Title:
Experimental and Computational Analysis of Multi-Rotor Aerodynamic Interactions
Experimental and Computational Analysis of Multi-Rotor Aerodynamic Interactions
Author(s)
Wylie, Daley
Advisor(s)
Rauleder, Juergen
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
This study aims to perform a comparative analysis of rotor–rotor aerodynamic interac-
tions. This work details an investigation that was performed into the aerodynamic inter-
actions between the rotors of multi-rotor vehicles in different configurations. The effects
of these interactions on the thrust and torque of all individual rotors were quantified in
wind tunnel tests. The effects of the changes in hub spacings, rotor rotational speeds, and
freestream velocities were investigated for isolated, tandem, quad-rotor plus and X con-
figurations. The maximum and minimum tip chord Reynolds numbers were 118,000 and
73,000, respectively.
In addition to the experimental work conducted, the investigation was completed com-
putationally as well. This served as a validation tool for the computational solver, a method
of looking deeper into the conclusions drawn from the experimental investigation, and a
way to investigate other phenomena not completed experimentally. Cartesian Grid Euler
Solver (CGE), an advanced adaptive CFD solver that rapidly resolves three-dimensional
configurations during design, was used. CGE uses state-of-the-art flux splitting routines,
implicit time marching algorithms, higher order interpolation methods and multigrid-based
acceleration schemes together with flow-based adaptive mesh routines. It has been vali-
dated for complicated geometries.
Experimental and computational results showed that the aft rotors experienced detri-
mental aerodynamic interactions in all configurations. In all examined multi-rotor config-
urations, an increase in the hub spacing caused a decrease in the thrust deficit between the
aft rotor and the isolated rotor. However, the differences in the configurations also affected
the measured loads. In the tandem configuration, the aft rotor experienced up to 24% re-
duction in the thrust coefficient at a hub spacing of 2.1R when compared to the isolated
rotor at the same rotor rotational speed and freestream velocity. The aft-most rotor in the
plus configuration experienced as large as a 28% decrease in the thrust coefficient when
compared to one of the aft rotors in the X configuration for the same hub spacing and flight
conditions.
Good correlation was found between these wind tunnel experiments and flight tests for
the fore and side rotors in X and plus configurations (7.9–14.2% difference), but a larger
difference of 30–41.9% was found for the aft rotors, which is due to the different rotor
trim conditions. The flow solver was found to over-predict the thrust and under-predict the
torque due to a thin airfoil assumption and the lack of implementation of a formal tip loss
function. Nevertheless, the same trends were followed as the experimental results.
The effects of the flight test vehicle fuselage were investigated computationally. It was
found that the aft rotors in both the plus and X configurations experienced a decrease in per-
formance when the fuselage was added to the computational model, with thrust decreases
of 4.4% and 7%, respectively. The results also show that there was a 37% difference be-
tween the flight tests and computational data when using the same trim conditions. This
indicates that the cause of the difference in wind tunnel experiments and flight tests remains
unknown, and should be further investigated.
Finally, the effects of the wind tunnel facility were investigated. The same conditions
were modeled with and without the presence of wind tunnel walls numerically, and it was
found that the presence of the wind tunnel surface mesh caused the actuator disks to be
less refined, as an artifact of the automatic mesh refinement in CGE associated with the
relationship it defines between the internal and external mesh. These challenges made it
difficult to compare the computational results with and without the wind tunnel test sec-
tion. It was shown that the presence of the side walls caused a drop in performance of the
side rotors. However, given the aforementioned meshing complications, this finding needs
further investigation, numerically and experimentally.
Sponsor
Date Issued
2023-06-29
Extent
Resource Type
Text
Resource Subtype
Thesis