Modified Hartman-Schijve fitting of mode I delamination fatigue data and the resulting variation in threshold values Gthr

With a view toward applications in structural composite design, mode I (tensile opening) fatigue delamination test data obtained for carbon fiber-reinforced polymer-matrix (CFRP) composites have recently been (re-)analyzed with different fitting approaches for the determination of threshold values G(thr). One approach that looks promising as an alternative to the conventional Paris-law type presentation of the data (a double logarithmic plot of delamination length increment per cycle da/dN versus G(imax) or Delta G(1)) is a modified Hartman-Schijve fit (also a double logarithmic plot of da/dN, but versus a square-root function of G) requiring four fitting parameters (labelled A, beta, D and G(thr)). The present paper details a procedure for selecting the fitting parameters A, beta, and D in order to determine G(thr) in a consistent way from the experimental data. For some design approaches, it is important to also have an estimate of the possible scatter in experimentally derived values of G(thr). The procedure chosen by the authors yields G(thr) values and a measure of the scatter due to experimental variation of the other fitting parameters. Selected mode I delamination fatigue data for IM7/8552 CFRP composites from literature are used to compare G(thr) values and their scatter from Paris-law type fitting with those from the modified Hartman-Schijve approach. Copyright (C) 2016 The Authors. Published by Elsevier B.V.