Date of Award


Document Type

Open Access Thesis


Civil and Environmental Engineering


College of Engineering and Computing

First Advisor

Paul Ziehl


Storage of spent nuclear fuel has become problematic in past decades due to delayed completion of long-term repositories for various reasons. Temporary storage containers called Dry Cask Storage Systems (DCSS) made of stainless-steel and surrounded by reinforced concrete have been in use but are exceeding their designed usage periods. Defects in canisters can be exacerbated in climates susceptible to high humidity and salinity levels. As inspection and relicensing of DCSS increases, more efficient monitoring techniques could save nuclear facilities valuable time and resources. Crack detection of the canister walls or welds in real time may be possible utilizing acoustic emission (AE) sensors. The capability to detect a partial-wall crack with an ASME-accepted, nondestructive testing method could prove useful for future DCSS inspection purposes.

The focus of this work, which is a part of a larger study for determining the feasibility of AE monitoring for DCSS, utilizes a small-scale type 304H stainless-steel specimen to monitor stress corrosion cracking. The plate specimen was statically loaded, creating a bending behavior which produced tensile stress at one face. A corrosive solution of potassium tetrathionate was then introduced to an electrical discharge machined starter notch and monitored for AE activity until crack initiation and propagation commenced. The raw wideband AE data was filtered and processed manually using proprietary software. Resonant AE data was filtered by hit count, duration and signal strength. Source location of the emissions was performed by triangulation of event arrivals at the AE sensors. Cracking was also observed via digital microscopic evidence throughout the testing period.

The waveform patterns observed after testing resembled AE related to crack initiation and propagation in various materials, where the amplitude of the waveform increases suddenly near the beginning of the signal and then dissipates with time as the wave travels through the medium.

The frequency spectrum of these waveforms was determined using the Fast Fourier Transform. The peak magnitude of the observed signal frequencies fell within range of 100-300 kHz for the small-scale specimens. The mean frequency was 224 kHz with a standard deviation of 52.1 kHz.

Future work should include investigation of a larger-scale specimen more representative of an actual DCSS canister, while using a similar procedure. This will more accurately replicate field conditions and determine the possibility of detecting an emission traveling the length of a canister (approximately 16 feet). Furthermore, this test would help determine the feasibility of real-time AE monitoring.


© 2018, Joseph W. Hill