Changes to solutionControl time consistency functions

I realised that we need separate treatment for transient and steady state
solvers (which do not have ddt equation). Hence, I provided necessary interface
and implementation for steady state solvers.
This commit is contained in:
Vuko Vukcevic 2017-01-04 10:15:34 +01:00
parent edb77356aa
commit ed03625c8b
3 changed files with 260 additions and 97 deletions

View file

@ -86,7 +86,7 @@ int main(int argc, char *argv[])
U = rAU*HUEqn.H();
// Consistently calculate flux
piso.calcTimeConsistentFlux(phi, U, rAU, ddtUEqn);
piso.calcTransientConsistentFlux(phi, U, rAU, ddtUEqn);
adjustPhi(phi, U, p);
@ -96,7 +96,7 @@ int main(int argc, char *argv[])
(
fvm::laplacian
(
rAU/piso.aCoeff(),
fvc::interpolate(rAU)/piso.aCoeff(),
p,
"laplacian(rAU," + p.name() + ')'
)
@ -119,7 +119,7 @@ int main(int argc, char *argv[])
# include "continuityErrs.H"
// Consistently reconstruct velocity after pressure equation
piso.reconstructVelocity(U, ddtUEqn, rAU, p, phi);
piso.reconstructTransientVelocity(U, ddtUEqn, rAU, p, phi);
}
runTime.write();

