Non-destructive sub-surface inspection of multi-layer wind turbine blade coatings by mid-infrared Optical Coherence Tomography

TL;DR

This study demonstrates that mid-infrared Optical Coherence Tomography (MIR OCT) with a ~4 μm center wavelength can non-destructively inspect the topcoat and primer layers of wind turbine blade coatings, achieving sub-surface visualization up to ~360 μm. By comparing MIR OCT to near-infrared OCT and X-ray CT, the authors establish MIR OCT’s superior penetration over NIR for organic coatings and its faster, higher-resolution capability for localized coating evaluation, while acknowledging X-ray CT’s deeper penetration for bulk inspection. The work introduces thickness estimation via attenuation-coefficient fitting and automatic interface detection, enabling thickness maps and defect identification (inclusions, cracks, holes) across the coating stack. Collectively, MIR OCT is shown as a complementary, non-destructive tool that could replace destructive dolly tests and enhance quality control in WTB production, with potential for full-blade scanning and AI-assisted analysis to accelerate throughput.

Abstract

Non-destructive inspection (NDI) is useful in the industrial sector to ensure that manufacturing follows defined specifications, reducing the quantity of waste and thereby the cost of production. In response to climate change and the need to reduce carbon footprints, there is a growing demand for NDI solutions that are primarily non-contact, as well as non-destructive, fast, and easily integrated. Optical Coherence Tomography (OCT), a well-known diagnostic technique in medical and biological research, is increasingly being used for industrial NDI. In the mid-infrared (MIR) wavelength range, OCT can be used to characterise parts and defects not possible by other industry-ready scanners, and enables better penetration than conventional near-infrared OCT. In this article, we demonstrate NDI of wind turbine blade (WTB) coatings using an MIR OCT scanner employing light around 4 $μ$m from a supercontinuum source. We inspected the top two layers of the coating (topcoat and primer) in two different samples. The first is to determine the maximum penetration depth, and the second one is to emulate defect identification. The results of our study confirm that MIR OCT scanners are highly suitable for coating inspection and quality control in the production of WTBs, with performance parameters not achievable by other technologies.

Figures (11)

• Figure 1: (a) Schematic of the layer composition of the samples visualised in (b,c). (b) Sample with aluminium foil patches placed at the interfaces between layers. (c) Sample with punch defects in different layers. For (b-c), (i) and (ii) show opaque and transparent 3D illustrations of the samples.
• Figure 2: (a) 3D representation of the MIR OCT system, SC: supercontinuum, LP: 3.2 $\mu$m long pass filter, BS: beam splitter, GM: gold mirror, Window: dispersion compensation. (b) Reference spectrum in red, measured interference spectrum in black, upconversion efficiency in blue, and (c) the corresponding calculated A-scan. (d) Schematic of the different OCT scans used in this article.
• Figure 3: (a-b) Comparison of penetration depth between B-scans in MIR OCT ($4~\mu$m centre wavelength) and (c-d) in NIR OCT ($1.3~\mu$m centre wavelength). The dashed green line in (c-d) shows the part where the coating was intentionally removed for the NIR OCT measurement to find the presence of aluminium foil. For each B-scan, the physical distance in depth was calculated with $n_{\mathrm{topcoat}} \simeq 1.49$. (e-f) Comparison of the A-scan average of each B-scan is depicted in (a-d). (e) Present the results with the Optical Path Delay ($OPD$) scale and (f) scale with $n_{\mathrm{topcoat}} \simeq 1.49$ (from 0 to 0.14 $\mu$m ) and then scale with $n_{\mathrm{primer}} \simeq 1.84$ (blue scale from 0.14 $\mu$m).
• Figure 4: Illustration of the impact of introducing punch defects at different layers in the production with (i) an ideal case without any reflow, and (ii) the worst case of reflow of wet paint, (a) in the topcoat, (b) in the primer, and (c) in the fine filling.
• Figure 5: Comparison between MIR OCT and X-ray CT measurements made of the defect made in the topcoat part with a diameter of $3$ mm. (a-b) En face view of the surface of the sample. (c-d) Selected B-scan in approximately the same place. The blue dashed lines in (a-b) show where the B-scans were extracted.