BUG: curvatureSeparation: incorrect modelling constant

The curvatureSeparation model of the finite-area framework is based on Owen & Ryley (1983), Eq. 11:

$$F_w = \frac{4}{3} \, \rho \, |\mathbf{u}|^2 \, \frac{R_1 - R_2}{R_1} - \frac{\rho |\mathbf{g}| cos(\theta +\beta) \left(R_1^2 - R_2^2 \right)}{2 R_1} + \frac{\sigma}{R_2}$$

where

$F_w$

is the radial stress at the wall, and the rest of the variables are well-defined in the original-research paper.

The implementation of the Foam::curvatureSeparation::correct function (as of December 2024, SHA:e1d45ec26e) is as follows:

$$F_w = \frac{72}{60} \, \rho \, |\mathbf{u}|^2 \, \frac{R_1 - R_2}{R_1} - \frac{\rho |\mathbf{g}| cos(\theta +\beta) \left(R_1^2 - R_2^2 \right)}{2 R_1} + \frac{\sigma}{R_2}$$

The original paper and implementation differ in terms of the leading coefficient of the inertial-force term, that is

$\frac{4}{3}$

vs

$\frac{72}{60}$

.

added bug finiteArea labels

assigned to @kuti

mentioned in merge request !714 (merged)

mentioned in commit d980a050

mentioned in commit dd0c70b4

mentioned in commit 1234ca9e

mentioned in commit 1e6d8d73

closed

mentioned in commit 45e7c4a0