No abstract

No abstract

The prognosis of thermo-acoustic/combustion instability is usually accomplished by applying a priori knowledge about features of unstable operation and measuring deviation from those features using point values. In the present work, we adopt a different methodology, whereby the presence and extent of the signature of unstable combustion are learnt as an anomaly from the distribution of pressure oscillations during stable operation across several protocols. The protocols involve a transition from stable to unstable operation in a swirl combustor. It is inferred that the stable combustion in the present case is stochastic noise with a normal distribution containing values comparable with root-mean-square values at unstable operation with a [Formula: see text] value 0.05–0.07. We exploit this feature to detect anomalies from flame intensity images, which represents the heat release rate fluctuations by manipulating their features to be a part of multivariate Gaussian distribution. To formulate this distribution, we employ a convolutional-neural-network-based variational auto-encoder (CNN-VAE) and express the associated reconstruction loss as an anomaly metric. The anomalies obtained through CNN-VAE and integrated intensity fluctuations are then evaluated for their sensitivity against the unsteady pressure data. The analysis reveals that the CNN-VAE metric performs better than the integrated intensity fluctuations for predominantly all [Formula: see text] values.

Industrial gas turbines commonly use lean premixed swirl stabilised combustors that are prone to thermo-acoustic instabilities. Combustor testing involves several steps ranging from atmospheric to high-pressure conditions. An open outlet boundary condition is maintained for atmospheric tests commonly, whereas high-pressure testing involves complex exit boundary conditions, which changes the reflection coefficient and can affect the nature of instability. Current work aims at studying how the change in outlet boundary affects the nature of instability due to changes in the exit reflection coefficient and acoustic mode shape inside the combustor. A laboratory-scale industrial swirl burner that uses partially premixed methane and air at atmospheric conditions is analysed for this purpose. A constant area contraction ratio of 6.5:1 is maintained at the exit of the combustor while varying the inlet Reynolds number at a constant global equivalence ratio. Flame dynamics based on conditionally phase averaged OH* chemiluminescence images and spatial Rayleigh index maps were used to compare different flowrate and exit boundary cases. The outlet contraction affects both the frequency and amplitude of the dominant thermoacoustic mode. The orifice plate at the exit reduces the outlet reflection coefficient leading to a change in acoustic mode shape inside the combustor. Overall, the instability amplitude is reduced considerably for cases with orifice plate at the exit boundary compared to open exit boundary cases. The results show the importance of defining outlet boundary condition as a parameter in combustion instability studies.

Current study focuses on the effect of multiple heat release mode interactions on the amplitude of unsteady pressure oscillation for a partially premixed radial swirl burner, which uses methane as fuel. A range of operating conditions based on inlet airflow rate and global equivalence ratio is considered for this purpose. The pressure time series shows amplitude modulation at the dominant frequency for all the conditions considered. Continuous wavelet transform based analysis illustrates the presence of heat release rate fluctuations at multiple frequencies other than the dominant mode of pressure oscillation, with this more pronounced at low frequencies. Spectral Proper Orthogonal Decomposition performed on time-resolved CH* chemiluminescence images reveal four dominant spatial modes, chosen based on the dominant wavelet coefficients for the same. The identified frequencies correspond to the duct-acoustic mode, helical mode (spectrally close to acoustic mode), low-frequency axisymmetric mode and low-frequency helical mode. The low-frequency helical mode (considered as the result of non-linear interaction between acoustic and helical mode) and the low-frequency axisymmetric mode (considered to have independent existence) have similar spectral content. Amplitude modulation of unsteady pressure is found to be a result of the superposition of duct-acoustic mode and low-frequency axisymmetric mode, whereas reduction in overall pressure amplitude with the decrease in global equivalence ratio is a result of an increase in the dominance of low-frequency helical mode. The relative dominance of low-frequency helical mode over dominant pressure mode reduces the overall pressure amplitude.