The effect of grain size on the defect detectability in copper components in ultrasonic testing

In metals ultrasonic waves attenuate in different ways. The attenuation can be divided in different categories: geometrical attenuation due to beam spreading due to the finite aperture of the transducer, scattering, and absorption. The scattering is caused mainly by grain boundary scattering and thus the grain size affects strongly the attenuation. Both the average grain size and the distribution of grain size have to be taken into account. Additionally the surface finish of the component affects amplitude variations in ultrasonic inspections. Not only the RMS-value of the surface roughness does play a role but also the form of roughness. This is not often recognized in inspections. Also the stress state of the component will affect the attenuation of ultrasonic waves mainly via the absorptive part (i.e. internal friction) of the material. In the copper components examined here, the grain size can vary locally between 50 µm to 2000 µm. Due to this large variation in the grain size distribution, the ultrasonic testing demands greater efforts than in normal polycrystalline materials, where the grain size is small compared to the ultrasonic wavelength. The used frequencies for copper in ultrasonic inspections vary between 1 MHz to 5 MHz and the corresponding wavelengths between 4.70 mm and 0.94 mm for longitudinal waves and between 2.26 mm and 0.45 mm for shear waves. This means that at least part of the grain sizes in the copper components examined are comparable to the ultrasonic wavelengths. Rarely higher frequencies than 5 MHz are used in thick copper component inspections because of their high attenuation. The attenuation in the copper components follows two regimes: Rayleigh scattering and resonance scattering, which both have different functional dependencies. The thickness calibration of the ultrasonic equipment is carried out with reference specimens. The time-of-flight is only little affected by the attenuation of the reference specimen, and hence they may serve well for the calibration of the time scale of the UT equipment, but the sensitivity calibration for the inspection is as important as the time-scale calibration. To get comparable inspection quality, the reference specimen should have similar attenuation than the component under test. Eight standard reference specimen made from copper were measured and the attenuation in these specimen were compared to each other. Additional nine reference specimens with variable grain size were also manufactured to study the effect of grain size on the attenuation. The grain size variation in these specimens was between 60 µm to 800 µm. The initial structure of the grain size of oxygen free copper after hot rolling was modified with variable hot pressing reduction from 20% to 61%. The evaluation of the average grain size is according to ASTM standards. The manufacturing goal was to produce a grain-size distribution as homogenous as possible. Studies of the attenuation of these samples are reported in this paper. The measured attenuation is compared to theories known in the literature. The attenuation is evaluated to optimize the ultrasonic inspection of copper components including their electron-beam welds (EB-welds) using phased array techniques. Similar copper samples have been studied by Stepinski