Quantitative evaluation of material properties with measurement grid and nonlinear multivariate searching algorithm

摘要

The micro damage and stress centralization on metal surface may cause huge damage to some airplane components, which can be found through quantitative evaluation of material properties, such as conductivity and permeability. In this paper, a novel planar electromagnetic sensor with spatially periodic exciting coil and spiral detection coils is introduced. It can be made in the flexible printed circuit board and applied to the inspection of components with different shape. The model of this sensor can be simulated and analyzed accurately and the nonlinear measurement grid can be established in advance, which is a database and reveals the relationship between the sensor output and several variates such as material properties and the liftoff. In order to ascertain the material property according to the measurement grid and the sensor output, a nonlinear multivariate searching algorithm is presented, which is a combination of rough searching and fine searching. Firstly, the initial grid cell is ascertained with dichotomy, and then the accurate target grid cell is searched through the judgment of location relationship between grid cell and the target point. This algorithm can not only shorten the searching time, but also reduce the failure ratio of searching. It provides a rapid and accurate way for quantitative evaluation of material properties.