Abstract
Modern gas turbines offer high potential for future energy generation using alternative fuels. Therefore, these machines need to be further developed to increase the service life of their built-in components and simultaneously increase their performance and efficiency. This can be achieved mainly through a higher compressor outlet pressure and a higher turbine inlet temperature, which lead to high stresses on the hot-gas-leading components in particular. These loads and environmental influences can lead to certain failures in gas turbines, including ruptures, creep deflections, high- and low-cycle fatigues, oxidation, corrosion, erosion, rubbing or wear, foreign object damage, thermal mechanical failures and combinations thereof. Since these factors can lead to cracks and component failure, reliable testing techniques are important for early maintenance. However, because of the complexity and costs associated with these techniques, it is desirable to use these components for as long as possible.The first research focus is on studying creep damage and early detection of cracks in turbine blades with modulated nonlinear ultrasonic methods. In general, nonlinear techniques have significant advantages over linear ones and enable earlier detection of failures. In a further development of the nonlinear method, ultrasonic waves were emitted with multiple frequencies, which generate harmonic and modulated response frequencies when the waves pass through failures in materials. This offers further possibilities for failure detection and classification. Therefore, the wave equation was solved analytically with multiple excitation frequencies to derive novel nonlinearity parameters. These parameters are the ratios of the measured amplitudes of the harmonic/modulated response frequencies and the fundamental frequencies, and thus they form the basis for failure evaluation. Their existence was verified numerically and experimentally, allowing cracks to be identified and their sizes estimated.
Plastic deformation forms during creep, which is one of the most common failure modes in gas turbines. In this type of failure, the components deform under a constant load depending on the time and temperature. Therefore, creep samples made of the Hastelloy X material were examined to obtain a better understanding of the damage process. The results showed that pores and microcracks propagate from the outside inwards. At the core of the samples, molybdenum was found to be primarily precipitated, while chromium was the main precipitation element in the edge area, leading to the formation of a chromium oxide layer. Moreover, modulated ultrasonic tests showed that
the harmonic nonlinearity parameters in particular react very sensitively to the damage sizes of the pores and microcracks.
If the cracks are closed, the ultrasonic waves can pass through the flaws unhindered, without the generation of harmonic or modulated response frequencies. Therefore, the second research focus, ultrasonically stimulated thermography, is on the detection of such cracks. With this technique, turbine blades with trailing edge cracks were vibrated to detect the increase in temperature as a result of frictional heat in the crack area with a thermal imaging camera. When a piezo actuator was used for excitation, the increase in temperature was determined as a function of the excitation frequency, which localised a hidden crack under the ceramic thermal barrier coating. In addition, with a constant excitation frequency and increasing energy input into the actuator, the temperature behaviour revealed whether the crack was constrained or preloaded. This was simulated with a developed finite element model to verify the experimental results.
In conclusion, the proposed techniques allow the sensitive and early detection of different types of cracks and creep damage with a high accuracy, which in turn allows a stable and safe use of the gas turbine components during operation.
| Date of Award | 22 Feb 2023 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Michele Meo (Supervisor) & Fulvio Pinto (Supervisor) |