Search

The Chair of Flight Dynamics at the RWTH Aachen University is conducting research on a method for identification of flight mechanical characteristics on free flying models in a wind tunnel. The main goal is to create a eproducible free flight environment for cost effective identification of important values even in an early design stage. The method will combine the advantages of free flight with wind tunnel techniques as it takes the free flight into a reproducible environment under laboratory conditions. The paper gives an overview of the project and provides insight into the work done so far.

Requirements for control systems in sailplanes and powered sailplanes to be equipped with servo-powered transmission and control elements have been analyzed to provide information regarding airworthiness of such controls. A flight dynamics dataset was acquired in flight tests and used to create a six degree-of-freedom simulation in which various potential actuator failure scenarios were investigated. An exemplary functional hazard analysis was conducted. To demonstrate a servo-boosted control system for the airbrakes, a system based on an electric actuator was designed, built, and evaluated in flight. The system kept the hand force below a specified level and allowed controlled deployment of the airbrakes over the whole speed range.

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2022; Aachen : RWTH Aachen University 1 Online-Ressource : Illustrationen, Diagramme (2022). = Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2022, The process of developing a new type of helicopter starts with the planning and design phase, in which modern methods such as model-based design are used. By this, conceptual errors in the preliminary design can be detected and eliminated at an early stage. This saves time and effort, which has a positive effect on costs. The basis of model-based design in helicopter construction is typically a parameterisable multibody model with a modular structure, which allows the modelling of any helicopter configuration. In the context of the present thesis, a multibody formalism is presented which automates the modelling of helicopters. The formalism allows the synthesis of all known helicopter configurations in a tree structure while adhering to the generally valid notation of flight mechanics (DIN 9300 and ISO 1151). Due to the tree structure, the algorithm is particularly well suited for model-based design and real-time simulations, as additional calculation steps through kinematic loops are not required. To ensure that the formalism is correct, two helicopter configurations (one standard and one coaxial) containing rotor hubs, rotor blades and four-point landing gear are modelled and subjected to plausibility tests. Techniques based on trim calculation and linearisation as well as numerical simulation of the nonlinear helicopter dynamics are used. For this purpose, an algorithm based on the convolution integral was developed to linearise the periodic and non-minimum-phase helicopter dynamics. The trim calculation is based on the multi-dimensional secant method, which transforms the inherent helicopter dynamics into stationary conditions for the flight dynamic analysis. The conditions include hover, horizontal flight and ground case, where the helicopter is still on the ground with the rotors not turning. Plausibility tests, in which numerics play a role, are performed with desktop or Pilot-in-the-Loop simulations. In the Pilot-in-the-Loop Simulation, the test pilot acts as a quality inspector whose task is to detect and identify abnormal behaviour due to model errors in the controlled helicopter dynamics. This is achieved by the use of validated state controllers and virtual reality techniques. For this purpose, the helicopter model is ported to a real-time computer connected to a radio control unit and virtual reality goggles to provide the test pilot with an extremely realistic test scenario. A total of six tests are carried out to verify the multibody formalism. Among them is an analytical test in which the linear flapping motion of the rotorblade is derived from the non-linear Equations of Motion of the multibody model, which can be seen as the first indication of a correct multibody formalism. The remaining tests are performed using trim calculation and linearisation as well as numerical simulation. These include the drop test, trim and stability tests with varying Neutral Point of the main rotor, and the Pilot-in-the-Loop test, which can be performed by an experienced test pilot. Analogous to the analytical test, the results of all plausibility tests confirm the correctness of the multibody formalism., Published by RWTH Aachen University, Aachen

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2020; Aachen 1 Online-Ressource (VI, iii, 87 Seiten) : Illustrationen, Diagramme (2020). = Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2020, Aircraft with vertical take-off and landing (VTOL) capability differ fundamentally from conventional airplanes due to their large flight envelope. This flight envelope ranges from the thrust-borne hover flight to the lift-borne aerodynamic flight, leading to highly variable, nonlinear aerodynamics. Modeling these aerodynamics requires elaborate computer simulations (i. e. CFD 1 ), wind-tunnel campaigns or flight experiments. When transitioning from the thrust-borne to the lift-borne flight regime, over-actuation often occurs, since both, the actuators required for thrust-borne and the actuators required for lift-borne flight, have a significant effectivity. This leads to the question of actuator allocation, which deals with the distribution of available actuators according to the commanded forces and moments. Over-actuation is the result of redundancy regarding the available actuators. While this redundancy complicates the controller design, in principle it also allows for fault-tolerance against actuator failure and should thus be considered in the controller design. In total, the inherent complexity due to aerodynamics, over-actuation and fault-tolerance are often reflected in the control system, creating a challenging control problem. With the goal of minimizing the required effort to design a controller, this work presents a control concept based on the principle of Incremental Nonlinear Dynamic Inversion (INDI). While classical Nonlinear Dynamic Inversion (NDI) is based on a comprehensive aerodynamic model, this dependency is largely eliminated when using the incremental form of NDI, as it only depends on the control effectivity. To enable a flexible implementation on different types of aircraft, the control effectivity is based on simple semi-empirical models and geometrical properties of the aircraft. Results from theory and practice show that this approach is sufficient in terms of controller performance. On top of that, the controller concept presented here incorporates a systematic approach to over- and under-actuation and thus enables a tolerance against actuator failure. The control concept is implemented on a tilt-wing aircraft. To validate the concept, both, simulation studies and flight tests are performed. Presenting selected failure modes in flight tests, the tolerance against actuator failure and the robustness against model uncertainties is confirmed., Published by Aachen

