*QUESTION:* Which material combinations would you like to see and why? Want to get your 3D prints to the next level already today? Check out our Heat Set Inserts and Tools at cnckitchen.store (Free shipping worldwide starting at €100).
maybe PLA with a core of water soluble support like polydissolve, maybe to do some really light weight prints, insted of solid lines you will get tubes of PLA
This process can indeed open a path to multimaterial filament such as continuous carbon/glass fiber core embedded (inside an easy to respool material at first, for test purposes), the non strech fiber mechanical properties might be a hassle to deal with though... PA6/12, PC, TPU or Nylon for high end continuous fiber embedded filament, but might be even trickier to process. Anyway, good luck if you try this. Great video as always !
Roughly every year or so, someone makes a completely game-changing 3D printing video showcasing an inventive new way of doing things which could open up so many new possibilities. This is one of those.
And at least a third of those seem to be done by CNC Kitchen. How many of the things Stephan has showcased over the years have become standard practice?
Yep I was wondering which 3D printer manufacturer is also watching this and thinking to add a filament-diameter scanner into the printhead to be able to automatically compensate for under/over extrusion during printing. Maybe this is even more important for flexible filaments to get a really consistent printing results?
Its interesting, but its already being done... Continuous Carbon Fiber filaments for high strength parts for example. "fiber" filaments rarely have increased strength, because the fibers have to be chopped up so small that they don't do much. So Continuous Carbon Fiber filament and the like were invented to change this. The game changer is being able to make multifilament at home, even as a proof of concept.
I work maintenance in a cable factory. The first 5 minutes of the video were pretty much most accessories of an extruder re-invented lol. You should really look up Kabmak extruders to get some inspirations because there are things like vacuum attachments and straightening rollers that could be added even to your machines. You could also add a mechanical brake to the pay-off spool to add pre-tensioning and it would make the plastic more consistent as it would feed through straight. The rollers serve the same purpose, but for wobbly lines. Another important addition would be a hot air blower to pre-heat the filament so you couldn't have to run a volcano type nozzle or anything special. You would need a lot less forces on that filament if it went into the nozzle already at say 50C. As for what you could be adding to it? Oh boy! Make 3 separate filament pay-offs with U-shaped "shells" on 2 of them and one filler. You could just add nylon strings or fishing line to the center spool and print multiple times as much filament at once. Hell, add a caterpillar-type tensioner brake to it and you're gonna have time to change filament on the go without interruption if you find a way to weld them in between.
I agree with all of these suggestions! Preheating the filament is a great idea, even with something simple as a halogen light bulb. And the roller suggestion is also great!
Preheat the filament too was first thing I thought when he was struggling to pull it through too. Can't believe he didn't think of that when he was basically already doing that to stretch down in size to insert it.
Off topic, but I think you might be the right person to ask this, if you don't mind helping out a random internet stranger :D I'm working on creating coaxial nylon-nichrome wires where I want to embed a nichrome wire in the core of a nylon monofilament line. The ultimate goal is to achieve this with 0.1mm D nylon and 0.01mm D nichrome wire, but I'll first try to do it with larger D nylon. What Stephan built here is roughly what I am also planning on doing, and in addition also feeding the nichrome wire through the core of the nozzle. Do you think this would be workable and do you have any tips? Thanks in advance!
"Yo Dawg, I heard you like layer adhesion problems. So I put layers in your filament so you can have layer adhesion problems while you're having layer adhesion problems."
@@stevrgrs candy makers dont make candy with two different plastics and different melting points. Stuffed candy is generally coated or injected, as for hard candy it's rarely if ever done with a mix of, let's say butter based caramel and hard candy, as the butter caramel would leak out of the strip before it hardens, ruining the pattern. Irony or not it's apple to oranges. he's right about temp differentials being a problem with multimaterial printing.
Hard candy is made in a similar fashion. All the design is created in a large (8-9" diameter) chunk then stretched into the final bite-size form. The pattern is retained through the stretching process same as with yours.
His old video about the PC core ABS he mentioned actually talks about the same thing. I believe the original researchers made large blanks with the desired pattern that they then had a machine to extrude into a filament. Exactly what you’re thinking! But requires specialized equipment. Assuming I’m remembering correctly! 😅
This TPU core experiment is just GENIUS! That's why I love your channel, you do science with your experiments, not just funny things for likes and subs
Why a solid core instead of the TPU and PLA emulsified together? People already use silk PLA for flexibility in certain applications since it already has TPU in it.
@@zackj997 The solid core unmixed turns it into a composite. When you have two materials with different properties together, their strengths can add together in ways that wouldn't if they were thoroughly mixed. In this case, the flexibility of the TPU functions like dampening rebar, preventing fracture propagation while preserving the stiffness that PLA offers.
It sounds genius if you know very little polymer science. PU and PLA are not chemically similar so they form what we call non-compatible blends. These blends are basically those that cause boundary layer separation because the two materials are not chemically bonding with each other, this forms a weak material that is worse in most aspects than the homopolymers themself. This is why when making polymers blends with dissimilar materials we use compatibilizers to act as a glue and bind both the materials together.
4:13 This phenomenon is actually a result of two combined effects: necking and strain hardening. When tension is applied to the nylon filament, it begins to undergo plastic deformation at its weaker points, leading to a reduction in diameter. At the same time, the nylon string strengthens in these areas of reduced thickness, which concentrates the strain on the thicker segments of the string. The combination of these effects causes the string to shrink to a new, constant diameter before ultimately breaking. Articles for more information: en.wikipedia.org/wiki/Necking_(engineering) en.wikipedia.org/wiki/Work_hardening
It's kind of right but not quite, the effect is due to the majority of the chains in the polimer getting aligned and because all the chains are in the same orientation they become stronger
@@hugofernandez8522 Yes, you are correct, in the case of polymers this effect is the result of chain alignment and not crystal shifts like in metals, but still, this effect is called strain hardening (or work hardening) in both cases
As others have mentioned, this is not work hardening, it's chain alignment. No, they are not the same thing. Work hardening is a crystallographic behavior (polymers do have crystal structure) while chain alignment is just generally the idea that you can do things to the material that cause the long molecules to be oriented in roughly the same direction. Chain alignment doesn't require work-hardening because polymers are viscous and have huge molecules. Things like melt-spinning can also cause higher chain alignment. Chain alignment also doesn't have to interact with the crystal structure of the plastic, and happens even in amorphous materials. It's just about the average direction the polymer chains are oriented.
