- Power calcs I'm bad at everything and misread the current limit of the buck converter. It's limited to 12A, so power efficiency was notably better than I mentioned (albeit still pretty bad). Oops 🙃 - Temperature! Sorry, I should have mentioned this: galinstan has a freezing point around -18 to -20C, which is near the lowest temperature experienced in LEO orbit. So it _might_ need a small heater to keep liquid, but maybe not depending on how much heat is retained from the sunwards pass of the orbit. The liquid metal will have a fairly large amount of thermal mass, and contained in an insulating polymer of some type. Also will experience some self-heating from joule losses during operation. Missions farther from the sun would need a heater to stay liquid however.
@@fluffy_tail4365 Kinda sorta. Won't eat the spacecraft, and lower freezing point which is great. But toxic so no one really wants to work with it if possible. It also ends up being not quite as efficient because it's less dense and less conductive (although lower viscosity too). Performance depends a lot on the specific geometry and magnet configuration, but mercury as a rough rule tends to be less efficient. But to be honest, I haven't seen many papers looking at it due to the toxicity problem so possible it could be optimized to run better... and just no one has because no one experiments with it.
Satellite reaction wheel bearings used to fail all the time, but the problem was solved in recent years. They've stopped using metal ball bearings and now use ceramic ones instead. The ceramic ones require no lubricants and are not susceptible to static electricity surface pitting slowly destroying them.
There's also the phenomenon of cold welding in the vacuum of space. I don't know enough about the clearances and lubricants but I'm sure the people who create satellite gyros do. Just wondering if it might be a contributing factor.
@@jimurrata6785it *is* a contributing factor but the pitting was significantly worse than initially theorized so thats why they *needed* to change rather than wanted to
@ker6349 well, ceramic bearings will certainly eliminate the static charge problem and are free to roll if not constantly subject to shock loads. It's a great solution for expensive assets that you cannot maintain. 👍
Are there any downsides to ceramic bearings? Do they have a lifespan or are they functionally "unbreakable" since they will never wear out before a satellite's end of life?
I've also heard of liquid metal bearings. They have really low friction once they are warmed up and spinning and run incredibly true. Their service life is so long, that the manufacturer deliberately went on to design them so they fail sooner.
I love this! However, it definitely requires mercury. I understand reluctance to use it, but Galinstan is neither particularly dense nor conductive, and its melting point is significantly higher than that of a mercury alloy.
Ionic liquids? Anyway I've never heard of this technology until now and it strikes me as being very important! Not merely because of the lack of bearing and wear or absence of vibration, but because, unless I am missing something important, it also obviates the need for magnetorquers such as on the Hubble to "push" against the earth's magnetosphere so that the reaction wheels can be periodically "desaturated" when they reach their rpm limit. It also eliminates the need for probes in deep space to burn fuel in the reaction control system to do the same thing! It seems to me this could significantly reduce the fuel consumption of long duration deep space missions.
You can find unprotected 18650 cells for sale at almost every local vape shop as well. 99% of "vape" batteries dont have protection because the mod or vape has those built into it by law, so it allows those cells to offer much higher constant current draw and many times that in peak current
Yup. 18650 cells do not have any kind of smarts. They are used in a wide range of applications, and all of those have their own protection. I was mildly surprised when I found them in a burglar alarm, as a back-up, in case mains were cut.
LiPo packs for RC also don't have protection circuits and they are designed for very high currents. Even the small ones I use in my combat robots which would fit perfectly on there can handle up to 40A.
@@andersjjensensome 18650 are protected(but arent really 18650s anymore since that is the dimensions). But you can usually tear off the wrapping to find the unprotected cell inside, then you can tear off the protection circuit. I did this once since my local electronics store only sold the protected type
9:05 The protection circuit is located at one end of the battery, under the insulation. It can safely be removed. I recommend getting high output cells, such as the ones used in power tool batteries.
@@BreakingTaps You can easily buy 18650 batteries without protection circuits, most are unprotected by default. Protected ones are usually explicitly labeled as such. Rule of thumb: battery length ~65mm = unprotected, ~68mm = protected.
@@BreakingTaps if it has a button it will have a protection circuit, flat top cells are always unprotected. the protection circuit is a small pcb on the positive end normally encased in a shield of the button.
My recommendation is molicel. Very high current batteries, and very affordable. I usually do through 18650batterystore. They ship properly, and their customer service is great.
Glad to see you closing the magnetic circuit with the iron. Even if there isn't enough, you're making a static field so hysteresis isn't really a big deal. Too many people just don't bother understanding a flux circuit and just rely on the unreasonable strength of rare earth magnets to leave all kinds of magnetic performance untapped, or just double the magnetic material required.
I remember learning the formulas in university. Even just a millimeter air gap is such a huge reluctance! Lacking a back iron really wastes a lot of potential.
In addition to reaction wheels, satellites also use control moment gyroscopes, which are a similar device that uses conservation of angular momentum to control satellite orientation. The key differences in performance between the two are that reaction wheels are more power efficient for smaller satellites, and larger manoeuvres, and CMGs are more power efficient for larger satellites and smaller manoeuvres. CMGs also provide more torque for less power. There's also a difference in how they operate. Reaction wheels change angular momentum by spinning up and down. CMGs change momentum by changing the wheel's axis of rotation without needing to change the angular velocity. Reaction wheels have their place, because many satellites are very small. But for larger craft, like the International Space Station, they use CMGs. The ISS has four.
There's also the magnetotorquer, which uses a dipole to generate torque against the planet's magnetosphere. It requires detailed knowledge of the local magnetic field gradient to calculate the torque correctly but they are commonly used for satellite attitude control and stabilization.
CMGs also have a lot less jitter than RWAs. So your platform is more stable, great for deep space telescopes where you need accurate pointing for long periods.
Yeah definitely. Refractory metals (tungsten in particular iirc) are fairly resistant, and some will "just" embrittle over time which might be fine once you're in an orbit and not experiencing much force. But definitely a concern for long missions. Considering the amount of current that's dumped through the electrodes something like graphite might be to inefficient unfortunately. Definitely possible with mercury, although you have to deal with the toxicity and it's not quite as efficient (density and conductivity). There are other options too, like ionic liquids or just really salty fluids, but they lose suffer due to lower conductivity and less mass. Very neat space to explore though!
@sammiller5509 It does! There are a bunch of pro/cons with different scaling losses, but yeah that approach can work. You can also run it as an AC conduction pump too (with both electric and magnetic field varying). Check out this paper (ntrs.nasa.gov/api/citations/20070022272/downloads/20070022272.pdf ) for more details, it's an interesting read!
We called'em mercury vortex engines. I've made them as far back as 30 years ago. I never thought about space application. I imagine friction makes conventional reaction wheels more energy efficient, than a fluid that is going to be continuously venting energy through heat.
@ I used mercury. Technically, you could lower friction of it by raising temperature or using gas as long as the density can form a conductive path, like in CFL bulbs, but the tradeoff is that with less density, there's less reactive mass. I mean, it's not like people haven't already thought of all the options in this video and rejected them for a good reason. Math just doesn't work out. My idea was to use this sort of propulsion to build a completely silent submarine. Went to science fairs with my prototype build, but never won anything.
The Round Shape can be optimized by making the shape not round , rather a saucer UFO shaped oblong. Try that shape, then laminar flow of the liquid metal seems to flow better in that shape. In theory, liquid metal could be sandwiched in UFO ships that spins at 50k RPS, to har that anti gravitic reaction.
This kind of tech could be expanded into a tube around the planet using the torque of the moving metal to keep said tube from falling to the ground and you could climb that tube to get to space
Here's an idea: You can reverse these devises to be a rotation detector. Any rotation with induce a current in the motor (this is how people detect rotation)
@@kellymoses8566 very poorly. Resonating structure gyros have achieved unbelievably exquisite levels of sensitivity and precision and can now reach FEMTORADIANS per second levels of rotational measurement accuracy.
I remember when I was a kid (a very very long time a go) there was a short article in Pop Sci or some such (paper) magazine about some scientists making a mercury "gyroscope" using magneto hydrodynamics to push the mercury around a metal tube loop. They hinted it could be used in place of spinning wheel gyros with then poor reliable ball bearings. This was before even the 1st satellites. Your video dredged up that long forgotten memory. What goes around comes around.
