Modeling A
==============


Characteristics of modeling
------------------------

3D modeling is used. For an interface and for the only "catalog of predefined shapes" and "fields at nodes" methods in DEFI_FISS_XFEM, we use a D_ PLAN modeling of the same geometry limited to the :math:`(X,Y)` plane because these methods are not available in 3D.


Characteristics of the mesh
----------------------------

The 3D mesh contains 64 elements of type HEXA8 of dimension :math:`\mathrm{1x1x0.5}\mathrm{mm}`:


.. image:: images/10000201000003590000025714B904F5D5CE2DF1.png
    :width: 3.0472in
    :height: 2.1299in


.. _RefImage_10000201000003590000025714B904F5D5CE2DF1.png:

The D_ PLAN mesh contains 16 elements of type QUAD4 of dimension :math:`\mathrm{1x1}\mathrm{mm}`:


.. image:: images/100002010000036D0000020436ADBED7B1B61F89.png
    :width: 3.1043in
    :height: 1.8264in


.. _RefImage_100002010000036D0000020436ADBED7B1B61F89.png:

The same mesh is used for both the structure and the grid.


Tested sizes and results
------------------------------

In each node, we test the difference between the level set on the mesh of the structure and the level set on the mesh of the grid. A value exactly equal to zero is obtained everywhere, which makes it possible to check the coincidence of fields at the nodes between structure and grid.


notes
---------

In the test we are forced to use a tolerance in TEST_TABLE. This tolerance is taken to be equal to 0.01. This has no influence on the test results because you are testing values that are always exactly equal to zero.