Experimental Levitating Rotor Machines
Construction Details
Construction Details
A Magnetic Levitating Torus Rotor
Within A Non Magnetic Metal Torus
Some possible variations.
I have made a "supermagnet" from 2 N35 NdFeB Rare Earth magnets 19mmdiam x 28.2mm long. Total length 56.4mm. Jaycar CAT.NO: LM1652. Price July 2022 $A23.70 a pair for 20+
These are made by retaining 2 magnets with like poles facing.
These will fit a bike rim 650mm diameter at magnet center if a 32 magnet 24 coil position layout is used. Circ = 2042.85/32 = 63.83 gap for each magnet.
The magnets are 56.4mm + room for a 5mm rivet = 61.4mm.
Alum tube 25mmOD x 3mm wall is a nice fit for the supermagnet retainers. This tube could be turned down slightly to locate further into the rim, as it may be better to reduce the magnet gap.
1.5mm gap reduction would make the magnet / torus diameter ????
These magnets will not fit the Torus outer shell which I have made which requires 36 poles.
When a pair of these magnets are joined magnetically with unlike poles facing, and dropped down an alum tube 26mm ID x 3mm wall, they take about 5.2 seconds to fall 1 meter.
When another pair is added the weight is doubled.Theoretically the field strength should double, but the fall time is reduced to 2.5 sec.
When the supermagnet is dropped the time taken to fall 1 meter is 8.7 sec. The drag is substantial.
When this supermagnet is dropped down a 1 meter length of 67mm ID x 4.5mm alum tube it tumbles and there is perhaps a slight drag. When another magnet is added to each end the tumbling stops. There is a noticeable drag, and the assemly takes approx 2 seconds to drop.
When the supermagnet is held central within the 67mm ID tube, Hall Effect sensor readings show the field extends 50mm beyond the OD. This indicates that the outer tube may need to be considerably larger.
Some choices available locally are 80mm OD x 3mm Wall Thickness
90mm OD x 3mm Wall Thickness
100mm OD x 3mm Wall Thickness
101.6mm OD x 6.35mm Wall Thickness
150mm OD x 3mm Wall Thickness
I think the 100 x 3 is a best guess. If it is too big, some pieces of the same could be softened and fitted inside to make the wall 6mm thick. There seems to be no option but to guess.
The magnet selection is also trial and error. I have made, and currently testing, a rotor made with 10mm x 10mm N35 cylinder magnets. This does not run as well as the ferrite rotor, and neither has reached levitation.
I think the best possibility may be 15mm dia x 20mm long N52 magnets fitted to the existing 36 pole 27 coil position Torus.
To substantially reduce weight AL magnet wire could be used. I only have a good supply of CU magnet wire so will continue using that.
Of real interest may be to construct the outer Torus of material which is claimed to be from a crashed UFO. Analysis shows this to be mostly magnesium, with small amounts of layered bismuth, and some zinc. This could perhaps be somewhat replicated using 3D metal printing. Mg is about 64% of the weight of Al, so the machine may be substantially lighter and have better rotor levitational properties. I have read that this material will levitate when subjected to terahertz frequencies. I have not heard of anyone exposing it to fast moving magnetic flux.
There are issues with ignition and flammability of magnesium, and availabiliy, which make it problematic for me.
It should be noted that the Torus machine I am experimenting with is based on the solid science and engineering of electro dynamic suspension. Like the Inductrac train on which it is based, the rotor / train must levitate when the correct combination of magnets, outer shell materials, dimensions and speed are found.
The only area of scientific conflict is what will happen when the levitating rotor speed is substantially increased.
Another build possibility is to increase the number of supermagnet poles for the same circumference machine. The current project could be fitted with a rotor having 72 radial supermagnet poles. Possible coil positions would increase to 54.
As an added benefit, no extra magnets would be required, and no extra weight gained, but a doubling of both available magnetic flux and rate of change of flux polarity would be obtained.
