An acoustic time-of-flight approach for unsteady temperature measurements: Characterization of entropy waves in a model gas turbine combustor
Within the framework of an AG Turbo 2020 project which deals with thermoacoustic instabilities in lean and premixed operated combustors, entropy waves are investigated. Such non-isentropic temperature fluctuations are caused by equivalence ratio perturbations at the burner inlet and trigger the so-called indirect combustion noise. This is because of the sound generation due to the acceleration of the entropy fluctuations at the first stage of the turbine. When reflected back to the flame, these acoustic waves affect the heat release rate and thus might close a thermoacoustic feedback loop that could lead to hazardous acoustic instabilities. Furthermore, the high pressure amplitudes limit the operational range of the combustion chamber and thus impede the operation in a low NOx regime.
In order to measure temperatures in a corrosive and high-temperature environment with a sufficient high time resolution, a novel measurement technology has been developed. The measurement method is based on a time-of-flight approach. An acoustic pulse is generated and by means of a microphone its traveling time is estimated. This provides the line-integrated inverse speed of sound and finally leads to an estimate of the line integrated temperature. The pulse is generated by means of an electric spark discharge which features high-temperature resistance, high repetition rates, a sufficiently high SPL, as well as proper detectability in the very noise environment of a gas turbine combustion chamber. By employing several acoustic receivers in one axial measurement plane in the combustor, tomographic methods allow the reconstruction of the time resolved 2-dimensional temperature field.
Based on the experimental results, analytical models for the prediction of the generation and the transport of entropy waves are derived and validated. Goal is to obtain the transfer function between equivalence ratio fluctuations in the burner and entropy fluctuations at the turbine inlet and thus to predict the relevance of indirect combustion noise.