Prediction of fracture toughness temperature dependence applying neural Network

0354379 - ÚFM 2011 FR eng C - Konferenční příspěvek (zahraniční konf.)
Dlouhý, Ivo - Hadraba, Hynek - Chlup, Zdeněk - Šmida, T.
Prediction of fracture toughness temperature dependence applying neural Network.
NT2F10 – New Trends in Fatigue and Fracture CongressMetz. Metz: LaBPS, 2010, s. 1-9.
[NT2F10 – New Trends in Fatigue and Fracture CongressMetz. Metz (FR), 30.08.2010-01.09.2010]
Grant CEP: GA ČR(CZ) GAP107/10/0361
Výzkumný záměr: CEZ:AV0Z20410507
Klíčová slova: brittle to ductile transition * fracture toughness * artificial neural network

J-integral are adopted for describing deformation and fracture of a solid with a notch. The notch failure assessment diagram is constructed as the unified diagram for a notch as well as a crack. Effects of local constraint were incorporated into the basic criteria equations which allow estimating the notch constraint-dependent fracture toughness and notch failure assessment diagrams for various components with a notch and various types of loading. The volumetric method suggested by Pluvinage has been discussed from a viewpoint of 3D finite element analysis of the elastic-plastic stress along the crack front under uni-axial and biaxial mode I loading. The distribution of the effective distance along the crack front was computed for the above-mentioned cases. Calculation of the J-integral in plates with lateral and central U- and V-blunt notches under mode I loading can be based on approximate analytical formulas taking into account both a linear elastic and non-linear elastic material.
Trvalý link: http://hdl.handle.net/11104/0193392