ABSTRACT
Pump performance optimization requires large degree of freedom modeling such as finite element structural analysis or fluid dynamic computational modeling. Loworder modeling is best suited for a system involving few discrete elements such as mass and resistance. This type of modeling is most ideal for describing the critical design aspects of the entire MP system. Large degree of freedom modeling is not suitable for design and optimization of time-dependent MP systems involving fixedvalve concepts, which typically operate in the kilohertz frequency range. Low-order nonlinear models are used for MPs incorporating mechanical valves. Present nonlinear models for diffuser pumps are very complex and require several assumptions and empirical coefficients, which cannot guarantee the modeling accuracy. Each pump can involve seven coefficients, five of which are based on steady-flow theory and two of which need adjustments for best fit to the data obtained for the turbulent flow. The velocity profiles for both the steady flow (laminar flow) and turbulent flow (unsteady flow) must be included in the modeling. Furthermore, any parameter adjustment in one area may yield errors in another parameter.
