Speaker
Description
This study demonstrates the practical application of thermal tomography based on the virtual wave concept for inspection tasks from the manufacturing industry as well as from the maintenance and repair (MRO) sector. The thermographic reconstruction of defects and damage located inside a specimen is generally a very challenging task, since information loss occurs due to entropy production during heat conduction. The experimental study on artificial defects and real components also reveals the limitations of the reconstruction process with regard to complex geometry, multilayered material systems, and disturbances due to convection in harsh environments.
In a first step, thermal waves are generated by a rectangular-shaped external optical excitation of the component surface. These broadband thermal wave signal diffuse through the test object, and the effect of the disturbed thermal waves on the surface temperature is captured contactless as thermal images using an infrared camera. The measured surface temperature signals are transformed locally into a virtual wave signal by means of a mathematical transformation in the form of a Fredholm integral of the first kind. Using novel iterative regularization methods that allow the incorporation of additional information about the experiment (e.g., detector noise characteristics, sparse reconstruction matrix, and non-negative temperature values), this ill-posed problem can be solved approximately. The depth-dependent positions and amplitudes of the sources and sinks in the virtual wave signal allow deductions about the type of interface layer and its depth.
As experimental results, we present various one-, two-, and three-dimensional reconstructions of real defects, primarily in composite components, such as delaminations, disbonds, inhomogeneous fiber distributions, water ingress, and more. One-dimensional virtual A-scans are used to explain in detail the behavior of the virtual wave signal due to various interfaces in the material with different effective thermal mismatch factors. Virtual B- and C-scans are used to visualize the internal interface layers. Examples are used to demonstrate how lateral heat flows can lead to systematic deviations in the determination of defect depth in the case of spacially limited defects or damage. The results of the thermal tomography experiments are subsequently interpreted and validated using 3D X-ray computed tomography data.