One problem is the space temperature, liquid metal need to be in a range of warm "ambient" temperature to be liquid, this device will need another system, like an temperature control to work properly, but it can be a little tricky in space, when we speak about fuel or energy
Thank you for the use of a back iron to redirect magmetic forces. I think a lot of people get discouraged from thier projects when they see the costs assoicated with the magnets per pound of strength and just give up then and there when really a lot of projects could be done will far lower costs by just using a back iron
Conversely, the relatively low cost and high availability of NdFeB magnets makes for a lot of silly project videos -- sure you can make a, say, windmill that makes a few dozen watts, but it's massive and full of air space and doesn't look anything like a commercial design! Why is that? Well, it's not much effort to go one more step and consider the back side of those magnets too!
If my memory serves me, Canon used air bearings on the laser assembly of their earlier color copiers. The model 100 I know had it, I believe the 1000 did as well. If you somehow come across one of these dead, might want to grab that assembly - it likely wouldn't work for this type of project, but be real fun for scattering a laser beam around.
A flywheel would keep a constant attitude with no extra power but such a liquid reaction wheel would start counter rotating as soon as the power is cut due to fluid friction loses slowing the inner liquid. interesting concept nonetheless
@@hashbrown777 Would be interesting how efficiency scales with size. As of now frictionless seems to be achievable with magnetic bearing flywheels but reducing fluid friction to negligible seems unlikely. A liquid metal's properties don't seem to help in that either
For the galinstan corrosion risk there’s always the option of option in a, potentially vacuum rated, epoxy. Especially if it’s not necessary to service it as it’ll be unlikely to react to any galinstan that escapes it’s primary containment
Can you imagine a science fiction senario, where a galinstan module breaks in orbit, creating tiny orbiting particules of galinstan, forcing all future space craft to have nonmetallic shielding? Damn, that would be a nightmare.
@Superwoodputtie whell there are already a lot of Natrium droplets from cold war atomic reactors circulating earth... But you don't have to care about corrosion with 9km/s...
I think he meant regarding electrodes. I'm not sure about Ga, but In and Sn at least have well understood solubility against base metals, like Fe or Cu -- for which, it's a fairly low fraction, and once the solution reaches saturation, that's it. For example, Cu can be used _at temperature_ to hold liquid Sn, albeit with notable erosion (Sn diffuses somewhat into Cu, forming intermetallics (IMC); the IMC dissolves, putting free Cu into Sn solution). Or the solubility of Fe in Sn is very small indeed, hence its popularity for soldering iron tip plating. In I think is similar, but I'm not sure about Ga. The tricky part is when these elements aren't quite soluble in the base metal, but diffuse into them anyway -- this is most apparent, and dangerous, for similar metals like Al and Ga, where the Ga diffuses into the crystal structure directly. I don't know the full extent/dynamics of this, but the diffusion rate of the elements through various alloys is the property you want to look for. I expect these data are available somewhere, well understood, and exotica aren't required. If nothing else, a (particularly sturdy: pinholes and fractures are a bane!) plating of a hard metal like Cr, Ir or W very likely handles it, or various compounds like conductive oxides, carbides, sulfides, etc. Other metals or alloys might be of interest; efficiency is directly related to resistivity, and I wonder which liquids are best. Mercury almost certainly is a poor choice, having much higher density and resistivity than most; NaK might be competitive with Galinstan. I wonder if molten Li might even be an option? Resistivity is probably worse at elevated temperatures though. (Incidentally, apparently Li and Na are immiscible? So there's no lighter analog to NaK.)
IT VIRKS! I used to work with a guy who was convinced in his idea for a reaction motor for linear motion in space. Basically an arm with a counterweight spinning around an axis fixed to the spacecraft body. The arm could move in and out of the axle - the idea being that as the counterweight spins, it is retracted and extended along the desired axis of motion using a rack and pinion setup, "push/pulling" the spacecraft forward in space. Needless to say, he didn't understand the physics of conservation of momentum.
There are some tether concepts using the gravitational gradient and something similar like you describe. Some where already tested in space, they work and do not contradict any laws. (see nasa tether propulsion handbook)
@@lukasskymuh5910 Nah, this guy's hypothesis was nothing nearly as subtle as that. It boils down to the notion that, say, if you're spinning a sledgehammer in space, you can pull yourself forward at 0 degrees and push yourself backwards at 180 degrees with a net forward linear motion with no external gravitational tethering.
@LilMissMurder3409 the weird thing is that such a simple mechanism this would actually increase your orbital in a gravitanional field with the correct timing. Unfortunately, not efficient enough to be interesting.
What about using a different conductive liquid with platinum or gold coated electrodes? If the chamber is sealed well enough, you could use something other than metal with a high vapor pressure. This might help with the low voltage/high current problem.
This tecnique is used in eautomated lectronic solderingmachines. They are called „Selektivlötanlagen“ a fountain of liquid Solder is produced by putting a coil around a tube in wich the melted solder is. With this fountain you than can solder THT Parts on to a pcb without heating the hole pcb or also can have smd components on the bottem
The reason you had such poor efficiencies with the magnetohydrodyanamic reaction wheel was that your base plate was of an inferior pre-famulated amulite composition. You should also test out a lotus-o-delta configuration to minimize waneshaft side fumbling.
In modern designs the Waneshaft effect can easily be dampened or in some cases completely offset via application of a diff-limited Laminal Heteroshyte vector. This is usually achieved with nickel-copium angleshunt diodes.
So..could you forego having 3 reaction wheels and just make one spherical reaction wheel in a double shell configuration (with 3 magnet inputs at right angles)?
Use two magnets to center a magnets on the shaft of a rotor disk that stores the angular momentum. Then create a rotating magnetic field that inductively interacts with the roto to make an induction motor. You might try shaded pole, or two coils with one whose current phase is shifted with a capacitor, or you can go with a three phase variable frequency driver.
Fun fact with the Galinstan; you don’t need to use a hot water bath. As long as the gallium is touching the indium and tin, the amalgam will start to form. It gets very cold and the heat helps speed things up, but it’s neat watching it happen on its own. Also if you put a layer of sulfuric or hydrochloric acid on top of the small set up, it keeps the oxides from forming. 0.1M or less would work. I’ve seen some papers on HCl vapors being used but the dilute aqueous is easier to work with IMO. Love the channel. I took the SEM class for my electives because of your microscopy videos. There’s I made a SEM video for this channel.
A big sphere of liquid metal would make a cool reaction wheel, if you had good control over the magnetic field generation you could have a few rivers of metal intersect and sum their respective torques assuming vortices don't easily form from the intersection.
7:36 I'm pretty sure I have the exact same buck converter and Im pretty sure its limited to around 8-9A and the diode shining red it's limiting current so it wouldn't be pumping out 20-30A
Oh, damn! I went back and checked the listing, theoretically it's limited to 12A. I could have sworn it was 30A limit, although I have no idea where I got that idea 🙃. That actually means it was performing a lot better than I was expecting. That's what I get for not double checking before calculating and filming a video haha. Thanks for the correction, I'll add it to the pinned note!
I’m in my last semester of college and am taking a nano manufacturing class as an elective. It’s a lot of fun learning more about all the things I’ve been watching on your channel over the years!
its the left hand rule. left hand rule for motors, right hand rule for generator. The index finger points in direction of magnetic filed, middle finger in direction of current, while thumb is in direction of force.
I'm pretty sure satellites now use black diamond bearings since they are effectively wear proof, chemically inert, and do not need lubrication. The process to make these bearings is fascinating. Manufacturers have these large machines that squeeze a mold at the center of them with so much force and heat that the resulting product is a polycrystalin diamond, or a black diamond. I can't wait to see what else this manufacturing technique will be used for in the future!
A small issue I foresee is the interaction of the magnetic field the reaction wheel requires to function interacting with the magnetosphere. Field gradients aren't uniform and you'll end up generating interfering torque forces of varying intensity as long as the magnet is on and the vessel in motion. Magnetotorquers use this principle without the reaction wheel component to maintain stability and attitude control for satellites via the magnetic dipole alone.
