Influence of turbulent shear stresses on the numerical blood damage prediction in a ventricular assist device

Abstract

The blood damage prediction in rotary blood pumps is an important procedure to evaluate the hemocompatibility of such systems. Blood damage is caused by shear stresses to the blood cells and their exposure times. The total impact of an equivalent shear stress can only be taken into account when turbulent stresses are included in the blood damage prediction. The aim of this article was to analyze the influence of the turbulent stresses on the damage prediction in a rotary blood pump’s flow. Therefore, the flow in a research blood pump was computed using large eddy simulations. A highly turbulence-resolving setup was used in order to directly resolve most of the computed stresses. The simulations were performed at the design point and an operation point with lower flow rate. Blood damage was predicted using three damage models (volumetric analysis of exceeded stress thresholds, hemolysis transport equation, and hemolysis approximation via volume integral) and two shear stress definitions (with and without turbulent stresses). For both simulations, turbulent stresses are the dominant stresses away from the walls. Here, they act in a range between 9 and 50 Pa. Nonetheless, the mean stresses in the proximity of the walls reach levels, which are one order of magnitude higher. Due to this, the turbulent stresses have a small impact on the results of the hemolysis prediction. Yet, turbulent stresses should be included in the damage prediction, since they belong to the total equivalent stress definition and could impact the damage on proteins or platelets.

Evaluation of clinically relevant operating conditions for left ventricular assist device investigations

“…Since similar static BC pairs are already used in other studies, 4,9,10,13 the in vivo determined, static BCs proposed in this article seem appropriate for first CFD design studies or in vitro tests during the early development stage of an arbitrary VAD. Nevertheless, these static BCs are not including the cardiac cycle with the systolic and diastolic conditions.…”

“…3,6,7,12,13 Also, the averaged pressure head of H = 62.0 mmHg was approximately situated in the range of many literature values. 4,9,10,13…”

“…This dynamic behavior also affects the flow-induced blood damage induced by the VAD. For example, the acceleration of flow during the heartbeat waveform leads to higher shear stresses near the walls (due to higher velocities), more turbulent stresses in the off-design conditions (due to stronger secondary flow structures), 10,21 and other dynamical flow conditions like alternating separation areas due to negative pressure gradients in decelerating phases. 22 But these effects cannot be investigated within analyses of single BCs.…”

“…Often, single values for the nominal flow rate in a narrow range between Q = 4.0–5.4 L/min are used. 3–13 The coupled pressure head values reveal a broader range between H = 65–100 mmHg for the associated flow rates. 3–13…”

Turbulence and turbulent flow structures in a ventricular assist device—A numerical study using thelarge‐eddysimulation

“…But, the described numerical blood damage analyses must be performed in such a way that all possible causes of blood damage within the VAD are considered. One of these causes is turbulence, since turbulent vortices inject additional (turbulent) shear stresses to the VAD flow 17 that could additionally damage the blood components 18 …”