Optimization of geometric indicators of a ventricular pump using computational fluid dynamics, surrogate model, response surface approximation, kriging and particle swarm optimization algorithm

Mohammad Saleh-Abadi,Ahmadreza Rahmati,Amirhamzeh Farajollahi,Ali Fatemi,Mohammad Reza Salimi
DOI: https://doi.org/10.1007/s40430-023-04355-y
IF: 2.361
2023-07-24
Journal of the Brazilian Society of Mechanical Sciences and Engineering
Abstract:A ventricular pump is a device used to treat patients with heart failure. In this device, the blood pressure is very important. On the other hand, the power use and efficiency are very important due to the battery volume reduction. In this study, it was attempted to introduce an optimum example of this pump using computational fluid dynamics (CFD) as an accurate method along with kriging and surrogate meta-models to reduce time and particle swarm optimization (PSO) algorithm. 5 indicators have been investigated as geometric optimization indicators in this study. Blade inlet angle, blade outlet angle, number of blades, the distance from the leading edge of the blade to the inlet eye of the pump and finally the width of the blade (viewed from the side to the blade) are the geometrical indicators analyzed in this study for their effects on the optimization. In this study, the power, efficiency, head, hemolysis and relative hemolysis have been analyzed, and hemolysis and efficiency have been introduced as the main objectives of PSO and pump head has been considered as a constraint. The introduced optimum example has reduced hemolysis more than 9% and power use more than 10% compared to the model investigated by the American Food and Drug Administration, while the efficiency has also increased by more than 2% and the related constraint to the pump head is maintained.
engineering, mechanical
What problem does this paper attempt to address?