That's insane, really, creathing a visible logo INSIDE of a filament spool, with hardware that's DIY and accessible to a hobbyist? Man, you just keep on delivering bangers, your ingenuity is amazing
This is why I love this channel, most of the other channels I used to watch are either all doing Voron builds or reviewing the same printer that's been done a hundred times already. This channel is unique!
You could think of the TPU as muscle and the PLA as bone and have them switch places. The tpu absorbing more impact on the outside and the PLA receiving less shock because of being shielded by the flexible TPU the same way that flesh protects bones most of the time. Awesome experiment!
The WInFiDEL sensor lets you measure filament diameter. Now you can implement a PID loop to automatically figure out the proper temperature at a given speed. If you want faster production, turn up the speed and the PID should correct for the change.
So many new ideas continue to pop up in 3D printing. Non-planar printing, arc overhangs, and now you are showing off multi-material filaments. Ten years from now, 3D printing is going to be absolutely off the charts.
I would love to see more videos on metamaterials! If the stiffness of the tpu-pla is close to normal pla it would truly be the king of all printing materials for parts that don't need high temperature resistance.
4:25 I remember something like this, a quick google search tells me it's called "necking". Don't remember that term but on the tensile strength curve it's the part where force stops increasing with deformation due to the section becoming smaller.
This is one of the coolest advancements in consumer 3D printing tech I've seen in a long time. What a great innovation. Even as a mechanical engineer I'm a little surprised how well you got this to work with such a simple setup, especially getting such disparate polymers into a single filament!
Every time you publish a Video it takes two days and your content is featured by all the 3D printing/ Maker Websites, this is amazing. Thanks for your R&D work and making it open to use for everyone Danke 🎉
I work at a company called Hills inc. where we do cross section logo fibers and other wild multi extrusion projects. Cross section logo fibers like that are used a decent amount in industry by basically making a dot matrix on the core of each fiber and routing the colored polymer to it's place in the matrix.
The nylon string reducing in diameter when tensile force is applied is called "necking". I got this definition from google, "In engineering and materials science, necking is a mode of tensile deformation where relatively large amounts of strain localize disproportionately in a small region of the material. The resulting prominent decrease in local cross-sectional area provides the basis for the name neck."
A PLA core with a water soluble PVA mantle could create interesting structures or grids when you print it and put it in water. Kind of like how aerogel is made, but bigger.
Unfortunately with the constant cross-section shown in the video I don't think this would give useable results. The core never bonds to itself, so if you dissolved the outer layer the entire print would just fall apart. It might work if you could vary the cross-section so the core sticks out in places, but that would be much harder to get working.
If you printed it with stripes of PVA that are diagonal edge to edge but left solid PLA layers between, the PVA might be able to dissolve out while leaving behind a structure that wouldn’t come apart.
I would drill that nozzle up to at least 1,8-1,9mm and pull the filament at above the breakpoint of the diameter/pulling force curve to achieve correct diameter and fuse the layers together properly. Also a loadcell controlled motor driver circuit would help keeping the pulling force constant.
This effect is called necking. It comes from the alignment of polymer chains. The modulus of elasticity of the material also greatly increases when necking occurs as the aligned chains creates higher crystallinity. This can also occur in multiple steps of necking down the diameter until the stress required to force the diameter down via aligning the chains further is greater than the yield stress of the filament at that point. PS: as a polymer engineering student, i love your videos.
Wow. So glad I saw this comment and went to check out your channel. As cool as the idea in this CNC kitchen video is (and no slight to Stefan, this really is a great idea), I think your pellet extruder has way more potential to revolutionize 3d printing. Amazing work.
I've recently seen your video about that and I think you deserve a lot more attention. What would you think about about combining your idea of direct printing from pellets with Stephan's filament-extruder and Thomas' thickness-sensor to regulate the speed of the extruder to match the amount of filament that the printer needs to build a printer which makes its own filament on the fly? That would solve the problems with the much heavier printhead and reduced printing-quality.
I think this idea might work, there are also ideas like implementing a real-time pressure sensor or AI camera to measure the flow and to automatically adjust it in real time. The only thing that needs to be done is the research & development which is costly and since I have spent most of my savings on R&D for the extruder itself... For this reason I keep all these things in mind but for now focus on releasing the extruder so everyone can benefit from it... 🙂
You have really outdone yourself with this video. The research you have done in this video is easily enough to write another master thesis. Outstanding!
Extrusion process engineer here. The phenomenon where you can stretch hot plastic is known as "drawdown." That's also the name of the process you are using. The simple explanation is that this phenomenon occurs due to internal tensions pulling material from the "corners" of the part to fill in the middle, where the material is splitting under tension. It helps to visualize it as a continuous landslide from the hilltops into a valley that keeps growing longer. You can also get something known as "die swell" as plastic expands due to the pressure drop at the die openings. Die swell isn't really something we have to consider with 3d printers, but it should be considered if you are extruding filament with an Archimedes type extruder. As for your filament layer bonding problems, you may be able to solve that by preheating the filament before drawing it down. Also try over-extruding your print by 3-5%, this should help layer bondability. Also, tension is everything. Pre-tensioning will help both the bondability, and help keep the filament from rotating during drawdown. The more aligned and balanced the tension is on either side of the nozzle, the more consistent the final filament turns out. Good luck, and keep up the amazing videos! "Like everything else in life, 3d printing is just a primitive, degenerate form of extruding." -Bender
@louisvaught2495 That's a great point, the production method can be improved. But the concept is still amazing. Hobbyist accessible metamaterials. That's still got me excited.
