Theory And Instrumentation For Pulsed Eddy Current Wall-thinning Measurement Of Ferromagnetic Objects

The ferromagnetic pressure equipments (e.g. pipes and vessels) with coatings often suffer thickness reduction from corrosion by aging. The failure of such equipments caused by wall thickness thinning will bring great loss of life and property. As an effective novel nondestructive testing (NDT) technology, pulsed eddy current (PEC) technology has become a research focus in the recent years. The PEC measurement theory, method and instrument for ferromagnetic pressure equipments are researched in the paper. Firstly a system model for PEC testing system is presented based on classic eddy current ring equivalent theory and system theory. The presented model simplifies the complicated pulsed eddy current behavior in the ferromagnetic specimen to the step response of a multiple-coil-coupling equivalent system. The model is easy to be derived and suitable for engineering application. A double-logarithmic domain median filtering pre-processing method for practical noisy PEC signal with large dynamic range is presented. Experiment result shows the method suppresses noise impacts with little signal distortion. Model fitting is then applied to the pre-processed signal and the estimated model parameters are analyzed. The parameter which is sensitive to specimen thickness variation and explains the signal characteristic best is selected as the feature value of the specimen thickness. Furthermore, the physical meaning of the feature value is explained. The experiment result verifies the effectivity of the pre-processing method and also the correctness of the presented model. To achieve a quantitative result of the testing result, a method based on cumulative integration-fitting-differential process is proposed. The method exploits cumulative integration process to eliminate the noise impacts because the mathematic expectation of the mean of a time domain noise is zero. By non-linear fitting to the cumulative integration curve of the original PEC signal, the model parameters are obtained. The fitted curve is then differentiated to achieve an estimation of the PEC signal waveform. The method is tested by simulated PEC signal and practical noisy PEC signal. The result shows the method is able to restore the signal with little distortion even in noisy environment. Some impact factors for quantitative result e.g. material property or probe lift-off distance are investigated. One the basis of above researches, a PEC thickness measurement instrument for ferromagnetic pressure equipments is developed. The key problem for power-amplifying stage in equipment development is overcome and discussed in detail. The performance of the developed instrument e.g. maximum probe lift-off distance and measurable thickness range is assessed by various experiments. The experiment result shows the error of the instrument's quantitative result is lower than 5% of the reference thickness. Therefore the error level is the same as the foreign instrument. The field application indicates the instrument is reliable and able to measure the thickness of the ferromagnetic equipment without removal of the coatings. The research offers a promising testing technology for petrochemical and power equipments and supports the edition of the relative draft standard.