I worked on almost exactly this concept 10 years ago. It's kind of interesting to see you guys go to the same problems i went through. I'll give you guys some freebie tips. Use a liquid hydrophobic mask to increase your resolution. Get rid of the sulfate electrolyte (you do not want sulfur in your deposit material). You want an electrolyte that doesn't have side chain reactions. I can tell you what I used, but it'll be more valuable if you figure it out for yourself. You can get rid of the dendrites and greatly speed up your process by physically accelerating your electrolyte. Brownian motion in the electrolyte is what forms the dendrites. Add a focused magnetic field. You have a magneto hydrodynamic cell but you're only controlling half of it. And lastly... The photoelectric effect is your friend... I'm not saying more unless I can be involved in the research.
In essence, it's simply a matter of putting together a more robust design, and then releasing it as an open source sort of affair for the sake of facilitating technological progress.
Defo. I have a student working on a more robust design where the aim is to put this into an open access repository for people to improve on it but we want to make it less buggy first :)
I've been imagining a similar setup but using the equivalent of a felt pen to convey the electrolyte. I'm super excited to see someone pursuing an electroplating-style metal printer. I figure it's potentially the most economical and simplest way to bring metal 3d printing to the masses.
Great to see the interest and hope it sparks further ideas. A felt tip could also work though you need to be careful about getting the deposition in the felt itself which is why we avoided this approach but I'm sure there are ways around it
@@BillyWu I wonder if the idea would work better upside down since you could counteract gravity if leakage is an issue. Also, since the resistance should change as the distance increases, I wonder if you could make the printer more precise by increasing or decreasing time spent in one area based on resistance readings.
@@ryelor123 Good comments. This current system works by having the electrolyte in the syringe with the sponge. Flipping it in this form would cause leakages. On the feedback mechanism, this is certainly possible and we had done some work on it and am sure this will be the way forward for smarter higher quality printing.
One way I had envisioned getting a metal part from a 3D printer was to print a porous conductive plastic model and then do electroplating and electric polishing. At some point we either use a solvent or heat to get rid of the plastic. And then more electroplating and electropolishing. As slow as this is it's probably still faster than my idea.
Certainly possible to do this and since the deposition area would be faster in some sort of electroforming process it would probably be faster. One thing to consider is that some conductive paints etc aren't 100% conductive when zoomed into the micro-scale since they're often a blend of conductive agents and binders. Some people also use electroless plating to get a more uniformly conductive layer
In Microelectronics and Microsystems we use electrostatic based 3D printing for prototyping (dacades old but still SOA) to move one atome at a time. Even for that industry, this method is very expensive and is used only in rare cases. This here is really a nice new development and is very interesting
This is incredible work here! As soon as I clicked on the video I thought “why didn’t I think of that?”. There are a few nuances I can think of, like whether the solution (be it the solution in the syringe or the tiny bit of solution touching the workpiece) changes in local concentration, or whether you need to introduce leaveners or other additives to the solution, or the inability to deposit some metals atop other metals. But I think all could be worked out. Also IIRC current is what matters more than voltage with electroplating, I’d be interested to see what constant gives you more consistent results. But what I’m most excited about is printing structural metals, and even alloys. Copper and nickel are one thing, but being able to deposit some sort of tough iron alloy would be incredible. I’ve seen a few papers on electroplating alloys before, I think in this case you’d just need to have both metals in the solution at once, and have two electrodes in the one syringe. Then the alloy ratio is determined by the ratio between the currents. Smart CAM software could even change the alloy composition dynamically as you print different parts, which could lead to an incredibly versatile process. A shame you can’t print titanium or aluminium (carbon too I guess), but I think from zinc all the way to platinum, there should be some reasonably strong alloys to be made. Perhaps you could make really compact thermoelectric coolers/generators, or even try depositing doped semiconductors. What I like most about this is how easy it would be to build from an existing FDM printer. I suspect this will be a trending topic among enthusiasts in 2-8 months. All it takes is someone with enough time and money to try and make this practical enough for hobbyist use. Are there any glaring flaws, big hurdles? Does the meniscus wick away down long thin cavities or drip away down the side of a tall print? Could it be better to print the whole thing submerged in the solution?
Definitely lots of ways to extend this. Electroplating of alloys is certainly possible and you can control the composition by changing the voltage. There was some recent work where they did this to make a thermocouple. You can also do the deposition in an electrolyte bath with a fine tipped electrode but the resolution isn't as good. Hopefully provide lots of ideas for others to build on. We're hopefully releasing some open source plans soon. Version 1 worked but was a little buggy lol
If a fine tipped electrode on its own had poor resolution, perhaps surrounding it with a ring at an opposite potential for shielding purposes might improve things? Of course you’d need to prevent the electroplating metal itself from depositing on it, but I think there should be metals with those properties, just as it’s impossible to plate iron onto copper. I’m still most concerned with how a system would handle tall structures and overhangs.
@@Scrogan if we used an insulated tip as you say, you could probably grow overhangs directly out wards (within a certain aspect ratio). Mabye even printing them bit by bit with every layer that is added (a kind on non planar printing to optimise giving support to overhangs to print them easier)
@@Scrogan and the closer we have the tip, the higher the current concentration directly bellow the tip, and the more reduced it is to the surrounding metal.
@@Scrogan You might even be able to get away with using an enamel coated copper wire. The enamel insulates it, and only the tip is conductive and used. As the wire is eaten away, the enamel should hopefully just flake off. Which admitedley might cause its own problems
@@weirdsciencetv4999 Thanks. Unfortunately, electrodeposition processes are a little bit slow so making larger items can be challenge but we're looking into ways of accelerating this. Certainty right now you can make very small functional items which still has a purpose
@@weirdsciencetv4999 Yeah, there's a few patents which use an array of electrodes which you switch on/off in a bath of electrolyte such that you print a complete 2D layer at once. Can be faster but features not as well defined and needs more expensive set-up. Companies such as Fabric8Labs are doing this www.eenewseurope.com/news/no-heat-no-laser-no-powder-reinventing-3d-metal-printing/page/0/1
Sounds like the main limitation is speed, but maybe also resolution, especially once the part is actually becoming "3d". And it requires a conductive build plate/base. Maybe a non-metal build plate allows separation of the part and reuse of the plate?
Yeah, stability of the liquid meniscus is also a challenge. Some people have demonstrated really good resolution with the use of nano-pipettes however at the cost of speed. The challenge with a non-metal base plate is that we need electrical contact for the deposition but you can deposit on glass if you have a conductive path from elsewhere.
Interesting with this set up you should be able to print (very simple) electronics unto surface. You could be able to print simple logic boards and even rfid
Did you think of using sponges brushes ? You syringe tip approach look indeed best for precision but since the process is slow, you could save a lot of time with a wider surface. The same way you have multi material the printer could use a second, wide, brush able to cover more area. Also maybe a truncated cone could be squished to have some control over the width of the contact patch.
Yeah, there's been some work with nano-pipettes to make very small structures. The challenge then is the volume deposition speed drops so in general there's a trade-off with speed and resolution
Very interesting discussion of electrochemical deposition process. As you mentioned in the video a big limitation is deposition speed because of dendrite formation at higher voltages. I can see parallels with traditional ink jet printer's which could address speed issues. A 2D matrix head, with individual channel deposition control would be possible. Limiting constraints would require high flatness, separation spacing and accuracy of the print head and deposition bed. I would be interested to know what the experimental deposition rate you've been able to achieve for copper.
Thanks. In previous work we demonstrated a volumetric deposition rate of about 20,000 um3/s with a 400 um tip. Sounds fast but is actually only ~0.0012 mm3/min lol. Could speed up with more active area and a few other tweaks but speed is a challenge with this approach but certainly a few applications where that's less of an issue.
Wow, that was awesome! Really good spoken and clearly understandable Video. This technology is like the reprap start. This will be going to revolutionize the 3d metal thinking 🤔. Top video extreme big thx for sharing! ♥️♥️♥️ Are you going to make something open source?
Should there be a small o-ring or seal around the sponge to isolate the electrolyte? Another thing I wonder about is if you could make this process more precise by adding it to a FDM printer and having the printer make a sort of shell of plastic around the metal to give precision?
Good questions. We tried to keep the system as simple as possible and found that with just the right compression and type of sponge that the back-pressure could be balanced with the electrolyte pressure to prevent leakage, however this was a fine balance that sometimes didn't work so improved electrolyte confinement would help. If you make an electrolyte guide as suggested that might help, though we wanted to see if we can achieve this without additional complexity, though perhaps supports might be needed
Containment of the electrolyte is moving in the wrong direction. You want to deplete or inactivate the electrolyte after consumption... Then it is self masking and regenerable.