@@ouansungyidan You can work a bit to remove the molecular alignment, but in general the process being used to *make* the structure is what limits the final properties. This particular process being so accessible and macro-scale means you don't have much molecular control over what's coming out. The part of the video where Stefan discusses having to keep the filament from getting too soft is a great indication that the process is really dependent on the molecular order of the polymer, rather than the process having any control over it.
For the layer adhesion issues why not pull it twice once to get the diameter and then again at a higher speed and hotter temp just to make sure it fused together.
I dare you to print on a regular fdm 3D printer using solder. I mean it would be liquid af, but if you could cool it fast enough, this could be revolutionary.
@@richardmillhousenixon I actually tried this, and it did seem to melt, but I wasn't able to find solder with a diameter of 1.75mm, or anything close to that, and my printer had it's extruder not attached to the print head, so I never got anywhere with it.
Such a cool concept! I wonder if you could print the pulled TPU-core filament into a second 4mm blank, and then repeat the pulling process to get even smaller TPU fibers
Please do a dedicated video about the strength and toughness of some filament combinations. I am definitely interested in the mechanical properties and think that it could lead to some promising future filaments
the phenomenon is yielding and chain alignment. when it stretches to the point that polymer chains are fully aligned, it reaches a higher modulus and strength and stops being the weak point so the thicker sections continue to stretch
I have an idea. Make the filament have a clear core with transparent coloured shell, print a cool vase model on vase mode and find a way to send light thru the extruded filament. Maybe one of those 60mm LED pucks from eBay under the vase would do it. On the other hand, a black core with transparent shell could be very interesting as well.
This is pretty cool! It's basically the same process confectioners use to make candies with logos and stuff in them, although they start with blanks like 10" in diameter and stretch them down by hand. Still very cool that this technique applies to plastics, too
You are a perfect example of what happens when you live and breathe a technology day in day out. Brand new ideas well beyond the reach of most people just keep popping up.
It’s actually 3 phenomena Necking: Localized thinning and formation of a narrow region that can propagate. Cold Drawing: Stretching that aligns polymer chains and reduces diameter in a wave-like manner. Viscoelastic Behavior: Combined elastic and viscous response leading to slow deformation.
This is absolutely amazing, I'm so impressed with what you came up with, so creative and well engineered at the same time. You ended on what you would try differently after exploring the reforming method and I can't wait for the next step.
stephan, this is phenomenal, I've been thinking about a way to get around the markforged patents for a while. I consistently enjoy watching your videos, i really like how thorough you are.
Your videos are always amazingly thought-provoking. You come up with truly innovative concepts for investigation that no one (or very few) have talked about. I love it.
The technique of stretching is very cool. It is also used in micro-optics. Where you take a piece of glass, heat and stretch it, and amazingly it keeps its cross sectional shape exactly, without any deformation or rounding.
There's so many interesting possibilities mixing different materials within one extrusion. This is definitely one of those ah-ha moments for the technology. CF Nylon with TPU or ASA with TPU could make interesting case studies. Delving further in mixing multiple materials to harness more properties is another product of these types of explorations. GJ Stefan always keeping us on our toes!
A couple of weeks ago I printed a TPU/PETG filament. It worked great for the printable balls! Improved bounce compared to TPU and much more durable than PETG.
Its always a good idea to look at prior art. The resizing you do is basically the same as getting metal wire to size using a pull plate. Going in steps, e.g. 3mm -> 2.5mm -> 2.0mm -> 1.7mm, would make the proces much easier and probably quicker.
You should try putting the Winfidel sensor in the feedback loop of an automatic controller for detecting filament thickness. It could actually be accurate compared with the pulling-a-paper-clip-and-hoping method.
This is very much how candy making works. They make the pattern really (like r e a l l y) big, and then strech it so that it gets substantially thinner, and the patterns remain. Candy makers can put entire figures within their pieces, and often make things like Marios, Disney characters and such.
I recommended you do this a while back, not sure if you saw my comment or not, but Im glad some testing is being done, I dont have the equipment to do it myself.
regarding the diameter issue, you are correct in that is because you are pulling the soft filament. This "could" be mitigated by cooling it directly at the nozzle output but I would recommend changing to a push method just like the hot ends of ffd printers. Edit; I didnt realize you had a different machine in the hot end setup, good idea.
4:27 The phenomenon is based on the crystallization of thermoplastics. To summarize: when a thermoplastic is pulled apart, the polymer chains in it lie next to each other, making the thermoplastic stronger, as more intermolecular attractive forces can be formed. As a result the crystalline part of the thermoplastic isn‘t pulled apart, but the weaker non-stretched gets stretched. Have a look here: de.wikipedia.org/wiki/Kristallisation_(Polymer)
The phenomenon is called (unofficially) spaghetti straightening. Where the crosslinked/entangled polymer chains are pulled straight (aligned). This is from tensile pulling and makes the polymer longer and tougher!
This is just like a stick of rock, and how a lot of polymer clay art works. Very cool. Especially from a cottage industry researcher! The tpu/pla matrix blew my mind.
Wait, if tricolor PLA were used through one of those experimental printers with a rotational axis extruder you could setup the slicer to change the rotational position of the extruder during printing in order to change and vary which color is visible on the outside of the print. This would actually allow you to print out those FEA models in full representative color. You would however have little or no control of the color output for dithering purposes beyond the available color gradient in the filament itself. Orientation would only allow you to control the portion of the filament that's shown on the outside of the part. So filament color blends with high contrast might allow for some in-model color variation without having to do any of the filament changes that a multi-material setup would otherwise need to perform to switch colors. Purge towers or other inbetween-material purge methods would no longer be required for switching colors. The major hurdles to the above would be setting up the slicer to understand how the rotation relates to the colors available, and more complicated from the hardware/firmware side would be the need to monitor the orientation of the filament color itself as it either enters or exits the extruder. It would be a very challenging but unique project.