For the battery solution, I had the same issues when trying to power a wifi router from a power bank. The solution I used was to add a NTC at the out put of the buck converter. I think the issue is the capacitors on the buck converter, they require large current as the charge up and work fine after charging.
Liquid metal bearings were used in xray tubes by Philips Medical after their patent around 1988-90. It was a game changer in extending the life of the tubes.
Man you are onto something with today tech focusing on making motion with minimal moving parts your idea certainly has a place I think you should keep exploring more ! All the best !
Carbon steel is suboptimal for magnetic conductor, The steel you want is called electric steel, it is either silicone steel (large domains of pure iron) or nickel iron alloy called permalloy. In this configuration I would simply used some electric steel from a transformer, such as microwave oven transformer.
About the corrosion you mentioned, there's actually already some pretty reliable passivation techniques used by the PC water cooling industry. Nickel-plating is considered a requirement when using liquid metal alloys as your thermal interface specifically because of the corrosion issues, and I'm sure there's some sort of plating concoctions that even your lab could work out to passivate against even more complex alloys like Gallinstan. I'm also sure with the various highly-precise machines and extreme cleanliness requirements of constructing things like satellites (especially the really high-end ones like JWST) could really push those coatings to last as long as possible, if not for longer than the life of the rest of the satellite.
I am wondering how big of an issue it is that these reaction wheels need to be actively heated? I am aware that satellites often have a heated partition for vulnerable electronics and measurement devices, but I believe pressurized air and some mechanical reaction wheels can operate at near space temperature, where this metal solidifies at relatively high temperatures...
I though i had worth mentioning is that if you already have a powerful electromagnet on board for a magnetorquer, the same hardware could do double duty as the field the electrified fluid pushes against for rotation
A friend of mine sadly gone, thought that if you have a torrid & magnetic coils that not only rotates the mercury round the inside of the torrid which will result it it being pushed to the middle of the inside wall, but also to rotate the mercury so whilst it pushes up against the wall it also tries to move it upwards. So two rotations, one round the torrid, the other up against the middle of the inner wall. I wondered if whether having "blades" on the outer inside wall might work like the paddles of a riverboat. Sorry for the bad wording. It's difficult to articulate this.
This is really awesome. I do wonder how viscous friction in the liquid metal compares to the rolling resistance of the bearings used in a traditional reaction wheel, my intuition is you would need a lot more power to maintain the velocity in the fluid. But then, intuition is moot when you don't have a fingertip feel for the dynamics of liquid metal. I wonder what the saturation condition is for this system too, will we ever see a video of a supersonic liquid metal jet escaping from one of these haha.
5:27 there is also solar flares that have caused arcing between the bearing races and balls depositing metal and prematurely causing bearing failure of reaction wheels
for powering this setup you can use big inductor around the machine and you can pickup from another inductor attached on your machine maybe not very efficient but it can deliver some good power
This is great. In college I tried to make a reaction wheel using ferrofluid and electro magnets, but had issues with sourcing supplies. I've always wanted to try again and see if I could produce a torque
I've got a couple of questions. First, why aren't you using mercury for your tests? It'd be far easier to work with and could potentially yield some rather extraordinary results. (Particularly in the field of electrogravitics.) Second, when I was growing up I had a BMX that had Gyro brakees on the front handlebars so the brake cables wouldn't get tangled. Could something like that be modified to work with electricity?
I have an interesting idea for a space project. Basically I'd like to fly FPV drones in space. These would be miniature spacecraft of course, since regular drones can't fly up there. But the purpose would be one for possible entertainment purposes such as space racing. Another is just to explore the possibilities of controlling miniature satellite-like spacecraft from earth.
Very goofy idea but it looked pretty fun to build. I wonder if better wall materials exist to speed up the flow. Mercury doesn't wet very well but there must be some super mercury-phobic material.
you should expand the small model to several concentric channels to create fine grain control and achieve higher mass for the reaction. each channel would be individually controlled and selectively cycled for small adjustmens or collectively cycled for large ones.
You should be able to make air bearings out of graphite relatively easily - Applied Science channel has a video about that. It's basically just machining a rough shape into a porous graphite block fixed in a 3d-printed mount with air supply and rubbing a steel ball bearing or whatever else you want to suspend to achieve precise matching of surfaces.
Point contact bearings would also solve all of these problems. In a brushless DC stator with shorted rotor configuration, I would think this is viable. It would be necessary to support the rotor during launch to avoid damaging the bearings. The density of mercury would be valuable in this application. A small series resistance would avoid the suppy's SC protection. (R is I to V converter...)
I'm very curious about the shape of the larger "wheel". Does the rounded square shape actually confer any advantages? It seems likes it could be losing some efficiency through having a higher radial acceleration through the tighter corners than a circle with the same outside dimensions. I haven't bothered to do the math so I suppose it's potentially not an issue, or at least not a big one.
Good eye! The research lab studying these things chose the rounded-square profile because it maximizes the effective diameter and amount of flowing metal that can fit in a cubesat shape. You're right though that the corners lose efficiency compared to a circular loop. So in real flight hardware you'd have to spend some time optimizing between a square (to maximize angular momentum) and a circle (to minimize fluid losses). There are a lot of other funky shapes I didnt talk about too, like L shapes (to get two axis in one loop) or stacked spirals (to get more angular momentum out of one device)
And we also did what we called crown-shaped FDAs for redundancy. The main reason for the rectangular-rectangular "toroid" was integration with CubeSat structures - you get a really bad volumetric efficiency when you try to combine a classical toroid with cuboids.
a quick way to tell whether an 18650 (or 21700) is protected or not is if it has a button top (the nub that makes it look like an oversized AA battery) or a flat top. button tops have the protection circuit underneath, flat tops are unprotected cells
@4:15 it ought to be fairly simple to calculate the resonance of the mass and wire, and drive the reaction wheel with a h-bridge circuit. @9:05 you could also pulse modulate the circuit to reduce the average current consumption. !!! Excellent work ❤
I am so happy this video posted because just this morning I was having a thought: Is it possible to exploit a hydrophobic surface to create a tiny, self centering and lubricating.... something? I have no idea what it could be used for but here is my thought: Get two cylinders, one just small enough to fit in the other while leaving a tiny gap for water. On the outside of the cylinder that fits on the inside, paint line from top to bottom in a hydrophobic layer, maybe measure out like 16 stripes? Not sure what the angle in degrees each would be off the top of my head. But, in my head, the hydrophobic surface will create like some tiny 'air shocks' to counter against the water sticky pull to drag the cylinders together? In my head I am picturing the example of dye in a similar set up but with corn syrup that shows viscosity of the liquid? Can spin it clockwise like 10 times then counter clockwise 10 times and watch the dye stretch out into thin smears then wind itself back into the starting blobs. I really hope someone reads this and can translate it from me monkey with rock talk to smart person speak.
The biggest problem I see with this this is that you would need constant current going in order to keep the thing spinning. The satellite only rotates while the wheel *changes* it's speed so over time velocity on the wheels builds up, and the fact that the wheels act as flywheels and keep spinning forever with (almost) no power loss is critical. In order to get this to work you'd need superconductors to keep the electric current going continuously forever, or use some of your power budget to maintain flow rates. Eventually satellites will do a wheel 'desaturation' burn where they use thrusters to counteract the spin as they spin down the wheels.
While I can see that with no bearings the reliability is much better. As long as the liquid metal doesn't freeze or leak out, should last a very long time. But what about power consumption? Traditional wheels can be spun up and they keep spinning without constant power input, right? The turbulence/ viscous nature of the liquid metal might mean you have to supply a continuous current to avoid having it just stop very quickly after shutting off.
Yeah the reason why flywheels work is because the wheel itself carries a lot of momentum. To achieve anything remotely similar means you need to have a lot of Liquid Metal, and then you're dealing with fluid motion inefficiencies. It's a great thought exercise, though not likely practical.
Multiple off-set electromagnets, even/uneven spaced ones & one or more submerged (in the liquid metal) vessels with little fins on them, might be a fun experiment to try. Should give an effect more like that of the vibration modules in mobile phones. Caveat: vibrations are horrible for telescopes, but can be tuned... the little finned vessels may impart better sudden stops, which can be beneficial.
