I suspect that it might be theoretically possible to build magnetic bearings with a strong enough field for this application, but you'd probably need superconducting magnets and the power source would be enormous. The several tonnes of additional equipment and fuel to power it would far overwhelm the miniscule efficiency increase of the axles. As Russ mentioned, you'd be better off streamlining where possible and make sure to keep your normal bearings properly lubricated.

The area of magnetic ball bearings is on going. We are always looking at that stuff for our engine applications. The thoght of having a highspeed, low heat generating bearing that does not require an oil system is a home run if we could get one. So the usage in our case is indeed justified by weight savings and increased performance.The major problem now is that the system is complicated because of the varying loads imposed during operations and the size of the set ups. I have not heard of one for a car, but I can see how that would probably be the easiest scenario to implement.

I am using solid neodymium magnets of 5300 gauss or .53 telsa (ithink) that are cylindrical 0.01m*0.01m and was wondering what equation i could use to work out the force acted upon the first nsk bearing (0.01m radius, 0.00351kg) in the series, which was given no extra velocity by me, only the velocity that the magnetic field acted upon it ( this ball bearing is attracted by the magnet when 0.02m away from it). 

I also would appreciate it if someone could explain how i could give/find the Potential Energy Difference of the first ball bearing (which would be at 0m due to its close proximity to the magnet) and the Potential Energy of the last Ball bearing in the system which changes from a displacement of 0.02m - 0.04m from the magnet (in the sereis of test due to a differnt number of 'buffer ballbearings). With the Kinetic energy gained from the 0.02m of acceleration of the first ball bearing (by the magnet) i could then subtract from it the Magnetic Potential Energy of the ball bearing to be launched to discover the total energy impacted onto the last ball bearing in the series (and do this for Displacements of 0.02,0.03 & 0.04m)

If there is no way to calculate the first part concerning the velocity of the ball bearing, i may simply time how long it takes for the ntn bearing to reach the magnet and use SUVAT to calculate the velocity* however the Magnetic potential energy part is still eluding my knowledge of physics and is my greatest priority. However if an equation or series of equations exists for the frist part, it would be helpful as it would be introduce a lesser degree of error than using SUVAT

20206

ER-16 bolt / right axis

ER-18