OCR Text |
Show 8, gives a very tight array. If the ports are spaced too closely, the jets from them merge too quickly, and the burner performance in respect to N O x emissions is likely to deteriorate. It may be noted that the X B M 1 , X B M 2 and X B M 3 only provide a single air port angle, O2 = 10°. The X B M 4 allows variation of the air port angle from 0° to 20°. 3. The CAGCT research furnace The CAGCT research furnace is described in detail in a CAGCT technical memorandum ( C A G C T 1993). The salient features are summarized in Table 2. Figure 2 gives an impression of the furnace body and of the locations of the burners, the recuperators, the probe ports and the sight ports. The manner of exhausting gas from the combustion chamber, through a uniform matrix of holes in the rear wall, provides a well-defined, simple exit boundary condition, kind to mathematical modelling, effectively approximating Vi'pU] - constant over that wall. The furnace shell is made of 6.35 m m thick steel plates bolted to a frame of I-beams. The joints everywhere are caulked and/or gasketed. Tests of gas-tightness are periodically conducted in which the burners are run slightly sub-stoichiometric and the furnace is put under strong negative draft. The appearance and level of O2 in the exhaust gases (together with significant C O ) are then indicative of the presence and amount of infiltration. W h e n significant leakage is detected, the problem is diagnosed and remedied. During normal burner/furnace trials when total absence of infiltration is not critical (as in heat transfer experiments), the furnace is operated under slightly negative draft (around -10 Pa time-average). However, when it is desirable to know the excess air arrival and its source very exactly, as in the present trials, with their emphasis on pollutant emissions and operation over a range of excess air levels including the very low, the furnace is operated under slightly positive draft (10 to 20 Pa average, with no negative-going fluctuations) to completely exclude infiltration. The furnace is designed to present typical features of industrial process furnaces: • Operation with side-mounted burners or roof burners • Operation with a single burner or with arrays of burners, involving burner interactions • Complex furnace aerodynamics involving jets, recirculation and cross-flows • Realistic regimes of turbulent flow, buoyancy, radiative and convective heat transfer, and chemical reaction • A wide range of furnace loads, varying in both magnitude and disposition The ftirnace is thus particularly suited for generating experimental data to guide and validate mathematical modelling. If a model works well for this furnace, then there is good reason to believe that it should provide realistic predictions for a variety of industrial furnaces that operate in similar domains. In general configuration, the furnace directly resembles typical large steel reheating furnaces, the chamber space and sink associated with each burner being similar in shape and around one-half of full scale. 4 |