Characterization of Different Actuator Designs for the Control of the Precessing Vortex Core in a Swirl-Stabilized Combustor

Abstract

The precessing vortex core (PVC) represents a helical-shaped coherent flow structure typically occurring in both reacting and nonreacting swirling flows. Until now, the fundamental impact of the PVC on flame dynamics, thermoacoustic instabilities, and pollutant emissions is still unclear. In order to identify and investigate these mechanisms, the PVC needs to be controlled effectively with a feedback control system. A previous study successfully applied feedback control in a generic swirling jet setup. The next step is to transfer this approach into a swirl-stabilized combustor, which poses big challenges on the actuator and sensor design and placement. In this paper, different actuator designs are investigated with the goal of controlling the PVC dynamics. The actuation strategy aims to force the flow near the origin of the instability—the so-called wavemaker. To monitor the PVC dynamics, arrays of pressure sensors are flush-mounted at the combustor inlet and the combustion chamber walls. The best sensor placement is evaluated with respect to the prediction of the PVC dynamics. Particle image velocimetry (PIV) is used to evaluate the passive impact of the actuator shape on the mean flow field. The performance of each actuator design is evaluated from lock-in experiments showing excellent control authority for two out of seven actuators. All measurements are conducted at isothermal conditions in a prototype of a swirl-stabilized combustor.

Excitation of the precessing vortex core by active flow control to suppress thermoacoustic instabilities in swirl flames

“…The disappearance of the natural PVC peak indicates the lock-in of the reacting flow to the helical actuation. 36…”

“…The parabolic shape of the centerbody provides unaffected PVC dynamics, which is mandatory to obtain adequate experimental conditions. 36

Figure 1.Test rig with measurement technique.

Figure 2.Detailed view of the burner.

…”

“…To quantify the actuator efficiency, it is common to relate the momentum of the actuation jets to the momentum of the main flow, yielding the momentum coefficient 41 where uRMS represents the RMS velocity at the actuator outlet, which was measured with a hot-wire anemometer, 36 G the outlet area of the actuator, A h the hydraulic cross section area, based on the hydraulic diameter of the mixing tube D h , and U 0 the bulk velocity of the main flow. In the present work, the maximum actuation amplitude was set to Cμ={1.19,0.83,0.49}% for the three considered Reynolds numbers, respectively (see Table 1).…”

“…For this purpose, the actuator was implemented in a centerbody upstream the flame and successfully tested at isothermal flow conditions. 36 Finally, the actuator and sensors were advanced to be heat resistant and successfully tested under reacting conditions. 37 With this newly developed technique, the idea is either to excite the PVC through open-loop control at conditions where is naturally suppressed or to damp the instability through closed-loop phasor control where it is naturally occurring.…”

Receptivity of the turbulent precessing vortex core: synchronization experiments and global adjoint linear stability analysis

“…In this work, we address two central questions: Firstly, we investigate where the PVC is most receptive to open-loop forcing within a fully turbulent flow. In contrast to previous work, we focus our attention to the flow region far upstream of the vortex breakdown bubble, which is expected to be most influential for the formation of the PVC, as indicated by previous studies (Kuhn et al 2016;Tammisola & Juniper 2016;Lückoff et al 2018;Kaiser et al 2018). Moreover, we validate the receptivity maps using active flow control, which directly leads to the second, more fundamental questions posed in this work: In what extend can the adjoint global mode based on the mean of a highly turbulent flow be used to guide and interpret flow control.…”