Cailiao gongcheng (Sep 2019)

Estimation and analysis of probability of PVD film sensor crack detection

  • LIU Kai,
  • CUI Rong-hong,
  • HOU Bo,
  • HE Yu-ting,
  • NIU Huan

DOI
https://doi.org/10.11868/j.issn.1001-4381.2018.000781
Journal volume & issue
Vol. 47, no. 9
pp. 160 – 166

Abstract

Read online

In order to quantitatively characterize the ability of PVD(physical vapor deposition) film sensors to detect cracks in metallic structures, the process parameters applied to LY12-CZ aluminium alloys were first optimized using orthogonal tests, and coin shape, 1mm concentric dual ring shape, 0.5mm concentric tricyclic shape PVD film sensors were prepared on three groups of specimen with central hole using these parameters. Subsequently, the on-line monitoring experiment of fatigue cracks was carried out under laboratory conditions, the potential output signals of the PVD film sensors and microscope observations were compared and analysed. Finally, the probability of detection curves of all kinds of PVD thin-film sensors and PVD film sensor in different shapes were plotted separately using the improved crack size separation method and binomial distribution detection model. PVD thin-film sensors have 93.56% detection probability for cracks with length greater than 0.99mm at 95% confidence level. Compared to coin-shaped thin-film sensors, concentric ring shape film sensors are more sensitive to cracks smaller than 0.5mm, and the sensors with finer channel width have the higher probability of detection on small size cracks.

Keywords