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2019; Aachen 147 pp. (2019). = Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2019, The wide flight envelope of tilt-wing aircraft includes VTOL and high flight speeds, making them attractive for various applications such as inspection, air delivery and air taxi. The different flight states -from thrustborne to liftborne- place high demands on the design of aircraft and control system. Thanks to current technologies, tilt-wing aircraft can be designed unstable to achieve high flight performance. Nevertheless, stabilization is a mandatory criterion which must always be fulfilled. The required flight controller must therefore ensure stabilization throughout the entire flight envelope while enabling high flight performance. In this work, a flight state controller is designed and evaluated which finds a compromise to achieve high flight performance while maintaining the required flying qualities for tilt-wing aircraft. This supports the ability to shift the design objectives of tilt-wing aircraft to good flight performance. The contradictory goals of high flight performance and good flying qualities on the one hand and the conflicting requirements for high flight performance in different flight conditions on the other are discussed. The designed flight state controller enables an optimal compromise between these contradictory goals based on the consideration of predefined criteria. Feedforwarding a trim control vector allows stationary flight in the entire flight envelope. For the usually large number of manipulated variables, a suitable trim control vector is selected on the basis of different criteria of flight performance and flying qualities. A control allocation preserves the control reserves required for stabilization and generates a compensated control deflection by applying the pseudoinverse. The reserve of the manipulated variable is taken into account implicitly by an artificial reduction of the control effectiveness. The designed approach of the flight state controller is applied to an example aircraft with high flight performance and a large number of control variables. The functionality of the approach is proven by means of a nonlinear simulation and final flight tests., Published by Aachen

RWTH Aachen University, Diss., 2017; Aachen, 1 Online-Ressource (xii, 140 Seiten) : Illustrationen, Diagramme(2017). = RWTH Aachen University, Diss., 2017, Design criteria are essential for qualitative and quantitative assessment, analysis and optimization of control systems. Flight control law design for conventional aircraft uses dedicated criteria in time and frequency domain. For tilt-wing aircraft in particular, criteria used for conventional aircraft cannot be directly adapted. Tilt-wing aircraft rotate their wings together with the propulsion system around the aircraft lateral axis. The resulting significant change in configuration influences the flight dynamics and aerodynamic characteristics. In order to enable seamless operation throughout the entire flight envelope from hover flight to energy efficient fast forward flight, an optimization based approach can significantly speed up tilt-wing system design.The main topic of this thesis is the derivation and systematic application of design criteria depending on specific flight dynamics and variable operating points for control law design and flight dynamics assessment of a tilt-wing aircraft configuration. To achieve this, the flight dynamics of an exemplary tilt-wing configuration is modeled in time and frequency domain to enable evaluation of criteria throughout the entire flight envelope. The criteria are analyzed systematically with regard to their suitability to assess and optimize a control system of tilt-wing aircraft. Furthermore, an optimization based approach is applied to validate and adapt the selected criteria and flight states.In a concluding verification step, a benchmark control system is enhanced by applying multi-objective optimization using selected criteria. Variations regarding performance, control and disturbance responses as well as stability are analyzed. Results show that the determined criteria are suitable for optimization of a tilt-wing flight control system. As the criteria are dependent on mission and configuration, it is shown how they can be adapted using a systematic approach for varying requirements, flight states and configurations of tilt-wing aircraft., Published by Aachen

