![Sample our Engineering & Technology journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5933/TOC-Subject-Sample-2021-Engineering-Tech.jpg"})
Abstract
This work evaluates use of a dimensionally adaptive (DA) meshing technique to speed CFD modeling of non-reacting pressurized flow in a cylindrical geometry with a center jet and non-axisymmetric secondary jets. The DA-capable approach incorporates an a priori dimensional boundary location and extension of 3-D RANS flow solvers. Recirculation zone length, not flow asymmetry, was the limiting factor in placement of the dimensional boundary between 3-D and axisymmetric meshes. Boundary placement was based on confined turbulent jet theory, which suggests jet attachment is a linear function of cylinder radius, L, expressed as 5.85L. CFD results for a turbulent axial pipe flow showed that as the dimensional boundary was moved upstream, computational time dropped quadratically with the reduction in mesh cells. Further simulations of a 0.2032-m diameter, 1.5-m long reactor with turbulent center primary jet and outer radius secondary jets placed the dimensional boundary at 0.7 m. Results for pressures of 20, 10, and 5 bar and mass flow rates of 75% and 50% of baseline demonstrated two key results. First, flow similarity was preserved across all cases and the predicted jet attachment location was approximately 0.61 m, compared with the theoretical value of 0.594 m. Second, runtime reductions of 79% (pressure) and 82% (flow rate) were seen for the DA meshes compared with fully 3-D meshes. Differences in average exit axial velocities between DA and fully 3-D meshes were 2–4% for the pressure cases and 0.7–1.5% for the flow rate cases. Lastly, comparison with 3-D quarter-geometry calculations showed the DA approach to be ∼10% faster with exit axial velocities within 3–5% of 3-D velocities.