View file

@ -238,6 +238,93 @@ const Foam::dimensionedScalar Foam::solutionControl::relaxFactor
}
void Foam::solutionControl::addDdtFluxContribution
(
surfaceScalarField& phi,
surfaceScalarField& aCoeff,
const surfaceVectorField& faceU,
const volVectorField& U,
const surfaceScalarField& rAUf,
const fvVectorMatrix& ddtUEqn
) const
{
// Add ddt scheme dependent flux contribution. Here: phi = H/A & Sf.
phi += fvc::ddtConsistentPhiCorr(faceU, U, rAUf);
// Add ddt scheme dependent contribution to diffusion scale coeff. Here:
// aCoeff = 1.
aCoeff += fvc::interpolate(ddtUEqn.A())*rAUf;
}
void Foam::solutionControl::addUnderRelaxationFluxContribution
(
surfaceScalarField& phi,
surfaceScalarField& aCoeff,
const volVectorField& U
) const
{
// Get under-relaxation factor used in this iteration
const dimensionedScalar alphaU(this->relaxFactor(U));
const dimensionedScalar urfCoeff((1.0 - alphaU)/alphaU);
// Add under-relaxation dependent flux contribution
phi += urfCoeff*phi.prevIter();
// Add under-relaxation dependent contribution to diffusion scale coeff
aCoeff += urfCoeff*aCoeff;
// Note: this is not 100% entirely consistent with the way FOAM handles
// under-relaxation. In fvMatrix::relax() member function, the diagonal
// dominance is first assured by limiting the diagonal and the matrix is
// then relaxed. However, this is a very minor inconsistency, especially
// since this term vanishes when non-linear iterations converge to a tight
// tolerance (achieving steady state in SIMPLE or converging PISO/PIMPLE in
// each time step). VV, 23/Dec/2016.
}
void Foam::solutionControl::correctBoundaryFlux
(
surfaceScalarField& phi,
const volVectorField& U
) const
{
// Get necessary data at the boundaries
const volVectorField::GeometricBoundaryField& Ub = U.boundaryField();
const surfaceVectorField::GeometricBoundaryField& Sb =
mesh_.Sf().boundaryField();
surfaceScalarField::GeometricBoundaryField& phib = phi.boundaryField();
// Correct flux at boundaries depending on patch type. Note: it is possible
// that we missed some important boundary conditions that need special
// considerations when calculating the flux. Maybe we can reorganise this in
// future by relying on distinction between = and == operators for
// fvsPatchFields. However, that would require serious changes at the
// moment. VV, 23/Dec/2016.
forAll (phib, patchI)
{
if (Ub[patchI].fixesValue())
{
// This is fixed value patch, flux needs to be recalculated
// with respect to the boundary condition
phib[patchI] == (Ub[patchI] & Sb[patchI]);
}
else if
(
isA<slipFvPatchVectorField>(Ub[patchI])
|| isA<symmetryFvPatchVectorField>(Ub[patchI])
)
{
// This is slip or symmetry, flux needs to be zero
phib[patchI] == 0.0;
}
}
}
// * * * * * * * * * * * * * * * * Constructors * * * * * * * * * * * * * * //
Foam::solutionControl::solutionControl(fvMesh& mesh, const word& algorithmName)
@ -273,7 +360,7 @@ Foam::solutionControl::~solutionControl()
// * * * * * * * * * * * * * * * Member Functions * * * * * * * * * * * * * //
void Foam::solutionControl::calcTimeConsistentFlux
void Foam::solutionControl::calcTransientConsistentFlux
(
surfaceScalarField& phi,
const volVectorField& U,
@ -284,7 +371,7 @@ void Foam::solutionControl::calcTimeConsistentFlux
// Store necessary fields for time consistency if they are not present
if (!aCoeffPtr_ && !faceUPtr_)
{
aCoeffPtr_ = new volScalarField
aCoeffPtr_ = new surfaceScalarField
(
IOobject
(
@ -316,20 +403,22 @@ void Foam::solutionControl::calcTimeConsistentFlux
{
FatalErrorIn
(
"tmp<surfaceScalarField> solutionControl::timeConsistentFlux"
"void solutionControl::calcTransientConsistentFlux"
"\n("
"\n surfaceScalarField& phi,"
"\n const volVectorField& U,"
"\n const volScalarField& rAU"
"\n const volScalarField& rAU,"
"\n const fvVectorMatrix& ddtUEqn"
"\n)"
) << "Either aCoeffPtr_ or faceUPtr_ is allocated while the"
<< " other is not. This must not happen."
<< endl
<< " other is not."
<< nl
<< " This must not happen in transient simulation. Make sure"
<< " that functions aiding consistency are called in the right"
<< " order (first flux and then velocity reconstruction)."
<< exit(FatalError);
}
// Now that we are sure that pointers are valid and the fields are there, we
// can do the calculation of the flux, while storing necessary data for
// future velocity reconstruction.
// Algorithm:
// 1. Update flux and aCoeff due to ddt discretisation
// 2. Update flux and aCoeff due to under-relaxation
@ -337,11 +426,12 @@ void Foam::solutionControl::calcTimeConsistentFlux
// remains consistent
// Get fields that will be updated
volScalarField& aCoeff = *aCoeffPtr_;
surfaceScalarField& aCoeff = *aCoeffPtr_;
surfaceVectorField& faceU = *faceUPtr_;
// Update face interpolated velocity field. Note: handling of oldTime faceU
// fields happens in ddt scheme when calling ddtConsistentPhiCorr
// fields happens in ddt scheme when calling ddtConsistentPhiCorr inside
// addDdtFluxContribution
faceU = fvc::interpolate(U);
// Interpolate original rAU on the faces
@ -353,76 +443,90 @@ void Foam::solutionControl::calcTimeConsistentFlux
// Calculate the ordinary part of the flux (H/A)
phi = (faceU & mesh_.Sf());
// Initialize aCoeff to 1
aCoeff = dimensionedScalar("one", dimless, 1.0);
// STAGE 1: consistent ddt discretisation handling
// Add consistent flux contribution due to ddt discretisation
phi += fvc::ddtConsistentPhiCorr(faceU, U, rAUf);
// Calculate aCoeff according to ddt discretisation
aCoeff = 1.0 + ddtUEqn.A()*rAU;
