In the present thesis, we discuss the statistics and dynamics of premixed flames in turbulent flows. We begin by experimentally studying the dynamics of a premixed CH4-air flame embedded in background turbulence. We discern nonlinear interference effects from convecting disturbances on the flame surface and quantify their impact
on the large-scale harmonic flame response. We also determine the local and global heat release rate response and explain its dependence on the local structure of the flame response. We measure the scale-dependent statistics of the fluctuating flame surface. We show that the large-scale statistics are significantly different from the small-scale statistics. In particular, we estimate the effect of small-scale turbulence on
the statistics of flame fluctuations. We observe that power-spectrum and the moments of the increments of flame fluctuations depict well-defined power-laws, implying selfsimilarity in the fluctuations over an intermediate range of scales not affected by viscosity and the restorative effect of the flame propagating normal to itself, termed as kinematic restoration. Further, we observe that scaling exponents of the structure function are anomalous as they do not grow linearly with the order of the structurefunctions, indicating the presence of small-scale intermittency in the power-law range. The presence of such small-scale intermittency has important implications in the modeling of turbulent premixed flames. We then analytically determine the effect of turbulence on the flame surface. For turbulent premixed flames, the fractal dimension is argued to be D = 7=3 for flamelet combustion and D = 8=3 for thickened flames (Da  O(1)) based on heuristic
scaling arguments. However, such scaling arguments do not consider the effect of the intermittent nature of turbulent kinetic energy dissipation on the flame surface.

We account for the effects of intermittent dissipation on the fractal dimension of thickened flames. Intermittent dissipation leads to variability in the inner cut-off, which then affects the scalar flux and total interface area. We account for these variabilities through two approaches: Coarse-grained approach based on the moments
of the dissipation and fine-scale analysis by adopting the multifractal formalism. We derive two corrections to the upper-limit of fractal dimension – D = 8=3+3=4(1