Computationally Efficient Lattice-Boltzmann Simulations of Multirotor Interactions in Boundary Effects

Rotor-rotor and rotor-boundary aerodynamic interactions of a quadrotor system without a fuselage in ground effect and ceiling effect for varying rotor-boundary distances and hub spacings were investigated. A GPU-accelerated Lattice-Boltzmann Method (LBM) was coupled to new unsteady actuator disk method (ADM) and actuator slice method (ASM) based rotor models for this purpose. Validation was conducted against experiments for both performance and particle image velocimetry flow field data. The trends in thrust and power were accurately predicted by both actuator methods, with high computational efficiency. Interactional flow physics were resolved, causing the consistent performance benefits very close to the ground, the performance penalties caused by the fountain flow effect between rotors occurring over a limited range of ground distances, and the persistent performance augmentation in ceiling effect. The ASM rotor model, with its individual blade representation, was found to predict rotor-rotor interactions more accurately than the ADM. However, it generally overpredicted aerodynamic loads, which was attributed to the Gaussian regularization applied along the blades. The results indicate that the GPU-accelerated LBM coupled with the introduced actuator methods is capable of computationally efficient prediction of multirotor aerodynamic performance and flow fields in ground and ceiling effects.