Session: 04-38 Combustion dynamics - flame response I

Paper Number: 122798

122798 - Inferring Flame Transfer Functions of Turbulent Conical Flames From Pressure Measurements 

We use approximate Bayesian inference, accelerated by adjoint methods, to construct a quantitatively accurate model of the thermoacoustic behaviour of a turbulent conical flame in a duct. We first perform a series of automated experiments to generate a data set. The data consists of time series pressure measurements from which we extract (i) the growth rate and (ii) natural frequency of oscillations; and (iii) the Fourier decomposed pressure at eight locations in the rig.  We then propose several candidate models of the rig's components and assimilate the data into each model to find the most probable parameters for that model. Using Bayesian model comparison, we rank the candidate models based on their marginal likelihood (evidence) and select the most likely model for each component. We begin this process by rigorously characterising the acoustics of the cold rig. We then introduce a series of different flames and infer their flame transfer functions with quantified uncertainty bounds. We do this with the flames in-situ, using only pressure measurements. We find that we can infer the flame transfer functions from pressure measurements alone, but if the thermoacoustic effect is weak or the experimental error is large, the uncertainties in the inferred flame transfer functions can be large.

Presenting Author: Matthew Yoko University of Cambridge

Presenting Author Biography: Matthew is a final year PhD candidate in the Engineering Department at the University of Cambridge.

Authors:

Matthew Yoko University of Cambridge
Matthew P. Juniper University of Cambridge