|
Executive summaries of Action Groups of Group of Responsables in Flight Mechanics
Introduction and Motivation Today's high performance aircraft are no longer flown directly by the pilot but have electronic flight control systems with sophisticated control algorithms that shape the pilot's commands to give good aircraft handling. This is especially true for fighter aircraft, which are often designed to be naturally unstable to improve performance. The European aeronautical industry is always looking for ways to improve the efficiency of the design of these systems. From 1994 until 1997, GARTEUR FM(AG08) has researched the application of robust control law synthesis. After FM(AG08) had finished its activities, a new and complementary challenge was defined for FM(AG11). This challenge focuses on the clearance of flight control laws, where it must be proven to the authorities that the flight controller is functioning correctly under all operational conditions, before flight tests are allowed. It was indicated by European aeronautical industries that clearance is a lengthy and expensive process. The number of cases that has to be investigated is huge, especially for fighter aircraft. All store configurations have to be cleared, over a flight envelope that may include transonic flight and high angles of attack. Parameter uncertainties in the aircraft model and the limited accuracy of air data measurements also have to be taken into account. A major improvement of today's clearance process can be expected by increased automation of the tools used for model-based analysis of the aircraft's behaviour. Over the past two decades, several mathematical techniques have been developed for the analysis of linear and nonlinear systems with uncertain parameters. At the start of AG11, it was difficult for the aeronautical industry to assess whether their application would improve the efficiency of the flight control law clearance process. Objectives and Description of the Work The main objective of the action group was to explore the potential benefits of using advanced analysis methods for the clearance of flight control laws, by demonstrating some of the most promising techniques on realistic flight clearance problems. As the analysis of a complex system produces complex results, good visualisation is essential to gain a deeper understanding of the clearance results ¾ upon which, important decisions about the airworthiness of an aircraft are to be based. For this reason, a secondary objective of the research activity was to explore, based on a "wish list" from industry, new tools that would improve the visualisation of clearance results. To achieve the objectives, seven major activities were defined: 1. Description of the current industrial clearance process. 2. Definition and modification of three fighter models to meet the group's needs: · HIRMplus (High Incidence Research Model "plus", a generic fighter model) · ADMIRE (Aero Data Model In a Research Environment, a delta-canard configuration) · Harrier Wide Envelope Model. 3. Application of new analysis techniques to the benchmark models (m-analysis, n-gap analysis, bifurcation analysis, a polynomial-based analysis method and optimisation-based clearance). The classical approach was also applied to provide a baseline reference. 4. Identification of requirements for visualisation tools. 5. Industrial evaluation of results. 6. Dissemination of results in a book and a final workshop. 7. Development of a GARTEUR Wide Area Network to support the group's activities. Main Technical Results Three models were developed in this project. In phase one of the analysis, each team applied their analysis techniques to the HIRMplus model, which was developed from the HIRM, which was available in a mature state from FM(AG08). In a second phase, analysis teams addressed one of the more realistic models, ADMIRE or HWEM. These models were only available at a later stage of the project, as they needed more development to adjust them to meet the group's needs. In particular, the ADMIRE and HWEM are of great potential value for future research activities, since such realistic nonlinear aircraft models are not usually available to universities. Five new analysis techniques have been applied, and compared with a baseline (classical) solution. In the industrial evaluation, the techniques were judged on generality, reliability, conservatism and the level of understanding and the learning effort required for the method. All the analysis methods have shown their strengths and potential to improve (part of) the clearance process, but also, drawbacks have been identified for each method which might act as drivers for future developments of the method. For all techniques, directions were given for future improvements, based on the experiences of the analysis teams and the industrial evaluation. Optimisation-based worst-case search was identified by industry as having the most potential to improve the current flight clearance process in the short term. The main reason for this was the method's flexibility in dealing with both linear and nonlinear analysis criteria, where most other methods focused mainly on linear analysis criteria. The optimisation-based method increases the analysis coverage, for a reasonable effort, compared to the current industrial method. As a side activity, requirements for visualisation tools have been identified and reported, based on inputs from the industrial members. Some of the analysis techniques also used promising concepts to visualise their results. Due to the limited time and effort available for this project, the ideas on visualisation could not be implemented and integrated to demonstrate their benefits. For illustration, figure 1 shows a clearance result of the polynomial-based clearance method, with an "adaptive-grid" algorithm and a clear visualisation of analysis results. The plot contains the result of the (linear) unstable eigenvalue criterion for HIRMplus at high angle of attack (27º). This method is able to clear whole regions of the flight envelope. For regions which cannot be indicated completely cleared (green) or completely non-cleared (red), the area is divided into four smaller regions, for which the process is repeated. Yellow indicates areas where the criterion is close to violation. In the white area on the bottom-right around flight condition FC7, the aircraft model cannot be trimmed at this angle of attack due to violation of the maximum load factor. Figure 1: Eigenvalue Stability criterion for the polynomial based method
The main
results of the project have been captured in a book, "Advanced
Techniques for Clearance of Flight Control Laws" [1], that was published
in September 2002 by Conclusions and Recommendations The research activity has been beneficial for all participants, with a healthy interchange of ideas and information. By undertaking this research, universities and research establishments have been able to familiarise themselves with an industrial task, and experience the practical problems that can arise. In return, European aircraft manufacturers have gained insight into the potential of the state-of-the-art analysis techniques that are available, and are still undergoing development within the scientific community. In this respect, the main objective of the project has been met. As expected, the analysis of (classical) nonlinear criteria with new analysis methods proved to be difficult. A future activity might investigate how nonlinear models, new analysis methods and (new) nonlinear criteria can be developed in such a way that they take each other's characteristics into account. Finally, it is noted that various combinations of methods could be beneficial and is worthy of further investigation. Reference [1] C.
Fielding, A. Varga, S. Bennani, M. Selier (Editors) "Analysis Techniques
for Clearance of Flight Control Laws", Springer-Verlag LNCIS Series No.
283,
|
