Straight-sided double cantilever beam (DCB) fracture specimens have be
en used previously for long-duration environmentally assisted crack gr
owth (EACG) testing. However, the applied strain-energy release rate G
decreases with crack extension. In this study the width taper require
d to achieve constant-G during crack growth in a double cantilever bea
m (DCB) fracture specimen is determined for the case of general loadin
g. The width taper b(a) is a function of the initial specimen complian
ce, C0, the crack length, a, and the stiffness of the loading frame, k
. The compliance of DCB specimens tapered for testing under three load
ing conditions were verified experimentally. These were constant-load
(k --> 0). constant-strain (k --> infinity), and critical spring loadi
ng (kC0 --> 0.5). Specimens designed for critical spring loading are s
uggested for use in long-term (EACG) experiments. Limitations on the p
roposed test method related to nonlinear material behavior (e.g., crac
k bridging) are discussed.