[xitao21]

Hello,everyone

I have a problem of having continuity so high that it is between 1e+00 and 1e-01 and not converged.here is my situation.

I am modeling the surface layer of an atmospheric boundary layer (on a 3km by 3km by 100m 3D grids where the cells are all 13 m by 13 m by about 2m averaged, uniform), over slightly complex terrain (roughness length = 0.02m).

I'm using wind horizontal velocity data at one single moment extracted from mesoscale and microscale meteorological model(WRF/CALMET) to initialize Fluent by the profile file.I don't have the vertical velocity,TKE and TDR data right now so I have to calculate them from mass conservation law and empirical parameterization. I checked the wind data put into Fluent and it is correctly interpolated.

I'm using the segregated steady solver, standard k-epsilon closure model with the pressure-based solver,pressure-velocity coupling is SIMPLEC, relaxation factors all set to 0.9. My spatial discretization settings are Pressure: PRESTO!, and Momentum, TKE, and TDR: 2nd order upwind.density is set to constant.

I had the convergence when using power law and log law of wind in the ideal experiment and the same empirical parameterization of TKE and TDR with the same mesh. the mesh quality is good by the way.

I have used the realizable and RNG k epsilon with the same data and it is still
not converged and I still got high continuity,lowering relaxation factor by 0.4 for each parameter including momentum doesn't help it converged.

I firstly thought it is the problem of the wind data complexity,
some of the wind direction interpolated to each side of the mesh have strong small-scale wind deflection with each other.

How can i get continuity converged? Any help or suggestion will be appreciated!

Xitao

[cylindrax]

Originally Posted by xitao21

Hello,everyone

I have a problem of having continuity so high that it is between 1e+00 and 1e-01 and not converged.here is my situation.

I am modeling the surface layer of an atmospheric boundary layer (on a 3km by 3km by 100m 3D grids where the cells are all 13 m by 13 m by about 2m averaged, uniform), over slightly complex terrain (roughness length = 0.02m).

I'm using wind horizontal velocity data at one single moment extracted from mesoscale and microscale meteorological model(WRF/CALMET) to initialize Fluent by the profile file.I don't have the vertical velocity,TKE and TDR data right now so I have to calculate them from mass conservation law and empirical parameterization. I checked the wind data put into Fluent and it is correctly interpolated.

I'm using the segregated steady solver, standard k-epsilon closure model with the pressure-based solver,pressure-velocity coupling is SIMPLEC, relaxation factors all set to 0.9. My spatial discretization settings are Pressure: PRESTO!, and Momentum, TKE, and TDR: 2nd order upwind.density is set to constant.

I had the convergence when using power law and log law of wind in the ideal experiment and the same empirical parameterization of TKE and TDR with the same mesh. the mesh quality is good by the way.

I have used the realizable and RNG k epsilon with the same data and it is still
not converged and I still got high continuity,lowering relaxation factor by 0.4 for each parameter including momentum doesn't help it converged.

I firstly thought it is the problem of the wind data complexity,
some of the wind direction interpolated to each side of the mesh have strong small-scale wind deflection with each other.

How can i get continuity converged? Any help or suggestion will be appreciated!

Xitao

Dear How did you create the UDF from the mesoscale meteorological model? I am trying to couple WRF and Coupling however I have difficulties creating the UDF. I have data point along the face but I don't know how to create the UDF files. Any suggestions?