Ultra-Low Friction Pivot Bearing?

Carbide tool inserts can be had for under $3; but you need both a flat and an edge, and they cannot be steel (magnetic forces are uncontrolled) nor should both be carbide (because the carbide is bonded with cobalt metal, and cobalt is magnetic). Hey, you DO want to control all the forces, nicht wahr?

I thought that you wanted it at the pivot of your pendulum (the only reason making sense to me). You really should use the twisted wire method.

Nick

Doug, et al,

If the goal is measuring the inertia of irregular parts, my experience is that you can do this without "nearly frictionless" bearings. Google the phrase, "trifilar pendulum" and consider using one to make your measurements.

A trifilar pendulum is basically three (ideally long) strings/wires supporting a hanging platform on which you place your part. Usually both the upper and lower ends of the wires are equalateral triangles. (Our platform was basically triangular, too; this simplifies things, IMHO) The part is placed with its mass center above the mass center of the hanging platform, and the whole thing is given an initial small twist and released. If you've set it up right, you get a gentle angular oscillation. From a measurement of the (rotational) period of the pendulum (and a few other masses, etc.), you can calculate the inertia. In my old lab, we did the period measurements with an optoswitch and a computer to record the timing. Averaging many periods helps get a nice accurate measurement.

Intuition to the contrary, the friction at the pivots (or air) will make next to no difference in the rotary period -- unless it's so bad that the oscillations die out too quickly to measure the period. We never had a problem with such losses. (Google will pull up some analyses that detail this, along with the formulas needed to calc inertias from the mass and period measurements.)

The mass of the platform is needed. You'll also need to find the center of mass of the part for each axis about which you want the inertia. You can find this pretty easily using a scale and a knife edge, by measureing reactions. It's a good idea to check that the plaftorm's mass center is where you expect it; again, pretty easy using a knife edge and a scale. Note that everything's steady during the reaction measurements, so an ultra-low-friction knife-edge isn't that critical.

It's helpful to be able to calibrate the pendulum, by adding a known inertia or two. We did this by making it easy to put ball bearings (uniform weight, easy-to-calc inertias) on our platform in a triangle, via sets of holes punched in the sheet metal.

HTH / Let me know if you'd like more details.

-- Larry Pfeffer

others have mentioned steel on agate, i would also suggest sapphire or ruby...

snip--

the trifilar pendulum might the best suggestion for this

a question arises: to what accuracy do you need the measurement?

for example: i have performed various experiments to measure the superfluid fraction of helium in various geometries

we were measuring changes in the period of perhaps a part in a billion, using tempered Be-Cu torsion rods in the form of hollow cylinders, capacitively driven and detected.

these measurements were made at low temperature (millikelvin) using several stages of vibration isolation, vacuum jackets, and refrigeration.

your requirements may differ...

in other applications i have used steel points on ruby, (superconducting motors designed to operate at millikelvin, with no more than one erg of heat dissipated per 3 degree step), steel on teflon (teflon gets quite rigid at low temperature) and steel on steel

at room temperature surprisingly good results can be attained with oil films (a dead weight tester is a good example) or magnetic suspension (as in a turbomolecular pump)

sidd

I'm considering this, th eonly tricky part is making the platform so that I can easily attach the test objects.

Doug White