While cardiovascular device-induced thrombosis is associated with negative patient outcomes, the convoluted nature of the processes resulting in a thrombus makes the full thrombotic network too computationally expensive to simulate in the complex geometries and flow fields associated with devices. A macroscopic, continuum computational model is developed based on a simplified network, which includes terms for platelet activation (chemical and mechanical) and thrombus deposition and growth in regions of low wall shear stress (WSS). Laminar simulations are performed in a two-dimensional asymmetric sudden expansion geometry and compared with in vitro thrombus size data collected using whole bovine blood. Additionally, the predictive power of the model is tested in a flow cell containing a series of symmetric sudden expansions and contractions. Thrombi form in the low WSS area downstream of the asymmetric expansion and grow into the nearby recirculation region, and thrombus height and length largely remain within 95 % confidence intervals calculated from the in vitro data for 30 min of blood flow. After 30 min, predicted thrombus height and length are 0.94 and 4.32 (normalized by the 2.5 mm step height). Importantly, the model also correctly predicts locations of thrombus deposition observed in the in vitro flow cell of expansions and contractions. As the simulation results, which rely on a greatly reduced model of the thrombotic network, are still able to capture the macroscopic behavior of the full network, the model shows promise for timely predictions of device-induced thrombosis toward optimizing and expediting the device development process.

Thrombosis and thromboembolization remain large obstacles in the design of cardiovascular devices. In this study, the temporal behavior of thrombus size within a backward-facing step (BFS) model is investigated, as this geometry can mimic the flow separation which has been found to contribute to thrombosis in cardiac devices. Magnetic resonance imaging (MRI) is used to quantify thrombus size and collect topographic data of thrombi formed by circulating bovine blood through a BFS model for times ranging between 10 and 90 min at a constant upstream Reynolds number of 490. Thrombus height, length, exposed surface area, and volume are measured, and asymptotic behavior is observed for each as the blood circulation time is increased. Velocity patterns near, and wall shear stress (WSS) distributions on, the exposed thrombus surfaces are calculated using computational fluid dynamics (CFD). Both the mean and maximum WSS on the exposed thrombus surfaces are much more dependent on thrombus topography than thrombus size, and the best predictors for asymptotic thrombus length and volume are the reattachment length and volume of reversed flow, respectively, from the region of separated flow downstream of the BFS.

Transitional and turbulent flow through a simplified medical device model is analyzed as part of the FDA's Critical Path Initiative, designed to improve the process of bringing medical products to market. Computational predictions are often used in the development of devices and reliable in vitro data is needed to validate computational results, particularly estimations of the Reynolds stresses that could play a role in damaging blood elements. The high spatial resolution of laser Doppler velocimetry (LDV) is used to collect two component velocity data within the FDA benchmark nozzle model. Two flow conditions are used to produce flow encompassing laminar, transitional, and turbulent regimes, and viscous stresses, principal Reynolds stresses, and turbulence intensities are calculated from the measured LDV velocities. Axial velocities and viscous stresses are compared to data from a prior inter-laboratory study conducted with particle image velocimetry. Large velocity gradients are observed near the wall in the nozzle throat and in the jet shear layer located in the expansion downstream of the throat, with axial velocity changing as much as 4.5 m/s over 200 μm. Additionally, maximum Reynolds shear stresses of 1000-2000 Pa are calculated in the high shear regions, which are an order of magnitude higher than the peak viscous shear stresses (<100 Pa). It is important to consider the effects of both viscous and turbulent stresses when simulating flow through medical devices. Reynolds stresses above commonly accepted hemolysis thresholds are measured in the nozzle model, indicating that hemolysis may occur under certain flow conditions. As such, the presented turbulence quantities from LDV, which are also available for download at https://fdacfd.nci.nih.gov/ , provide an ideal validation test for computational simulations that seek to characterize the flow field and to predict hemolysis within the FDA nozzle geometry.

Platelet adhesion to a polyurethane urea surface is a precursor to thrombus formation within blood-contacting cardiovascular devices, and platelets have been found to adhere strongly to polyurethane surfaces below a shear rate of approximately 500 s−1. The aim of the current work is to determine platelet adhesion properties to the polyurethane urea surface as a function of time varying shear exposure. A rotating disk system is used to study the influence of steady and pulsatile flow conditions (e.g. cardiac inflow and sawtooth waveforms) for platelet adhesion to the biomaterial surface. All experiments retain the same root mean square angular rotation velocity (29.63 rad/s) and waveform period. The disk is rotated in platelet rich bovine plasma for two hours with adhesion quantified by confocal microscopy measurements of immunofluorescently labeled bovine platelets. Platelet adhesion under pulsating flow is found to exponentially decay with increasing shear rate. Adhesion levels are found to depend upon peak platelet flux and shear rate regardless of rotational waveform. In combination with flow measurements, these results may be useful for predicting regions susceptible to thrombus formation within ventricular assist devices.

Reynolds shear stress (RSS) has served as a metric for the effect of turbulence on hemolysis. Forstrom (1969) and Sallam and Hwang (1984) determined the RSS threshold for hemolysis to be 50,000 and 4,000 dyne/cm, respectively, using a turbulent jet. Despite the order of magnitude discrepancy, the threshold by Sallam and Hwang has been frequently cited for hemolytic potential in blood pumps. We recreated a Sallam apparatus (SA) to resolve this discrepancy and provide additional data to be used in developing a more accurate hemolysis model. Hemolysis was measured over a large range of Reynolds numbers (Re) (Re = 1,000-80,000). Washed bovine red blood cells (RBCs) were injected into the free jet of phosphate buffered saline, and hemolysis was quantified using a percent hemolysis, Hp = h (100 - hematocrit [HCT])/Hb, where h (mg/dl) is free hemoglobin and Hb (mg/dl) is total hemoglobin. Reynolds shear stress was calculated using two-dimensional laser Doppler velocimetry. Reynolds shear stress of ≥30,000 dyne/cm corresponding to Re of ≥60,000 appeared to cause hemolysis (p < 0.05). This RSS is an order of magnitude greater than the RSS threshold that Sallam and Hwang suggested, and it is similar to Forstrom's RSS threshold. This study resolved a long-standing uncertainty regarding the critical values of RSS for hemolysis and may provide a foundation for a more accurate hemolysis model.