Optimization of a low Nq number pump based on experimentation and numerical simulation

The aim of this work is to analyze the functioning of an industrial centrifugal pump and try to improve its efficiency according to the specific working conditions. The machine is produced by a German company that sells it in a number of several thousand pieces per year and this, despite the small dimension of the pump, makes appealing the possibility of enhancing the efficiency characteristic. To decide on which parameters to act and how to influence the fluid dynamic behavior, it is important to completely understand the functioning of the machine; this will be done by conducting an analysis from four different prospective to get a unified prospective on the problem: - The experimental analysis produces accurate characteristics curves that are required to validate the numerical model other than to be compared with the commercial data sheet. -The use of semi-empirical models known in literature makes possible to calculate the amount of losses caused by each component of the pump and most important of all, the relative ratio that occurs among them. - The simple calculation of the velocity triangles along with the use of some empirical relations is used to provide a good first try in the optimization process. Moreover the distance, for certain geometric dimensions, between the suggested value and the real one could imply the presence of some problems. -The numerical analysis is done by means of a Computational Fluid Dynamic method and allows taking a look into the flow inside the pump. The CFD program, which is validated through the confrontation with experimental results, represents the main tool used for the iterative optimization process. The outputs of the whole procedure are two new impeller versions: Model 8 presents a rise of efficiency of almost 4,5% on all the useful range and has a two dimensional geometry of the blades that is easy and cheap to manufacture. Model 10 attains an efficiency increase of around 7% in the same range but the blades are three dimensional and more expensive to produce. The relevance of the results, for the company purposes, is proven by their decision of proceeding with the experimental test of the new model; this will be the last step before the commercialization of the product.