| Popis: |
Lean Direct Injection (LDI) burners are a promising technology aimed at reducing NOx emissions in new generation aeroengines. However, one of the main drawbacks of this technology is the appearance of combustion instabilities at certain operating conditions. In order to investigate these issues, a confined, atmospheric, swirl-stabilized LDI burner has been set up at the Institute CMT-Motores Térmicos. In this configuration, air mass flow, temperature, and fuel mass flow rate, which is controlled by the injection pressure, can be independently modified to reach different combustion states. In this paper, a parametric study of the equivalence ratio (0.3 ≤ Φ ≤ 0.8), air temperature (50, 100, 150 °C) and fuel mass flow rate (200, 250, 300, 335, 370 mg/s) has been performed to assess their influence on the dynamics of the system through the evaluation of the pressure signals inside the chamber. These signals have been acquired with two piezoresistive sensors flush-mounted to the combustor wall at the same axial distance but in opposite sides of the chamber. Large-amplitude unsteady oscillations are detected for some combinations of the variables of interest. Equivalence ratio variations are shown to affect deeply the dynamic features of the pressure signal, obtaining more stable configurations either close to the lean limit (Φ = 0.3) or at the richest condition tested (Φ = 0.8). Mid-range equivalence ratio values (0.4 ≤ Φ ≤ 0.7) are shown to display the most unstable behavior, featuring large pressure oscillations that remain nearly constant over time (quasi-periodic states) or signals that experience sudden variations in their amplitude (intermittent states). Since turbulent swirl-stabilized spray flame combustors may experience complex flow-flame interactions that could lead to nonlinear behavior of these combustion regimes, several signal processing techniques such as three-dimensional phase space reconstruction or recurrence plots have been applied to the experimental data in order to obtain better insight into these highly dynamic features. |
