Comprehensive study on simulation, performance evaluation and optimization strategies for blood pumps

This study aims to develop simulation methods, performance evaluation methods, and optimization strategies for blood pumps. Computational fluid dynamics, in-vitro hemocompatibility experiments and clinical data were utilized to analyze 8 clinical blood pumps across approximately 150 conditions. Results indicated that using the transient RANS and LES models with over 20 million mesh cells improves predictions of hydraulic performance and haemocompatibility. Optimal haemolysis model coefficients were identified to align closely with experimental results. A method to assess thrombosis, bleeding, and stroke risks caused by blood pumps was developed and validated with clinical data. Single haemolysis indicators were found insufficient, highlighting the need for a multi-indicator method to evaluate haemocompatibility. An energy loss-based optimization strategy was proposed and applied to enhance the haemocompatibility of a blood pump. This study provides robust simulation, evaluation, and optimization methods applicable to blood pumps, facilitating their performance assessment and optimization.