Aboussalih Amira is a teacher at the Institute of Applied Sciences at Constantine 1 Algeria University, with the grade of Lecturer "A". A good part of her professional experience was in the field of civil engineering, in particular soil study. She continued her studies at university and later obtained a Magister degree and a Doctorate in Mechanical Engineering. The side of scientific research occupies a primordial place in her university life, in particular the study of the behavior of materials and their fatigue by cyclic stresses. She has published a few articles in international journals, and communications in this context.
The present work, is concerned with the evaluation of the nonlinear behavior of 316L steel used in several fields, industry, automotive, aeronautics, medicine .... subjected to sinusoidal movements. This investigation allowed us to analyze the different responses of steel in low cycle fatigue, and to perceive the phenomenon of ratcheting leading, to the damage of the structure by an excess of plastic deformation, under loading cycles. The prediction of the behavior of the steel studied, reveals that the ratcheting increases according to the average stress when the stress amplitude is constant, and decreases according to the attenuation of the stress amplitude when the average stress is constant.
The simulation campaign allowed us to present an assessment, of the conditions favoring the acceleration of the ratcheting, includes several scenarios of periodic loadings, with strain and stress imposed under uniaxial tension/compression and biaxial tension/torsion loading.
The quality of the model to predict, the macroscopic elastoplastic behavior of 316 L is considered as a basic step for the description, of the hardening produced during cyclic loading. In the present work we have chosen the Chaboche model, this model is capable of translating the real behavior of materials; it is based on the use of a law combining an isotropic hardening and a nonlinear kinematic hardening.The predictions of this model are evaluated by subjected to cyclic loading within the software ZébuLon.