Validation of numerically simulated ventricular flow patterns during left ventricular assist device support.

Area of Science:

• Biomedical Engineering
• Fluid Dynamics
• Medical Devices

Background:

• Computational fluid dynamics (CFD) is used to study intraventricular flow patterns during left ventricular assist device (LVAD) support.
• CFD simulations aid in developing parameters for thrombus formation risk analysis.
• Experimental validation is crucial for the accuracy of complex CFD simulations.

Purpose of the Study:

• To experimentally validate CFD simulations of intraventricular flow.
• To compare the accuracy of different numerical methods in simulating flow patterns within a left ventricle model.
• To assess the suitability of various CFD models for LVAD flow analysis.

Main Methods:

• A ventricular model with a defined inflow section was analyzed using particle image velocimetry (PIV).
• Transient CFD simulations were performed using four numerical methods: laminar, standard k-omega, shear-stress transport (SST), and renormalized group k-epsilon (RNG k-epsilon).
• Simulated flow patterns were compared to PIV measurements across 46 planes in the left ventricle.

Main Results:

• CFD simulations using laminar, standard k-omega, and SST methods showed comparable results to PIV measurements.
• The RNG k-epsilon method exhibited a higher absolute error (17.8 mm/s) compared to other methods.
• The laminar model demonstrated the best transient behavior among the validated methods.
• Absolute errors for laminar, standard k-omega, and SST were 10.5, 11.3, and 11.3 mm/s, respectively.

Conclusions:

• The laminar, standard k-omega, and shear-stress transport CFD methods provide acceptable accuracy for simulating intraventricular flow under LVAD support.
• The laminar model offers the most accurate transient flow behavior in this specific experimental setup.
• Experimental validation using PIV is essential for refining CFD models in cardiovascular device research.