Simulation Study On Simultaneous Measurement Of Gas Temperature And Concentration

The high-temperature gas produced by fossil fuels combustion exist in boilerfurnaces, industrial furnaces, gas turbines, engines and many other energy&power devices.Accurate measurement of gas temperature and species concentration is very important tothe study of combustion mechanism, fuel combustion characteristics, pollutant formationand energy&power devices optimization. In this thesis, the technique for gas temperature,carbon dioxide and water vapor concentration simultaneous measurement, which is basedon multi-band gas radiation modeling, is studied. The details are as follows:Firstly, the multi-band gas radiation modeling method is analyzed basing on therelationship between band emissivity and gas column density. The binary polynomialfitting using for modeling is presented. Then the theory of gas temperature, CO₂and H2Oconcentration simultaneous measurement technique is studied and an inversion algorithmwhich is appropriate to the models is presented. A band selection method which is basedon the band emissivity curve, band emissivity versus temperature curve and the spectralbandwidth of spectrometer is also presented.A simulation study on simultaneous measurement of temperature, CO₂and H2Oconcentration in the flue gas of a laboratory ethylene laminar diffusion flame is carriedout. Firstly, the models are built on2340cm-1,3900cm-1and3740cm-1band respectivelywith the use of HITEMP-2010database. Then the models and inversion algorithm areexamined via several cases. The calculation results of cases in which the gas isisothermal and homogeneous demonstrate that the maximum relative errors oftemperature, CO₂and H₂O concentration are2.44%,1.69%and1.0%when there are noerrors in signals, and6.34%、11.08%、3.80%when there are±5%errors in signals. Thecalculation results of cases in which the gas is non-isothermal and homogeneousdemonstrate that the temperature values are in well agreement with the set values, andthe maximum relative errors of CO₂and H₂O concentration are0.69%and1.38%whenthere are no errors in signals, and-5.24%and2.55%when there are±5%errors insignals. This study demonstrates that the models and inversion algorithm have good accuracy and anti-error capacity, and they also can be used to measure non-isothermaland homogeneous gas.