addDdtFluxContribution(phi, aCoeff, faceU, U, rAUf, ddtUEqn);
// STAGE 2: consistent under-relaxation handling
// Get under-relaxation factor used in this iteration
const dimensionedScalar alphaU = this->relaxFactor(U);
// Helper variable
const dimensionedScalar urfCoeff = (1.0 - alphaU)/alphaU;
// Add consistent flux contribution due to possible under-relaxation
phi += urfCoeff*fvc::interpolate(aCoeff)*phi.prevIter();
// Add under-relaxation contribution to aCoeff
aCoeff += urfCoeff*aCoeff;
// Note: this is not 100% entirely consistent with the way FOAM handles
// under-relaxation. In fvMatrix::relax() member function, the diagonal
// dominance is first assured by limiting the diagonal and the matrix is
// then relaxed. However, this is a very minor inconsistency, especially
// since this term vanishes when non-linear iterations converge to a tight
// tolerance (achieving steady state in SIMPLE or converging PISO/PIMPLE in
// each time step). VV, 23/Dec/2016.
addUnderRelaxationFluxContribution(phi, aCoeff, U);
// STAGE 3: scale the flux and correct it at the boundaries
// Scale the flux
phi /= fvc::interpolate(aCoeff);
// Get necessary data at the boundaries
const volVectorField::GeometricBoundaryField& Ub = U.boundaryField();
const surfaceVectorField::GeometricBoundaryField& Sb =
mesh_.Sf().boundaryField();
surfaceScalarField::GeometricBoundaryField& phib = phi.boundaryField();
// Correct flux at boundaries depending on patch type. Note: it is possible
// that we missed some important boundary conditions that need special
// considerations when calculating the flux. Maybe we can reorganise this in
// future by relying on distinction between = and == operators for
// fvsPatchFields. However, that would require serious changes at the
// moment. VV, 23/Dec/2016.
forAll (phib, patchI)
{
if (Ub[patchI].fixesValue())
{
// This is fixed value patch, flux needs to be recalculated
// with respect to the boundary condition
phib[patchI] == (Ub[patchI] & Sb[patchI]);
}
else if
(
isA<slipFvPatchVectorField>(Ub[patchI])
|| isA<symmetryFvPatchVectorField>(Ub[patchI])
)
{
// This is slip or symmetry, flux needs to be zero
phib[patchI] == 0.0;
}
}
phi /= aCoeff;
correctBoundaryFlux(phi, U);
}
void Foam::solutionControl::reconstructVelocity
void Foam::solutionControl::calcSteadyConsistentFlux
(
surfaceScalarField& phi,
const volVectorField& U
) const
{
// Store aCoeff field for time consistency if it is not present. Note: faceU
// does not need to be stored
if (!aCoeffPtr_ && !faceUPtr_)
{
aCoeffPtr_ = new surfaceScalarField
(
IOobject
(
"aCoeff",
mesh_.time().timeName(),
mesh_,
IOobject::NO_READ,
IOobject::NO_WRITE
),
mesh_,
dimensionedScalar("zero", dimless, 0.0)
);
}
else if (faceUPtr_)
{
FatalErrorIn
(
"void solutionControl::calcSteadyConsistentFlux"
"\n("
"\n surfaceScalarField& phi,"
"\n const volVectorField& U,"
"\n const volScalarField& rAU"
"\n)"
) << "faceUPtr_ is allocated."
<< nl
<< " This must not happen in steady simulation. Make sure"
<< " that the right functions aiding consistency are called in the"
<< " right order (first flux and then velocity reconstruction)."
<< exit(FatalError);
}
// Algorithm:
// 1. Update flux and aCoeff due to under-relaxation
// 2. Scale the flux with aCoeff, making sure that flux at fixed boundaries
// remains consistent
// Get aCoeff field
surfaceScalarField& aCoeff = *aCoeffPtr_;
// Store previous iteration for the correct handling of under-relaxation
phi.storePrevIter();
// Calculate the ordinary part of the flux (H/A)
phi = (fvc::interpolate(U) & mesh_.Sf());
// Initialize aCoeff to 1
aCoeff = dimensionedScalar("one", dimless, 1.0);
// STAGE 1: consistent under-relaxation handling
addUnderRelaxationFluxContribution(phi, aCoeff, U);
// STAGE 2: scale the flux and correct it at the boundaries
phi /= aCoeff;
correctBoundaryFlux(phi, U);
}
void Foam::solutionControl::reconstructTransientVelocity
(
volVectorField& U,
const fvVectorMatrix& ddtUEqn,
@ -444,15 +548,16 @@ void Foam::solutionControl::reconstructVelocity
{
FatalErrorIn
(
"void solutionControl::reconstructVelocity"
"void solutionControl::reconstructTransientVelocity"
"\n("
"\n volVectorField& U,"
"\n const volVectorField& ddtUEqn,"
"\n const volScalarField& rAU,"
"\n const volScalarField& p"
"\n const surfaceScalarField& phi"
"\n) const"
) << "faceUPtr_ not calculated. Make sure you have called"
<< " calculateTimeConsistentFlux(...) before calling this function."
<< " calcTransientConsistentFlux(...) before calling this function."
<< exit(FatalError);
}
surfaceVectorField& faceU = *faceUPtr_;
@ -474,15 +579,31 @@ void Foam::solutionControl::reconstructVelocity
}
const Foam::volScalarField& Foam::solutionControl::aCoeff() const
void Foam::solutionControl::reconstructSteadyVelocity
(
volVectorField& U,
const volScalarField& rAU,
const volScalarField& p
) const
{
// Reconstruct the velocity field
U -= rAU*fvc::grad(p);
U.correctBoundaryConditions();
// Note: no need to store and update faceU field for steady state run
}
const Foam::surfaceScalarField& Foam::solutionControl::aCoeff() const
{
if (!aCoeffPtr_)
{
FatalErrorIn
(
"const volScalarField& solutionControl::aCoeff() const"
"const surfaceScalarField& solutionControl::aCoeff() const"
) << "aCoeffPtr_ not calculated. Make sure you have called"
<< " calculateTimeConsistentFlux(...) before calling aCoeff()."
<< " calcTransientConsistentFlux(...) or "
<< " calcSteadyConsistentFlux(...) before calling aCoeff()."
<< exit(FatalError);
}

