3D Evolutionary Reconstruction of Scalar Fields in the Gas-Phase

Abstract

An evolutionary reconstruction technique (ERT) was developed for three-dimensional (3D) reconstruction of luminescent objects, in particular turbulent flames for the first time. The computed tomography (CT) algorithm is comprised of a genetic algorithm (GA) and a ray-tracing software. To guide the reconstruction process, a mask is introduced. It uses a Metropolis algorithm (MA) to sample locations where specific genetic operators can be applied. Based on an extensive parameter study, performed on several types of phantoms, the ability of our algorithm for 3D reconstructions of fields with varying complexities is demonstrated. Furthermore, it was applied to three experiments, to reconstruct the instantaneous chemiluminescence field of a bunsen flame, a highly turbulent swirl flame and the turbulent Cambridge-Sandia stratified flame. Additionally, we show direct and quantitative comparison to an advanced computed tomography of chemiluminescence (CTC) method that is based on an algebraic reconstruction technique (ART). The results showed good agreement between CTC and ERT using both phantom data from flame simulations, and experimental data.

Keywords

• tomography
• combustion
• genetic algorithm
• ray-tracing
• Monte–Carlo sampling