Numerical Simulation of Complex Air Flow in an Aeroengine Gear Box

Abstract

The internal gear box (IGB) of an aeroengine represents a severe challenge in computational fluid dynamics (CFD). In the present study, an axisymmetric CFD model was assessed to investigate the complex internal air flow in an aeroengine IGB. All the non-axisymmetric components and geometry features inside the gear box, such as bearings, gears, bolts and slots, as well as the radial drive system and vent pipes, were simulated using porous media models. Their flow resistance was estimated either by empirical correlations or by preparatory CFD studies and comparison with measurements. To evaluate the CFD technique adopted in the present investigation, a separate bolt windage study was conducted using a similar axisymmtric CFD model with the porous media approach. Good agreement of the bolt windage with other workers’ rig test data was observed. The present application of the porous media approach into a complex gear box flow represents a first attempt to use state of art CFD to assist an industrial design. Both maximum take-off (MTO) and ground idle (GI) running conditions were investigated. The complex flow patterns in the gear box were obtained. The results show a similar dimensionless performance of intermediate pressure (IP) and high pressure (HP) gears between the two operating conditions. For the present gear box arrangement under investigation, the CFD results suggest that the airflows induced by the HP gear and HP bearing are higher than their IP counterparts. A comparison with power absorption rig test data for the similar HP crownwheel in isolation shows that an assumption of pressure loss coefficient of 10 for the porous media of bevel gears may be appropriate, as the HP gear torque coefficient obtained in the CFD prediction is equal to 0.05, very close to its expected value. In addition, the effects of an assumed stationary IP gear and a large seal clearance on the HP gear performance were also investigated. The numerical results show that their impacts are insignificant, probably due to the strong pumping effects of the HP gear. Further discussion on the possible influence of the airflow on the oil motion within the gearbox and assistance to improve the traditional internal airflow models used for bearing chamber sealing analysis was also made. Three dimensional geometry modeling and inclusion of the oil phase are considered feasible. Such further investigations would aid the understanding of the interaction between the induced airflow due to the rotating components and oil motion, and their impact on oil scavenging behaviour and ‘windage’ contribution to heat oil.