Alloy compatibility with LiF--BeF₂ salts containing ThF₄ and UF₄

Abstract

Various compatibility tests between LiF- BeF/sub 2/ fused salts and Hastelloy N and certain steels were conducted in thermal convection loops. Temperature gradient mass transfer, as noted by weight losses in the hot leg and weight gains in the cold leg, was evidenced in all tests. The weight changes of corrosion specimens increased with increasing temperature and time. From an operational standpoint, some difficulty was encountered in melting LiF-- BeF/sub 2/- ThF/sub 4/ mixtures without mechanically failing the loop. The difficulty stemmed from the high melting point of the salts and the high melting point of centain phases in the salts. All the salts tested were compatible with Hastelloy N. Bismuth in contact with a fuel salt had no effect on mass transfer in a Hastelloy N loop. It was established that electrochemical methods to determine the oxidation potential of molten fluoride salts could be used to predict the corrosion behavior of a thermal convection loop. The values obtained by the electrochemical methods correlated well with specimen weight change data. It was shown that tellurium (as plated on Hastelloy N specimens) does mass transfer in a molten fluoride salt system and that an equilibrium between tellurium in the salt and tellurium on the alloy can be established. A type 304L stainless exposed to a fuel salt for 9.5 years in a type 304L stainless steel loop showed a maximum uniform corrosion rate of 0.86 mil/year. Voids extended 10 mils into the Matrix. Chromium depletion was found. The corrosion resistance of maraging steel (12% Ni- 5% Cr--3% Mo-bal Fe) at 662 deg C was better than type 304L stainless but worse than Hastelloy N under equivalent conditions. The unlform corrosion rate was 0.55 mil/ year. Subsurface voids were seen in the microstructure of the specimens after 5700 h, and microprobe analysis disclosed depletion of chromium and iron. Tyne 316 stainless steel exposed a a fuel salt in a type 316 stainless steel loop showed a maximum uniform corrosion rate of 1 mil/year for 4298 h. Mass transfer of chromium and iron did occur in the system. For selected nickel- and iron-base alloys a direct correlation was found between corrosion resistance in a molten fluoride salt and the chromium and iron content of an alloy. The more chromium and iron in the alloy, the less the corrosion resistance. (auth)