Multicopter flow fields and their influence on a spray released from multicopters. (2021)
Type of ContentTheses / Dissertations
Thesis DisciplineMechanical Engineering
Degree NameDoctor of Philosophy
Multicopters are remote-controlled vertical take-off and landing unpiloted aerial vehicles (UAVs). When used for releasing particulates (crop seeding, targeted fertiliser, and aerial spraying), they are a convenient tool for farms situated in rocky or mountainous terrain that does not allow for the use of helicopters or aeroplanes. Their small size and high manoeuvrability are also attractive for spraying near sensitive areas (e.g., riverbeds, lakes, native forest, residential areas).
Understanding the behaviour of spray is crucial for targeted spray dispersal and for the protection of sensitive areas. This research studies multicopter wake and its influence on the performance of spraying liquid.
The primary experimental technique used for the study of multicopter wake was stereo particle image velocimetry (SPIV), supplemented by constant temperature anemometry (CTA) with a three-axial probe.
The study analysed the isolated rotor wakes of the APC 1047 (127 mm radius, 119.3 mm pitch), APC 1045 (127 mm radius, 114.3 mm pitch), APC 1040 (127 mm radius, 101.6 mm pitch) and DJI E7000 (420 mm radius, 230 mm pitch). The isolated wake vector field, normalized by rotor tip velocity, was found to remain similar for each rotor with changing rotational speed.
Multicopters can be divided into two-rotor sections, allowing a simplified experimental setup using two rotors instead of four or more. APC 1045 counter-rotating coplanar rotor pairs were used for the analysis of multicopter rotors in hover, at rotor arc spacings of 02R, 0.36R and 0.55R (R=rotor radius). SPIV experimental analysis shows the rotor wakes tilting towards each other. The tilt angle decreases with increased tip spacing. Two counter rotating coplanar DJI E7000 rotors with 0.2R rotor tip spacing also demonstrated tilted wakes. The tilt may be explained by interaction between wakes creating a low-pressure zone between them.
A region of upwash was detected with APC 1045 rotors at rotor arc distances of 0.2R, 0.36R and 0.55R. The upwash region was observed at every rotational position and phase difference for DJI E7000 rotors at an arc distance of 0.2R. Upward velocity magnitude is dependent on the angular position of the rotor, peaking when one of the rotor tips is at its closest approach to the neighbouring rotor’s arc. It is weakly dependent on rotational speed over the range of 1740- 2150 rpm. Upwash generation may be explained by interaction between the tip vortices and wakes of the two rotors.
APC 1045 counter-rotating rotor pairs were analysed in the presence of lateral velocity. In streamwise configuration, the wake of the windward rotor tilts 20-25º more than that of the leeward rotor at lateral velocities of 6 m/s, 10 m/s and 14 m/s. The shading of the leeward wake by the windward wake is the cause of the difference in tilt angles.
In the presence of lateral velocity with a streamwise rotor configuration, the roll-up vortex is attached to the windward side of the windward rotor disk and extends in the direction of the airflow relative to the multicopter. With a spanwise rotor configuration, the roll-up vortex is attached to the windward side of both rotors and extends in the direction of the airflow near the free side of the rotor disk. However, in the presence of another counter-rotating rotor, the upwash region does not have a downstream lateral component at lateral velocities between 2- 6 m/s.
Based on SPIV analysis data used to track spray deposition near the rotors, it is recommended to avoid placing the spray nozzle immediately under the arc swept by the rotor tip (0.8R-1R), especially in the zone between rotors. This draws some spray upwards, decreasing spraying efficiency and potentially entering the multicopter’s electrical components. The recommended nozzle position is the zone of strongest downwash (0.5-0.7R).
A fast-computing model for spray pattern prediction was developed in OpenFoam, using rotor disk simplification as a boundary condition inside the domain. The velocity field boundary condition was obtained from SPIV data. The rotor boundary condition used the turbulence kinetic energy data obtained via CTA. The atmospheric wind model was incorporated into the model and can be used on-demand. The effect of plant canopy was introduced with a porous medium model.
Two DJI E7000 coplanar counter-rotating rotors were modelled in hovering flight. The modelled velocity field below the rotors was within one standard deviation of SPIV experimental results. The modelled velocity field between rotors was not within one standard deviation. The upward velocity region was not reproduced in the model.
Two APC 1045 rotors were modelled at 2 m/s, 6 m/s, 10 m/s lateral velocity in streamwise and spanwise configurations. In streamwise configuration, the leeward rotor is shaded by the windward rotor, therefore the inclination angle of the leeward rotor is smaller than that of the windward rotor. The roll-up vortices are observed in the model. The location of the roll-up vortices is similar in the model and experiment in both spanwise and streamwise configurations.
A DJI Agras MG-1 multicopter was modelled to allow comparison of swath patterns in the model and experimental results. Two roll-up vortices are present in the multicopter and extend in a streamwise direction. The model output was used for spray pattern prediction by applying Lagrangian particle tracking. An evaporation model was implemented in a particle tracking algorithm.
The spray footprints of two nozzle positions were modelled in hovering flight and compared to experimental results, revealing that the model can be used for spray footprint evaluation. Differences between the model and the experiment may be explained by absence of tip vortices in the model.
The swath pattern in the wake of a DJI Agras MG-1 multicopter in three different flights (true airspeed 3.627 m/s, ground speed 2.85 m/s, and crosswind speed 0.736 m/s; true airspeed 3.234 m/s, ground speed 2.9 m/s, and crosswind speed 2.164 m/s; true air speed 4.88 m/s, ground speed 4.88 m/s, and crosswind speed 0.04 m/s), is comparable in the model and experiment. The effective swath width (30% line separation) is within one standard deviation of the model.
In all flight trials, the modelled swath was closest to the experimentally obtained swath when the surface roughness of the ground was equal to 0.5 m (bushes) and the rotational speed of all rotors was equal to 2475 rpm with 0.75R (0.2m) tall plant canopy (grass) introduced to the model.
The model can be used to evaluate the swath pattern left on the ground by the multicopter. It showed acceptable validity for hovering flight and flight velocities of up to 2.8-5 m/s when flight parameters can be approximately estimated. The computational time of the model is 12 minutes.
RightsAll Right Reserved
Showing items related by title, author, creator and subject.
Chyrva I; Kabaliuk N; Dunker C; Strand T; Geoghegan P; Jermy, Mark (The University of Queensland, 2020)Multirotor unpiloted aerial vehicles are increasingly used for agricultural spraying. Spray dispersion is complicated by the interaction of adjacent rotor flows. Stereo PIV was used to map the flow field of a multicopter ...
Influence of contact parameters on Discrete Element method (DEM) simulations of flow from a hopper: Comparison with magnetic resonance imaging (MRI) measurements Danczyk M; Meaclem T; Mehdizad M; Clarke D; Galvosas P; Fullard L; Holland D (Elsevier BV, 2020)Here the effect of contact properties on discrete element method (DEM) simulations of dry granular flow was investigated. The particle contact properties were first measured using particle tracking velocimetry. DEM simulations ...
Wotherspoon NE; Jones OJ; Bhatti YA; Williams JHT; Mackie SL; Mulcahy JP; Revell, Laura (American Geophysical Union (AGU), 2021)Marine aerosol production is influenced by wind speed, particularly over the Southern Ocean which is the windiest region on Earth year-round. Using climate model simulations with artificially enhanced sea spray aerosol ...