[syavash]

Hi Foamers,

I looked over GenSGSStress class under incompressible LES. Why nuSgs is defined in this class, while not read nor written? AFAIK this class only deals with sub-grid stress tensor B, and not sub-grid scale viscosity nuSgs. However, the latter has been implemented in calculating divDevReff which has confused me so much!

Now I appreciate if anyone could give an explanation.

Thanks,
Syavash

[syavash]

This is really important to me,
Please, anyone got an idea??!

Thanks

[syavash]

Could anyone give me an explanation please??!

Thanks,
Syavash

[desanga]

Hi,

I'm not sure this reply would be sufficient to resolve your problem but let me give a try.

GenSgsStress is just a base class and inside this class nuSgs is initialized with nu (or laminar viscosity). if you have a look inside LRRDiffStress which is a derived class of GenSgsStress, you will see nuSgs_ is calculated during the runtime.

inside the momentum equation, divDevReff(U) will invoke divDevReff(U) in GenSgsStress and turbulence->correct() in pisoFoam/pimpleFoam will execute correct(const tmp<volTensorField>& tgradU) (see turbulence.H) which will lead to runtime calculation of nuSgs_.

Hope this will help you a little

[syavash]

Originally Posted by desanga

Hi,

I'm not sure this reply would be sufficient to resolve your problem but let me give a try.

GenSgsStress is just a base class and inside this class nuSgs is initialized with nu (or laminar viscosity). if you have a look inside LRRDiffStress which is a derived class of GenSgsStress, you will see nuSgs_ is calculated during the runtime.

inside the momentum equation, divDevReff(U) will invoke divDevReff(U) in GenSgsStress and turbulence->correct() in pisoFoam/pimpleFoam will execute correct(const tmp<volTensorField>& tgradU) (see turbulence.H) which will lead to runtime calculation of nuSgs_.

Hope this will help you a little

Thanks for your reply

Honestly, I do not still get why nuSgs is used in GenSGSStress to calculate divDevReff, as it is supposed to use only sub-grid stress tensor B. This would matter if I want to develop a SGS model that only calculates B, and not nuSgs. I guess I should manually set nuSgs to zero in current implementation.
Ah..., about this LRRDiffStress class, does it exist in OF 2.3.1?! What is it supposed to do??

Thanks,
Syavash

[desanga]

Originally Posted by syavash

Thanks for your reply

Honestly, I do not still get why nuSgs is used in GenSGSStress to calculate divDevReff, as it is supposed to use only sub-grid stress tensor B. This would matter if I want to develop a SGS model that only calculates B, and not nuSgs. I guess I should manually set nuSgs to zero in current implementation.
Ah..., about this LRRDiffStress class, does it exist in OF 2.3.1?! What is it supposed to do??

Thanks,
Syavash

check
src/turbulenceModels/incompressible/LES/LRRDiffStress/
OR
https://github.com/OpenFOAM/OpenFOAM.../LRRDiffStress


try to get the full picture with this derived class because base class itself does not give full meaning. most of the sgs models implement through sgs viscosities which are calculated using "B" . if you set nusgs manually to zero, this may result laminar simulation.

However, I'm not very familiar with this particular LES model. try to find and read "Launder, Rece & Rodi 1975" (I think you can download this using "http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=386515&fileId= S0022112075001814" but double check)