Plasma-assisted combustion (PAC) technologies are being explored to address various challenging combustion regimes, including lean flammability limit extension [1–3], suppression of dynamic combustion instabilities [4–6], and acceleration of ignition for low-residence time combustion [7, 8]. In many applications, and especially as the technology progresses towards implementation in practical systems, it becomes essential to consider complex combustion environments; this can include turbulence, multi-scale structures, inhomogeneous gases, two-phase flow, propagating flame fronts, etc. It is therefore necessary to understand the two-way interaction of plasmas and flames during combustion transients and when the location of the plasma relative to the flame can vary with time [9–11].

A common experimental platform to study some of these dynamic environments uses a quartz reactor, for optical access, with electrodes attached to the dry-side of the reaction cell (external surface, not in contact with the process gas), so that the discharge actuates in a Dielectric Barrier Discharge (DBD) configuration. Although initially used for its uniform properties to validate chemical kinetic mechanisms [12, 13], the DBD discharge often appears in microdischarge regimes. A similar setup has also been used to study the thermal-chemical instability controlling the transition from a stable homogeneous discharge to unstable microdischarges in different mixtures [14–16]; as well as plasma regime transitions, between uniform and microdischarge modes, during passage of a transient laminar flame [17]. Comparable arrangements have also been used to study plasma-assisted detonation [18, 19]. In all of these works, imaging and optical diagnostics of discharge and combustion is performed through the narrow side of the discharge cell, with the line-of-sight contained in a plane parallel to the flat electrodes. This view-angle hinders the quantification of several properties of plasma and combustion, including: the volumetric density of microdischarges present in the cell, as well as the 2D or 3D plasma and combustion dynamics beyond a spatially-integrated 1D representation.

The objective of this brief research report is to showcase a novel imaging technique that allows for the direct observation of 2D interactions between a propagating flame front and microdischarges triggered by Nanosecond Repetitively Pulsed (NRP) dielectric barrier discharges (DBD). The imaging is enabled by the use of transparent electrodes arranged orthogonal to the flame front plane, with a camera mounted to view along the axis of the applied electric field. Several complex interactions between the flame and plasma are revealed that were hidden in prior studies visualizing the interaction from the side. Although the use of transparent electrodes has been common in DBD research, particularly to study DBD uniformity and patterning [20–22], when used in a system combining a DBD with a flame front they allow for studying both the structure of the DBD and the flame, and how they interact with one another.

A schematic of the experimental setup is shown in Figure 1. It consists of a tapered quartz burner fed with premixed methane-air mixtures that has been previously described in [17]. The mixture enters at the narrow end of the channel at below the flame speed and is ignited at the wide end via a conventional spark plug, causing a flame to flash back through the channel before being quenched by the narrowing walls. The reactor is open to atmospheric pressure at the wide end and its dimensions are given in Figure 1. In the absence of plasma, the flame has a one or two-lobe structure, in the x-y plane, as a consequence of the finite channel dimensions and the development of fluid instabilities; this is thought to be a tulip-flame structure [23, 24]. Electrodes are mounted on the dry walls of the channel (external surface, not in contact with the process gas, Figure 1) so as to apply an electric field in the z-direction and a NRP DBD is triggered as the flame approaches the discharge volume. The ground electrode is an aluminum plate that spans the length of the reactor. The powered electrode is a glass slide of 1 mm thickness coated on the external side with a thin film of Indium Tin Oxide (ITO). High voltage is applied using a nanosecond pulse generator (Transient Plasma Systems SSPG-20X-HP1). The ITO electrode is transparent, allowing the camera to be positioned to view along the z-direction (see Figure 1) which is aligned with the dominant electric fields and the axis of any microdischarges forming.

Experimental parameters are chosen as follows. The applied voltage pulses have an applied peak voltage measured at the load of 22.5 kV and are repeated at frequencies up to 8 kHz. Trains of pulses are applied continuously with the start of the pulse-train synchronized with the triggering of the flame front, at variable delays. Premixed methane/air mixtures at an equivalence ratio of 1.06 (close to stoichiometric) are used, with a total flow rate of 400sccm.

Imaging is performed using high-speed videography. Two cameras are used: a Photron Fastcam Mini AX100 is used for the colour images, Figure 2, and an Edgertronic SC2+ is used for the monochrome images used to quantify flame and microdischarge dynamics, Figures 4, 6. Color images allow for visual separation of the flame (blue) and discharge (violet) to distinguish the different structures which may be overlapping. Monochrome images have greater luminosity and allow for higher frame rates, up to the point where the camera frame rate is twice the pulse repetition rate which allows for alternating frames of flame only and flame plus discharge. This provides an alternate method of distinguishing flame from discharge and enables flame front and microdischarge tracking.

Several images from a single test are shown in Figure 2 with test conditions given in the caption. The images show a flame approaching the discharge region from the left. Before the flame reaches the inter-electrode gap, the discharge is turned on and initially takes a uniform appearance. As the pulses accumulate, the discharge transitions to microdischarges. While the flame is inside the discharge volume, a uniform discharge exists in the burned gas, while microdischarges persist ahead of the flame. The flame front is considerably more wrinkled than it was prior to reaching the discharge volume. After the flame leaves the discharge volume, the discharge transitions back to microdischarges, with a uniform discharge persisting in regions of higher gas temperature. The use of the transparent electrodes enables several observations that were not possible previously with a conventional camera view along the y-axis (see [17]). A couple of novel observations, enabled by the ITO electrodes, are briefly discussed in the following paragraphs, with a quantitative analysis left to a future work.

In conclusion, a mesoscale channel-burner was used to investigate the interaction between a propagating premixed laminar flame front and an NRP DBD. The use of transparent ITO electrodes facilitated a comprehensive evaluation of the bidirectional interactions between the plasma and the flame. The discharge was observed to alter flame propagation by locally accelerating it and changing its geometry, through curvature and surface area increases. Simultaneously, the flame influenced the uniformity of the discharge and the overall motion of the microdischarges. Recognizing and quantifying this two-way interaction is key to the design of plasma-assisted combustion systems, particularly those involving propagating flame fronts and in-situ plasma. The use of transparent electrodes in plasma assisted combustion, ignition and detonation, opens up an opportunity to visualize and quantify 2D and 3D features of the interaction that were previously inaccessible, when using a viewpoint orthogonal to the electric field direction.