Ph.D. Defense
Mahmoud A. Hayajnh
(Advisor: Prof. Prasad)
"Development of Parametric Rotor Control Equivalent Turbulence Input (RCETI) Models"
Monday, January 27
3:30 p.m.
Montgomery Knight 317
Summary
Turbulence modeling and simulation is crucial for evaluating the ability of rotary wing vehicles to perform tasks in adverse weather conditions. For this, models that replicate the effect of turbulence on vehicles, and that are relevant to flight testing and simulations are required. This study focuses on exploring the methodology to develop Rotor Control Equivalent Turbulence Input (RCETI) models. These models aim to generate control inputs that produce rotor responses, and thus, vehicle responses, that are stochastically similar to the response in atmospheric turbulence. Unlike existing vehicle-specific Control Equivalent Turbulence Input (CETI) models, the rotor-specific models developed in this study offer a more generalized approach. Since these models are rotor-specific, parametric models can be developed as function of rotor parameters only. This approach effectively reduces the parameter space for the generalization of these models to be scaled and applied to various rotorcraft configurations. Utilizing these models allows for simulation of the behavior of different rotorcraft configurations in various flight scenarios, without the need for actual flight testing in turbulent conditions.
Currently, Control Equivalent Turbulence Input (CETI) models have been developed for specific vehicles, and they are not transferable to other types of rotorcraft. Creating parametric models that can be scaled and applied to multiple types of rotorcraft can reduce the time and cost of developing models for new rotorcraft vehicles, as they can be adapted and scaled to various configurations. By utilizing hub-loads as outputs and swashplate deflections as inputs, rotor-specific CETI models can be developed. This recognizes that the stochastic characterization of vehicle response to turbulence is primarily driven by stochastic characterization of rotor hub loads. The use of hub loads as outputs captures the aerodynamic forces and moments induced by turbulence, while the swashplate angles serve as control inputs influencing the blade pitch angles. This focused approach reduces the parameter space and enables efficient scalability of the models to different rotorcraft types. Additionally, these models can be extended to multi-rotor vehicles by applying them to each rotor individually.
To demonstrate the feasibility and effectiveness of the RCETI approach, FLIGHTLAB is utilized to develop a comprehensive nonlinear helicopter model representative of a UH-60 Blackhawk helicopter. The model incorporates a 33-state inflow model and accounts for the elasticity of the rotor blades. This nonlinear model accurately captures the dynamic loads experienced by the rotor system. By performing a linearization around a periodic equilibrium, first-order linear time periodic (LTP) approximations, which account for the coupled dynamics of the body, rotor, and inflow, are derived from the nonlinear model. To facilitate the analysis of developing RCETI models, the LTP approximations are further transformed into linear time-invariant (LTI) approximations using the harmonic decomposition methodology. The fidelity of the resulting LTI approximation is evaluated by comparing its predictions with data from nonlinear simulations, both in the frequency and time domains. This assessment ensures the accuracy and reliability of the LTI approximations for subsequent analysis. A comparison of vehicle response spectra to hub load spectra due to turbulence revealed that the rotor is the primary load-producing element in turbulence, supporting the use of hub loads as outputs in the RCETI models. Using the developed LTI approximations, RCETI models are created by employing swashplate angles as inputs. It is found that RCETI models can effectivity replicate the vehicle response to turbulence.
After establishing that rotor is the main element to produce loads due to turbulence, the analysis shifts to isolated rotor models. Time-domain simulations were conducted to capture the effects of turbulence using two-dimensional turbulence model to find blade-element sample fluctuations. The effect of turbulence rotational sampling was assessed using the isolated rotor model. It is seen that hub-fixed sampling is inadequate for low speed scenarios, and that blade-element sampling is needed to capture the effect of turbulence. Next, different isolated rotor models, representing various configurations selected through design of experiment method, were created in FLIGHTLAB, and subsequently used to develop parametric RCETI models. A neural network framework is used to develop parametric RCETI model, which relates rotor parameters to the coefficients of RCETI model transfer functions. These models incorporate rotor parameters such as rotor solidity, lock number, etc. This approach allows for scalable modeling across a broad range of rotor configurations. The scalability of these neural network-based models is evaluated for various rotor parameters and configurations, including those that were not initially considered during the model development phase.
The research further extends RCETI methodology to encompass multi-rotor vehicles, considering both variable-RPM and variable-pitch rotor configurations. The applicability of RCETI models in multi-rotor systems was demonstrated under uniform and differential turbulence scenarios, where time delays are utilized to simulate spatially varying turbulence effects. The findings demonstrate that RCETI models can effectively replicate stochastic vehicle response to turbulence, emphasizing their adaptability to multi-rotor configurations.
In summary, the development of generalized RCETI models offers significant benefits in terms of efficiency and cost-effectiveness for the design, development, and testing of vertical lift platforms. By utilizing these models, the need for extensive flight tests in turbulent conditions to gather data and develop specific CETI models can be eliminated, thereby minimizing the inherent risks associated with conducting such tests.
Committee
• Prof. J.V.R. Prasad – School of Aerospace Engineering (advisor)
• Prof. Marilyn J. Smith – School of Aerospace Engineering
• Prof. Lakshmi N. Sankar – School of Aerospace Engineering
• Dr. Mark J. Lopez – Technology Dev. Directorate, U.S. Army Combat Capabilities Dev. Command Aviation & Missile Center
• Prof. Umberto Saetti – Department of Aerospace Engineering, University of Maryland