I work in an industry where we make plastic film. The way we get a different thickness with the same pressure from out extruders is to pull it at different speeds. If you could find a rudimentry way to force the large pieces in, you should be able to pull the final stock at higher speeds an making this more viable for larger projects.
This method of creating patterns in the cross section of filaments really reminds me of old school taffy candy shops, where they stretch the candy while preserving the pattern
I've found that abs and tpu bond very well, if you're wanting to continue experimenting with that. Something like the Phaetus 78D TPU inside some ABS or ASA might make for incredibly strong parts
The narrowing is called “necking”. And yes, the nozzle diameter IS related to the final filament diameter - it needs to be a bit bigger so it necks down to the correct diameter.
4:05 The pulling force plays indeed a major role, in real filament factories they also have a measurement device at the end of the line which adjusts the pull speed to get the filament to the correct diameter.
I think one interesting way to use this, is with multi-extrusion/dual-nozzle printers, themselves, printing a multi-layer spool as the nozzles are fixed, extruding from both, would mean the base printed spool would already have a changing pattern when printed with, but also, it should probably help with metamaterials, as you can have each material as their own perfect temp, without letting the other material cool down, both the changing position, not being in perfectly straight layers, and having both of them be hot while laid down, should greatly decrease the risk of delaminating, imho
PLA with TPU core is something I would really like to play with. The strength test didn’t really surprise me, because that’s a genius design and it should probably be producible with the right tools (as semi proven here). People would definitely pay double for a full roll like that
I love the update with an older experiment I think it would be cool to pull long strands of each material and line them up to form a pattern as a practical mass production setting. Also I think the thinning effect you talked about in the beginning is called strain elongation
Yoooooo this is amazing. So many possibilities! Can even make longer rolls by fusing multiple coils together by melting hem together. Then the resizer does it's magic to smooth out the joint
4:20 the term for the thinning of the wire is called necking. Its an extremely common phenomenon when any material undergoes plastic deformation (permanent warping) through tensile stress.
I think you should try TPU with a core of PETG for printing airless basketballs. So the TPU would get the perfect layer adhesion, while the PETG would make it bounce
Ill have to look up when I get home, but theres an open source repo for an optical filament width sensor with 3 mirrors in a half hexagon configuration because you really need at least 3 different cross sections to get an accurate filament volume because roundness matters (if you want irresponsible accuracy)
4:27 Mostly used to describe metals metals, this is called necking. The material becomes thinner as it is stretched longer, eventually leading to a total failure when the necked area breaks. The break happens at the thinnest section because the surface area becomes too small to support the tensile forces
I would love to try some of the TPU filled PLA to see how well it works in some unusual applications. It almost sounds like the ideal material for high impact prints.
Had to check out this video after talking to you at RMRRF. This is super cool, now I'm going down the rabbit hole of what it would take for us to produce something like this at scale here at Polar Filament
![ร้อยพ่อพันแม่ - WanMai [Official MV]](http://i.ytimg.com/vi/XY2pNYASEbg/mqdefault.jpg)
*QUESTION:* Which material combinations would you like to see and why?
Want to get your 3D prints to the next level already today? Check out our Heat Set Inserts and Tools at cnckitchen.store (Free shipping worldwide starting at €100).
Mix carbon fiber PETG with ninja flex TPU
ABS + chocolate 😋
Fiber glass composite pla 😋😋😋
maybe PLA with a core of water soluble support like polydissolve, maybe to do some really light weight prints, insted of solid lines you will get tubes of PLA
This process can indeed open a path to multimaterial filament such as continuous carbon/glass fiber core embedded (inside an easy to respool material at first, for test purposes), the non strech fiber mechanical properties might be a hassle to deal with though...
PA6/12, PC, TPU or Nylon for high end continuous fiber embedded filament, but might be even trickier to process.
Anyway, good luck if you try this. Great video as always !
Roughly every year or so, someone makes a completely game-changing 3D printing video showcasing an inventive new way of doing things which could open up so many new possibilities. This is one of those.
And at least a third of those seem to be done by CNC Kitchen. How many of the things Stephan has showcased over the years have become standard practice?
Yep I was wondering which 3D printer manufacturer is also watching this and thinking to add a filament-diameter scanner into the printhead to be able to automatically compensate for under/over extrusion during printing.
Maybe this is even more important for flexible filaments to get a really consistent printing results?
Its interesting, but its already being done...
Continuous Carbon Fiber filaments for high strength parts for example.
"fiber" filaments rarely have increased strength, because the fibers have to be chopped up so small that they don't do much. So Continuous Carbon Fiber filament and the like were invented to change this.
The game changer is being able to make multifilament at home, even as a proof of concept.
I work maintenance in a cable factory. The first 5 minutes of the video were pretty much most accessories of an extruder re-invented lol. You should really look up Kabmak extruders to get some inspirations because there are things like vacuum attachments and straightening rollers that could be added even to your machines. You could also add a mechanical brake to the pay-off spool to add pre-tensioning and it would make the plastic more consistent as it would feed through straight. The rollers serve the same purpose, but for wobbly lines. Another important addition would be a hot air blower to pre-heat the filament so you couldn't have to run a volcano type nozzle or anything special. You would need a lot less forces on that filament if it went into the nozzle already at say 50C. As for what you could be adding to it? Oh boy! Make 3 separate filament pay-offs with U-shaped "shells" on 2 of them and one filler. You could just add nylon strings or fishing line to the center spool and print multiple times as much filament at once. Hell, add a caterpillar-type tensioner brake to it and you're gonna have time to change filament on the go without interruption if you find a way to weld them in between.