This is a brilliant idea, kudos! Looks absolutely fascinating. Two questions immediately come to mind: 1. Would precise control over electrode distance and a higher current efficiency electrolyte (nitrates perhaps?) allow you to operate directly in a bath rather than your syringe? McGeough's (U. Edinburgh) comments on the relationship between efficiency and resolution for electrochemical machining come to mind. 2. Would you be able to take advantage of the double-layer through the use of pulsed current, akin to Pulse-ECM? This may be able to improve resolution and perhaps allow for direct in-bath printing as well.
Thanks for the very interesting comments. There have been a few other works which use fine tipped electrodes to do the electrochemical printing but the deposition tends to be a bit more spread out. You can address this somewhat from having a finer electrode which is closer to the substrate but we wanted to try to use lower cost components and avoid costly position controllers. On pulsing, yes this should be possible and have seen some work where people do pulsed deposition to get favourable morphologies. Lots of potential variables to optimise further so am always happy when this might inspire someone to give it a try with a new approach.
Electrochemical method for 3d printing is too slow for now. You could do an analysis for femtosecond laser technology for combining electrical conductivity of femtosecond laser with electrochemical deposition.
Thanks. We tried to capture these types of deposition techniques under the "directed energy deposition" category but agreed that there are new forms of printing constantly coming out so always great to have a discussion around the newest approaches. The pdf of the overview in the video can be found at if interested am-hub.dk/wp-content/uploads/2018/01/IMSE-Briefing-paper-2-AM.pdf
@@iro3d There's a few companies such as Fabric8Labs which seem to have a reasonably mature product. They use an array type approach so might be a bit faster but with challenges around resolution. I suspect prototypes of this form are reasonably mature but the challenge is whether there would be sufficient consumer interest for a printer like that with the known challenges of scale fabric8labs.com/
Do it! I so need a metal 3d printer I can afford! I would be thrilled with an aluminum printer for maybe a few thousand. Or even a nylon SLS printer for that.
Is there a limit to what materials you can use... obviously you can use copper, nickel, and chromium but for example metals like steel, aluminum, titanium, and a multitude of other metals, basically what im asking is are there limits to which materials you can use, also you could probably make an alloyed rod to going into the syringe to get alloyed printing? idk just think here
Great question. In principle any metal you can electrodeposit you can use this approach, however one thing to bear in mind is that we designed this to use relatively low cost and accessible components. Some electrodeposition processes use more aggressive electrolytes and temperatures to get a good deposition and therefore would need modification to make safe. In some cases you should be able to control the alloy composition by varying the deposition potential and control.
Thanks for the interest in the project. I had a couple of student projects to open source the printer but it's not quite ready for release yet but hopefully soon.
Amazing approach! If the speed limitation could be overcome it could spark almost another industrial revolution, or at least help bring manufacturing back to EU and USA. I was wondering if localized electrochemical deposition could be used to join 2 wires (e.g. welding wire) of the same metal together, essentially speed up the process significantly, for the cost of some precision?
In our case, the anode which was a copper rod was periodically replaced manually. We were mostly printing very small items so the mass of the copper rod was far in excess of the printed item so we didn't have any problems with running out of material, but its a simple case of connecting a new rod and putting it in solution.
I was just wondering if this concept would work! Very cool, and well explained, too. Is it possible to print in alloys, or does the plating process limit the user to elemental metals?
Thanks and great question. Yes, it's possible to print alloys. In principle anything you can electroplate, you can print with however in the case of some metals the electrolyte and conditions start to become a bit more challenging to handle. We've mostly focused of aqueous systems (water based) but lots of non aqueous systems exist for more challenging materials.
Normally we just need to remove the part from the substrate. In our case we printed on aluminium foil and then were able to just manually peel off. If really stuck in theory you can just dissolve the aluminium away
Hi, I love it. I was big in Electroplating of high aspect ratio structures back in my research days. Used to use Megaherz ultrasonics to break up the boundary boundary layer. But I was wondering if you are controling your deposition spread just via the size of the meniscus or optimizing your frequency and PWM too. So you are basically using the the charging of your boundary layer to focus your mass transport to the close areas?
Thanks Jens. Yeah there's a few approaches for enhancing the mass transport properties such as electrolyte flow systems and as you say ultrasonic systems. We control the size of the deposition mostly through the size of the nozzle, it's separation from the build plate and the deposition conditions mostly but are also looking at other approaches too which we hope to share soon :)
Good question. We did some indentation tests and generally for the copper this was harder/stiffer than bulk cast copper which we attributed to the fine grain structure when printing with electrodeposition, where smaller grains generally give harder/stiffer materials but are more brittle
Or the syringe could implant ions into a permeable dielectric substrate; coated epoxy-fiberglass? Washing the substrate in a chemical bath would initiate 'electro-less' metal plating on the ions. Plated-thru vias in PCB's are made this way. Printing the pattern would be fast; the time-consuming plate-up process is done in batch mode; thousand of parts in the same bath. What do you think?
Yeah potentially printing a pattern first and then increasing the thickness could work. Definitely lots of different ways to take this forward so more people working in this would be great
This is extremely interesting. For the sake of complexity reduction could this not work with the part and nozzle submerged instead of maintaining a meniscus? I am assuming the current would still be constrained by the nozzle wall and predominantly favor the the shortest path. With a bath you'd be able to circulate electrolyte through the nozzle though I'm not sure if that is actually useful. Apologies if you addressed this in your first paper. I don't have access to that journal.
Hi Alexander. You're right that this approach would also work with a bath of electrolyte. Others have done this with sharp electrodes to help focus the current. Having a nozzle in the bath may help to also focus the current however the challenge would the be the difficulty in doing multiple materials, where a bath change would be needed. Definitely worth looking at though.
How are the dendrites negatively effecting the metal? Do they also cause poor adhesion? Just wondering if dendrites are acceptable trade off for example some applications.
Hi. As others have mentioned dendrites generally result in a mossy and porous structure which has poor mechanical properties, so ideally we avoid this by going slower.
@@BillyWu Thank you for your answer. Poor mechanical poperties is kinda of subjective term, though. Since it really is depends on the application you be using for. If it's really porous it might be good for heat shielding or something. Really cool technique since are also kind of able to also determine the density of the metal through the voltage. Or is that little bit to far fetched.
Could you submerge the entire build palte in the copper sulfate soloution and jsut use an electrode that very very localy applies the voltage and builds up the part?
Definitely, that would work too, however the features would be less well defined, though some use very fine electrodes. The meniscus means that you only get deposition where there is liquid.
@@BillyWu But what do you do with the menuscus once you have printed one layer, ir even want to put down a line next to the current one? It will start to pool, leading to a reduction of accuracy. Could you also point me to a paper or two on submerged electroplate 3d printing I could read? I am very interested in this process. Thanks
Good question. Close control of the meniscus is key. This article talks about the work from a team ETH Zurich who show the meniscus in more complex 3D prints www.azom.com/news.aspx?newsID=57437. This review paper talk about localised electrodeposition link.springer.com/article/10.1007/s00170-020-05799-5
@@BillyWu Have you read this paper, very intesting. That high plating rates (50um/s for a 0.5mm nozzle) could enable macro scale 3d printing. I just put it into my slicer, and you could 3d print a 20x10x5mm object in under 24hrs
Yup. We modified a standard FDM printer to achieve this by replacing the print head with a syringe holder and filling this with an electrolyte and a rod of copper/nickel. You'll also need a power supply to enable the deposition.
Is it possible to put multiple electrodes with different metals in a single syringe and then apply a voltage to the electrode with the metal you want? Would it be possible to make alloys like this by using multiple electrodes at once with different levels of voltage?
There's a few other studies that have done co-electrodeposition which by controlling the deposition voltage you can control the composition. For example in this paper pubs.acs.org/doi/10.1021/acsami.0c01100 they do a nickel-copper deposition with controllable composition which is quite nice for making sensors. The challenge is making sure a constant amount of metal ions stay in solution
We haven't done gold electroplating yet but I think it's slightly faster than nickel, however it does normally use a cyanide based electrolyte so we avoid this where possible.
Yeah. There are 2 main approaches for electrochemical additive manufacturing. What we showed here is the meniscus confinded approach where you have a bead of electrolyte to focus the deposition. You can also have a bath of electrolyte and use an array of electrodes. This needs more electrolyte and th resolution of the features can potentially be less well defined, but it can be done.
A very interesting concept, thank you for sharing! However, am I correct in assuming that this can not print "islands", structures which are not supported by already printed material underneath? Conventional 3D printing solves this by printing supports underneath, however I assume they might adhere to much when using metals. To me, it seems like that's where printing in a powder bed or gel matrix excels. No supports needed and even "floating", unconnected parts won't fall of until they're fused.
