Honestly, your channel is such a gold mine. This is one of the best demonstrations of the engineering process at work: you find the issue, you think of a solution, and then you tinker and optimize until you've achieved what you've set out for. You demonstrate this whole process so eloquently in your videos that it's inspiring
I especially liked the process of finding an issue, fixing it, begin happy it worked and then figuring out that it worked for completely different reason than what you thought you were doing.
Offsetting your carbon footprint is shooting yourself in the foot. CO2 is plant food, more of it better it is. Here where I live people of your age smoke marijuana as recreational pastime and as you know marijuana growers keep their greenhouse at 2000 PPM CO2 level, that is where marijuana does its best. Marijuana is a C3 photosynthesis plant, it was evolved when CO2 was at a high level compared to today's. Of all known plants on earth, C3 plants are 90%. CAM and C4-species constitute about ~10% and 1%, respectively. So reducing CO2 level as you propagandize will be damaging to our planet, because C3 photosynthesis plants are staving of CO2 at a level lower than 2000 PPM. I know that advertising on your channel is a source of income, but at what cost?
DemolitionRanch has a video on basically that. Can't post links on youtube without it getting flagged as spam, but he calls it a railgun in his title if you want to look it up. Its pretty cool.
Calculate how long it takes for the coil to reach maximum magnetic power, use this to determine how far the triggering hall effect sensor needs to be to optimize energizing the coils as the velocity increases. Also, when you lower the winding count, increase the wire size. Remember, wind count has an effect on the speed at which the magnetic field builds to maximum. I was involved in the original development of the brushless motors in today's electric planes and drones.
I built a mass driver as a science fair project back in the '80's, I got around the curve flattening by progressively spacing the driver coils further apart as they went down the track. My reasoning was like a transmission in a car the impulses then match the speed of the projectile as it accelerates.
I don't think increasing the spacing of later coils would help in Tom's contraption, since each coil is still the same size and still only has the same limited time to impart a kick. Increasing the spacing only makes the contraption take up more space and experience more track-friction losses.
@@gluino Yeah, but there's also delay in turning on the coils to full power, so I think that would be necessary to get sufficient higher and higher spacing for the sensors.
Since this is getting a little interest, I'll elaborate a bit.. mine powered all the electromagnets at once and then cut power to them in series when the projectile passed a photoresistor in a socket. It fired tape wrapped paperclips faster than the speed of sound! (it made a loud CRACK!)
@@Coneshot huh. Common in military scifi is a switch from gunpowder rifles to flechette rifles, firing 5mm or less darts electrically at supersonic speeds, packing up to several hundred rounds in the space of what is currently a standard capacity 30 round magazine. What you made could be considered an early prototype to such a device, with that supersonic speed and an extremely small projectile. Interesting :P
A way to know which reason is the most limiting would have been to make a regression, it the curve is close to an exponential it's mostly a back EMF reason, if the curve is close to a square root, it's mostly due to kinetic energy (harder and harder to increase for a given power).
You can improve even more replacin alluminium rod with plastic, a magnet sliding on metal tends to reduce the speed because of the inducted current generated by the movement.
A few ideas: - You don't have to reverse polarity to also push the sled. You can just use a second Hall sensor located after the coil, and a second winding on the coil wired up in inverted polarity, so when the second Hall sensor turns on the second coil, it will push. With the position of the sensors, I think it can be timed perfectly. Yes it is 2X the components (and thus cost and power draw), but also roughly 2x the force at each magnet, resulting in a speed increase. - With a lubricated sled (and the plastic part of it shaped aerodynamically), most of the drag on the sled would be eliminated. - I've also noticed on the high-speed footage that the coils lean in a bit when active. This is expected as per Newton's third law, but the coild might need to be mounted more sturdily.
the problem with double winding is that they would induct current into itself and thus lose much of their efficiency and magnetic field. and it would be easier to reverse the polarity than doing this. it isn't really complex to reverse it but it is at least double the wire and parts and the power draw is still an issue here. if the voltage drops this much even without the reverse, then it isn't logical to implement it because you would lose even more voltage if you have two sets of coils on at the same time and thus the power is even more reduced.
@@dovos8572 The coils wouldn't be on at the same time. The reverse coil would ideally only turn on when the other turns off. But yeah it probably wouldn't increase the speed enough to justify the added complexity.
you would always have two sets of coils on. the one that is pulling and the one that is pushing. and one set on is already enought for such a huge voltage drop. so the voltage drop would be way higher when you add another 2 coils that are on at the same time.´and thus the power behind all coils would drop drastically.
@@matekovacs2696 it doesn't matter. the reverse coil would induce a current in the forward coil, losing much of the energy generated, especially since the coils are coaxial.
The air gap is still ~40mm long. The other side of the electromagnet is open, so the magnetic flux has to flow to the next electromagnet via air. The overall reluctance is still quite high and can be greatly reduced by tying backs of electromagnets with an iron bar and using a magnet that partially overlaps 2 electromagnets.
Also by reducing the air gap? I used to wind rc brushless motoers. Getting the air gap as small as possible without interference was the best way to increase efficiency.
I would look at replacing each coil pair with a single coil and an iron yoke under / through the track. Probably use a sled with just 1 magnet instead of 2 side by side.
I love how this channel started as modelling planes and various flying ... things (some of them is hard to describe) but then evolved into trebuchets and now its about building a coilgun.
In weapon terminology what you made is a coil gun. Lots of people confuse rail guns and coil guns despite coil guns being significantly more complicated (due to the need for rapid switching as you are clearly aware)
Brilliant video Tom. Even when I cannot quite follow all the info you are giving. You still make it all clear enough to make it really interesting. Thank you.....
Offsetting your carbon footprint is shooting yourself in the foot. CO2 is plant food, more of it better it is. Here where I live people of your age smoke marijuana as recreational pastime and as you know marijuana growers keep their greenhouse at 2000 PPM CO2 level, that is where marijuana does its best. Marijuana is a C3 photosynthesis plant, it was evolved when CO2 was at a high level compared to today's. Of all known plants on earth, C3 plants are 90%. CAM and C4-species constitute about ~10% and 1%, respectively. So reducing CO2 level as you propagandize will be damaging to our planet, because C3 photosynthesis plants are staving of CO2 at a level lower than 2000 PPM. I know that advertising on your channel is a source of income, but at what cost?
@@gregpetrics5669 by reading your comment i died of laughter XDDDD because me as a czech guy who doesnt understand what means "shareholders" and doesnt know whats sarcasm even in czech language, this comment for no reason killed me because i have broken humor also and when something is nonsense for me then i laugh af
For those still confused about the difference between railguns and coilguns: Railguns use two rails and lots of electricity to launch a projectile with the Lorentz force. Coilguns have a projectile that goes through the center of a group of coils stacked like donuts, with the ones in front of the projectile attracting it and pulling it forward, and ideally, the ones behind it repelling it and pushing it forward. The magnetic launcher in the video uses the same principals as a coilgun, but with the way the coils are oriented, calling it a magnetic launcher would be more accurate since its primary purpose and design is for it to launch stuff from the sled rather than simply using the sled itself as a projectile like a standard coilgun would.
yeah but no ones ever heard of coil gun, everyone says railgun regardless of the precise arrangement of the magnets. Its like everyone says 'literally' to mean actually... but it really means 'as per the literature'
Currently taking physics 2 as a senior in high school. Gotta say, after doing an entire magnetism and electromagnetism unit and learning everything theoretically, it’s really cool to see these concepts be implemented in such an easily understandable real-world example!
I used to wind and assemble r/c brushless motors. As far as I can tell, they are pretty much identical to your linear motor. The fewer the winds (w/heavier wire) the faster the motor, but the lower the torque. What would happen if you had progressively fewer winds as the carrier went down the track? It would start off at low speed w/high torque, then go progressively to high speed with low torque. Much like gears in a car. The last few coils could be only a few winds (with heavy wire OR multiple strands in parallel) and this would really kick up the speed. Also, work to get the air gap between the coils and the magnet as small as possible. The magnetic attraction (reluctance) drops off as a square of the distance. In motor construction, this is the easiest way to increase efficiency. This is hard to do in mass production, but with a DIY project the effort really pays off. Reduce the gap by a power of two and get four times the output power!
3 роки тому+3
I actually was sure the video was going this route, was surprised to learn it didn't.
@@jebimasta4604 He did design a new sled to reduce this gap. To further reduce it he could reduce the thickness of the plastic between the coil and magnet
Love how the spreadsheet graph matches your test data. You might try flipping the sensors so they tell you when the sled is leaving the coil. Then play with successively larger distances between coils as many have commented. The spreadsheet will help maximize acceleration per coil prior to a rebuild. A fixed distance between coils locks in a maximum speed since pulse timing is critical as speed increases.
