You might find polymers such as PEEK interesting. It's a super polymer usable in 3d printers with a usable part temp of 482 F and a melting point of 662 F Your average hobby soldering iron doesn't even get hot enough to melt this plastic. It has a tensile strength comparable to some types of steel.
@@Hughjanus720 You can, you just needed one with a heated chamber and a high temp hot end. And ender 3 can literaly print peek if you mod it. You will only get the less capable amorphous peek unless your chamber can break 113 C which gets you crystalline peek. That takes water cooling though so most printers can't do crystalline.
For balancing rotating parts like the impeller, you could use the method my father used to balance his model aircraft propellers. He had a shaft the size as the center hole in the propeller. The shaft came to close to a needle point on either end and was carefully balanced at either end on a metal plate so that it would spin as freely as possible. The propeller was then gently spun and allowed to come to a rest. If the propeller ended up one blade down, that blade was marked and spun again to see if the same blade ended up pointing down. If was the same blade again on the next try, a little was shaved from the end and the whole process repeated until it seemed to be random as to which blade was pointing down.
That only balances the weight, not the thrust and this impeller is that small that the weight imbalance would be tiny, you would struggle to find anything low friction enough to do the test the way you described.
@@lukearts2954 honestly his comment is not even close to ChatGPT-levels of unhuman writing. I'd be surprised if ChatGPT or similar wrote that without some prompt engineering
Ya know Daniel, I've been subbed to you for a really long time. You were a kid back then and your videos were kind of cute. Kid makes cool things out foam cuz he is a kid and has no money but here we are all these years later and honestly you are a true gift. Everything you make is interesting even your ad/sponsor parts of videos. So I just want to say thank you for all you do.
Likewise. I’ve been watching him since he was destroying things in his parents garage and neighborhood. Very entertaining. I wish I had even half the fabrication knowledge he has learned over the years.
I used to do motor R&D for Shark Vacuums. Your rate of improvement is great and watching this took me back to some of my favorite engineering. Love your work.
You certainly have Dyson beat in the "explosive potential" category. 😂 The moment you talked about the close clearance and high RPM, my first thought was "it's about to explode".
The motors seem to need a thrust washer/bearing as the axle seems to be pushing out which pushes the impellor into the house. You can see 13:00 after the impellor explodes that the axle retracts when the motor slows.
Great work! It looks like your impeller spins the wrong way for the thread handedness, which if true is the reason for a number of the failures: The impeller is wanting to spin the nut off the motor shaft at higher loads. Easy fix is to print the mirror of the impeller, so that it runs in the opposite direction, and reverse the motor rotation. When I have dynamically balanced things in the past, I’ve used a flexible beam (thin long bar of aluminum) that I can clamp up on to vary its length, made a motor mount to attach to it, and put a dial test indicator (DTI) on the free end. Gives extremely precise measurement and repeatability for balancing.
I also think there's something to be said for the plasticity level of the impeller material. After awhile the friction I'm sure causes material warping which expands the impeller and pushes it against the snail housing walls. Similar to blasting a skateboard wheel with compressed air/water and it expanding.
Instead of a scroll housing I'd suggest taking the air all around circumference back over the motor for cooling, also exhausting in same direction as intake airflow should give better thrust. Early gets worked this way as do most model gas turbines now
The Dyson vacuum motor is aurally a little different from a typical brushless motor. It is a switched reluctance motor if I remember right, they are good at reaching insane speeds.
Psst! If your injection molding is accurate enough, maybe you could make him some more balanced impellers? Or even machine some out of aluminum or another metal?
The higher speed listed for the Dyson is with the inlet blocked. This creates a partial vacuum, which means less resistance on the impeller => higher speed
Daniel this is some awesome engineering! Love your iterations and fearless testing. And your very well rounded skills in Cad and 3D printing. Your steadily increasing physics, systems and shop knowledge help you to quickly recover and test new ideas. Love it, man!
Some of us learned about how Dyson motors are designed, ways to measure thrust and vacuum pressure with basic metering devices, and what it takes to 3D print and manually balance impellers that are spun at high RPM. Others just enjoyed seeing Daniel blow stuff up.
If you want pressurized air to be more efficient you should check out the inside of a compressor housing on a turbo. They don’t use director blades to force the air in the direction of the outlet to increase velocity. It will help a lot.
Last year in my design internship at a quantum cooling startup I got introduced to onshape. Coming vom Rhino7, which of course has so much functions and his hard to master, I quickly found a liking to onshape. The fact, that you can open, view and edit your file from almost any device from anywhere you can imagine because it's cloud based is so tremendous. Plus in a company you can manage projects and put single files together in a huge asssembly is amazing. The company also had formlabs Form3+ printer. Its a gamechanger compared to my own ender 3v2. No hazzle, no tweaking. The print always came out perfect and smooth.
