DESIGN AND PERFORMANCE ANALYSIS OF MULTISTAGE CENTRIFUGAL PUMP USING CFD

Abstract

A multistage centrifugal pump is a specific type of centrifugal pump featuring multiple impellers arranged in series on a single shaft. While a typical centrifugal pump uses a volute casing, a multistage pump employs a diffuser that serves a similar function. To guide the flow to the next impeller, a return passage channel is utilized. To conduct numerical flow simulations for multistage pumps, it is necessary to first perform an analysis of a single-stage pump and then further for the multistage pump. A single-stage model is created in Creo-parametric software, where each domain, i.e., impeller, diffuser, and return passage, is modeled separately. The model is then imported into ANSYS 22R1 for part meshing of each domain. Subsequently, the model is imported into CFX –PRE, where boundary conditions at the inlet and the outlet are applied, and the domains are interfaced with each other. The flow simulation is carried out using the k-ε turbulence model. The flow analysis is performed for different numbers of blades (5, 6, and 7) at varying speeds from 1450 rpm to 1900 rpm while maintaining a constant discharge. The results show that the pump performance increases with speed, with the highest efficiency observed for 1900 rpm. The response surface approach was employed to enhance the hydraulic performance of the pump at the rated point. Specifically, an approximate link between the design head and efficiency of the single-stage centrifugal pump and the parameters of the impeller’s design was established. The first step in creating a one-factor experimental design involved selecting significant geometric variables as factors. Decision variables such as the number of blades, flow rate, and rotation were chosen due to their significant impact on hydraulic performance, while head and efficiency were considered as responses. Subsequently, the best-optimized values for each level of the parameters were identified using response surface analysis and a central composite design. The impeller schemes of the Design-Expert software were evaluated for head and efficiency using Computational fluid dynamics, and a total of 20 experiments were conducted. The simulated results were then validated with experimental data. Through the analysis of the individual parameters and the approximation model, the ideal parameter combination that increased head and efficiency by 7.90% and 2.06%, respectively, at the rated value was discovered. It is worth noting that in cases of a high rate of flow, the inner flow was also enhanced. xix After analyzing the single-stage pump, the three best-performing configurations based on speed are selected for further flow simulation of a two-stage centrifugal pump for different blades, varying from 5, 6, and 7, with the speed and mass flow kept constant at 1900 rpm and 128.8 kg/s, respectively. It is observed that the two-stage pump with a 7-blade impeller rotating at 1900 rpm provides the highest head among all configurations. It is also observed that losses in the diffuser of the 1st stage and 2nd stage are almost the same, but head losses in the return passage of the 1st stage and 2nd stage have differences of 19.99%, 28.5%, and 23.59% for 5, 6, and 7 blades, respectively. For compacting the design structure of the Centrifugal Pump, three different types of return guide vanes (mixed return guide vane, radial return guide vane, and distorted return guide vane) are modeled and analyzed using ANSYS 22R1 with the same configurations used in a two-stage pump. The results show that the radial return guide vanes give the best vane efficiency and lower loss coefficients. Further, the analysis is done on a radial guide vane with varying vane outlet angles (20, 25, and 30). From this analysis, we found that the radial guide vane with 25 degrees gives the best vane efficiency and less loss coefficient. This thesis provides valuable insights for optimizing multistage centrifugal pump designs, enhancing efficiency and performance in industries like water treatment, oil and gas, and chemical processing. By applying these findings, engineers can achieve higher energy efficiency, reduced operational costs, and improved reliability in pump systems.