Yeah, to deposit material we need electrical contact. You can create some "island" structures by printing in a different material and then selectively dissolving the support material in a similar way to how water soluble supports work in FDM printing. True that powder bed fusion for polymers can make these freestanding structures. Much more difficult with metals though, due to higher thermal stresses. Just messes up the print bed.
Certainly. If you can electrodeposit a material it should also work in a localised deposition printing approach. Some people use ionic liquids to do deposition of metals so more exotic materials than copper are possible.
Good question. We've mostly focused on pure metals such as copper and nickel. There are some alloys that you can electroplate, however since steel is an alloy of iron and carbon, it could be a bit more tricky since this is normally manufactured at high temperatures through mixing and solidification.
There are a few fundemental problems with this. First, electroplating is slow, and I dont know of ang ways you can speed up the process by several orders of magnitude. At the voltages you describe, it would take weeks to complete a decent sized print. Second, strength. The faster you grow your structure, the weaker it will be. Electroplated structures are more similar to wet sponge than anything else
Fair points. Speed is still an issue with this approach but there are some niche applications where the scales are much smaller and multi-material might be useful, but still at the R&D stage
@@BillyWu multi-material work is limited by the chemical reactivity of the plated element and electrolyte. You really cant do anything with it. You cant use alloys, and the whole thing will corrode because you have two different metals in close contact. This is a step backward 1000 years metalergy wise
Thanks for presenting your progress. I wonder if a wax layer or oil would keep the dendrites forming as an interference compound over the base material. Something the nozzle could move thru but still make a deposit later.
Sure. You can find more details about the set-up in our 2 papers www.researchgate.net/publication/319315170_A_Low_Cost_Desktop_Electrochemical_Metal_3D_Printer www.nature.com/articles/s41598-019-40774-5 We're hopefully going to be releasing an open source version of this soon once we iron out some of the initial bugs.
There's a few works which use an array of electrodes which you turn on and off which can accelerate the process but generally this has to be immersed in a bath of electrolyte. There's also a few other things you can do to increase the deposition speed such as higher temperatures and different electrolytes. We just used a low cost solution as copper sulphate is quite cheap. Some root killers use copper sulphate
Thanks for the interest Stefan. From the tests we previously did, we can see that the hardness of the material is generally higher than cast copper due to the small grains that we form during deposition. We haven't compared this with direct laser melted copper, but would imagine that its favourable since we have low defects, though our sample size remains relatively small.
Do copper and nickel deposit at the same voltage? Given two electrolytes don’t cause some undesired reaction, could you use a single syringe and apply alternating voltages to perform multi-material deposition in one go? (Not a chemist). Follow up: would aluminum sulphate work as an electrolyte? Could you print aluminum parts?
Voltage is a relative measure between 2 different points. Since we use copper/nickel as the anode and cathode the voltage we apply to get the deposition is similar but generally nickel is a bit slower without any additives. It is possible to do a co-deposition if you get the mixtures right and you can electrodeposit alloys in this way but its a bit more tricky. Aluminium can also be electrodeposited, but often required more challenging electrolytes and we wanted to stick with water based solutions. For example, we stayed away from gold since its more expensive for prototyping but also many deposition solutions use cyanide containing components. Essentially, nearly all materials you can electrodeposit with you can print with.
@@BillyWu thanks for answering my questions. Really novel approach. What are your + your team’s current thoughts on ways to speed up deposition? I understand that y’all were able to increase deposition speeds of Cu by 34% and Ni by 85% by replacing the sponge with a nanofibre.
@@davidlanday6102 There's a few potential avenues. Certainty moving to other non-aqueous systems can potentially accelerate this but will be a bit more tricky to handle the electrolyte. Additives can help and also print head designs that minimise the mass transport problems I think are ways forward.
Yeah, you're right. This approach of printing metals with electrodeposition is a bit slow for large parts. There are a few tricks to speed it up but at the moment the main applications are likely to be smaller items, where the ability to put down metals means you can start to build conductive pathways
Great question. The components themselves aren't very costly and mirror that of standard plastic 3D printers with the stepper motors, frame and control electronics. The different components include the pipette, electrolyte and power supplies which are all generally low cost. The challenge is that speed is still quite limited in terms of volumetric deposition rate so still needs some work.
I am looking for a technology solution for aerospace for printing fuselage and a wing in one piece. Can you see It's doable for a bigger printable object like an airplane? Thank you in advance.
Unfortunately, this approach isn't really suitable for large scale structures. I know there are some "big area additive manufacturing" (BAAM) approaches such as www.e-ci.com/baam but these mostly print with plastics.
It's a bit tricky as it needs to be printed on something conductive to have access to a flow of electrons, but what some people have done is start from a point and then print a circuit from there, so in theory you could do that
what about using ordinary cheap copper wire at the extruder with a extruder made of plastic or something nonconductive with a flat tip of the wire dipped in a hydrochloric solution with a thin metal plate at the bottom. that should allow the copper to dissolve rapidly and build up on the plate. can also increase the voltage to that of the amps to accelerate the electrons in the solution. with about less than a milimetre of gap distance chould reduce random ion floating outside the printing alignment. the rate of print is only limited to that of conductivity in the solution. printing speed is something that can be perfected. you still need the the feeder mechanism of the copper spool to keep the gap between the wire tip and the plate fixed as the plate or wire moves.
Yeah, we considered using a wire feed as well but wanted to try and keep it as simple as possible, but yeah there are lots of alternative ways to do this. Would be great to see what other designs people can make
Doesn't work well. Already tried it. It does work up to a point, but past a certain speed the water starts to boil and it destroys the print quality. To go faster would involve non-aqueous electrolytes and high temperatures... Which are decidedly unfriendly to work with and come with all kinds of surprises like thermal warping and cracking. Basically, wire is a no go. You'd be better off by modifying a traditional welding process... Which isn't as hard as it sounds. Miniature welding is totally possible and at reasonable speeds and temperatures.
What's the largest object printed with this technique and what time did it take to print? Edit: I see from other comments that it would be really small
Yeah, unfortunately it is quite small in terms of volume. It's relatively each to make mm-scale objects but with limited depth. We demonstrated some very thin lettering in the work that was several mm. Could potentially be used to make circuit tracks... slowly, or sensors
But how are you going to get liquid stainless steel and liquid aluminum? Makers, Designers, Builders and Inventors aren't clamoring to to print copper parts. I suppose there's a possible chromium nickel solution that might be somewhat viable... but maybe not as a functional part itself. What I picture as a possible low budget solution, is suspended microfine powder (agitated upon nozzle dispersion) in a UV curable medium. Much in the same way that flat bed UV printers can print white ink (titanium dioxide) through piezo electric discharge. Yeah, you are back to the binder removal process, then the furnace/kiln firing in an inert chamber thing... but honestly, that's a device that could be cheap if produced at mass scale. We aren't talking about much complexity beyond a PLC, heat bricks and heating elements. The door just needs an airtight seal, and a gas valve solenoid. It's literally not beyond the skillset of a competent fabricator. (So, somewhat DIY to a degree) I think we need to explore industrial print heads, and their maximum viscosities they can push, and what mesh powder can we squeeze through them before clogging is an issue. I mean we seem to be quite OK with a 0.4mm nozzle hols for FDM printing, provided the z-step is minimal. I'm willing to bet that quite a bit of suspended metal powder could squeeze through a pressurized print head with nozzles that large. The UV curing could be localized too... utilizing a fiber optic approach, you could get the light right down by the nozzle with a light pipe, and not reflected anywhere else, like back up the nozzle hole.
Nice suggestions. There are also metal jetting systems which are starting to emerge for larger scale metal printing but you're right this approach has some limitations
Unfortunately this approach is a quite slow for large objects. In our previous paper we showed that for a 400 μm nozzle, we can achieve a deposition rate of 20,000 μm3/s. So for a 10 mm cube with a 10% infill that would take approximately 58 days to complete. There are a few things you can do to accelerate this, but this is the reason why this is still mostly at the R&D stage. However, there are applications for smaller scale structures which are electrically conductive.
@@BillyWu This is very good for Circuits and creating own IC's and Transistors. Possibilities are endless and this opens field in new direction. Imagine 200 or 300 printing same model at same time like an transistor or an IC which have new capabilities which we had never thought of!! The production can easily be achieved and there will be low cost of creating dust free labs for such printers and robots or automated arms just change the build plate and stack them for someone to collect the pile at the end of hours or days!! Its a breakthrough in nano technology. The projection of this printer wrong, it should be projected as Nano Metal 3D Printer. Keep it up and one day it will be high in Demand like a 48MP camera sensor in Smartphones! :)
@@himanilsharma2147 Thanks. There's been a few examples of applications of this approach in electronics, so you're right this is likely to be the initial application area if it takes off.
