Hydrodynamic bearings have a key role in the functioning of heavy duty gas turbines: they join a great vibration absorption with an efficient power dissipation by means of a film oil inserted between the turbo machine axes and the bearing case. A classical approach for studying the functioning and the performance of this kind of bearings is to solve the so called Reynolds’ equation, which is obtained from the Navier-Stokes equations under simplifying assumptions. As a result the pressure field is derived, the fluid film being considered isothermal: dissipation effects have to be estimated a posteriori in a postprocessing procedure. On the other hand a fluid environment having to be taken into account, a direct approach is carried out by the time consuming CFD analysis. After defining an appropriate mesh and choosing the appropriate solver, an almost exact solution of the entire flow field is obtained, that is pressure, velocity and temperature distributions. The main drawback is that the required time is several order greater than that required for the solution of the Reynolds’ equation. In the present work an alternative strategy is proposed, which consists of an iterative procedure: at each step the Reynolds’ equation is solved in order to obtain the pressure field; a 1D energy balance is then applied along the length of the bearing for computing the temperature field. In this way the close relationship between pressure and temperature is modelled, the former depending on the oil viscosity locally changing with temperature, and the latter depending on the local oil mass flow and on dissipated power strictly correlated to the pressure distribution. The upgrading of both the entities ends when the convergence is reached. Comparisons with literature test cases reveal the efficiency of the proposed technique: treating the interaction between pressure and temperature gives a solution which is very close to industrial configurations investigated, and at the same time the computational load is as light as that needed for the solution of the only Reynolds’ equation. The performance of the above coupled solver can be greatly emphasised applying it to the bearing design. An integration with a multi-objective genetic optimization process is proposed, taking as objects to be optimized both geometrical and environmental variables. Application examples are shown about an industrial Ansaldo Energia lemon bore hydrodynamic bearing: given a currently applied configuration, possible improvements are suggested. Results are presented.

This content is only available via PDF.
You do not currently have access to this content.