Thank you so much. This is exactly what I wish every comment on TH-cam was like
I agree with all of these suggestions! Preheating the filament is a great idea, even with something simple as a halogen light bulb. And the roller suggestion is also great!
Preheat the filament too was first thing I thought when he was struggling to pull it through too. Can't believe he didn't think of that when he was basically already doing that to stretch down in size to insert it.
Just putting a hot air blower or halogen lamp at the entry is not very energy efficient, for a process that takes hours this is something to consider
Off topic, but I think you might be the right person to ask this, if you don't mind helping out a random internet stranger :D
I'm working on creating coaxial nylon-nichrome wires where I want to embed a nichrome wire in the core of a nylon monofilament line. The ultimate goal is to achieve this with 0.1mm D nylon and 0.01mm D nichrome wire, but I'll first try to do it with larger D nylon. What Stephan built here is roughly what I am also planning on doing, and in addition also feeding the nichrome wire through the core of the nozzle. Do you think this would be workable and do you have any tips? Thanks in advance!
Wow, this is insane. Mans maxed out his engineering level and just grinding sidequests now.
He’s completed the CNC part of his username and will soon begin the Kitchen stage.
Edit: tbh he’s already cooking
He is manufacturing a new reality for earth in his kitchen !!!
Lol
😂😂
If you think our man has reached his final form you underestimate him.
"Yo Dawg, I heard you like layer adhesion problems. So I put layers in your filament so you can have layer adhesion problems while you're having layer adhesion problems."
You are having adhesion problems with pla?
@@ericwheelhouse4371 14:28
@@ericwheelhouse4371 never under estimate a bad printer (my 2018 wanhao duplicator i3)
So basically you’re doing what candy makers have done for over 100 years :P 😂
@@stevrgrs candy makers dont make candy with two different plastics and different melting points.
Stuffed candy is generally coated or injected, as for hard candy it's rarely if ever done with a mix of, let's say butter based caramel and hard candy, as the butter caramel would leak out of the strip before it hardens, ruining the pattern.
Irony or not it's apple to oranges. he's right about temp differentials being a problem with multimaterial printing.
Hard candy is made in a similar fashion. All the design is created in a large (8-9" diameter) chunk then stretched into the final bite-size form. The pattern is retained through the stretching process same as with yours.
Same with “millefiori” glass art.
This was the first thing that came to mind. Watching how candy is made, is extremely similar and follow the same principles.
His old video about the PC core ABS he mentioned actually talks about the same thing. I believe the original researchers made large blanks with the desired pattern that they then had a machine to extrude into a filament.
Exactly what you’re thinking! But requires specialized equipment.
Assuming I’m remembering correctly! 😅
Beat me to it! I was about to say this is how they do designs in saltwater taffy too
Funnily enough, this is mentioned in the subtitles at 13:58, but he doesn't say it out loud
This TPU core experiment is just GENIUS! That's why I love your channel, you do science with your experiments, not just funny things for likes and subs
Absolutely genius!
Why a solid core instead of the TPU and PLA emulsified together? People already use silk PLA for flexibility in certain applications since it already has TPU in it.
@@zackj997 The solid core unmixed turns it into a composite. When you have two materials with different properties together, their strengths can add together in ways that wouldn't if they were thoroughly mixed. In this case, the flexibility of the TPU functions like dampening rebar, preventing fracture propagation while preserving the stiffness that PLA offers.
@@zackj997 Really? Most of my silk prints seem more brittle, not less. I'm a sucker for off-brand budget filament deals so that might be a factor.
It sounds genius if you know very little polymer science. PU and PLA are not chemically similar so they form what we call non-compatible blends. These blends are basically those that cause boundary layer separation because the two materials are not chemically bonding with each other, this forms a weak material that is worse in most aspects than the homopolymers themself. This is why when making polymers blends with dissimilar materials we use compatibilizers to act as a glue and bind both the materials together.
4:13 This phenomenon is actually a result of two combined effects: necking and strain hardening. When tension is applied to the nylon filament, it begins to undergo plastic deformation at its weaker points, leading to a reduction in diameter. At the same time, the nylon string strengthens in these areas of reduced thickness, which concentrates the strain on the thicker segments of the string. The combination of these effects causes the string to shrink to a new, constant diameter before ultimately breaking.
Articles for more information:
en.wikipedia.org/wiki/Necking_(engineering)
en.wikipedia.org/wiki/Work_hardening
aka continuous necking
It's kind of right but not quite, the effect is due to the majority of the chains in the polimer getting aligned and because all the chains are in the same orientation they become stronger
In amorfous materials like polymers you don't have strain hardening like in metallic materials
@@hugofernandez8522 Yes, you are correct, in the case of polymers this effect is the result of chain alignment and not crystal shifts like in metals, but still, this effect is called strain hardening (or work hardening) in both cases
As others have mentioned, this is not work hardening, it's chain alignment.
No, they are not the same thing.
Work hardening is a crystallographic behavior (polymers do have crystal structure) while chain alignment is just generally the idea that you can do things to the material that cause the long molecules to be oriented in roughly the same direction.
Chain alignment doesn't require work-hardening because polymers are viscous and have huge molecules. Things like melt-spinning can also cause higher chain alignment.
Chain alignment also doesn't have to interact with the crystal structure of the plastic, and happens even in amorphous materials. It's just about the average direction the polymer chains are oriented.
That's insane, really, creathing a visible logo INSIDE of a filament spool, with hardware that's DIY and accessible to a hobbyist? Man, you just keep on delivering bangers, your ingenuity is amazing
This is why I love this channel, most of the other channels I used to watch are either all doing Voron builds or reviewing the same printer that's been done a hundred times already.
This channel is unique!
You could think of the TPU as muscle and the PLA as bone and have them switch places. The tpu absorbing more impact on the outside and the PLA receiving less shock because of being shielded by the flexible TPU the same way that flesh protects bones most of the time. Awesome experiment!
