OCR Text |
Show Ereaut and Gover report that the probe cooling was satisfactory for periods of around 10 minutes in the most intense part of the flame, and that the front window required cleaning after periods of 20 minutes, although this depended considerably on the probe location. This paper presents the design of a water-cooled probe developed at the IFRF to permit LOV measurements in semiindustrial test furnaces. The objectives of this paper are: (1) - to present the details of the probe design and to address the questions of LOV probe cooling and prevention of window contamination, (2) - to discuss the effect of probe intrusiveness and purge flow on the measurement accuracy, (3) - to present LOV measurements performed in natural gas, coal and oil flames with thermal inputs from 2 to 12 MW in test furnaces with diameters of 2 and 3.5 m, (4) - to assess the validity of a particle size discrimination method based on signal amplitude. EXPERIMENTAL The laser velocimeter is a two-color, back-scatter, fiber-optics system from TSI. The LOV operates with a 4 Watt Argon laser. The 20 meter long fiber optics cable enables the laser, most of the LOV optics, and the signal processor to be kept in a control room next to the furnace. Therefore, only the 25 mm diameter LOV probe needs to be protected from the harsh flame and furnace environment. In flame access is provided by enclosing the small LOV transmitting and receiving optics in a 63 mm outside diameter water-cooled jacket, as shown in Figure 2. The transmitting optics features a 20 mm diameter front lens with a 135 mm focal length which places the measurement volume at 110 mm from the probe tip. Signal processing is performed with a DSA, a frequency domain fast Fourier transform (FFT) processor from Aerometrics. The OSA has a maximum selectable bandwidth of SO MHz and a 40 MHz frequency shifting for measuring flow reversals This frequency range, together with a transfer function of 4.64 ms-1/MHz, provides a maximum measurable velocity range of -92.S to 27S.4 mis, far in excess of the velocities measured in the near field of a swirl burner. Probe Design Features The 3.6 m long water- cooled probe has an outer diameter of 63 mm and provides a 26 mm diameter housing for introducing the LOV optics (025 mm). As can be seen in Figures 2 and 3, a removable threaded tip enables the introduction of two quartz windows and one rubber seal from the probe front side. A holder maintains a 2 mm spacing and allows a small water flow between the two windows. After positioning the inside parts, the threaded tip is mounted and presses against the two windows, separator and rubber seal which seals the probe front end. The probe tip also provides a small annula~ ring for radial purge flow injection. 77 to 145 mm primary water-cooling threaded tip nitrogen purge quartz windows Figure 2: LDV water-cooled probe Probe Cooling The probe cooling consists of two independent cooling circuits. The primary circuit provides cooling to the probe body and is identical to other cooling jackets used in IFRF probes for temperature or gas composition measurements. Unfiltered water is injected at a pressure of 3 to 6 barg with a flow rate of roughly 2 m3 /h. A water pressure of 5 barg was suitable for cooling a 5 m long probe in flows with peak temperatures around IS00°C. The secondary cooling circuit which protects the LOA front optics from the incident flame radiation consists of a film of drinking water flowing between two quartz windows (see Figures 2 and 3). The 2 mm thick windows are made of sapphire glass, which was chosen for its very good mechanical, optical and thermal properties. The water film between the windows has a thickness of 2 mm and flows over the whole window surface with a flow rate of about one liter per minute. Experiments in high temperature, very luminous coal flames has shown that this water-cooled probe protects the LOV optics satisfactorily. No heating of the LOV optics was observed, even after hours of measurements in flames with radiative heat fluxes as high as 300 kW 1m2. LDV probe quartz ~..J,.';"-_~~_I---~ ___ "";,-=,,water flow J purge flow Figure 3: Modelling of purge flaw inside probe tip Probe Purging Because of the presence of seeding or fuel particles in the flow, the probe tip must incorporate a purging system to prevent contamination of the probe front window. An important design constraint is to keep the probe tip short in order to position the measurement volume as far away of the probe as possible to minimize the probe intrusiveness. An initial design based on radial air injection in a conical tip proved unsuccessful because it produced a non uniform flow with local suction at the probe outlet. This recirculation provided a mechanism to suck particles inside the probe tip and resulted in a fast contamination of the probe optics. Understanding of the problem was gained by using a computational fluid dynamics package (FLUENT) to model the flow inside the probe tip and improve the tip design. Figure 3 displays the final tip geometry together with the predicted velocity vectors. It can be seen that the curvature of inside shape helps avoid separation of the flow from the walls and leads to a uniform flow profile at the tip outlet cross-section. Operation of the probe in very high particle loading coal and heavy fuel oil flames showed that this final purging design prevents window contamination effectively. The purge flow outlet velocity is of the order of 2 to 3 m/s. Typically, the probe can be used for mapping the velocity profiles of a complete 2.5 MW coal flames (5 or 6 traverses of 25 points) without requiring periodic cleaning of the front window. The purge fluid can be compressed air or nitrogen, although compressed air is not recommended since it often contains oil which can contaminate the probe optics. |