Molten sand and ash may deposit on turbine blades and other hot section components of gas turbine engines, forming a glassy material composed mainly of calcium magnesium alumino-silicates (CMAS) that can cause damage. CMAS accumulates on the thermal barrier coating (TBC) and infiltrates the pores, causing the coating to crack and spall. MetroLaser is developing optical coherence tomography (OCT) as a non-destructive method to inspect engine components, revealing CMAS damage and helping to determine the remaining lifetime of a component.
Figure 1 shows a three dimensional (3D) volumetric OCT scan on a thermal barrier coating of a heavily CMAS-infiltrated turbine blade. The interface between the TBC and the substrate can be clearly seen, appearing as a dim line running parallel to the top surface. The signal intensity as a function of depth into the sample is affected by the absorption and scattering coefficients of the CMAS/TBC layers, and contains information on the infiltration depth. MetroLaser is exploiting these relationships to enable a damage assessment tool.
We are investigating OCT for its ability to detect CMAS beneath the surface of a TBC. A commercial OCT system operating at a wavelength of 1300 nm (Thorlabs Telesto II, shown in Figure 2) is being employed to investigate turbine blades and test coupons. OCT can be thought of as similar in principle to ultrasound, but using light. The use of a light source with short temporal coherence produces depth resolution on the micrometer scale that can be used for volumetric imaging. Typically, OCT is used as a filter that suppresses multiple light scattering and preserves the single-scattering component characterized by well-defined scattering angles and polarization. This allows three-dimensional rendering of structures in semi-transparent media. However, it is also possible to investigate the multiple-scattering regime of wave propagation using OCT, which as shown below has intriguing possibilities for depth profiling of CMAS.
Three TBC-coated turbine blades were examined. The blades included a clean blade with no prior service (“pristine”), a clean blade that had been distressed and was partially spalled, and a blade taken out of service that had obvious signs of CMAS infiltration.
Figure 3 shows processed OCT images for each case, in which the horizontal axis represents the position along the red arrow (which in this case is chosen to be more or less parallel to the physical surface of the object) and the vertical axis represents the optical depth into the object. The optical depth is the product of the physical depth and the refractive index of the medium. Looking first at the pristine blade, two more or less parallel bright lines can be seen in the image, the brightest of which (seen at the top) corresponds to the top surface of the TBC, and the other to the TBC-substrate interface. The OCT signal can be seen to continue below this interface, giving the appearance that light extends into the metal. However, since light cannot penetrate metal, this part of the signal is known to be due to multiple scattering, either within the TBC or at the bond coat.
Several qualitative differences can be immediately seen between the pristine blade and the CMAS blade in Figure 3. For the CMAS blade, the TBC-substrate interface is all but invisible. Furthermore, the bright line marking the top surface is not as sharp as it is for the pristine blade, and it fades more gradually with depth. We envision that effects such as these can be exploited to determine the extent of CMAS damage to thermal barrier coatings in advanced aircraft engines.