@@BillyWu You know I personally believe 3D printing is next Industrial revolution like lathe machines did and current technologies of 3D printing are primitive and we will see more advanced techniques and technologies in coming future. Where everything will be automated like we see as VMC, if someone in 1920 or 30 had told about VMC to someone then People had laughed, same like Miracast and Airplay like technologies which enable to do video chat on TV!! and this technique which you demonstrated in video is Nokia 3310 which enabled anyone to play game in Cellphone first time in the human World, that too Snake game in almost a device having pixel bigger than 1mm, and within just 20 years children are playing PubG like game in their own smartphone worrying about latency!! Its a huge breakthrough and I personally believe we will see that soon very soon just within next 10-15 years.
@@BillyWu I'm not know scientist or physicist I don't even have a GED or high school diploma I have dyslexia I'm one of these people that had no opportunity when I was a kid and could have done great things in my life if I had standard opportunities like most people in the world but because I was abused since I was a kid I can't do stuff to change the world but I don't care I will tell people things I I know that will change the world
Thank you. Very interesting information. Now, if I could get a useful metal printer for about $10,000, without needing a machine to wash the parts out, and not gave to sinter, I’d be very happy. Butblow cost here no doubt means at least several times that.
Uhm, basically a facility employing this type of electrochemical additive machine in industrial scale is going to produce the same type of waste as an electroplating facility. Therefore, it's going to be subject to similar environmental and waste management regulations as such. I'm working in the electroplating industry and I can tell you de-ionizing waste water processing systems are extremely expensive to buy and to maintain. Oh and in most countries you need to apply for a permit to run a facility that involves heavy metal ion solutions, and that can also be hugely expensive depending on where you are located. I guess in small scale for DIYers or small offices where they'd just dump the waste water into the drain would be fine. But if somehow it got extremely popular and start having an effect on the environment, it could ended up being banned by the government.
There's a few patents around in this area with different variations of electrochemical additive manufacturing and a few companies are trying to take this to market such as Fabric8Labs
Compared to FDM? Yes, but it doesn't have to be super slow. The whole process is only limited by heat. Heat changes the chemistry, so there's an upper limit and a lower limit. The basic problem is how fast can you dump the heat from the reaction into the environment without changing the quality of your results.
Yes, very fair point. Electrodeposition is quite a slow process. There are ways to increase the speed with different electrolytes and designs but will likely always be a challenge, however there are some applications where this might be suitable given the ability to do multi-material. Still R&D at this stage but it's good fun research
You could have an array of electrodes, and thus deposit material at many places at once. 10 electrodes would reduce print-time ~tenfold ... and so forth.
Skip to 4:21, where the video actually starts. This is not low cost, its unbelievably slow, lithographic masking techniques followed by various deposition or erosion options are faster, cheaper and more precise. DLP activation of resists is a good area to develop a micro 3d print version of this. Or maybe a way to make conductive FDM filament that you can electroplate. I see no viable path for the method shown in the video to compete with lithographic, DMLS or pla / liquid binder then oven centered 3d metal printing (whatever that's called). I'm not saying it was a bad research project, but in science sometimes things fail, this is a thing that while technically feasible is not economically feasible nor do I see any way for it to become so. You have a layer height of like 6nm and a linear speed of 0.4 mm/s this is ridiculous. And to call this low cost... I can take some sharpie ink and draw on a substrate then dunk it in an electro plate bath for like $10 with scrap anyone has lying around. And then to brag about it being 4d printing because you can get some mems behavior.... OK I get it you did not coin the term but it really needs to stop. It's a retarded thing to call a 3d printed object that can be manipulated. You could call tweezers a 4d object by that definition, it's just bad and anyone who uses the term should feel bad. RUclips comment section, where the real peer review happens.
Some fair comments. It is relatively slow and no doubt not the solution for everything but the approach has certain merits which warrant further research in terms of increasing the speed and multi-material aspects. Welcome comments from all asides positive and negative.
@zyxwvutsrqponmlkh you lack imagination electrochemical printing opens the door to modular self replicating devices (which can be exponential therefore solve the speed problem ) which is lot harder to achieve with light or thermal based systems. in here all you need is electricity insulated capillary(that can be printed) and an ionic liquid. It opens the door to manipulate electromechanical properties without post processing by selectively depositing metals with different propreties perhaps turning them into insulators or electro deposit ceramics such as calcium carbonate . I see the future of this from a bottom up perspective rather than a top down approach . As for 4d printing I see this as one of the most promising tool to achieve some types of 4d printing for the reasons stated above unless you want to narrow the definition of 4d printing to some passive environmentally triggered assembly a la skylar tibbit .I have used selective electrochemical depositing (not a 3d printer a 2.5d) to print programmable joule heating activated trusses that assembles into 3d structures. we even printed bimetal "memristors" or switches on them to program how they assemble.
Lookup the Harvard milidelta . A variant of it can be entirely be electrochemicaly printed subtracted from the conductive substrate and ready to electrochemical print automatically .It parts can then be assembled into something else.
@@sidneo14 "self replicating" ... stops there ruclips.net/video/DtudW-sfSSE/видео.html There's like a bazillion things wrong with that buddy. Your sponge and syringe gonna wear out, it only prints materials that can be electroplated, the tolerance is all over the place. You wanna self replicate machines that print anything maybe you should start by learning how to stick molecules on the tip of an atomic force microscope or something.
You mean 3D printing processes that have had decades of research and commercial investment are more refined and faster than a new technique? Colour me shocked. There are multiple avenues by which the speed and quality could be improved. Give it time, then judge.
![[DRAGON BALL LEGENDS REVEALS ＆ STUFF] LEGENDS FESTIVAL 2024 Special Edition Part 1](http://i.ytimg.com/vi/mV5bGIswTtQ/mqdefault.jpg)
This is more accomplishable for hobbyists than lasers and electron beams .. keep it up!
Thanks. Yeah, hopefully it's quite accessible. Hopefully, we can share some updates soon on progress on this approach
@@BillyWu is this process fast enough for large part printing ?
I worked on almost exactly this concept 10 years ago.
It's kind of interesting to see you guys go to the same problems i went through.
I'll give you guys some freebie tips.
Use a liquid hydrophobic mask to increase your resolution.
Get rid of the sulfate electrolyte (you do not want sulfur in your deposit material). You want an electrolyte that doesn't have side chain reactions. I can tell you what I used, but it'll be more valuable if you figure it out for yourself.
You can get rid of the dendrites and greatly speed up your process by physically accelerating your electrolyte.
Brownian motion in the electrolyte is what forms the dendrites. Add a focused magnetic field. You have a magneto hydrodynamic cell but you're only controlling half of it.
And lastly... The photoelectric effect is your friend...
I'm not saying more unless I can be involved in the research.
Thanks for the comments
So it's literally ECM in reverse. That genious!
In essence, it's simply a matter of putting together a more robust design, and then releasing it as an open source sort of affair for the sake of facilitating technological progress.
Defo. I have a student working on a more robust design where the aim is to put this into an open access repository for people to improve on it but we want to make it less buggy first :)
@@BillyWu Thank you so much for your work & publishing it in such a fashion. Following, appreciate your contributions to a "Makers" world.
@@BillyWu Thank you
@@BillyWu is there any progress on this?
Cool, but a demonstration would be nice
I've been imagining a similar setup but using the equivalent of a felt pen to convey the electrolyte. I'm super excited to see someone pursuing an electroplating-style metal printer. I figure it's potentially the most economical and simplest way to bring metal 3d printing to the masses.
Great to see the interest and hope it sparks further ideas. A felt tip could also work though you need to be careful about getting the deposition in the felt itself which is why we avoided this approach but I'm sure there are ways around it
@@BillyWu I wonder if the idea would work better upside down since you could counteract gravity if leakage is an issue. Also, since the resistance should change as the distance increases, I wonder if you could make the printer more precise by increasing or decreasing time spent in one area based on resistance readings.
@@ryelor123 Good comments. This current system works by having the electrolyte in the syringe with the sponge. Flipping it in this form would cause leakages. On the feedback mechanism, this is certainly possible and we had done some work on it and am sure this will be the way forward for smarter higher quality printing.
love to see the experiments and the outcomes you have. amazing concept
Thanks Anton. Hopefully can update soon on our progress
One way I had envisioned getting a metal part from a 3D printer was to print a porous conductive plastic model and then do electroplating and electric polishing. At some point we either use a solvent or heat to get rid of the plastic. And then more electroplating and electropolishing. As slow as this is it's probably still faster than my idea.
Certainly possible to do this and since the deposition area would be faster in some sort of electroforming process it would probably be faster. One thing to consider is that some conductive paints etc aren't 100% conductive when zoomed into the micro-scale since they're often a blend of conductive agents and binders. Some people also use electroless plating to get a more uniformly conductive layer
In Microelectronics and Microsystems we use electrostatic based 3D printing for prototyping (dacades old but still SOA) to move one atome at a time. Even for that industry, this method is very expensive and is used only in rare cases. This here is really a nice new development and is very interesting
Thanks for the comment. Great to hear that there's potential applications in all these areas
A quite low-powered green or blue diode laser can create a local dipole, resulting in the deposition of copper from an electroplating solution.
