| dc.description.abstract | The dynamic and static performance of a flexure-pivot tilting pad bearing is presented at
a load between pad configuration for various load and speed combinations. A similar
work performed on the same bearing at lower loads ranging from 0-1 MPa (0-150 psi) by
Al-Ghasem was tested, whereas the current work investigates effects in the load range
between 1-2.2 MPa (150-320 psi). The bearing design parameters include: 4 pads with
pad arc angle 72º and 50% pivot offset, pad axial length 0.0762 m (3 in), pad radial
clearance 0.254 mm (0.010 in), bearing radial clearance 190.5 µm (0.0075 in), preload
0.25, and shaft nominal diameter of 0.11684 m (4.600 in). An important distinction
between the two sets of tests is the difference in experimental bearing radial clearance,
which for this case measured 208 µm (0.00082 in), and for Al-Ghasem’s was 165.1 µm
(0.0065 in). The rotordynamic coefficients are determined experimentally using a test rig
equipped with motion and load sensors. The rig is modeled using Newton’s laws, which
is converted from the time to frequency domain using Fourier Transform to give complex
dynamic stiffnesses. From the resulting complex dynamic stiffnesses the associated real
and imaginary components are plotted as a function of excitation frequency and curve
fitted via linear regression to give the rotordynamic coefficients. The primary objectives
were to determine whether the real component of the complex dynamic stiffnesses could
be better modeled with or without the mass coefficient and to contrast the rotordynamic
coefficients with an analytical model. Only in the load range of 1 to 2.2 MPa were the
unloaded direct mass coefficients near or at 0, which would allow for a [K][C] model to
be used. The remaining real components are better represented with the mass term. The
analytical model generally overpredicted the stiffness, damping and mass coefficients,
especially for the direct components; the trends were generally consistent. | en |
