Thermodynamic Assessment of Hot Corrosion Mechanisms of Superalloys Hastelloy N and Haynes 242 in Eutectic Mixture of Molten Salts KF and ZrF4

Abstract

This computational work models the behavior of Haynes 242 and Hastelloy N in the binary molten salt eutectic mixture KF-ZrF 4 . Limiting the principal alloying elements to four-nickel, chromium, molybdenum, and iron-allows reasonable conclusions to be made about the nature of the hot corrosion resistance of these materials. It suggests that alloy composition optimization work could be conducted to further enhance alloy hot corrosion resistance.The onset of hot corrosion for both alloys is associated with chromium leaching and the formation of CrF 2 in relatively mild oxidizing conditions. However, the onset of hot corrosion for Alloy N requires harsher conditions than for Haynes 242. Thermodynamic data for these systems need to be generated in future research efforts in order to get a clearer picture of the mechanisms of hot corrosion onset. Firstprinciples atomistic simulations can illuminate those cases where experimental thermodynamic data is absent.Another important direction of research is coupling, in real time, the hydrodynamics of molten salt flow inside the reactor with a thermodynamic assessment of corrosion made under the assumption of local equilibrium on a sufficiently fine spatio-temporal grid. This would pave the way to understanding how corrosion develops in real space and time in heat exchanger loop components made of superalloys and steels. A possible application of this new technology would be the analysis of pyrolytic processes, when molten salts are used for catalytic purposes (e.g., biomass reactors, obtaining quality fuel out of heavy fractions of oil refining, etc.).Continuing this computational research effort will allow nuclear engineers to approach the problem of down selecting materials for NGNP AHTRs with greater confidence, less effort spent on expensive experimental work, and in shorter periods of time.

Hot Corrosion Monitoring in Boilers Using a Nonlinear Estimator and Electrochemical Noise-Based Corrosion Sensors

“…99 Most first-principles models require information about variables like stress and strain factors and the concentration of soluble species at the metal−gas interface, which cannot be obtained in this work. 42,43,46,48 Mohanty and Shores used a parabolic rate law, but no model was proposed for the parabolic rate constant, and values for it were calculated experimentally. Our modeling approach seeks to develop a phenomenological model that can capture the unique "bell"shaped response corrosion rate with respect to temperature.…”

“…Thermodynamics-based models were also developed to study the weight gain due to corrosion scale formation in high-temperature corrosion. 48,49 These thermodynamics-based models mainly provide information about equilibrium composition and are not suitable where the transient evolution of corrosion scale is of interest, such as for process monitoring. Kumari et al developed an artificial neural network (ANN) model for the deposit-induced fireside corrosion on superheater tubes by using the industrial data of SO x concentration, fly ash composition, etc., as input variables.…”