ENDF Cross Sections are not Uniquely Defined

Abstract

Most evaluated data that is coded into the ENDF format [1] does not uniquely define cross sections, because the evaluator defined total is not equal to the sum of evaluator defined partial cross sections, i.e., the total is not equal to elastic plus capture, etc. So we have always had the question: What is the correct total cross section? This is not a new problem; it has existed since the very beginning of ENDF over forty years ago. It is a problem that is periodically discussed and apparently handled, only to have it pop up again every ten years or so, as we have the next generation of ENDF format users who are not aware of the problem. See the Appendices for a summary of the differences that exist today for the ENDF/B-VII.0 (Appendix C), JEFF- 3.1(Appendix D), JENDL-3.3 (Appendix E), and CENDL-3.1 (Appendix F) data libraries. For use in our application we need consistent, unique data. To accomplish this for decades we [2, 3] have been ignoring the evaluator defined total, and re-defining it as equal to the sum of its evaluator defined parts. This has never been completely satisfactory to us, because we have been doing this without consulting evaluators, or obtaining their approval, so that the data we actually use in our applications may or may not be what the evaluators intended.