have it! The crazy Über-Volcano performed seriously well. We’ve also seen that polymer melts are hard to disturb, and you should be able to use this fancy dual-color filament on any FDM 3D printer you have. Though, do you think investigating a mixing nozzle for filament 3D printers might be interesting? Could this be used to perform full-color CMYK 3D printing with one of these multiple-in, one-out hotends? Leave your thoughts and ideas down in the comments!