Plastic constraint considerations at the prediction of wide plate failure behaviour from small scale testing, like c.o.d. determinations in s.e.n.b.-specimens

Abstract

The stress intensity factor, crack opening displacement, J-integral, etc. will depend, besides on load, reduced (physical) cracklength and dimensions also on the plastic constraint situation near the crack tip. Assumedly their critical values are also dependent on this, which can imply an appreciable quantitative consequence for uniaxial gross stresses larger than ( 2 3 ) Y, as elaborated. This might interfere when trying to predict large sale behaviour of notched plates from small scale behaviour of specimens with the same thickness, as this presumbly does not guarantee equal plastic constraint. An average plastic constraint factor (p.c.f.) estimate for contained plastic flow can be indicated from simultaneous readings of s.i.f. or c.o.d. and gross stress during loading and its decrease followed. This allows to predict the failure conditions at the end of the experiment, provided the dependence of critical c.o.d. on this p.c.f. was previously determined. For arbitrary notched specimens a priori predictions of failure from the known critical c.o.d./p.c.f.-curve can be envisaged, using analytical and numerical evaluations of p.c.f. development at loading. The transition temperature concept is discussed with respect to the p.c.f. The failure prognosis for wide plate testing from small scale testing is reconsidered for deformation controlled loading: for contained plastic flow an alternative for the design-curve is suggested, using the deformation of the gauge length of the specimen defined as the distance between the localisations of load application. Finally the allowable deformation of plates, indicated for the required load in absence of a defect, is analysed for plates with a defect present.