.. _U2.06.04:

**u2.06.04** Instructions for the construction of dynamic scale models
========================================================================

**Summary:**

The models used for linear dynamic response calculations are getting bigger and bigger. To significantly improve calculation times, it is possible to build a small-sized subspace that makes it possible to project the complete problem, and thus accelerate the various response calculations. These reduced component models can also be used in sub-structuring approaches, and certain non-linear analyses (shocks, non-linear behavior). A methodology is also proposed to "condense" internal variables for linear dynamic analysis. This manual aims to help users build dynamic scale models and provides a framework for evaluating the quality of built scale models.

**We emphasize the value of the process of orthogonalization of projection bases (operator** DEFI_BASE_MODALE **, keyword** ORTHO_BASE **) making it possible to eliminate vectors that are too collinear, resulting from the concatenation of several families of vectors. A family containing almost collinear vectors will lead, in the worst case, to singular projected matrices and to a poorly posed problem.**




.. toctree::
    :hidden:

    self

.. toctree::
    :maxdepth: 2
    :numbered:

    Introduction
    R_duction_de_mod_le___principe_de_la_correction_statique
    Prise_en_compte_d_efforts_ext_rieurs__de_chocs__de_comportements_non-lin_aires
    R_duction_de_mod_le_pour_la_sous_structuration_dynamique_
    Construction_d_une_base_adapt_e_au_probl_me_dissipatif
    Cas_test_pour_Code_Aster
    Conclusions_sur_l_utilisation_des_m_thodes_de_r_duction
    R_f_rences