Using statistical compatibility to derive advanced probabilistic fatigue models

0355184 - ÚFM 2011 RIV NL eng J - Journal Article
Fernández-Canteli, A. - Castillo, E. - López-Aenlle, M. - Seitl, Stanislav
Using statistical compatibility to derive advanced probabilistic fatigue models.
Procedia Engineering. Roč. 2, č. 1 (2010), s. 1131-1140. E-ISSN 1877-7058.
[Fatigue 2010. Praha, 06.06.2010-11.06.2010]
Institutional research plan: CEZ:AV0Z20410507
Keywords : Fatigue models * Statistical compatibility * Functional equations
Subject RIV: JL - Materials Fatigue, Friction Mechanics

Heuristic deterministic models are often proposed to reproduce S−N and ε−N fields as well as crack growth curves aiming at providing basic material characterization to be used in lifetime design. Usually, they arise from the intuition of the researchers and are supported by experimental data, sometimes being complemented by micromechanical considerations based on good knowledge of physical and metallurgical properties of the material. In spite of their utility, this procedure implies serious limitations. In this work, some new methodological suggestions are presented allowing the functional structure of the problem under consideration to be derived, in an attempt of avoiding arbitrary assumptions. The relevant variables involved in the problem are recognized from experimental evidence, a reduced set of dimensionless variables is selected using dimensional analysis, a set of compatibility conditions are established and the constraints to be fulfilled by the model considered.
Permanent Link: http://hdl.handle.net/11104/0194014