As others have pointed out, you need to either increase coil spacing with distance, or increase your control speed, which means hitting the coils harder down the line. One way to do that is to distribute the capacitor bank down the length of the gun, adding caps as you approach the end. Your existing system is running out of power as the projectile advances, when you want to be increasing power as it advances in a system with constant spacing. One way to advance the design is to simulate the system. There are SPICE models for coils and moving magnets that can help simulate this whole thing. There are also tricks to shunting flyback into the next coil…
You are so charming…and brilliant, 10:20: “…so I designed a mini 3D printable glider…”. And you are so calm, and patient, and joyful. I am constantly impressed. I hope you have a lot of children. 10:38 The Nikon camera shows it’s outstanding resolution, excellent depth of field (yes, there is a ton of light), and remarkable frame capture rate (slow motion), which is a powerful “message from our sponsor” with no words said! Love that!
I would like to say that you did an excellent job with clarity in this video. Your slow, even tone and stating things planely made it very easy to follow. Also, NOT throwing lots of numbers out. Many presenters think they need to say every number of the screen for clarity when this does the opposite. Stay with m/s or km/hr. Convert the number but leave them on the screen.
I do love projects like this and have a multi-stage coil gun that I've built myself to explore this fascinating area of physics. I do find your design to be quite interesting in general with having the idea of "un-rolling" a standard electric motor to make a linear motor. The way I had managed to maximize my projectile velocity was to move away from super capacitors and move towards high voltage caps with a much lower capacitance. Since P = IV or I^2*R, and capacitor potential energy is a function of both the capacitance and the voltage squared I found that operating with more voltage and a capacitor that drained very quickly with P also being equal to J/s I managed very high speeds with far less coils. Perhaps moving towards high voltage and lower capacitance could yield an even higher end velocity with your design. Oh and I found it quite interesting on how many people thought that your device was a "rail gun". While both your device and a rail gun are both linear motors, you system is far more similar to a coil gun since rail guns operate on quite different systems. I guess this may be a result of people hearing a term tossed around online without understanding what that device is. I hope to see more wonderful projects from you in the future.
A very good presentation that avoids the deepest of detail and still covers the salient calculations. Motors don't have small gaps just because... You can get monolithic H bridges to switch polarity etc.. It is still a far better mag rail than any I've made. Very, well done sir.
Yes of course. Maybe, maybe. It's mega cool! Look here: ! New: Apocalypse Coilgun, Fallout Coilgun ua-cam.com/video/G9-jKWPyMFo/v-deo.html See also: Portable Super Coilgun: ua-cam.com/video/rclLsQ9nyeg/v-deo.html&feature=youtu.be ua-cam.com/video/-mjDqp_oWZk/v-deo.html ua-cam.com/video/YLOYxsqvjw0/v-deo.html !Coilgun and Survival in the Zombie Apocalypse ua-cam.com/video/_9WUcGwmS5s/v-deo.html
Well done . . . includes formulas and concepts I've forgotten over the past 70 years. It would be interesting to see how the system described here scales up to the General Atomics EMALS system now in use on the USS Gerald R. Ford Carrier. EMALS has had an extended development period as the issues and weak points are now being explored in use on a ship. One significant difference between the system described here is that it uses uses capacitors and/or batteries for energy storage while the EMALS system uses flywheels. [Based on my career I would have thought EMALS would have used capacitors.] So far EMALS seems to be beating the steam catapults in use now although problems are still being discovered.
minus some hiccups the system seems to be in it's final stage, especially with the increased testing over the last year and a half while the ford was at sea almost continuously. During this timeframe they also broke the record for most sorties in a day.
You could also try launching steel, while it won’t have as much of a generated magnetic field, I feel as if it may give you a acceleration advantage because it’s not canceling itself out therefore allowing more coils(and unlimited amount of coils) throughout the launcher Also you can try dry graphite on your launcher which will make the friction very minimal on your sled(trick used from pinewood derby cars)
If you use steel, it doesn't generate its own magnetic field so the electromagnets would have to provide more power. Not only this, how would you detect and accurately time the coils as hall sensors are now worthless?
I think you should consider finding a way to conserve the inertia the sled has when it falls away. Essentially, the energy you put into moving the sled forward is energy the isn’t pushing the glider forward. If you designed the sled to “push off” of the plane as it falls away, perhaps with a spring or pair off magnets with like poles facing each other, it might result in better overall performance. If the sled falls straight down after leaving the rail, you’ll know it has transferred its inertia to the glider, in much the same way that water falls straight down after striking an optimized scoop in a turbine. I don’t know if the spring/opposed magnet idea would work, but I’m sure you’re ingenious enough to find SOME way to extract that last little bit of inertia.
Except for hydropower (Pelton) turbines, when the water comes to a stop after hitting the bucket, the bucket is moving at half the speed of the water jet. If the bucket is moving at the speed of the water jet (glider moving at the speed of the sled), there is no energy transfer. Like in his trebuchet videos, he would have to come up with a solution which can transfer the kinetic energy of the sled to the glider by slowing the sled down, which means IT can't be the one pushing on the glider, but instead some lever arm or such.
4:30 I did physics this year and used these a lot, but I've never heard them be called "SUVAT" I assume some call it that because those are the variables it takes? I think we just called them "equations for constant acceleration"
We just don't call them somehow special. We just calculate the integral of the acceleration to get the velocity. The benefit is, that the formula found like that is also true for not constant acceleration.
Tom, you're my hero! With strong affinity for the mechanical/physical realm, and unfortunately now quadriplegic, I live vicariously through your videos! So now I'm back in school for computer science/software engineering. You're a huge inspiration and I look forward to all your videos Please don't stop uploading.
2 thoughts: Instead of velocity vs distance graph, try using an energy/time graph. Kinetic energy is equal to velocity squared, so if energy increase is constant, then a velocity/time graph would look like sqrt(t). Using distance instead of time also skews things, and you switch to using time instead of distance for the battery graphs. My guess is that a good curve to have in energy/time graph is a straight line. I love your constant attempts to use super capacitors only to find out batteries work better. Definitely a consistent theme of the channel. I think we both just think super capacitors are so cool and want to find uses for them :) I wonder - capacitors definitely have faster response times than batteries when load changes, while batteries can sustain the load better. Their strengths and weaknesses compliment each other. Did you try connecting both the capacitors and the batteries in parallel?
That's a good idea. The standard technique on PCBs is to add decoupling capacitors as physically close to the sink of current (IC pins typically, but here would be the coils/FETs) as possible to reduce the loss associated with impedance from the relatively large distance back to the power supply/battery.
The string of 8 super-caps will have much higher ESR than 3 lithium-ion packs in parallel. Their only advantage here is superior charging speed. You can't simply state that "capacitors have faster response times" without considering the capacitors and their configuration. To increase the power output of the batteries with capacitors, I think you'd be looking at a wild amount of low-ESR capacitors in parallel. I reckon (by simulation) at least 500mF to make a noticeable difference to the first pulse in Tom's launcher.
first thought went over my head, but I noticed the same thing for thought bubble two. the combination of both super cap and battery might provide a further stable voltage and current demand
@@tboy32 decoupling capacitors are not used as a means to reduce power loss, they act as semi current source when the chip sinks more current(inductance of PCB trace resists this.). You are essentially inductors in this case no point using capacitors.
Many, many years ago I made one of these with my physics teacher at school, I was happy when we exceeded 5m/s and that's where it ended (along with the school year!). This thing is awesome and all props but more to the point, I finally have a video to show people to explain the principle!
As others have mentioned, kinetic energy goes up with the square of the velocity. Another way to increase power is to add a flux return path for the magnetic fields. There might be a solution in chaining linearly but, especially for this setup, I think just a chunk of iron in like a "C" shape, oriented Up, around the track, would get you ~3x improved magnetic performance. You'd wrap the iron with your coils and trigger it the same way. Iron is good for these low frequencies but higher frequencies need different core materials. It also makes the steel the "force carrier", as in what actually gets pushed on. It can be a lot easier to hold onto steel than to coils. Notice the individual wire of the foreground coil moving at 10:01 in the video. Eventually that wire will fail as copper fatigues/work hardens quickly...
Laminated electrical steel would be a good improvement over simple iron. It's a big reason why mains transformers in everyday devices can be so efficient.
From what i know, the coils need some time to energize fully, so it might be worth to put the activation sensor further back the rail, and then use software delay so you can adjust timing, and find sweet spot. This might be especially important for the last few coils because when the projectile is travelling with high speed, the coils might not have enough time to energize properly.
@@TheWondermittens the coils rely on the back emf (the voltage of the sled passing by) to limit the input power, connecting them to the power supply without the sled moving past would cook them very quickly. As he says in the video each coil would draw over 1000W which might cause some issues lol
What about the inductance of the coils? Current flowing through a coil can't ramp up infinitely fast. The more windings, the higher the inductance and the slower is the ramp up. This may be an additional or even the main reason why less turns give you a higher end velocity. At the end of the track the sled may be already too fast to allow the current through the coils to ramp up to maximum before the sled has passed the coil. So I'd suggest to measure the inductance and tune it by adjusting the number of windings in order to better match the current ramp up to the position and velocity of the sled.
Or, adjusting the distance between the hall sensor and the coil. That would give the later coils enough time for he current to ramp up. That said, if it was taking a long time for the current in the coils to ramp up, we wouldn't be seeing those sharp voltage drops.