The outside of the filament would also be bending more than the core, so having the outside made of tpu must be beneficial too
Embedding your logo in the filament is such a flex and extra way of tagging your prints as genuine.
I love that, watermarked 3D prints
Except that everyone can do it.
The WInFiDEL sensor lets you measure filament diameter. Now you can implement a PID loop to automatically figure out the proper temperature at a given speed. If you want faster production, turn up the speed and the PID should correct for the change.
So many new ideas continue to pop up in 3D printing. Non-planar printing, arc overhangs, and now you are showing off multi-material filaments. Ten years from now, 3D printing is going to be absolutely off the charts.
I would love to see more videos on metamaterials! If the stiffness of the tpu-pla is close to normal pla it would truly be the king of all printing materials for parts that don't need high temperature resistance.
4:25 I remember something like this, a quick google search tells me it's called "necking". Don't remember that term but on the tensile strength curve it's the part where force stops increasing with deformation due to the section becoming smaller.
Thanks
This is one of the coolest advancements in consumer 3D printing tech I've seen in a long time. What a great innovation. Even as a mechanical engineer I'm a little surprised how well you got this to work with such a simple setup, especially getting such disparate polymers into a single filament!
Every time you publish a Video it takes two days and your content is featured by all the 3D printing/ Maker Websites, this is amazing. Thanks for your R&D work and making it open to use for everyone Danke 🎉
Always pushing the limits of 3d printing, great job. The watermark filament honestly blew my mind.
Dude you are the GOAT of 3d printing experiments. Is there anyone else out there doing stuff as cool?
I work at a company called Hills inc. where we do cross section logo fibers and other wild multi extrusion projects. Cross section logo fibers like that are used a decent amount in industry by basically making a dot matrix on the core of each fiber and routing the colored polymer to it's place in the matrix.
How about filaments with an embedded cf or Kevlar strand. Is that done?
The nylon string reducing in diameter when tensile force is applied is called "necking". I got this definition from google, "In engineering and materials science, necking is a mode of tensile deformation where relatively large amounts of strain localize disproportionately in a small region of the material. The resulting prominent decrease in local cross-sectional area provides the basis for the name neck."
Can this be modeled in a bottom-up simulation?
A PLA core with a water soluble PVA mantle could create interesting structures or grids when you print it and put it in water. Kind of like how aerogel is made, but bigger.
3d printed sponge lol
Aero light plastic for rc plane wings or something
Unfortunately with the constant cross-section shown in the video I don't think this would give useable results. The core never bonds to itself, so if you dissolved the outer layer the entire print would just fall apart. It might work if you could vary the cross-section so the core sticks out in places, but that would be much harder to get working.
If you printed it with stripes of PVA that are diagonal edge to edge but left solid PLA layers between, the PVA might be able to dissolve out while leaving behind a structure that wouldn’t come apart.
Oh god i was so young back then... excellent technique! love the 4mm to 1.75 downsizing!
That TPU/PLA is so cool! what a great experiment
You are a pioneer in expanding the possibilities for 3D printer users.
A CNC-Kitchen Video 🎉 my Evening is saved.. Why my evening? Because each one of them teaches me enough to continue reading for hours… thank you!
What a crazy idea PLA + TPU core, amazing job man! This is surely a neat improvement for prints that will face a lot of physical stress
I would drill that nozzle up to at least 1,8-1,9mm and pull the filament at above the breakpoint of the diameter/pulling force curve to achieve correct diameter and fuse the layers together properly. Also a loadcell controlled motor driver circuit would help keeping the pulling force constant.
This effect is called necking. It comes from the alignment of polymer chains. The modulus of elasticity of the material also greatly increases when necking occurs as the aligned chains creates higher crystallinity. This can also occur in multiple steps of necking down the diameter until the stress required to force the diameter down via aligning the chains further is greater than the yield stress of the filament at that point.
PS: as a polymer engineering student, i love your videos.
What a coincidence that we both made a video about mixing TPU with PLA at the same time hahaha 😅
But I was a bit faster 🙃
Upvote for more awareness. Your idea to mount a corkscrew extruder on a 3D printer could be a real game changer!
Not to brag or anything
Wow. So glad I saw this comment and went to check out your channel. As cool as the idea in this CNC kitchen video is (and no slight to Stefan, this really is a great idea), I think your pellet extruder has way more potential to revolutionize 3d printing. Amazing work.
I've recently seen your video about that and I think you deserve a lot more attention.
What would you think about about combining your idea of direct printing from pellets with Stephan's filament-extruder and Thomas' thickness-sensor to regulate the speed of the extruder to match the amount of filament that the printer needs to build a printer which makes its own filament on the fly? That would solve the problems with the much heavier printhead and reduced printing-quality.
I think this idea might work, there are also ideas like implementing a real-time pressure sensor or AI camera to measure the flow and to automatically adjust it in real time. The only thing that needs to be done is the research & development which is costly and since I have spent most of my savings on R&D for the extruder itself...
For this reason I keep all these things in mind but for now focus on releasing the extruder so everyone can benefit from it... 🙂
You have really outdone yourself with this video. The research you have done in this video is easily enough to write another master thesis. Outstanding!
I think 3d printer experts should learn more from candy makers...the more i learn the more i am convinced...
Extrusion process engineer here. The phenomenon where you can stretch hot plastic is known as "drawdown." That's also the name of the process you are using. The simple explanation is that this phenomenon occurs due to internal tensions pulling material from the "corners" of the part to fill in the middle, where the material is splitting under tension. It helps to visualize it as a continuous landslide from the hilltops into a valley that keeps growing longer. You can also get something known as "die swell" as plastic expands due to the pressure drop at the die openings. Die swell isn't really something we have to consider with 3d printers, but it should be considered if you are extruding filament with an Archimedes type extruder.
