Novel Methods for Recording Stress-Strain Curves in Proton Irradiated Material.

Proton irradiation is often used as a proxy for neutron irradiation but the irradiated layer is typically <50 µm deep; this presents a problem when trying to obtain mechanical test data as a function of irradiation level. Two novel methodologies have been developed to record stress-strain curves for thin proton-irradiated surface layers of SA-508-4N ferritic steel. In the first case, in-situ loading experiments are carried out using a combination of X-ray diffraction and digital image correlation on the near surface region in order to measure stress and strain, thereby eliminating the influence of the non-irradiated volume. The second approach is to manufacture small-scale tensile specimens containing only the proton irradiated volume but approaching the smallest representative volume of the material. This is achieved by high-speed focused ion beam (FIB) milling though the application of a Xe+ Plasma-FIB (PFIB). It is demonstrated that both techniques are capable of recording the early stage of uniaxial flow behaviour of the irradiated material with sufficient accuracy providing a measure of irradiation-induced shift of yield strength, strain hardening and tensile strength.

Quantitative methods for the APT analysis of thermally aged RPV steels.

Atom Probe Tomography (APT) is extensively used for the analysis of RPV steels. However, many different analysis methods and cluster search parameters are used, making comparisons between different datasets difficult. Suitable d(max) and N(min) parameters for the maximum separation method are investigated. In a randomised distribution of solute there is a finite probability that a group of more than N(min) solute ions exists within the d(max) distance. The same is true for experimental datasets from samples which have been thermally aged or irradiated, however these background clusters are not the result of ageing, they are purely statistically random co-incidences. A method is presented for identifying such "background" statistical clusters in real APT data sets, based upon their size and composition, which allows for improved sensitivity to small clusters.