OCR Text |
Show 1. At the higher temperature levels, there is typically no flame visible to the naked eye beyond the pilot flame.The main reaction zone is quite invisible against the hot refractory. This is no doubt a result of, (i) maximum temperature levels being relatively low, due to the extensive admixture of both fuel and air, before they meet each other, with recirculating product gases that have lost enthalpy to furnace heat transfer, and (ii) the reaction zone being spread over a larger volume than usual. Possibly the u.v. flame detectors also have some difficulty seeing a flame, the u.v. emission likely being feebler and more diffused than with conventional burners. There were many occasions when the control system suddenly shut the burners off even though combustion had been extremely steady and no visible failure had occurred. 2. It is quite possible that the sightlines provided for the detectors in the burners tested were not optimal.. 7.5. Other results Copious data were routinely gathered, through the data acquisition system, on heat transfer under the various conditions employed in the trials. The measurements include sink heat flux densities and refractory surface temperature distributions, all of which are available to be employed, for example, in validating the modelling of radiative transfer in furnace cavities. 8. Discussion 8.1. Combustion aerodynamics: the strong-jet/weak-jet model The behaviour of the CGRI burner combustion field or "flame", as with most non-premix burners, is clearly strongly shaped by turbulent jet flow and mixing. It should be noted that it is a little hard, here, to speak of flame in the traditional sense when often none is visible, in which case the bounds of the field can be found only by probing. W e will, for simplicity, nevertheless continue to sometimes speak of this field as the flame (as in flame length and flame volume).. The main features of the flow and mixing are well captured by a simple model that focusses initial attention upon the minimum realization of the situation. The C G R I burners used in the present work are all with N =1 (seven fuel ports and seven air ports, all in a ring). In the elementary model, w e consider the practically least interesting but theoretically very illuminating case N = \. The analysis is developed in detail in a companion paper (Grandmaison, Yimer, Sobiesiak and Becker 1996) where it is called the strong-jet/weak-jet model. The situation posed is illustrated in Fig. 14. The burner wall is imagined to be virtually infinite in extent. The strong jet, which is of air in the present context, issues from a port whose axis is normal to the wall, so &i = 0. The weak jet, of fuel gas, issues from a port whose axis is at angle 0\ to the normal and coplanar with the air-port axis. The angle between the port axes is in general denoted by J3\2, but here P\2 = 0\. The distance between the ports at centre of exit is d\2. The air jet is relatively "strong", and the fuel jet "weak", because of combustion stoichiometry. So long as the "burner" is operated around stoichiometric or with excess air, the mass flux of air greatly exceeds that of fuel, m2 » mx (the notation, with subscripts 1 and 2 for fuel and air, instead of/and a, is that of the cited paper). However, in typical C G R I burner operation, the port velocities U\ = GJm and U2 = G2/m2 are of similar magnitude. Thus the air momentum flux, G2 , is typically large 14 |