As for your filament layer bonding problems, you may be able to solve that by preheating the filament before drawing it down. Also try over-extruding your print by 3-5%, this should help layer bondability. Also, tension is everything. Pre-tensioning will help both the bondability, and help keep the filament from rotating during drawdown. The more aligned and balanced the tension is on either side of the nozzle, the more consistent the final filament turns out.
Good luck, and keep up the amazing videos!
"Like everything else in life, 3d printing is just a primitive, degenerate form of extruding." -Bender
Omg he made his own metamaterial at home. Kind of freaking out right now. That's a research gold mine.
It's not particularly useful due to the high shear during production. That imparts unfavorable structure on the material.
@louisvaught2495 That's a great point, the production method can be improved. But the concept is still amazing. Hobbyist accessible metamaterials. That's still got me excited.
@@ouansungyidan You can work a bit to remove the molecular alignment, but in general the process being used to *make* the structure is what limits the final properties. This particular process being so accessible and macro-scale means you don't have much molecular control over what's coming out.
The part of the video where Stefan discusses having to keep the filament from getting too soft is a great indication that the process is really dependent on the molecular order of the polymer, rather than the process having any control over it.
For the layer adhesion issues why not pull it twice once to get the diameter and then again at a higher speed and hotter temp just to make sure it fused together.
I dare you to print on a regular fdm 3D printer using solder. I mean it would be liquid af, but if you could cool it fast enough, this could be revolutionary.
Solder melts significantly hotter than FDM filament, it's not really that feasible
@@richardmillhousenixon I actually tried this, and it did seem to melt, but I wasn't able to find solder with a diameter of 1.75mm, or anything close to that, and my printer had it's extruder not attached to the print head, so I never got anywhere with it.
Such a cool concept! I wonder if you could print the pulled TPU-core filament into a second 4mm blank, and then repeat the pulling process to get even smaller TPU fibers
Babe wake up! CNC Kitchen just uploaded
Please do a dedicated video about the strength and toughness of some filament combinations. I am definitely interested in the mechanical properties and think that it could lead to some promising future filaments
And again Stephan changed the 3D printing world again! What a legend!
Keep innovating Stefan, you're one of the main drivers helping to push open source additive manufacturing towards the future
If its ever sold that TPU/PLA combo should be named Wagyu beef!
Having a flexible core and a hard surface is what makes teeth so tough... so it could be called Wagyu Teeth! 😂
This is probably the most impactful 3d printing video of the year and and it's just chilling in our feeds.
Hexagons are the best-agons! CGP Grey Reference!!!
the phenomenon is yielding and chain alignment. when it stretches to the point that polymer chains are fully aligned, it reaches a higher modulus and strength and stops being the weak point so the thicker sections continue to stretch
I have an idea. Make the filament have a clear core with transparent coloured shell, print a cool vase model on vase mode and find a way to send light thru the extruded filament. Maybe one of those 60mm LED pucks from eBay under the vase would do it.
On the other hand, a black core with transparent shell could be very interesting as well.
Man, this is genius! The most interesting use case I've seen for the Prusa XL and this really inspires!
Imagine finding a way to embed continious carbon fiber in a filament this way!
This exists already in the high end commercial printer world.
Yeah, thats the way! And opensource it
This is pretty cool! It's basically the same process confectioners use to make candies with logos and stuff in them, although they start with blanks like 10" in diameter and stretch them down by hand. Still very cool that this technique applies to plastics, too
Now that is neat.
You are a perfect example of what happens when you live and breathe a technology day in day out. Brand new ideas well beyond the reach of most people just keep popping up.
The wilder the hair gets, the better the engineering?
It’s actually 3 phenomena
Necking: Localized thinning and formation of a narrow region that can propagate.
Cold Drawing: Stretching that aligns polymer chains and reduces diameter in a wave-like manner.
Viscoelastic Behavior: Combined elastic and viscous response leading to slow deformation.
I love the amount of innovation and experimentation in videos like this.
That twisted 2 color where the colors shift is pretty awesome, that effect with another color could be really cool
This is absolutely amazing, I'm so impressed with what you came up with, so creative and well engineered at the same time.
You ended on what you would try differently after exploring the reforming method and I can't wait for the next step.
stephan, this is phenomenal, I've been thinking about a way to get around the markforged patents for a while. I consistently enjoy watching your videos, i really like how thorough you are.
The behavior of the material stretching and thinning out before breaking is called necking
The amount of time that you put into these videos is truly astounding. Thank you so much for all of the dedication you have to 3D Printing!
Love to watch you push the envelope in a meaningful way!
Use the multicoulered matirial to print a fillament for recising and you will get an amazing fillament
Your videos are always amazingly thought-provoking. You come up with truly innovative concepts for investigation that no one (or very few) have talked about. I love it.
Abs outside, pva or any other soluble material inside could be a crazy tubing situation worth to research!
wow your english pronunciation is really well. you'r the first one english speaker I'v get to understand on high speed playback
The technique of stretching is very cool. It is also used in micro-optics. Where you take a piece of glass, heat and stretch it, and amazingly it keeps its cross sectional shape exactly, without any deformation or rounding.
There's so many interesting possibilities mixing different materials within one extrusion. This is definitely one of those ah-ha moments for the technology. CF Nylon with TPU or ASA with TPU could make interesting case studies. Delving further in mixing multiple materials to harness more properties is another product of these types of explorations. GJ Stefan always keeping us on our toes!
A couple of weeks ago I printed a TPU/PETG filament. It worked great for the printable balls! Improved bounce compared to TPU and much more durable than PETG.
Its always a good idea to look at prior art. The resizing you do is basically the same as getting metal wire to size using a pull plate. Going in steps, e.g. 3mm -> 2.5mm -> 2.0mm -> 1.7mm, would make the proces much easier and probably quicker.
You should try putting the Winfidel sensor in the feedback loop of an automatic controller for detecting filament thickness. It could actually be accurate compared with the pulling-a-paper-clip-and-hoping method.
