Experimental and Numerical Analysis of Dye Washout in Ventricular Assist Devices

PurposeVentricular Assist Devices (VADs) are associated with complications due to the damage done to the blood. This damage relates to shear stress and exposure time. Designing VADs to minimise exposure time, in addition to reducing stress, should then reduce damage. The aim of this work was to use washout of a dye to investigate local and overall residence times in VADs.MethodsA flow rig was developed such that initially dyed glycerol circulated and a three way tap was used to switch to a transparent solution. A camera was focused on the transparent VAD outlet tube, and the colour change in the images calibrated to dye concentration. The concentration change at the outlet, ‘dye washout curve’, was used as an indicator of the residence times in the VAD. Dye washout curves were measured for VADs both in development and commercially available.Computational Fluid Dynamics (CFD) was used to calculate dye washout curves. An established flow field was first obtained with a transient simulation. A scalar variable was introduced to represent the dye, and a transport equation was used to model its convection. Simulated dye washout curves were compared with experiment, and simulations were used to understand fluid dynamic factors determining the different curves.ResultsDifferences in experimental washout were clearest at a flow rate of 1 l/min. The time taken for the concentration to drop from 100 to 20 %, T20, was an initial indicator of residence time. T20 was 2.1 s in the CentriMag (magnetic suspension) and just 0.9 s in the HeartMate II (axial & contact bearings). This difference may be explained by the larger volume of the CentriMag. Despite a similar internal volume to the HeartMate II, the Calon MiniVAD (centrifugal & magnetic suspension) had T20 of 1.4 s. The difference is likely due to the secondary flow in the MiniVAD which is required for suspension.ConclusionThe secondary flow path plays a critical role in determining blood residence time in the VAD. Ongoing work is focussed on investigating the importance of secondary flow path parameters.