Is there any relation between this limit and the accuracy of the numerical prediction of fluid-film bearing dynamic coefficients?
I've already heard about smaller limits, even under 0.5. However, I'm not quite sure if this is a practical or theoretical limit. Does it have any scientific basis?
Before we decide anything about fluid-film bearings and if we do not have a computer program for the calculations we can run manually the solution of Reynolds equation as given by Raimondi and Boyd. This example with its tables may be found in all books of Machine Design.
By saying that we can run manually the solution of Reynolds equation, you mean only for short bearings, don't you? Actually I'm looking for a recommended limit for the eccentricity ratio of general fluid-film bearings, mainly cylindrical plain bearings and elliptical bearings. The numerical simulations allow any value between zero and one. However, they aren't very accurate for high eccentricity ratios.
I've implemented the program in Matlab by using the FEM, but it seems that the numerical predictions for high eccentricity ratio aren't so reliable. In practical application do you recommend operation above 0.7 or 0.9?
I would run the program for at least three values 0.3 0.5 0.7 to have curves for the parameters I am interested in and see how they come up. Then, I would choose the most effective values of the parameters and decide for the most promising value of eccentricity ratio.
To answer on practical point of view, the relative eccentricity of the shaft in a cynlindrical plain journal bearing could be close to 1, but of course less than 1 (without considering bearing deformations). The limit is not on the eccentricity but on the minimum film thickness (which depends on the relative eccentricity of the journal). There are given values in the ISO Standards. For safe operating conditions under hydrodynamic regime, the minimum film thickness should be (much) greater than 3 times the combined surface roughness of the journal and the bushing.
Thank you so much, Dr. Michel! This practical view is very important when dealing with simulations, to avoid irrelevant results.
On the other hand, from the theoretical point of view, what do you think about the traditional way of calculating the dynamic coefficients of plain journal bearings (numerical solution of Lubrication Equation + small perturbation theory) ?
The linear dynamic coefficients obtained based on the small perturbation theory are very useful and convenient as long as the shaft orbit or trajectory is relatively small. For medium and large shaft orbits, it is necessary to use a nonlinear modeling.
- Badges
- Science topic
- Similar topics
- Engineering
- Thermal Engineering
- Thermofluid
- Fluid Mechanics
- Fluid Dynamics