Which turbulence models you would like to combine?

In principle so-called zone-dependent physics modeling is not possible in Fluent. But there is an exception.

The exception is called ELES - Embedded Large-Scale Eddy Simulation. In this case you can have 2 or 3 zones. From a defined inlet you can start e.g. with a two-equation turbulence model like SST, followed by a zone with scale-resolving turbulence model and again followed by a downstream far-field zone, where the modeling is again switched to a 2-eq. turbulence model formulation. The modeled turbulent kinetic energy from the near-inlet zone is converted at the zone interface into resolved scales for the LES-method by the use of a synthetic turbulence generator (vortex generator). Further details on the ELES method in both CFX and Fluent can be found in the product manual.

Which turbulence models you would like to combine?

In principle so-called zone-dependent physics modeling is not possible in Fluent. But there is an exception.

The exception is called ELES - Embedded Large-Scale Eddy Simulation. In this case you can have 2 or 3 zones. From a defined inlet you can start e.g. with a two-equation turbulence model like SST, followed by a zone with scale-resolving turbulence model and again followed by a downstream far-field zone, where the modeling is again switched to a 2-eq. turbulence model formulation. The modeled turbulent kinetic energy from the near-inlet zone is converted at the zone interface into resolved scales for the LES-method by the use of a synthetic turbulence generator (vortex generator). Further details on the ELES method in both CFX and Fluent can be found in the product manual.

If you use the SST model (shear stress Transport model). It combines the ke,epsilon-model that is better in far wall fields and the k,omega-model which is better in wall near fields.

I think for modeling different turbulence models for two different zones, use the FFI simulation in the ADINA software is better than ANSYS. I have articles related to FSI simulation that this simulation is similar to FFI simulation

Yes it is possible. However, how much approximation is allowed to the specific challenge dictates this.

If you are able to approximate in the macro level it is possible.

Just to add that using in the same flow problem different models of the same type of formulation is somehow congruent whilst using different formulations requires much more care in the interfaces of the zones.

I agree with the comment on SST model by @ Frank Ulrich Rückert. SST is the main workhorse for (U)RANS of turbulent flows, since it combines the advantages of k-epsilon and k-omega formulations within one (U)RANS model. It should mainly be used, if the time, money or computational resources (or all together) are not sufficient for a more sophisticated analysis (like e.g. SAS-SST or SBES or any other kind of (E)LES / DDES approaches).

I agree with upon comment