The pressure vessel surveillance program is mainly based on the transition temperature change assessment, a change which is induced by neutron irradiation. Uncertainties in Charpy test measurements are well known; however, we are less familiar with uncertainties due to general procedures governing experiments, which can be significant and therefore must be taken into account. In fact, procedures specify neither the number of specimens needed to obtain a transition curve, nor the choice of test temperatures, nor the fitting method for the transition curve.
A study has been conducted to determine the influence of the experimental procedure on the accuracy of transition temperature determination, and the initial results are presented in this paper. Two EDF laboratories performed Charpy tests on the surveillance program reference metal, using 8, 16, 24, 32 and 64 specimens to evaluate how the number of specimens affects the transition temperature. The influence of the scatter of mechanical properties has also been studied at two levels of irradiation. We have evaluated the effect of different sampling strategies and investigated a new fitting method, which is based on a simultaneous fitting of all curves with common constraints on parameters.