@@robertbackhaus8911 He could have an Arduino electronically switch the magnets on/off based on the sled's last three known positions. That way you could have the magnets on before the sled was anywhere near the magnet. For that matter, he could have all the magnets on and switch them off mechanically as the sled passes. Also, isn't there a more novel way of measuring the position of the sled? I mean, the magnets on the sled must change the voltage or current through the coils as it passes, right? Electromagnets aren't my wheelhouse, so I don't know what I'm talking about.
In an electronically commuted washing machine they advance and extend the commutation as the speed increases in the spin cycle. This allows for the slow current build due to the inductance, and the inductance then helps keep the current flowing as the peak back EMF can exceed the commutation voltage.
The number of windings of the coil don't only affects to the ramp up inductance; Also to the Magnet Resilience, or demagnetized of the coil, what afects to the atraction / repulsion of permanent magnets of the proyectile. This takes a time; The more windings, higher the time what the coil remains magnetized (slower magnetizacion due to highter inductance, and greater resilience), and contrariwise.
This for me helps to solves the problem for electric powered flight, and so it is interesting to see you take up this method. I have seen gliders use this method on take off using a pulley line, which on reaching a certain height drops away, or aircraft carriers that use steam and magnetic rails.
use a prediction time for actuating a coil: use a estimated time of actuation of the coil instead of the sensor, use every sensor reading to adjust the prediction for the next coil, like an advance for a piston engine, if you tweak the activation time maybe you can squeeze a little more performace.
IMO this is something that should be added once more lucrative optimizations are made. Optimizations to the hardware design should come before spending a lot of time optimizing software, to avoid wasting time on hardware that doesn't ultimately match your performance expectations.
Very cool!! The switching topology can also make a big difference to linear accelerator performance. The rise & fall time of the magnetic field also affects performance, so on & off time have to be adjusted, the faster the projectile travels. Later stages benefit from less turns to decrease inductive reactance. 😊😊
Great design! I think having the plane have its wings deployed while being shot down the track might be contributing heavily to a performance hit. Having some sort of retaining mechanism on the wings or even a mechanism that tilts them from 0 lift/negligible drag to take off position with some sort of trigger on the end of the track or attached to the the sled dropping away might massively improve flight performance
The performance gains of such systems are never worth it. As an example flaps increase drag on full sized aircraft. Yet they set flaps to takeoff rather than deploying them right before rotation, even when the field is short and the plane needs all the performance it can get.
An intuitive way (without just plugging in kinematics) to think about why it _looks_ like each subsequent coil imparts less velocity is this: while they all accelerate by the same amount, the projectile is moving faster past later coils, and so accelerates for less time.
When I was in High School in the late 60s I took a course called Industrial Physics, One of the pieces of equipment we had was called an Air Track which allowed a small cart to float over an aluminum track. We experimented with linear motors until we blew up the electromagnetic coils by using four times the voltage and amperage the equipment was designed for, we also but a hole in one of the concrete block walls of the school .
Interesting project! Thanks @Tom Stanton - here is a project for you : research and calculate the amount of energy required to melt silicon+doping and grow and ingot. As I recall (worked at ARCO/Siemens Solar), growing a full 35kg crucible of silicon into a125mm diameter x 1.2m single crystal silicon ingot took about 36 hours non-stop. Each ingot, when sliced with a wire saw, yielded 1200-1500 wafers to be processed. Then tell us the carbon footprint. That would make roughly 1 ordinary house-worth of solar panels.
Greater acceleration per coil could be achieved by having steel "C" cores reduce the reluctance of each magnetic circuit. Using steel cores would increase the magnetic flux for a given current.
You could use supercaps and lithium batteries in parallel, just for the ultimate low ESR power supply :) This is a very fun setup to watch, I'd love another episode about it!
You may be able to increase the final velocity, by spacing the coils farther apart, further down the rail. This will make the launcher longer, but should also make it faster.
Great job! Have you tried to reduce the friction of the sled using an air cushion (like it is used with air hockey) or magnets? That might boost the top speed further! I'd also suspect that the wiring of the energy supply causes some of the voltage sags due to the back EMF of the wires - try to add some caps (ceramics if you can) in close electrical proximity to the individual coils. Another idea would be to keep the coils constantly on and just turn them off when the sled passes to keep the voltage at the starting level - I suspect you may see the inverse on the voltage profile though when the coils turn off so be careful and expect some overvoltage! Also, if you do that watch out for the current/power draw in your coils and the temperature
If the coils are already pulling over 1000w then the small amount of friction does not matter. It would be more beneficial to add a top guide to reduce the air gap between magnet and coil and to make the sled lighter. But that would also be only a marginal improvement. It would be best to use higher voltage and cool the coils with liquid nitrogen to lower resistance and prevent them from evaporating. Maybe run 4 big car batteries in series. Also have to get bigger mosfets. 10kw to the coils will put a smile on your face.
I like how all of your videos have to do with science and are getting people excited abought the science behind magnets. Great job on the video and keep up the great work.
Awesome work. Your intuition that the distance between the permanent magnet and electromagnet is spot on. I used to work on the design team for a company designing motors for the robotics industry. The gap was one (if not the most) important aspect to control. You've got me wanting to make one of these now! I may have missed it but if you are not already using N52 magnets, see if you can get some of those. Don't worry about the characters following the number, such as N52H. The (H) is to do with the maximum operating temperature.
Tom I find this very interesting, the thought that came to my mind is what happens if you increase the center to center distances of your coils say .125 mm with each coil. This would in effect change the winding timing, this is done on some EV motors to gain rpm for a given voltage.
Always amazed by your projects. I love optimization. In my own physics degree I designed something called a Zeeman Slower which basically the same thing but for atoms and does the opposite ie slows things down instead of speeding them up. I loved tweaking the design for optimum performance
I think the curve flattens out because you increase not speed but kinetic energy. And as we now it has power of two. And so you get the inversed function
It's a curve because it's velocity vs. distance, so even if it's constantly accelerating, it spends less time in each successive unit of distance and so accelerates less from coil to coil. If you plot velocity vs. time you'd have a straight line with constant acceleration.
Very cool and exactly the discussion I need for designing a mass driver for the moon. The cost is getting it built, the energy is there and the re-entry craft can be built with a 3D printer on the moon using basaltic rock melted to make basaltic thread to weave the re-entry craft with some simple ideas to slow the incoming packages so that they can be decelerated to land on any kilometer on earth.
Some precision here: The firing interval needs to be adapted with velocity. The time of travel from Hall sensor to coil is decreasing with projectile speed but the current increasing time of the coil is constant. By reducing the number of turn at the end coils, the rising time was shortened leading to better timing. The back EMF is present yes but I'm not sure whether it is really significant. You can unplug the last coil and do a back EMF measurement at that coil during launching to check it the voltage supply is enough. If it is EMF limited at the end, the design is very efficient and performing closed to optimal. About the curve fitting, it is better to do a linear fitting with a transformed signal (e.g. log() if exponential signal expected). Human perception of fitting is not perfect... You did a very decent job!
You can try using those black powdered metal ferrite cores. Using normal steel or iron impedes with the instant flux change, but those ferrite cores are pretty fast.
Love following your projects; super inspiring! I believe all your conclusions regarding the power delivered to the coils are in the ballpark, but the inductance of the coil will affect how the current is increasing from zero when turned on. In the end there are several trade-offs to be made when choosing number of windings, wire gauge and the geometry of your coils. Anyhow, very cool end result and a satisfying swooosh when launching :D
If you really really want to optimize it, you could match the coild with capacitors, creating an LC oscillator at each step, the resonant frequency tuned to the speed of the sled at each step so that the positive half-period is when the sled approaches, and the negative half is when it has passed the coil. This eliminates the need for second coils, and complicated polarity-switching circuitry, but lets the magnets pull and the also push the sled. But finding the right capacitors is tedious work, and probably past the point of diminishing returns.
@@matekovacs2696 And abouth the Magnet Resilience, from magnetized / demagnetized of the coil? It afects to the atraction / repulsion of permanent magnets of the proyectile. This takes a time! The more windings, higher the time what the coil remains magnetized (slower magnetizacion due to highter inductance, and greater resilience), and contrariwise. Also the magnetic permeability of the core can be optimiced, but to this 'modest' proyect it's a nice performance just like it's now
Reducing amount of turns in the windings also reduces inductivity of the coil. It means it will produce less force per amp of current. But also inductivity makes it longer for the coil to go full power. Just like electric motor with less windings will have higher KV (rpm at curtain voltage) you made your end coils to respond quicker. Which was right decision but for a different reason :)
Yes of course. Maybe, maybe. Look here: ! New: Apocalypse Coilgun, Fallout Coilgun ua-cam.com/video/G9-jKWPyMFo/v-deo.html See also: Portable Super Coilgun: ua-cam.com/video/rclLsQ9nyeg/v-deo.html&feature=youtu.be ua-cam.com/video/-mjDqp_oWZk/v-deo.html ua-cam.com/video/YLOYxsqvjw0/v-deo.html !Coilgun and Survival in the Zombie Apocalypse ua-cam.com/video/_9WUcGwmS5s/v-deo.html
You're an absolute lad of a gentleman for making a video even better than the last one! I put off watching this because I didn't want to be disappointed as the first one was so well done imo
I've seen some interesting schematics on such launchers back in the days. Try to experiment with different coils. IIrc it was recommended to use bigger coils at the start, with thicker wire, so projectile start faster.
