Study on two-dimensional tomography algorithm for gas temperature distribution based on TDLAS

Abstract

In the combustion flow field, the concentrations of temperature and water vapor are very important in determining combustion efficiency. The traditional contact measurement will induce shock so as to disturb the flow field, and most of the probe can’t be used in the high temperature air. So the existing contact measurement method can't meet the measurement requirements of the combustion field, but the tunable laser absorption spectrum technology (TDLAS) can realize non-contact nondestructive measurement of the combustion flow field. Various parameters such as temperature, gas composition and concentration, flow velocity, can be measured at the same time. And there is no temperature limit. It is very good at measuring combustion field parameters in the high temperature and high speed environment. TDLAS can calculate the gas temperature in real-time by scanning both absorption signal of gas absorption lines, but this is one-dimensional path integral measurement, can’t reflect the real information of the combustion field. So it can't be used to measure objects with distinct temperature gradient. In order to overcome this deficiency, tunable laser absorption spectrum technology combined with computer tomography technology (called TDLAT) is used to realize the measurement of the two dimensional temperature distribution in the burning flow field. In this paper, the measurement principle and algorithm of the two dimensional temperature field distribution are put forward. In TDLAT system, the measured area is divided into many grids. TDLAS is used to get the laser path integral spectrophotometry along the grid line. In succession, deeply grid information is gotten by non-negative constrained least squares. Thus, assuming that temperature measurement plane within is in smooth transition, interpolation algorithm is used to recreate the high spatial resolution of the two dimensional temperature field distribution. According to the measuring principle and measuring objects, the model established is used to get the simulation result. The algorithm for TDLAT system was also verified to determine where it is consistent or not. As a result, the deviation value is less than 3% between the result of the temperature distribution and original hypothesis spectrophotometry, which shows that the algorithm is self-consistent.