Your turbine impeller exploding is due to unbalanced parts and large tolerances on the interfaces. You need to do two things to address this: 1) balance the rotating assembly, and 2) ensure that there is no movement of the impeller on the shaft. 1) is relatively easy: you construct a balancing spindle as described by another commenter. A precision shaft that the impeller mounts to, which tapers to needle points on both ends. The unit is suspended between angled plates, allowing it to spin freely. Then the whole assembly is spun and the downward facing blades marked. Repeat this cycle to ensure that the same blade is heavy, and then shave until the blade down is not repeatable. 2) is harder because of the materials you're working with and the parts. The threaded shaft on the motor is the main problem. Even if you print a tight fit on the impeller, if it is unbalanced at all then the forces will be amplified by the extreme rotation and cause the plastic to dig into the threads of the shaft, thus introducing slop. This slop allows more movement which allows more vibration and more slop until the whole thing goes boom. To eliminate this, you will need to sleeve the shaft with a hard metal (not brass or copper or aluminum, regular steel should be fine doesn't need to be hardened), and then bring your impeller bore to final size using a reamer instead of relying on the printers finish. Be sure to do the reaming before balancing. You may want to press in a hard bushing/mount on the impeller instead of sleeving the shaft, but again take care to ensure a very close fit on the shaft. You can keep using the nut clamp system, just under size the length of the sleeve so that the nuts still clamp the impeller directly. If you're using washers, you need to ensure a tight tolerance on the washers and consider balancing them as well. A final practical matter is airflow and cooling: you probably want to direct airflow over the motor for cooling. This will probably cost some thrust, but is critical for motor longevity especially at the high power levels you are using. Looking forward to the next installment!
you did an amazing job with these kind of things so far. You definitely showed somebody like me who is an engineer about these kind of things how difficult it really is to make these kind of things efficient enough
$150/L the resins are expensive, but you're only using a few cc's for each print. I'm no resin printing guru, but the tools all have a calculator that you can put the price into, and it'll spit out the volume, and the cost of the part. I'll bet it was a buck or two per impeller? The shell probably $8? Wild ass guesses here.
I do love this video which is awesome but not quite far to compare thrust on a dyson motor which is strictly designed for being a vacuum and not for pushing air out so its really not a far comparison, but they also are designing it for better battery consumption as well. I am so impressed you were able to make a 3d printed impeller spin that fast without vaporizing, that's incredible
The snail housing: i only know of the ones that have a very small outer diameter directly behind the exit (rotation wise). Due to the same diameter much of the air gets just spun around in circle until the centrifugal forces prevent it from making the turn. Hmm... thats probably the diffrence, between max flow or max pressure on the outlet. Yours is probably set up for pressure. You can try to balance the impeller statically by mounting it to a thin needle and place that on a post. That should reduce friction and should turn the impeller with the heavy side down.
Seems like you might have significant losses due to the design of your impeller housing. Normal centrifugal pump housings have a volute that gathers the air flung by the impeller to guide it out of the pump. Your fully toroidal housing might be losing a significant amount of thrust just swirling air around inside... Also, those internal guide vanes probably just represent losses to the airflow coming off the impeller - the angle of the outflowing air should be controlled using blade angle on the impeller, not external vanes. Would be interesting to see the effect a refined housing has on performance!
One way of using the dyson impeller is making something more like a jet engine, you might get more of a ram effect and air multiplication if the exhaust is made right
You can cut down on vibrations a bit by mounting the impeller differently. Threaded rods and nuts are very imprecise, using the nut to center the impeller will result in it being off center and at higher RPMs it will just vibrate like crazy. Minimize the impellers runout (how off center it is) and it will make balancing it much easier.
2:00 about KV rating: KV is for motor RPM what pitch speed is for a prop: when a prop is going at it's pitch speed, it's no longer producing ANY thrust. Same with a brushless, if it's going at it's theoretical RPM (KV x Volts) it's no longer producing ANY torque. When calculating the useful rpm of your motor, you have to multiply by a 'slip' factor: typically around 0.75 or 0.8 depending on alot of things ; ) So your motor should give you 118 000 * 0.75 = +- 90 000 RPM
I used to work in a Turbo shop, and you can properly balance an impeller with a prop balancer. You would know the ones: they magnetically levitate the Prop/fan on a shaft for almost zero friction. Grab one of those to balance your impeller!
The reason why they are using only 2 pole pairs is likely also due to the electronic commutation limit. With an excess of 100000 rpm many cheaper controller are not capable of commutating motors with a high pole pair count. 100000 electronic communications a second is a common limit of those controllers. And because they practically have no load on the motor, it’s forgivable to only have two pole pairs.
no, it is rare used SRM motor with quadrature sequence, very neat design for robust very high RPM cases en.wikipedia.org/wiki/Switched_reluctance_motor
@@AABB-px8lc Fair enough, no magnets for the rotor, makes it a bit cheaper. And with the volume of motors they need to produce it's probably a good move. You'd still ned electronics for commutation though.
Dyson uses a reluctance motor according to my research. Rather than permanent magnets, the rotating core just has a magnetic material like iron that seeks to center itself in a magnetic field. It works like a familiar linear solenoid but in rotation. There is no field reversal in the iron so this supposedly gives an RPM advantage.
