8.3 Thermografische Verfahren
Filtern
Sprache
- Englisch (3) (entfernen)
Referierte Publikation
- nein (3) (entfernen)
Schlagworte
- Active thermography (3) (entfernen)
Organisationseinheit der BAM
Eingeladener Vortrag
- nein (2)
Learned block iterative shrinkage thresholding algorithm for photothermal super resolution imaging
(2020)
Block-sparse regularization is already well-known in active thermal imaging and is used for multiple measurement based inverse problems. The main bottleneck of this method is the choice of regularization parameters which differs for each experiment. To avoid time-consuming manually selected regularization parameters, we propose a learned block-sparse optimization approach using an iterative algorithm unfolded into a deep neural network. More precisely, we show the benefits of using a learned block iterative shrinkage thresholding algorithm that is able to learn the choice of regularization parameters. In addition, this algorithm enables the determination of a suitable weight matrix to solve the underlying inverse problem. Therefore, in this paper we present the algorithm and compare it with state of the art block iterative shrinkage thresholding using synthetically generated test data and experimental test data from active thermography for defect reconstruction. Our results show that the use of the learned block-sparse optimization approach provides smaller normalized mean square errors for a small fixed number of iterations than without learning. Thus, this new approach allows to improve the convergence speed and only needs a few iterations to generate accurate defect reconstruction in photothermal super resolution imaging.
Lock-in- and flash thermography are standard methods in active thermography. They are widely used in industrial inspection tasks e.g. for the detection of delaminations, cracks or pores. The requirements for the light sources of these two methods are substantially different. While lock-in thermography requires sources that can be easily and above all fast modulated, the use of flash thermography requires sources that release a very high optical energy in the very short time.
By introducing high-power vertical cavity surface emitting lasers (VCSELs) arrays to the field of thermography a source is now available that covers these two areas. VCSEL arrays combine the fast temporal behavior of a diode laser with the high optical irradiance and the wide illumination range of flash lamps or LEDs and can thus potentially replace all conventional light sources of thermography.
However, the main advantage of this laser technology lies in the independent control of individual array areas. It is therefore possible to heat not only in terms of time, but also in terms of space. This new degree of freedom allows the development of new NDT methods. We demonstrate this approach using a test problem that can only be solved to a limited extent in active thermography, namely the detection of very thin, hidden defects in metallic materials that are aligned vertically to the surface. For this purpose, we generate destructively interfering thermal wave fields, which make it possible to detect defects within the range of the thermal wave field high sensitivity. This is done without pre-treatment of the surface and without using a reference area to depths beyond the usual thermographic rule of thumb.
Active thermography with flash and halogen light excitation is used as a method for non-destructive testing of 3D-printed polymer components. Test specimens with artificial defects have been generated, using laser sintering and fused layer modeling. These test specimens have been investigated in different measurement configurations with both excitation methods. Afterwards, the different measurement conditions were compared regarding their capability to detect the defects. Furthermore, advanced analysis methods are used, to fully exploit the capabilities of these techniques.