The present article reports experimental and numerical analyses of the macrostructures featured by a stratified swirling flame for varying stratification ratio (SR). The studies are performed with the Beihang Axial Swirler Independently-Stratified (BASIS) burner, a novel double-swirled full-scale burner developed at Beihang University. Experimentally, it is found that depending on the ratio between the equivalence ratios of the methane-air mixtures from the two swirlers, the flame stabilizes with three different shapes: attached V–flame, attached stratified flame and lifted flame. In order to better understand the mechanisms leading to the three macrostructures, large eddy simulations (LES) simulations are performed via the open source Computational Fluid Dynamics software OpenFOAM using the incompressible solver Reacting Foam. Changing the SR, simulation results show good agreement with experimentally observed time-averaged flame shapes, demonstrating that the incompressible LES are able to fully characterize the different flame behaviours observed in stratified burners. When the LES account for heat loss from walls, they better capture the experimentally observed flame quenching in the outer shear layer. Finally, insights into the flame dynamics are provided by analysing probes located near the two separate streams.

No abstract

No abstract

In this paper, deep learning is involved to comprehend thermoacoustic instability more deeply and achieve early warning more reliably. Flame images and pressure series are acquired in model combustors. A total of seven data domains are obtained by changing the combustor structural parameters. Then, the pre-trained model TIPE (Thermoacoustic Image-Pressure Encoder), containing an image encoder with ResNet architecture and a pressure encoder with Transformer architecture, is trained through the contrastive self-supervised task of aligning the image and pressure signals in the embedding space. Furthermore, transfer learning in thermoacoustic instability prediction is performed based on k-nearest neighbors. Results show that the pre-trained model can better resist the negative effect caused by class imbalance. The weighted F1 score of the pre-trained model is 6.72% and 2.61% larger than supervised models in zero-shot transfer and few-shot transfer, respectively. It is inferred that the more generic features encoded by TIPE result in superior generalization in comparison with traditional supervised methods. Moreover, our proposed method is insensitive to the thresholds of determining thermoacoustic states. Principal component analysis reveals the physical interpretability preliminarily through the connection between feature principal components and pressure fluctuation amplitudes. Finally, the key spatial region of flame images and temporal interval of pressure series are visualized by class activation map and global attention scores.

The emission characteristics of a model centrally staged lean premixed prevaporized (LPP) combustor was investigated under near-critical and supercritical main fuel injections. The Chinese aviation kerosene, RP-3, with its critical temperature and pressure of 651 K and 2.35 MPa, was preheated from 500 to 740 K and pressurized from 2.0 to 3.5 MPa before being injected into the combustor. The combustor liner consists of ceramic matrix composites (CMC), which are installed on a water-cooling frame. Therefore, the combustor features a high dome air ratio (95% of the total air) by removing both primary and dilution holes and redirecting the liner cooling air to the dome. The overall fuel-to-air ratio was varied from 0.030 to 0.053. The emissions at the combustor outlet were measured at various operating conditions in the range of inlet air temperatures from 600 to 840 K and pressures from 2.0 to 2.8 MPa. The results showed that EINOx decreases about 40% as the injection temperature increase from 500 K to 740 K at 2.0 to 2.4 MPa injection pressure. It indicates that the transition from liquid fuel to supercritical fuel drastically reduces fuel density and surface tension. Increasing injection fuel temperature significantly improves the fuel/air mixing and avoids hot spot formation that favors NOx formation. Both EICO and EIUHC decrease slightly with increasing fuel injection temperature, suggesting a weak relation between the combustion efficiency and fuel thermodynamic state. The finding of the current study suggests that the NOx emissions are affected by the premixing quality of the main injector and may be reduced by injecting supercritical kerosene.