Your ability to crank out worthy content is mesmerizing! Thank you! It’s my hope it continues to remain more on the enjoy side than the work side for you:)
Hi, i was thinking about your balancing problem. 1. Perhaps ~ if the impler/disc/whatever was marked into four sections. Red, Green, Blue, Yellow with a dot marking the top. 2. With a constant speed of rotation was maintained 3. You could then use 4 flashes of light in sequence, 1,2,3,4 so that those positions are captured on video, freezing the wheel in a series of four frames. 4. You would then be able to see which of the coloured parts of the wheel where out of alignment. This would help calibrate small wheel of any size :) Looking sideways on the wheel, and zooming in like: R1 Good G2 out B3 Good Y4 out, even worse
How cool. You are one of the few show ng how functional prints can be used for a myriad of applications. Well done keep it up I want to see lawn mower v2
I think you could improve the results with a better snail housing. The radius should increase closer to the exit nozzle. (so the axle of the impeller is eccentric to the housing, while in your design it is nicely centered)
Vibration is definitely a huge thing to deal with... You have to balance the impeller for static and dynamic vibration... But one more thing to consider is physical expansion due to high centrifugal force... Materials tend to expand at such high RPMs/Forces...so you have to adjust you tolerances accordingly.
you need to set impeller tip clearances to the shroud and volute. Also the blade pitch and load matter which comes down to setting the correct rpm to the impeller design. The motor direct drive will not reduce the rpm. You need a secure fit to maintain the correct clearances, and it will reduce vibration. which will increase efficiency and reduce load on your bearings. If your impeller clearance is too tight.. you will get rub which will heat the impeller and thermal expansion will detonate the machine. Yes your bearings are faaarked from repeated detonations and vibration. Your alignment is also terrible and without good alignment they will keep exploding. All these things matter and are critical to reliability which in power generation is the most important factor. Sorry if that all sounds very critical but those are the problems evident here. I did really enjoy the video and enjoyed seeing your designs detonate. Thanks for sharing and going to the time to make this.
One concept I've always wanted to see is using one of these impellers to create a pseudo "fanwing" design. Where air is compressed then blown over the surface of a wing using a slit, to see if the fast moving air over the top of the aerofoil generates extra lift at low relative speeds.
Really great effort, the balancing of high speed parts seems to need computer accuracy to be successful. Dyson may have a few more people, and a lot more money, to develop these systems, but I LOVE that you gave it a shot !!!
They use 4 coils and a hall effect sensor for multiple reasons : 1- ensure the right rotation way of the motor and so of the impeller : they need the position of the rotor (not necessarily precisely) and then they will start only two of the coils to be sure of the rotation way, then inertia and the hall effect sensor take the relay i guess or maybe they just use two coils when the magnets of the rotor are closed to their end position, better see with a scope to be sure haha… 2- the hall sensor as in three phased esc is used to have feed back and so send the rights pulses at the right moments to have a stable torque ans so make their product less louder. But in the end for a 2 phased motor to work you don’t need all this, you’ll just won’t know the rotation way and the position of the rotor so you will just need to respect the KV rating to make it work
You could probably use a bearing closer to the impeller. In the car world we would use a run out dial indicator and make sure the back side and all the tips of the blades are spinning within .001 and if your outside rear edge and all the tips of the blades are running at .001 it should be balanced. Adding hot glue will interfere with the blade effectiveness so just try taking of smooth layers on the meat or within the central part on the back of the impeller to balance just a lot of back and forth on the micrometer
I never thought about the axial thrust load on the impeller shaft when it's sucking an air. There has to be a force pushing on the impeller blade. That's awesome
A disk of weight behind the impeller to act as a flywheel will improve the reciprocating motion and reduce vibration. A bigger housing encapsulating the electric motor should aid cooling and any heat inside the pressurized area will help thrust by expanding the volume of air, even if only marginally.
A better way to radially balance the impeller is to set up a set of parallel bars the are on the same plane that are level. Run a straight and true rod all the way through the impeller and place on the parallel bars. Nudge it to get it to roll. When it stops the heaviest side will be towards the floor. Grind a little off and repeat until the impeller doesn't come to a definitive stop, it just seems like it wants to keep rolling. At that point you'll have a balanced impeller that didn't require some special electronics set up.
If the motor was on the intake side with larger inlet that allows / forces the airflow around the outside of the motor it could cool itself down while it draws in the air. 🤔😉👍 Those resin printed parts look absolutely amazing.
Next time you make a housing put the outlet of the impeller at the bottom of the snail. You want to induce air so it flows around the outside of the donut. By dumping air into the center like you are, you are inducing turbulence and friction in the airflow. You can actually compress air in the housing (which requires more flow) if you do it like this which will actually result in more thrust.
It's probably a slightly dumb question, but why not add a thin bar in front that supports the shaft at thd front also? It would cause minimal air intake losses I think. Just a straight line across the tip that connects to the front of the shaft with a small bearing to keep it from shaking.
Every issue here is quite easily fixable. Next step is to either go back to a one piece housing or have a two-piece that bolts together around its perimeter, a glued butt joint will always be problematic. Similarly, you need to beef up the motor attachment, again properly bolting things together. Some vibration dampening would help, for resonance purposes though at the same time better coupling everything (frequent bolts, wicking threadlocker, etc.) would also do the job, but also DIYing a dynamic balancer would help as well. Another thing that could help is having a bearing flange within the intake nozzle, similar to how PC fans locate their bearing within the fan's mounting square, with a full through-shaft for the impeller, this way you're also eliminating any movement within your printed plastic parts, especially when vibration and harmonics do become problematic. I'd also look around for a different motor or getting a different mount for that motor, that thing is long enough to become problematic, not to mention a better housing would have a wider mounting footprint, one that would otherwise be much more compatible with the project; also, a custom housing would provide the ability to better cool the motor, as heat will also become problematic, a simple fan with airflow would suffice, akin to how various power tool motors have a cooling fan on them. Ideally you would also key the impeller to the shaft, rather than relying on threaded nuts to locate it; or, thread the impeller, and jamnut both ends, and again use _wicking_ threadlocker to ensure a fuller seal on the threads of every threaded component- though, threaded components also often have issues with concentricity, this is why keyed precision shafts between located bearings is far superior.