72 Magnet poles. 54 Coil Positions. 650 Diameter. Two 15mm dia x 10mm long magnets makes one 20mm long supermagnet. The space between supermagnets is 7.7mm so 15mm x 5mm long magnets will fill the gaps nicely. Could use the existing Torus shell by winding intermediate coils and relocating one of the hall sensors. This will mean a doubling of the switching speed. I have tested to see if there is any loss of field strength / distance using 10mm dia supermagnet radial pole magnets at both 26mm & 52mm spacing, and there is no appreciable difference.
However, tests using readily available N50 grade 15mm diameter magnets show a noticeable drop in flux strength at tube wall distance with 26mm spacing.
But at 10mm from the magnet the flux is both strong and identical at 26mm & 52mm spacing. This seems to indicate that using the closer spacing is a good choice, as it will be more likely for the rotor to levitate at a lower speed. It should be possible to fit an AL packer say 3mm x 15mm in the invert of the outer torus between the wheels, bringing the rotor a bit closer to the non magnetic conductor, and therefore a little more likely to levitate.
The most important thing is to get rotor levitation. I believe that when levitation occurs, there should be a weight reduction of the overall machine, at least equal to the weight of the rotor.
The magnet retainers could be made from 20mm x 3mm wall AL tube about 4mm longer than 2 magnets. This would need boring out to 15mm for a loose fit, and turning down the ends to about 1.25 wall thickness to bend over and hold the magnets. More work but less likely to come apart.
The rotor could use an AL 26" bike rim. 20mm tube fits nicely into the rim. 2 layers of AL sheet metal around the outside of the magnets, and some fiberglass should hold it all together.
The only problem I see is that the Hall Sensors may be difficult to adjust with only about 26mm between alternating magnet poles. They may need relocating inside the torus shell.
Another problem with closer spacing of the magnets is the fabrication of the outer torus. My build was tedious and difficult to weld. The shape is lobsterback. More coil walls and less space means that casting may be my only option for further development. A mold of say 6 sections, when cut in half would make 2 complete molds. The molds would need to be made of steel for multi use. Each casting would be about 450 grams, so only a small forge would be required.
Welding the castings together would be fairly simple.
A lobsterback torus made with 54 sections would have a much more circular or smooth internal surface than 27 sections. I have no idea which might be better for creating electrodynamic suspension.
These are made by retaining 2 magnets with like poles facing.
These will fit a bike rim 650mm diameter at magnet center if a 32 magnet 24 coil position layout is used. Circ = 2042.85/32 = 63.83 gap for each magnet.
The magnets are 56.4mm + room for a 5mm rivet = 61.4mm.
Alum tube 25mmOD x 3mm wall is a nice fit for the supermagnet retainers. This tube could be turned down slightly to locate further into the rim, as it may be better to reduce the magnet gap.
1.5mm gap reduction would make the magnet / torus diameter ????
These magnets will not fit the Torus outer shell which I have made which requires 36 poles.
When a pair of these magnets are joined magnetically with unlike poles facing, and dropped down an alum tube 26mm ID x 3mm wall, they take about 5.2 seconds to fall 1 meter.
When another pair is added the weight is doubled.Theoretically the field strength should double, but the fall time is reduced to 2.5 sec.
When the supermagnet is dropped the time taken to fall 1 meter is 8.7 sec. The drag is substantial.
When this supermagnet is dropped down a 1 meter length of 67mm ID x 4.5mm alum tube it tumbles and there is perhaps a slight drag. When another magnet is added to each end the tumbling stops. There is a noticeable drag, and the assemly takes approx 2 seconds to drop.
When the supermagnet is held central within the 67mm ID tube, Hall Effect sensor readings show the field extends 50mm beyond the OD. This indicates that the outer tube may need to be considerably larger.
Some choices available locally are 80mm OD x 3mm Wall Thickness
90mm OD x 3mm Wall Thickness
100mm OD x 3mm Wall Thickness
101.6mm OD x 6.35mm Wall Thickness
150mm OD x 3mm Wall Thickness
I think the 100 x 3 is a best guess. If it is too big, some pieces of the same could be softened and fitted inside to make the wall 6mm thick. There seems to be no option but to guess.