As always, lovely storytelling and engineering Tom. Of course you used neodymium magnets right? What about removing the slay friction? Would love to see an optimising an optimised magnetic launcher! :D
I was wondering Tom, what’d happen if you factor the exposure time of the each magnet on the slid. More precisely, what if you increase exponentially the distance between the magnets without increasing the number of turns of the coil windings. I expect this leading to the same results without increasing the power of the magnets.
Pradyumn meet aircraft carrier. Zero to 150 mph in approximately 2 seconds. Pilots and many civilians do it every day. But yes, for most it would not be acceptable and there would be injuries and sickness.
@@willierants5880 You make a good point, and I'm not arguing with you, but can you tell us more please about "many civilians" being shot off steam cats?
@@eduardopupuconnot quite you may increase friction tho ..! The same principle as a commutator on the rota of a electrical motor. There will always be inefficiency related to movement, you could do the experiment in a vacuum that would remove drag. However the experiment is over complicated buy the use of too many electrical components, and much the same result would be achieved with switches, the advantage of switches is the coils and voltage could then be increased on a separate circuit.
If you want to increase the speed you're going to have to start increasing the Gap in between coils. Not opposing coils, but in between coil pairs. It's like adding a muzzle brake on a rifle. It's much easier to time. A 4-pack at the end in a slightly more eccentric Gap pattern gradually increasing to the end may allow you to time the coils. May also stabilize the sled.
Thanks that is a god idea. You can see very well in the voltage curve the picks due to the lower current consumption of the coils by the self-induction. There, the power and correspondingly the electronic field is small and reaches its maximum only in the voltage plateau. With the last magnets, this is only very short. Unless the MOSFET can drive even fewer windings and thus the inductance is lower, an extension of the periods is the best choice.
Hi Tom, GREAT project! I did not check through all the comments to see if someone may have already suggested this, but with respect to the idea of adding 2 magnets, instead of using two magnets of the same size as the one magnet, how about using two magnets of half the size each of the one magnet? If the magnets are saturated, then halving them does not necessarily reduce the magnetic field by a factor of two, but you will have reduced the mass by a factor of two. So by this means you might get get even more velocity. Keep on experimenting! :-)
How much does the friction of the slider effect the gain of velocity forward? If you could eliminate any friction due to movement, that might give you an increase in forward motion, but then the gain might not be worth the effort of eliminating the friction. :)
@@cuteanimegril7012 WD-40 is for cleaning not lubricating. The WD stands for "water displacer" the 40 stands for "formula 40". I'd go with a silicone or Teflon lube. I prefer a carbon lube, but conductive lubes and electronics don't mix.
This is awesome, It's worth trying to increase the distance between the magnets relative to the increase in velocity, so that each magnetic, no matter how fast the sled is travelling, has the same length of time focused on the sled.
Tom, first thanks for another interesting video. I was curious (cant recall if you mentioned this) but would changing the spacing (increase the gap) as you go along the launch track given it will be going quicker would that give them more chance to pull? Would you want bigger and stronger magnetic forces too the further along to keep up with the speed increase and impart more effect? Would be interesting to see the plots for various gap deltas.
I feel that your coils are too close together, and when the speed rises, they cannot push the magnet enough because it passes so quick to the next coil that it's turned off too early. Could it be that your electronics struggle to push current agains impedance of the coils when they need to switch too fast Can you calculate the efficiency of the system? Kinetic energy out / electric energy input It would be cool if you could take away half of the coils, or mounting them in a longer rail. Maybe less coils make it less powerfull but more efficient, few power losses, and getting more time to be accelerated would increase efficiency too. I know it wouldnt work for making a plane launcher, but what if you made around a tube to accelerate a ball bearing or some kind of projectile? The magnetic field would be stronger at the center, will it make it more efficient?
Very interesting and cool ideas here Tom. But when fast switching on coils, you gotta consider coil inductance and the way it's limiting di/dt (rate of current rise). And current fall! Best of luck with all your projects mate!
Tom, this is a cool project! I hesitate to offer a suggestion, because I only barely understand the principles ... but I keep wondering: would it help any to space the coils further apart as you go down the rail? I'm thinking that would give each coil a more equal amount of time to act on the shuttle as it accelerates. Again, I don't know enough to know whether that would make any difference ...
You could maybe combine the LiPo battery packs and capacitor banks in parallel to maximize current delivery capabilities. Also you could use the hall effect sensor of the 2nd coil to trigger reverse current on the 1 coil and so on. This would likely consume only a few high current mosfets per coil but also double your current load at any point which your power solution right now might not be able to handle
Some more ways you could improve this: • Batteries and Supercapacitors working together • active cooling (liquid-cooled sensors and coils) • better electrical conductors (how expensive is enamel-coated silver wire?) • stronger ferromagnetic cores for the coils (tiny rare-earth magnets) • coat that sled and rail in teflon (should vastly lower friction) • focused magnetic fields (via steel plates around the launcher to act as magnetic lenses; see Tech Ingredients) • finally build the complex polarity-switching thing • implement some optical sensors as well
Tom, your launcher is perfect, but you need to balance your plane. It flies up very fast at big angle, and slows down. If you balance your plane correctly (centre of mass need to be under front part of wing) plane will fly further.
Honestly, your channel is such a gold mine. This is one of the best demonstrations of the engineering process at work: you find the issue, you think of a solution, and then you tinker and optimize until you've achieved what you've set out for. You demonstrate this whole process so eloquently in your videos that it's inspiring
my thoughts exactly. the scientific approach is elegant and inspiring.
I especially liked the process of finding an issue, fixing it, begin happy it worked and then figuring out that it worked for completely different reason than what you thought you were doing.
But I'll bet it wasn't that easy.
I like that he has brought up excel for all the crap I've built I haven't used it since high school was a great reminder haha!!
M Sachin
I honestly find myself impressed half the time I'm watching.
That slow motion shot of the launch sled falling away from the glider looked fantastically elegant, awesome project!
Problem is the sled carries away a whole lot of the gained momentum, rather than it being transferred to the plane itself.
Looks like footage of a spacecraft launch sequence where it's separating, so cool.
Offsetting your carbon footprint is shooting yourself in the foot.
CO2 is plant food, more of it better it is.
Here where I live people of your age smoke marijuana as recreational pastime and as you know marijuana growers keep their greenhouse at 2000 PPM CO2 level, that is where marijuana does its best.
Marijuana is a C3 photosynthesis plant, it was evolved when CO2 was at a high level compared to today's.
Of all known plants on earth, C3 plants are 90%. CAM and C4-species constitute about ~10% and 1%, respectively.
So reducing CO2 level as you propagandize will be damaging to our planet, because C3 photosynthesis plants are staving of CO2 at a level lower than 2000 PPM.
I know that advertising on your channel is a source of income, but at what cost?
It’s all fun and games at the Nerf match until Tom shows up with a handheld version of this with an automatic trigger and backpack mounted battery
That would make for a great video..... cough cough hint hint......
I really love your DRehm FC, i've been trying to add more function such as PID Auto Tune
"I am heavy weapons guy, and this, is my weapon"
he actually in orbit prepping the Nerf Ion Cannon
DemolitionRanch has a video on basically that. Can't post links on youtube without it getting flagged as spam, but he calls it a railgun in his title if you want to look it up. Its pretty cool.
Wow the slow-mo of that glider taking off was glorious. And the release mechanism working exactly as intended!
it helped me visualize the concept of center of mass traveling in 2 dimensions
Calculate how long it takes for the coil to reach maximum magnetic power, use this to determine how far the triggering hall effect sensor needs to be to optimize energizing the coils as the velocity increases. Also, when you lower the winding count, increase the wire size. Remember, wind count has an effect on the speed at which the magnetic field builds to maximum.
I was involved in the original development of the brushless motors in today's electric planes and drones.
This needs more likes
Twinkle Twinkle Little Star
Power Equals I Squared R
I'll remember it forever because of this.
Thanks destin
👍
Ayo turbulent flow is the best flow aight
Thanks Destin from SmarterEveryDay.
I'll be sure to use this at the next opportunity. Thanks, Destin!
I built a mass driver as a science fair project back in the '80's, I got around the curve flattening by progressively spacing the driver coils further apart as they went down the track. My reasoning was like a transmission in a car the impulses then match the speed of the projectile as it accelerates.
I postulated this as soon as i saw this utube and before i read your missive
I don't think increasing the spacing of later coils would help in Tom's contraption, since each coil is still the same size and still only has the same limited time to impart a kick. Increasing the spacing only makes the contraption take up more space and experience more track-friction losses.
@@gluino Yeah, but there's also delay in turning on the coils to full power, so I think that would be necessary to get sufficient higher and higher spacing for the sensors.
Since this is getting a little interest, I'll elaborate a bit.. mine powered all the electromagnets at once and then cut power to them in series when the projectile passed a photoresistor in a socket. It fired tape wrapped paperclips faster than the speed of sound! (it made a loud CRACK!)