Well done on the experiments and great job explaining it! I especially appreciate explaining the back iron - you rarely see those used let alone explained in most videos. You have a nice voice and are good at explaining things with just the right balance of simplicity and completeness.
the big issue with this design is obvious : the viscosity, internal turbulence create friction, but reaction wheel don't work with impulses, they continuously turn when not affecting the moment of the object. The low friction is necessary to be useful. This design will succeed in affecting the moment but fail at not affecting the moment when powered down/standby mode. It means that they continuously require power, way more power than conventional reaction wheel when standing by. They do present an advantage in terms of bearing life span, because they use none, but since bearing tech is mature and pretty good, it's a slim advantage. They would be useful for extreme lifespan satellites or craft that have continuous and generous power.
bro ive been thinking of this since watching 911 conspiracy videos on alien scientist channel about the na sea bell proj and his theories about a mercury powered antigravitiy device lol. idk this was like a blast form the past of old school fringe yt channel but actually well recorded and presented and edited. thanks
@@ExtantFrodo2 That could have been from the loneliness alone. Can you imagine months at a time in a solitude? I dont think I'd stay sane even with no mercury...
I've not heard of Galllinstan being used in space, but mercury is used in space in the opposite mechanical arrangement, where a liquid metal is used to measure angular momentum, rather than generate angular momentum. Generally liquid metal is a bad idea in a spacecraft with no humans to maintain it anywhere nearby.
You forgot about two inventions that really help you... One a compass 🧭 with a needle that floats on cork as a simple flotation device (but you only need the floating (almost frictionless ideal) for showcasing your next "rotation" effects of the gyroscopic force... You'll be able to show us how much power each model you make has/had... And TWO... a battery so it doesn't have wires messing the whole thing up :)
Consider trying the same concept but with a mganetc pumb and thick ferro fluid. that set up also has no moving parts except the ferro fluild which is mover by a series of solenoids
The oxide skin could be a problem, not just for seeing the liquid, but bc of the increased surface tension slowing flow at the very top of the metal layer. I'd try filling it in a vacuum glovebox, though redesigning it for air-proofing could be difficult. edit: I actually just found out that in the new nvidia 5090 founder's edition card, they use liquid metal for thermal transfer, and it is sealed in a triple-layer hermetic seal not only to keep it in place, but to prevent oxides from forming.
I thought of a way to transfer axis rotation to a different axis using these liquid metal wheels with the housing of the liquid metal varying it's surface friction to apply force to the housing instead of at the magnet. It would allow force to be applied in one direction with the effect adding up over time. Tada, transportation between star systems
Man i hate that i didnt htink of these as reaction wheels in terms of satilites. Years ago when I was in college I built almost exactly the motor you built here. i use to sit in the engineering lab and use the laser cutters for personal projects and I would see people screw around with stuff all the time. one day someone came in working on what was essentially a balancing cube. he didn't want to use a standard fly wheel and we were all brain storming on different ideas. having just build the little motor I suggested that as a spitball idea (I knew it wouldn't work considering the weight and lack of actual torq required), but that's how brain storming works. Absoloutley love this video!
Dude I've been working on something similar for 2 years now, just mine has a second phase that uses suspended particles to generate voltage. Cool to see something in action!
I like this idea there is nothing to wear out technically. But it has really big drawbacks. First would be fricton the fluid still needs to slide over surfaces and that causes energy loss. Normal reaction wheel when you remove power it will keep applying torque as it will keep spinning for hours as ceramic bearings have super low fricton. This means that you really only need to apply power to change reaction wheels torque and tiny amount of power to compensate for parasitic loss, meanwhile for liquid metal reaction wheel you need to apply lot of power constantly to keep metal spinning at constant speed due to it's massive parasitic losses. Maybe the technology will get better but for now the ceramic bearings fixed pitting problems that liquid metal reaction wheel is not really worth it at this time.
btw instead of a permanent magnet why not use an electromagnet, then it can be switched off to avoid ongoing field toque with external fields, plus when running it can be intermittently reversed at the same time as reversing the electrode voltage to provide torque in the same direction while reversing & on average cancelling the active state external field coupling. Plus for the liquid metal would NaK be a more efficient fluid?
Afik some do, but you need supports for launch, backup bearings in case of power or system failure, the control system is pretty complex, and the low stiffness of magnetic bearings probably puts a strict limit on the maximum turn rate.
Also you want satellites to not have any intrinsic magnetic field at all, since that would interact with earth's magnetic field. The most simple way is to not use any permanent magnets on board, or be super careful about their placement. I imagine the rotating magnets of such a bearing would be extra hard to design with.
I had a similar nerd snipe moment when I was asked about if it would be possible to correct for the angular incidence of sunlight and make some weird perfect reflection. Needless to say, I now have lots of weird wooden jigs around the house and piles of broken mirrors from overtorque'd attempts to make minute adjustments to curving mirror surfaces.
Fascinating concept! High current RC drone batteries would be ideal for this application. They can pull in the hundreds of amps and have no inbuilt protection circuitry.
I've had problems with SMPS and batteries. My issue was that the even though the average current was below the max current the battery could supply, the SMPS would draw current in bursts, and the peak current of those bursts was tripping the PTC in the battery.
- Power calcs
I'm bad at everything and misread the current limit of the buck converter. It's limited to 12A, so power efficiency was notably better than I mentioned (albeit still pretty bad). Oops 🙃
- Temperature!
Sorry, I should have mentioned this: galinstan has a freezing point around -18 to -20C, which is near the lowest temperature experienced in LEO orbit. So it _might_ need a small heater to keep liquid, but maybe not depending on how much heat is retained from the sunwards pass of the orbit. The liquid metal will have a fairly large amount of thermal mass, and contained in an insulating polymer of some type. Also will experience some self-heating from joule losses during operation. Missions farther from the sun would need a heater to stay liquid however.
wouldn't mercury work better for space?
Thanks, was the first thing that came to mind
How did the Berlin team address this problem?
@@fluffy_tail4365 Kinda sorta. Won't eat the spacecraft, and lower freezing point which is great. But toxic so no one really wants to work with it if possible. It also ends up being not quite as efficient because it's less dense and less conductive (although lower viscosity too). Performance depends a lot on the specific geometry and magnet configuration, but mercury as a rough rule tends to be less efficient. But to be honest, I haven't seen many papers looking at it due to the toxicity problem so possible it could be optimized to run better... and just no one has because no one experiments with it.
@@BreakingTapsmercury (13.55g/cm³) is much denser than galinstan (6.44g/cm³)
Satellite reaction wheel bearings used to fail all the time, but the problem was solved in recent years. They've stopped using metal ball bearings and now use ceramic ones instead. The ceramic ones require no lubricants and are not susceptible to static electricity surface pitting slowly destroying them.
There's also the phenomenon of cold welding in the vacuum of space.
I don't know enough about the clearances and lubricants but I'm sure the people who create satellite gyros do.
Just wondering if it might be a contributing factor.
@@jimurrata6785it *is* a contributing factor but the pitting was significantly worse than initially theorized so thats why they *needed* to change rather than wanted to
@ker6349 well, ceramic bearings will certainly eliminate the static charge problem and are free to roll if not constantly subject to shock loads.
It's a great solution for expensive assets that you cannot maintain. 👍
Are there any downsides to ceramic bearings? Do they have a lifespan or are they functionally "unbreakable" since they will never wear out before a satellite's end of life?
I've also heard of liquid metal bearings. They have really low friction once they are warmed up and spinning and run incredibly true.
Their service life is so long, that the manufacturer deliberately went on to design them so they fail sooner.
I love this! However, it definitely requires mercury. I understand reluctance to use it, but Galinstan is neither particularly dense nor conductive, and its melting point is significantly higher than that of a mercury alloy.
As expected, the mercury king is already here!
Ionic liquids? Anyway I've never heard of this technology until now and it strikes me as being very important! Not merely because of the lack of bearing and wear or absence of vibration, but because, unless I am missing something important, it also obviates the need for magnetorquers such as on the Hubble to "push" against the earth's magnetosphere so that the reaction wheels can be periodically "desaturated" when they reach their rpm limit. It also eliminates the need for probes in deep space to burn fuel in the reaction control system to do the same thing! It seems to me this could significantly reduce the fuel consumption of long duration deep space missions.
