In the previous steps, it was ensured that the boundary nodes of the partial model correspond geometrically with the inner boundary nodes of the overall model. Now the attribution of the boundary nodes of the partial model is analysed with regard to the effect in the merged model.

• The flow over the model boundary determined from the overall model is specified on the boundary nodes of the partial model. These boundary conditions must be removed from the partial model, as otherwise non-existent inflows or outflows will be simulated.

• Formally, it must also be checked for the partial model whether line data extends beyond the edge.

• It is also necessary to check whether the same attributes are defined on corresponding nodes in both the partial and overall model. For data types that can only be present once on a node, this step is of lesser importance, as in this case the value of the imported mesh overwrites the existing value of the source mesh.

• For attributes that are permitted more than once on a node (e.g. line-wise leakage coefficients), the corresponding attribute must be removed in one of the two meshes. Otherwise, for example, a receiving water course that runs along the edge will have a double leakage effect after the integration of the submodel. This must be avoided in any case to ensure data consistency.

• It is advantageous to obtain an overview of the existing attributes on the collapsing edges in order to recognise any inconsistencies and to be able to clarify the cause before certain data on the overall mesh is finally lost.

Once the nodes of the coinciding edges in the two models have been adjusted with regard to both the geometric position and the attribution, the models are merged using SPRING. The following strategy is advantageous:

Firstly, the mesh of the overall model is opened, from which the area of the partial model is removed. The mesh of the sub-model is then imported into the overall mesh (File Import Submesh...). If an attribute is already defined on an edge node during this import, the value from the partial model overwrites the existing value from the overall model. As the submodel is usually up to date, this assignment is desirable.

If node and element numbers are to remain unchanged in one of the model parts, this model part must always be read in first and then the other part imported, as the node numbering of the imported part will be changed.

If the partial model is opened first and the complete model is imported, the 3D boundary must be recalculated (Mesh 3D 3D boundary Mesh boundary =3D boundary).

After saving the integrated model, there is a new model mesh without inner edges

Carrying out consistency and plausibility checks