Use of combined linear and nonlinear ultrasound to examine microstructural and microchemical variations in highly irradiated 304 stainless steel

Abstract This research combines linear and nonlinear ultrasound to examine the microstructural and microchemical changes in five large, highly irradiated cold-worked ANSI 304 stainless steel coin specimens cut from two hexagonal cross section blocks with radiation damage levels ranging from ~0.4 to ~33 dpa following irradiation in the EBR-II fast reactor. Both linear (velocity) and nonlinear ultrasonic (β) measurements were conducted in a hot cell using an automated fixture device specially designed to hold the sensor for repeatable measurements in an environment where direct human access is limited, demonstrating their potential to more fully investigate the complex defect ensemble introduced by irradiation. The measurements were performed at the Westinghouse Churchill site hot cell facility. The linear measurements agree very well with previous measurements employing a different sensor application technique. The ultrasonic nonlinearity parameters measured after irradiation increased by more than 100% from the unirradiated state and show a level of random variation below 4.7%. These ultrasonic nonlinearity parameters also show a spatial dependence on measurement location across each of the specimens in agreement with the spatial inhomogeneity of microstructure and microchemical distribution previously observed in adjacent, nominally identical specimens using transmission electron microscopy (TEM) and atom probe tomography (APT). It is proposed that the inhomogeneous distribution of microstructural and microchemical features including Frank loops, and both intragranular and grain boundary precipitates, as well as their different nonlinearity generation efficiencies are responsible for the spatial dependence of the measured ultrasonic nonlinearity parameters. The major and dominant component of the microstructure of these specimens at 20–30 dpa is vacancy agglomerations called voids, previously observed by microscopy and measurable using either linear ultrasonic velocity or linear attenuation, but which are not readily visible using nonlinear ultrasound. Components such as Frank loops, which are one of the other major microstructural components, do not contribute significantly to changing the linear ultrasonic velocity. Using this combination of linear and nonlinear measurements allows for the examination of microstructural/microchemical components whose linear ultrasonic interactions are overshadowed by voids, especially Frank loops and various radiation-induced precipitates. The changes in ultrasonic nonlinearity parameters in these specimens are thought to arise from two major sources. The smaller contribution is due to carbide pinning of dislocations in the grain boundary walls and the few line dislocations present in the grain interior. The largest contribution appears to come from segregation of Ni and Si to the perimeter of the very high density of Frank loops, leading to the formation of Ni3Si precipitates which are thought to be very efficient at pinning the Frank loops. These results on two hex blocks provide an ideal, though not completely realistic, test case to demonstrate the sensitivity of nonlinear ultrasound to two different microstructure sources.