Characterization and quantification of magnetic remanence in unexploded ordnance

Under the Range Rule (1997), the Department of Defense defines unexploded ordnance (UXO) as a piece of ordnance that has been deployed, but has not exploded. This study investigated a subset of UXO including artillery shells and mortars (projectiles), but excluding landmines. UXO contaminates approximately 15 million acres of land in the United States alone. The geophysical tools most frequently used for detection are electromagnetic and magnetic methods. These methods, however, produce a “hit” for any metallic object in the ground, not just UXO. In a given survey, this results in a large number of anomalies, only a small subset of which are actually UXO. Current practice dictates that most anomalies are dug up and identified, and the UXO is exploded in place. The estimated cost to remediate existing UXO, in the United States, with current methods, is tens to hundreds of billions of dollars. Discrimination is the process in which UXO (hazardous) is differentiated from non-UXO (non-hazardous), prior to anomaly excavation. Methods to discriminate the UXO greatly reduce the cost of remediation. Remanent magnetization in UXO is one quantity that must be understood to improve discrimination methodologies in the future. Upon impact, a shock wave propagates through a projectile UXO, randomizing the directionality of the magnetic domains, and erasing a portion of the original remanence. Due to this shock demagnetization, the typical UXO item exhibits smaller remanence than other metallic debris. Thus, if the amount of remanence contributing to a magnetic anomaly can be measured, and is found to be comparatively small, that item is more likely to be UXO. For the seven stages in the history of a UXO projectile (unformed, unfired, fired, exploded, storage, transport, and unexploded/waste), I determine the origin of remanent magnetization. I quantify viscous remanent magnetization (VRM) for the first four stages, and find it to be less than 5% for up to 1000 years for all samples, and thus determine remanence to be stable in UXO within that time period. I show the ColeDavidson α-parameter to be a new discrimination parameter in the nonlinear forward modeling of the VRM data. I develop a procedure to recover the remanent dipole moment of 34 UXO items using data from the Mobile Remanence Interrogation Platform