I designed turbos for 5 years. The motor very likely spins faster without the housing. In applications like this it's an interesting design problem as you're effectively trying to design the most efficient compressor for a given specific power output. You have to balance flow, pressure, and rpm.
I can vouch for the tenacity of Phrozen-Nylon Green Tough resin. I mostly print with my FDM printers and recently I laser too but the Phrozen-Nylon is very resilient, definitely a better bet than the standard tough resins. Nylon actually has more resilience with a bit of it's hygroscopicity satisfied. Very dry and it is more brittle.
Maximizing thrust involves different aerodynamics from maximizing suction. And, maximizing static thrust is different from high speed thrust. The small diameter orifice would be good for a fast jet-like aircraft. A slower flying aircraft would have a larger diameter orifice, moving a larger mass of air at a slower speed, and would have more static thrust.
you can use high speed camera for balancing of the rotor. make a few marks that will be distinguishable when spinning and you can just chase the mark that oscilates the rotor furthest.
The faster you try to move air (eg higher rpm), the denser the air gets. Adding the 2nd set of poled increases the torue the motor can sustain, and at a perfect 180⁰ out of phase they can be aligned perfectly to never fight against one another. The reason it isn't always done is cost. Of parts, of upgrading manufacturing, or the cost of a more complex control curcuit. Just some food for thought 🙂
![⭐️( ลอยกระทง 2024 ) Ver. แดนซ์ Remix BY [ ดีเจกิต รีมิกซ์ ]](http://i.ytimg.com/vi/rMWOLhgXUHc/mqdefault.jpg)
The fact that you can actually 3D print an impeller rotating at such a speed with a plastic is insane
Makes me dizzy thinking about it
You might find polymers such as PEEK interesting.
It's a super polymer usable in 3d printers with a usable part temp of 482 F and a melting point of 662 F
Your average hobby soldering iron doesn't even get hot enough to melt this plastic. It has a tensile strength comparable to some types of steel.
@@nocare good luck printing that with a 3D printer at home…..
@@Hughjanus720 You can, you just needed one with a heated chamber and a high temp hot end.
And ender 3 can literaly print peek if you mod it.
You will only get the less capable amorphous peek unless your chamber can break 113 C which gets you crystalline peek.
That takes water cooling though so most printers can't do crystalline.
@@nocare I could build a whole car at home too, just don’t think it’s worth it
The dyson's impeller is used in a vacuum that sucks up dust and the one this guy made creates dust and efficiently spreads it around everywhere.
Fresh micro-plastics for your lungs 🤤
The world requires balance.
Anti-dyson expeller
and the circle closes
@@AnIdiotwithaSubaru ??? There is hepa filter at the end
For balancing rotating parts like the impeller, you could use the method my father used to balance his model aircraft propellers. He had a shaft the size as the center hole in the propeller. The shaft came to close to a needle point on either end and was carefully balanced at either end on a metal plate so that it would spin as freely as possible. The propeller was then gently spun and allowed to come to a rest. If the propeller ended up one blade down, that blade was marked and spun again to see if the same blade ended up pointing down. If was the same blade again on the next try, a little was shaved from the end and the whole process repeated until it seemed to be random as to which blade was pointing down.
Wonder if ya.could use water to float it check for lower/higher side
That's how to balance everything that rotates actually
That only balances the weight, not the thrust and this impeller is that small that the weight imbalance would be tiny, you would struggle to find anything low friction enough to do the test the way you described.
@@lukearts2954 honestly his comment is not even close to ChatGPT-levels of unhuman writing. I'd be surprised if ChatGPT or similar wrote that without some prompt engineering
@@lukearts2954 It wasn’t chat GPT, I’m just neuro divergent and struggle to organise a coherent sentence 😂
Ya know Daniel, I've been subbed to you for a really long time. You were a kid back then and your videos were kind of cute. Kid makes cool things out foam cuz he is a kid and has no money but here we are all these years later and honestly you are a true gift. Everything you make is interesting even your ad/sponsor parts of videos. So I just want to say thank you for all you do.
This is such a nice comment. Awwww!!
wholesome comment
Likewise. I’ve been watching him since he was destroying things in his parents garage and neighborhood. Very entertaining. I wish I had even half the fabrication knowledge he has learned over the years.
Always amazed me as to how he could glue a few bits of foam together and have such a good flying rc plane
I also love how he lets shit blow up and scuff his hand, but doesn't care and goes on with the project. It's aerospace grade chaos, no drama!🤣
Knowing a little about high rpm’s, it is ALLL about balance. It’s quite impressive to get such results in a home lab setting 👍
it's not quiet , it's loud and impressive nonetheless , i see u man have understanding gut in you big balls , basketballs i mean
Bro turbo’s his vacuuum
@@KingSkrap bro should put a mini v12 in his vacuum
I used to do motor R&D for Shark Vacuums. Your rate of improvement is great and watching this took me back to some of my favorite engineering. Love your work.