Dissertation, RWTH Aachen University, 2017; xvii, 120 Seiten ; Illustrationen, Abbildungen(2017). = Dissertation, RWTH Aachen University, 2017, Information about the flightmechanical characteristics of a new airplane is necessary from design through development until certification.Free-flight tests with dynamically similar scaled models can provide a useful contribution in all stages.In addition to geometric scaling, dynamic similarity requires the same relative airplane mass density and Froude number of the scaled free flight model and the unscaled airplane.This thesis demonstrates that the method of dynamically scaled free flight models can in particular be applied to general aviation airplanes and their certification according to CS-23.To this end, the requirements of dynamic scaling were investigated with respect to the applicability to models with different weights.It could be shown that free flight models with a maximum weight of 25 kg are particularly useful to build dynamically scaled models of general aviation airplanes.Within the scope of the European research project CESAR a dynamically scaled model of the EV-55 Outback was designed and built.Flight test maneuvers were chosen from maneuvers which are also used during the certification process according CS-23.A parameter identification of the model was executed based on the output error method.The parameter identification results of the model were converted to the unscaled airplane and an evaluation of the flightmechanical characteristics was performed with respect to the CS-23 certification requirements.The result of this work is that dynamically scaled free flight models are very useful to investigate CS-23 certification requirements on static and dynamic stability as well as some flight performances. Even though for some parameters the flight test maneuvers could not excite the proper motion of the model for an estimation of all parameters.For other very generally described requirements more informations are required to evaluate the requirements. Without a comparison to an already certified airplane and its dynamically scaled model an evaluation is only possible with limitations., Published by Aachen

As a result of complex and aging structures an ever increasing number of infrastructures prone to damage exists, especially in industrialized countries. At the same time there is a lack of adequate and automated methods for the inspection and monitoring of these infrastructures due to the exposed location as well as the specific design of these objects. On the basis of current developments in the field of robotics, new possibilities are emerging for the use of Remotely Piloted Aircrafts (RPA), which in the case of airborne inspection can be used as flying sensor platforms allowing a less time-consuming data acquisition while saving costs for personnel and materials. The prerequisite for a safe and economical application is the existence of a precise as well as reliable navigation method as a basis for reproducible results of such automated inspection systems.This thesis discusses the safe and defined navigation procedures for the approach, positioning and control of RPA systems in the close vicinity to structures. At first, five airborne non-destructive testing procedures are examined hereto, which are basically available for integration into RPA systems. Based on the explanations of the pursued inspection process, the requirements for the entire system are then formulated. In order to be able to assign a specific RPA platform to the navigation method, a system evaluation is carried out for various RPA configurations based on their characteristics, which concludes with the selection of an appropriate RPA system for the final validation of the navigation method. Equivalent to the RPA, three measuring methods basically available are presented and evaluated with regard to their usage as airborne navigation sensors. The final sensor technology is then determined in a direct comparison followed by the selection of a specific navigation sensor. The selected sensor is subjected to an in-depth evaluation providing information about the exact, individual sensor behavior. The knowledge of these sensor characteristics is essential for the development of a complex sensor model. The core aspect of the close-range navigation is first formulated in a general approach, which is then subsequently adapted to the previously identified inspection criteria as well as the selected navigation sensor. In a next step the integration concept is presented which specifically contains the modeling of the position control on the basis of flight modes as well as the modeling of a complex LiDAR sensor model based on the evaluation results. The integration concept is evaluated in a subsequent simulation using an individually created scenario under laboratory conditions. Finally the basic function of the indicated close-range navigation method is demonstrated during flight tests using a real RPA system.

This work describes the design and implementation of a real-time aerodynamic parameter estimation algorithm on a small remotely piloted aircraft. The Extended Kalman Filter (EKF) is adapted for aerodynamic parameter estimation. A formulation is given which is similar to the recursive least squares (RLS) algorithm but uses noise covariances instead of a forgetting factor for tuning. Optimization for low computing power hardware is discussed. A demonstrator aircraft based on an R/C model is equipped with the required hardware (air and inertial data sensors, onboard processor, telemetry). Wind tunnel tests and calculations produce a reference data set for aerodynamics and propulsion of this aircraft, which is then used for simulation. This simulation allows to prove the performance of the parameter identification (PID) algorithm and predict the set of parameters which is identifiable with the given hardware. The main influence on identifiability is the relative contribution of a derivative to a coefficient in relation to the output noise level. Correlation issues are identified which arise because of the very fast rolling motion (and somewhat less the pitching motion) in comparison to the achievable update rates. Two sorts of flight test results are presented: post-flight analyses of logged flight data and identified parameters from the working real-time algorithm. Although some minor derivatives are not identifiable, the results prove the general feasibility of the approach.

