Skip to main content

Linear eddy viscosity models

Linear eddy viscosity turbulence model selections include:

Background

Under the Boussinesq hypothesis 7, the deviatoric anisotropic stress is considered proportional to the traceless mean rate of strain:

ρRdev=ρuu+23ρkI=μt[2S(23u)I] - \rho \tensor{R}_\mathit{dev} = - \rho \av{\u' \otimes \u'} + \frac{2}{3} \rho k \tensor{I} = \mu_t \left[ 2\, \tensor{S} - \left( \frac{2}{3} \div \u \right) \tensor{I} \right]

where S\tensor{S} is the symmetric tensor

S=12(u+(u)T) \tensor{S} = \frac{1}{2}\left( \grad \av{\u} + \grad \left( \av{\u} \right)^T \right)

leading to:

ρRdev=μt(u+(u)T)+μt(23u)I - \rho \tensor{R}_\mathit{dev} = \mu_t \left( \grad \av{\u} + \grad \left( \av{\u} \right)^T \right) + \mu_t \left( \frac{2}{3} \div \u \right)\tensor{I}

where μt\mu_t is the dynamic eddy viscosity. The momentum equation therefore becomes:

t(ρu)+(ρuu)=gp+(μeffu)+[μeffdev2((u)T)] \ddt{\rho \av{\u}} + \div \left( \rho \av{\u} \otimes \av{\u} \right) = \vec{g} - \grad \av{p}' + \div \left( \mu_\mathit{eff} \grad \av{\u} \right) + \div \left[ \mu_\mathit{eff} \, \mathrm{dev2}\left(\left(\grad \av{\u}\right)^T \right) \right]

where μeff\mu_\mathit{eff} is the effective dynamic eddy viscosity:

μeff=μ+μt \mu_\mathit{eff} = \mu + \mu_t

i.e. the sum of the laminar and turbulent contributions.