Molecular Transport Effects of Hydrocarbon Addition on Turbulent Hydrogen Flame Propagation

We analytically investigated the influence of light hydrocarbons on turbulent premixed H2/air atmospheric flames under lean conditions, in view of safe handling of H2 systems, applications in H2 powered IC engines and gas turbines, and also with an orientation towards modelling of H2 combustion. For this purpose, an algebraic flame surface wrinkling model included with pressure and fuel type effects is used. The model predictions of turbulent premixed flames are compared with the set of corresponding experimental data of Kido et al. (Kido, Nakahara et al. 2002). These expanding spherical flame data include H2–air mixtures doped with CH4 and C3H8, while the overall equivalence ratio of all the fuel/air mixtures is fixed at 0.8 for constant unstretched laminar flame speed of 25 cm/s, by varying N2 composition. The model predictions show that there is little variation in turbulent flame speed ST for C3H8 additions up to 20-vol%. However, for 50 vol% doping, flame speed decreases by as much as 30 % from 250 cm/s that of pure H2–air mixtures for turbulence intensity of 200 cm/s. With respect to CH4, for 50 vol% doping, ST reduces by only 6 %, cf. pure H2/air mixture. In the first instance, the substantial decrease of ST with C3H8 addition may be attributed to the increase in the Lewis number of the dual-fuel mixture and proportional restriction of molecular mobility of H2. That is, this decrease in flame speed can be explained using the concept of leading edges of the turbulent flame brush (Lipatnikov and Chomiak 2005). As these leading edges have mostly positive curvature (convex to the unburned side), preferential-diffusive-thermal instabilities cause recognizable impact on flame speed at higher levels of turbulence, with the effect being very strong for lean H2 mixtures. The lighter hydrocarbon substitutions tend to suppress the leading flame edges and possibly transition to detonation in confined structures and promote flame front stability of lean turbulent premixed flames. Thus, there is a necessity to develop a predictive reaction model to quantitatively show the strong influence of molecular transport coefficients on ST.