[Weiqiang Liu]

Hi all,

I am modeling a hydrogen combustion in porous zone with fluent. I used a UDS to solve solid energy equation and also the energy equation of fluid is modified.

For fluid energy equation, I modified thermal conductivity of fluid to an effective thermal conductivity. Specificly, effective thermal conductivity is mass weighted thermal conductivity added with an additional term.

Then I need to define source term for both solid and fluid energy equation and fluid thermal conductivity will be used in the source terms as well.

My question is: for the thermal conductivity emerged in source term, should I should use the value return by DEFINE_PROPERTY which is effective thermal conductivity or the original mass weighted averaged velocity? According to the symbols emerged in paper, it seems like original mass weighted thermal conductivity should be used. However, I am not very sure.

Can anybody give me some suggestions?

Best regards

Weiqiang

[vinerm]

DEFINE_PROPERTY returns value to Fluent, not to UDF. If you want to fetch value in UDF, use C_K_L(c,t). Which value to use depends on the source term. Fluent uses averaged properties only with equilibrium model. With non-equilibrium model, individual material properties are used. So, if you have different source terms for fluid and solid, you should use fluid and solid thermal conductivity, resp., not averaged. But, as mentioned above, it depends on source term formulation; the phenomenon which is the cause for the source.

[Weiqiang Liu]

Originally Posted by vinerm

DEFINE_PROPERTY returns value to Fluent, not to UDF. If you want to fetch value in UDF, use C_K_L(c,t). Which value to use depends on the source term. Fluent uses averaged properties only with equilibrium model. With non-equilibrium model, individual material properties are used. So, if you have different source terms for fluid and solid, you should use fluid and solid thermal conductivity, resp., not averaged. But, as mentioned above, it depends on source term formulation; the phenomenon which is the cause for the source.

Do you mean C_K_L(c,k) access volume averaged conductivity not fluid conductivity?

I tried to compute non equilibrium phenomenon with my own UDF and still kept the in built non equilibrium model in fluent disabled

[vinerm]

C_K_L returns property being applied to the cell. If you have equilibrium model enabled, then it will return averaged property since there is only one cell zone. If you enable non-equilibrium model, then C_K_L will return property depending upon the cell zone that it addresses. Do note that by disabling the NET model, ET is still enabled and thermal equation is still based on porosity averaged properties of fluid and solid. So, your fluid thermal field is not really fluid thermal field, but thermal field based on fluid-solid averaged zone. If you really want only UDS to be active for solid, then set the porosity to 1. It will have no effect on momentum since porosity does not play a role there, but it will affect superficial velocities and reactions.

[Weiqiang Liu]

Originally Posted by vinerm

C_K_L returns property being applied to the cell. If you have equilibrium model enabled, then it will return averaged property since there is only one cell zone. If you enable non-equilibrium model, then C_K_L will return property depending upon the cell zone that it addresses. Do note that by disabling the NET model, ET is still enabled and thermal equation is still based on porosity averaged properties of fluid and solid. So, your fluid thermal field is not really fluid thermal field, but thermal field based on fluid-solid averaged zone. If you really want only UDS to be active for solid, then set the porosity to 1. It will have no effect on momentum since porosity does not play a role there, but it will affect superficial velocities and reactions.

what if I use C_K_L to access volume averaged thermal conductivity and then compute fluid thermal conductivity by myself according to the volume averaged equation since I know solid thermal conductivity as a constant.

[vinerm]

You can do that; however, fluid's thermal conductivity is a user-specified property. So, until and unless this is a very complicated function requiring gradients of fields, you can directly calculate the thermal conductivity in your UDF.

[Weiqiang Liu]

Originally Posted by vinerm

You can do that; however, fluid's thermal conductivity is a user-specified property. So, until and unless this is a very complicated function requiring gradients of fields, you can directly calculate the thermal conductivity in your UDF.

Do you mean I can just calculate thermal conductivity in DEFINE_SOURCE where I need to use fluid thermal conductivity?

[vinerm]

If you have provided constant thermal conductivity for fluid, then you don't need to fetch or calculate. However, if you are using some profile or UDF, then you already know the equation and you can calculate it within UDF.

[Weiqiang Liu]

Originally Posted by vinerm

If you have provided constant thermal conductivity for fluid, then you don't need to fetch or calculate. However, if you are using some profile or UDF, then you already know the equation and you can calculate it within UDF.

yes, I know the equation. it's just a mass weighted average plus some additional terms. However, I think I can still use C_K_L to access thermal conductivity of cell and then use volume averaged equation to get gas conductivity. I mean I just need to calculate thermal conductivity in DEFINE_PROPERTY macro for one time. For other macros in which I need to use fluid thermal conductivity. I just need to use C_K_L(c,t) to fetch the value, right?