Gotta do your velocity triangles and angular momentum flux conservation.
I appreciate your work at Shark. Impressive power from my Shark!
shark motors die so soon compared to dyson :(
You certainly have Dyson beat in the "explosive potential" category. 😂
The moment you talked about the close clearance and high RPM, my first thought was "it's about to explode".
The motors seem to need a thrust washer/bearing as the axle seems to be pushing out which pushes the impellor into the house. You can see 13:00 after the impellor explodes that the axle retracts when the motor slows.
Was just about to mention that, he's got a lot of thing's to fix to make something remotely long lasting/usable
I could be wrong, but I don't think the shaft moved it was the nuts spinning down from the inertia.
@@41A2E true, the shaft never moves only the nuts. Zoom in to make it easier to see
Great work! It looks like your impeller spins the wrong way for the thread handedness, which if true is the reason for a number of the failures: The impeller is wanting to spin the nut off the motor shaft at higher loads. Easy fix is to print the mirror of the impeller, so that it runs in the opposite direction, and reverse the motor rotation.
When I have dynamically balanced things in the past, I’ve used a flexible beam (thin long bar of aluminum) that I can clamp up on to vary its length, made a motor mount to attach to it, and put a dial test indicator (DTI) on the free end. Gives extremely precise measurement and repeatability for balancing.
I also think there's something to be said for the plasticity level of the impeller material. After awhile the friction I'm sure causes material warping which expands the impeller and pushes it against the snail housing walls.
Similar to blasting a skateboard wheel with compressed air/water and it expanding.
Instead of a scroll housing I'd suggest taking the air all around circumference back over the motor for cooling, also exhausting in same direction as intake airflow should give better thrust. Early gets worked this way as do most model gas turbines now
The Dyson vacuum motor is aurally a little different from a typical brushless motor. It is a switched reluctance motor if I remember right, they are good at reaching insane speeds.
yeah. surprised he got something as basic as this wrong
But the Dyson motor has a permanent magnet as the rotor so it's not a switched reluctance motor, it's just the same old BLDC
@@hOZish7 it has no permanent magnet
@@disp3rsion neodymium magnets are permanent and dyson uses those
This is really SOMETHING! Thanks soo much for bringing us this, Daniell🥳🥳🥳
Psst! If your injection molding is accurate enough, maybe you could make him some more balanced impellers? Or even machine some out of aluminum or another metal?
@@bjrn-oskarrnning2740 An aluminum impeller sounds pretty sick
it's insane what your company offer :)
The higher speed listed for the Dyson is with the inlet blocked. This creates a partial vacuum, which means less resistance on the impeller => higher speed
Daniel this is some awesome engineering! Love your iterations and fearless testing. And your very well rounded skills in Cad and 3D printing. Your steadily increasing physics, systems and shop knowledge help you to quickly recover and test new ideas. Love it, man!
Dyson fan multirotor seems eerily close to fruition 😮
do it!
This is EXACTLY what I've been looking for, an insanely high pressure blower fan that I can 3d print cheaply. Thanks!!!
Today we are going to see how loud, inefficient, and unreliable we can make an impeller
Some of us learned about how Dyson motors are designed, ways to measure thrust and vacuum pressure with basic metering devices, and what it takes to 3D print and manually balance impellers that are spun at high RPM. Others just enjoyed seeing Daniel blow stuff up.
@@28th_St_Air the 3rd one for me
@@28th_St_Air I did both. 🤷🏻♂️
"Sucking at something is the first step to being sorta good at something." Winston Churchill
“Experience is what you get right after you needed it.” - Me
If you want pressurized air to be more efficient you should check out the inside of a compressor housing on a turbo. They don’t use director blades to force the air in the direction of the outlet to increase velocity. It will help a lot.
Last year in my design internship at a quantum cooling startup I got introduced to onshape. Coming vom Rhino7, which of course has so much functions and his hard to master, I quickly found a liking to onshape. The fact, that you can open, view and edit your file from almost any device from anywhere you can imagine because it's cloud based is so tremendous. Plus in a company you can manage projects and put single files together in a huge asssembly is amazing.
The company also had formlabs Form3+ printer. Its a gamechanger compared to my own ender 3v2. No hazzle, no tweaking. The print always came out perfect and smooth.
Your turbine impeller exploding is due to unbalanced parts and large tolerances on the interfaces. You need to do two things to address this: 1) balance the rotating assembly, and 2) ensure that there is no movement of the impeller on the shaft.
1) is relatively easy: you construct a balancing spindle as described by another commenter. A precision shaft that the impeller mounts to, which tapers to needle points on both ends. The unit is suspended between angled plates, allowing it to spin freely. Then the whole assembly is spun and the downward facing blades marked. Repeat this cycle to ensure that the same blade is heavy, and then shave until the blade down is not repeatable.
2) is harder because of the materials you're working with and the parts. The threaded shaft on the motor is the main problem. Even if you print a tight fit on the impeller, if it is unbalanced at all then the forces will be amplified by the extreme rotation and cause the plastic to dig into the threads of the shaft, thus introducing slop. This slop allows more movement which allows more vibration and more slop until the whole thing goes boom. To eliminate this, you will need to sleeve the shaft with a hard metal (not brass or copper or aluminum, regular steel should be fine doesn't need to be hardened), and then bring your impeller bore to final size using a reamer instead of relying on the printers finish. Be sure to do the reaming before balancing. You may want to press in a hard bushing/mount on the impeller instead of sleeving the shaft, but again take care to ensure a very close fit on the shaft.
