Multi-domain analysis and prediction of the light emitted by an inductively coupled plasma jet

TL;DR

This paper investigates how torch power and chamber pressure affect plasma jet behavior in an inductively coupled plasma wind tunnel, using high-speed imaging and Gaussian Process Regression to predict jet profiles under various conditions.

Contribution

It introduces a multi-domain analysis combined with a data-driven Gaussian Process Regression model to predict plasma jet emissions across different power and pressure settings.

Findings

High-speed imaging reveals jet dynamics under different conditions.

Gaussian Process Regression accurately predicts jet profiles at unseen parameters.

Understanding plasma jet behavior aids in material testing and simulation development.

Abstract

Inductively coupled plasma wind tunnels are crucial for replicating hypersonic flight conditions in ground testing. Achieving the desired conditions (e.g., stagnation-point heat fluxes and enthalpies during atmospheric reentry) requires a careful selection of operating inputs, such as mass flow, gas composition, nozzle geometry, torch power, chamber pressure, and probing location along the plasma jet. The study presented herein focuses on the influence of the torch power and chamber pressure on the plasma jet dynamics within the 350 kW Plasmatron X ICP facility at the University of Illinois at Urbana-Champaign. A multi-domain analysis of the jet behavior under selected power-pressure conditions is presented in terms of emitted light measurements collected using high-speed imaging. We then use Gaussian Process Regression to develop a data-informed learning framework for predicting Plasmatron X jet profiles at unseen pressure and power test conditions. Understanding the physics behind the dynamics of high-enthalpy flows, particularly plasma jets, is the key to properly design material testing, perform diagnostics, and develop accurate simulation models