Correlation and Regression Method of Centrifugal Pump Geometry Optimization

Abstract

Correlation and regression analysis is the effective means to study the narrowness and analytical relationships of the factors to be analyzed. It can be applied in technics as a result of gathering and processing of operational data or results of tests of already existing samples of machinery, for example, for centrifugal pumps upon hydraulic tests. This article covers the procedure of correlation and regression analysis of centrifugal pump head and efficiency dependencies on its impeller geometry in meridional and circumferential planes. The purpose of the analysis is to optimize the impeller geometry in order to achieve the maximum pump head ratio and efficiency. This article provides a detailed description of the proposed optimization method of the design. Geometric factors more significantly effecting the energy parameters of the pump are included in the mathematical models connecting the energy parameters of the centrifugal pump with the relative geometric dimensions of the impeller. Based on the obtained models, optimization of the impeller geometry is performed by adjusting the absolute dimensions of the impeller interblade channels to the values corresponding to the optimum ones determined by correlation and regression analysis. This approach allows modification of existing equipment based on the mathematical processing of empirical data appearing only after the beginning of operation. The correlation and regression method of design optimization is one of the few means for efficient use of operational data not only in the field of centrifugal pumps but also in other hydraulic machines and devices as well. This method is informative, does not require intensive experimental studies, and provides for a modification of the equipment with the minimum material and financial expenses.