View file

@ -145,7 +145,7 @@ protected:
) const;
// Time and under-relaxation consistency
// Time and under-relaxation consistency helper functions
//- Get relaxation factor for velocity field. Overriden in
// pimpleControl since different relaxation factor may be used for
@ -155,6 +155,34 @@ protected:
const volVectorField& U
) const;
//- Add consistent flux contribution arising from time derivative
// term. Note: aCoeff is parameter for clarity
void addDdtFluxContribution
(
surfaceScalarField& phi,
surfaceScalarField& aCoeff,
const surfaceVectorField& faceU,
const volVectorField& U,
const surfaceScalarField& rAUf,
const fvVectorMatrix& ddtUEqn
) const;
//- Add consistent flux contribution arising from the
// under-relaxation. Note: aCoeff is parameter for clarity
void addUnderRelaxationFluxContribution
(
surfaceScalarField& phi,
surfaceScalarField& aCoeff,
const volVectorField& U
) const;
//- Correct flux at the boundaries
void correctBoundaryFlux
(
surfaceScalarField& phi,
const volVectorField& U
) const;
private:
@ -163,7 +191,7 @@ private:
// Fields necessary for time and under-relaxation consistency
//- A coeff (A^~) arising from consistency formulation
mutable volScalarField* aCoeffPtr_;
mutable surfaceScalarField* aCoeffPtr_;
//- Face velocity needed for consistent formulation
mutable surfaceVectorField* faceUPtr_;
@ -230,16 +258,14 @@ public:
inline bool consistent() const;
// Time and under-relaxation consistency. Note: these functions could be
// made virtual and specific to SIMPLE, PISO and PIMPLE, but we will
// make them general in sense they account for arbitrary ddt scheme
// and arbitrary under-relaxation (e.g. if someone runs SIMPLE in
// transient mode or PISO with under-relaxation)
// Time and under-relaxation consistency.
// Note: argument matching U parameter needs to be equal U = H/A, where
// H and A come from convection-difussion equation only (without time
// derivative term)
//- Calculate and return time consistent flux (before pressure
// equation). At point of call, U = H/A where H and A come from
// convection-diffusion equation only.
void calcTimeConsistentFlux
//- Calculate consistent flux for transient solvers (before pressure
// equation).
void calcTransientConsistentFlux
(
surfaceScalarField& phi,
const volVectorField& U,
@ -247,10 +273,17 @@ public:
const fvVectorMatrix& ddtUEqn
) const;
//- Reconstruct velocity (after pressure equation). At point of
// call, U = H/A where H and A come from convection-diffusion
// equation only
void reconstructVelocity
//- Calculate consistent flux for steady state solvers (before
// pressure equation).
void calcSteadyConsistentFlux
(
surfaceScalarField& phi,
const volVectorField& U
) const;
//- Reconstruct velocity for transient solvers (after pressure
// equation and flux reconstruction).
void reconstructTransientVelocity
(
volVectorField& U,
const fvVectorMatrix& ddtUEqn,
@ -259,8 +292,17 @@ public:
const surfaceScalarField& phi
) const;
//- Reconstruct velocity for steady state solvers (after pressure
// equation and flux reconstruction).
void reconstructSteadyVelocity
(
volVectorField& U,
const volScalarField& rAU,
const volScalarField& p
) const;
//- Const access to aCoeff (needed for pressure equation)
const volScalarField& aCoeff() const;
const surfaceScalarField& aCoeff() const;
// Evolution