Power losses in high-speed gears come from the friction between the teeth (sliding and rolling), the lubrication process (dip or jet lubrication), the pumping of a gas-lubricant mixture during the meshing and the losses associated with windage effects. The objective of this paper is to present a number of preliminary experimental and theoretical findings on the prediction of windage losses. Experiments were conducted on a test bench whose principle consists in driving a gear to a given speed and then measuring its deceleration once it has been disconnected from the motor. The transmission between the motor and the rotor is ensured by a friction wheel which also plays the role of speed multiplier. A pneumatic jack either imposes a sufficient contact pressure between the driving wheel and the rotor (transmission of rotation), or is used for separating the parts when the maximum speed is reached. A disk and 4 different gears were tested in the absence of a lubricant at speeds ranging from 0 to 12 000 rpm. Two different theoretical approaches have been developed: i) a dimensional analysis based upon the dimensionless groups of terms which account for the flow characteristics (Reynolds number), the gear geometry (tooth number, pitch diameter, face width) and the speed, ii) a quasi-analytical model considering in detail the fluid flow on the gear faces and inside the teeth. It is found that both approaches give good results in comparison with the experimental evidence and two analytical formulae aimed at predicting windage losses in high-speed gears are proposed.

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