Modeling (α,n) neutron source terms in the ALEPH depletion code

The modeling of ( α ,n) neutron source terms in the ALEPH depletion code is assessed for two contrasting systems: spent nuclear fuel, a heavy actinide-oxide matrix, and an americium-beryllium (AmBe) source, a light low-Z matrix. In each, the source intensity and alpha spectrum are computed both by the analytic thick target yield and by direct alpha transport. For the spent fuel the yield from MCNP, ALEPH’s transport solver, falls 16% below the thick target yield, whereas for AmBe the two agree to within about 1%. The discrepancy is traced to MCNP’s internal stopping power table, which deviates from the validated SRIM and NIST ASTAR data in the heavy matrix but not the light one. The thick target yield is therefore at present the only reliable route to ( α ,n) source terms within ALEPH: MCNP alpha transport should not be used in heavy matrices, and an explicit transport estimate should come from a code with validated stopping powers, such as PHITS. A separate, matrix-independent energy-binning artifact in the MCNP alpha fluence spectrum is also identified; it redistributes fluence between neighboring bins without biasing integral quantities, and matters only when a differential spectrum is tallied. The neutron emission spectra from JENDL-5, JEFF-4.0 and TENDL-2025 are assessed through their effect on the ambient dose equivalent rate H*(10), computed for the AmBe source in a PHITS model of a shielded calibration laboratory and benchmarked against the reference ISO 8529-1 spectrum. The comparison supports JENDL-5 as the ALEPH default.