[GeoMechanicsApplication] Improve performance of DGeoSettlement workflow

[rfaasse]

As a user, I want the DGeoSettlement workflow to run as fast as possible.

Similar to the performance improvements of the DGeoFlow workflow (see #12811), the DGeoSettlement workflow can be profiled and (probably improved). Especially since settlement has multiple stages and a time stepping scheme, this will be an educational exercise about what impacts the performance of a more complex workflow.

Acceptance Criteria

• The learned lessons from the DGeoFlow issue [GeoMechanicsApplication] Performance issue in D-GeoFlow #12811 are applied for the components in the settlement workflow
• A VTune analysis is done for DGeoSettlement, based on which performance bottlenecks are tackled where possible.
• The improved performance after issue completion is documented