Congratulations! This looks very promising.
Thanks. Appreciate the comments
This general idea is awesome, I am anxious to see where this leads.
Thanks. We have a few things in progress to improve the system and will update once finalised
@@BillyWu Great work in general =) I have checked out some of your other videos.
This is incredible work here! As soon as I clicked on the video I thought “why didn’t I think of that?”. There are a few nuances I can think of, like whether the solution (be it the solution in the syringe or the tiny bit of solution touching the workpiece) changes in local concentration, or whether you need to introduce leaveners or other additives to the solution, or the inability to deposit some metals atop other metals. But I think all could be worked out. Also IIRC current is what matters more than voltage with electroplating, I’d be interested to see what constant gives you more consistent results. But what I’m most excited about is printing structural metals, and even alloys. Copper and nickel are one thing, but being able to deposit some sort of tough iron alloy would be incredible. I’ve seen a few papers on electroplating alloys before, I think in this case you’d just need to have both metals in the solution at once, and have two electrodes in the one syringe. Then the alloy ratio is determined by the ratio between the currents. Smart CAM software could even change the alloy composition dynamically as you print different parts, which could lead to an incredibly versatile process. A shame you can’t print titanium or aluminium (carbon too I guess), but I think from zinc all the way to platinum, there should be some reasonably strong alloys to be made. Perhaps you could make really compact thermoelectric coolers/generators, or even try depositing doped semiconductors.
What I like most about this is how easy it would be to build from an existing FDM printer. I suspect this will be a trending topic among enthusiasts in 2-8 months. All it takes is someone with enough time and money to try and make this practical enough for hobbyist use.
Are there any glaring flaws, big hurdles? Does the meniscus wick away down long thin cavities or drip away down the side of a tall print? Could it be better to print the whole thing submerged in the solution?
Definitely lots of ways to extend this. Electroplating of alloys is certainly possible and you can control the composition by changing the voltage. There was some recent work where they did this to make a thermocouple. You can also do the deposition in an electrolyte bath with a fine tipped electrode but the resolution isn't as good. Hopefully provide lots of ideas for others to build on. We're hopefully releasing some open source plans soon. Version 1 worked but was a little buggy lol
If a fine tipped electrode on its own had poor resolution, perhaps surrounding it with a ring at an opposite potential for shielding purposes might improve things? Of course you’d need to prevent the electroplating metal itself from depositing on it, but I think there should be metals with those properties, just as it’s impossible to plate iron onto copper.
I’m still most concerned with how a system would handle tall structures and overhangs.
@@Scrogan if we used an insulated tip as you say, you could probably grow overhangs directly out wards (within a certain aspect ratio). Mabye even printing them bit by bit with every layer that is added (a kind on non planar printing to optimise giving support to overhangs to print them easier)
@@Scrogan and the closer we have the tip, the higher the current concentration directly bellow the tip, and the more reduced it is to the surrounding metal.
@@Scrogan You might even be able to get away with using an enamel coated copper wire. The enamel insulates it, and only the tip is conductive and used.
As the wire is eaten away, the enamel should hopefully just flake off. Which admitedley might cause its own problems
this was exactly one of my to many ideas
The same
I feel you
Very interesting i hope we hear more from this printing technique.
A few things in the pipeline. #watchthisspace :)
@@BillyWu Ok I am very curious, subscribed
I had this idea a while back, never got around to trying it. So cool to know it would have worked!! Now I gotta find time and energy to make this.
Great. I think lots of different ways to take this forward so the more people working on it the better :)
@@BillyWuI think what you have demonstrated is really cool! Any fundamental scaling issues, or can this make a decent sized object?
@@weirdsciencetv4999 Thanks. Unfortunately, electrodeposition processes are a little bit slow so making larger items can be challenge but we're looking into ways of accelerating this. Certainty right now you can make very small functional items which still has a purpose
@@BillyWu an array of independently switched nozzles riding on the same head, kinda like an inkjet head comes to mind
@@weirdsciencetv4999 Yeah, there's a few patents which use an array of electrodes which you switch on/off in a bath of electrolyte such that you print a complete 2D layer at once. Can be faster but features not as well defined and needs more expensive set-up. Companies such as Fabric8Labs are doing this www.eenewseurope.com/news/no-heat-no-laser-no-powder-reinventing-3d-metal-printing/page/0/1
Awesome! You are doing a very good job!
Thanks. Appreciate it
Thank you for sharing this video. Well done.
Excellent work, great idea.
Thanks Javier
Sounds like the main limitation is speed, but maybe also resolution, especially once the part is actually becoming "3d". And it requires a conductive build plate/base. Maybe a non-metal build plate allows separation of the part and reuse of the plate?
Yeah, stability of the liquid meniscus is also a challenge. Some people have demonstrated really good resolution with the use of nano-pipettes however at the cost of speed. The challenge with a non-metal base plate is that we need electrical contact for the deposition but you can deposit on glass if you have a conductive path from elsewhere.
this is pretty smart i kinda wanna try this.
Great!
Interesting with this set up you should be able to print (very simple) electronics unto surface. You could be able to print simple logic boards and even rfid
Definitely
Did you think of using sponges brushes ?
You syringe tip approach look indeed best for precision but since the process is slow, you could save a lot of time with a wider surface.
The same way you have multi material the printer could use a second, wide, brush able to cover more area.
Also maybe a truncated cone could be squished to have some control over the width of the contact patch.
There's definitely more improvement and different ways of doing the printing tip. Certainly a larger print head area could help.
This technique is good for micro or nano 3D printing!!!
Yeah, there's been some work with nano-pipettes to make very small structures. The challenge then is the volume deposition speed drops so in general there's a trade-off with speed and resolution
Hecking cool and smart!
I wish I was smart enough to come up with these things!
Very interesting discussion of electrochemical deposition process. As you mentioned in the video a big limitation is deposition speed because of dendrite formation at higher voltages. I can see parallels with traditional ink jet printer's which could address speed issues. A 2D matrix head, with individual channel deposition control would be possible. Limiting constraints would require high flatness, separation spacing and accuracy of the print head and deposition bed. I would be interested to know what the experimental deposition rate you've been able to achieve for copper.
Thanks. In previous work we demonstrated a volumetric deposition rate of about 20,000 um3/s with a 400 um tip. Sounds fast but is actually only ~0.0012 mm3/min lol. Could speed up with more active area and a few other tweaks but speed is a challenge with this approach but certainly a few applications where that's less of an issue.
Wow, that was awesome! Really good spoken and clearly understandable Video. This technology is like the reprap start. This will be going to revolutionize the 3d metal thinking 🤔. Top video extreme big thx for sharing! ♥️♥️♥️ Are you going to make something open source?
Thanks. The first designs were a bit rough but we're hoping to refine and open source it
@@BillyWu really really nice! :-)
Should there be a small o-ring or seal around the sponge to isolate the electrolyte? Another thing I wonder about is if you could make this process more precise by adding it to a FDM printer and having the printer make a sort of shell of plastic around the metal to give precision?
Good questions. We tried to keep the system as simple as possible and found that with just the right compression and type of sponge that the back-pressure could be balanced with the electrolyte pressure to prevent leakage, however this was a fine balance that sometimes didn't work so improved electrolyte confinement would help. If you make an electrolyte guide as suggested that might help, though we wanted to see if we can achieve this without additional complexity, though perhaps supports might be needed
Containment of the electrolyte is moving in the wrong direction. You want to deplete or inactivate the electrolyte after consumption... Then it is self masking and regenerable.
This is a brilliant idea, kudos! Looks absolutely fascinating. Two questions immediately come to mind:
1. Would precise control over electrode distance and a higher current efficiency electrolyte (nitrates perhaps?) allow you to operate directly in a bath rather than your syringe? McGeough's (U. Edinburgh) comments on the relationship between efficiency and resolution for electrochemical machining come to mind.
2. Would you be able to take advantage of the double-layer through the use of pulsed current, akin to Pulse-ECM? This may be able to improve resolution and perhaps allow for direct in-bath printing as well.
Thanks for the very interesting comments. There have been a few other works which use fine tipped electrodes to do the electrochemical printing but the deposition tends to be a bit more spread out. You can address this somewhat from having a finer electrode which is closer to the substrate but we wanted to try to use lower cost components and avoid costly position controllers. On pulsing, yes this should be possible and have seen some work where people do pulsed deposition to get favourable morphologies. Lots of potential variables to optimise further so am always happy when this might inspire someone to give it a try with a new approach.
Really perfection..!!
Thanks a lot!
Thank you ☺️
Glad you found it interesting
Electrochemical method for 3d printing is too slow for now. You could do an analysis for femtosecond laser technology for combining electrical conductivity of femtosecond laser with electrochemical deposition.
femtosecond laser doesnt form a plasma channel in water last i checked?
EXCELLENT - Waiting for outcome please , Best of Luck
Thanks. Hopefully update soon on progress
Very interesting video. Though the overview is missing the newest and the most promising metal 3D printing method - Selective Powder Deposition (SPD)
Thanks. We tried to capture these types of deposition techniques under the "directed energy deposition" category but agreed that there are new forms of printing constantly coming out so always great to have a discussion around the newest approaches. The pdf of the overview in the video can be found at if interested am-hub.dk/wp-content/uploads/2018/01/IMSE-Briefing-paper-2-AM.pdf
@@BillyWu Thanks Billy, that's very interesting. Any idea when Electrochemical 3D printers might appear on the market?
@@iro3d There's a few companies such as Fabric8Labs which seem to have a reasonably mature product. They use an array type approach so might be a bit faster but with challenges around resolution. I suspect prototypes of this form are reasonably mature but the challenge is whether there would be sufficient consumer interest for a printer like that with the known challenges of scale fabric8labs.com/
Do it! I so need a metal 3d printer I can afford! I would be thrilled with an aluminum printer for maybe a few thousand. Or even a nylon SLS printer for that.
Mind blown!
İ thought this kind of electrolysis for iron, ironically stopped after i learned how much current required for a gear.
Is there a limit to what materials you can use... obviously you can use copper, nickel, and chromium but for example metals like steel, aluminum, titanium, and a multitude of other metals, basically what im asking is are there limits to which materials you can use, also you could probably make an alloyed rod to going into the syringe to get alloyed printing? idk just think here
Great question. In principle any metal you can electrodeposit you can use this approach, however one thing to bear in mind is that we designed this to use relatively low cost and accessible components. Some electrodeposition processes use more aggressive electrolytes and temperatures to get a good deposition and therefore would need modification to make safe. In some cases you should be able to control the alloy composition by varying the deposition potential and control.
BRILLIANT.
Thanks. Hope it was useful
Hello Dr. Wu - any updates on this? Curious if you have made any advancements from the time this video was posted.
Thanks for the interest in the project. I had a couple of student projects to open source the printer but it's not quite ready for release yet but hopefully soon.
@@BillyWu awesome! Where is the best place to be on the lookout for the release of the open source project? Would love to see what comes of it
Amazing approach! If the speed limitation could be overcome it could spark almost another industrial revolution, or at least help bring manufacturing back to EU and USA.
I was wondering if localized electrochemical deposition could be used to join 2 wires (e.g. welding wire) of the same metal together, essentially speed up the process significantly, for the cost of some precision?
How do you manage the replenishment of the anode?
In our case, the anode which was a copper rod was periodically replaced manually. We were mostly printing very small items so the mass of the copper rod was far in excess of the printed item so we didn't have any problems with running out of material, but its a simple case of connecting a new rod and putting it in solution.
Это отличная идея, когда она будет реализована на практике?
I was just wondering if this concept would work! Very cool, and well explained, too. Is it possible to print in alloys, or does the plating process limit the user to elemental metals?
Thanks and great question. Yes, it's possible to print alloys. In principle anything you can electroplate, you can print with however in the case of some metals the electrolyte and conditions start to become a bit more challenging to handle. We've mostly focused of aqueous systems (water based) but lots of non aqueous systems exist for more challenging materials.
@@BillyWu Nice! Thanks for the reply.
Can you print at an angle with this technique or are you restricted to sharp straight structures?
Angles are definitely possible. We have some new results but it's taking awhile to get out. Watch this space :)
If you use a bath, you could have a bed of nails style print head with hundreds of heads which could speedup the printing process.
Definitely. I believe a company called Fabric8Labs is taking this approach with an array of electrodes to hopefully help speed this up.
what kind of post processing would parts made with this process need?
Normally we just need to remove the part from the substrate. In our case we printed on aluminium foil and then were able to just manually peel off. If really stuck in theory you can just dissolve the aluminium away
Hi, I love it. I was big in Electroplating of high aspect ratio structures back in my research days. Used to use Megaherz ultrasonics to break up the boundary boundary layer.
But I was wondering if you are controling your deposition spread just via the size of the meniscus or optimizing your frequency and PWM too. So you are basically using the the charging of your boundary layer to focus your mass transport to the close areas?
Thanks Jens. Yeah there's a few approaches for enhancing the mass transport properties such as electrolyte flow systems and as you say ultrasonic systems. We control the size of the deposition mostly through the size of the nozzle, it's separation from the build plate and the deposition conditions mostly but are also looking at other approaches too which we hope to share soon :)
Did you ever end up writing a paper about this. Am very interested in your results
you really should check out ir03d's selective powder deposition printer. needs a kiln, but does wonders
What are the material properties of resulting prints?
Good question. We did some indentation tests and generally for the copper this was harder/stiffer than bulk cast copper which we attributed to the fine grain structure when printing with electrodeposition, where smaller grains generally give harder/stiffer materials but are more brittle
Or the syringe could implant ions into a permeable dielectric substrate; coated epoxy-fiberglass?
Washing the substrate in a chemical bath would initiate 'electro-less' metal plating on the ions.
Plated-thru vias in PCB's are made this way.
Printing the pattern would be fast; the time-consuming plate-up process is done in batch mode; thousand of parts in the same bath.
What do you think?
Yeah potentially printing a pattern first and then increasing the thickness could work. Definitely lots of different ways to take this forward so more people working in this would be great
Well-Done
Thanks
This is extremely interesting. For the sake of complexity reduction could this not work with the part and nozzle submerged instead of maintaining a meniscus? I am assuming the current would still be constrained by the nozzle wall and predominantly favor the the shortest path. With a bath you'd be able to circulate electrolyte through the nozzle though I'm not sure if that is actually useful. Apologies if you addressed this in your first paper. I don't have access to that journal.
Hi Alexander. You're right that this approach would also work with a bath of electrolyte. Others have done this with sharp electrodes to help focus the current. Having a nozzle in the bath may help to also focus the current however the challenge would the be the difficulty in doing multiple materials, where a bath change would be needed. Definitely worth looking at though.
Very cool - you have just given me an idea,,, :)
Great!
very interesting but I get the impression there is still a way to go for optimizing this .. right ??
Thanks. Definitely room for lots of improvement and loads of potentially different approaches to solve the problems :)
How are the dendrites negatively effecting the metal? Do they also cause poor adhesion? Just wondering if dendrites are acceptable trade off for example some applications.
Dentrites would result in high porosity, i.e. low strength.
Hi. As others have mentioned dendrites generally result in a mossy and porous structure which has poor mechanical properties, so ideally we avoid this by going slower.
@@BillyWu Thank you for your answer. Poor mechanical poperties is kinda of subjective term, though. Since it really is depends on the application you be using for. If it's really porous it might be good for heat shielding or something. Really cool technique since are also kind of able to also determine the density of the metal through the voltage. Or is that little bit to far fetched.
Could you submerge the entire build palte in the copper sulfate soloution and jsut use an electrode that very very localy applies the voltage and builds up the part?
Definitely, that would work too, however the features would be less well defined, though some use very fine electrodes. The meniscus means that you only get deposition where there is liquid.
@@BillyWu But what do you do with the menuscus once you have printed one layer, ir even want to put down a line next to the current one? It will start to pool, leading to a reduction of accuracy.
Could you also point me to a paper or two on submerged electroplate 3d printing I could read? I am very interested in this process. Thanks
Good question. Close control of the meniscus is key. This article talks about the work from a team ETH Zurich who show the meniscus in more complex 3D prints www.azom.com/news.aspx?newsID=57437. This review paper talk about localised electrodeposition link.springer.com/article/10.1007/s00170-020-05799-5
@@BillyWu Have you read this paper, very intesting. That high plating rates (50um/s for a 0.5mm nozzle) could enable macro scale 3d printing.
I just put it into my slicer, and you could 3d print a 20x10x5mm object in under 24hrs
This mathode is possible to use in existing FDM printer ?
Yup. We modified a standard FDM printer to achieve this by replacing the print head with a syringe holder and filling this with an electrolyte and a rod of copper/nickel. You'll also need a power supply to enable the deposition.
Very interesting stuff.
What is the ideal & maximum volumetric deposition rate of this technique?
Is it possible to put multiple electrodes with different metals in a single syringe and then apply a voltage to the electrode with the metal you want? Would it be possible to make alloys like this by using multiple electrodes at once with different levels of voltage?
There's a few other studies that have done co-electrodeposition which by controlling the deposition voltage you can control the composition. For example in this paper pubs.acs.org/doi/10.1021/acsami.0c01100 they do a nickel-copper deposition with controllable composition which is quite nice for making sensors. The challenge is making sure a constant amount of metal ions stay in solution
Do you think gold metal would take faster?
