Prediction of the pressure–flow characteristic of a Belleville washer-based damper valve

Abstract

Flow control valves used in automotive suspension dampers are almost universally based upon the restriction of flow using Belleville washers. These washers can best be described as a non-flat washer of conical shape and uniform thickness. They are also known as disc springs and are commonly used for load bearing in which their compactness and ability to produce a wide range of load–deflection characteristics provide an alternative to conventional coil springs. Due to the non-linear behaviour of Belleville washers, and the complex flow paths in an automotive damper valve, it is difficult to predict the resulting pressure–flow characteristic. As a result, damper valve design is often considered to be a ‘black art’ best carried out by an experienced engineer using an experimental-based trial-and-error approach. Clearly, without the aid of an analytical tool to predict the behaviour of prototype systems it is difficult for the designer to fully exploit the functionality of a particular design. The current paper describes theoretical and experimental studies undertaken to investigate the pressure–flow characteristic of a Belleville washer-based damper valve. Existing load–deflection theory was extended to account for the hydraulic pressure acting on the surfaces of the washer. However, poor agreement was obtained between the measured pressure–flow data using a uniform pressure distribution across the washer diameter. For this reason the model was developed further to include the effect of a non-uniform pressure distribution within the valve. Computational fluid dynamics software was used to predict the pressure regime within the valve and this, in turn, was used to calculate the forces acting on the surface of the Belleville washer. The improved model was found to match the experimental behaviour with good accuracy, both for a range of spring pre-load conditions and for applications where several washers are stacked up in parallel. However further work would be required to develop a reliable simulation tool as an element of trial and error was necessary to estimate some simulation parameters.