@@Coneshot huh. Common in military scifi is a switch from gunpowder rifles to flechette rifles, firing 5mm or less darts electrically at supersonic speeds, packing up to several hundred rounds in the space of what is currently a standard capacity 30 round magazine.
What you made could be considered an early prototype to such a device, with that supersonic speed and an extremely small projectile. Interesting :P
A way to know which reason is the most limiting would have been to make a regression, it the curve is close to an exponential it's mostly a back EMF reason, if the curve is close to a square root, it's mostly due to kinetic energy (harder and harder to increase for a given power).
Vous, ici!? :D
wow ça fait longtemps que j'ai pas vu INCROYABLE EXPERIENCE, peut être 8 ans.
@@nobodyelse-h6h après la 20 000eme itération de la batterie, ils se sont perdus dans la sauce
You could also use the same equations in reverse to get something that should ideally look like a line.
Don't know what he said, but he used big words so it must be right!
You can improve even more replacin alluminium rod with plastic, a magnet sliding on metal tends to reduce the speed because of the inducted current generated by the movement.
also the friction is there hence the rod must be of some smooth material, like glass with lubricants
Some lubricating oil would help a lot too
I love listening to the audio on the slow mo of the plane launch at the end. You can hear each individual pull or push of the magnet on the sled.
A few ideas:
- You don't have to reverse polarity to also push the sled. You can just use a second Hall sensor located after the coil, and a second winding on the coil wired up in inverted polarity, so when the second Hall sensor turns on the second coil, it will push. With the position of the sensors, I think it can be timed perfectly. Yes it is 2X the components (and thus cost and power draw), but also roughly 2x the force at each magnet, resulting in a speed increase.
- With a lubricated sled (and the plastic part of it shaped aerodynamically), most of the drag on the sled would be eliminated.
- I've also noticed on the high-speed footage that the coils lean in a bit when active. This is expected as per Newton's third law, but the coild might need to be mounted more sturdily.
Sounds like a challenge to me!
the problem with double winding is that they would induct current into itself and thus lose much of their efficiency and magnetic field. and it would be easier to reverse the polarity than doing this. it isn't really complex to reverse it but it is at least double the wire and parts and the power draw is still an issue here. if the voltage drops this much even without the reverse, then it isn't logical to implement it because you would lose even more voltage if you have two sets of coils on at the same time and thus the power is even more reduced.
@@dovos8572 The coils wouldn't be on at the same time. The reverse coil would ideally only turn on when the other turns off.
But yeah it probably wouldn't increase the speed enough to justify the added complexity.
you would always have two sets of coils on. the one that is pulling and the one that is pushing. and one set on is already enought for such a huge voltage drop. so the voltage drop would be way higher when you add another 2 coils that are on at the same time.´and thus the power behind all coils would drop drastically.
@@matekovacs2696 it doesn't matter. the reverse coil would induce a current in the forward coil, losing much of the energy generated, especially since the coils are coaxial.
The air gap is still ~40mm long. The other side of the electromagnet is open, so the magnetic flux has to flow to the next electromagnet via air. The overall reluctance is still quite high and can be greatly reduced by tying backs of electromagnets with an iron bar and using a magnet that partially overlaps 2 electromagnets.
Also by reducing the air gap? I used to wind rc brushless motoers. Getting the air gap as small as possible without interference was the best way to increase efficiency.
Yes. Completing the flux path would undoubtedly yield great benefits.
I would look at replacing each coil pair with a single coil and an iron yoke under / through the track.
Probably use a sled with just 1 magnet instead of 2 side by side.
or... a greater density of electromagnets
Introducing an iron flux guide would likely increase the inductance of the coil. Not sure, if it would still turn on fast enough.
I love how this channel started as modelling planes and various flying ... things (some of them is hard to describe) but then evolved into trebuchets and now its about building a coilgun.
cough cough rail gun cough cough
@@Snakeman-jr6xz it's not RAILGUN, he even mentioned it specifically in this video.
@@quint3ssent1a cough cough *joke goes woosh* cough cough
@@kindlin haha i’m going to say something that’s wrong and say it’s a joke. very funny.
@@quint3ssent1a Ты типо шутки не понимаешь?
In weapon terminology what you made is a coil gun. Lots of people confuse rail guns and coil guns despite coil guns being significantly more complicated (due to the need for rapid switching as you are clearly aware)
Brilliant video Tom. Even when I cannot quite follow all the info you are giving. You still make it all clear enough to make it really interesting.
Thank you.....
cool project! I've always been a fan of linear induction motors!
Technically isn't this a linear PMDC motor? Not induction
I prefer linear synchronous motors, those power most modern launch coasters.
Now just make it a rail gun lol
@@thescubersteve5214 Yes, Mary is right. This is more of a LSM.
This is not an induction motor.
Tom is a young man who gives me a positive feeling toward the future. We need to support people like him.
But how does it produce value for shareholders?!
Offsetting your carbon footprint is shooting yourself in the foot.
CO2 is plant food, more of it better it is.
Here where I live people of your age smoke marijuana as recreational pastime and as you know marijuana growers keep their greenhouse at 2000 PPM CO2 level, that is where marijuana does its best.
Marijuana is a C3 photosynthesis plant, it was evolved when CO2 was at a high level compared to today's.
Of all known plants on earth, C3 plants are 90%. CAM and C4-species constitute about ~10% and 1%, respectively.
So reducing CO2 level as you propagandize will be damaging to our planet, because C3 photosynthesis plants are staving of CO2 at a level lower than 2000 PPM.
I know that advertising on your channel is a source of income, but at what cost?
@Tom R wow really I never would have guessed 😐
These vids do a nice positive touch, which is obviously important.
@@gregpetrics5669 by reading your comment i died of laughter XDDDD because me as a czech guy who doesnt understand what means "shareholders" and doesnt know whats sarcasm even in czech language, this comment for no reason killed me because i have broken humor also and when something is nonsense for me then i laugh af
For those still confused about the difference between railguns and coilguns:
Railguns use two rails and lots of electricity to launch a projectile with the Lorentz force.
Coilguns have a projectile that goes through the center of a group of coils stacked like donuts, with the ones in front of the projectile attracting it and pulling it forward, and ideally, the ones behind it repelling it and pushing it forward.
The magnetic launcher in the video uses the same principals as a coilgun, but with the way the coils are oriented, calling it a magnetic launcher would be more accurate since its primary purpose and design is for it to launch stuff from the sled rather than simply using the sled itself as a projectile like a standard coilgun would.
Thank you for nicely explaining the concept
yeah but no ones ever heard of coil gun, everyone says railgun regardless of the precise arrangement of the magnets.
Its like everyone says 'literally' to mean actually... but it really means 'as per the literature'
@@hindugoat2302 ? A coilgun still isn't the same as a railgun tho.
@@hindugoat2302 Plenty of people say gauss rifle/pistol for coilguns, especially because of Fallout.
Currently taking physics 2 as a senior in high school. Gotta say, after doing an entire magnetism and electromagnetism unit and learning everything theoretically, it’s really cool to see these concepts be implemented in such an easily understandable real-world example!
I'm amazed by the perfect separation of sled and printed plane. Were there more iterations of it before it worked like this?
I used to wind and assemble r/c brushless motors. As far as I can tell, they are pretty much identical to your linear motor. The fewer the winds (w/heavier wire) the faster the motor, but the lower the torque. What would happen if you had progressively fewer winds as the carrier went down the track? It would start off at low speed w/high torque, then go progressively to high speed with low torque. Much like gears in a car. The last few coils could be only a few winds (with heavy wire OR multiple strands in parallel) and this would really kick up the speed. Also, work to get the air gap between the coils and the magnet as small as possible. The magnetic attraction (reluctance) drops off as a square of the distance. In motor construction, this is the easiest way to increase efficiency. This is hard to do in mass production, but with a DIY project the effort really pays off. Reduce the gap by a power of two and get four times the output power!
I actually was sure the video was going this route, was surprised to learn it didn't.
I'm a bit confused, is this not what he did? apart from increasing the wire size. He reduced the gap and the number of windings.
@@jebimasta4604 He did design a new sled to reduce this gap. To further reduce it he could reduce the thickness of the plastic between the coil and magnet
Love how the spreadsheet graph matches your test data. You might try flipping the sensors so they tell you when the sled is leaving the coil. Then play with successively larger distances between coils as many have commented. The spreadsheet will help maximize acceleration per coil prior to a rebuild. A fixed distance between coils locks in a maximum speed since pulse timing is critical as speed increases.
As others have pointed out, you need to either increase coil spacing with distance, or increase your control speed, which means hitting the coils harder down the line. One way to do that is to distribute the capacitor bank down the length of the gun, adding caps as you approach the end. Your existing system is running out of power as the projectile advances, when you want to be increasing power as it advances in a system with constant spacing.
One way to advance the design is to simulate the system. There are SPICE models for coils and moving magnets that can help simulate this whole thing. There are also tricks to shunting flyback into the next coil…
This guy used the word 'shunt' therefore he either knows his shit or is really bad at insulting people.
@@DCsk8rgoelz One or the other, yep.
something from ArcFlash Labs Maybe? xD
Not a single sentence that you wrote is correct. Not trying to make fun of you, but its not.
