# Mesh type and element size#

Triangular meshes are flexible in terms of geometry (e.g. rounded, pitched-in corners) and in general produce a regular mesh. Triangular elements are suitable for RC constructions, since their components usually feature a relativeley low slenderness.

Quadrilateral meshes are less flexible, but tend to produce better results for thin surfaces (high slenderness), like steel plates.

Combined meshes try to combine the advantages of triangular and quadrilateral meshes. But in many cases they produce an irregular mesh. Combined meshes are recommended for thin surfaces with complex geometry.

# Empirical values for element size#

For buildings in concrete and masonry, triangular meshes with the following element sizes have proven to serve the purpose in many cases.

• Floors 1,50 … 2,00 m
• Walls 1,00 … 1,50 m

For very narrow components it can be of advantage to choose a smaller element size. If possible it is not recommended to use element sizes below 0.50 m. Depending on the calculation, individual domains can be meshed with considerably greater element sizes (e.g. floors in seismic calculations).

The mesh adjustments (“fit mesh to loads”, “adjust mesh to column head”) forces mesh nodes an lines at defined places.

• Load application point for nodal loads (mesh node)
• Load application line for line loads (mesh line)
• Contour of load application area for surface loads (mesh line)
• Corners of column cross-sections (mesh node)

The fitting of the mesh to loads is meant to guarantee a most accurate consideration of the loads. If a load acts on a point inside an element, it is spread out for calculation purposes to the nearest nodes and is (taken literally) not a point load any more, but a ‘surface load’ over a small area.

If the mesh is adjusted to column heads, the elastic peaks in the moment distribution can me smoothed out automatically. To do so, it is necessary to introduce a mesh node at the corners of the column cross-section. This leads to a lot of nodes closely together for round columns, which is not ideal. In extreme cases the mesh generation can be impeded by this situation.

To use the mesh adjustment for thin walled cross-sections (e.g. steel profiles) and still get a good mesh, the corners of the convex contour of the cross-section are used to adjust the mesh.

Source: http://www.ingware.com