Double-suction centrifugal pumps are widely used in industrial and agricultural applications since their flow rate is twice that of single-suction pumps with the same impeller diameter. They usually run for longer, which makes them susceptible to cavitation, putting the downstream components at risk. A fast approach to predicting the Net Positive Suction Head required was applied to perform a multi-objective optimization on the double-suction centrifugal pump. An L32 (84) orthogonal array was designed to evaluate 8 geometrical parameters at 4 levels each. A two-layer feedforward neural network and genetic algorithm was applied to solve the multi-objective problem into pareto solutions. The results were validated by numerical simulation and compared to the original design. The suction performance was improved by 7.26%, 3.9%, 4.5% and 3.8% at flow conditions 0.6Qd, 0.8Qd, 1.0Qd and 1.2Qd respectively. The efficiency increased by 1.53% 1.0Qd and 1.1% at 0.8Qd. The streamline on the blade surface was improved and the vapor volume fraction of the optimized impeller was much smaller than that of the original impeller. This study established a fast approach to cavitation optimization and a parametric database for both hub and shroud blade angles for double suction centrifugal pump optimization design.
Vertical inline pump is a single-stage single-suction centrifugal pump with a bent pipe before the impeller, which is usually used where installation space is a constraint. In this paper, with three objective functions of efficiencies at 0.5 Qd, 1.0 Qd, and 1.5 Qd, a multi-objective optimization on the inlet pipe of a vertical inline pump was proposed based on genetic algorithm with artificial neural network (ANN). In order to describe the shape of inlet pipe, the fifth-order and third-order Bezier curves were adopted to fit the mid curve and the trend of parameters of cross sections, respectively. Considering the real installation and computation complexity, 11 variables were finally used in this optimization. Latin hypercube sampling (LHS) was adopted to generate 149 sample cases, which were solved by CFD code ANSYS cfx 18.0. The calculation results and design variables were utilized to train ANNs, and these surrogate models were solved for the optimum design using multi-objective genetic algorithm (MOGA). The results showed the following: (1) There was a great agreement between numerical results and experimental results; (2) The ANNs could accurately fit the objective functions and variables. The maximum deviations of efficiencies at 0.5 Qd, 1.0 Qd, and 1.5 Qd, between predicted values and computational values, were 1.94%, 2.35%, and 0.40%; (3) The shape of inlet pipe has great influence on the efficiency at part-load and design conditions while the influence is slight at overload condition; (4) Three optimized cases were selected and the maximum increase of the efficiency at 0.5 Qd, 1.0 Qd, and 1.5 Qd was 4.96%, 2.45, and 0.79%, respectively; and (5) The velocity distributions of outflow in the inlet pipe of the three optimized cases were more uniform than the original one.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.