Premium
Crack Identification as an Optimization Task
Author(s) -
Engelhardt M.,
Stavroulakis G.E.,
Antes H.
Publication year - 2003
Publication title -
pamm
Language(s) - English
Resource type - Journals
ISSN - 1617-7061
DOI - 10.1002/pamm.200310525
Subject(s) - identification (biology) , generalization , boundary element method , waveform , task (project management) , artificial neural network , computer science , element (criminal law) , nondestructive testing , boundary (topology) , structural engineering , acoustics , algorithm , finite element method , artificial intelligence , engineering , mathematics , mathematical analysis , physics , telecommunications , systems engineering , radar , botany , quantum mechanics , law , political science , biology
In this paper, crack and hole identification problems are studied. From the knowledge about the reflected wave forms of an embedded crack, an identification may be performed by an evolutionary strategy or by a neural network. This approach corresponds to a generalization of the impact‐echo nondestructive evaluation technique. The numerical solution of the two‐dimensional elastodynamical problem, i.e., to determine the reflected waveforms, is performed by means of boundary element techniques.