Validation of a CFD model of an orbiting culture dish with PIV and analytical solutions

Abstract

Particle image velocimetry (PIV) and an extended solution of Stokes’ second problem were used to validate a computational fluid dynamics (CFD) model of flow in an orbiting dish. Velocity vector components throughout one complete orbit differed between CFD and PIV by less than 5%. Computational velocity magnitudes averaged over the interior 20% radius, the region where the analytical solution is most applicable, were 0.3% higher than the analytical values, while the experimental values in the same region were 2.4% higher. Velocity profiles in the center of the dish across normalized heights that most influence wall shear stress varied on average by ∼–0.00046 for the normalized radial component and by ∼0.0038 for the normalized tangential component compared to the analytical solution. These results represent the most comprehensive validation to date for computational models of the orbiting dish system. © 2017 American Institute of Chemical Engineers AIChE J, 63: 4233–4242, 2017

Orbitally shaken shallow fluid layers. II. An improved wall shear stress model

“…In the boundary layer model, the thickness of the boundary layer is characterised by ν ω . At this z level, the theory (14) predicts that V (z = ν ω ) = 0.7V s = 0.66V max . By means of the inverse operation, we define the boundary layer thickness δ in the simulations as the z level where the simulated velocity is 0.66 of the simulated maximum velocity, i.e.…”

“…The velocity profile obtained in the simulations has been normalised by the characteristic thickness δ and plotted in the vertical direction up to the free surface for each point in the different regions of the container (Fig. 6); this allows for comparison to the theoretical profile shape (14) which is indicated by a discontinuous line. In the centre region, the theoretical profile correctly describes the velocity shape except for some points where the maximum peaks at the free surface instead of z max = 3.2δ, as predicted by the theory ( § II A).…”

“…In the intermediate region, the theoretical profile correctly describes the velocity slope at the bottom with very few exceptions; however for z > δ, the profiles show multiple behaviours. In the edge region, most profiles have a shape (14).…”

Understanding mechanobiology in cultured endothelium: A review of the orbital shaker method

“…is uniform throughout the well, unaffected by its side walls and independent of the starting height of the fluid. All three of these properties are incorrect [ [29] , [30] , [31] ] and the calculated WSS magnitude can be in error by an order of magnitude [ [31] , [32] , [33] ].…”

Mode decomposition and Lagrangian structures of the flow dynamics in orbitally shaken bioreactors

“…Froude number based either on the cylinder diameter, d i , or orbital diameter, d o . Analytical solutions of the flow in orbiting cultures based on Stokes' second problem were compared to PIV and CFD results by Thomas et al, 15 while potential flow functions were derived by Reclari et al 16 and Bouvard, Herreman, and Moisy 17 and validated with PIV measurements. Similarly a potential sloshing model of the free surface was formulated and compared against free surface wave measurements by Reclari et al, 16 who identified the presence of different modal responses inducing different flow regimes in a shaken cylindrical container, while Discacciati et al 18 developed a pressure correction method for CFD simulations, to best capture the free surface deformation and assess the shear stress levels for a highly viscous fluid.…”