Nondestructive Evaluation of Coating Defects and Uniformity Based on Terahertz Time-Domain Spectroscopy

Abstract

Structural coatings are widely used because of their excellent mechanical and thermal properties. To evaluate defects and uniformity in coatings, both qualitatively and quantitatively, a terahertz time-domain spectroscopy (THz-TDS) detection technique is proposed. The thermal barrier coating is selected as a typical single-layer coating structure for quantitative defect detection. A wavelet noise reduction method is used on the acquired raw signals to eliminate noise while retaining detailed information. The peak value of the preprocessed signal is used as a feature parameter for imaging, and the automatic binarization threshold segmentation technique is used to describe the defects quantitatively. The automotive coating is selected as a typical multilayer coating structure for uniformity detection. The time-frequency characteristics of a strongly superimposed signal are analyzed; the peak-to-peak value is used as a feature parameter for imaging, and the peak-to-peak 3D imaging is then used to characterize the coating uniformity, enabling fast and intuitive acquisition of the coating state. The statistical characteristics of the standard deviation and range are used to evaluate the uniformity of each layer of the automotive coating. The results show that the uniformity of the clean coating is optimal. The results of a subsequent thickness inspection using an eddy current gauge are consistent with those of the terahertz technique. The results demonstrate that THz-TDS can effectively detect defects and uniformity in coatings.