Simultaneous reconstruction of 3D temperature distribution and radiative properties of participating media based on the multi-spectral light-field imaging technique

Abstract

A method based on multi-spectral light-field imaging technique is developed for the simultaneous reconstruction of the multi-dimensional inhomogeneous temperature distribution and radiative properties (scattering and absorption coefficients) of participating media. The generalized sourced multi-flux method (GSMFM) is applied to simulate the exit radiation intensities in arbitrary directions at the boundary surfaces of absorbing, emitting, and scattering media, which are used as inputs for the inverse analysis. Based on the forward GSMFM solution, a hybrid least-square QR decomposition–conjugate gradient (LSQR–CG) algorithm is developed to reconstruct simultaneously the multi-dimensional temperature distribution and the absorption and scattering coefficients of cylindrical participating media. The retrieval results indicate that the temperature distribution and radiative properties can be recovered even with noisy data, and that the reconstruction results of the temperature distribution are more accurate than those of the radiative properties. An experiment is performed to validate the suitability of the proposed method, and the temperature distribution and radiative properties of the ethylene diffusion flame with Al2O3 particles are reconstructed using the LSQR–CG algorithm. Moreover, the thermocouple is applied to measure the temperature of flame at different positions, and the measurement results are compared with the reconstructed results. All the experimental results show that the temperature distribution and radiative properties of ethylene flame can be reconstructed reasonably using the proposed LSQR–CG method combined with the multi-spectral light-field imaging technique, which has the potential to be widely applied to the non-contract measurement of the temperature distribution and radiative properties in practical engineering applications.