A model for the natural vortex length in centrifugal separators

The aim of this project is to formulate a knowledge-based model for the "natural vortex length" in reverse-flow centrifugal separators, to replace the well-known and much used, but seriously inadequate, empirical model of Alexander from 1949. The "natural vortex length" is the length between the gas outlet and a spontaneous instability that may exist low in the separator known as the "vortex end". The model will be based on fundamental mathematical modelling in combination with an experimental research pr ogram. The mathematical model must describe instability in a confined vortex flow, and the existence of two possible flow patterns, both of which may be stable under certain operating conditions: a) a straight vortex core, coinciding with the cyclone axi s and b) a vortex core bending to, and precessing around, the wall of the separator. For the theoretical description a special technique, known as "continuation methods" will be applied, and the theory will be used to crystallize out a simple and useabl e, but knowledge-based model for the vortex length. The experiments will cover the relevant ranges of a number of operational and configurational parameters known from earlier research to be important in determining the occurrence of the instability and t he position of the vortex end. The position of the end of the vortex is crucial for the design of centrifugal separators, such as cyclones.