Computational and experimental study of geometry modifications inside a flow-blurring injector

Abstract

Liquid fuel atomization is widely used for combustion in many applications. With the strong emphasis on emissions regulations coupled with the ever increasing drive to improve energy efficiency, all aspects of combustion are being thoroughly researched. One key way to achieve the above goals is further improvement in the liquid fuel atomization process. Better atomization improves mixing of fuel and air, thus results in lower emissions, whereas improved liquid fuel injector designs can improve energy efficiency. The flow-blurring (FB) atomization technique, developed recently and investigated at the University of Alabama, has shown promise in both these areas. Previous research has shown that the FB injector produces smaller droplets and a more desirable droplet distribution than the commercial air-blast injector. In addition, the FB injector is able to successfully atomize a wider range of fuels, and it is much less susceptible to the change in surface tension or viscosity of the liquid fuel. In this study, a computational fluid dynamics (CFD) model is created to mimic the mixing of the fuel and air inside the injector, and thus, understand the underlying physics of the FB atomization process. The 2D model is assumed to be asymmetric and incompressible, and it uses the mixture model for the two-phase flow. A transient solution is found and analyzed revealing a recirculation zone, due to a stagnation point near the exit, is formed within the fuel tube of the injector. The recirculation zone is responsible for the mixing of fuel and air and the formation of bubbles. Prior experimental research conclusions are compared with the model as various operating conditions are implored for verification of the models accuracy. Finally, the model is utilized by simulating and studying the effect of geometric modifications within the wall gap of the FB injector. An inner-slant wall gap provides promising results compared to the original geometry. The geometry modifications are then implemented in an actual injector tested in an atmospheric burner. Emissions measurements, thermal imaging of the combustor surface, and OH* chemiluminescence imaging of the flame are used to first verify proper operation of the combustor and then to characterize the flame structure. Several operating conditions are altered and the influence of these changes is studied. OH* chemiluminescence images reveal the flame is stable and a better distribution of OH* signals represents improved atomization. Finally, the geometric modifications to the injector are tested to determine the performance improvements with respect to the baseline design. Experimental results of the different geometries indicate the injector with inner-slant seems to improve the atomization process. The inner-slant injector has lower emissions for a range of ALR values and a lower pressure drop though the injector compared to the original geometry.