OCR Text |
Show Tests were conducted ' using aviation fuel JP-5. An industrial swirl-type pressure atomizer (#1.65 Monarch, 80 deg angle) was utilized in all of the experiments. The sprays selected for investigation were studied in their burning forms in the rig shown in Fig.1. The sprays were injected horzontally in the center of a steel combustor model with 140 mm diameter and length 700 mm. The fuel (supplied by China Petroleum Corporation) had the following properties: gravity API at 60 F: 45.4, viscosity at -4 F : 4.8 cSt and net heat of combustion 18,646 Btu/lb. The combustor model was positioned by the 3-D traverse system. Probes access to the combustor model was from the radial direction. In order to extract samples and temperature throughout the point of interest, the combustor model capable of movement in axial and radial directions was required. Thus the combustor model was mounted in the 3-D traverse system. It is essential that the composition of the gas sample should not be changed during its transfer from the sampling point to the gas analysis instruments. This requires that the composition should be "frozen" by rapid cooling within the sampling line and keeping constant temperature to avoid the possible chemical reactions. Some difficulties were encounted in obtained representative gas samples when probing in turbulent, recirculating flows. Isokinetic sampling rela tive to the mean flow velocity was not performed. There exist some significant influences were caused by the velocity fluctuation and flow directions in the recirculation zone. For such conditions it may be impossible to obtain a accurate sample. The compositions of CO,C02,02 and unburned hydrocarbon inside the combustor flame zone were measured using a radial traversing sampling probe coupling to on-line gas instrumentation. The superalloy Inconel gas sampling probe has an outside diameter 5.0 mm. Samples analysis was performed by NDIR gas analyser for CO and C02,paramagnetic for oxygen an Flame ionization for unburned hydrocarbon. Temperature was measured with K-type thermocouple coated ~ith silica. The thermocouple was shielded and supported with a small Inconel tube inserted radially from taps onto the combustor wall. The junction was about 5 mm from the end of the Inconel tube. The time-averaged temperature reading was obtained in period of 3 minutes for a measured point. No corrections were made for conduction and radiation losses. - 2 - |