Figure 2。1  Twin disk test machine used in this study。

Figure 2。2   A top view schematic of two-disk test machine used in this study [17]。

speeds, and  (in rad/s), to achieve any rolling and sliding speeds at the roller contact interface。

The disk-side drive shaft was rigidly supported by a pair of rolling element bearings within their housing。 The roller shaft was supported in the loading fixture by a pair of high precision ball bearings on one side, and one needle roller bearing on the other。 Figure 2。3 shows a close up view of this arrangement。 The roller shaft was threaded on one side to receive a washer and locking nut to axially fix the position of the roller。 In its nominal unloaded position, the roller was located at a very small distance from the disk with no contact。 The loading lever shown in Figure 2。3 pushed the roller and its shaft towards the disk to initiate contact, in the process, offsetting the roller shaft axis from the roller-side drive shaft。 In order to accommodate this offset while loading, a flex-coupling (helical spring type) was used to connect the roller shaft and the roller-side drive shaft as shown in Figure 2。3。 The disk was shrunk fit onto the shaft and a locking nut was tightened to prevent axial motion。 This shaft was rigidly supported by four angular contact ball bearings as well as a needle thrust bearing。

A small, low-flow-rate lubrication system was used with the tests。 The oil was heated inside the reservoir and pumped through flexible hosing into the test area。 The maximum obtainable temperature for the oil was 150˚C。 Any constant flow rate was possible within the flow rate range of 0。4 to 3。3 lpm。  Much lower flow rates, less   than

0。1 lpm were also achieved by restricting certain sections of the hose and manipulating the nozzle。

Figure 2。3 Close up view of the test pair, roller and disk shafts, roller-side flexible coupling and the loading arm mechanism。

The loading arm applied a normal force through a controlled proportional valve leading to a pneumatic cylinder。 The pressure regulator provided the cylinder with a pressure with up to 550 kPa, corresponding to a normal load of 4,450 N。 This force was monitored by a button head-type load cell with a threshold of 1,112 N, which was inserted at the end of the loading arm, in a small recess。  A torque-meter with a capacity

of 3。5 Nm and ±   % resolution of full scale was utilized between the disk shaft and

its motor through two helical style flexible couplings to measure torque within the

system。 The same LabVIEW program used for controls also monitored and  recorded the torque throughout each test。

Two K-type thermocouples were used to measure the bulk surface temperature of the specimen。 Mounted on the test box cover through a drilled out bolt with an integrated spring, this arrangement allowed for the thermocouple tips to come into contact with the test specimen, yet not induce a measureable amount of downward force。 A third thermocouple was placed inside the oil reservoir to measure supply temperature。 The temperature measurements were also monitored and recorded in the LabVIEW program。

In order to contain heat and add locations for thermocouple placement, as well as to contain oil spray and maintain safety, a safety cover was built。 The cover was a sheet metal frame with Lexan windows。 This cover not only contained the test specimen, but also the flexible coupling for the roller shaft。 Separate safety covers were in place to shield other rotating components such as belts, sheaves, and couplings。

2。2 Test Specimens

The test specimens were comprised of two cylinders。 The roller had an outside radius of  r                            87   mm, with no lead crown and a face width of 7。6 mm。  The   disk,