Premium
A neural network approach for predicting steel properties characterizing cyclic Ramberg–Osgood equation
Author(s) -
GHAJAR R.,
NASERIFAR N.,
SADATI H.,
ALIZADEH K. J.
Publication year - 2011
Publication title -
fatigue and fracture of engineering materials and structures
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.887
H-Index - 84
eISSN - 1460-2695
pISSN - 8756-758X
DOI - 10.1111/j.1460-2695.2010.01545.x
Subject(s) - artificial neural network , exponent , strain hardening exponent , monotonic function , hardening (computing) , ultimate tensile strength , test data , structural engineering , materials science , experimental data , approximation error , correlation coefficient , mathematics , computer science , engineering , mathematical analysis , composite material , artificial intelligence , statistics , philosophy , linguistics , layer (electronics) , programming language
This paper attempts to demonstrate the applicability of artificial neural networks to the estimation of steel properties, cyclic strain‐hardening exponent and cyclic strength coefficient, characterizing cyclic Ramberg–Osgood equation on the basis of monotonic tensile test properties. For this purpose, steel tensile data were extracted from the literature and two separate neural networks were constructed. One set of data was used for training the two networks and the remaining for testing purposes. Regression analysis and mean relative error calculation were used to check the accuracy of the system in the training and testing phases. Comparison of the results obtained from the neural networks and the values obtained from direct fitting of experimental data, indicated the reasonable prediction of cyclic strain‐hardening exponent and cyclic strength coefficient, which are often used to characterize the cyclic deformation curve by a Ramberg–Osgood type equation.