1. Overview of order PROJ_CHAMP

The PROJ_CHAMP command allows you to project fields to nodes or fields by elements to another mesh.

For the fields at the nodes, 4 methods are available:

METHODE: “COLLOCATION”

METHODE: “NUAGE_DEG_0”

METHODE: “NUAGE_DEG_1”

METHODE: “COUPLAGE”

The “COLLOCATION” method uses the shape functions of the \(\mathit{ma1}\) mesh elements. It is detailed in paragraph [§ 2].

The other 2 methods use smoothing the values of the field in the vicinity of the point where you want to project the field. These 2 methods are detailed in paragraph [§ 3].

Finally, the “COUPLAGE” method is a variant of the “COLLOCATION” method. It is specially developed for fluid-structure coupling with*Code_Saturne* and should therefore not normally be used outside this framework.

For fields with “ELNO” elements, you can use the “COLLOCATION” method. The problem is the same as for a field with nodes but the projected field is always continuous while the initial field is discontinuous. In addition, if a node in mesh \(\mathit{ma2}\) is located at the border of 2 elements in mesh \(\mathit{ma1}\), the projected value will be random: it will come from either side of the border according to the numbering of the elements. This is why it is not recommended to project the “ELNO” fields.

For fields with “ELEM” elements, you can still use the “COLLOCATION” method. The value of the field (constant per element) is duplicated on the nodes of the element and we are back to the case of the “ELNO” fields.

For fields by elements” ELGA “, a dedicated method (” ECLA_PG “) is described in paragraph [§ 4]

Regardless of the method, the user has the option of projecting only a « piece » of field onto a « piece » of the \(\mathit{ma2}\) mesh. This possibility is offered by the keyword factor VIS_A_VIS. A field chunk is the restriction of the field on a set of nodes (or meshes) in mesh \(\mathit{ma1}\). A piece of mesh \(\mathit{ma2}\) is a subset of the nodes in \(\mathit{ma2}\).

In the rest of the document, we will no longer talk about a subset of a mesh, we will act as if we were projecting all the \(\mathit{ma1}\) mesh onto the whole \(\mathit{ma2}\) mesh.