EXPERIMENTS AND ANALYSIS OF HYBRID-ELECTRIC POWER FOR MULTI-ROTOR UNMANNED AIR VEHICLES

Abstract

This research developed a fundamental understanding of a hybrid-electric powerplant for an unmanned rotorcraft powered by proprotors through model testing and comprehensive analysis. The key questions answered were by how much should the rotor tip speed be reduced in forward flight and by what mechanism? The correct answers were shown to be crucial to establish the payload, range, and gross weight of such aircraft. These answers could not be found without detailed engine-generator fuel consumption data, and without its inclusion in a detailed comprehensive analysis of such aircraft. This dissertation filled these gaps --- in a small scale within the resources available.

A $3.43$ kW $4$-stroke piston engine was coupled to a generator and rectifier to acquire basic data on power generation and engine specific fuel consumption. The isolated engine, generator, and coupled engine-generator efficiencies were characterized. A powertrain efficiency model was synthesized and validated, and partly calibrated, from the data. The model predicted engine specific fuel consumption as a function of engine torque and speed in a non-dimensional form. Thus, the model was rubberized for quantifying variations in engine efficiency with operating torque and speed in a scale-invariant manner. The model captures sensitivities to the engine operating conditions.

An in-house comprehensive analysis University of Maryland Advanced Rotorcraft Code II, UMARC-II, was significantly expanded to support the analysis of unique design elements of hybrid-electric aircraft. These refinements were validated with data from several wind tunnel test campaigns of the full-scale 26-ft diameter Boeing proprotor (M222), and the model-scale 4.75-ft diameter Maryland Tiltrotor Rig (MTR). UMARC-II was then used to understand rotor-wing dynamic and aerodynamic interactions, and was used to predict the power, L/D, and rotor torque and speed of a 50 lb quadrotor biplane. It was found that for a rotor mounted on a fixed wing generating lift, the oscillatory blade bending moments and the mean hub moments were amplified compared to an isolated rotor. The increase was proportional to how much lift the wing generated. Additionally, the wing aerodynamics were found to interact with the rotor aerodynamics in a way that accelerated the onset of stall at low speed for the quad biplane.

These quantities were coupled to the engine model through a rotorcraft sizing code. A rotorcraft sizing code was developed expressly to design electric and hybrid-electric VTOL. The aircraft was sized and a sensitivity analysis conducted to quantify the impact of engine specific fuel consumption assumptions. Once verified it was used to find the answer to the key question: by how much should the rotor tip speed be reduced? Next, several methods of engine speed control to meet the rotor operating torque and speed combinations were explored, compared to one another in complexity and value, and the best mechanism to reduce the rotor tip speed proposed. Thus, the key questions were answered.

The principal conclusions follow. For correct prediction of the optimal rotor tip speed reduction in cruise --- a key attribute and benefit of distributed propulsion --- a detailed engine specific fuel consumption model that captures the variation in speed and torque separately is required. Whereas aerodynamics alone suggest an optimal tip speed reduction to 40\% of hover, the correction reduction to minimize gross weight is much higher around 60\%-70\% of hover. Reduction in rotor tip speed should not be carried out by the electric motors alone. The engine should be controlled independently to minimize the fuel consumed at the desired power output. Based on the resulting bus voltage, the motors then control the rotors to achieve the desired tip speed reduction. Complexity is added to the control scheme, but significant operational gains are attained.