No primary source located confirms atmospheric air/oxygen-blowing entrainment as steel's cobalt-60 contamination pathway; the primary literature instead documents furnace-gauging and scrap-feedstock routes

materialsnuclearcontaminationradiation-detectionlow-background-physicssteelmakingcobalt-60verificationargument-from-silencenegative-citation

question-verify-low-background-steel-fallout-contamination-mechanism asks whether any primary source confirms the mechanism the vault's existing claim carries from Wikipedia: that Bessemer/basic-oxygen air- or oxygen-blowing draws atmospheric fallout radionuclides directly into the melt. Two Tier-1 physics primaries were located and read directly this session, and both discuss the origin of anthropogenic cobalt-60 in commercial iron and steel in some technical detail — and neither names blown air or oxygen entraining fallout particulate as the pathway. Heusser (1995) attributes it to furnace-lining wear-monitoring gauge sources; Zhang et al. (2016) attributes it to recycled scrap feedstock.

This does not disprove an atmospheric-entrainment contribution — trace radionuclide deposition from ambient air onto any large open melt is physically plausible, and neither paper explicitly rules it out — but it means the specific, popularly repeated mechanism this vault's existing claim-note carries is not the mechanism the primary literature located this session actually documents when explaining why modern steel is contaminated.

Grading this as an argument from silence (per moc-argument-from-silence's McGrew framework): the record term is reasonably strong — both papers discuss contamination origin at some technical length, the place they would plausibly mention an atmospheric pathway if it were the operative one. The notice term is weaker — neither paper is framed as a survey of all contamination pathways, so their silence on one specific pathway is less decisive than a defining text's silence would be. The survive term is the weakest: only two papers were read this session; several further leads (Maneschg et al. 2008, Kohler et al. 2004, the 1983 Ciudad Juárez incident's primary record) remain unread. Graded soft-to-medium, comparable to claim-no-primary-source-documents-ach-abduction-similarity rather than to the exhaustive-reference-list instances in the MOC's cluster.