In surveillance programs for some older nuclear power plants, the orientation of specimens is longitudinal to the main working direction. According to presently valid codes and standards, the specimen orientation should be transverse. The material toughness is lower in the transverse direction. However, according to former test results, there was common acceptance that the shift as a result of irradiation is independent of orientation. The results of more recent research on trepan material taken from a decommissioned power plant seemed to give reason for doubt in this respect.

Siemens/KWU has simultaneously irradiated Charpy-V specimens made of steel 22 NiMoCr 3 7 (similar to ASTM A 508 cl. 2) with both orientations. Furthermore, transverse specimens were made out of the broken specimen halves from irradiation experiments with only longitudinal specimens using electron beam welding techniques.

The available results of this research and an ample number of relevant data from literature were evaluated. From this it is concluded that independent of a wide variation of parameters the 41-J transition temperature of transverse specimens is on average about 11°C higher than that of longitudinal specimens and the irradiation-induced shift ΔT₄₁ is independent of specimen orientation. The upper shelf energy (USE) shows a wider scatter. As a lower bound, the proposal of the U.S. Nuclear Regulatory Commission (NRC) standard review plan to substitute 65% of longitudinal USE for unavailable transverse values is confirmed by 90% of all data for unirradiated and irradiated material. The deviations, especially of the Gundremmingen trepan results, obviously depend on fabrication inhomogeneity rather than on irradiation effects.