You can keep using the nut clamp system, just under size the length of the sleeve so that the nuts still clamp the impeller directly. If you're using washers, you need to ensure a tight tolerance on the washers and consider balancing them as well.
A final practical matter is airflow and cooling: you probably want to direct airflow over the motor for cooling. This will probably cost some thrust, but is critical for motor longevity especially at the high power levels you are using.
Looking forward to the next installment!
It's easier to remove support material before curing 👍
Also, using nyloc® nuts for the motor will keep them tight under vibrations
yep. my recommended workflow is to remove the part from the BP and take off the supports between Wash and Cure.
"digital motor" gotta love the marketing department and their brilliant naming ideas
Engineering, with comedy, is always the best form of entertainment! Thanks again Daniel, I always enjoy your content. Bye. :)
you did an amazing job with these kind of things so far. You definitely showed somebody like me who is an engineer about these kind of things how difficult it really is to make these kind of things efficient enough
It's fun to see how you and great scott differ in procedure. The hammer to open it in particular 😄
I can’t imagine the prices of those formlabs resins. How long does 1 Liter go for you? How much did the impeller take?
$150/L the resins are expensive, but you're only using a few cc's for each print. I'm no resin printing guru, but the tools all have a calculator that you can put the price into, and it'll spit out the volume, and the cost of the part. I'll bet it was a buck or two per impeller? The shell probably $8? Wild ass guesses here.
I do love this video which is awesome but not quite far to compare thrust on a dyson motor which is strictly designed for being a vacuum and not for pushing air out so its really not a far comparison, but they also are designing it for better battery consumption as well. I am so impressed you were able to make a 3d printed impeller spin that fast without vaporizing, that's incredible
I like that you're willing to take things you've worked hard on and test them to destruction.
The snail housing: i only know of the ones that have a very small outer diameter directly behind the exit (rotation wise). Due to the same diameter much of the air gets just spun around in circle until the centrifugal forces prevent it from making the turn. Hmm... thats probably the diffrence, between max flow or max pressure on the outlet. Yours is probably set up for pressure.
You can try to balance the impeller statically by mounting it to a thin needle and place that on a post. That should reduce friction and should turn the impeller with the heavy side down.
Looks like it could be a great electric turbo for a car, bike, boat, or lawn motor...
Also, what a sexy 3D print project!
Seems like you might have significant losses due to the design of your impeller housing. Normal centrifugal pump housings have a volute that gathers the air flung by the impeller to guide it out of the pump. Your fully toroidal housing might be losing a significant amount of thrust just swirling air around inside... Also, those internal guide vanes probably just represent losses to the airflow coming off the impeller - the angle of the outflowing air should be controlled using blade angle on the impeller, not external vanes. Would be interesting to see the effect a refined housing has on performance!
The fact that it didn't blow up within 5 seconds impressed me!! I don't think I could ever make something spin that fast and not go pop!
One way of using the dyson impeller is making something more like a jet engine, you might get more of a ram effect and air multiplication if the exhaust is made right
You can cut down on vibrations a bit by mounting the impeller differently. Threaded rods and nuts are very imprecise, using the nut to center the impeller will result in it being off center and at higher RPMs it will just vibrate like crazy. Minimize the impellers runout (how off center it is) and it will make balancing it much easier.
2:00 about KV rating: KV is for motor RPM what pitch speed is for a prop:
when a prop is going at it's pitch speed, it's no longer producing ANY thrust.
Same with a brushless, if it's going at it's theoretical RPM (KV x Volts) it's no longer producing ANY torque.
When calculating the useful rpm of your motor, you have to multiply by a 'slip' factor: typically around 0.75 or 0.8 depending on alot of things ; )
So your motor should give you 118 000 * 0.75 = +- 90 000 RPM
I used to work in a Turbo shop, and you can properly balance an impeller with a prop balancer. You would know the ones: they magnetically levitate the Prop/fan on a shaft for almost zero friction. Grab one of those to balance your impeller!
Videos like these and the influence of my dad is why I'm getting into engineering lol
You can use air flow meter from the car, to measure air mass.
The reason why they are using only 2 pole pairs is likely also due to the electronic commutation limit. With an excess of 100000 rpm many cheaper controller are not capable of commutating motors with a high pole pair count. 100000 electronic communications a second is a common limit of those controllers.
And because they practically have no load on the motor, it’s forgivable to only have two pole pairs.
no, it is rare used SRM motor with quadrature sequence, very neat design for robust very high RPM cases en.wikipedia.org/wiki/Switched_reluctance_motor
@@AABB-px8lc Fair enough, no magnets for the rotor, makes it a bit cheaper. And with the volume of motors they need to produce it's probably a good move. You'd still ned electronics for commutation though.
Dyson uses a reluctance motor according to my research. Rather than permanent magnets, the rotating core just has a magnetic material like iron that seeks to center itself in a magnetic field. It works like a familiar linear solenoid but in rotation. There is no field reversal in the iron so this supposedly gives an RPM advantage.
Your ability to crank out worthy content is mesmerizing! Thank you!
