In turbulent premixed combustion, local turbulent burning velocities are dependent on the structure and wrinkling of flames, which are affected by various parameters including local turbulence levels, shear layer geometry, flame attachment locations, and flame-flame interactions [1]. The impact of these parameters is not universal–their effects on flame structure and wrinkling are different for different flame configurations. Several studies have experimentally investigated the impact of turbulence on flame structure and wrinkling [2-16] and have provided a foundational understanding of the behavior of flame structure and its impact on burning velocity and flame propagation. While these studies provide insightful information, the results obtained in these studies are limited to single-flame configurations. Present day combustion devices, such as jet engine combustors and augmentors, power generation gas turbines, and industrial boilers and furnaces, utilize multiple adjacent turbulent flames. In these devices, the interaction of individual flow fields and scalar fields from multiple flames changes the flame dynamics. Additionally, within the interaction zones of these flames, the flame wrinkling process is affected by changes in the local turbulence levels and flame structure.

Capturing the behavior of flame-flame interactions is a critical first step in the development of sub-grid models for some common large eddy simulations (LES). Many of these sub-grid scale turbulent combustion models use flame surface density (FSD) transport equations to model the flame behavior [17-23]. Appropriate prediction of flame surface generation and destruction is important to these models as the turbulent flame speed is known to directly correlate with FSD [4, 18]. Flame surface destruction is known to occur due to flame surface quenching and mutual flame-flame interactions in turbulent flames [18]. Several models have been proposed in the past to account for these destruction mechanisms [17, 19], although very few experimental studies have been performed that can be used to validate these models, including work done by Skiba et al.[24] and Worth and Dawson [25]. These flame-flame interaction dynamics are particularly important to capture in V-flames as V-flames have been shown to have significant wrinkling as a result of the mean shear generated by the separating boundary layer from the bluff body. Studies of turbulent V-flames by Kheirkhah and Gülder [8-10, 26] have shown that topology of the flame front significantly varies with changes in the inlet turbulence intensities. Formations of flame cusps are enhanced as the inlet turbulence levels increase to moderate levels. Additionally, freely propagating flames, flame pockets and localized extinctions are also reported to be observed in V-flames at high inlet turbulence levels [9]. In one of their studies, Kheirkhah and Gülder showed that flame front velocity statistics vary as a function of streamwise location from the flame holder location [26]. Their results showed that the transverse component of the reactants velocity in the vicinity of the flame front drives the fluctuation of the flame front and the flame front velocity. This mechanism also increases the flame front velocity as a function of downstream distance and at larger downstream distances, the flame front velocity is shown to be much greater than the corresponding laminar flame speed. These studies show that wrinkling is enhanced in V-flames at larger downstream distances and results in greater flame front propagation speeds.