Sub-model analysis – Points to ponder

The errors in the sub-model results could be because of following issues.1. Errors due to approximations/assumptions in the full model Harmonic model.2. Errors due to approximations/assumptions in the sub-model analysis.3. Stiffness change in the sub-model analysis due to fine meshing and contacts

The force problem becomes displacement problem in sub-model. Any changes in the stiffness due to fine meshing (Stiffness reduces due to refinement) or introduction of contacts (Contacts increase stiffness along the surfaces and reduce stiffness in couples DOF) will lead to change in the forces and leads to wrong stresses.

F

d

F

F1

F2

ᵟ

K

K1

K2

Force-Displacement curve

4. No C1 and C2 continuity at the cut boundaries due to imposed displacements.C1 continuity might be taken care-off by the solver but C2 continuity which is required for bending problems might be missing.

5. If contacts are not fully covered in the submodel, it might give rise to spurious or meaning-less stresses.

6. If the cut-boundary is close to fixity or coupling location, the stresses in the sub-model will be singular or meaning less.

Original element

Deformed element

Cut boundary

Coupling @ nodesContact @ surfaces

Sub-model domain Harmonic domain

7. Body forces and external loads of the sub-model geometry are missing ?Presently in our analysis procedure we are giving average acceleration for the required geometry. This we found to be in agreement with harmonic results by trial and error method. But for more accurate results, forces and moments from the full harmonic model are better bet.

Displacements @ cut boundary

Body forces and external forces @

sub-model