So if you were to test this, what about floating it on a sea of mercury rather than an air bearing?
@@HighOnTacos :D but it will have same issues as airbag but instead you will get spilled mercury everywhere when your balance fails xD
We use mercury, here in Tartary...
You can find unprotected 18650 cells for sale at almost every local vape shop as well. 99% of "vape" batteries dont have protection because the mod or vape has those built into it by law, so it allows those cells to offer much higher constant current draw and many times that in peak current
Yup. 18650 cells do not have any kind of smarts. They are used in a wide range of applications, and all of those have their own protection. I was mildly surprised when I found them in a burglar alarm, as a back-up, in case mains were cut.
LiPo packs for RC also don't have protection circuits and they are designed for very high currents. Even the small ones I use in my combat robots which would fit perfectly on there can handle up to 40A.
@@andersjjensen some do actually come with an additional protection cirtuit spot welded to the top.
Yeah, I made the same suggestion.
@@andersjjensensome 18650 are protected(but arent really 18650s anymore since that is the dimensions). But you can usually tear off the wrapping to find the unprotected cell inside, then you can tear off the protection circuit. I did this once since my local electronics store only sold the protected type
9:05 The protection circuit is located at one end of the battery, under the insulation. It can safely be removed. I recommend getting high output cells, such as the ones used in power tool batteries.
Aha, thanks!
@@BreakingTaps If this is of use to you, I'm glad to have helped!
@@BreakingTaps You can easily buy 18650 batteries without protection circuits, most are unprotected by default. Protected ones are usually explicitly labeled as such.
Rule of thumb: battery length ~65mm = unprotected, ~68mm = protected.
@@BreakingTaps if it has a button it will have a protection circuit, flat top cells are always unprotected. the protection circuit is a small pcb on the positive end normally encased in a shield of the button.
My recommendation is molicel. Very high current batteries, and very affordable. I usually do through 18650batterystore. They ship properly, and their customer service is great.
Glad to see you closing the magnetic circuit with the iron. Even if there isn't enough, you're making a static field so hysteresis isn't really a big deal. Too many people just don't bother understanding a flux circuit and just rely on the unreasonable strength of rare earth magnets to leave all kinds of magnetic performance untapped, or just double the magnetic material required.
I remember learning the formulas in university. Even just a millimeter air gap is such a huge reluctance! Lacking a back iron really wastes a lot of potential.
In addition to reaction wheels, satellites also use control moment gyroscopes, which are a similar device that uses conservation of angular momentum to control satellite orientation. The key differences in performance between the two are that reaction wheels are more power efficient for smaller satellites, and larger manoeuvres, and CMGs are more power efficient for larger satellites and smaller manoeuvres. CMGs also provide more torque for less power.
There's also a difference in how they operate. Reaction wheels change angular momentum by spinning up and down. CMGs change momentum by changing the wheel's axis of rotation without needing to change the angular velocity.
Reaction wheels have their place, because many satellites are very small. But for larger craft, like the International Space Station, they use CMGs. The ISS has four.
There's also the magnetotorquer, which uses a dipole to generate torque against the planet's magnetosphere. It requires detailed knowledge of the local magnetic field gradient to calculate the torque correctly but they are commonly used for satellite attitude control and stabilization.
CMGs also have a lot less jitter than RWAs. So your platform is more stable, great for deep space telescopes where you need accurate pointing for long periods.
Metal electrodes are doomed in this application sooner or later. Maybe try some graphite? Something not dissolving in metal (mercury?)
Yeah definitely. Refractory metals (tungsten in particular iirc) are fairly resistant, and some will "just" embrittle over time which might be fine once you're in an orbit and not experiencing much force. But definitely a concern for long missions. Considering the amount of current that's dumped through the electrodes something like graphite might be to inefficient unfortunately. Definitely possible with mercury, although you have to deal with the toxicity and it's not quite as efficient (density and conductivity).
There are other options too, like ionic liquids or just really salty fluids, but they lose suffer due to lower conductivity and less mass. Very neat space to explore though!
does induction work on liquid metal?
if so there would be no need for wires and no corrosion
Induction?
Set it up with a series wound field coil and run it off AC? Probably easier to find metal alloys that are innert in galistan.
@sammiller5509 It does! There are a bunch of pro/cons with different scaling losses, but yeah that approach can work. You can also run it as an AC conduction pump too (with both electric and magnetic field varying). Check out this paper (ntrs.nasa.gov/api/citations/20070022272/downloads/20070022272.pdf ) for more details, it's an interesting read!
We called'em mercury vortex engines. I've made them as far back as 30 years ago. I never thought about space application. I imagine friction makes conventional reaction wheels more energy efficient, than a fluid that is going to be continuously venting energy through heat.
Yeah my first reaction was that friction might be a serious efficiency loss here.
Makes me wonder how good could this work with teflon tube or another low friction material for flow region?
@ I used mercury. Technically, you could lower friction of it by raising temperature or using gas as long as the density can form a conductive path, like in CFL bulbs, but the tradeoff is that with less density, there's less reactive mass. I mean, it's not like people haven't already thought of all the options in this video and rejected them for a good reason. Math just doesn't work out. My idea was to use this sort of propulsion to build a completely silent submarine. Went to science fairs with my prototype build, but never won anything.
Heat build up would be less of a thing in space with proper shielding, no?
The Round Shape can be optimized by making the shape not round , rather a saucer UFO shaped oblong. Try that shape, then laminar flow of the liquid metal seems to flow better in that shape.
In theory, liquid metal could be sandwiched in UFO ships that spins at 50k RPS, to har that anti gravitic reaction.
Wait, did you essentially make a liquid metal railgun moshpit ?
That's one way to put it
more like a circlejerk
This kind of tech could be expanded into a tube around the planet using the torque of the moving metal to keep said tube from falling to the ground and you could climb that tube to get to space
Here's an idea: You can reverse these devises to be a rotation detector. Any rotation with induce a current in the motor (this is how people detect rotation)
How would this compare to laser ring gyroscopes or fiber laser gyroscopes?
@kellymoses8566 i don't know
@@kellymoses8566 different, that's all. I could see these being VERY low power too as they generate some current; oh and super super tiny..
@@seejjordanSmaller than a MEMS gyroscope?
@@kellymoses8566 very poorly. Resonating structure gyros have achieved unbelievably exquisite levels of sensitivity and precision and can now reach FEMTORADIANS per second levels of rotational measurement accuracy.
I remember when I was a kid (a very very long time a go) there was a short article in Pop Sci or some such (paper) magazine about some scientists making a mercury "gyroscope" using magneto hydrodynamics to push the mercury around a metal tube loop. They hinted it could be used in place of spinning wheel gyros with then poor reliable ball bearings. This was before even the 1st satellites. Your video dredged up that long forgotten memory. What goes around comes around.
One problem is the space temperature, liquid metal need to be in a range of warm "ambient" temperature to be liquid, this device will need another system, like an temperature control to work properly, but it can be a little tricky in space, when we speak about fuel or energy
Also, the current of the magnetic field, can provide the warm necessary to maintain it liquid, but, as you said, it need a great current to work
Thank you for the use of a back iron to redirect magmetic forces. I think a lot of people get discouraged from thier projects when they see the costs assoicated with the magnets per pound of strength and just give up then and there when really a lot of projects could be done will far lower costs by just using a back iron
Conversely, the relatively low cost and high availability of NdFeB magnets makes for a lot of silly project videos -- sure you can make a, say, windmill that makes a few dozen watts, but it's massive and full of air space and doesn't look anything like a commercial design! Why is that? Well, it's not much effort to go one more step and consider the back side of those magnets too!
If my memory serves me, Canon used air bearings on the laser assembly of their earlier color copiers. The model 100 I know had it, I believe the 1000 did as well. If you somehow come across one of these dead, might want to grab that assembly - it likely wouldn't work for this type of project, but be real fun for scattering a laser beam around.
