The developing brownout cloud beneath a hovering microrotor is simulated using a hybrid free-vortex methodunsteady Reynolds-averaged Navier-Stokes solver coupled with a Lagrangian sediment-tracking algorithm. The present work attempts to examine the capabilities of a high-fidelity computational fluid dynamics analysis together with a sediment-tracking algorithm to analyze the evolution of brownout clouds. Both the single-phase and dual-phase flowfields are simulated and compared to available experimental data. The variation of the phenomenological attributes of the cloud evolution as a function of particle size was analyzed and found to agree with experimental evidence. To enable fast single-desktop simulations, all components of this methodology are designed to run on double-precision, programmable graphics processing units. This hybrid implementation uses a multigranular parallel approach to demonstrate significant performance gains over equivalent single-core central processing unit simulations even when computations involve the solution of sparse implicit systems.
Nomenclature
A= rotor-disk area, πR 2 , m 2 A R = aspect ratio, R∕c a = speed of sound, m · s −1 C P = rotor power coefficient, P∕ρAΩ 3 R 3 C Poge = rotor power coefficient out-of-ground effect C T = rotor thrust coefficient, T∕ρAΩ 2 R 2 C Toge = rotor thrust coefficient out-of-ground effect c = rotor chord, m d p = particle diameter, μm h = height of rotor aboveground, m M = Mach number, V∕a R = rotor radius, m r = radial distance, m V = velocity, m · s −1 z = distance above the ground plane, m Θ = particle launch angle, rad ρ = flow density, kg · m −3 ρ p = particle density, kg · m −3 ψ = wake age, deg Ω = rotational frequency, rad · s −1