We haven't done gold electroplating yet but I think it's slightly faster than nickel, however it does normally use a cyanide based electrolyte so we avoid this where possible.
What if make printing head like flour sieve (with gates or something) and print lot of particles at once ?
Yeah. There are 2 main approaches for electrochemical additive manufacturing. What we showed here is the meniscus confinded approach where you have a bead of electrolyte to focus the deposition. You can also have a bath of electrolyte and use an array of electrodes. This needs more electrolyte and th resolution of the features can potentially be less well defined, but it can be done.
A very interesting concept, thank you for sharing!
However, am I correct in assuming that this can not print "islands", structures which are not supported by already printed material underneath?
Conventional 3D printing solves this by printing supports underneath, however I assume they might adhere to much when using metals.
To me, it seems like that's where printing in a powder bed or gel matrix excels. No supports needed and even "floating", unconnected parts won't fall of until they're fused.
Yeah, to deposit material we need electrical contact. You can create some "island" structures by printing in a different material and then selectively dissolving the support material in a similar way to how water soluble supports work in FDM printing. True that powder bed fusion for polymers can make these freestanding structures. Much more difficult with metals though, due to higher thermal stresses. Just messes up the print bed.
@@BillyWu
That does make sense. Thank you for the answer! :)
And again thanks for sharing this process with us!
@MichaelKingsfordGray
...come again?
Is it possible to deposit metal from molten salts?
That could enable 3D printing of metals such as aluminum or magnesium.
Certainly. If you can electrodeposit a material it should also work in a localised deposition printing approach. Some people use ionic liquids to do deposition of metals so more exotic materials than copper are possible.
Could this be used to print something like 4140?
Good question. We've mostly focused on pure metals such as copper and nickel. There are some alloys that you can electroplate, however since steel is an alloy of iron and carbon, it could be a bit more tricky since this is normally manufactured at high temperatures through mixing and solidification.
There are a few fundemental problems with this.
First, electroplating is slow, and I dont know of ang ways you can speed up the process by several orders of magnitude. At the voltages you describe, it would take weeks to complete a decent sized print.
Second, strength. The faster you grow your structure, the weaker it will be. Electroplated structures are more similar to wet sponge than anything else
Fair points. Speed is still an issue with this approach but there are some niche applications where the scales are much smaller and multi-material might be useful, but still at the R&D stage
@@BillyWu multi-material work is limited by the chemical reactivity of the plated element and electrolyte. You really cant do anything with it. You cant use alloys, and the whole thing will corrode because you have two different metals in close contact. This is a step backward 1000 years metalergy wise
Thanks for presenting your progress. I wonder if a wax layer or oil would keep the dendrites forming as an interference compound over the base material. Something the nozzle could move thru but still make a deposit later.
Do you have a list of parts to purchase for someone to construct the same setup you have?
Sure. You can find more details about the set-up in our 2 papers www.researchgate.net/publication/319315170_A_Low_Cost_Desktop_Electrochemical_Metal_3D_Printer
www.nature.com/articles/s41598-019-40774-5
We're hopefully going to be releasing an open source version of this soon once we iron out some of the initial bugs.
Could you use an array of say 1000 head outlets to all be printing in parallel to speed it up?
There's a few works which use an array of electrodes which you turn on and off which can accelerate the process but generally this has to be immersed in a bath of electrolyte. There's also a few other things you can do to increase the deposition speed such as higher temperatures and different electrolytes. We just used a low cost solution as copper sulphate is quite cheap. Some root killers use copper sulphate
Very interesting stuff! How robust is the printed structure? Is it as strong as the Laser melting method?
Thanks for the interest Stefan. From the tests we previously did, we can see that the hardness of the material is generally higher than cast copper due to the small grains that we form during deposition. We haven't compared this with direct laser melted copper, but would imagine that its favourable since we have low defects, though our sample size remains relatively small.
@@BillyWu Wow, that's great! I didn't expected that. Thanks for the info :-)
This is pretty great
What is the top-speed limit of your method?
Do copper and nickel deposit at the same voltage? Given two electrolytes don’t cause some undesired reaction, could you use a single syringe and apply alternating voltages to perform multi-material deposition in one go? (Not a chemist). Follow up: would aluminum sulphate work as an electrolyte? Could you print aluminum parts?
Voltage is a relative measure between 2 different points. Since we use copper/nickel as the anode and cathode the voltage we apply to get the deposition is similar but generally nickel is a bit slower without any additives. It is possible to do a co-deposition if you get the mixtures right and you can electrodeposit alloys in this way but its a bit more tricky. Aluminium can also be electrodeposited, but often required more challenging electrolytes and we wanted to stick with water based solutions. For example, we stayed away from gold since its more expensive for prototyping but also many deposition solutions use cyanide containing components. Essentially, nearly all materials you can electrodeposit with you can print with.
@@BillyWu thanks for answering my questions. Really novel approach. What are your + your team’s current thoughts on ways to speed up deposition? I understand that y’all were able to increase deposition speeds of Cu by 34% and Ni by 85% by replacing the sponge with a nanofibre.
@@davidlanday6102 There's a few potential avenues. Certainty moving to other non-aqueous systems can potentially accelerate this but will be a bit more tricky to handle the electrolyte. Additives can help and also print head designs that minimise the mass transport problems I think are ways forward.
Doesn't it take to infinity to finish a 20cm high print ?
What are the cons in here ?
Yeah, you're right. This approach of printing metals with electrodeposition is a bit slow for large parts. There are a few tricks to speed it up but at the moment the main applications are likely to be smaller items, where the ability to put down metals means you can start to build conductive pathways
Just out of curiosity, what do you believe the price point of a ECAM printer once fully developed?
Great question. The components themselves aren't very costly and mirror that of standard plastic 3D printers with the stepper motors, frame and control electronics. The different components include the pipette, electrolyte and power supplies which are all generally low cost. The challenge is that speed is still quite limited in terms of volumetric deposition rate so still needs some work.
I am looking for a technology solution for aerospace for printing fuselage and a wing in one piece. Can you see It's doable for a bigger printable object like an airplane?
Thank you in advance.
Unfortunately, this approach isn't really suitable for large scale structures. I know there are some "big area additive manufacturing" (BAAM) approaches such as www.e-ci.com/baam but these mostly print with plastics.
Amazing!
buen trabajo
Could this be combined with plastic printing to make 3D embedded circuits in stuff?
It's a bit tricky as it needs to be printed on something conductive to have access to a flow of electrons, but what some people have done is start from a point and then print a circuit from there, so in theory you could do that
what about using ordinary cheap copper wire at the extruder with a extruder made of plastic or something nonconductive with a flat tip of the wire dipped in a hydrochloric solution with a thin metal plate at the bottom. that should allow the copper to dissolve rapidly and build up on the plate. can also increase the voltage to that of the amps to accelerate the electrons in the solution. with about less than a milimetre of gap distance chould reduce random ion floating outside the printing alignment. the rate of print is only limited to that of conductivity in the solution. printing speed is something that can be perfected. you still need the the feeder mechanism of the copper spool to keep the gap between the wire tip and the plate fixed as the plate or wire moves.
Yeah, we considered using a wire feed as well but wanted to try and keep it as simple as possible, but yeah there are lots of alternative ways to do this. Would be great to see what other designs people can make
Doesn't work well. Already tried it. It does work up to a point, but past a certain speed the water starts to boil and it destroys the print quality.
To go faster would involve non-aqueous electrolytes and high temperatures... Which are decidedly unfriendly to work with and come with all kinds of surprises like thermal warping and cracking.
Basically, wire is a no go. You'd be better off by modifying a traditional welding process... Which isn't as hard as it sounds. Miniature welding is totally possible and at reasonable speeds and temperatures.
Papers are one thing, but I want to see it in action.
What's the largest object printed with this technique and what time did it take to print?
Edit: I see from other comments that it would be really small
Yeah, unfortunately it is quite small in terms of volume. It's relatively each to make mm-scale objects but with limited depth. We demonstrated some very thin lettering in the work that was several mm. Could potentially be used to make circuit tracks... slowly, or sensors
If I had the money, Billy, I would love to invest in your work. If I could make a bunch of laboratories for rent, I would do so.
Thanks. We aim to share our work as widely and openly as possible :)
This is epic
Thanks for the comment
But how are you going to get liquid stainless steel and liquid aluminum? Makers, Designers, Builders and Inventors aren't clamoring to to print copper parts.
I suppose there's a possible chromium nickel solution that might be somewhat viable... but maybe not as a functional part itself.
