The non-uniform distribution of atherosclerosis within the arterial tree has been attributed to variation in wall shear stress (WSS). However, the metric leading to disease initiation is disputed. The problem has been extensively investigated using the swirling well method, which exposes cultured endothelial cells (EC) to spatially varying shear conditions. However, the CFD used to characterize the flow has rarely been validated experimentally, and it is hard to distinguish effects on EC of WSS magnitude from effects of flow directionality. This study aims to improve, validate and adapt the existing CFD model of a swirling 12-well plate on an orbital shaker. The results were used to test the currently accepted theory that cells align with the mean WSS direction.

If was found that fluctuations in rotational speed can be ignored if they remain below ±19 RPM but surface tension/wetting effects need to be incorporated in the CFD model as they have a significant effect on wave breaking and velocity magnitude: the agreement of experimental particle imaging velocimetry (PIV) and height with CFD was improved when surface tension effects were incorporated into the CFD model.

Modifications were made by increasing culture medium viscosity and, volume, and by geometrical changes to the 12-well. Adding a central cylinder produced uniaxial flow whereas suspending a cylinder above the base produced multidirectional flow. Increasing the volume produced low but relatively constant magnitudes whereas viscosity changes influenced only the magnitude of WSS and not directionality. Tilting the well removed radial symmetry of the flow. EC were subjected to control or modified flows and nuclear morphology was assessed by confocal microscopy. Contrary to the conventional view, EC did not always align with the mean WSS vector; nuclei aligned with the modal WSS vector when the time average WSS was 0.3-0.5 Pa, but aligned so as to minimise the transverse WSS when the time average WSS was 0.6 Pa.