Assessment of the Flow Field in the HeartMate 3 Using Three-Dimensional Particle Tracking Velocimetry and Comparison to Computational Fluid Dynamics

Abstract

Flow fields in rotary blood pumps (RBPs) have a significant influence on hemocompatibility. Because flow characteristics vary with flow rate, different operating conditions play a role. Furthermore, turbulence is crucial in the evaluation of blood damage potential, but the level of turbulence in implantable RBPs is still unknown. In this study, we addressed both research aspects and for the first time measured turbulent flow fields in the HeartMate 3 (HM3) at different operating flows. The averaged, three-dimensional velocity field including fluctuating velocity components in a HM3 with a transparent lower housing was measured using three-dimensional particle tracking velocimetry (3D-PTV). In vitro results were compared with computational fluid dynamic (CFD) simulations for two flow cases, representing the lower and upper physiologic flow range (2.7 and 5.7 L/min), using two different turbulence models that account for fluctuating velocity fields: the k-ω shear stress transport and the Reynolds stress model (RSM). The measurements revealed higher mean and turbulent kinetic energies (TKEs) for the low-flow condition especially within the gap beneath the impeller. Computed mean fields agree well with 3D-PTV for both models, but the RSM predicts the TKE levels better than the k-ω model. Computational fluid dynamic results further show wall shear stresses higher than 150 Pa, a commonly used damage threshold, in the bottom gap for the lower flow condition. In conclusion, the low-flow condition was found to be more prone to blood damage. Furthermore, CFD predictions for turbulence must be carefully experimentally validated.

Linking Hydraulic Properties to Hemolytic Performance of Rotodynamic Blood Pumps

“…For flow analysis based on computational fluid dynamics (CFD), the pump geometries of both RBPs were reverse engineered as previously described. [14,15] Corresponding priming volumes were spatially discretized to high-quality computational grids (HM3: ≈9.9 million cells; HVAD: ≈8 million cells) as presented in Escher et al, [16] and grid quality was verified a posteriori (see Supporting Information). Both grid generation and in silico flow analysis were performed in the commercial software StarCCM+ (Siemens, Munich, Germany).…”

Hemocompatibility and hemodynamic comparison of two centrifugal LVADs: HVAD and HeartMate3

“…For each device individually, higher hemolysis index is detected when increasing the rotational speed. Furthermore, a potentially higher risk of hemolysis is detected operating at low-flow conditions, as encountered by other authors (Granegger et al 2020;Thamsen et al 2020). Operating at higher rotational speed the shear stresses are greater promoting higher levels of hemolysis, while the increase in hemolysis at low-flow conditions is a consequence of the larger residence times of a blood cell within the pumps.…”

High‐speed visualization of ingested, ejected, adherent, and disintegrated thrombus in contemporary ventricular assist devices

“…The centrifugal pumps (HVAD and HM3) were fitted with a flat acrylic viewing window for visualization of the blood flow path (See Figure 1). The dimensions of each of the 3 VAD models were reverse engineered from explanted pumps and confirmed with the literature 20‐22 . Table 1 summarizes the key dimensions, including fluid volume, impeller diameter and thickness, and gap between the impeller and pump housing.…”