Thermoacoustic Combustion Instabilities Research Articles

Application of neural dynamic optimization to combustion-instability control

The suppression of thermoacoustic combustion instabilities represents one of the main goals in the design of reliable high-performances combustion chambers. Unstable dynamics arise when a non-linear coupling is established between the acoustic field and the flame front generating high-amplitude and low-frequency pressure and heat release oscillations, associated with the excitation of the combustor’s natural modes. Temperature and pressure peaks due to these phenomena are particularly harmful for the structural damage they can cause as well as for performance degradations and increase of pollutant emissions. Due to the non-linear nature of the phenomenon, relevant problems arise when it is necessary to define model-based control-systems. The aim of this study is to define a control strategy, based on the application of recent results in the field of neural control of non-linear systems. The proposed strategy is an application of an innovative neural-network-based technique, namely Neural Dynamic Optimization, which is able to exploit the potential of optimal control strategies in dealing with complex non-linear systems and the flexibility and the generalisation properties of neural networks. Reported simulations show the satisfactory performance of the proposed controller in suppressing undesired thermoacoustic combustion instabilities.

Neural prediction of combustion instability

Combustion instabilities consist in self–exited oscillations in combustion chambers and can cause structure degradations. When combustion instabilities occur, the combustion process is characterised by the coupling of various non-linear phenomena, which can lead to the formation of either a limit cycle or chaos. The intrinsic highly non-linear behaviour, especially when chaos arises, poses a major problem for the formulation of good predictions and the design of reliable control systems. Due to the relevant number of degree of freedom and to the non-linear coupling of different phenomena, the mathematical modeling of combustion instabilities is computationally heavy and may produce an unsatisfactory correspondence between simulated and experimental data. Analogous problems arise also from the uncertainty for the parameters of the process, such as the local velocity of the flame front (and, hence, the reaction rate and the exhausted temperatures), and their unpredictable variations. In fact, note that most of the chemical and thermo-physical variables both strongly depend and influence the instantaneous displacement of the flame front, which is positioned on the unstationary eddies external surface. In order to obtain a reliable model for thermo-acoustic combustion instabilities, a different approach was chosen in this paper. The black-box identification of an experimental combustion chamber was obtained by means of a generalized NARMAX model. The model was implemented by training a Multilayer Perceptron artificial neural network with input-output experimental data. The main advantages of the proposed approach consisted in the natural ability of neural networks in modeling nonlinear dynamics in a fast and simple way and in the possibility to address the process to be modeled as an input-output black box, with little or no mathematical information on the system. Satisfactory agreement between simulated and experimental data was found and results show that the model successfully predicted the temporal evolution of thermo-acoustic combustion instabilities.

Experimental analysis of thermo-acoustic combustion instability

Thermo-acoustic instabilities are dynamic phenomena that represent a major threat for most modern combustion systems. Many studies, mainly undertaking a linear analysis of experimental data, have been carried out to provide a deeper understanding of the underlying phenomena. However, linear analysis may lead to an oversimplified view of the problem, which involves many complex non-linear interactions, and a more detailed non-linear analysis may be necessary. This paper presents both linear and non-linear analyses of experimental measurements observed in a methane-fuelled laboratory combustor. The linear analysis aims to verify the existence of thermo-acoustic instabilities and consists of the elaboration of both power-spectral density distribution and Rayleigh Index of the experimental time series. Non-linear analysis aims to investigate dynamic behaviours by means of the deterministic chaos theory. Results of the analyses show that combustion instabilities occur in all the experimental operating conditions. Moreover, the existence of a chaotic source in the combustion system under study is demonstrated.

RECONSTRUCTION OF ACOUSTIC TRANSFER MATRICES BY INSTATIONARY COMPUTATIONAL FLUID DYNAMICS

Thermoacoustic combustion instabilities are a frequently encountered problem in the operation of combustion equipment. The “brute-force” application of computational fluid dynamics to the analysis of thermoacoustic instabilities is estimated to be forbiddingly expensive for many systems of technical interest due to the high computational demands of a time- and space-accurate simulation of a (low Mach number) compressible reacting flow in a complex geometry. Thermoacoustic systems can be modelled efficiently as networks of acoustic multi-ports, where each multi-port corresponds to a certain component of the system, e.g., air or fuel supply, burner, flame, combustor and suitable terminations, and is represented mathematically by its transfer matrix. For some multi-ports, the transfer matrix can be derived analytically from first principles: i.e., the equations of fluid motions and suitable approximations. However, the acoustic behavior of more complicated components, e.g., a burner or a flame, has to be determined by empirical methods, by using a “black box” approach common in communications engineering. In this work, a method is introduced which allows one to reconstruct the transfer matrix of an acoustic two-port from an instationary computation of the response of the two-port to an imposed perturbation of the steady state. Firstly, from the time series data of fluctuating velocity and pressure on both sides of the two-port, the auto- and cross-correlations of the fluctuations are estimated. Then, the unit impulse responses of the multi-port are computed by inverting the Wiener–Hopf equation. Finally, the unit impulse responses are z -transformed to yield the coefficients of the transfer matrix. The method is applied to the one-dimensional model of a heat source with time delay placed in a low-Mach-number compressible flow, for which an analytical description can be derived from first principles. Computational predictions of the transfer matrix have been validated successfully against these analytical results. Furthermore, a comparison of the novel approach presented in this paper with a method for computing the frequency response of a flame by Laplace-transforming its step response is carried out.

Nonlinear Heat-Release/Acoustic Model for Thermoacoustic Instability in Lean Premixed Combustors

Lean premixed combustors, such as those used in industrial gas turbines to achieve low emissions, are often susceptible to the thermoacoustic combustion instabilities, which manifest themselves as pressure and heat release oscillations in the combustor. These oscillations can result in increased noise and decreased durability due to vibration and flame motion. A physically based nonlinear parametric model has been developed that captures this instability. It describes the coupling of combustor acoustics with the rate of heat release. The model represents this coupling by accounting for the effect of acoustic pressure fluctuations on the varying fuel/air ratio being delivered to the flame, causing a fluctuating heat release due to both fuel air ratio variations and flame front oscillations. If the phasing of the fluctuating heat release and pressure are proper, an instability results that grows into a limit cycle. The nonlinear nature of the model predicts the onset of the instability and additionally captures the resulting limit cycle. Tests of a lean premixed nozzle run at engine scale and engine operating conditions in the UTRC single nozzle rig, conducted under DARPA contract, exhibited instabilities. Parameters from the model were adjusted so that analytical results were consistent with relevant experimental data from this test. The parametric model captures the limit cycle behavior over a range of mean fuel air ratios, showing the instability amplitude (pressure and heat release) to increase and limit cycle frequency to decrease as mean fuel air ratio is reduced.