Three nominally-identical, 4-pad flexure-pivot-pad bearings (FPBs) were manufactured, varying only in their (dimensionless) pad preloads, namely, 0.264, 0.511, and 0.695. A hybrid hydrostatic bearing (HBB) was also manufactured with the same nominal length, diameter, and clearances. Water was used as the test fluid. The FPBs were tested at the following three constant supply pressures (Ps): 0.689, 2.76, and 5.52 bar. The HBB was tested with a supply pressure that increased linearly to these same three terminal pressures. A magnetic bearing was used to load all bearings with an applied unit-load profile that increased linearly with time to reach the following two peak values: 0.745 and 1.38 bar. The design speed ωd was the maximum planned speed for a test run, and it was varied over 3, 3.5, 4, and 4.5 krpm. The tests aimed to produce a linearly increasing speed profile, ω(t) = At, before reaching ωd. For the HBB, bearing Ps also nominally increased linearly with ω.
At lift-off, the shaft leaves the bearing surface during the shaft’s angular acceleration and remains separated and supported over a finite time and speed range. In some circumstances, FPB bearings lifted off and were then forced back into contact with the shaft due to the linearly increasing applied load. Once lifted off, the HBB always remained separated from the bearing surface.
The peak load capacity was the maximum load supported by a bearing once lifted off (even if it was subsequently forced back into contact).
FPBs have been used successfully in commercial turbomachinery handling low viscosity fluids. The results reported here indicate that preloads of m = 0.264 and 0.695 have comparable lift-off and peak load-capacity performance, substantially better than the m = 0.511 bearing. The FPB data also show a surprising steady drop in lift-off speeds and peak-load capacity with increasing Ps values, presumably because of end seals that were provided. From these results, an FPB should be used with end seals and preloads of m = 0.264 or 0.695.
The HBB lift-off ω values also dropped with increasing terminal Ps values. For example, at ωd = 4.5 krpm, an increase in terminal Ps from 0.689 bar to 2.76 bar dropped lift-off ω from ∼3 krpm to ∼ 300 rpm. The supply pressure provided to the HBB increased linearly with time and, consequently, also increased linearly with ω. In reality, a centrifugal pump Ps would be proportional to ω2. To the extent that HBB lift-off ω performance depends on the supply pressure at lift off, lift-off ω and load-capacity performance of the present HBB would be worse with Ps ∼ ω2, since the same required lift-off supply pressure would occur only at a higher ω.
Except at ωd = 3.5 krpm, the FPBs had better (lower) liftoff speeds than the HBB. The HBB tested maintained a consistently better peak-load capacity than the FPB. In some cases, the shaft lifted off the FPB and then returned to contact. In all cases, the HBB remained in a lifted-off condition.
As a final conclusion, the present data provide no clear basis for choosing between the FPBs and the HBB.