GfsAdaptError

From Gerris

GfsAdaptError uses an a posteriori error estimate of a given field as a cost function for adaptation. The error is estimated by computing the norm of the Hessian matrix of the given field, estimated using third-order-accurate discretisation operators.

The syntax in parameter files is:

[ GfsAdaptGradient ]

The cmax value of the corresponding GfsAdapt is the maximum acceptable estimated error.

Examples

• Viscous folding of a fluid interface

    AdaptError { istep = 1 } { cmax = 1e-2 maxlevel = max_level } U

    AdaptError { istep = 1 } { cmax = 1e-2 maxlevel = max_level } V

• Forced isotropic turbulence in a triply-periodic box

  AdaptError { istep = 1 } { cmax = 5e-2 maxlevel = 7 } U/Unbar

  AdaptError { istep = 1 } { cmax = 5e-2 maxlevel = 7 } V/Unbar

  AdaptError { istep = 1 } { cmax = 5e-2 maxlevel = 7 } W/Unbar

• The 2004 Indian Ocean tsunami

    AdaptError { istep = 1 } {
    	cmax = 0.05
    	minlevel = 5 maxlevel = LEVEL 
    } Pmax

• Cyclone-generated wave field

    AdaptError { istep = 1 } { cmax = 0.1 minlevel = 4 maxlevel = LEVEL c = Hse } Hs

    AdaptError { istep = 1 } { 
        cmax = 0.2 minlevel = 4 maxlevel = LEVEL c = Ve 
    } sqrt(U10*U10 + V10*V10)

• Time-reversed advection of a VOF concentration

    AdaptError { istart = 1 istep = 1 } { cmax = 1e-3 maxlevel = LEVEL } C

    AdaptError { istart = 1 istep = 1 } { cmax = 1e-3 maxlevel = LEVEL } G

• Equilibrium of a droplet suspended in an electric field

    AdaptError { istep = 1 } { cmax = 2e-4 maxlevel = 7 } U

    AdaptError { istep = 1 } { cmax = 2e-4 maxlevel = 7 } V