Modeling issues in the computation of plume radiation signatures

Multi-spectral radiation images have been obtained of Atlas rocket plumes under flight conditions. These spatially resolved data have been used to validate computational methods and models for plume simulations. However, recent computational fluid dynamics calculations of these flows showed poor agreement with the experimental radiation data. It is proposed that this lack of agreement is due to two factors: poor grid resolution of the flowfield and the presence of soot in the Atlas plume. Grid resolution studies on an axisymmetric plume show that grid sizes of 250,000 points with strong clustering in the barrel shock impingement region are required to capture the features of these flows. Solutions on less resolved grids can lead to incorrect impingement location and an underprediction of the post shock temperature of several thousand degrees. In addition, it is observed that a regular reflection can suddenly jump to a Mach reflection when the grid is resolved, which has a major impact on the post shock flow and radiation signature, and could instigate afterburning in some cases. Three dimensional simulations of the Atlas plume were modified to include small concentrations of soot as an overlay species. Radiation calculations based on the soot-laden flowfield show much better agreement to the experimental data, indicating that soot could be a major radiator in these flows.