Using Air bearings in the vacuum of outer space might not be a good idea.
@BigA1 air bearings are only used in this context to remove friction for testing the reaction motor. Definitely wouldn't work in space.
A flywheel would keep a constant attitude with no extra power but such a liquid reaction wheel would start counter rotating as soon as the power is cut due to fluid friction loses slowing the inner liquid. interesting concept nonetheless
He mentions that
At 10:00
@@hashbrown777 Would be interesting how efficiency scales with size. As of now frictionless seems to be achievable with magnetic bearing flywheels but reducing fluid friction to negligible seems unlikely. A liquid metal's properties don't seem to help in that either
Also note that reaction wheels have bearing friction and cogging torque which will slow down the flywheel too. No free lunch :)
For the galinstan corrosion risk there’s always the option of option in a, potentially vacuum rated, epoxy. Especially if it’s not necessary to service it as it’ll be unlikely to react to any galinstan that escapes it’s primary containment
Can you imagine a science fiction senario, where a galinstan module breaks in orbit, creating tiny orbiting particules of galinstan, forcing all future space craft to have nonmetallic shielding?
Damn, that would be a nightmare.
@Superwoodputtie whell there are already a lot of Natrium droplets from cold war atomic reactors circulating earth...
But you don't have to care about corrosion with 9km/s...
I think he meant regarding electrodes. I'm not sure about Ga, but In and Sn at least have well understood solubility against base metals, like Fe or Cu -- for which, it's a fairly low fraction, and once the solution reaches saturation, that's it. For example, Cu can be used _at temperature_ to hold liquid Sn, albeit with notable erosion (Sn diffuses somewhat into Cu, forming intermetallics (IMC); the IMC dissolves, putting free Cu into Sn solution). Or the solubility of Fe in Sn is very small indeed, hence its popularity for soldering iron tip plating. In I think is similar, but I'm not sure about Ga.
The tricky part is when these elements aren't quite soluble in the base metal, but diffuse into them anyway -- this is most apparent, and dangerous, for similar metals like Al and Ga, where the Ga diffuses into the crystal structure directly. I don't know the full extent/dynamics of this, but the diffusion rate of the elements through various alloys is the property you want to look for. I expect these data are available somewhere, well understood, and exotica aren't required.
If nothing else, a (particularly sturdy: pinholes and fractures are a bane!) plating of a hard metal like Cr, Ir or W very likely handles it, or various compounds like conductive oxides, carbides, sulfides, etc.
Other metals or alloys might be of interest; efficiency is directly related to resistivity, and I wonder which liquids are best. Mercury almost certainly is a poor choice, having much higher density and resistivity than most; NaK might be competitive with Galinstan. I wonder if molten Li might even be an option? Resistivity is probably worse at elevated temperatures though. (Incidentally, apparently Li and Na are immiscible? So there's no lighter analog to NaK.)
IT VIRKS!
I used to work with a guy who was convinced in his idea for a reaction motor for linear motion in space. Basically an arm with a counterweight spinning around an axis fixed to the spacecraft body. The arm could move in and out of the axle - the idea being that as the counterweight spins, it is retracted and extended along the desired axis of motion using a rack and pinion setup, "push/pulling" the spacecraft forward in space. Needless to say, he didn't understand the physics of conservation of momentum.
Hey, it worked in ksp!
There are some tether concepts using the gravitational gradient and something similar like you describe. Some where already tested in space, they work and do not contradict any laws. (see nasa tether propulsion handbook)
@@lukasskymuh5910 Nah, this guy's hypothesis was nothing nearly as subtle as that. It boils down to the notion that, say, if you're spinning a sledgehammer in space, you can pull yourself forward at 0 degrees and push yourself backwards at 180 degrees with a net forward linear motion with no external gravitational tethering.
@LilMissMurder3409 the weird thing is that such a simple mechanism this would actually increase your orbital in a gravitanional field with the correct timing. Unfortunately, not efficient enough to be interesting.
6:55 If the sides of the horseshoe shaped parts tapered out slightly, the parts could be stacked and very little sheet metal would go to waste.
What about using a different conductive liquid with platinum or gold coated electrodes? If the chamber is sealed well enough, you could use something other than metal with a high vapor pressure. This might help with the low voltage/high current problem.
11:08 Beesat-9 was a nice reminder of the recent Beesat-1 resurrection video (38c3).
Yes, sadly BEESAT-9 re-entered last year together with most of the other BEESATs.
You could add another capacitor along the line or reduce the wire loop area to ease the strain on the batteries
This tecnique is used in eautomated lectronic solderingmachines. They are called „Selektivlötanlagen“ a fountain of liquid Solder is produced by putting a coil around a tube in wich the melted solder is. With this fountain you than can solder THT Parts on to a pcb without heating the hole pcb or also can have smd components on the bottem
8:40 alternatively, you may consider placing the air bearing on top and hanging the payload table from it
The reason you had such poor efficiencies with the magnetohydrodyanamic reaction wheel was that your base plate was of an inferior pre-famulated amulite composition. You should also test out a lotus-o-delta configuration to minimize waneshaft side fumbling.
I’m gunna include this in a turbo encabulator joke.
Oh, doh! Totally forgot about the waneshaft side fumbling! 🤦 What a rookie mistake 😄
In modern designs the Waneshaft effect can easily be dampened or in some cases completely offset via application of a diff-limited Laminal Heteroshyte vector. This is usually achieved with nickel-copium angleshunt diodes.
I... I am scared. What are all these words?
It's times like these that I realise how much I have yet to learn.
@@thomascorbett709 Im no scientist but I think he is just making stuff up. lol
So..could you forego having 3 reaction wheels and just make one spherical reaction wheel in a double shell configuration (with 3 magnet inputs at right angles)?
You could! There are some papers looking at exactly that, although I think it's very much early R&D for that kind of actuator
This is researched currently at TU Berlin by my former team members.
Use two magnets to center a magnets on the shaft of a rotor disk that stores the angular momentum. Then create a rotating magnetic field that inductively interacts with the roto to make an induction motor. You might try shaded pole, or two coils with one whose current phase is shifted with a capacitor, or you can go with a three phase variable frequency driver.
Fun fact with the Galinstan; you don’t need to use a hot water bath. As long as the gallium is touching the indium and tin, the amalgam will start to form. It gets very cold and the heat helps speed things up, but it’s neat watching it happen on its own.
Also if you put a layer of sulfuric or hydrochloric acid on top of the small set up, it keeps the oxides from forming. 0.1M or less would work.
I’ve seen some papers on HCl vapors being used but the dilute aqueous is easier to work with IMO.
Love the channel. I took the SEM class for my electives because of your microscopy videos. There’s I made a SEM video for this channel.
I was actually lucky enough to operate TechnoSat for a while at the TU Berlin. To now see it being featured in one of your videos is just awesome 🤩
A big sphere of liquid metal would make a cool reaction wheel, if you had good control over the magnetic field generation you could have a few rivers of metal intersect and sum their respective torques assuming vortices don't easily form from the intersection.
The more fancy words the more fun it is!
Technobabble... Verbal SciFi Gold
can't believe you didn't show a screenshot of the red october when talking about the application with sea water :D
haha oh no! Was it in the movie? It's been literal ages since I've seen it 😅
@@BreakingTaps It's the entire reason everyone wants the Red October XD. "Engage the silent drive!"
7:36 I'm pretty sure I have the exact same buck converter and Im pretty sure its limited to around 8-9A and the diode shining red it's limiting current so it wouldn't be pumping out 20-30A
Oh, damn! I went back and checked the listing, theoretically it's limited to 12A. I could have sworn it was 30A limit, although I have no idea where I got that idea 🙃. That actually means it was performing a lot better than I was expecting. That's what I get for not double checking before calculating and filming a video haha. Thanks for the correction, I'll add it to the pinned note!
Isn't it "cold" up there? Depends on the sun a lot so the metal doesn't freeze
I’m in my last semester of college and am taking a nano manufacturing class as an elective. It’s a lot of fun learning more about all the things I’ve been watching on your channel over the years!
its the left hand rule. left hand rule for motors, right hand rule for generator. The index finger points in direction of magnetic filed, middle finger in direction of current, while thumb is in direction of force.
