| OCR Text |
Show Paper presented at the 6th international symposium on Applications of laser techniques to Fluid Mechanics. Lisbon 1992 LASER VELOCIMETRY IN SEMI-INDUSTRIAL NATURAL GAS, OIL AND COAL FLAMES BY MEANS OF A WATER-COOLED LDV PROBE Jacques Ougue and Roman Weber International Flame Research Foundation PO Box 10000 1970 CA IJmuiden The Netherlands ABSTRACT A water-cooled jacket enclosing a small fiber-optics LOA probe has been designed to perform laser velocimetry measurements in semi-industrial scale flames. The paper presents details of the probe design for front optics cooling and prevention of window contamination. Experiments in natural gas, heavy fuel oil and coal flames have demonstrated fast and reliable operation. Laser velocimetry measurements are performed with the same speed as standard suction pyrometry. Operation for long periods in high temperature flames with high particle loading showed no heating of the LOV optics and no fouling of the front window. The effects of the probe intrusiveness and probe purge flow on the measurements accuracy were assessed and shown to be minor, provided that the measurement volume is located at least one and a half probe diameter from the probe tip. INTRODUCTION For the last thirty years, the measurement techniques used in semi-industrial scale flames (thermal input of 1 MW or higher) have mainly consisted of sampling techniques for temperature, gas composition and heat transfer. Although velocity and turbulence measurements are essential to understand the effect of burner aerodynamics en the flame chemistry and pollutant formation, characterization of flame aerodynamics has remained difficult at scales above a few hundred kilowatts. The high flame temperature, turbulence level and (in coal and oil flames) particle loading have been the main causes for the scarcity of detailed in-flame velocity and turbulence measurements. Over the last decade, however, advances in laser velocimetry and signal processing technology have enabled applications to larger combusting flows [I-8]. These experiments have always been performed with a laser velocimeter operating in the back-scatter mode and with a long focal length so that the LOY optics remained outside the hot flow. One of the unavoidable limitations of this approach, however, is the decrease in signal quality associated with the beam steering fluctuations which are induced by temperature gradients. The beam steering fluctuations increase with the distance travelled by the laser beams in the hot flow until crossing at the measurement volume. The beam steering fluctuations eventually put a limit on the maximum furnace size in which non-intrusive measurements are practical. Figure 1 displays a typical example of the sharp decrease of validated data rates as the measurement volume penetrates farther in the furnace. This data from Ougue and Abbott (1989) was generated using a back-scatter LOY equipped with a 1.5 m focal length front lens. Signal processing was performed with an FFf processor (Oantec BSA). The measurements were taken in a 2.5 MW natural gas flame seeded with zirconium oxidepowder. The furnace had an internal diameter of 2 m and a wall thickness of 50 em. At 20 an from the furnace wall, the data rate was of the order of 100 Hz. Beyond 50 em from the furnace wall, the data rate dropped below 5 Hz. Measurements taken in a 2.5 MW coal flame showed similar data rates as obtained in the natural gas flame. The typical example of Figure 1 shows the inherent limitation of non-intrusive velocimeters where the optics remains outside the flow and the beams travel a long distance in a hot flow. Experience shows that this configuration is still suitable for measurements in the center of a 0.5 m diameter combustion chamber, but is not satisfactory for fast, routine measurements in a 2 m internal diameter furnace. 1000 ~----------------------------~ (~O ° , Q) ~ 100 ~ 1 o ° '0 \ ° I 20 ° backscatter LOA focal length 1.5 m 2.5 MW gas flame 000 °0 °0 °0 I I 40 60 ° .° 0 ,...,. 80 100 Distance from furnace wall [cm] Figure 1: Dntn rate versus mensurement volume positioll ill [unlnce The introduction in 1986-1987 of laser velocimeters with fiber links provided new solutions for measurements in large scale combusting flows in industrial environments. Most et a1. (1989,1990) used a 030 mm LOY probe enclosed in a 3 m long, o 70 mm water-cooled jacket to perform measurements in a 3 MW pulverized coal flame and in an industrial cement precalciner. The LOY probe front end was thermally protected by two quartz windows separated by a film of water. The probe focal length of 16 em resulted in a measurement volume position 6 to 12 an away from the probe tip, depending on the tip employed. Ereaut and Gover (1991) used a 025 mm probe enclosed in a 70 mm outside diameter, 5.2 m long water-cooled jacket to perform measurements in a coal-fired power station. The probe front end was air cooled with a flow introduced at pressures of 0.21 to 1.4 barg, depending on the cooling needs. The air flow was also used to prevent window contamination. |