To achieve identification of the flight mechanical parameters of aircraft experimental techniques like free flight or wind tunnel tests are still essential, even if the numerical methods are getting better. But the mentioned techniques have both their disadvantages: On the one hand because the mounting of the model during wind tunnel measurements causes interference to the flow. On the other hand coupling effects caused by the movement of the aircraft in the flow can only determined with high costs – in some cases it is even not possible. The disadvantages of the free flight techniques consist of no reproducible conditions (atmospheric disturbances), high costs and the risk of manned test flights. The goal of this work which was conducted at the Chair of Flight Dynamics at the RWTH Aachen University was the creation of a reproducible free flight environment for the cost effective Identification of main flight mechanical parameters even in an early design stage. Therefore the advantages of the free flight techniques will be combined with the advantages of the Wind Tunnel techniques by bringing the free flight into the wind tunnel under laboratory conditions. The position and attitude of the aircraft is affiliated by a 3D Camera System with high frequency and accuracy. Hence, the aircraft must only be equipped with sensor for the control surface positions and the revolution of the engine. The aircraft and the whole process are controlled by a real time system, which is implemented in Matlab/Simulink before. An adaptive identification algorithm, based on a regression in the frequency domain, generates the required excitation manoeuvre and analyses them in real time. Because the algorithm is adaptable, it needs only little a priori knowledge of the aircraft characteristics. The manoeuvre will be adapted until they fit most exactly to the eigenfrequency of the aircraft. The control algorithm has the function to reposition the aircraft after an identification manoeuvre and to trim. Additional, the algorithm takes control of the aircraft if it reaches the border of the free stream and tries to reposition it in the center of the wind tunnel. Because of the high non linearity and the agility of the aircraft the dynamic inversion is used. This is enhanced by a pseudo control hedging to avoid non linear rate- and deflection limits and to avoid the inversion of the dynamic of the actuator. As this combination is not very robust against uncertainties in the parameters, as they appear especially in this case, additionally adaptive terms in form of neural networks where included. After the free flight experiment an identification based on offline algorithms is conducted. Because of the higher resources even nonlinear models and better filter algorithms can be used. Because of the sensor concept, drift and bias in the signals do not exist. Therefore an equation error method is adequate. The Validation exemplary takes place with an aircraft in the wind tunnel of the Chair of Flight Dynamics.

Enhanced methods for the determination of flightdynamic parameters are presented in this study. Because of their simple setup this study concentrates on two methods that have to be modified to study the unstable spaceplane configuration PHOENIX. In the wind tunnel method of free oscillation the wind tunnel model is equipped with electrically driven control surfaces and a feedback control system to operate the model in the test section. For the second method a radio controlled model of the spaceplane is dropped from another unmanned aerial vehicle to collect data during its flight. The application of feed forward- and feed back control systems enhances this method to make it applicable to unstable configurations with a high wing loading. In a first step an equation error method is used to identify the parameters of quasi-stationary aerodynamic models. It is shown, that the parameters of the equations for lift, pitch moment and rolling moment show a significant dependency on the frequency of motion. Therefore unsteady aerodynamic models are introduced that integrate these frequency effects with the help of transfer functions. The identified parameters of these unsteady models are shown to be independent from the frequency of motion. A comparison of wind tunnel data with data from the flight tests makes clear, that it is possible to identify precise parameters of quasi-stationary models with this kind of flight tests. The discrepancies between results from the two methods are in the same order than in much more ambitious flight test campaigns.

This thesis describes a method for a mission oriented multidisciplinary design for controlled Micro Air Vehicles (MAV). MAV’s are unmanned air vehicles with maximum dimensions smaller than 50 centimetres, whereby their dimensions are much smaller than the ones a unmanned air vehicles used up to now. Because of the manifold applications for Micro Air Vehicles the mission requirements differ very much. To tap the full potential of a MAV, it is necessary to adapt the Vehicle to the specific mission requirements. For this reason the presented method was developed offering the possibility to adapt the Micro Air Vehicle to user specified mission requirements. Therefore the method performs the aerodynamic design of the configuration, the choice of all necessary hardware components and the determination of the on-board flight controller, which is necessary to improve the flying qualities and to reduce the pilot’s work load. Using an integrated optimization algorithm, which estimates all subsystem similar and at the same time, the best compromise to fulfil the mission requirements is determined. Using several examples of use the function of the design method could be approved. For the final validation of the whole design procedure two Micro Air Vehicles were realized and flight tested. Based on the flight test data the predicted results of the design procedure could be confirmed.

A method for the nonlinear analysis of dynamic systems and its application to aircraft is presented. In extension of existing averaging methods this procedure follows the linearizing eigenvalue analysis, but defines the eigenvalue and eigenvector as functions of the deflection from the examined equilibrium. This approach allows the determination of the systems stability properties even beyond the close vicinity of the equilibrium. A test case to verify the robustness of the method and its application to the model of an agile aircraft with highly nonlinear aerodynamics and a large aircraft with flexible degrees-of-freedom is shown. The analysis shows very good agreement with the results of bifurcational analyses and allows a quantitative prediction of the systems stability behavior even between the branches of the bifurcation diagram. The unstable spiral mode of the flexible aircraft shows the existence of additional stable equilibria in the vicinity of the examined equilibrium due to system nonlinearities.