I'm pretty sure satellites now use black diamond bearings since they are effectively wear proof, chemically inert, and do not need lubrication. The process to make these bearings is fascinating. Manufacturers have these large machines that squeeze a mold at the center of them with so much force and heat that the resulting product is a polycrystalin diamond, or a black diamond. I can't wait to see what else this manufacturing technique will be used for in the future!
A small issue I foresee is the interaction of the magnetic field the reaction wheel requires to function interacting with the magnetosphere. Field gradients aren't uniform and you'll end up generating interfering torque forces of varying intensity as long as the magnet is on and the vessel in motion. Magnetotorquers use this principle without the reaction wheel component to maintain stability and attitude control for satellites via the magnetic dipole alone.
For the battery solution, I had the same issues when trying to power a wifi router from a power bank. The solution I used was to add a NTC at the out put of the buck converter. I think the issue is the capacitors on the buck converter, they require large current as the charge up and work fine after charging.
Liquid metal bearings were used in xray tubes by Philips Medical after their patent around 1988-90.
It was a game changer in extending the life of the tubes.
Man you are onto something with today tech focusing on making motion with minimal moving parts your idea certainly has a place I think you should keep exploring more !
All the best !
Carbon steel is suboptimal for magnetic conductor, The steel you want is called electric steel, it is either silicone steel (large domains of pure iron) or nickel iron alloy called permalloy.
In this configuration I would simply used some electric steel from a transformer, such as microwave oven transformer.
I like it that you aren't shouting hurray all the time but discuss possible roadblocks, too - well done!
About the corrosion you mentioned, there's actually already some pretty reliable passivation techniques used by the PC water cooling industry. Nickel-plating is considered a requirement when using liquid metal alloys as your thermal interface specifically because of the corrosion issues, and I'm sure there's some sort of plating concoctions that even your lab could work out to passivate against even more complex alloys like Gallinstan. I'm also sure with the various highly-precise machines and extreme cleanliness requirements of constructing things like satellites (especially the really high-end ones like JWST) could really push those coatings to last as long as possible, if not for longer than the life of the rest of the satellite.
I am wondering how big of an issue it is that these reaction wheels need to be actively heated? I am aware that satellites often have a heated partition for vulnerable electronics and measurement devices, but I believe pressurized air and some mechanical reaction wheels can operate at near space temperature, where this metal solidifies at relatively high temperatures...
The first electric motor ever was made by Michael Faraday with rotating mercury in 1821
I though i had worth mentioning is that if you already have a powerful electromagnet on board for a magnetorquer, the same hardware could do double duty as the field the electrified fluid pushes against for rotation
A friend of mine sadly gone, thought that if you have a torrid & magnetic coils that not only rotates the mercury round the inside of the torrid which will result it it being pushed to the middle of the inside wall, but also to rotate the mercury so whilst it pushes up against the wall it also tries to move it upwards. So two rotations, one round the torrid, the other up against the middle of the inner wall. I wondered if whether having "blades" on the outer inside wall might work like the paddles of a riverboat. Sorry for the bad wording. It's difficult to articulate this.
This is really awesome. I do wonder how viscous friction in the liquid metal compares to the rolling resistance of the bearings used in a traditional reaction wheel, my intuition is you would need a lot more power to maintain the velocity in the fluid. But then, intuition is moot when you don't have a fingertip feel for the dynamics of liquid metal. I wonder what the saturation condition is for this system too, will we ever see a video of a supersonic liquid metal jet escaping from one of these haha.
5:27 there is also solar flares that have caused arcing between the bearing races and balls depositing metal and prematurely causing bearing failure of reaction wheels
You beat me to it!
for powering this setup
you can use big inductor around the machine and you can pickup from another inductor attached on your machine
maybe not very efficient but it can deliver some good power
This is great. In college I tried to make a reaction wheel using ferrofluid and electro magnets, but had issues with sourcing supplies. I've always wanted to try again and see if I could produce a torque
For some reason I cannot stop imagining how you could incorporate an electromagnet into the design. There is probably a very clever way of doing it.
I've got a couple of questions. First, why aren't you using mercury for your tests? It'd be far easier to work with and could potentially yield some rather extraordinary results. (Particularly in the field of electrogravitics.) Second, when I was growing up I had a BMX that had Gyro brakees on the front handlebars so the brake cables wouldn't get tangled. Could something like that be modified to work with electricity?
I have an interesting idea for a space project. Basically I'd like to fly FPV drones in space. These would be miniature spacecraft of course, since regular drones can't fly up there. But the purpose would be one for possible entertainment purposes such as space racing. Another is just to explore the possibilities of controlling miniature satellite-like spacecraft from earth.
Very goofy idea but it looked pretty fun to build. I wonder if better wall materials exist to speed up the flow. Mercury doesn't wet very well but there must be some super mercury-phobic material.
you should expand the small model to several concentric channels to create fine grain control and achieve higher mass for the reaction. each channel would be individually controlled and selectively cycled for small adjustmens or collectively cycled for large ones.
You should be able to make air bearings out of graphite relatively easily - Applied Science channel has a video about that. It's basically just machining a rough shape into a porous graphite block fixed in a 3d-printed mount with air supply and rubbing a steel ball bearing or whatever else you want to suspend to achieve precise matching of surfaces.
Point contact bearings would also solve all of these problems. In a brushless DC stator with shorted rotor configuration, I would think this is viable. It would be necessary to support the rotor during launch to avoid damaging the bearings.
The density of mercury would be valuable in this application.
A small series resistance would avoid the suppy's SC protection. (R is I to V converter...)
You know what would be a great experiment... connect and IMU to the device to see if can hold a position. Great video. Thx for sharing.
I'm very curious about the shape of the larger "wheel". Does the rounded square shape actually confer any advantages? It seems likes it could be losing some efficiency through having a higher radial acceleration through the tighter corners than a circle with the same outside dimensions. I haven't bothered to do the math so I suppose it's potentially not an issue, or at least not a big one.
Good eye! The research lab studying these things chose the rounded-square profile because it maximizes the effective diameter and amount of flowing metal that can fit in a cubesat shape. You're right though that the corners lose efficiency compared to a circular loop. So in real flight hardware you'd have to spend some time optimizing between a square (to maximize angular momentum) and a circle (to minimize fluid losses). There are a lot of other funky shapes I didnt talk about too, like L shapes (to get two axis in one loop) or stacked spirals (to get more angular momentum out of one device)
And we also did what we called crown-shaped FDAs for redundancy.
The main reason for the rectangular-rectangular "toroid" was integration with CubeSat structures - you get a really bad volumetric efficiency when you try to combine a classical toroid with cuboids.
You could put two metal rings on the cylindrical bearing to act line a slip ring so it can rotate freely. This is still very cool.
a quick way to tell whether an 18650 (or 21700) is protected or not is if it has a button top (the nub that makes it look like an oversized AA battery) or a flat top. button tops have the protection circuit underneath, flat tops are unprotected cells
incredible! congratulations!
Cool to see this, I’ve been wanting to learn more about Liquid Metal pumping for salt based nuclear reactors. Different applications same technology
Nah, as an affectionate sci-fi writer I gotta do something with this concept before it completely takes off, it's super inspiring
You should apply force in orthogonal direction to the ring plane for getting reaction force. It's a gyroscope made of liquid metal.
@4:15 it ought to be fairly simple to calculate the resonance of the mass and wire, and drive the reaction wheel with a h-bridge circuit.
@9:05 you could also pulse modulate the circuit to reduce the average current consumption.
!!! Excellent work ❤
I am so happy this video posted because just this morning I was having a thought:
Is it possible to exploit a hydrophobic surface to create a tiny, self centering and lubricating.... something? I have no idea what it could be used for but here is my thought: Get two cylinders, one just small enough to fit in the other while leaving a tiny gap for water. On the outside of the cylinder that fits on the inside, paint line from top to bottom in a hydrophobic layer, maybe measure out like 16 stripes? Not sure what the angle in degrees each would be off the top of my head. But, in my head, the hydrophobic surface will create like some tiny 'air shocks' to counter against the water sticky pull to drag the cylinders together?
