Gaussian windowed frequency-modulated thermal wave imaging for Non-Destructive testing and evaluation of steel materials

Maintaining the structural integrity of mild steel components is essential in numerous industrial applications, as subsurface flaws can severely affect the material’s mechanical strength and its ability to withstand loads. Traditional thermal non-destructive testing (TNDT) techniques, including conventional active infrared thermography methods such as pulsed thermography (PT), lock-in thermography (LT), and pulse-phase thermography (PPT), face limitations in terms of resolution, energy dispersion, and defect detectability. To address these challenges, this study employs spectral reshaping of linear frequency modulated waveform using Gaussian windowing for enhanced subsurface defect detection. The research primarily focuses on detecting subsurface blind hole defects at varying depths in mild steel samples. Post-processing approaches including frequency-domain analysis, time-domain phase analysis, and correlation-based processing were applied, along with Wiener filtering, to improve defect visibility. While frequency-domain analysis was limited by energy dispersion and fixed resolution, time-domain analysis showed better performance by concentrating energy with higher resolution. However, correlation-based time-domain processing using pulse compression produced the best results by significantly enhancing defect contrast and resolution. Cross-correlation coefficient (CCC) analysis with Gaussian windowing frequency modulated thermal wave imaging (GWFMTWI) clearly revealed all six subsurface defects, outperforming conventional linear frequency modulated thermal wave imaging (FMTWI). The improvement was further confirmed by the signal-to-noise ratio (SNR) as a figure of merit, providing a quantitative measure of defect detection reliability.