Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Bearings for 200,000 rpm and up

  1. Mar 20, 2014 #1

    we are interested in high-speed bearings -- 200,000 rpm and up
    bearing size:
    inner diameter: 0.5cm - 1cm
    outer diameter: up to 3cm
    thickness: up to 1cm

    Many sites offer high-speed dental bearings, but they are quite small in size...

    Would anyone have an idea about a site/catalog for such bearings.
  2. jcsd
  3. Mar 20, 2014 #2


    User Avatar
    Science Advisor

    The forces and friction on balls or rollers in larger bearings at those speeds will be destructive.
    You should consider magnetic or fluid bearings.
    What are you spinning, how big is it, what drives it and how long must it operate?
  4. Mar 20, 2014 #3


    User Avatar
    Gold Member

    At 200,000 rpm for a device requiring a 1cm inner diameter, the material constraints are stiff.
    Magnetic bearings in a vacuum seem the most plausible, but data will be hard to come by, because this is enrichment centrifuge territory.
  5. Mar 21, 2014 #4


    User Avatar
    Gold Member

    "In this study, we presented the design procedure and experimental results of the rotor and bearings for 500 W class MTG with the rated speed of 400,000 rpm. The rotordynamic analysis model for MTG rotor is reliably constructed by using impact test for the test rotors. The motoring test rig to evaluate the radial air foil bearings is fabricated, and the motoring test was performed up to 280,000rpm."

  6. Mar 21, 2014 #5
    My thinking is, since dental bearings are good for up to 500,000 rpm, there must be larger ones with the same speed rating...
  7. Mar 21, 2014 #6


    User Avatar
    Science Advisor
    Homework Helper

    I think there is less to that than meets the eye. This is a poster presentation at a conference, not a peer reviewed published paper (not even a paper presented at the conference), and the research has nothing to do with Imperial College London, except one of the conference organizers happens to work there. http://cap.ee.ic.ac.uk/~pdm97/powermems/2011/

    The way I read it, they tested three designs of rotor by measuring one vibration mode at zero RPM, and then tweaked the rotor model to give exactly the same frequency (table 1). That doesn't give you any real validation of how the dynamic behavior changes with RPM, especially considering they estimate that the accelerometer perturbed the vibration frequency by about 10% (table 2). Why didn't they use a non-contact method to (e.g. an LDV or even a capacitor proximity probe) to measure the unperturbed frequency?

    They tested the rotating structure, but they could only run up to 70% of the design speed (or 60% of their calculated critical rotor speed) because of high vibration levels. On the basis of that result, they claim the bearings worked OK.

    That's a bit like saying "we designed a table to carry a load of 1000 pounds. We could only test it up to 500 pounds because the table top started bending, but the legs worked OK."

    Nice try, but no cigar :smile:
    Last edited: Mar 21, 2014
  8. Mar 21, 2014 #7
    Turbochargers have shaft speeds on that order. Don't know much in terms of specifics, but it may be worth a look.
  9. Mar 21, 2014 #8


    User Avatar
    Science Advisor

    If it was load that you were carrying you could simply stack dental bearings along the shaft. So I assume that you need bigger bearings because you are transmitting torque along the supported shaft through the centre of the bearing. Power = torque * RPM, so by using higher speed you can reduce torque for the same power, which reduces bearing size. Ball bearing RPM is inversely proportional to diameter. At 200k RPM with a 10 mm diameter, bearing surface speeds will be around 360 km/hr. You are getting too close to the edge of the envelope for reliable cool intermittent operation of ball bearings.

    The bearing operational envelope decides what is available. Bigger dental bearings will not need to be made if there is a better bearing technology for those dimensions. Each technology develops it's niche, there are so many dental bearings made that they have finally optimised the design. Dental bearings are not rated for continuous operation at those speeds. Outside the dental sized ball bearing there will be very little experience and so they will be unreliable. It will depend on the duty cycle and longevity you require. I suspect you may be looking at speeds and powers that will require either magnetic bearings or bearings separated and cooled by hydrogen gas pressure.

    There is also the problem of balancing the rotor and of rotor vibration. Magnetic and fluid bearings have a much greater tolerance for rotor asymmetry than do ball bearings.

    Without a better idea of the application, power and geometry, it is hard to narrow down the search or invert the design.
  10. Mar 28, 2014 #9
    We got one some time ago, it was rated for up to 220,000 rpm, but never ran past 100,000 rpm :smile: It didn't have bearings, but one lubricated bushing, which is some sort of a solution at these speeds :)
  11. Mar 28, 2014 #10


    User Avatar
    Science Advisor

  12. Mar 29, 2014 #11
    We could manage to find a way to decrease the rpms into a range where plenty of bearings are available (say, below 20,000)
    Sorry haven't been able to keep up with all questions
    Thanks to all who contributed
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook