Study On Tomography Of Temperature And Concentration In Combustion Based On Absorption Spectrum Technology

Tunable diode laser absorption spectroscopy (TDLAS) technique, with the superiority in fast response, simplicity, adaptation to environment and wide measuring range, can realize non-intrusive measurement. It's suitable for the measurements of high temperature gas in combustion. Due to the line of sight property of TDLAS, we can only obtain the average measurement along the light direction. Because the flow disturbance, non-uniform distribution of fuel in the process of combustion, the theperature and concentration of the target gas is non-uniform. The average measurement along the light direction does not accurately reflect the combustion state. For the spatial distribution information of gas parameters in combustion, the development of effective two-dimensional diagnosis technology is needed. This thesis developed a series of research on non-uniform temperature field and concentration field reconstruction based on TDLAS by measuring water vapor in combustion products. The outlines and conclusions are giving as follows:Numerical simulation models and experimental system are established for one-dimensional uniform field to demonstrate the ability of part of the point of unimodal fitting method is able to achieve accurate measurement. The latest HITRAN2012 spectral database provide 224515 H2O absorption lines within 0?25711cm-1. The complete isolated and meeting the demand of measurement absorption lines is difficult to find in these lines. So part of the point of unimodal fitting method is developed to reduce the interference from nearby absorption peaks. Two lines,7185.59cm-1 and 7444.36cm'1, are used to demonstrate the accuracy of this method by fitting the simulate absorbance curve. The result prove that the above fitting scheme can effectively improve the computing speed by reducing the independent variables and the measured the temperature and concentration consistent with the set value. To verify the reliability of fitting scheme used in the actual temperature measurement, a high temperature tube furnace measuring system is builded and a series of different temperature are measured. The experimental results show that the measured results and the set value agree within 3%.Two-dimensional reconstruction algorithms using two absorption line are modified and used to the measurement of actual flame. Designed a multipass measurement system for straight association-like aeroengine. The reliability of measurement system is verified by experiment. Simultaneous algebraic reconstruction technique, considering the contribution of different optical path that through the same grid, has a certain ability to resist noise. Based on the algorithm, a nonnegative constraint and the smoothness regularization are added considering the real physical absorption process. To verify the effectiveness of the modified algorithm, two kinds of temperature and concentration distribution model are set for reconstruction. Compared with ART algorithm and SART algorithm, we can find the modified SART algorithm reconstructive precision is the highest and the noise immunity has been further improved on the basis of the original. The temperature and water vapor concentration distribution of the furnace flame plane are reconstructed with measured projection from 8 angles,38 light of every angles. The reconstruct temperature distribution agree well with the thermocouple measurement. On this basis, a multipass measuring system for Straight association-like aeroengine is designed. To verify the reliability of the measuring system, various gas-oil ratio combustion are measured. Reconstruction results can accurately reflect the flame position.Alternating iterative optimization algorithm of temperature and concentration imaging has been proposed, solving a low utilized efficiency of the multispectral absorption. Firstly, a linearization scheme was used in calculation of concentration field, weakening the nonlinearity of the original system; secondly, an alternately iterative scheme is used to reduce half the unknown variable number in each iteration; Finally, a penalty term is introduced in the optimization process to overcome the illness problem of reconstruction. To verify the validity of the algorithm, two different temperature and concentration distribution models are set and compared the reconstructed results from above algorithm with simulated annealing algorithm ones. The results show that the two algorithms are also be able to achieve accurate reconstruction of temperature field and concentration field, while the alternating iterative optimization algorithm require less time. To test the anti-noise ability of the algorithm, the measurements used for calculated are superimposed with different levels of Gaussian noise. Reconstruction results show that the algorithm has strong noise immunity. The alternating iterative optimization algorithm are used for flat flame furnace temperature and concentration distribution. Absorbance of ten different lines in two orthogonal directions were measured using time division multiplexing technology. The reconstruction results are in good agreement with the corrected thermocouple measurements.