This work examines the variability in grain boundary segregation measuremen
ts in proton irradiated iron-chromium-nickel alloys made using Anger electr
on spectroscopy (AES) and scanning transmission electron microscopy with en
ergy dispersive X-ray spectroscopy (STEM-EDS). Variability occurs in segreg
ation measurements made on different boundaries in the same sample and made
on different samples irradiated under the same conditions. Variability occ
urs in each measurement technique, but is greater for AES. A portion of the
greater variability in the AES measurements occurs because only the concen
trations calculated from AES measurements are sensitive to changes in the s
hape of the energy intensity peak. A statistical analysis technique for tes
ting the consistency of experimental conditions demonstrated that the varia
bility is not attributable to uncertainty in irradiation temperature, dose,
or material condition. Finally, the analysis of grain boundary composition
distributions in Fe-20Cr-9Ni indicates that attempts to minimize environme
ntal cracking by controlling grain boundary composition need to focus on bo
th the average grain boundary compositions and the shape of the concentrati
on distributions. (C) 2000 Elsevier Science B.V. All rights reserved.