Premium
Locating the Crack Tip Using Displacement Field Data: A Comparative Study
Author(s) -
Zanganeh M.,
LopezCrespo P.,
Tai Y. H.,
Yates J. R.
Publication year - 2013
Publication title -
strain
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.477
H-Index - 47
eISSN - 1475-1305
pISSN - 0039-2103
DOI - 10.1111/str.12017
Subject(s) - displacement (psychology) , displacement field , simplex algorithm , digital image correlation , position (finance) , stress intensity factor , genetic algorithm , stress (linguistics) , newton's method , simplex , algorithm , field (mathematics) , computer science , mathematics , structural engineering , mathematical optimization , geometry , fracture mechanics , physics , optics , finite element method , engineering , nonlinear system , linear programming , philosophy , psychotherapist , linguistics , psychology , quantum mechanics , finance , economics , pure mathematics
The evaluation of stress intensity factors from experimentally determined crack‐tip stress or displacement fields almost always requires that the location of the crack tip is identified beforehand. In this work, a study has been performed to compare how different methods to locate the crack‐tip position from the displacement field around a crack tip obtained by digital image correlation influence the estimation of stress intensity factors. The methods used were two constrained Newton type methods: the trust‐region reflective Newton method and quasi‐Newton method; an unconstrained direct search method: the Nelder–Mead Simplex method; a constrained genetic algorithm; and a constrained Pattern Search (PS) method. It is shown that the Newton type methods are less accurate compared with the direct search methods studied. The PS method was found to be the most accurate. Furthermore, the PS method was found to be about twice as fast as the Simplex method and 10 times faster than a Genetic Algorithm for the same computing hardware and the same input data.