Study On Temperature And Volume Fraction Fields Reconstruction Of Multiple Particles In Complex Combustion Media Based On Inverse Radiation Problem

Nanofluid refers to uniform and stabilized suspensions containing metal or nonmetal nanoscale additives in dilute concentrations.Conventional liquid fuels with the addition of nanoparticles can decrease ignition delay time and ignition temperature,reduce soot and gaseous pollutant(such as NO_x)emissions,and increase volumetric energy density and burning rate.Until now,studies on the combustion process and combustion characteristics of nanofluid fuel are still in a primary stage,which is largely limited by the development of combustion diagnostic techniques.The nanofluid fuel flame is a typical case of complex combustion systems,and the co-existence of various nanoparticles brings difficulties for combustion diagnostics.It is of great scientific significance and practical value to establish effective measurement techniques for multi-particles temperature and concentration fields of complex combustion flames to deeply study on the nanofluid fuel combustion.In this thesis,reconstruction models of multi-particles temperature and concentration fields in complex combustion flames are developed based on inverse radiation problem,and these temperature and concentration fields are obtained using inverse problem solving algorithm from the knowledge of flame emission radiation intensities.It provides the theoretical basis for the combustion diagnostics of nanofluid flame.Firstly,the theoretical basis and the inverse problem solving algorithm for the simultaneous reconstruction of multi-particles temperature and concentration fields are systematically introduced.The integral expression for radiation intensity in arbitrary directions is derived and radiation characteristics of individual particle and sparse particle systems are introduced,which provides the theoretical basis for solving direct radiation problem to obtain emission radiation intensities from the flame boundary.The principles of the truncated singular value decomposition algorithm(TSVD)and the least squares QR decomposition algorithm(LSQR)used to solve inverse problems are introduced in details.The reconstruction models of temperature and concentration fields are presented in a symmetric optically thin flame containing two kinds of particles.The concentration fields of different particles are decoupled through obtaining the volume fraction ratios between these particles.According to the technique for obtaining the volume fraction ratios,reconstruction models combining with thermophoretic sampling particle diagnostic-transmission electron microscope(TSPD-TEM)technique or based on multispectral technique are respectively proposed.Specifically,TSPD-TEM technique is used to directly measure the volume fraction ratios between different particles,and then LSQR algorithm or TSVD algorithm is used to simultaneously retrieve the multi-particles temperature and concentration fields from the knowledge of the radiation intensities at two wavelengths.Considering disturbances to the flame flow field by the contact TSPD-TEM technique,LSQR algorithm combined one-dimensional searching algorithm or nonlinear optimization algorithm based on the multispectral technique is further used to directly reconstruct these temperature and concentration fields without the prior knowledge of volume fraction ratios by the TSPD-TEM technique.Numerical results prove that the two reconstruction methods are accurate and effective,and the reconstruction accuracy is acceptable even with low signal to noise ratio of the measurement system.The simultaneous reconstruction methods for multi-particles temperature and concentration fields in complex combustion symmetric flame are improved by considering self-absorption effects of multi-particles.The attenuated radiation intensities are corrected to unattenuated radiation intensities using a formula dealing with the self-absorption effect of soot in sooting flame,and the multi-particles temperature and concentration fields are obtained by an iteration algorithm.As for the reconstruction method combined with the TSPD-TEM technique,the self-absorption effects are easily embedded into the inverse reconstruction model.As for the direct reconstruction method based on the multispectral technique,the self-absorption effects are hardly embedded into the inverse reconstruction model because there are so many unknown parameters that it is difficult to run the iterative algorithm.Therefore,a separation method for radiation intensity matrixes is proposed,and the volume fraction ratios are searched by a one-dimensional searching algorithm based on the separated radiation intensity matrixes ring by ring.After that,the multi-particles temperature and concentration fields are retrieved by the iteration algorithm combined with LSQR algorithm.The reconstructed results show that,as for optically thin flames,reconstruction accuracies of all the unknown fields can be improved by considering the self-absorption effects,and as for optically thick flames,the reconstruction may fail without consideration of the self-absorption effects.Further,extend the inverse reconstruction models used in symmetric flames containing two kinds of particles to apply in complex combustion flames containing three or four kinds of particles.A multi-dimensional optimization algorithm is introduced to solve multi-groups volume fraction ratios based on an objective function composed of emission source terms at multi-wavelengths,and the multi-particles temperature and concentration fields are obtained according to the radiation intensities at two wavelengths.The effects of searching ranges of the multi-dimensional optimization algorithm,wavelength combinations,numbers of detection lines and radiation intensity measurement errors on reconstruction accuracies are investigated.And the reconstructed results show that the reconstruction method has strong robustness,and high stability and accuracy.Finally,the simultaneous reconstruction models of multi-particles temperature and concentration profiles in asymmetric optically thin and optically thick flames are presented.A method to deal with the self-absorption effects of multi-particles in complex combustion asymmetric flames is explored and the influence of the self-absorption on reconstruction accuracies is investigated.