Bugs of `dfSteadyFoam` using `flareFGM` combustion model

[TimoLin]

Some bugs related to the flareFGM model in steady state solver.

Test case: twoD_SandiaD_flareFGM from examples, the settings has been adapted according to the steady solver dfSteadyFoam.

1. ddt term

ddt term of each flamelet variable's equation should be removed foir the steady solver

deepflame-dev/src/dfCombustionModels/FGM/baseFGM/baseFGM.C

Lines 778 to 789 in 1be82b6

	fvScalarMatrix ZvarEqn
	(
	fvm::ddt(rho_,Zvar_)
	+(
	buffer_
	? scalarUWConvection->fvmDiv(phi_, Zvar_)
	: fvm::div(phi_, Zvar_)
	)
	-fvm::laplacian( mut/Sct_+mu/Sc_, Zvar_)
	-(2.0*mut/Sct_*(fvc::grad(Z_) & fvc::grad(Z_)))
	+(2.0*rho_*chi_Z_)
	);

2. Treatment of chi_zc for 4D table

If the table only has 4 dimensions (Z, Zvar, C, Cvar), then the contribution of Chi_ZC shouldn't be added to the source term of C, either set Chi_ZC value to 0 or don't add it to the omega_c term.

deepflame-dev/src/dfCombustionModels/FGM/flareFGM/flareFGM.C

Lines 296 to 305 in 1be82b6

	+ (
	tableSolver::scaledPV_
	? ( this->chi_ZCells_[celli]*this->cCells_[celli]
	*this->lookup3d(this->NH,this->h_Tb3,hLoss,this->NZ,this->z_Tb3,this->ZCells_[celli],
	this->NGZ,this->gz_Tb3,gz,this->d2Yeq_Tb2)
	+ 2*this->chi_ZcCells_[celli]*this->lookup3d(this->NH,this->h_Tb3,hLoss,
	this->NZ,this->z_Tb3,this->ZCells_[celli],
	this->NGZ,this->gz_Tb3,gz,this->d1Yeq_Tb2) )
	: 0.0
	);

3. relaxation of flamelet variable equations

The relaxation_ switch seems to be duplicated and may cause misunderstanding.

deepflame-dev/src/dfCombustionModels/FGM/baseFGM/baseFGM.C

Lines 518 to 536 in 1be82b6

	if(buffer_)
	{
	Info<< "UW schemes used for scalars" << endl;
	fvScalarMatrix ZEqn
	(
	fvm::ddt(rho_,Z_)
	+scalarUWConvection->fvmDiv(phi_, Z_)
	-fvm::laplacian( mut/Sct_ + mu/Sc_ , Z_)
	==
	- SZ
	);
	if(relaxation_)
	{
	ZEqn.relax();
	}
	ZEqn.solve();
	}

If user want's to apply relaxation to Z, the following settings should be both set in constant/combustionProperties and system/fvSolution:

// combustionProperties
flareFGMCoeffs
{  
...
  relaxation       true;
...
}

// fvSolution
relaxationFactors
{
...
    equations
    {
...
        "(Z|Zvar|c|cvar|Zcvar|Ha).*"                       0.5;
...
    }
}

[mactone]

May I ask if there's any new progress? Still waiting for dfSteadyFoam example. Thank you.

[TimoLin]

May I ask if there's any new progress? Still waiting for dfSteadyFoam example. Thank you.

Sorry for the late reply.
The bug fixes and enhancements of dfSteadyFoam are still work in progress. But for now this issue is not in a high priority.
I'll keep you noticed once we have updates.