WIT Press


Numerical Optimization Of A Distributor Valve

Price

Free (open access)

Paper DOI

10.2495/OP050321

Volume

80

Pages

11

Published

2005

Size

936 kb

Author(s)

L. Dahlén & P. Carlsson

Abstract

In this paper, a non-linear optimization method is used to improve the design of a distributor valve. The distributor valve is an important component in a radial piston hydraulic motor, and optimization of the design to minimize power losses is an interesting way to increase efficiency. The main function of a distributor valve is to supply the pistons with a pressurized flow and to return oil during rotation. At the same time the distributor valve acts as an externally pressurized lubricated thrust bearing, in order to separate the rotating parts from the motor case. The bearing acts as a hydrostatic annular multi-recess plane thrust bearing, with different recess pressures. The separating force of the bearing is balanced hydrostatically by the pressure that is applied and springs. Losses will occur in the contact between the parts in the distributor valve, due to friction and leakage. This paper shows that modern optimization methods can be used as an effective tool to create new designs and to modify the existing design of the bearing surface geometry of the distributor. A finite element method has been used to simulate the contact, and the program is linked to an optimization routine to perform the optimization. The results of the optimized design show a significant decrease in power loss, compared to the existing design in the operating range. Keywords: radial piston hydraulic motor, simulations, optimization, power loss. 1 Introduction Hydraulic pumps and motors are expected to work under a wide range of operating conditions, and finding the optimum design of components can be difficult. When optimizing a hydraulic machine for maximum efficiency, an accurate model of the physical behaviour governing flow and torque losses is

Keywords

radial piston hydraulic motor, simulations, optimization, power loss.