Experimental and Computational Study of Nonreacting Vortex Breakdown in a Swirl-Stabilized Combustor

Abstract

Nonreacting flow experiments are conducted in a swirl-stabilized combustor with several configurations of a triple annular research swirler fuel injector. Particle imaging velocimetry is used to measure mean axial, radial, and tangential distribution of the velocity field, from which the swirl ratio and the position of vortex breakdown are calculated for each injector configuration. Numerical simulations based on the Reynolds-averaged Navier-Stokes model equations of nonreacting flows provide insight into the characteristics of the axisymmetric vortex breakdown phenomenon in a circular, finite-length pipe with a sudden expansion downstream of a concentric, circular inlet pipe. Results show that flow states with vortex breakdown can be simulated numerically. Good agreement is found between the results of the simulations and available theoretical predictions for the first appearance of breakdown downstream of the pipe expansion plane, as well as its first appearance in the inlet pipe section as the inlet swirl level is increased. Good agreement is also shown between the measured experimental data, simulations, and theoretical predictions. It is demonstrated that the theoretical criteria and numerical simulations can be used to predict the appearance and location of vortex breakdown for several different nozzles of varying swirl numbers relevant to the stability of lean, premixed combustion.

Wall-separation and vortex-breakdown zones in a solid-body rotation flow in a rotating finite-length straight circular pipe

“…for all time t 0, ψ x (L, y, t) = 0 for 0 y 1 2 . This condition is supported by the experimental data shown in Umeh et al (2010), where the mean swirling flow has established a parallel flow state after several radii from the inlet. This is also supported by the simulations of Meliga & Gallaire (2011).…”

“…We note that the experimental studies of Sarpkaya (1971Sarpkaya ( , 1974Sarpkaya ( , 1995, Garg & Leibovich (1979), Brucker & Althaus (1995), Mattner et al (2002) and Umeh et al (2010) used guiding vanes to generate an inlet circumferential velocity that is represented by the Q-vortex model with either small or large core sizes. In the limit, and through a careful fine design of the turning vanes, the inlet circumferential velocity may approach the solid-body rotation flow (which is a special case of a vortex flow with a core size that is the same as the pipe radius).…”

An active feedback flow control theory of the axisymmetric vortex breakdown process

“…This condition is relevant for a sufficiently long pipe (where L 1 but L is less than the entrance length for a fully developed flow). The experimental results of Liang & Maxworthy (2005) and Umeh et al (2010) also support these outlet conditions at distances greater than six times the pipe radius. Other boundary conditions at the outlet may also be assumed to reflect various devices at the outlet, such as contracting nozzles or expanding diffusers.…”

“…The case of a non-zero ψ xx (0, y, t) was studied in Rusak (1998). The experimental results in figures 8-19 of Liang & Maxworthy (2005) and figures 18 and 19 of Umeh et al (2010) support these inlet conditions as long as swirl level, ω, is up to 50 % above critical, ω 1 . Other inlet conditions may be considered to reflect special vortex generators such as a honeycomb system just in front of the pipe (see Leclaire, Sipp & Jacquin 2007;Leclaire & Sipp 2010).…”

“…In this process, an axisymmetric decelerating-flow wave grows in size and is accompanied by the growth of a spiral wave along the outer boundary of the breakdown zone, the appearance of significant radial velocity at and near the inlet of the test chamber and flow radial expansion around a large breakdown zone. Figures 18 and 19 in Umeh et al (2010) showed at Re ∼ 40 000 the significant radial expansion of the swirling flow when swirl is above critical, the formation of a large open breakdown (near-stagnation) zone and a parallel flow state at distances of six radii and above from the inlet. In addition, Sarpkaya (1995) and Novak & Sarpkaya (2000) studied vortex breakdown in high-Re flows (Re > 50 000) in pipes.…”