friction coefficient of the contact was then computed as

Figure 2。10 illustrates this procedure used to remove bearing traction。 In Figure 2。10(a), a raw      measurement at a given      is shown。  As both      and      constitute this     signal, the traction data is not skew-symmetric about the origin of the graph at

R    。  In Figure 2。10(b), the corresponding      measurement (at            ) within  the

same R range (representing the effect of the variation of      ) is shown。   Here,             is

observed to increase linearly with increasing R (or with increasing )。 Finally, Figure 2。10(c) represents the roller contact traction curve that is obtained by using Figures 2。10(a) and (b) according to Eq。 (2。4)。

According to Figure 2。6(b), the test machine with the specified roller geometries was capable of a maximum contact stress up to 2。6 GPa。  A maximum contact stress

Figure 2。10  An example set of measured (a)      , (b)     , and (c) the resultant   

calculated from (a) and (b)。

value of 2。0 GPa and an intermediate value of 1。2 GPa were selected as the load conditions in traction tests。

The lubrication system was turned on initially at a flow rate of 1 lpm to heat the specimens to the set test temperature value, while the test specimens were rotated slowly under no load during this heating period time。 Once the specimens reach the desired initial surface temperatures, then the traction tests described above were performed at a test flow rate of 0。5 lpm。

2。3。2 Wear Tests

The wear test procedure was defined with multiple interim inspections。 A maximum test duration of 30 million roller cycles was considered。  This   corresponded

to 70 hours of testing at                 m/s and R   —      。  This high-speed and high-sliding文献综述

condition was found to be suitable for inducing wear within the total loading cycles

defined。 The largest load increment corresponding to a maximum contact stress of 2 GPa was utilized by the same reasoning。 Interval for interim inspections  was set initially at 10 million cycles, while in certain tests, this was reduced to every 2 million cycles to more accurately document the wear rate。 A suspension criterion of maximum wear depth of 25 µm across the profile of either the disk or roller was adapted。 Any wear accumulation beyond this value within the first 30 million roller cycles was not allowed。  The lubrication flow rate was set at 0。5 lpm for the duration of the wear  tests,

and was maintained at this level。 Oil inlet temperature was maintained at 9 ª。

2。3。3 Scuffing Tests

The scuffing tests were performed under two different conditions: (i) fully lubricated and (ii) starved。 Both types of tests were carried out the same way, with only lubrication conditions being different。 Fully lubricated tests had an oil flow rate of 0。5 lpm, while the flow rate for the starved condition was limited to 0。01 lpm。 In order to achieve such a low flow rate, the pump motor was turned to its lowest setting and the nozzle was restricted as to create a small drip periodically。

Before testing began, the specimens were brought to a constant surface temperature under the same procedure as the traction tests。  The tests were performed at

         m/s   and   R  —      , again to create a challenging contact environment。   The