Innovative nature of the project, originality of the objectives and the methodology
The originality of the project is that it focuses on fundamental issues in the combustion dynamics of sustainable aviation fuels (SAFs) and more specifically on nonlinear mechanisms driving thermo-acoustic instabilities of multi-component fuel systems representative of SAFs. It proposes a combined experimental and computational approach, by combining expertise of CORIA, EM2C and CERFACS, in which the same injection units will be tested in four complementary facilities (see Fig.1) while large eddy simulations, will be carried out by setting a constant dialog between experiments and simulations. The main issues can be summarized as follows:
- Two-phase combustion still needs to examine the dynamics of multi-component liquid fuel and air swirling spray flames. Experiments and simulations of these flames are scarce and difficult. The impact of the droplets comprising several components with different evaporation rates on the interaction between the unsteady flow, combustion dynamics and acoustics is missing.
- The most dangerous instabilities take place in the azimuthal direction with complex interactions between the injector, spray and flame responses to waves. These longitudinal and azimuthal/transverse couplings between the flow dynamics and acoustics were not studied in the case of multi-component fuels.
- Flame and flow interactions take place between neighboring injectors and the response of multiple
This type of methodology will ease the prediction of these undesirable phenomena at the early stage of design of new combustors fed by SAF fuels derived from biomass. This original approach, validated in FASMIC for single-component fuels, will be applied to SAF surrogates with swirling injectors and to droplet spray combustion in the presence of azimuthal (standing or spinning) waves.
Figure 1: The four test facilities that will be used in the project. From left to right: SICCA, MICCA, TICCA (EM2C-CNRS) and TACC (CORIA).
Position in relation to the state of the art
Research in combustion/acoustics instabilities has been quite substantial with notable progress. Nevertheless, azimuthal instabilities have only been studied recently due to their complex interactions with flames. The present consortium has brought crucial elements in the case of the coupling between azimuthal/transverse modes and swirling single-component fuel flames (Fig.2).
Figure 2: (a) Local phase-averaged Rayleigh index R, calculated with the pressure and OH* intensity fluctuations, as a function of the height z above the burner and of the phase angle α over a cycle. R is synchronized with the acoustic pressure cycle reported at the bottom. The experimental conditions in TACC-Spray correspond to those in which instabilities were observed in MICCA-Spray for n-heptane fuel (flame power P = 7.5 kW, forcing frequency f f = 732 Hz and Prrms = 1030 Pa at a pressure antinode of the transverse wave); (b) Space-time evolution of the Rayleigh source term during 3 cycles by LES (P = 6.4 kW, ff =461 Hz and P rrms = 618 Pa) for a longitudinal forced SICCA-Spray configuration with n-heptane.
One of the many results obtained by the present group concerns a phenomenon designated as dynamical blow out, a process induced by large amplitude oscillations. This phenomenon was revealed in experiments carried out in MICCA (an experimental annular combustor designed and operated by EM2C). It was shown by CORIA using their TACC facility that blow out could be achieved by large amplitude transverse modulation of a linear array of injectors of the same type as those used in MICCA (Fig.3).
Figure 3: Dynamical blow-out (DBO) induced by oscillations: (a) Stable operation in MICCA, (b) DBO in MICCA under high amplitude azimuthal oscillations, (c) Stable operation in TACC, (d) DBO in TACC under large amplitude transverse oscillations.
If works on the physico-chemical properties of multi-component surrogates emulating alternative jet fuels can be found, the dynamics of such fuel spray flames and their thermoacoustic instabilities are not well documented. Our own past investigation indicates that the flame dynamics and domains of instability are modified by changing the single-component fuel while global conditions (equivalence ratio, power, swirl number…) are maintained. Studies carried out at EM2C and CORIA with different fuels show that instability characteristics and flame describing functions (FDF) depend on the fuel used. The fuel influences the gain of the flame as it responds to incident perturbation while the phase reflects changes associated with different time lags.
The Euler/Euler approach applied with LES to study thermoacoustic instabilities in real configurations gave satisfactory results when it comes to the main flow properties. Limitations however remain due to the difficulty of representing numerically the complex multi-size droplet clouds. The Lagrangian framework solves the trajectory for each particle with a given initial diameter. The Lagrangian approach has been extensively developed recently in FASMIC proving its capability and advantages in the context of thermoacoustic combustion instabilities coupled with single-component fuels.. The importance of correctly modelling the interaction between liquid particles and internal elements of the combustor to reproduce thermoacoustic oscillations in the SICCA-spray combustor. A sensitivity analysis of the numerical predictions to the spray injection angle was also performed revealing that this parameter is for the SICCA-spray burner a key parameter that controls the thermoacoustic coupling even if it has a small impact on the mean flame in non-oscillating conditions. The modifications imposed to the injected fuel are indeed shown to potentially stabilise the system opening new possibilities and strategies for the control of unwanted instabilities.
Finally, a joint experimental/numerical study has also been performed in collaboration with EM2C to understand flame dynamics under a self-sustained limit cycle. A typical comparison between experimental data and calculations is shown in Fig. 4. In all of these studies, a simplified approach consisting in directly injecting the experimentally measured diameter distribution was used to model the dispersed phase. CERFACS is currently involved in research projects aiming at correctly reproducing the complex atomization process inside SICCA-spray without imposing any experimental droplets diameter PDF.
Note that despite all these efforts and progress around the Lagrangian approach, additional challenges have been identified so the approach is applicable to multi-component SAF fuels.
Figure 4: Experimental (Left: two rows: mean phase averaged heat release rate and droplet spatial distribution) and numerical (Right: two rows, same diagnostics as in the experiment) single component spray and flame dynamics during limit cycle in SICCA- spray combustor. Results from ANR project FASMIC.
i. Classical globally reduced chemistry usually comprising two reactions (the fuel oxidation and the CO- CO2 equilibrium) are clearly insufficient to properly describe complex bi-fuel flames and emitted pollutants. A complete chemical description is available with detailed chemical schemes, but these are not affordable for LES of complex configurations. They can be advantageously replaced by Analytically Reduced Chemistry (ARC) [23] which provides the same accuracy on targeted quantities at a much lower computational cost. However, a novel ARC scheme able to correctly reproduce the flame properties for multi-component SAF surrogate will need to be derived and used in all the numerical simulations proposed in FlySAFe.
ii. Specifying the adequate droplet distribution at injection will have to be ensured. This can be guaranteed thanks to experimental characterizations of the different sprays issued by the chosen liquid fuel injection system / swirler while changing the initial liquid fuel composition and therefore properties. These parameters impact the spray so they will have to be evaluated. Likewise, the capacity of existing modelling to account for such effects will need to be assessed numerically for proper understanding and prior to fully reacting simulations.
One important challenge is to describe the evaporation of a complex multi-component liquid fuel mixture and represent changes in evaporation rates that appear and differentiate one SAF from another. Lighter components are expected to evaporate first leaving the heavier components to evaporate much later. Such an effect can be addressed numerically by introducing the so-called Discrete or Continuous Multi-component evaporation models (noted respectively CME or DME models) as demonstrated in [24] using DME. To conclude and based on the present state of the art resulting from the FASMIC project, FlySAFe offers a unique framework to address SAFs issues experimentally as well as numerically and to examine their impact on the thermoacoustic stability of aeronautical combustors. It furthermore confirms the necessity to fill the lack of knowledge about the combustion dynamics of multi-component mixtures and to investigate the flames submitted to large pressure oscillations as well as their robustness to blow out in an azimuthal/transverse acoustic field.
Haut de page
Les partenaires