@@CookiePsycho The hallmark of the drive-by troll: "everything you said is wrong, but I'm not going to say where you're wrong."
I love watching the progression by tweaking certain factors. I also learned a lot as you walked us through the process. Keep up the great work!
You are so charming…and brilliant, 10:20: “…so I designed a mini 3D printable glider…”. And you are so calm, and patient, and joyful.
I am constantly impressed. I hope you have a lot of children.
10:38 The Nikon camera shows it’s outstanding resolution, excellent depth of field (yes, there is a ton of light), and remarkable frame capture rate (slow motion), which is a powerful “message from our sponsor” with no words said! Love that!
I feel like randomly seeing someone using a trebuchet in England wouldn't look that out of place 🤣🙏👍
I would like to say that you did an excellent job with clarity in this video. Your slow, even tone and stating things planely made it very easy to follow. Also, NOT throwing lots of numbers out. Many presenters think they need to say every number of the screen for clarity when this does the opposite. Stay with m/s or km/hr. Convert the number but leave them on the screen.
I do love projects like this and have a multi-stage coil gun that I've built myself to explore this fascinating area of physics. I do find your design to be quite interesting in general with having the idea of "un-rolling" a standard electric motor to make a linear motor. The way I had managed to maximize my projectile velocity was to move away from super capacitors and move towards high voltage caps with a much lower capacitance. Since P = IV or I^2*R, and capacitor potential energy is a function of both the capacitance and the voltage squared I found that operating with more voltage and a capacitor that drained very quickly with P also being equal to J/s I managed very high speeds with far less coils. Perhaps moving towards high voltage and lower capacitance could yield an even higher end velocity with your design.
Oh and I found it quite interesting on how many people thought that your device was a "rail gun". While both your device and a rail gun are both linear motors, you system is far more similar to a coil gun since rail guns operate on quite different systems. I guess this may be a result of people hearing a term tossed around online without understanding what that device is. I hope to see more wonderful projects from you in the future.
Totally excellent. Especially the slo-mo videography of the glider launches at the conclusion. Thanks!
A very good presentation that avoids the deepest of detail and still covers the salient calculations.
Motors don't have small gaps just because...
You can get monolithic H bridges to switch polarity etc..
It is still a far better mag rail than any I've made. Very, well done sir.
Yes of course. Maybe, maybe. It's mega cool! Look here:
! New: Apocalypse Coilgun, Fallout Coilgun
ua-cam.com/video/G9-jKWPyMFo/v-deo.html
See also: Portable Super Coilgun:
ua-cam.com/video/rclLsQ9nyeg/v-deo.html&feature=youtu.be
ua-cam.com/video/-mjDqp_oWZk/v-deo.html
ua-cam.com/video/YLOYxsqvjw0/v-deo.html
!Coilgun and Survival in the Zombie Apocalypse
ua-cam.com/video/_9WUcGwmS5s/v-deo.html
The way the sled detached from the glider was amazing!
Well done . . . includes formulas and concepts I've forgotten over the past 70 years. It would be interesting to see how the system described here scales up to the General Atomics EMALS system now in use on the USS Gerald R. Ford Carrier. EMALS has had an extended development period as the issues and weak points are now being explored in use on a ship. One significant difference between the system described here is that it uses uses capacitors and/or batteries for energy storage while the EMALS system uses flywheels. [Based on my career I would have thought EMALS would have used capacitors.] So far EMALS seems to be beating the steam catapults in use now although problems are still being discovered.
minus some hiccups the system seems to be in it's final stage, especially with the increased testing over the last year and a half while the ford was at sea almost continuously. During this timeframe they also broke the record for most sorties in a day.
You could also try launching steel, while it won’t have as much of a generated magnetic field, I feel as if it may give you a acceleration advantage because it’s not canceling itself out therefore allowing more coils(and unlimited amount of coils) throughout the launcher
Also you can try dry graphite on your launcher which will make the friction very minimal on your sled(trick used from pinewood derby cars)
If you use steel, it doesn't generate its own magnetic field so the electromagnets would have to provide more power. Not only this, how would you detect and accurately time the coils as hall sensors are now worthless?
I think you should consider finding a way to conserve the inertia the sled has when it falls away. Essentially, the energy you put into moving the sled forward is energy the isn’t pushing the glider forward. If you designed the sled to “push off” of the plane as it falls away, perhaps with a spring or pair off magnets with like poles facing each other, it might result in better overall performance. If the sled falls straight down after leaving the rail, you’ll know it has transferred its inertia to the glider, in much the same way that water falls straight down after striking an optimized scoop in a turbine. I don’t know if the spring/opposed magnet idea would work, but I’m sure you’re ingenious enough to find SOME way to extract that last little bit of inertia.
Except for hydropower (Pelton) turbines, when the water comes to a stop after hitting the bucket, the bucket is moving at half the speed of the water jet. If the bucket is moving at the speed of the water jet (glider moving at the speed of the sled), there is no energy transfer. Like in his trebuchet videos, he would have to come up with a solution which can transfer the kinetic energy of the sled to the glider by slowing the sled down, which means IT can't be the one pushing on the glider, but instead some lever arm or such.
I really enjoy it when you improve your projects, how an engineer thinks and solves problems.
4:30 I did physics this year and used these a lot, but I've never heard them be called "SUVAT"
I assume some call it that because those are the variables it takes?
I think we just called them "equations for constant acceleration"
I like SUVAT more though
TUVAS sounds more organic to me :)
In my 1960's Grammer School applied maths we called them the projectile formulae.
To me they’re called kinematics equations
We just don't call them somehow special. We just calculate the integral of the acceleration to get the velocity. The benefit is, that the formula found like that is also true for not constant acceleration.
Tom, you're my hero! With strong affinity for the mechanical/physical realm, and unfortunately now quadriplegic, I live vicariously through your videos! So now I'm back in school for computer science/software engineering. You're a huge inspiration and I look forward to all your videos Please don't stop uploading.
2 thoughts:
Instead of velocity vs distance graph, try using an energy/time graph. Kinetic energy is equal to velocity squared, so if energy increase is constant, then a velocity/time graph would look like sqrt(t). Using distance instead of time also skews things, and you switch to using time instead of distance for the battery graphs. My guess is that a good curve to have in energy/time graph is a straight line.
I love your constant attempts to use super capacitors only to find out batteries work better. Definitely a consistent theme of the channel. I think we both just think super capacitors are so cool and want to find uses for them :) I wonder - capacitors definitely have faster response times than batteries when load changes, while batteries can sustain the load better. Their strengths and weaknesses compliment each other. Did you try connecting both the capacitors and the batteries in parallel?
A Capacitor + Battery mix should perform better.
That's a good idea. The standard technique on PCBs is to add decoupling capacitors as physically close to the sink of current (IC pins typically, but here would be the coils/FETs) as possible to reduce the loss associated with impedance from the relatively large distance back to the power supply/battery.
The string of 8 super-caps will have much higher ESR than 3 lithium-ion packs in parallel. Their only advantage here is superior charging speed. You can't simply state that "capacitors have faster response times" without considering the capacitors and their configuration.
To increase the power output of the batteries with capacitors, I think you'd be looking at a wild amount of low-ESR capacitors in parallel. I reckon (by simulation) at least 500mF to make a noticeable difference to the first pulse in Tom's launcher.
first thought went over my head, but I noticed the same thing for thought bubble two. the combination of both super cap and battery might provide a further stable voltage and current demand
@@tboy32 decoupling capacitors are not used as a means to reduce power loss, they act as semi current source when the chip sinks more current(inductance of PCB trace resists this.). You are essentially inductors in this case no point using capacitors.
That transition between project and the sponsor/ad was very smooth ngl.
Many, many years ago I made one of these with my physics teacher at school, I was happy when we exceeded 5m/s and that's where it ended (along with the school year!). This thing is awesome and all props but more to the point, I finally have a video to show people to explain the principle!
As others have mentioned, kinetic energy goes up with the square of the velocity.
Another way to increase power is to add a flux return path for the magnetic fields. There might be a solution in chaining linearly but, especially for this setup, I think just a chunk of iron in like a "C" shape, oriented Up, around the track, would get you ~3x improved magnetic performance. You'd wrap the iron with your coils and trigger it the same way. Iron is good for these low frequencies but higher frequencies need different core materials. It also makes the steel the "force carrier", as in what actually gets pushed on. It can be a lot easier to hold onto steel than to coils. Notice the individual wire of the foreground coil moving at 10:01 in the video. Eventually that wire will fail as copper fatigues/work hardens quickly...
Laminated electrical steel would be a good improvement over simple iron. It's a big reason why mains transformers in everyday devices can be so efficient.
From what i know, the coils need some time to energize fully, so it might be worth to put the activation sensor further back the rail, and then use software delay so you can adjust timing, and find sweet spot. This might be especially important for the last few coils because when the projectile is travelling with high speed, the coils might not have enough time to energize properly.
Would it make sense to energize all of the coils at the beginning, and just turn them off as the sled passes by?
@@TheWondermittens my thought exactly!
