Nowadays computational fluid dynamics (CFD), due to the ever rising computing power, can be considered both a fundamental tool in basic research and a useful industrial design tool. Indeed the development cost of a combustion device can be sensibly lowered by means of numerical simulations instead of expensive experiments and tests. On the other hand, numerical simulations can help the physical investigation of combustion phenomena. Major difficulties in understanding premixed turbulent combustion are associated with turbulent burning being a highly nonlinear and multi-scale phenomenon, which locally involves various processes, such as chemical reactions, heat release and thermal expansion, molecular transport, and turbulence.
In this context, a relatively unexplored subject that has found
a growing interest is the role played by intrinsic flame instabilities, as Darrieus-Landau (DL), in the turbulent propagation of a premixed flame. Intrinsic instabilities can largely modify laminar flame propagation and can affect the interaction between turbulence and flame structure. In fact, the DL mechanism appears to play a substantial role in premixed turbulent combustion by controlling the growth of unburned mixture fingers and therefore causing oscillations of the flame surface area, turbulent burning velocity, and mean flame brush thickness. This research project focalizes on such interplay, aiming to characterize all of the main features of turbulent premixed flame propagation whenever, under equivalent laminar flow, the operative conditions would favor the onset of instabilities. As a result, reliable predictions of the burning speed based on the turbulent and thermochemical conditions ahead of the flame has been a central problem in combustion science. Turbulent premixed combustion can be studied using direct numerical simulations, taking advantage of present High Performance Computer (HPC) resources using and modifying an highly scalable open source code.
The intrinsic flame instability in premixed turbulent combustion is a relatively unexplored field of study [39, 40]. Intrinsic instabilities can largely modify laminar flame propagation and can possibly greatly affect the interaction between turbulence and flame structure in the context of turbulent propagation. Certain experimental data [13] are often claimed to show an important role played by the DL instability in premixed turbulent burning. We note that the exact domain of influence of these instabilities in turbulent premixed combustion remains unknown [40] and is certainly a matter of investigation. Recently, Creta et al. [22] performed 2D simulations of the slot Bunsen in weakly turbulent flow field. In a mildly turbulent setting, a dramatic morphological dichotomy between stable and unstable flames is observed, which suggests that the statistical properties of curvature can act as an unambiguous marker for DL effects. In particular, the skewness of curvature probability density function, measuring its asymmetry about the mean, is an observable expected to act as such marker. These findings [22] suggest that DL effects on the turbulent propagation of a premixed flame are clearly evident and noticeable albeit strictly confined to a low turbulence intensity regime. The study of the coupling effects between the two known kinds of instability, namely the thermo-diffusive (discernible when dealing with Lewis numbers smaller than unity) and the hydrodynamic instability (due to the thermal expansion) on the flame propagation can be used as a important new results to calculate the turbulent flame speed. The novelties of this research project are the development of the DNS database with both simplified and detailed chemistry in 2D and 3D. The database can be used to generate a closure a-priori models, inclusive of intrinsic instability effects, to filtered simulations as RANS and LES which are a practical tools for a variety of combustion problems of technical interest. To the best of our knowledge intrinsic flame instability effects haven't been, consistently involved in combustion models for LES/ RANS approaches.
References
[1] P.F. Fischer et al. http://nek5000.mcs.anl.gov, 2008.
[2] S.G. Kerkemeier et al. Direct numerical simulation of the autoignition of a ¿hydrogen plume in a turbulent coflow of hot air. J. of Fluid Mechanics, 720:424-456, 2013.
[3] P.E. Lapenna, R. Lamioni, G. Troiani, and F. Creta. Darrieus-landau induced regime of propagation of turbulent premixed flames in bunsen configurations. XXXIX Meeting of the Italian Section of the Comb. Inst. 2016.
[4] P.E. Lapenna, R. Lamioni, and F. Creta. Lng/lox mixing and combustion under supercritical conditions. Space Propulsion Conference, 2016.
[5] P.E. Lapenna F. Creta.Mixing under transcritical conditions: an a-priori study using direct numerical simulation. J. of Supercritical Fluids, 2017.
[6] G. Darrieus. La Technique Moderne,1945.
[7] L. Landau. Acta Physics and Chemistry, 77-85, 1944.
[8] G. Markstein. Mac-Millan, 1964.
[9] H. Guenoche, M. Jouy. Symp (Int) on Comb. Inst. 4:403-407,1952.
[10] R. Petersen et al. Physics of Fluids, 4:456,1961.
[11] C. Clanet et al. Physics Review Letter, 80:3867, 1998.
[12] Y. A. Gostintsev et al. Combustion, Explosion and Shock Waves, 24:563-569, 1988.
[13] T. Wu et al. Proceeding Combustion Institute, 34:937, 2013.
[14] H.Kobayashi et al.XXVI Symp (Intl) on Comb. Inst. 89-396,1996.
[15] R. N. Paul N. C. Bray. XXVI Symp. (Intl) on Comb. Inst. 259-266,1996.
[16] H. Kobayashi et al. XXVII Symp. (Intl) on Comb. Inst. 941-948,1998.
[17] H. Kobayashi H. Kawazoe. Proceeding Combustion Institute, 28:375-381, 2000.
[18] A. S. Savarianandam et al. Combustion and Flames, 146:1-18, 2006.
[19] M. Matalon, F. Creta. Comptes Rendus-Mecanique, 340:845-858, 2012.
(Continue in next section)