Flutter Flight Tests
In this chapter, we consider the general problem of lifting-surface flutter. Several types of flutter analysis were presented, including the p method, classical flutter analysis, к method, and p-к method. The application of classical flutter analysis to discrete one – and two-degree-of-freedom wind-tunnel models was presented. Students were exposed to Theodorsen’s unsteady thin-airfoil theory along with the more modern finite-state thin-airfoil theory of Peters et al. Application of the assumed- modes method to construct a flutter analysis of a flexible wing was demonstrated as well. The important parameters of the flutter problem were discussed, along with current design practice, flight testing, and certification. With a good understanding of the material presented herein, students should be sufficiently equipped to apply these fundamentals to the design of flight vehicles.
Moreover, with appropriate graduate-level studies well beyond the scope of material presented herein, students will be able to conduct research in the exciting field of aeroelasticity. Current research topics are quite diverse. With the increased sophistication of controls technology, it has become more common to attack flutter problems by active control of flaps or other flight-control surfaces. These so-called flutter-suppression systems provide alternatives to costly design changes. One type of system for which flutter-suppression systems are an excellent choice is a military aircraft that must carry weapons as stores. These aircraft must be free of flutter within their flight envelope for different configurations, sometimes many different configurations. At times, avoidance of flutter by design changes is simply beyond the capability of designers for such complex systems. There is also a body of research to determine in flight when a flutter boundary is being approached. This could be of great value for situations in which damage had altered the properties of the aircraft structure—perhaps unknown to the pilot—thus shifting the flutter (or divergence) boundary and making the aircraft unsafe to operate within its original flight envelope. Other current problems of interest to aeroelasticians include improved analysis methodology for prediction of flutter, gust response, and limit-cycle oscillations; design of control systems to improve gust response and limit-cycle oscillations; and incorporation of aeroelastic analyses at an earlier stage of aircraft design.
In this step, we take off and fly at the lowest speed at low altitude and in level flight. We then measure the response due to air turbulence and conduct FFT and PSD analyses. We determine whether these results match our analytical predictions. If they do not, we must again correct and/or adjust the model until we obtain agreement. When they eventually agree, we then may proceed to activate the actuators for an impulse command. Next, we determine whether there is sufficient damping. If there is, we conduct a sweep of frequencies and conduct FFT and PSD analyses of frequencies and damping. If these results match the analyses, then we may (cautiously!) activate actuators at the calculated flutter frequency and conduct FFT and PSD analyses of the response. When the damping measurements match the theoretical predictions, we activate actuators for a step command. Next, we determine whether the response matches the analyses. When it does, we collect and store data in the form of U – g – g’ diagrams. Only if and when there is reasonable agreement with analyses may we proceed to perform a 2g maneuver at the same speed and altitude. At this level of gs, we activate the actuators for an impulse command to see whether there is sufficient damping. If there is, we move on to 3g, then 4g, and so on, all at the same airspeed and altitude, until we have tested the complete set of specified load factors within the flight envelope.
If our model is still in agreement with the results obtained at any stage, only then is it safe to go to a higher speed (e. g., 25 knots faster) at the same altitude. At this point, we repeat all of the steps and collect data in the form of U – g – g’ diagrams. We systematically and cautiously increase the speed up to its maximum, checking at every increment to ensure that our analysis is valid. Similarly, we systematically increase altitude to its maximum and repeat the regimen. We stop (i. e., reduce speed to the previous safe speed) immediately whenever any one of the following happens:
1. There is even the slightest indication that the damping coefficient g decreases below 2% (5% in a civil aircraft).
2. Oscillations in at least one of the measurements diverge and tend to grow beyond preapproved limits.
3. The dominant frequency deviates from the predicted mode frequency.
Thus, it is observed that the analysis of flutter is strongly based on theoretical studies. The theory is the work tool for analysis and decisions about critical configurations and flight conditions. Ground-vibration experiments are used to tune the analysis to yield accurate structural-dynamics modes and wind-tunnel experiments to tune the unsteady-aerodynamics code.
Flutter flight tests are extremely dangerous. Real-time measurements and various actuation techniques are used to estimate the damping of the airplane at various flight altitudes, speeds, and load factors, moving from one point to another with caution. Analysis, ground experiments, and flight tests always go together to provide full clearance for flight without any flutter problems.