Open-Source 3D CFD of a Quadrotor Cyclogyro Aircraft

Abstract

This chapter provides a detailed method for building an unsteady 3D CFD model with multiple embedded and adjacent rotating geometries. This is done relying solely on open source software from the OpenFOAM package. An emphasis is placed on interface meshing and domain decomposition for parallel solutions. The purpose of the model is the aerodynamic analysis of a quadrotor cyclogyro. The challenging features of this aircraft consist of a series of pairwise counter-rotating rotors, each consisting of blades that oscillate by roughly 90 • about their own pivot point. The task is complicated by the presence of solid features in the vicinity of the rotating parts. Adequate mesh tuning is required to properly decompose the domain, which has two levels of sliding interfaces. The favored decomposition methods are either to simply divide the domain along the vertical and longitudinal axes or to manually create sets of cell faces that are designated to be held in a single processor domain. The model is validated with wind tunnel data from a past and finished project, for a series of flight velocities. It agrees with the experiment in regard to the magnitude of vertical forces, but only in regard to the trend for longitudinal forces. Comparison of past wind tunnel video footage and CFD field snapshots validates the features of the flow. The model uses the laminar Euler equations and gives a nearly linear speedup on up to 4 processors, requiring one day to attain periodic stability.

Aerodynamic Analysis of an Unmanned Cyclogiro Aircraft

“…It is thus reasonable to plan a turnover time of 10 days per case. To reproduce the model, the reader is encouraged to read the relevant publication [35].…”

“…It evolved from a previously validated rotor model [30] by carefully designing the airframe mesh [35]. The resulting model is validated qualitatively, as reported in the companion paper [35]. To further ensure validity of the model, this section presents both an investigation on the rear rotor velocity and a verification that the trends of the generated forces follow the experimental data as the forward flight velocity increases.…”

“…Once a stable analysis method was thus obtained, the geometry and simulation parameters were validated against the wind tunnel data. The thorough description of the model, boundary conditions, numerical methods, and mesh generation process are given in a different publication [35].…”

“…Thus, an attentive iterative design procedure was presented so far. It evolved from a previously validated rotor model [30] by carefully designing the airframe mesh [35]. The resulting model is validated qualitatively, as reported in the companion paper [35].…”

“…Regardless of this, the calculated longitudinal forces carefully follow the trend of the experimental data at various flight velocities. A mesh refinement study was undertaken prior to choosing to final mesh size and is reported with more details in the companion paper [35]. It covered both the airframe alone with grids having 0.3 to 1.7 million cells and the whole aircraft with grids having 3.7 to 5.7 million cells.…”