H. Heshmat et De. Brewe, PERFORMANCE OF A POWDER LUBRICATED JOURNAL BEARING WITH WS2 POWDER - EXPERIMENTAL-STUDY, Journal of tribology, 118(3), 1996, pp. 484-491
Previous investigations have demonstrated the principle and feasibilit
y of quasi-hydrodynamic lubrication using powders. With conventional l
ubricants, the hydrodynamic behavior of a journal bearing can be summa
rized in the form of a plot of friction coefficient against a nondimen
sional load parameter usually referred to as a Stribeck curve. One of
the main features of the curve is that, as the thickness of the lubric
ant film increases, the friction coefficient initially decreases until
it reaches a minimum and then increases due to shearing of the lubric
ant film. Previous tests on a three pad journal bearing using molybden
um disulfide powder ns the lubricant showed that its friction/load cha
racteristics followed the Stribeck form. The experimental work reporte
d herein describes further tests carried out on a three-pad journal be
aring to evaluate its thermal stability and long teem went characteris
tics using tungsten disulfide powder as the lubricant. Dry tungsten di
sulfide powder lubricant has been shown to provide excellent wear resi
stance. Its friction properties are similar to molybdenum disulfide bu
t it can be used at much higher temperatures. Bearing tests were carri
ed out nt speeds up to 30,000 rpm and lends Lip to 236 N. Over the loa
d/speed, range stable thermal operation was achieved From friction tor
que measurements it was found that the coefficient of friction for the
bearing was essentially constant over the range of lends and speeds.
Unlike liquids, powders have a limiting shear strength property. When
the limiting shear stress is reached in a powder flow, the powder esse
ntially shears along a plane and does not pow in a viscous manner. The
constant friction coefficient seen with the tungsten disulfide powder
suggests that the powder had a low limiting shear strength and that t
his dominated the operation of the bearing in the rests. The total acc
umulated test time was more than eighteen hours. Periodically, the tes
t journal and bearing pads were inspected but there was no measurable
wear or visible damage.