HTS rf-SQUID-Planargradiometer mit langer Basislänge für die Inspektion von Flugzeugfelgen
Maus, M. (Corresponding author)
Publikationen vor 2000; PRE-2000; Retrocat
Jülich Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag 2000
III, 136 p.
Report
Book
Addenda
Berichte des Forschungszentrums Jülich 3779
OpenAccess
OpenAccess
Please use the identifier: http://hdl.handle.net/2128/23983 in citations.


An important task in the field of nondestructive evaluation (NDE) of aircraft components is the reliable detection of small eracks hidden very deep in the material. This will become even more important in the future, because the thickness of aircraft parts is going to increase furthen Conventional eddy current testing systems reach their Limit here, whereas SQUIDs ($\underline{S}$uperconducting $\underline{QU}$antum $\underline{I}$nterference $\underline{D}$evices) show an excellent sensitivity for the lowexcitation frequencies needed. As it is required to suppress large magnetic disturbances without magnetic shielding, a gradiometer SQUID is utilized for these measurements. By caleulations, it is proved in this work that an optimum Signal-to-noise ratio is reached when the baseline of the SQUID gradiometer is enlarged from the so far used 3.6 mm to about 10 mm. However, an enhancement of the baseline leads to a significant increase of the SQUID inductance and hence of the white noise. Consequently, the SQUID sensor cannot be operated reliably in the flux locked loop any more. lt is shown that the inductance may bereduced significantly by plaeing a superconducting cover strip on top of the gradiometer slit.Thus, the gradiometer slit cannot be penetrated by magnetic flux any more. Calculations ofthe geometric and kinetic inductance confi rm that a distinct reduction of the inductance can bereached this way. Furthermore, the calculations show that it is possible to raise the baseline to9 mm without an increase of the inductance, compared to an uncovered gradiometer with3.6 mm baseline. The experimental realization of the gradiometer is very complicated because it is necessary touse three epitaxial layers for the fabrication and multilayer technology for high temperaturesuperconductors is still under development. lt is especially important to avoid electrical shortsbetween the superconducting layers. In order to achieve that, gradiometer structures withvarying steepness at the edges as well as different widths of the cover strip are examined. Inorder to exclude electrical shorts, caused by partieles on top of the surface of the lowersuperconducting layer, sputtered YBaCuO thin films are used. They typically form verysmooth surfaces almost free of particles. The experiment shows that by applying the superconducting cover strip, the white noise of thegradiometer is reduced from 144 $\mu\Phi_{0}/\surd$Hz to 73 $\mu\Phi_{0}/\surd$Hz, nearly by a factor of 2.Simultaneously, the fieldgradient-to-flwo-coefficient rises from 15.1 nT/($\Phi_{0}$ cm) to 33.4 nT/($\Sigma_{0}$ cm). This proves that the Iower white noise is really due to a smaller pick-up area for the magnetic flw(, and thus the shielding strip works. The common-mode-rejection-ratio(CMRR) of the gradiometer shows a decrease to 10, compared to an undercovered sample havinga typical value of about 1000. On one hand, this can be partially attributed to an observedlateral shift of the cover strip caused by inevitable alignment errors of the photolithographicprocess. Consequently, both pick-up areas for the magnetie flux slightly vary in size. On theother hand, scanning electron microscope (SEM) photographs show that the crystal quality ofthe upper YBaCuO thin film is not distributed homogeneously. This leads to a differentshielding effect an both sides of the cover strip.The work shows that planar rf SQUID gradiometers with lang baseline can be realized bycovering the gradiometer slit with a superconducting film. However, some technologicaldifficulties have to be solved before the gradiometers can go into production. Futuredevelopment should aim especially at the improvement of the crystal quality of the upperYBaCuO thin film. Optimization of the multilayer technology for high temperaturesuperconductors is a key issue in future SQUID development.