It’s my hope it continues to remain more on the enjoy side than the work side for you:)
Hi, i was thinking about your balancing problem.
1. Perhaps ~ if the impler/disc/whatever was marked into four sections. Red, Green, Blue, Yellow with a dot marking the top.
2. With a constant speed of rotation was maintained
3. You could then use 4 flashes of light in sequence, 1,2,3,4 so that those positions are captured on video, freezing the wheel in a series of four frames.
4. You would then be able to see which of the coloured parts of the wheel where out of alignment. This would help calibrate small wheel of any size :)
Looking sideways on the wheel, and zooming in like:
R1 Good
G2 out
B3 Good
Y4 out, even worse
How cool. You are one of the few show ng how functional prints can be used for a myriad of applications. Well done keep it up
I want to see lawn mower v2
I love the how to basic style of deconstruction on the vacuum
Thank you for doing what you do never stop making these videos.
I think you could improve the results with a better snail housing. The radius should increase closer to the exit nozzle. (so the axle of the impeller is eccentric to the housing, while in your design it is nicely centered)
"I used the built-in lathe" got a good chuckle from me. I'm still watching, just wanted to mention this.
Vibration is definitely a huge thing to deal with... You have to balance the impeller for static and dynamic vibration... But one more thing to consider is physical expansion due to high centrifugal force... Materials tend to expand at such high RPMs/Forces...so you have to adjust you tolerances accordingly.
you need to set impeller tip clearances to the shroud and volute. Also the blade pitch and load matter which comes down to setting the correct rpm to the impeller design. The motor direct drive will not reduce the rpm. You need a secure fit to maintain the correct clearances, and it will reduce vibration. which will increase efficiency and reduce load on your bearings. If your impeller clearance is too tight.. you will get rub which will heat the impeller and thermal expansion will detonate the machine. Yes your bearings are faaarked from repeated detonations and vibration. Your alignment is also terrible and without good alignment they will keep exploding. All these things matter and are critical to reliability which in power generation is the most important factor. Sorry if that all sounds very critical but those are the problems evident here. I did really enjoy the video and enjoyed seeing your designs detonate. Thanks for sharing and going to the time to make this.
Holy CRAP! The way this first iteration blew itself up was just... my reaction was like, "WHAAAH!"
Cool beans.
I have 18 partially completed projects, but I think you have me beat in terms of competed & more to come.
One concept I've always wanted to see is using one of these impellers to create a pseudo "fanwing" design. Where air is compressed then blown over the surface of a wing using a slit, to see if the fast moving air over the top of the aerofoil generates extra lift at low relative speeds.
Really great effort, the balancing of high speed parts seems to need computer accuracy to be successful. Dyson may have a few more people, and a lot more money, to develop these systems, but I LOVE that you gave it a shot !!!
They use 4 coils and a hall effect sensor for multiple reasons :
1- ensure the right rotation way of the motor and so of the impeller : they need the position of the rotor (not necessarily precisely) and then they will start only two of the coils to be sure of the rotation way, then inertia and the hall effect sensor take the relay i guess or maybe they just use two coils when the magnets of the rotor are closed to their end position, better see with a scope to be sure haha…
2- the hall sensor as in three phased esc is used to have feed back and so send the rights pulses at the right moments to have a stable torque ans so make their product less louder.
But in the end for a 2 phased motor to work you don’t need all this, you’ll just won’t know the rotation way and the position of the rotor so you will just need to respect the KV rating to make it work
You could probably use a bearing closer to the impeller. In the car world we would use a run out dial indicator and make sure the back side and all the tips of the blades are spinning within .001 and if your outside rear edge and all the tips of the blades are running at .001 it should be balanced. Adding hot glue will interfere with the blade effectiveness so just try taking of smooth layers on the meat or within the central part on the back of the impeller to balance just a lot of back and forth on the micrometer
I never thought about the axial thrust load on the impeller shaft when it's sucking an air. There has to be a force pushing on the impeller blade. That's awesome
13:13
Just look at this beautiful vortex just spiraling in. Forming and dissolving. Nature is so beautiful.
Love the pcb way ad read while wearing the on-shape tee-shirt. Hitting two birds with one stone
A disk of weight behind the impeller to act as a flywheel will improve the reciprocating motion and reduce vibration.
A bigger housing encapsulating the electric motor should aid cooling and any heat inside the pressurized area will help thrust by expanding the volume of air, even if only marginally.
Please, please do a deep dive on magneto hydrodynamics… your the only one that will go down a rabbit hole and come back with answers!!😊
"it's loud, inefficient and likely to explode" you mean it's perfect, thanks for the video!
A better way to radially balance the impeller is to set up a set of parallel bars the are on the same plane that are level. Run a straight and true rod all the way through the impeller and place on the parallel bars. Nudge it to get it to roll. When it stops the heaviest side will be towards the floor. Grind a little off and repeat until the impeller doesn't come to a definitive stop, it just seems like it wants to keep rolling. At that point you'll have a balanced impeller that didn't require some special electronics set up.
If the motor was on the intake side with larger inlet that allows / forces the airflow around the outside of the motor it could cool itself down while it draws in the air. 🤔😉👍 Those resin printed parts look absolutely amazing.
3:10 I got a free sample of formlabs printing capabilities a few years ago.
