G modeling
==============


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

This modeling is similar to modeling A. The main difference is in the management of the time step. A first calculation is carried out with a time discretization approximately 2 times coarser than that of modeling A. Then, the list of the moments actually calculated is extracted from the result (taking into account any sub-divisions of the time step during the first calculation) and a second list of moments is created, 2 times finer than this extract. Finally, we perform a 2nd calculation, identical to the first, but with the more detailed list of moments.


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

The tests focus on deformations at the end of load :math:`{\varepsilon }_{\mathrm{xx}}(B)` and :math:`{\varepsilon }_{\mathrm{xy}}(B)` for both calculations. The tolerances of the 1st calculation are 2 times looser than those of the 2nd calculation.


Calculation with the rough list
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


.. csv-table::

    "**Identification**", "**Reference Type**", "**Value**", "**Tolerance**"
    ":math:`{\varepsilon }_{\mathrm{xx}}` at the moment :math:`A` ", "'ANALYTIQUE'", ":math:`\mathrm{3,5265}{.10}^{\mathrm{-}2}` "," 0.4%"
    ":math:`{\varepsilon }_{\mathrm{xy}}` at the moment :math:`A` ", "'ANALYTIQUE'", ":math:`\mathrm{2,0471}{.10}^{\mathrm{-}2}` "," 1.2%"


These tests coupled with non-regression tests.


Calculation with the refined list
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


.. csv-table::

    "**Identification**", "**Reference Type**", "**Value**", "**Tolerance**"
    ":math:`{\varepsilon }_{\mathrm{xx}}` at the moment :math:`A` ", "'ANALYTIQUE'", ":math:`\mathrm{3,5265}{.10}^{\mathrm{-}2}` "," 0.2%"
    ":math:`{\varepsilon }_{\mathrm{xy}}` at the moment :math:`A` ", "'ANALYTIQUE'", ":math:`\mathrm{2,0471}{.10}^{\mathrm{-}2}` "," 0.6%"