Investigation of Surface Peening Technologies for Addressing Stress Corrosion Cracking in Used Fuel Dry-Cask Storage Stainless Steel Canisters

Abstract

Nuclear power serves as a vital and sustainable global energy source. The safe, secure, and enduring storage of spent nuclear fuel is an important focus of the nuclear industry. In the United States, as in many nations world-wide, the absence of established permanent geological repositories has necessitated interim storage of used fuel in dry-cask storage systems (DCSS) for extended periods of time. Central to the DCSS architecture is the austenitic stainless steel (316L or 304L grade) canister, which serves as a primary barrier between the stored fuel and the external environment. The fusion-welded regions of the canister, particularly in the heat affected zone (HAZ) are prone to corrosion and eventually chloride-induced stress corrosion cracking (CISCC) in the long-term. No siting of CISCC has been reported, nevertheless concerns exist given the potentially long-term duration of interim storage. This research focuses on the investigation of three potential of surface peening technologies to address the challenge of CISCC in DCSS canisters. These include pulsed water jet peening (PWJP), ultrasonic nanocrystalline surface modification (UNSM), and laser shock peening (LSP). The LSP technology was investigated both with and without ablative coating. For each technology, the most significant process parameter was varied, namely water jet traverse speed for PWJP, power density of LSP, and applied load for UNSM. The substrates for this research were sensitized 0.5” thick 304H stainless steel samples. Sensitization was conducted at 650 °C for 48 hours to simulate the microstructure of the HAZ of the weld, and the higher carbon ‘H’ grade was selected to accentuate the effects of sensitization. Testing and characterization was performed using SEM, XRD for both phase identification and use of sin2Ψ method for residual stress measurements, XPS, microhardness testing, profilometry, and electrochemical anodic polarization testing in 3.5wt% NaCl solution. For PWJP water jet traverse speeds of 300 mm/s, 400 mm/s, and 1000 mm/s were investigated. The resulting surface microstructure showed discernible deformation zones, and slip bands extending several hundred microns below the surface. More plastic deformation as well as greater erosion was observed at lower traverse speeds due to greater residence time of the water jet on the samples. Strain-induced martensite was observed at the intersection of slip bands. The near-surface hardness was significantly enhanced due to the creation of high dislocation density and possibly due to martensite formation. The process introduced high compressive stress in the near-surface regions (e.g., -579.41 MPa at 400mm/s), but at lower speeds surface erosion led to removal of some regions of high compressive stresses. UNSM, performed at applied loads of 4 kg, 8 kg, and 12 kg, imparted severe plastic deformation in the near-surface regions of the substrate, where it resulted in grain refinement and sub-micron scale microstructure. Complete transformation of austenite phase to strain-induced martensite was observed in the near-surface regions of the substrate. UNSM-treated samples also exhibited very high compressive residual stresses (e.g., -864.05 MPa at 4 kg applied load), in the near-surface regions with the magnitude increasing with increasing load. For LSP processing without ablative coating, an oxide layer formed on the samples’ surface due to the heating from the laser, as established by XPS depth profiling. The ablative coating results in pressure amplification, and therefore the power density was adjusted to impart pressures similar to those without the ablative coating. Slip band formation was observed for samples without the ablative coating, and the presence of the ablative coating promoted the formation of a sub-micron grain structure in the near-surface region. The extent of the austenite-to-martensite transformation was higher formation for samples treated with the ablative coating because lower temperatures were maintained in this case. Samples treated without the ablative coating exhibited a tensile surface residual stress state due to thermal expansion from laser heating, transitioning to a compressive state below the surface. Samples with the ablative coating exhibited high compressive residual stresses, up to -600 MPa, but at the highest power density of 10 GW/cm2 relaxation of the compressive residual stress state was observed. Electrochemical anodic polarization testing in 3.5%NaCl solution demonstrated enhanced corrosion performance for all three peening technologies based on increases in open circuit potential and pitting potential to more positive values and a lowering of corrosion current density. Microstructural and compositional homogenization, fine grain structures, and the presence of compressive residual stresses all contributed to the superior corrosion performance by way of delaying pit nucleation and promoting and stabilizing the passive oxide film. The presence of martensitic regions in an austenitic matrix was found to be detrimental to corrosion resistance, however when present as a sole phase (as in the case of UNSM) can be beneficial to corrosion resistance. Post-corrosion pit depth and pit number density measurements indicated the UNSM treatments to result in lower pit depths. The pit number density decreased for PWJP and UNSM-treated samples, with increasing traverse speed and applied load, respectively. The pit number density for LSP-treated samples decreased with increasing imparted pressure up to a threshold power density of 5 GW/cm2. Overall, the research demonstrates the feasibility of using various peening technologies for the DCSS application from the standpoint of microstructure, compressive stresses, and superior corrosion resistance. However, further optimization studies specific to DCSS application as well as an understanding of CISCC initiation and growth are required for implementation of these peening technologies.