[jairoandres]

Dear all

I have been using the DESHybrid scheme for a long time without any issues (PimpleFoam, PimpleDyMFoam, sprayFoam). I am moving now to use DPMFoam family of solvers and I have an issue with the scheme:

"--> FOAM FATAL ERROR:
Scheme requires a turbulence model to be present. Unable to retrieve the turbulence model from the mesh database"

The DESHybrid reads like this for U.air:

default none;
div(alphaPhi.air,U.air) Gauss DEShybrid
linear // scheme 1
linearUpwind grad(U.air) // scheme 2
delta.air // LES delta name OF 1912
0.61 // DES coefficient, typically = 0.61
12 // Reference velocity scale
0.05 // Reference length scale
0 // Minimum sigma limit (0-1)
1 // Maximum sigma limit (0-1)
1e-3; // Limiter of B function, typically 1e-03

"delta.air" is defined (from OpenFOAM 1906) as otherwise, the hybrid scheme would not find the delta.air field.

I have also modified the DPMTurbulenceModels.C file and recompiled to include the hybrid LES/RANS models as they are not available for the DPMFoam based solvers, i.e:

#include "kOmegaSSTDDES.H"
makeLESModel(kOmegaSSTDDES);

I am looking at the DESHybrid.H library to see the possible localization of the error, but I am not having luck with this.

I tried modifying the DEShybrid, to let it know the actual names of the fields, i.e:
// **** MODIFICATION ********
// tmp<volTensorField> gradU(fvc::grad(U)); original
tmp<volTensorField> gradU(fvc::grad(U.air)); // DPMFoam

but then an error is produced in the compilation:
error: 'const volVectorField' has no member named 'air'
tmp<volTensorField> gradU(fvc::grad(U.air)); // DPMFoam


Please let me know if you have any ideas regarding how to fix this error and modify the DESHybrid scheme in order to be able of using hybrid turbulence models in DPMFoam and other similar solvers. I am almost positive the error is produced for adding the .air tag to the fields and somewhere the DESHybrid scheme is not reading well this information. The DESHybrid.H file can be found here:

https://www.openfoam.com/documentati...8H_source.html

Thank you a lot.

[HPE]

I would create a feature request ticket in GitLab (not GitHub) if that solver should work with DESHbrid scheme.

[jairoandres]

Thank you. I will do that as I was unable of modify the hybrid scheme to work. Meanwhile I am working with the LUST scheme, which in theory should produce acceptable results in LES zones.