4.1 A pressure vessel surveillance program requires a methodology for relating radiation-induced changes in materials exposed in accelerated surveillance locations to the condition of the pressure vessel (see Practice E853). An important consideration is that the irradiation exposures be expressed in a unit that is physically related to the damage mechanisms.

4.2 A major source of neutron radiation damage in metals is the displacement of atoms from their normal lattice sites. Hence, an appropriate damage exposure index is the number of times, on the average, that an atom has been displaced during an irradiation. This can be expressed as the total number of displaced atoms per unit volume, per unit mass, or per atom of the material. Displacements per atom is the most common way of expressing this quantity. The number of dpa associated with a particular irradiation depends on the amount of energy deposited in the material by the neutrons, and hence, depends on the neutron spectrum. (For a more extended discussion, see Practice E521.)

4.3 No simple correspondence exists in general between dpa and a particular change in a material property. A reasonable starting point, however, for relative correlations of property changes produced in different neutron spectra is the dpa value associated with each environment. That is, the dpa values themselves provide a spectrum-sensitive index that may be a useful correlation parameter, or some function of the dpa values may affect correlation.

4.4 Since dpa is a construct that depends on a model of the neutron interaction processes in the material lattice, as well as the cross section (probability) for each of these processes, the value of dpa would be different if improved models or cross sections are used. The calculated dpa cross section for ferritic iron, as given in this practice, is determined by the procedure given in 6.3. The currently recommended dpa cross sections in this practice (Table 1) were generated using the iron ENDF/B-VI iron cross section (1).3 A recent calculation using ENDF/B-VII produced identical results.(2,3) Although the ENDF/B-VI based dpa cross section differs from the previously recommended ENDF/B-IV dpa cross section (1) by about 60 % in the energy region around 10 keV, by about 10 % for energies between 100 keV and 2 MeV, and by a factor of 4 near 1 keV due to the opening of reaction channels in the cross section, the integral iron dpa values are much less sensitive to the change in cross sections. The update from ENDF/B-IV to ENDF/B-VI dpa rates when applied to the H. B. Robinson-2 pressurized water reactor resulted in “up to ∼4 % higher dpa rates in the region close to the pressure vessel outer surface” and in “slightly lower dpa rates ... close to the pressure vessel inner wall” (4, 5). Table 2 presents a comparison of a previous edition (Practice E693-94) and currently recommended dpa estimates for several neutron spectra.