The magnet selection is also trial and error. I have made, and currently testing, a rotor made with 10mm x 10mm N35 cylinder magnets. This does not run as well as the ferrite rotor, and neither has reached levitation.
I think the best possibility may be 15mm dia x 20mm long N52 magnets fitted to the existing 36 pole 27 coil position Torus.
To substantially reduce weight AL magnet wire could be used. I only have a good supply of CU magnet wire so will continue using that.
Of real interest may be to construct the outer Torus of material which is claimed to be from a crashed UFO. Analysis shows this to be mostly magnesium, with small amounts of layered bismuth, and some zinc. This could perhaps be somewhat replicated using 3D metal printing. Mg is about 64% of the weight of Al, so the machine may be substantially lighter and have better rotor levitational properties. I have read that this material will levitate when subjected to terahertz frequencies. I have not heard of anyone exposing it to fast moving magnetic flux.
There are issues with ignition and flammability of magnesium, and availabiliy, which make it problematic for me.
It should be noted that the Torus machine I am experimenting with is based on the solid science and engineering of electro dynamic suspension. Like the Inductrac train on which it is based, the rotor / train must levitate when the correct combination of magnets, outer shell materials, dimensions and speed are found.
The only area of scientific conflict is what will happen when the levitating rotor speed is substantially increased.
Another build possibility is to increase the number of supermagnet poles for the same circumference machine. The current project could be fitted with a rotor having 72 radial supermagnet poles. Possible coil positions would increase to 54.
As an added benefit, no extra magnets would be required, and no extra weight gained, but a doubling of both available magnetic flux and rate of change of flux polarity would be obtained.
72 Magnet poles. 54 Coil Positions. 650 Diameter. Two 15mm dia x 10mm long magnets makes one 20mm long supermagnet. The space between supermagnets is 7.7mm so 15mm x 5mm long magnets will fill the gaps nicely. Could use the existing Torus shell by winding intermediate coils and relocating one of the hall sensors. This will mean a doubling of the switching speed. I have tested to see if there is any loss of field strength / distance using 10mm dia supermagnet radial pole magnets at both 26mm & 52mm spacing, and there is no appreciable difference.
However, tests using readily available N50 grade 15mm diameter magnets show a noticeable drop in flux strength at tube wall distance with 26mm spacing.
But at 10mm from the magnet the flux is both strong and identical at 26mm & 52mm spacing. This seems to indicate that using the closer spacing is a good choice, as it will be more likely for the rotor to levitate at a lower speed. It should be possible to fit an AL packer say 3mm x 15mm in the invert of the outer torus between the wheels, bringing the rotor a bit closer to the non magnetic conductor, and therefore a little more likely to levitate.
The most important thing is to get rotor levitation. I believe that when levitation occurs, there should be a weight reduction of the overall machine, at least equal to the weight of the rotor.
The magnet retainers could be made from 20mm x 3mm wall AL tube about 4mm longer than 2 magnets. This would need boring out to 15mm for a loose fit, and turning down the ends to about 1.25 wall thickness to bend over and hold the magnets. More work but less likely to come apart.
The rotor could use an AL 26" bike rim. 20mm tube fits nicely into the rim. 2 layers of AL sheet metal around the outside of the magnets, and some fiberglass should hold it all together.
The only problem I see is that the Hall Sensors may be difficult to adjust with only about 26mm between alternating magnet poles. They may need relocating inside the torus shell.
Another problem with closer spacing of the magnets is the fabrication of the outer torus. My build was tedious and difficult to weld. The shape is lobsterback. More coil walls and less space means that casting may be my only option for further development. A mold of say 6 sections, when cut in half would make 2 complete molds. The molds would need to be made of steel for multi use. Each casting would be about 450 grams, so only a small forge would be required.
Welding the castings together would be fairly simple.
A lobsterback torus made with 54 sections would have a much more circular or smooth internal surface than 27 sections. I have no idea which might be better for creating electrodynamic suspension.
