PNNL

Abstract

Ultrasound sensors and inspection systems are frequently used to generate acoustic waves in metal structures capable of detecting and characterizing cracks, pits, erosion, inclusions, weld anomalies, and other material and structural features. One significant problem with piezoelectric transducers is the difficulty to achieve good coupling between the transducer and the surface being examined. This is particularly true in harsh conditions with high temperatures, cyclical hot and cold temperatures, highly radioactive fields found near nuclear reactors or spent nuclear fuel, caustic or corrosive fluids, and other extreme environmental conditions, as well as in long-term monitoring applications where repair or replacement of the sensor is difficult or expensive. Typically, coupling between the surface and the transducer is achieved with water, gel, grease, viscous shear coupling material, or pressure, which might not be possible or appropriate for long-term applications in which the impedance-matching materials wear away, evaporate, or simply stop functioning due to changes in surface conditions. Fluid couplings can evaporate or drain away from the transducer-substrate interface; glue-based couplings may foul or fail and are notoriously unreliable at high temperatures and in radioactive environments.This work explores the behavior of a magnetostrictive cold-spray patch that is metallurgically bonded to a stainless steel inspection target surface, and compares it to the performance of a standard adhesively-bonded ferrous-cobalt magnetostrictive strip solution. Cold-spray is a coating process where 10–100 micron diameter powdered metal is accelerated to Mach 2 to Mach 3 (2–3 ? speed of sound) and impacted on the surface to be coated. Each powder particle forms a kinetic bond with the substrate or other coating particles to produce a metallurgically bonded layer. If the powder is nickel or cobalt with high magnetostrictive coefficients, this surface can serve as the base of a magnetostrictive sensor suitable for crack or pitting-damage inspection and monitoring that is not subject to temporal or environmental degradation. Two rounds of testing were performed in this study. The first round sample set consisted of a commercially pure nickel (CPNi) cold-spray patch-based sensor on a 0.6 x 1.2m, 64mm and 127mm (2 ? 4 ft., ¼ and ½ in.) thick plate that was contrasted with a more conventional FeCo adhesive strip on the same plates. Tests assessed the feasibility and relative efficacy of the two magnetostrictive substrates. Influences of magnetic bias and cold-spray patch thickness were also explored in this study. The initial comparison results showed a strong response in the SH0 mode however there was an undesirable echo response from the edge of the nominally 1 and 2mm coatings. The coating thickness was reduced to ~ 0.5mm then 0.25mm resulting in the patch echo being substantially eliminated without reducing the edge reflection amplitude. The magnetostrictive coefficient of CPNi is reported as 25–60 ppm while cobalt has a magnetostrictive coefficient of 40–120. Moreover, the FeCo strip is prepared with a magnetic bias treatment. As might be anticipated, the FeCo strip showed ~7–19 dB stronger edge-wall reflection than the CPNi patch. The reduced amplitude response of the CPNi patches relative to the FeCo strip was well within the ability of the EMAT instrument to detect the plate edge response. A second round of tests included lower pressure N2 and air cold spray samples, different powders, and different vendors plus an assessment of Lamb wave as well as SH0 wave modes.Generally both the SH0 mode and Lamb mode signals from the initial high pressure N2 cold spray samples and the subsequent range of coating samples were readily detectable. The different cold spray processes had a significant influence on the edge amplitude responses.