In my head I am picturing the example of dye in a similar set up but with corn syrup that shows viscosity of the liquid? Can spin it clockwise like 10 times then counter clockwise 10 times and watch the dye stretch out into thin smears then wind itself back into the starting blobs.
I really hope someone reads this and can translate it from me monkey with rock talk to smart person speak.
The biggest problem I see with this this is that you would need constant current going in order to keep the thing spinning. The satellite only rotates while the wheel *changes* it's speed so over time velocity on the wheels builds up, and the fact that the wheels act as flywheels and keep spinning forever with (almost) no power loss is critical. In order to get this to work you'd need superconductors to keep the electric current going continuously forever, or use some of your power budget to maintain flow rates. Eventually satellites will do a wheel 'desaturation' burn where they use thrusters to counteract the spin as they spin down the wheels.
While I can see that with no bearings the reliability is much better. As long as the liquid metal doesn't freeze or leak out, should last a very long time.
But what about power consumption? Traditional wheels can be spun up and they keep spinning without constant power input, right? The turbulence/ viscous nature of the liquid metal might mean you have to supply a continuous current to avoid having it just stop very quickly after shutting off.
Yeah the reason why flywheels work is because the wheel itself carries a lot of momentum. To achieve anything remotely similar means you need to have a lot of Liquid Metal, and then you're dealing with fluid motion inefficiencies. It's a great thought exercise, though not likely practical.
Multiple off-set electromagnets, even/uneven spaced ones & one or more submerged (in the liquid metal) vessels with little fins on them, might be a fun experiment to try. Should give an effect more like that of the vibration modules in mobile phones. Caveat: vibrations are horrible for telescopes, but can be tuned... the little finned vessels may impart better sudden stops, which can be beneficial.
Well done on the experiments and great job explaining it! I especially appreciate explaining the back iron - you rarely see those used let alone explained in most videos. You have a nice voice and are good at explaining things with just the right balance of simplicity and completeness.
the big issue with this design is obvious : the viscosity, internal turbulence create friction, but reaction wheel don't work with impulses, they continuously turn when not affecting the moment of the object. The low friction is necessary to be useful. This design will succeed in affecting the moment but fail at not affecting the moment when powered down/standby mode. It means that they continuously require power, way more power than conventional reaction wheel when standing by. They do present an advantage in terms of bearing life span, because they use none, but since bearing tech is mature and pretty good, it's a slim advantage. They would be useful for extreme lifespan satellites or craft that have continuous and generous power.
Brilliant work. I'm impressed you were able to do all this.
bro ive been thinking of this since watching 911 conspiracy videos on alien scientist channel about the na sea bell proj and his theories about a mercury powered antigravitiy device lol. idk this was like a blast form the past of old school fringe yt channel but actually well recorded and presented and edited. thanks
Add mineral oil on top of the liquid metal, or inject it below the oil to start without oxyde layer and reduce surface tension
You forgot a turbo encabulator. It's cancels the reluctance and fluxes so side fumbling and sinusoidal replaneration are effectively reduced
If you are bucking to a lower voltage by did you hook the batteries in series? Leave them at ~3.7V and hook them all in parallel.
This is the technology ancient light houses used to rotate the mirrors. Mercury was used for reaction and as float bearing.
No wonder the lighthouse keepers were mad.
@@ExtantFrodo2 That could have been from the loneliness alone. Can you imagine months at a time in a solitude? I dont think I'd stay sane even with no mercury...
@arp_catchall It's certainly not a job most people would choose, but there's always the exception.
I've not heard of Galllinstan being used in space, but mercury is used in space in the opposite mechanical arrangement, where a liquid metal is used to measure angular momentum, rather than generate angular momentum. Generally liquid metal is a bad idea in a spacecraft with no humans to maintain it anywhere nearby.
I imagine a simple toroid or donut shaped channel would be more efficient.
Could you remove the permanent magnet issue by using electromagnets?
You forgot about two inventions that really help you... One a compass 🧭 with a needle that floats on cork as a simple flotation device (but you only need the floating (almost frictionless ideal) for showcasing your next "rotation" effects of the gyroscopic force... You'll be able to show us how much power each model you make has/had... And TWO... a battery so it doesn't have wires messing the whole thing up :)
Consider trying the same concept but with a mganetc pumb and thick ferro fluid. that set up also has no moving parts except the ferro fluild which is mover by a series of solenoids
The oxide skin could be a problem, not just for seeing the liquid, but bc of the increased surface tension slowing flow at the very top of the metal layer. I'd try filling it in a vacuum glovebox, though redesigning it for air-proofing could be difficult.
edit: I actually just found out that in the new nvidia 5090 founder's edition card, they use liquid metal for thermal transfer, and it is sealed in a triple-layer hermetic seal not only to keep it in place, but to prevent oxides from forming.
generally you want flywheels large, but with these its a balance between moment of inertia and drag
I thought of a way to transfer axis rotation to a different axis using these liquid metal wheels with the housing of the liquid metal varying it's surface friction to apply force to the housing instead of at the magnet. It would allow force to be applied in one direction with the effect adding up over time. Tada, transportation between star systems
Man i hate that i didnt htink of these as reaction wheels in terms of satilites. Years ago when I was in college I built almost exactly the motor you built here. i use to sit in the engineering lab and use the laser cutters for personal projects and I would see people screw around with stuff all the time. one day someone came in working on what was essentially a balancing cube. he didn't want to use a standard fly wheel and we were all brain storming on different ideas. having just build the little motor I suggested that as a spitball idea (I knew it wouldn't work considering the weight and lack of actual torq required), but that's how brain storming works. Absoloutley love this video!
You could probably make a very rudimentary slip ring to eliminate the supply wire twisting problem if you want
Dude I've been working on something similar for 2 years now, just mine has a second phase that uses suspended particles to generate voltage. Cool to see something in action!
Serious potential in this concept...Great work
1:08 No they all look like X :D Great video - nice footage and narration :)
I like this idea there is nothing to wear out technically. But it has really big drawbacks.
First would be fricton the fluid still needs to slide over surfaces and that causes energy loss.
Normal reaction wheel when you remove power it will keep applying torque as it will keep spinning for hours as ceramic bearings have super low fricton.
This means that you really only need to apply power to change reaction wheels torque and tiny amount of power to compensate for parasitic loss, meanwhile for liquid metal reaction wheel you need to apply lot of power constantly to keep metal spinning at constant speed due to it's massive parasitic losses.
Maybe the technology will get better but for now the ceramic bearings fixed pitting problems that liquid metal reaction wheel is not really worth it at this time.
btw instead of a permanent magnet why not use an electromagnet, then it can be switched off to avoid ongoing field toque with external fields, plus when running it can be intermittently reversed at the same time as reversing the electrode voltage to provide torque in the same direction while reversing & on average cancelling the active state external field coupling. Plus for the liquid metal would NaK be a more efficient fluid?
Wonder why they don't use magnetic bearing for flywheels instead of ball bearings that would fail over time?
Afik some do, but you need supports for launch, backup bearings in case of power or system failure, the control system is pretty complex, and the low stiffness of magnetic bearings probably puts a strict limit on the maximum turn rate.
Also you want satellites to not have any intrinsic magnetic field at all, since that would interact with earth's magnetic field. The most simple way is to not use any permanent magnets on board, or be super careful about their placement.
I imagine the rotating magnets of such a bearing would be extra hard to design with.
i love this channel so much these projects are always so fascinating in a unique way each time
yea you got me thinking now. I love this sort of stuff. and have noticed a bit of an emergence in gyroscopic stabilization/manipulation solutions.
I had a similar nerd snipe moment when I was asked about if it would be possible to correct for the angular incidence of sunlight and make some weird perfect reflection.
Needless to say, I now have lots of weird wooden jigs around the house and piles of broken mirrors from overtorque'd attempts to make minute adjustments to curving mirror surfaces.
Fascinating concept! High current RC drone batteries would be ideal for this application. They can pull in the hundreds of amps and have no inbuilt protection circuitry.
I've had problems with SMPS and batteries. My issue was that the even though the average current was below the max current the battery could supply, the SMPS would draw current in bursts, and the peak current of those bursts was tripping the PTC in the battery.