It was a sample of filament that can be used for springy things
Next time you make a housing put the outlet of the impeller at the bottom of the snail. You want to induce air so it flows around the outside of the donut. By dumping air into the center like you are, you are inducing turbulence and friction in the airflow. You can actually compress air in the housing (which requires more flow) if you do it like this which will actually result in more thrust.
I am amazed how long the Dyson motors last. They are really made well. period !!!
8:13 “This video is sponsored by PCBWay” he says while wearing an Onshape shirt
Amazing, almost feel like putting on some safety goggles just looking at the high RPM testrun.
"It is loud, inefficient, and it is likely to explode." I think with that you've offically crossed the line into Mad Science.
Amazing guy. A pleasure to listen to.
The parts on the impeller, including the hex nut, must be balanced, just like in a turbocharger on a gasoline engine.
Broooo, you should make a “hundred thousand miles test on rc cars” but with higher quality cars. That would be awesome, I love that video
Wow Never new that pcbway could Machine big parts
loved the video- - loved the microplastics on your hands they will be some where else soon
The howtobasic-style hammering at the beginning made me chuckle :D
Get PCBWay to metal print an impeller for you! Awesome video
It's probably a slightly dumb question, but why not add a thin bar in front that supports the shaft at thd front also? It would cause minimal air intake losses I think. Just a straight line across the tip that connects to the front of the shaft with a small bearing to keep it from shaking.
Every issue here is quite easily fixable. Next step is to either go back to a one piece housing or have a two-piece that bolts together around its perimeter, a glued butt joint will always be problematic. Similarly, you need to beef up the motor attachment, again properly bolting things together. Some vibration dampening would help, for resonance purposes though at the same time better coupling everything (frequent bolts, wicking threadlocker, etc.) would also do the job, but also DIYing a dynamic balancer would help as well. Another thing that could help is having a bearing flange within the intake nozzle, similar to how PC fans locate their bearing within the fan's mounting square, with a full through-shaft for the impeller, this way you're also eliminating any movement within your printed plastic parts, especially when vibration and harmonics do become problematic. I'd also look around for a different motor or getting a different mount for that motor, that thing is long enough to become problematic, not to mention a better housing would have a wider mounting footprint, one that would otherwise be much more compatible with the project; also, a custom housing would provide the ability to better cool the motor, as heat will also become problematic, a simple fan with airflow would suffice, akin to how various power tool motors have a cooling fan on them. Ideally you would also key the impeller to the shaft, rather than relying on threaded nuts to locate it; or, thread the impeller, and jamnut both ends, and again use _wicking_ threadlocker to ensure a fuller seal on the threads of every threaded component- though, threaded components also often have issues with concentricity, this is why keyed precision shafts between located bearings is far superior.
You also need to weigh the external ESC when calculating thrust to weight since the Dyson has a built in ESC.
Congratulations on getting 3 different sponsors in one video
Well to be fair part of that weight for the dyson includes the speed controller etc.
You can actually rig a mobile phone to the mount to test the vibrations in IPS with an app. It's super sensitive so might help with balancing
"Loud, inefficient, and likely to explode" - that just makes it better!
I designed turbos for 5 years. The motor very likely spins faster without the housing. In applications like this it's an interesting design problem as you're effectively trying to design the most efficient compressor for a given specific power output. You have to balance flow, pressure, and rpm.
Yes, and if you don't care about pressure, get rid of the vaned diffusor. Less likely to surge.
Idea for a video, Making different styles of 3D printed water pumps & seeing which one is better
❤ PLEASE BUILD A PROPANE AFTERBURNER FOR THIS 🔥
I would assume they could go to a 4 pole motor on newer models due to faster switching semi conductors in the motor control.
that engineering resin is awesome, didnt know that was a thing
Technically, when comparing weight you need to put your ECS on there too since their motor includes the speed controller PCB.
Best ad ever! I wanted some pcbs printed.
This video had some really cool shots!
You've come a long way from that kid bashing foam in the school yard! Right on!
I can vouch for the tenacity of Phrozen-Nylon Green Tough resin. I mostly print with my FDM printers and recently I laser too but the Phrozen-Nylon is very resilient, definitely a better bet than the standard tough resins. Nylon actually has more resilience with a bit of it's hygroscopicity satisfied. Very dry and it is more brittle.
Maximizing thrust involves different aerodynamics from maximizing suction. And, maximizing static thrust is different from high speed thrust. The small diameter orifice would be good for a fast jet-like aircraft. A slower flying aircraft would have a larger diameter orifice, moving a larger mass of air at a slower speed, and would have more static thrust.
you can use high speed camera for balancing of the rotor. make a few marks that will be distinguishable when spinning and you can just chase the mark that oscilates the rotor furthest.
The faster you try to move air (eg higher rpm), the denser the air gets. Adding the 2nd set of poled increases the torue the motor can sustain, and at a perfect 180⁰ out of phase they can be aligned perfectly to never fight against one another. The reason it isn't always done is cost. Of parts, of upgrading manufacturing, or the cost of a more complex control curcuit.
Just some food for thought 🙂
Fun thing. Being Finnish the KV actually works directly in Finnish language as "kierrosta (rounds) voltille (per volt)"
Ive been wanting to make one of these for a project, this is amazing
Really cool video. Thank you for sharing the cad file, was cool and helpful to look through and see how you made the it, especially the blades. : )