OCR Text |
Show Effect of Probe Intrusiveness and Purge Flow A serious concern associated with probe measurements of any kind is the possibility of probe induced perturbations of the flow field. With the probe described in this paper, we can recognize two main sources of disturbance of the near probe flow field: one is the intrusiveness of the probe itself, the other is the effect of the purge flow. The effect of both factors onto the velocity measurement accuracy at the measurement volume has been assessed. This was achieved by performing the velocity measurements with a non-intrusive LOV probe kept in a fixed location and by positioning the water-cooled probe at various distances from the measurement volume. The experimental set-up is displayed in Figure 4. The furnace model had an inner diameter of 40 em and was equipped with a quart and a swirl generator. The velocity measurements shown in Figures 5 and 6 were performed on the centerline of a non swirling flow and at two quarl diameters downstream of the quarl outlet. non-intrusive LDV optics 0400mm d Sf ~ I m water-eooled probe with purge flow Figure 4: Experimental set-up to test probe intrusiveness Figures 5 and 6 display the mean (U, W) and fluctuating (u',w') axial and tangential velocities measured on the model centerline with the water-cooled probe positioned at various distances from the measurement volume. A sample size of at least 4000 validated signals was acquired at each position in at least 15 seconds in order to minimize the statistical errors. The furnace model had a radius of 20 cm; therefore, the measurements with the water-cooled probe outside the flow or close to the wall can be considered non-intrusive. Figure 5 shows measurements taken without purge flow through the water-cooled probe tip whereas Figure 6 displays measurements taken with a purge flow of 5 m/ s average outlet velocity. This purge flow rate is about twice higher than the flow rate used during in-flame measurements. _ Analysis of Figures 5 and 6 show that a clear deviation from the non-intrusive measurements is only visible when the measurement volume is less than 3 em away from the probe tip. Since the s'mall LOV probe is fitted with front lenses with focal lengths of 102, 135 or 170 mm, which place the measurement volume between 77 and 145 mm from the watercooled probe tip, the effect of the probe intrusiveness on the measurements accuracy can be considered minor. Comparison of the measurements obtained with purge flow through the water-cooled probe (Figure 6) or without purge flow (Figure 5) does not reveal any noticeable effect on the measured velocities. This indicates that the purge flow rate used during actual in-flame measurements has no significant effect on the measurements accuracy. 8r-----~------~------_r----__, 6 ..... .... :. .... ..... ;. .L ..I__~ _~ _' ___~ _ _: _'~ ee e : e : e ee: ee e .: eee.', eee e ... ... ... ........... ...... : ... ... .................. . . ~ ......... ... ......... ... ...... <.. ......................... . o -2 L...-___ .l....-___ ..L...-___ ..L...-__ ---' o 5 10 15 20 Distance between probe tip and measurement volume [cm] Figure 5: Effect of probe intrusiveness on measurement accuracy inlet swirl number = 0, no purge flaw Because flows with high swirl levels are known to be more sensitive to distortion by probes (Bilger, 1976), the measurements were repeated for different inlet swirl levels. Figure 7 displays velocity measurements obtained in a flow with an inlet swirl level of 0.6. The purge flow outlet velocity was maintained at 5 mis, and the measurement volume was positioned at 2 cm from the centerline. This position was selected because the probe disturbance was expected to be more pronounced inside a recirculation zone. The measurements at distances of 23,23.5 and 24 em were actually taken with the water-cooled probe outside the flow and with the access hole closed. Thus, these three measurements give reference velocities obtained in fully non-intrusive conditions. Analysis of Figure 7 show little data scattering when the measurement volume is more than 12 an away from the probe tip, and a clear deviation from the non-intrusive results when the measurement volume is less than 10 em away from the probe tip. This indicates that for the experimental conditions reported here, (measurements in the recirculation zone of a flow with an inlet swirl number of 0.6, water-cooled probe diameter of 63 mm), the probe intrusiveness can be minimized by keeping a distance of at least 100 to 120 mm between the measurement volume and the probe tip. 8~---~------r-------r_----~ lo euu, AWl . ~ w' 6 ............ ...... ...... ... .; ...... .................. ...... ~ ... ......... ......... ... ...... -: ... .... ... .... - - eee e:' e ee .' eeee:e ee e : : e ~ e ' , , e oS 4 e· ······ ·· ;· ·········: ······· · ·· ·;·· · ···· ··· , , , ~ 0 0000000000000000 ~ 2 . ~&A~AAAA~~A~A~~~A~ o .. ~& ... +A;A~ .A.~. A ....... . ~A .A ·A· A. A -2 'A- _____. L..-______. .L...-______. l....-____- --J o 5 10 15 20 Distance between probe tip and measurement volume [cm] Figure 6: Effect of probe tip purge flaw on measurement accuracy inlet swirl number = 0, purge flaw outlet velocity = 5 m/s |