0003374: The calculation of alphaEff() in rhoCentralFoam is not alway right for laminar or turbulent flows - OpenFOAM Issue Tracking

ID	Project	Category	View Status	Date Submitted	Last Update

0003374	OpenFOAM	Bug	public	2019-10-20 10:55	2019-10-21 08:26

Reporter	haoyuan	Assigned To	henry
Priority	high	Severity	major	Reproducibility	always
Status	closed	Resolution	duplicate
Platform	GNU/Linux	OS	Ubuntu	OS Version	16.04

Summary	0003374: The calculation of alphaEff() in rhoCentralFoam is not alway right for laminar or turbulent flows
Description	The thermal diffusivity of a mixture, alphaEff(), is not right for laminar flow in OpenFOAM V5.0 and above, but right for turbulent flows. However, this calculation is right for laminar flow in OpenFOAM 4.1 or below but wrong for turbulent flows. When it is laminar, alphaEff() is calculated from (src/TurbulenceModels/compressible/ThermalDiffusivity/ThermalDiffusivity.H) //- Return the effective turbulent thermal diffusivity for enthalpy 158 // [kg/m/s] 159 virtual tmp<volScalarField> alphaEff() const 160 { 161 return alpha(); 162 } This part of calculation is the same in both V4.1 and V5.0 (or above). The alpha(), if we choose Sutherland transport mode, is calculated from (src/thermophysicalModels/specie/transport/sutherland/sutherlandTransportI.H) template<class Thermo> 154 inline Foam::scalar Foam::sutherlandTransport<Thermo>::alphah 155 ( 156 const scalar p, 157 const scalar T 158 ) const 159 { 160 161 return kappa(p, T)/this->Cpv(p, T); 162 } The line of 161 is "Cpv" in OpenFOAM 4.1 but "Cp" in OpenFOAM 5.0 or above. This is not consistent. In OpenFOAM 4.1 the alpha() calculated could be either the thermal diffusivity of enthalpy or inner energy, which depends on what thermal type we choose,(for instance "sensibleEnthalpy" or "sensibleInternalEnergy"). However, in OpenFOAM 5.0 or above, it could only return the thermal diffusivity of enthalpy. So, if we run a laminar simulation with rhoCentralFoamm for high-speed flows and choose the "sensibleInternalEnergy" package, the alphaEff() returned is the thermal diffusivity of inner energy in OpenFOAM V4.1, which is what we want, but it returns the thermal diffusivity of enthalpy in OpenFOAM V5.0, which is not right. If we run a turbulent simulation, the situation is just the opposite. The alphaEff() is calculated from (src/TurbulenceModels/compressible/EddyDiffusivity/EddyDiffusivity.H) //- Return the effective turbulent thermal diffusivity for enthalpy 133 // [kg/m/s] 134 virtual tmp<volScalarField> alphaEff() const 135 { 136 return this->transport_.alphaEff(alphat()); 137 } and if we choose "hePsiThermo" package (src/thermophysicalModels/basic/heThermo/heThermo.C) template<class BasicThermo, class MixtureType> 785 Foam::tmp<Foam::volScalarField> 786 Foam::heThermo<BasicThermo, MixtureType>::alphaEff 787 ( 788 const volScalarField& alphat 789 ) const 790 { 791 tmp<Foam::volScalarField> alphaEff(this->CpByCpv()*(this->alpha_ + alphat)); 792 alphaEff.ref().rename("alphaEff"); 793 return alphaEff; 794 } because alphat is always the turbulent thermal diffusivity of enthalpy, so if "this->alpha_" is for enthalpy, this calculation is self-consistent. As we have known, the calculation of "this->alpha_" could be either for enthalpy of inner energy, so it may be not self-consistent under certain condition.
Steps To Reproduce	A 2D hypersonic laminar cylinder flow case is attached. With V4.1 and V5.0, we would get different results on wall heat transfer rate but almost identical results on wall pressure.
Tags	No tags attached.

Date Modified	Username	Field	Change
2019-10-20 10:55	haoyuan	New Issue
2019-10-20 10:55	haoyuan	File Added: 2dCylinder.zip
2019-10-20 15:18	haoyuan	Note Added: 0010846
2019-10-21 08:25	henry	Relationship added	duplicate of 0003025
2019-10-21 08:26	henry	Assigned To	=> henry
2019-10-21 08:26	henry	Status	new => closed
2019-10-21 08:26	henry	Resolution	open => duplicate