@@TheWondermittensI think it would suck a lot of current, which battery would not be able go provide, and its voltage would go super low
@@TheKapicraftPL Maybe just the next two coils then, to give the second one more time to warm up.
@@TheWondermittens the coils rely on the back emf (the voltage of the sled passing by) to limit the input power, connecting them to the power supply without the sled moving past would cook them very quickly.
As he says in the video each coil would draw over 1000W which might cause some issues lol
What about the inductance of the coils? Current flowing through a coil can't ramp up infinitely fast. The more windings, the higher the inductance and the slower is the ramp up. This may be an additional or even the main reason why less turns give you a higher end velocity. At the end of the track the sled may be already too fast to allow the current through the coils to ramp up to maximum before the sled has passed the coil. So I'd suggest to measure the inductance and tune it by adjusting the number of windings in order to better match the current ramp up to the position and velocity of the sled.
Or, adjusting the distance between the hall sensor and the coil. That would give the later coils enough time for he current to ramp up. That said, if it was taking a long time for the current in the coils to ramp up, we wouldn't be seeing those sharp voltage drops.
@@robertbackhaus8911 He could have an Arduino electronically switch the magnets on/off based on the sled's last three known positions. That way you could have the magnets on before the sled was anywhere near the magnet.
For that matter, he could have all the magnets on and switch them off mechanically as the sled passes.
Also, isn't there a more novel way of measuring the position of the sled? I mean, the magnets on the sled must change the voltage or current through the coils as it passes, right? Electromagnets aren't my wheelhouse, so I don't know what I'm talking about.
In an electronically commuted washing machine they advance and extend the commutation as the speed increases in the spin cycle. This allows for the slow current build due to the inductance, and the inductance then helps keep the current flowing as the peak back EMF can exceed the commutation voltage.
@@RobertKreegier Arduino not be a bit slow for the requirements of the end of the coil-gauss accelerator?
The number of windings of the coil don't only affects to the ramp up inductance; Also to the Magnet Resilience, or demagnetized of the coil, what afects to the atraction / repulsion of permanent magnets of the proyectile. This takes a time; The more windings, higher the time what the coil remains magnetized (slower magnetizacion due to highter inductance, and greater resilience), and contrariwise.
This for me helps to solves the problem for electric powered flight, and so it is interesting to see you take up this method. I have seen gliders use this method on take off using a pulley line, which on reaching a certain height drops away, or aircraft carriers that use steam and magnetic rails.
The slow motion video of the glider launch and the magnet(s) dropping away - excellent!
That take off was amazing. Looked like a real hi tech launching system.
use a prediction time for actuating a coil: use a estimated time of actuation of the coil instead of the sensor, use every sensor reading to adjust the prediction for the next coil, like an advance for a piston engine, if you tweak the activation time maybe you can squeeze a little more performace.
IMO this is something that should be added once more lucrative optimizations are made. Optimizations to the hardware design should come before spending a lot of time optimizing software, to avoid wasting time on hardware that doesn't ultimately match your performance expectations.
I’d love to see the flipped polarity. Even if it does take some time to perfect, I’d be amazing to see you make that in the next video.
I want my son, and aspiring engineer, to watch the "real" world of engineering and problem solving: this is great stuff!
What a glorious feeling it must have been when you first saw a successful launch! Very impressive, looking forward to more!
Very cool!!
The switching topology can also make a big difference to linear accelerator performance.
The rise & fall time of the magnetic field also affects performance, so on & off time have to be adjusted, the faster the projectile travels.
Later stages benefit from less turns to decrease inductive reactance. 😊😊
Great design! I think having the plane have its wings deployed while being shot down the track might be contributing heavily to a performance hit. Having some sort of retaining mechanism on the wings or even a mechanism that tilts them from 0 lift/negligible drag to take off position with some sort of trigger on the end of the track or attached to the the sled dropping away might massively improve flight performance
The performance gains of such systems are never worth it.
As an example flaps increase drag on full sized aircraft. Yet they set flaps to takeoff rather than deploying them right before rotation, even when the field is short and the plane needs all the performance it can get.
An intuitive way (without just plugging in kinematics) to think about why it _looks_ like each subsequent coil imparts less velocity is this: while they all accelerate by the same amount, the projectile is moving faster past later coils, and so accelerates for less time.
When I was in High School in the late 60s I took a course called Industrial Physics, One of the pieces of equipment we had was called an Air Track which allowed a small cart to float over an aluminum track. We experimented with linear motors until we blew up the electromagnetic coils by using four times the voltage and amperage the equipment was designed for, we also but a hole in one of the concrete block walls of the school .
Interesting project! Thanks
@Tom Stanton - here is a project for you : research and calculate the amount of energy required to melt silicon+doping and grow and ingot. As I recall (worked at ARCO/Siemens Solar), growing a full 35kg crucible of silicon into a125mm diameter x 1.2m single crystal silicon ingot took about 36 hours non-stop. Each ingot, when sliced with a wire saw, yielded 1200-1500 wafers to be processed. Then tell us the carbon footprint. That would make roughly 1 ordinary house-worth of solar panels.
The electrical genius of this guy is deeply underappreciated.
Greater acceleration per coil could be achieved by having steel "C" cores reduce the reluctance of each magnetic circuit. Using steel cores would increase the magnetic flux for a given current.
but it would also drasticaly increase the inductance of the coils leading to slower current rise times, which have to be countered with higher voltage
You could use supercaps and lithium batteries in parallel, just for the ultimate low ESR power supply :)
This is a very fun setup to watch, I'd love another episode about it!
I'm surprised by how versatile excel spreadsheet actually is.
You may be able to increase the final velocity, by spacing the coils farther apart, further down the rail. This will make the launcher longer, but should also make it faster.
Great job!
Have you tried to reduce the friction of the sled using an air cushion (like it is used with air hockey) or magnets? That might boost the top speed further!
I'd also suspect that the wiring of the energy supply causes some of the voltage sags due to the back EMF of the wires - try to add some caps (ceramics if you can) in close electrical proximity to the individual coils. Another idea would be to keep the coils constantly on and just turn them off when the sled passes to keep the voltage at the starting level - I suspect you may see the inverse on the voltage profile though when the coils turn off so be careful and expect some overvoltage! Also, if you do that watch out for the current/power draw in your coils and the temperature
I mean if all else fails you can just apply lubricant like krytox to the sled or rails in order to reduce friction
If the coils are already pulling over 1000w then the small amount of friction does not matter. It would be more beneficial to add a top guide to reduce the air gap between magnet and coil and to make the sled lighter. But that would also be only a marginal improvement. It would be best to use higher voltage and cool the coils with liquid nitrogen to lower resistance and prevent them from evaporating. Maybe run 4 big car batteries in series. Also have to get bigger mosfets. 10kw to the coils will put a smile on your face.
I like how all of your videos have to do with science and are getting people excited abought the science behind magnets. Great job on the video and keep up the great work.
If you do run into a back EMF limitation as you keep improving the speed, you could try advancing the timing of the later coils.
The elegance of that little release mechanism was a thing of beauty.
Awesome work. Your intuition that the distance between the permanent magnet and electromagnet is spot on. I used to work on the design team for a company designing motors for the robotics industry. The gap was one (if not the most) important aspect to control. You've got me wanting to make one of these now! I may have missed it but if you are not already using N52 magnets, see if you can get some of those. Don't worry about the characters following the number, such as N52H. The (H) is to do with the maximum operating temperature.
Tom I find this very interesting, the thought that came to my mind is what happens if you increase the center to center distances of your coils say .125 mm with each coil. This would in effect change the winding timing, this is done on some EV motors to gain rpm for a given voltage.
A great velocity can be achieved; unimaginable to grasp if the material can hold its durability.
Always amazed by your projects. I love optimization. In my own physics degree I designed something called a Zeeman Slower which basically the same thing but for atoms and does the opposite ie slows things down instead of speeding them up. I loved tweaking the design for optimum performance
I think the curve flattens out because you increase not speed but kinetic energy. And as we now it has power of two. And so you get the inversed function
It's a curve because it's velocity vs. distance, so even if it's constantly accelerating, it spends less time in each successive unit of distance and so accelerates less from coil to coil. If you plot velocity vs. time you'd have a straight line with constant acceleration.
People like you are the future of this world
Very cool and exactly the discussion I need for designing a mass driver for the moon. The cost is getting it built, the energy is there and the re-entry craft can be built with a 3D printer on the moon using basaltic rock melted to make basaltic thread to weave the re-entry craft with some simple ideas to slow the incoming packages so that they can be decelerated to land on any kilometer on earth.
Some precision here:
The firing interval needs to be adapted with velocity.
The time of travel from Hall sensor to coil is decreasing with projectile speed but the current increasing time of the coil is constant. By reducing the number of turn at the end coils, the rising time was shortened leading to better timing.
The back EMF is present yes but I'm not sure whether it is really significant. You can unplug the last coil and do a back EMF measurement at that coil during launching to check it the voltage supply is enough.
If it is EMF limited at the end, the design is very efficient and performing closed to optimal.
About the curve fitting, it is better to do a linear fitting with a transformed signal (e.g. log() if exponential signal expected). Human perception of fitting is not perfect...