This is very much how candy making works. They make the pattern really (like r e a l l y) big, and then strech it so that it gets substantially thinner, and the patterns remain. Candy makers can put entire figures within their pieces, and often make things like Marios, Disney characters and such.
I recommended you do this a while back, not sure if you saw my comment or not, but Im glad some testing is being done, I dont have the equipment to do it myself.
That was super interesting! I think you're onto something here with the TUP/PLA. Hope you pursue this further and provide some updates. Thanks.
regarding the diameter issue, you are correct in that is because you are pulling the soft filament. This "could" be mitigated by cooling it directly at the nozzle output but I would recommend changing to a push method just like the hot ends of ffd printers.
Edit; I didnt realize you had a different machine in the hot end setup, good idea.
4:27 The phenomenon is based on the crystallization of thermoplastics. To summarize: when a thermoplastic is pulled apart, the polymer chains in it lie next to each other, making the thermoplastic stronger, as more intermolecular attractive forces can be formed. As a result
the crystalline part of the thermoplastic isn‘t pulled apart, but the weaker non-stretched gets stretched. Have a look here: de.wikipedia.org/wiki/Kristallisation_(Polymer)
The phenomenon is called (unofficially) spaghetti straightening. Where the crosslinked/entangled polymer chains are pulled straight (aligned). This is from tensile pulling and makes the polymer longer and tougher!
This is just like a stick of rock, and how a lot of polymer clay art works. Very cool. Especially from a cottage industry researcher!
The tpu/pla matrix blew my mind.
Wait, if tricolor PLA were used through one of those experimental printers with a rotational axis extruder you could setup the slicer to change the rotational position of the extruder during printing in order to change and vary which color is visible on the outside of the print. This would actually allow you to print out those FEA models in full representative color. You would however have little or no control of the color output for dithering purposes beyond the available color gradient in the filament itself. Orientation would only allow you to control the portion of the filament that's shown on the outside of the part.
So filament color blends with high contrast might allow for some in-model color variation without having to do any of the filament changes that a multi-material setup would otherwise need to perform to switch colors. Purge towers or other inbetween-material purge methods would no longer be required for switching colors.
The major hurdles to the above would be setting up the slicer to understand how the rotation relates to the colors available, and more complicated from the hardware/firmware side would be the need to monitor the orientation of the filament color itself as it either enters or exits the extruder. It would be a very challenging but unique project.
I work in an industry where we make plastic film. The way we get a different thickness with the same pressure from out extruders is to pull it at different speeds. If you could find a rudimentry way to force the large pieces in, you should be able to pull the final stock at higher speeds an making this more viable for larger projects.
This method of creating patterns in the cross section of filaments really reminds me of old school taffy candy shops, where they stretch the candy while preserving the pattern
I've found that abs and tpu bond very well, if you're wanting to continue experimenting with that. Something like the Phaetus 78D TPU inside some ABS or ASA might make for incredibly strong parts
The narrowing is called “necking”. And yes, the nozzle diameter IS related to the final filament diameter - it needs to be a bit bigger so it necks down to the correct diameter.
so cool. next step is using the klipper thing which adjusts flow depending on live pfilament diameter
4:05 The pulling force plays indeed a major role, in real filament factories they also have a measurement device at the end of the line which adjusts the pull speed to get the filament to the correct diameter.
Tried to reach this clip from YT main page but got strange information page about ads several times. Searched for CNC kitchen and now it works. FYI
I think one interesting way to use this, is with multi-extrusion/dual-nozzle printers, themselves, printing a multi-layer spool
as the nozzles are fixed, extruding from both, would mean the base printed spool would already have a changing pattern when printed with, but also, it should probably help with metamaterials, as you can have each material as their own perfect temp, without letting the other material cool down, both the changing position, not being in perfectly straight layers, and having both of them be hot while laid down, should greatly decrease the risk of delaminating, imho
PLA with TPU core is something I would really like to play with. The strength test didn’t really surprise me, because that’s a genius design and it should probably be producible with the right tools (as semi proven here).
People would definitely pay double for a full roll like that
I love the update with an older experiment I think it would be cool to pull long strands of each material and line them up to form a pattern as a practical mass production setting.
Also I think the thinning effect you talked about in the beginning is called strain elongation
Yoooooo this is amazing. So many possibilities!
Can even make longer rolls by fusing multiple coils together by melting hem together. Then the resizer does it's magic to smooth out the joint
4:20 the term for the thinning of the wire is called necking. Its an extremely common phenomenon when any material undergoes plastic deformation (permanent warping) through tensile stress.
I think you should try TPU with a core of PETG for printing airless basketballs. So the TPU would get the perfect layer adhesion, while the PETG would make it bounce
TPU is a game changer material, it can have 0.1% shrinkage, HDT 100deg and hardness almost 90D.
I believe the thinning phenomena that occurs when pulling polymers like TPU is called "necking"
I would love to see more experiments with multi-material filament, it seems to have a lot of potential.
Ill have to look up when I get home, but theres an open source repo for an optical filament width sensor with 3 mirrors in a half hexagon configuration because you really need at least 3 different cross sections to get an accurate filament volume because roundness matters (if you want irresponsible accuracy)
4:27
Mostly used to describe metals metals, this is called necking. The material becomes thinner as it is stretched longer, eventually leading to a total failure when the necked area breaks. The break happens at the thinnest section because the surface area becomes too small to support the tensile forces
Doing material research at home with cheap equipment and just an insane amount of smartness and dedication? Very cool.
I would love to try some of the TPU filled PLA to see how well it works in some unusual applications. It almost sounds like the ideal material for high impact prints.
This is quite possibly the coolest video you have ever made, and that's saying a lot! Wow!
Had to check out this video after talking to you at RMRRF. This is super cool, now I'm going down the rabbit hole of what it would take for us to produce something like this at scale here at Polar Filament
Composite filaments could be a game changer for engineering applications.