New documents recently published

Mechanical behavior for numerical simulations R5.03.27

The material behavior is an essential input data in most mechanical simulation studies. In a numerical simulation this behavior is taken into account through a model more or less complex, assumed sufficiently accurately reproduce the behavior of the real material. The parameters controlling the model have been adjusted in advance from experimental data. Any model is a simplified and inaccurate representation of reality, it is essential to ensure that the choice of formulation and the selected parameter set is relevant to the entire field of stresses characterizing the study.
Apart from the simple case where we can settle for a linear elastic behavior or plastic behavior with isotropic hardening, the choice of a constitutive law for a given material is a long and delicate process, requires the availability of relevant experimental data that it is not always easy to collect.
This document is complementary to U2.04.03 who advises a user wishing to perform calculations with non-linear elastic-plastic type, or visco-elastic-plastic behavior to choose a constitutive law adapted to models considered.

Implementation of a calculation of eigenmodes of a structure U2.06.01

This document presents an overview of the different approaches available in Code_Aster to calculate eigenmodes of a mechanical structure. These approaches are described starting from the most simple to implement, for standard studies and moving progressively towards more sophisticated implementations for advanced studies. Necessary sequences of Code_Aster operators is presented without going into the details of each operator.

PERF014 – Parallel solving of contact between two plates V1.01.311

The objective of this test case is to measure the performance of a parallel calculation using a large number of degrees of freedom of contact. It consists of two plates in contact with a rigid foundation.
This test case is available in three models almost identical. The differences are related to the refinement of the meshes used and the number of processors :

• 1) Modeling A: 8900 elements on one processor,
• 2) Model B: 40000 elements on 4 processors,
• 3) Modeling C: 63000 elements on eight processors.