Multi-Objective CFD Optimization of an Intermediate Diffuser Stage for PediaFlow Pediatric Ventricular Assist Device.

Area of Science:

• Biomedical Engineering
• Fluid Dynamics
• Medical Devices

Background:

• Ventricular assist devices (VADs) utilize computational fluid dynamics (CFD) for design, balancing performance and biocompatibility.
• Pediatric VADs face unique challenges due to size constraints, necessitating specialized design optimization.
• This study focuses on optimizing the PediaFlow pediatric VAD's diffuser stage to enhance pressure recovery.

Purpose of the Study:

• To perform automated CFD-driven shape optimization of a novel intermediate diffuser stage for the PediaFlow pediatric VAD.
• To maximize pressure recovery while simultaneously minimizing hemolysis in the VAD.
• To improve the overall performance and hemocompatibility of pediatric VADs.

Main Methods:

• A multi-objective optimization approach was employed, evaluating over 450 design variants using Sobol sequence.
• The diffuser stage was isolated for efficient CFD analysis, generating a Pareto front of optimal solutions.
• A local T-search algorithm refined the best candidate design, followed by CFD verification and in vitro validation.

Main Results:

• Critical dependencies were identified: longer blades increased pressure recovery but also hemolysis.
• Fewer blades (2 or 3) consistently outperformed designs with more blades (4 or 5) in both pressure recovery and hemolysis metrics.
• The optimized two-blade design improved hydraulic efficiency from 26.3% to 32.5% and reduced hemolysis by 31% at lower operating speeds.

Conclusions:

• Multi-objective CFD optimization systematically explores complex design spaces for pediatric VADs.
• This approach effectively balances competing priorities of hydraulic performance and hemocompatibility.
• The optimized diffuser design offers a promising solution for improving pediatric VADs.