Kilian,
Thank you, it seems to be what I need indeed. I will take a closer look.
Best regards,
Dmitry
On 26.05.2020 15:06, Kilian Weishaupt wrote:
Hi Dmitry,
we have implemented an outflow condition for Box in the 1p2c
isothermal test.
Maybe you can adapt it to your needs (see below)
Best wishes
Kilian
template<bool useBox = isBox, std::enable_if_t<useBox, int> = 0>
NumEqVector neumann(const Element& element,
const FVElementGeometry& fvGeometry,
const ElementVolumeVariables& elemVolVars,
const ElementFluxVariablesCache&
elemFluxVarsCache,
const SubControlVolumeFace& scvf) const
{
// no-flow everywhere except at the right boundary
NumEqVector values(0.0);
const auto xMax = this->gridGeometry().bBoxMax()[0];
const auto& ipGlobal = scvf.ipGlobal();
if (ipGlobal[0] < xMax - eps_)
return values;
// set a fixed pressure on the right side of the domain
static const Scalar dirichletPressure = 1e5;
const auto& volVars = elemVolVars[scvf.insideScvIdx()];
const auto& fluxVarsCache = elemFluxVarsCache[scvf];
// evaluate the pressure gradient
GlobalPosition gradP(0.0);
for (const auto& scv : scvs(fvGeometry))
{
const auto xIp = scv.dofPosition()[0];
auto tmp = fluxVarsCache.gradN(scv.localDofIndex());
tmp *= xIp > xMax - eps_ ? dirichletPressure
: elemVolVars[scv].pressure();
gradP += tmp;
}
// compute flux
auto phaseFlux = vtmv(scvf.unitOuterNormal(),
volVars.permeability(), gradP);
phaseFlux *= useMoles ? volVars.molarDensity() :
volVars.density();
phaseFlux *= volVars.mobility();
// set Neumann bc values
values[contiH2OEqIdx] = phaseFlux;
// emulate an outflow condition for the component transport on
the right side
values[contiN2EqIdx] = phaseFlux * ( useMoles ?
volVars.moleFraction(0, N2Idx) : volVars.massFraction(0, N2Idx) );
return values;
}
On 26.05.20 13:41, Dmitry Pavlov wrote:
I see that dirichletAtPos() is available for Box method, too, but in
fact, I must use neumann() because by boundary conditions are more
complex than just the constant pressure. I have a two-phase flow and
need to implement a certain equation on the outflux of different
phases, depending on the pressure gradient.
Best regards,
Dmitry
On 26.05.2020 14:08, Dmitry Pavlov wrote:
Hello,
After transition from the CC method to the Box method in my 1d
porous medium flow problem, I am wondering how I should specify a
constant pressure boundary condition.
With the CC method, it was relatively easy: I could either implement
dirichletAtPos(), or implement neumann() instead and calculate the
flux across the boundary from the pressure gradient as
gradp = ([boundary pressure] - [pressure in the center of the cell])
/ (half-width of the cell)
With the Box method, though, it is no longer obvious to me. I will
appreciate any pointers. I see addCCDirichletFluxDerivatives in the
DuMux code, but no similar method for Box method.
Best regards,
Dmitry
_______________________________________________
Dumux mailing list
[email protected]
https://listserv.uni-stuttgart.de/mailman/listinfo/dumux
_______________________________________________
Dumux mailing list
[email protected]
https://listserv.uni-stuttgart.de/mailman/listinfo/dumux
