Fast framing imaging and modelling of vapour formation and discharge initiation in electrolyte solutions

Abstract

The formation of vapour and initiation of electrical discharges in normal saline and other NaCl solutions has been observed and modelled. Millisecond pulses of −160 to −300 V were applied to a sharp tungsten carbide electrode immersed in the liquid. A fast framing camera was used to observe vapour layer growth around the electrode by shadowgraphy. Images were also taken without backlighting to observe emission accompanying breakdown. The conductivity of the vapour layer has been estimated by comparison of experimentally measured impedances with impedances calculated by finite element modelling of the liquid and vapour around the electrode observed by shadowgraphy. The conductivity of the vapour layer is estimated to vary between ∼1 and 10−3 S m−1, which are orders of magnitude higher than expected for water vapour. The reason for the high conductivity is not clear, but may be due to injection of charge carriers into the vapour by corona discharge at the sharp tip and/or the presence of cluster ions in the vapour. Alternatively, the vapour layer may be a foam mixture of high conductivity liquid and low conductivity vapour bubbles. Discharge formation does not occur primarily at the sharp tip of the electrode, but rather at the side. Finite element models of the vapour layer show that the highest electric fields are found close to the sharp tip of the electrode but that the field strength drops rapidly moving away from the tip. By contrast slightly lower, but consistently high, electric fields are often observed between the side of the electrode and the liquid vapour boundary where plasma emission is mostly observed due to the shape of the vapour liquid boundary. The electron number density in the discharge is estimated to be ∼1021 m−3. There is evidence from the shadowgraphy for different boiling mechanisms; nucleate boiling at lower voltages and film boiling at higher voltages. A rule of thumb based on a simplistic model is suggested to predict the time to discharge based on electrode area, initial current, and electrolyte temperature, density, conductivity and heat capacity.