H modeling
==============


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

We use a AXIS model (quadratic elements).


.. image:: images/Object_16.svg
    :width: 484
    :height: 259


.. _RefImage_Object_16.svg:

The continuous contact formulation is used. This test case presents flip-flop phenomena (oscillations in the status of the contact linked to a "grazing" type of contact), to avoid this problem, we advance the piece by 10mm at the first step, the zone BC compressing the piece, there is no longer a problem.


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


.. image:: images/Object_17.png
    :width: 5.5201in
    :height: 1.5209in


.. _RefImage_Object_17.png:

Number of knots: 1174

Number of meshes: 510 TRIA6 and 152 SEG3

Number of nodes in contact: 21


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

The reference values are considered to be those of the C modeling.

The following displacement :math:`Y` of the point :math:`K` of the piece of paper with respect to the surface :math:`\mathrm{ABCDEFG}` of the die is tested.

.. csv-table::

    "**Identification**", "**Reference type**", "**Reference value**", "**Tolerance**"
    "Point :math:`K`/Point :math:`B` - :math:`\mathrm{DY}` ", "'NON_REGRESSION'", "10.12"," 0.1%"
    "Point :math:`K`/Point :math:`F` - :math:`\mathrm{DY}` ", "'NON_REGRESSION'", "155.08"," 0.1%"



notes
---------

The calculation is carried out by imposing a displacement on the back side of the block :math:`\text{(MN)}`.

The calculations do not converge with the default keywords used for convergence in STAT_NON_LINE, because the default value of RESI_GLOB_RELA = 1.E-6 is too restrictive (the forces to which the piece is subjected are initially relatively weak). To overcome this problem, use the keyword RESI_GLOB_MAXI = 1.E-6.