| OCR Text |
Show High temperature suction pyrometer In-flame temperature measurements in air flames have traditionally been performed with suction pyrometers. These devices consist of a Platinum/Rhodium thermocouple protected by an alumina sheath, which is itself placed within one or two concentric shields. Suction velocities above 200 m/s around the sheath and between the shields are applied to maximize convective heat transfer within the pyrometer head and minimize difference between gas and thermocouple temperatures [5]. A well designed and operated suction pyrometer yields a thermocouple temperature less than 50°C lower than the true gas temperature. However, the thermal limitations of alumina and platinum/rhodium thermocouple limit the maximum operating temperature to about 1750°C. The high temperature pyrometer used at Air Liquide was supplied by the IFRF and is capable of operating in flames up to about 2300°C. Its design results from developments at the IFRF prior and throughout the O X Y F L A M research program [9, 11]. This modified pyrometer is still based on a type B thermocouple within an alumina sheath, but the outer ceramic shields have been replaced by one water-cooled steel jacket. The tip of the alumina sheath is retracted from the suction hole to avoid radiation exchange between the thermocouple and the furnace environment. The advantage of this design is to apply onto the sheath and thermocouple a forced radiation cooling depending only on the pyrometer head inner geometry and the mass flow rate aspirated through the probe. Therefore, the thermocouple junction remains several hundred degrees below the true gas temperature. The correction factor that relates the thermocouple temperature to the gas temperature was obtained by calibration with C A R S measurements. Such measurements were carried out by the D L R Insitut fur Physikalische Chemie der Verbrennung (Stuttgart) at the IFRF in 1996 [12], and repeated at higher temperature in 1998. Figure 6 displays an example of radial temperature profiles measured at three locations in the IFRF furnace no2 (the thermocouple temperature is shown uncorrected). It can be noted that at the first traverse (22 c m from a 1.0 M W pipe-in- pipe burner), temperature gradients of several hundreds of degrees per millimeter are measured. These results shows that the new high temperature suction pyrometer is capable of prov iding detailed temperature profiles with high spatial resolution. Gas analysis Using conventional flue gas analysis equipment is generally limited to specific species such as. CO. CO:. NOx. SOx and O;. Extending the range to include other species can be done using a G C or M S (mass spectroscopy) instrument coupled to the extraction probe. However, each of these techniques has drawbacks. G C analysis suffers from slow response time with discrete sampling making time averaged measurements difficult. For M S analysis interpretation of the mass spectrum can be complicated by overlapping masses due to species having similar fragments. Applying in-line FTIR (Fourier Transform Infrared) gas analysis provides an alternative technique to extend the scope of species to monitor. The system prov ides fast response time (2 sec/scan @ 4 cm"1 resolution) and specificity in spectral regions free of interference. A n example FTIR spectrum obtained for an in-flame measurement of 10 averaged spectra at 1 cm"1 resolution is shown in Figure 7. In this case, extractive sampling was conducted with continuous gas flow at 1 L/min through a 10 c m cell (optical path length). The small sample cell and high flow rate provides an additional benefit by time averaging the sample. Before entering the cell, the gas passed through a condenser set at 2°C for removal of water. For the example spectrum shown the major bands are clearly identified and efficient removal of water is indicated by the small absorbance signal observed. The following example demonstrates the use of extractive sampling FTIR measurements in the flame region. In this experiment, the objective was to identify soot precursor formation from the ALGLASS FC™ operating with natural gas and oxygen. Before sampling in the flame, a calibration is conducted to relate absorbance of a species to concentration. These calibrations are performed using a standard with dilution at various levels for construction of a calibration curve as shown in Figure 7 for acetylene. From the acetylene calibration, a detection limit of 1000 p p m is obtainable for 10 averaged spectra and the small sample cell. For the other hydrocarbon species observed, detection limits of 200 p p m and 400 p p m were obtained for C H 4 and C^rL respectively. Detection limits can easily by improved by increasing the sample cell pathlength, but this will also increase the residence time of the gas in the cell. By probing the flame at different locations, a m a p of the acetylene concentration was obtained, as shown in Figure 8. This burner uses separate injection of natural gas and oxygen allowing N G pyrolysis to occur forming soot precursors. The increased concentration of soot precursors results in higher soot volume fractions, as will be discussed later, producing 4 |