What I picture as a possible low budget solution, is suspended microfine powder (agitated upon nozzle dispersion) in a UV curable medium. Much in the same way that flat bed UV printers can print white ink (titanium dioxide) through piezo electric discharge. Yeah, you are back to the binder removal process, then the furnace/kiln firing in an inert chamber thing... but honestly, that's a device that could be cheap if produced at mass scale. We aren't talking about much complexity beyond a PLC, heat bricks and heating elements. The door just needs an airtight seal, and a gas valve solenoid. It's literally not beyond the skillset of a competent fabricator. (So, somewhat DIY to a degree)
I think we need to explore industrial print heads, and their maximum viscosities they can push, and what mesh powder can we squeeze through them before clogging is an issue. I mean we seem to be quite OK with a 0.4mm nozzle hols for FDM printing, provided the z-step is minimal. I'm willing to bet that quite a bit of suspended metal powder could squeeze through a pressurized print head with nozzles that large.
The UV curing could be localized too... utilizing a fiber optic approach, you could get the light right down by the nozzle with a light pipe, and not reflected anywhere else, like back up the nozzle hole.
Nice suggestions. There are also metal jetting systems which are starting to emerge for larger scale metal printing but you're right this approach has some limitations
How long it will take for print a 10 mm cube ( 10 % filler , 0.2 layer height ) ?
Unfortunately this approach is a quite slow for large objects. In our previous paper we showed that for a 400 μm nozzle, we can achieve a deposition rate of 20,000 μm3/s. So for a 10 mm cube with a 10% infill that would take approximately 58 days to complete. There are a few things you can do to accelerate this, but this is the reason why this is still mostly at the R&D stage. However, there are applications for smaller scale structures which are electrically conductive.
@@BillyWu This is very good for Circuits and creating own IC's and Transistors. Possibilities are endless and this opens field in new direction. Imagine 200 or 300 printing same model at same time like an transistor or an IC which have new capabilities which we had never thought of!! The production can easily be achieved and there will be low cost of creating dust free labs for such printers and robots or automated arms just change the build plate and stack them for someone to collect the pile at the end of hours or days!! Its a breakthrough in nano technology. The projection of this printer wrong, it should be projected as Nano Metal 3D Printer. Keep it up and one day it will be high in Demand like a 48MP camera sensor in Smartphones! :)
@@himanilsharma2147 Thanks. There's been a few examples of applications of this approach in electronics, so you're right this is likely to be the initial application area if it takes off.
@@BillyWu You know I personally believe 3D printing is next Industrial revolution like lathe machines did and current technologies of 3D printing are primitive and we will see more advanced techniques and technologies in coming future. Where everything will be automated like we see as VMC, if someone in 1920 or 30 had told about VMC to someone then People had laughed, same like Miracast and Airplay like technologies which enable to do video chat on TV!!
and this technique which you demonstrated in video is Nokia 3310 which enabled anyone to play game in Cellphone first time in the human World, that too Snake game in almost a device having pixel bigger than 1mm, and within just 20 years children are playing PubG like game in their own smartphone worrying about latency!! Its a huge breakthrough and I personally believe we will see that soon very soon just within next 10-15 years.
@@himanilsharma2147 Thanks for your words of encouragement. Hopefully we can update soon on some work in progress in our labs.
what if you take potonion and uranium and stack on each other layer by layer in between the layers will be will be a insulator
Haven't worked with uranium so aren't that familiar with its properties unfortunately :)
@@BillyWu I'm not know scientist or physicist I don't even have a GED or high school diploma I have dyslexia I'm one of these people that had no opportunity when I was a kid and could have done great things in my life if I had standard opportunities like most people in the world but because I was abused since I was a kid I can't do stuff to change the world but I don't care I will tell people things I I know that will change the world
@@BillyWu here's something for you ruclips.net/video/9nfaYAU92tY/видео.html I thought this years ago
Thank you. Very interesting information. Now, if I could get a useful metal printer for about $10,000, without needing a machine to wash the parts out, and not gave to sinter, I’d be very happy.
Butblow cost here no doubt means at least several times that.
Uhm, basically a facility employing this type of electrochemical additive machine in industrial scale is going to produce the same type of waste as an electroplating facility. Therefore, it's going to be subject to similar environmental and waste management regulations as such. I'm working in the electroplating industry and I can tell you de-ionizing waste water processing systems are extremely expensive to buy and to maintain. Oh and in most countries you need to apply for a permit to run a facility that involves heavy metal ion solutions, and that can also be hugely expensive depending on where you are located. I guess in small scale for DIYers or small offices where they'd just dump the waste water into the drain would be fine. But if somehow it got extremely popular and start having an effect on the environment, it could ended up being banned by the government.
*metal 3D printing companies sweating*
BILLY WU FOR PRESIDENT
3D printers for all
Is this patented?
There's a few patents around in this area with different variations of electrochemical additive manufacturing and a few companies are trying to take this to market such as Fabric8Labs
@@BillyWu Oh well ... then we won't see this technology at a price worth paying, for the next 20 years :(
Patents are pure evil!
Shut up and take my money!
I suppose it's an extremely slow process.
Yeah, unfortunately the process is a bit slow but there are ways to speed this up which are being researched
Compared to FDM? Yes, but it doesn't have to be super slow.
The whole process is only limited by heat. Heat changes the chemistry, so there's an upper limit and a lower limit.
The basic problem is how fast can you dump the heat from the reaction into the environment without changing the quality of your results.
Electroplating to anything other than microscopic parts will be extremely slow.
Yes, very fair point. Electrodeposition is quite a slow process. There are ways to increase the speed with different electrolytes and designs but will likely always be a challenge, however there are some applications where this might be suitable given the ability to do multi-material. Still R&D at this stage but it's good fun research
You could have an array of electrodes, and thus deposit material at many places at once.
10 electrodes would reduce print-time ~tenfold ... and so forth.
This process is so slow your talking about 1mm per hundreds of hours
Skip to 4:21, where the video actually starts.
This is not low cost, its unbelievably slow, lithographic masking techniques followed by various deposition or erosion options are faster, cheaper and more precise. DLP activation of resists is a good area to develop a micro 3d print version of this. Or maybe a way to make conductive FDM filament that you can electroplate.
I see no viable path for the method shown in the video to compete with lithographic, DMLS or pla / liquid binder then oven centered 3d metal printing (whatever that's called).
I'm not saying it was a bad research project, but in science sometimes things fail, this is a thing that while technically feasible is not economically feasible nor do I see any way for it to become so.
You have a layer height of like 6nm and a linear speed of 0.4 mm/s this is ridiculous. And to call this low cost... I can take some sharpie ink and draw on a substrate then dunk it in an electro plate bath for like $10 with scrap anyone has lying around.
And then to brag about it being 4d printing because you can get some mems behavior.... OK I get it you did not coin the term but it really needs to stop. It's a retarded thing to call a 3d printed object that can be manipulated. You could call tweezers a 4d object by that definition, it's just bad and anyone who uses the term should feel bad.
RUclips comment section, where the real peer review happens.
Some fair comments. It is relatively slow and no doubt not the solution for everything but the approach has certain merits which warrant further research in terms of increasing the speed and multi-material aspects. Welcome comments from all asides positive and negative.
@zyxwvutsrqponmlkh you lack imagination electrochemical printing opens the door to modular self replicating devices (which can be exponential therefore solve the speed problem ) which is lot harder to achieve with light or thermal based systems. in here all you need is electricity insulated capillary(that can be printed) and an ionic liquid. It opens the door to manipulate electromechanical properties without post processing by selectively depositing metals with different propreties perhaps turning them into insulators or electro deposit ceramics such as calcium carbonate . I see the future of this from a bottom up perspective rather than a top down approach . As for 4d printing I see this as one of the most promising tool to achieve some types of 4d printing for the reasons stated above unless you want to narrow the definition of 4d printing to some passive environmentally triggered assembly a la skylar tibbit .I have used selective electrochemical depositing (not a 3d printer a 2.5d) to print programmable joule heating activated trusses that assembles into 3d structures. we even printed bimetal "memristors" or switches on them to program how they assemble.
Lookup the Harvard milidelta . A variant of it can be entirely be electrochemicaly printed subtracted from the conductive substrate and ready to electrochemical print automatically .It parts can then be assembled into something else.
@@sidneo14 "self replicating" ... stops there ruclips.net/video/DtudW-sfSSE/видео.html
There's like a bazillion things wrong with that buddy. Your sponge and syringe gonna wear out, it only prints materials that can be electroplated, the tolerance is all over the place.
You wanna self replicate machines that print anything maybe you should start by learning how to stick molecules on the tip of an atomic force microscope or something.
You mean 3D printing processes that have had decades of research and commercial investment are more refined and faster than a new technique? Colour me shocked.
There are multiple avenues by which the speed and quality could be improved. Give it time, then judge.
terrible audio. quiet and muffled like most english people ????