You did a very decent job!
You can try using those black powdered metal ferrite cores. Using normal steel or iron impedes with the instant flux change, but those ferrite cores are pretty fast.
Love following your projects; super inspiring! I believe all your conclusions regarding the power delivered to the coils are in the ballpark, but the inductance of the coil will affect how the current is increasing from zero when turned on. In the end there are several trade-offs to be made when choosing number of windings, wire gauge and the geometry of your coils. Anyhow, very cool end result and a satisfying swooosh when launching :D
If you really really want to optimize it, you could match the coild with capacitors, creating an LC oscillator at each step, the resonant frequency tuned to the speed of the sled at each step so that the positive half-period is when the sled approaches, and the negative half is when it has passed the coil. This eliminates the need for second coils, and complicated polarity-switching circuitry, but lets the magnets pull and the also push the sled. But finding the right capacitors is tedious work, and probably past the point of diminishing returns.
@@matekovacs2696 And abouth the Magnet Resilience, from magnetized / demagnetized of the coil? It afects to the atraction / repulsion of permanent magnets of the proyectile. This takes a time! The more windings, higher the time what the coil remains magnetized (slower magnetizacion due to highter inductance, and greater resilience), and contrariwise. Also the magnetic permeability of the core can be optimiced, but to this 'modest' proyect it's a nice performance just like it's now
Reducing amount of turns in the windings also reduces inductivity of the coil. It means it will produce less force per amp of current. But also inductivity makes it longer for the coil to go full power. Just like electric motor with less windings will have higher KV (rpm at curtain voltage) you made your end coils to respond quicker. Which was right decision but for a different reason :)
Yes of course. Maybe, maybe. Look here:
! New: Apocalypse Coilgun, Fallout Coilgun
ua-cam.com/video/G9-jKWPyMFo/v-deo.html
See also: Portable Super Coilgun:
ua-cam.com/video/rclLsQ9nyeg/v-deo.html&feature=youtu.be
ua-cam.com/video/-mjDqp_oWZk/v-deo.html
ua-cam.com/video/YLOYxsqvjw0/v-deo.html
!Coilgun and Survival in the Zombie Apocalypse
ua-cam.com/video/_9WUcGwmS5s/v-deo.html
You're an absolute lad of a gentleman for making a video even better than the last one! I put off watching this because I didn't want to be disappointed as the first one was so well done imo
Excellent video, Tom. Very interesting and clearly presented data that you've provided. Keep up the great work. Slow-mo was amazing.
Next - this, on a remote controlled aircraft carrier
I've seen some interesting schematics on such launchers back in the days. Try to experiment with different coils. IIrc it was recommended to use bigger coils at the start, with thicker wire, so projectile start faster.
Very interesting! It reminded me of the V1 launch footage, the way the carriage drops away.
I'm glad you weren't around when they designed the V1. Excellent stuff chap.
As always, lovely storytelling and engineering Tom. Of course you used neodymium magnets right? What about removing the slay friction? Would love to see an optimising an optimised magnetic launcher! :D
I was wondering Tom, what’d happen if you factor the exposure time of the each magnet on the slid. More precisely, what if you increase exponentially the distance between the magnets without increasing the number of turns of the coil windings.
I expect this leading to the same results without increasing the power of the magnets.
A very good concept but imagine the level of g force the passengers would experience during takeoffs if this model is implemented.
Pradyumn meet aircraft carrier. Zero to 150 mph in approximately 2 seconds. Pilots and many civilians do it every day. But yes, for most it would not be acceptable and there would be injuries and sickness.
@@willierants5880 And the rest of us would happily pay extra
the newest aircraft carriers already use a similar electromagnetic system to launch planes
@@willierants5880 You make a good point, and I'm not arguing with you, but can you tell us more please about "many civilians" being shot off steam cats?
Very true, this is a coilgun, not a railgun.
Use the projectile as a contact/switch to activate each coil as it progresses along the rail completing the coil circuits moves forward. Simples
contact switch would increase drag tho
@@eduardopupuconnot quite you may increase friction tho ..! The same principle as a commutator on the rota of a electrical motor. There will always be inefficiency related to movement, you could do the experiment in a vacuum that would remove drag.
However the experiment is over complicated buy the use of too many electrical components, and much the same result would be achieved with switches, the advantage of switches is the coils and voltage could then be increased on a separate circuit.
If you want to increase the speed you're going to have to start increasing the Gap in between coils. Not opposing coils, but in between coil pairs. It's like adding a muzzle brake on a rifle. It's much easier to time. A 4-pack at the end in a slightly more eccentric Gap pattern gradually increasing to the end may allow you to time the coils. May also stabilize the sled.
Thanks that is a god idea. You can see very well in the voltage curve the picks due to the lower current consumption of the coils by the self-induction. There, the power and correspondingly the electronic field is small and reaches its maximum only in the voltage plateau. With the last magnets, this is only very short. Unless the MOSFET can drive even fewer windings and thus the inductance is lower, an extension of the periods is the best choice.
I thought this too
@@sylvan186 pretty much a gauss rifle.
Hi Tom, GREAT project! I did not check through all the comments to see if someone may have already suggested this, but with respect to the idea of adding 2 magnets, instead of using two magnets of the same size as the one magnet, how about using two magnets of half the size each of the one magnet? If the magnets are saturated, then halving them does not necessarily reduce the magnetic field by a factor of two, but you will have reduced the mass by a factor of two. So by this means you might get get even more velocity. Keep on experimenting! :-)
How many times can you slice the magnet before the velocity goes down? I.e. what is the optimum amount of slices?
How much does the friction of the slider effect the gain of velocity forward? If you could eliminate any friction due to movement, that might give you an increase in forward motion, but then the gain might not be worth the effort of eliminating the friction. :)
I was thinking exactly the same thing. Where is the lube?
WD40: Allow me to introduce myself.
@@cuteanimegril7012 WD-40 is for cleaning not lubricating. The WD stands for "water displacer" the 40 stands for "formula 40".
I'd go with a silicone or Teflon lube. I prefer a carbon lube, but conductive lubes and electronics don't mix.
@@jtjames79 As a former mechanic I cringe when people even think of pulling out WD40. It's just not really good for anything.
What about graphite powder?
This is awesome, It's worth trying to increase the distance between the magnets relative to the increase in velocity, so that each magnetic, no matter how fast the sled is travelling, has the same length of time focused on the sled.
I've always wanted to work on a rail coil launcher and your videos are a 100% my inspiration !
Tom, first thanks for another interesting video. I was curious (cant recall if you mentioned this) but would changing the spacing (increase the gap) as you go along the launch track given it will be going quicker would that give them more chance to pull? Would you want bigger and stronger magnetic forces too the further along to keep up with the speed increase and impart more effect? Would be interesting to see the plots for various gap deltas.
I feel that your coils are too close together, and when the speed rises, they cannot push the magnet enough because it passes so quick to the next coil that it's turned off too early. Could it be that your electronics struggle to push current agains impedance of the coils when they need to switch too fast
Can you calculate the efficiency of the system? Kinetic energy out / electric energy input
It would be cool if you could take away half of the coils, or mounting them in a longer rail. Maybe less coils make it less powerfull but more efficient, few power losses, and getting more time to be accelerated would increase efficiency too.
I know it wouldnt work for making a plane launcher, but what if you made around a tube to accelerate a ball bearing or some kind of projectile? The magnetic field would be stronger at the center, will it make it more efficient?
There's a account on tiktok stealing your content and posting as theirs.
Ok
The take-off in slow motion - BEAUTY!
Very interesting and cool ideas here Tom. But when fast switching on coils, you gotta consider coil inductance and the way it's limiting di/dt (rate of current rise). And current fall! Best of luck with all your projects mate!
Tom, this is a cool project! I hesitate to offer a suggestion, because I only barely understand the principles ... but I keep wondering: would it help any to space the coils further apart as you go down the rail? I'm thinking that would give each coil a more equal amount of time to act on the shuttle as it accelerates. Again, I don't know enough to know whether that would make any difference ...
++
Ok the first time in this early...but tom i just absolutely love ur vids and u r my role model....wanna build stuff with u some day ✌️✌️😉
You could maybe combine the LiPo battery packs and capacitor banks in parallel to maximize current delivery capabilities.
Also you could use the hall effect sensor of the 2nd coil to trigger reverse current on the 1 coil and so on. This would likely consume only a few high current mosfets per coil but also double your current load at any point which your power solution right now might not be able to handle
Some more ways you could improve this:
• Batteries and Supercapacitors working together
• active cooling (liquid-cooled sensors and coils)
• better electrical conductors (how expensive is enamel-coated silver wire?)
• stronger ferromagnetic cores for the coils (tiny rare-earth magnets)
• coat that sled and rail in teflon (should vastly lower friction)
• focused magnetic fields (via steel plates around the launcher to act as magnetic lenses; see Tech Ingredients)
• finally build the complex polarity-switching thing
• implement some optical sensors as well
Tom, your launcher is perfect, but you need to balance your plane. It flies up very fast at big angle, and slows down. If you balance your plane correctly (centre of mass need to be under front part of wing) plane will fly further.