During the last few years, the g-heating issue has gained in stature, mainly for thesafety of the 3rd-generation reactors in which a stainless steel reflector is inserted. Thepurpose of this work is the qualification of the needed tools for calculation of the g-heating inthe nuclear reactors.In a nuclear reactor, all the photons are directly or indirectly produced by the neutronmatterinteractions. Thus, the first phase of this work is a critical analysis of the photonproduction data in the standard nuclear data library. New evaluations have been proposed tothe next version of the JEFF library after that some omissions have been found. They havepartly been accepted for JEFF-3.2.Two particle-transport codes are currently developed in the CEA: the deterministiccode APOLLO2 and the Monte Carlo code TRIPOLI4. The second part of this work is thequalification of both these codes by interpreting an integral experiment called PERLE. Theexperimental set-up is made by a LWR pin assembly surrounded by a stainless steelreflector in which the g heating is measured by Thermo-luminescent Detector (TLD). Acalculation scheme has been proposed for both APOLLO2 and TRIPOLI4 in order tocalculate the TLD’s responses.Comparisons between calculations and measurements have shown that TRIPOLI4gives a satisfactory estimation of the g heating in the reflector. These discrepancies arewithin the experimental 1s uncertainty. Before the qualification, APOLLO2 has beenpreviously validated against TRIPOLI4 reference calculation. This validation gives anestimation of the bias due to the deterministic approximations of the transport equationresolution. The qualification has shown that the discrepancies between APOLLO2predictions and TLD’s measurements are in the same range as experimental uncertainties.
