Prediction of Fracture Toughness Transition from Tensile Test Parameters Applying Artificial Neural Networks

0354571 - ÚFM 2011 RIV CZ eng K - Conference Paper (Czech conference)
Dlouhý, Ivo - Hadraba, Hynek - Chlup, Zdeněk - Kozák, Vladislav - Šmida, T.
Prediction of Fracture Toughness Transition from Tensile Test Parameters Applying Artificial Neural Networks.
New Methods of Damage and Failure Analysis of Structural Parts. Ostrava: VŠB - TU Ostrava, 2010 - (Strnadel, B.), s. 207-215. ISBN 978-80-248-2265-5.
[New Methods of Damage and Failure Analysis of Structural Parts. Ostrava (CZ), 06.09.2010-10.09.2010]
R&D Projects: GA ČR(CZ) GAP108/10/0466
Institutional research plan: CEZ:AV0Z20410507
Keywords : thermal ageing * brittleness * fracture
Subject RIV: JL - Materials Fatigue, Friction Mechanics

Reference temperature localizing the fracture toughness temperature diagram on temperature axis was predicted based on tensile test data. Regularization artificial neural network (ANN) was adjusted to solve the interrelation of these properties. For analyses, 29 data sets from low-alloy steels were applied. The fracture toughness transition dependence was quantified by means of master curve concept enabling to represent it using one parameter - reference temperature. Different strength and deformation characteristics from standard tensile specimens and notched specimens, instrumented ball indentation test etc. have been applied. A very promising correlation of predicted and experimentally determined values of reference temperature was found.
Permanent Link: http://hdl.handle.net/11104/0193545