Skip to content

Why the implementation of the solid-fluid coupling in the specfem2d and specfem3d_globe is different? #821

Description

@lyuchao

In the specfem2D/compute_coupling_viscoelastic_ac.f90, we can find that the definition of pressure is minus potential_dot_dot_acoustic. This can be understood due to the definition relation between pressure (P) and the potential of displacement(phi): P = - /partial_{tt} potential. We have benchmarked the specfem2d with our current DFDM 2D global, the waveforms are the same.

! compute pressure on the fluid/solid edge
  pressure = - potential_dot_dot_acoustic(iglob)

However, in the specfem3D/compute_coupling.f90, the pressure = - RHO_TOP_OC * accel_outer_core.
My understanding is that: accel_outer_core = /partial_{tt} phi (second derivatives with respect to time). This implementation is different from the specfem2d case. And it is also different from the previous velocity potential formulation (-rho* partial_t velocity potential).

pressure = - RHO_TOP_OC * accel_outer_core(iglob_oc)

So I feel confused.

After checking the source codes, it seems that RHO_TOP_OC is a ratio between RHO_TOP_OC and average density. That makes sense.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions