Validation of the moving-boundary Lagrangian-mean formulation
Establish, using a deforming-mesh computation for the circular cavity with an oscillating boundary, whether transferring the wall condition to the reference circle and imposing no slip on the Lagrangian mean correctly represents the second-order moving-boundary problem.
References
Two premises stay untested: transferring the wall condition to the reference circle, and no slip on the Lagrangian mean as the correct second-order condition. A deforming-mesh computation would settle both, and what it costs is set by the ratio of the mean to the oscillation, $\varepsilon\,\mathrm{Wo}2$. At the operating point of Sec.~\ref{sec:scope}, where $\varepsilon = 0.01$ and $\mathrm{Wo}2 = 0.053$, that ratio is $5.3\times10{-4}$. It compares the two velocity scales rather than the two measured speeds: the peak Lagrangian speed there is $0.00401$ in the Eulerian scale, which against the wall speed $U$ is $2.1\times10{-6}$, and how finely a computation must resolve the mean follows from its own error analysis rather than from the scale ratio. At $\varepsilon = 0.05$ and $\mathrm{Wo}2 = 1$ the scale ratio is $5\times10{-2}$, about ninety times larger, and Sec.~\ref{sec:finiteWo} supplies the closed form there as well, so that point tests the same premises. The computation is therefore affordable at finite Womersley number, and it was not performed here.