Experimental determination of stirring factors generated by straight and swirling jets in isothermal combustion-chamber models

Abstract

Stirring factors generated by three different burners have been measured in isothermal models using the “cut-off tracer” technique in conjunction with internal sampling in the model. The three burners produced jets with swirl levels varying from zero (a straight jet) to a high level. The results based on about 2500 measurements showed that, in all cases, a clear physical demarkation could be identified inside the model separating a “stirred” region at the burner end of the model from a plug-flow region further downstream. The plug-flow region extended to the exit, with some modification due to exit-corner recirculation. The “stirred” region, however, showed up as one of two extreme types of backmix flow, or a mixture of both, interpreted in terms of a concept of “homogeneous” or “heterogeneous” backmix. An example of the latter is the classic corner-recirculation flow generated by an unswirled jet, with major differences in concentration and direction of flow across a radius. Homogeneous backmix resulted in a well-stirred reactor condition inside which was identified a measurable mixing-delay time that was constant throughout the well-stirred region. Using this in conjunction with the stirred-reactor time measured for the region, a stirring factor was calculated. The stirring factor could only be defined for highly swirled systems. For the three systems tested, the measured stirring factors were: 0.815 for a vane-swirled jet; 0.615 for a tangential entry jet; and 0.675 as the maximum for a countervortex system measured previously (Ref. 3). The mixing delay time is interpreted as being the consequence of finite time for travel in the backmix streams, or the difference in forward bulk and backmix velocity.