This invention relates to a fluid film bearing and a process for evaluating pneumatic hammer instability. Based on the stability criteria, a method was developed using 3D CFD (computational fluid dynamics) evaluation dynamics to evaluate the onset of pneumatic hammer instability in a fluid-film bearing. The pneumatic hammer instability criteria and the 3D CFD evaluation processes are available for use on any programs using bearings operating with a compressible fluid.
Design tools that can accurately predict the performance of fluid-film bearings are critical to the successful implementation of this technology. Bearing designers have generally relied on their “rule of thumb” to evaluate bearing stability. High-speed turbopumps of the future will require fluid-film bearings operating at a higher Reynolds number and incorporating 3D effects, which makes current pneumatic hammer instability criteria obsolete. An accurate computer program accounting for the major geometric and fluid effects within the bearing is necessary.
Based on stability criteria, in order to maintain stability, the following inequality must be true:
(α + β)/Θ > q/s
where α is the variation of mass flow rate into the recess pressure, at equilibrium; β is the variation of mass flow rate out of the recess with recess pressure, at equilibrium; Θ is the variation of mass flow rate out of the recess with annulus height, at equilibrium; q is the variation of the fluid mass in the bearing (recess plus lands) with respect to recess pressure, at equilibrium, and s is the variation of fluid mass in the bearing (recess plus lands) with respect to annulus height, at equilibrium. In a journal bearing, Θ and s are now related to a change in eccentricity instead of annulus height. These criteria can only be evaluated numerically. The subject invention describes a process developed to evaluate the criteria using a 3D CFD program. Submodels are used due to the difficulty in defining certain boundary conditions in a full 360º model.
In the process, two CFD submodels are used with a first model to evaluate α and a second model to evaluate β. To calculate α, it is necessary to use the top half of the bearing 3D model. It is necessary to impose a periodic boundary condition on both the bottom and top models. For the bottom half of the model, the inlet boundary condition is set using a total pressure profile solved from the original 360º solution. The 360º model is used to evaluate the effect of a changing eccentricity. After α and β are found, q and s are determined. The equations are then re-solved for the same cases where shaft rotation was neglected in order to evaluate the effect on flows, pressures, and fluid mass.