[heisenmech]

Hi all,

Hope everyone is well. Can anyone please guide me on extracting turbulent heat flux <u'T'> from k-w SST and LES?

Thanks.

[vinerm]

Do you want this at a boundary or at interior faces of the mesh?

[heisenmech]

Originally Posted by vinerm

Do you want this at a boundary or at interior faces of the mesh?

Hi Vinerm,

Actually both as I want turbulent heat fluxes along a line normal to the wall.

Many thanks.

[vinerm]

For LES, this definition makes sense but has neither of these fields available. So, you have to define it either using k or subtract heat flux due to molecular diffusion from total to fetch turbulent heat flux.

[heisenmech]

Originally Posted by vinerm

For LES, this definition makes sense but has neither of these fields available. So, you have to define it either using k or subtract heat flux due to molecular diffusion from total to fetch turbulent heat flux.

Sorry for replying a bit late. I think that'd a bit tricky and unreliable to reconstruct turbulent heat fluxes from turbulent kinetic energy so I'll probably stick to unsteady simulations. For LES or URANS, should I directly multiply RMS u and RMS T to get <u't'>? or is there any alternative way to do so?

Many thanks

[vinerm]

For RANS, you can determine u' using k,



Whether LES or RANS, turbulent heat flux is product of u' with mean T and not T'. Another method of determining turbulent heat flux in RANS is via and mean temperature gradient. You can use this for LES as well, however, for LES, you directly have u'. I am not sure but Fluent may directly have turbulent heat flux field available.

[heisenmech]

Originally Posted by vinerm

For RANS, you can determine u' using k,



Whether LES or RANS, turbulent heat flux is product of u' with mean T and not T'. Another method of determining turbulent heat flux in RANS is via and mean temperature gradient. You can use this for LES as well, however, for LES, you directly have u'. I am not sure but Fluent may directly have turbulent heat flux field available.

I've further looked into the definition of turbulent heat fluxes but I'm actually a bit confused with your comment as I think it must be T' -fluctuating temperature. Can you please refer me a book or paper where it says it is mean T rather than T'?

Many thanks.

[vinerm]

You are correct, in similarity with Reynolds stress, turbulent heat flux uses temperature fluctuation and not it's mean. It's only if turbulent diffusivity is used that the gradient of mean makes sense for turbulent heat flux determination.